Conference Proceedings, Volume 1

Transcription

Proceedings of the Conference
2015
SCience in TEchnology
www.scinte.gr
SCinTE
November 5-7, 2015
Athens, Greece
Volume 1
Topic A: Applied Mechanics, Civil and Energy Engineering
Topic B: Earth and Environmental Sciences
Topic C: Arts & Humanities
Οrganized by
Supported by
Technological
Educational Institute
(TEI) of Athens
Research Funding
Program
ARCHIMEDES III
2015
SCinTE
Proceedings of the Conference “SCience in TEchnology”
Copyright 2015, by TEI of Athens
Copyright belongs to the Technological Educational Institute (TEI) of Athens. All rights reserved. No part of this
publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic,
mechanical, photocopying, recording, or otherwise, without the prior written permission of the Editors.
DESIGNED BY EFI PANAGIOTIDI
2015
SCinTE
Proceedings of the Conference “SCience in TEchnology”
Athens, Greece / November 5-7, 2015
Οrganized by
Supported by
Technological
(TEI) of Athens
Research Funding
Program
ARCHIMEDES III
Editor:
Dimos Triantis, Technological Educational Institute of Athens, Greece
Associate Editors:
Ioannis Kandarakis, Technological Educational Institute of Athens, Greece
Stavros Kourkoulis, National Technical University of Athens, Greece
Filippos Vallianatos, Technological Educational Institute of Crete, Greece
Bessie Argiropoulou, Technological Educational Institute of Athens, Greece
Ilias Stavrakas, Technological Educational Institute of Athens, Greece
International Scientific Committee
Applied Mechanics, Civil and Energy Engineering
Zacharias Agioutantis, University of Kentucky, USA
Konstantinos Belibassakis, National Technical University of Athens, Greece
Harris Douglas, Heriot-Watt University, UK
Wolf Gerrit Fruh, Heriot-Watt University, UK
Ioannis Kaldelis, TEI of Piraeus, Greece
Dimitris Karalekas, University of Piraeus Greece
Kostas Karamanos, Technological Educational Institute of Athens, Greece
Tassos Karayiannis, Brunel University, UK
Antonios Kyriazopoulos, Technological Educational Institute of Athens, Greece
Eric Landis, University of Maine, USA
Georgios Livanos, Technological Educational Institute of Athens, Greece
Nikos Pnevmatikos, Technological Educational Institute of Athens, Greece
Costas Politis, Technological Educational Institute of Athens, Greece
Tian Tang, University of Alberta, Canada
Stavros Tsantzalis, University of Patras, Greece
Earth & Environmental Sciences
Marios Anagnostou, National Observatory of Athens, Greece
Torok Akos, Budapest University of Technology and Economics, Hungary
Seraphim Alvanides, Northumbria University, UK
Michele Campagna, Universita di Cagliari, Italy
Angelo De Santis, Istituto Nazionale di Geofisica e Vulcanologia, Italy
Georgia Fotopoulos, Queen’s University, Canada
Antonios Giannopoulos, The University of Edinburgh, UK
Vassilios Karakostas, Aristotle University of Thessaloniki, Greece
Vincenzo Lapenna, Institute of Methodologies for Environmental Analysis, Italy
Wen Nie, Chengdu University of Technology, China
Vassilios Pagounis, Technological Educational Institute of Athens, Greece
Dimos Pantazis, Technological Educational Institute of Athens, Greece
Eleftheria Papadimitriou, Aristotle University of Thessaloniki, Greece
Stefan van der Spek, T.U.Delft, Nederland
Luciano Telesca, Institute of Methodologies for Environmental Analysis, Italy
Maria Tsakiri, National Technical University of Athens, Greece
Andreas Tsatsaris, Technological Educational Institute of Athens, Greece
Grigoris Tsokas, Aristotle University of Thessaloniki, Greece
Andreas Tzanis, National and Kapodistrian University of Athens, Greece
Proceedings of the International Conference “SCience in TEchnology - SCinTE - 2015”
Arts & Humanities
Athina Alexopoulou, Technological Educational Institute of Athens, Greece
Nikitas Chiotinis, Technological Educational Institute of Athens, Greece
Christian Degrigny, Haute Ecole Arc Conservation-restauration, Switzerland
Andreas Karydas, Demokritos Institute of Nuclear Physics, Greece
Vasilios Lampropoulos, Technological Educational Institute of Athens, Greece
Ian D. MacLeod, Fremantle Museums & Maritime Heritage, Western Australian Maritime
Museum, Australia
Paraskevi Pouli, Institute of Electronic Structure and Lasers, Greece
Wafaa El Saddik, Children’s Alliance for Tradition and Social Engagement, Germany
Brian Singer, University of Northumbria at Newcastle, UK
8
ISBN: 978-960-98739-8-7
Volume 1 / Topic A: Applied Mechanics,
Volume 1 Civil and Energy Engineering
Table of Contents
Topic A: Applied Mechanics, Civil and Energy Engineering
17
Pressure Stimulated Currents and Acoustic Emission combined recordings
for the detection of compressive strength of Dionysos marble I. Stavrakas, D. Triantis, A. Kyriazopoulos, K. Ninos 19
Acoustic Emissions Experimental Technique used to verify Kaiser effect during
compressional tests of Dionysos marble.
D. Triantis and I. Stavrakas
23
Experimental evaluation of a retrofitted PV/T collector
Evangelos Sakellariou, Petros Axaopoulos 27
An Internet-Accessible Experiment Using a Remote Photovoltaic Laboratory
Emmanouil D. Fylladitakis, Petros J. Axaopoulos, Michael Theodoridis
And Konstantinos Moutsopoulos
31
Determination of Fracture Loads for Mode I Fracture Toughness Calculations
for CCNBD Marble Specimens by Acoustic Emission Signals K. Kaklis, S. Mavrigiannakis, V. Saltas, A. Daskalaki, F. Vallianatos, Z. Agioutantis
35
Evaluation of LNG Bunkering Concept for Greek Sea Territory G.A. Livanos, S. Dimitrellou, E. Strantzali, G. Theotokatos
40
Energy management in the domestic sector - Peculiarities and chances in the Greek case John Gelegenis, George Mavrotas and George Giannakidis
44
Ship Shape Optimization using a BEM-Isogeometric solver for wave resistance
K.V. Kostas, A.I. Ginnis, C.G. Politis and P.D. Kaklis
48
Energy harvesting from sea waves and currents using oscillating hydrofoils
K. Belibassakis, T. Gerostathis, E. Filippas, J. Touboul and V. Rey
54
Application of a 3D coupled-mode model to the hydroelastic analysis of very large
floating bodies over inhomogeneous seabeds
Th.P. Gerostathis, K.A. Belibassakis and G.A. Athanassoulis
58
Degradation effects in field-aged c-Si solar panels of a grid-tie photovoltaic plant Konstantinos Gkarakis, Konstantinos Loukidis, Petros Axaopoulos
63
Acoustic and electrical emissions from sandstone under uniaxial compression
V. Saltas, I. Fitilis, J. P. Makris and F. Vallianatos 67
ISBN: 978-960-98739-8-7
9
The displacement field in a circular disc made of either orthotropic
or transversely isotropic material under parabolic radial pressure Christos F. Markides
71
Rejoining fragmented elements of stone monuments: Critical aspects
from the experience of the Athens Acropolis restoration project
Stavros K. Kourkoulis
76
An experimental study of the pull-out mechanism of threaded metallic bars
embedded in marble blocks
I. Dakanali, I. Stavrakas, D. Triantis, E.D.Pasiou, S.K. Kourkoulis
81
Electrical and Acoustic Emissions during Three Point Bending Tests
of Pre-Notched Marble Specimens
Ermioni D. Pasiou, Ilias Stavrakas, George Hloupis, Stavros K. Kourkoulis
and Dimos Triantis
86
Simple Probabilistic Slope Stability Analysis
Georgios Belokas
91
Embodied energy and embodied CO2 of building constructive materials
in a typical Hellenic dwelling
Georgios Syngros and Dimitrios Koubogiannis
95
Comparison of different Air Conditioning Systems implemented in a typical
office building in Greece
Ilia Lampro and Dimitrios Koubogiannis
99
Optimal Flow Control of Laminar Vortex Shedding around a Cylinder Ioannis Bonis, Haralambos Sarimveis and Dimitrios Koubogiannis
104
Using b-value analysis to estimate pre-failure characteristics of cement mortar
beams subjected to Three-Point-Bending
I. Dakanali, A. Kyriazopoulos, I. Stavrakas, D. Triantis
109
Green Islands in Europe and Prospects for Greek Islands. The Tilos Project
J.K. Kaldellis, G. Salagiannis, N.C. Ilia, P. Stinis, K. Dimakis
113
Electromobility in EU: Current status and future prospects for the Greek market
J.K. Kaldellis, St. Liaros, G. Spyropoulos
118
Application of recurrence quantification analysis in wind time series from wind farms
A.Κ. Charakopoulos, T.E. Karakasidis and I. Sarris
124
Non-linear Time series Methods Applications on Transport Data A.D.Fragkou, T.E. Karakasidis and E. Nathanail
10
128
ISBN: 978-960-98739-8-7
Effect of carbon nanotubes addition on the mechanical performance of co-polyimide
hollow fiber membranes
Filippos D. Gegitsidis, Nikolaos D. Alexopoulos, Evangelos P. Favvas and
132
Corrosion-induced fracture toughness degradation of artificially aged 2024
aluminum alloy specimens Paraskevi Liberaki, Nikolaos D. Alexopoulos, Constantinos I. Stergiou
and Stavros K. Kourkoulis
135
The impact of corrosion-induced micro-cracking on the structural integrity
of ultra-thin sheets of aeronautical aluminum alloy 2024
Christina Margarita Charalampidou, Nikolaos D.Alexopoulos, Panagiotis Skarvelis,
Charis Dalakouras and Stavros K. Kourkoulis
138
Fracture toughness of electron beam welded specimens of aluminum alloy 6156
Theano Examilioti, Nikolaos D. Alexopoulos, George Stefanou, Vasilis Stergiou,
142
Polyvinyl alcohol - carbon nanotubes fiber as an embedded sensor in GFRP under
bending loads
Panagiota Lymperta, Stavros K. Kourkoulis, Philippe Poulin and Nikolaos D. Alexopoulos 144
The effect of graphene nanoplatelets addition on the mechanical performance
of epoxy resin
Zafeiroula Paragkamian, Nikolaos D. Alexopoulos, Philippe Poulin and S. K. Kourkoulis 148
Effect of prior solid solution heat treatment on the mechanical behavior
of artificially aged aluminum alloy 2024 specimens
Antonios I.Pasoudis, Paraskevi Liberaki, Alexis Kermanidis and Nikolaos D. Alexopoulos
150
Nano-additions of multiwall carbon nanotubes in white cement for restoration
of monuments of Cultural Heritage
Zoi Metaxa, Spyridoula Boutsioukou, Stefanos Nitodas and Stavros K. Kourkoulis
153
Hydroelastic analysis of VLFS elastically connected to the seabed in shallow
wave conditions E. Karperaki, K. A. Belibassakis, T. K. Papathanasiou and S. I. Markolefas
156
Topic B: Earth & Environmental Sciences
161
Coastal Transport Integrated System (Co.Tr.I.S) Dimos N. Pantazis, Panagiotis Stratakis, Vassilis Moussas, Elias Lazarou,
Dimitris Stathakis, Eleni Gkadolou, Charalampos Karathanasis, Costas Politis,
Anna Christina Daverona, Vassiliki Kyriakopoulou, Eleni Babalona, John Chrysoulakis,
Maria Paralika, John Tyrinopoulos 163
ISBN: 978-960-98739-8-7
11
3D Heritage Recording using Terrestrial Laser Scanning Techniques Vassilios D. Andritsanos, Michail Gianniou, Vassilios Pagounis, Maria Tsakiri 167
A Shale Brittleness Index Based on Shale Peak Strength and Strain Yanhu Tan, Fenglin Xu, Lin Chen, Qiao Chen, Linhong Zhu, Lisha Wang and Wen Nie
171
Urban run-off management in a Greek coastal city: Citizens’ awareness,
attitudes and proposed solutions
Kalderis D., Stavroulakis G. and Diamadopoulos E.
174
Forest fires: a phenomenon with serious environmental and other impacts
Georgia Tsakni
178
Spatio–temporal cluster analysis of seismicity using a modified density–based
clustering algorithm
Dionysios Mountakis
181
Advances in Satellite Remote Sensing of major Natural, Environmental and
Industrial Hazards
V. Tramutoli and N. Pergola
185
A non-extensive statistical physics view in Earth Physics from laboratory
to planetary scale: A review.
Filippos Vallianatos 187
A Non Extensive Statistical Physics view on the spatiotemporal distribution
of Earth‘s seismicity
Kalliopi Chochlaki, Filippos Valianatos and George Michas
191
Paikon and Tzena terranes (Axios zone, Greece): Can they be correlated? Katrivanos E., Kilias A. and Mountrakis D.
195
Temporal Properties of Seismicity in the Corinth Rift: Are Earthquakes Random?
Georgios Michas, Filippos Vallianatos and Peter Sammonds
199
The Hellenic Seismological Network of Crete (HSNC): Monitoring results
and the new strong motion network
Georgios Chatzopoulos, Ilias Papadopoulos and Filippos Vallianatos
203
Studying Lithosphere-Atmosphere-Ionosphere Coupling at the south
front of the Hellenic Arc
John P. Makris, Bruno Zolesi, Massimo Chiappini, Filippos Vallianatos,
Hercules Rigakis, Maxim Smirnov, Andreas Tzanis, Despina Kalisperi,
Fragkiskos Pentaris,Vassileios Saltas, Ioannis Barbounakis, Ilias Papadopoulos,
George Hloupis, Paolo Spalla, Adriano Nardi, Luca Spogli, Roberto Carluccio,
Enrico Zuccheretti, Lucilla Alfonsi, Pantelis Soupios and Eleni Kokinou
207
12
ISBN: 978-960-98739-8-7
Spectral fractal analysis of MHz electromagnetic signals Dimitrios Nikolopoulos, Petraki Ermioni, Cantzos Demetrios, Grigorios Coulouras
and Constantinos Nomicos
212
ImproDeProF Project: Recent Advances and New Challenges in the development
of the DeProF tentative theory for steady-state two-phase flow in porous media Marios S. Valavanides
216
Microseismicity study in the area of Florina (Western Macedonia, Greece)
Mesimeri M., Karakostas V., Papadimitriou E. and Tsapanos T.
220
Evidence of accelerating deformation prior to large earthquakes: a case study
on the southern Hellenic Arc of Greece
G. Minadakis,G. Chatzopoulos and F. Vallianatos
224
Aftershock statistics of large earthquakes in Greece G. Minadakis, F. Vallianatos and I. Baskoutas
229
Automatic Image Orientation for Accurate Texture Mapping of 3D City Models L. Grammatikopoulos, I. Kalisperakis and E. Petsa
234
Exploration of spatial variability of cycling in a British Urban Environment Evidence from North East England
Seraphim Alvanides, Godwin Yeboah
238
Stress and Moment Release Models for Seismic Hazard Assessment in Corinth Gulf
Mangira Ourania, Tsaklidis Georgios and Papadimitriou Eleftheria
242
A MATLAB routine for earthquake epicentre determination using macroseismic data Vasiliki Kouskouna and Georgios Sakkas
246
A homogeneous Mw equivalent earthquake catalog for Greece 1990-2015
Karakostas Vasileios, Papadimitriou Eleftheria and Mesimeri Maria
250
“Earthshield”: An implementation for automatic macro-seismic calculations
P. Loumpardias, K. Chimos , T. Karvounidis, A. Ganas, .C. Douligeris
254
Small-world property and distinct evolution of complex networks at periods
of main shocks from historical earthquake records in Greece
D. Chorozoglou, D. Kugiumtzis and E. Papadimitriou
258
Design and Development of Embedded Systems for Precision Agriculture
Applications: The Case of Monitoring Dacus Oleae Population and Bait Sprays George N. Fouskitakis, Lefteris D. Doitsidis, Hercules Rigakis, Kyriaki Varikou,
Ioannis Sarantopoulos
263
Seismicity Memory Properties in Corinth Gulf (Greece)
Gkarlaouni Charikleia, Lasocki Stanislaw and Papadimitriou Eleftheria
267
ISBN: 978-960-98739-8-7
13
The joint modeling of earthquake time series using bivariate Poisson
Hidden Markov Models
Katerina Orfanogiannaki and Dimitris Karlis
272
An overview of landslide detection and monitoring using geodetic satellite observations AC Peidou, G Fotopoulos
276
Problems of Fracture Mechanics and Statistical Physics
Peter Sammonds and Clare Matthews
280
Topic C: Arts & Humanities
283
Paradigm Shift in Innovative Design: Use of Form and Form of Use
Dr Athanasios Kouzelis, Professor
284
Technology as a Creative Factor
Prof. Dr Tomas Kačerauskas
289
Science and Technology in Archaeological Museums: The Re-Exhibition of the Archaeological Museum of Tegea Gregory Grigorakakis, Panagiotis Ioannidis and Andreas Tsatsaris
293
Oil paintings on paper support: Determination of condition criteria
via non destructive testing and microanalysis
A. Alexopoulou, B. Singer, P. Banou, S. Zervos, A. Kaminari, A. Moutsatsou, A. Terlixi,
E. Tziamourani, A. Karabotsos, M. Doulgeridis
299
Swelling inhibition of clay-bearing building materials N.A. Stefanis, M. Stefanidou, A. Kalagri, I. Karatasios, V. Kilikoglou, I. Papagianni
and P. Theoulakis
303
Testing composites for the in situ protection of wooden shipwrecks
Alexios-Nikolaos Stefanis and Anastasia Pournou
306
Polypeptide models of keratin and collagen as a tool for the destruction of works of art
E. Fotou, E. Panou-Pomonis, E. Ioakeimoglou, E. Malea, G. Panagiaris and
M. Sakarellos-Daitsiotis
309
Assessing Environmental Effects on Organic Materials in Cultural Heritage:
Chemical Deterioration of Artificially Aged Bone Stamatis Boyatzis, Eleni Ioakimoglou, Yorgos Facorellis, ClioVossou,
Maria Sakarellou, Eugenia Panou-Pomoni, Fotou Evmorfia, Efrosini Karantoni,
Angelos G. Kalampounias, George A. Voyiatzis, Soghomon Boghosian,
Jane Richter, Athanasios Karampotsos, Eleni Tziamourani and George Panagiaris
14
312
ISBN: 978-960-98739-8-7
Development of a standard laser technology for cleaning evidence of proteinaceous
Cultural and Natural Heritage
Athanassia Papanikolaou, Kristallia Melessanaki, Alexandra Alexandropoulou,
Konstantinos Giannaras, Athanasios Karambotsos, Effrosyni Karantoni,
Ekaterini Malea, Stavroula Rapti, Nikolaos-Alexios Stefanis, Yorgos Facorellis,
Stamatis Boyatzis, Paraskevi Pouli and George Panagiaris
317
Removing iron corrosion products from museum artefacts: Investigating
the effectiveness of innovative green chelators Stavroula Rapti, Shayne Rivers and Anastasia Pournou
322
On the role and significance of cultural heritage for a positive national identity
Some educational remarks on museum education
Wafaa El Saddik
325
Preliminary investigation of L-cysteine and mature tobacco as corrosion inhibitors
for marine composite artefacts containing copper or iron alloys.
V. Argyropoulos, S.C. Boyatzis, M. Giannoulaki, A. Malea, A. Pournou, S. Rapti,
A. Zacharopoulou, and Elodie Guilminot
328
Experimental design of natural and accelerated bone and wood ageing
(Project INVENVORG)
Y. Facorellis, A. Pournou, J. Richter, E. Karantoni, C. Vossou, N. Androutsopoulos
and G. Panagiaris
332
Bioart: Borders and definitions. Visitors’ survey at Bioart exhibition
S.O. Papaioannou, E. Karantoni, Z. Tsourti, E. Kokkinidou, St. Drakos, N. Panourgia ,
V. Papakiriakou, G. Panagiaris, Th. Avaritsiotis
337
ISBN: 978-960-98739-8-7
15
Topic A
Applied Mechanics, Civil and Energy Engineering
18
ISBN: 978-960-98739-8-7
Volume 1 / Topic A: Applied Mechanics, Civil and Energy Engineering
Pressure Stimulated Currents and Acoustic Emission combined
recordings for the detection of compressive strength of Dionysos
marble.
I. Stavrakas, D. Triantis, A. Kyriazopoulos, K. Ninos
Laboratory of Electronic Devices and Materials, Technological Educational Institute of
Athens, 12210, Athens, Greece, *[email protected]
Keywords: Pressure Stimulated Currents, Acoustic Emissions, Dionysos marble
compressive strength, b value analysis.
Abstract
The estimation of the compressive strength of geomaterials is of significant importance for
various applications ranging from civil engineering to natural hazards. In this work prismatic
Dionysos marble specimens where subjected to constantly increasing compressive mechanical
stress up to their fracture. Concurrently to the mechanical stress application weak electrical currents and Acoustic Emission events were detected using high specifications laboratory equipment. The signatures of the electrical signals and the Acoustic Emissions were analysed and
studied under the concept of detecting characteristic behaviours that may be used to indicate
an upcoming event of fracture. Thus, the temporal variation of the electrical currents and the
statistical quantity known as Improved b-value were combined and the results clearly show that
such measurements may be used to indicate the existence of dynamic processes that guide the
fracture of the specimen.
Introduction
In order to understand the phenomena that are related to electrical earthquake precursors a
large number of laboratory experiments have been conducted, which have demonstrated the
generation of transient electric potential prior to rupture in both dry and wet rock specimens
(Nitsan, 1997; Ogawa and Miura, 1985; Enomoto and Hashimoto, 1990; O’Keefe και Thiel,
1995; Takeuchi and Nagahama, 2001; Triantis et al., 2006). An interesting experimental technique that is used to record electrical signals emissions mechanical stress is applied with various modes on rock specimens, like marble and amphibolite, and artificial materials like cement
based materials are has been developed during the latest years. The electrical signal is detected
by recording a weak electrical current using a sensitive electrometer. The detection sensor comprises a pair of gold plated electrodes that is attached at the sides of the specimen under study.
The above experimental technique is known as Pressure Stimulated Currents technique (Triantis et al., 2006), while the emitted weak electrical currents are known by the term Pressure
Stimulated Currents (PSC).
Additionally, the scientific literature records several studies and recordings of Acoustic Emissions that are caused by externally applied mechanical stress leading to microcracking on rocks
(Lockner, 1993). Towards this direction the Acoustic Emissions (AE) experimental technique
is used and continuously developing in order to constitute a valuable tool for the monitoring
and understanding of the mechanisms of dynamic processes, and for detection of an upcoming
mechanical failure (Rao and Lakschmi, 2005). Considering the AE analysis a parameter that
is most commonly studied in order to investigate a critical condition regarding its mechanical
status is the b-value (Rao and Lakschmi, 2005; Colombo et al., 2003; Carpinteri et al., 2008)
which systematically changes during the different stages of the failure process. The AE based
ISBN: 978-960-98739-8-7
19
b-value analysis and the study of the b-value variation, have attracted researchers working in
the engineering field (Sagar et al., 2013; Sagar et al., 2012). The b-value analysis have been
recently improved by the use of statistical distribution parameters (mean value and standard
deviation) and the new method in known as Improved b-value (Ib-value) and has been proposed
by Rao and Lakschmi, 2005. The formalism that is used to calculate the improved b-value is
defined as follows:
Ib =
log N ( µ − α1σ ) − log N ( µ + α 2σ )
(α1 + α 2 ) ⋅ σ
(1)
where m is the mean amplitude of the acoustic events, the standard deviation, and α1 and α2 are
constant that are regularly take the value of 1 (Rao and Lakschmi, 2005).
This work presents a combined correlation between the PSC signal and the Ib-value after analyzing the AE amplitudes during compressional stress tests at constant stress rate on dry marble
specimens up to its fracture.
Sample and experimental technique
The specimen (Dionysos marble) was shaped prismatic of dimensions 35mm×35mm×75mm.
The physical and chemical properties of this kind of marble have already been presented in
bibliography (Kourkoulis et al., 1999; Stavrakas et al., 2003). A description of the experimental
arrangement for PSC measurements as well as of all measurement procedures has been thoroughly presented in the previous works (Stavrakas et al., 2003; Anastasiadis et al., 2007a). The
strain was measured using Kyowa strain gauges attached on the Microlink-770, 120Ω resistor
bridge. During the tests the released AE events were monitored using a Physical Acoustic Corporation (PAC) Mistras Systems. The AE sensor was installed at the center of the side area of
the specimen. The AE threshold for detecting acoustic events was set at 40dB. Further details
on the monitoring of the ΑΕ are described in an earlier paper (Stergiopoulos et al., 2013).
Experimental results – discussion
The specimen was subjected to axial compressional stress at a constant rate of 320kPa/s up to
its fracture that occurred at a stress level of 82MPa. The stress-strain curve shows a wide linear
region ranging from 5MPa up to 60MPa and a calculated Young’s modulus of 75GPa that is
supported by the existing relevant literature (Kourkoulis et al., 2007). Figure 1, presents the temporal development of the emitted PSC (curve a) in the marble specimen. In the same diagram,
curve b describes the temporal evolution of the strain of the specimen. A weak electric signal
below 5pA appears up to 190s approximately. It must be stressed that until this moment the
strain rate is constant and the material is within the linear range with respect to its mechanical
behaviour because the specimen is being loaded at a linearly increasing uniaxial compressional
stress. Observing the PSC recording the following remarks can be drawn. An intense increase
of the PSC is observed after the time of 190s when the increasing rate of the strain is already
evident. In this case, the applied stress is greater than the yield stress (60MPa approximatelly),
the Young’s modulus gradually decreases and the material enters the damage range. This fact
supports the consideration that the PSC recording may provide and evidence of critical damage
initiation. The PSC reaches a maximum value of 15pA just before the failure of the specimen
and this constitutes a serious indication that the material has suffered extensive damages.
Regarding the AE during the complete experimental procedure 3400 AE events were detected.
Figure 2 shows the temporal recording of the PSC is shown with the corresponding cumulative AE energy from the first AE event up to the fracture of the specimen. The abrupt increases
of the cumulative AE energy for times longer than 235s are clearly related to the intense PSC
emissions peaks.
20
ISBN: 978-960-98739-8-7
Figure 1: (a) The temporal variation of the PSC enission and (b) the corresponding temporal variation of the
mechanical strain.
Figure 2: The temporal variation of the PSC enission and
the corresponding temporal variation of the cumulative
energy of the recorded AE.
In order to show up the variability of the Ib-value, during the complete experimental procedure
the total number of AE was divided into groups of successive AE events. Each group included
n=200 AE events from which the 100 first belonged to the previous group (i.e. each group was
overlapping the last 100 AE events of the previous). This way, 33 sequential Ib-values were
calculated according to equation 1. The calculated Ib-values were related to the stress that corresponds to the 200th AE event of each group.
Figure 3: The temporal variation of the PSC enission and the corresponding temporal variation of the Improved
b-value of the recorded AE.
The results are shown as Ib-values versus time in Figure 3 with respect to the corresponding
temporal behaviour of the recorded PSC. It is clear that during the early stages (t<190s) the
Ιb-values are continuously increasing which is an indicative fact of prevalence of microcracks
(Rouchier et al., 2012). During this stages the PSC signal is also maintained at low levels
showing a slow gradual increase. Sequentially, for longer times (t>190s) when stress exceeds
60MPa approximately, the Ιb-values show an intense decay that is characteristically related to
the intense PSC increase. When the PSC tends to show a peak (t>240s) the Ib-value reaches a
value of 1.5 and sequentially decreases until the fracture of the specimen where it becomes 0.9
approximately.
Concluding Remarks
Both the recordings of the emitted electrical current (PSC emission) and the corresponding AE
through the Ib-value analysis, validate that the microfractue phenomena start to interact guiding
ISBN: 978-960-98739-8-7
21
the forming the necessary conditions of fracture, when the applied stress becomes higher than
70% of the strength of the specimen.
Acknowledgements
This research has been co‐financed by the European Union (European Social Fund – ESF) and
Greek national funds through the Operational Program “Education and Lifelong Learning” of
the National Strategic Reference Framework (NSRF) ‐ Research Funding Program: THALES.
Investing in knowledge society through the European Social Fund.
References
1. Anastasiadis, C., Triantis, D., Stavrakas, I., Vallianatos, F., 2004. Pressure stimulated currents (PSC) in marble
samples after the application of various stress modes before fracture. Ann. Geophys. 47 (1), 21–28.
2. Anastasiadis, C., Triantis, D., and Hogarth, C. A.: Comments on the phenomena underlying pressure stimulated
currents (PSC) in dielectric rock materials, J. Mater. Sci., 42, 2538–2542, 2007a.
3. Carpinteri, A. Lacidogna, G. Niccolini, G. and Puzzi, S. Critical defect size distributions in concrete structures
detected by the Acoustic Emission technique. Meccanica, 43(3), 349-363 (2008).
4. Colombo, S., Main, I., G. and Forde, M. C. Assessing damage of Reinforced Concrete Beam using ‘‘b-value’’
Analysis of Acoustic Emission signals. J. Mat. Civil Eng. (ASCE), 15, 280-286 (2003).
5. Enomoto, J. and Hashimoto H.: Emission of charged particles from indentation fracture of rocks, Nature, 346,
641–643, 1990.
6. Kourkoulis, S. K., Exadaktylos, G. E., and Vardoulakis, I.: Unotched Dionysos-Pentelicon marble beams
in three point bending: The effect of nonlinearity, anisotropy and microstructure, Int. J. Fracture, 98, 369–
392,1999.
7. Kourkoulis, S.K., Moupagitsoglou and N.L. Ninis. 2007. The influence of the flutes on the mechanical
behaviour of the ancient Greek columns. Experimental Analysis of Nano and Engineering Materials and
Structures, Proc. Of the 13th International Conference on Experimental Mechanics, Alexandroupolis, Greece,
July 1-6. pp. 955
8. Lockner, D. 1993. The role of acoustic emission in the study of rock fracture, International Journal of Rock
Mechanics and Mining Sciences & Geomechanics Abstracts Volume 30, Issue 7, December 1993, Pages 883899
9. Nitsan, U., 1997. Electromagnetic emission accompanying fracture of quartz-bearing rocks. Geophys. Res.
Lett. 4, 333–337.
10. Ogawa, T.K., Miura, T., 1985. Electromagnetic radiation from rocks. J. Geophys. Res. 90, 6245–6249.
11. O’Keefe, S.G., Thiel, D.V., 1995. A mechanism for the production of electromagnetic radiation during fracture
of brittle materials. Phys. Earth Planet. Int. 89, 127–135.
12. Rao, M.V.M.S. and Lakschmi, P.K.J. Analysis of b-value and improved b-value of acoustic emissions
accompanying rock fracture. Current Science, 89, 1577-1582 (2005).
13. Rouchier, S. Foray, G. Godin, N. Woloszyn, M. Roux J.J. 2012. Damage monitoring in fibre reinforced mortar
by combined digital image correlation and acoustic emission, Construction and Building Materials 38, pp.
371-380.
14. Sagar, V.R., Prasad, R.B.K., Rao MVMS, 2013. Microcracking and fracture process in cement mortar and
concrete: a comparative study using acoustic emission technique. Experimental Mechanics, 53:1161–1175 .
15. Sagar, V.R., Prasad, R.B.K., Kumar, S.S. (2012). An experimental study on cracking evolution in concrete
and cement mortar by the b-value analysis of acoustic emission technique, Cement and Concrete Research 42
1094–1104.
16. Stavrakas, I., Anastasiadis, C., Triantis, D., and Vallianatos, F., 2003. Piezo stimulated currents in marble
samples: precursory and concurrent-with-failure signals, Nat. Hazards Earth Syst. Sci., 3, 243–247.
17. Stavrakas, I., Triantis, D., Agioutantis, Z., Maurigiannakis, S., Saltas, V., Vallianatos, F., Clarke, M., 2004.
Pressure stimulated currents in rocks and their correlation with mechanical properties. Nat. Hazards Earth
Syst. Sci. 4, 563–567.
18. Stergiopoulos, C., Stavrakas, I., Hloupis, G., Triantis, D., Vallianatos, F., 2013. Electrical and acoustic
emissions in cement mortar beams subjected to mechanical loading up to fracture, Engineering Failure
Analysis, 35, 2013 pp. 454–461.
19. Triantis, D., Stavrakas, I., Anastasiadis, C., Kyriazopoulos, A., Vallianatos, F., (2006). An analysis of pressure
stimulated currents (PSC), in marble samples under mechanical stress, Phys. Chem. Earth 31, 234–239.
22
ISBN: 978-960-98739-8-7
Acoustic Emissions Experimental Technique used to verify Kaiser
effect during compressional tests of Dionysos marble.
D. Triantis* and I. Stavrakas
Laboratory of Electronic Devices and Materials, Technological Educational Institute of
Athens, 12210, Athens, Greece, *[email protected]
Keywords: Acoustic Emissions, Dionysos marble, Kaiser effect, Felicity Ratio.
Abstract
The damage development due to externally applied mechanical stress is a hot topic of interest
involving several applications of everyday life like civil engineering, monument restoration,
construction evaluation and health monitoring. Repetitive loadings on brittle materials cause
internal damages that gradually extend leading to inevitable failures. Such processes are studied under the concept of the material mechanical memory effect that is widely known as Kaiser effect. The Kaiser effect, states that a structure will only suffer further internal damaging
when exposed to applied stresses higher than previously encountered. Certain conditions lead
to a violation of the Kaiser effect, known as the Felicity effect, quantitatively measured using
the Felicity Ratio. This work presented the scientific results of two experiments demonstrating repetitive mechanical load loops on marble specimens and concurrent Acoustic Emission
measurements. The collected Acoustic Emission data are studied in combination with the mechanical data like mechanical stress and strain under the frame of the Kaiser effect. It is clearly
seen that the Felicity ratio strongly depends on the stress range the material is subjected to, with
regard to the elastic or plastic deformation region.
Introduction
The scientific literature reports several studies and recordings of Acoustic Emissions (AE) that
are attributed to mechanical stresses adequate to cause microcracking phenomena from rocks
(Tonolini et al., 1987). Towards this direction the Acoustic Emission (AE) detection experimental technique has been evaluated and developed in order to constitute a valuable tool for the
monitoring and interpretation of the underlying physical mechanisms of mechanical dynamic
processes and the detection of an upcoming event of mechanical failure (Rao and Lakschmi,
2005).
The Kaiser effect is an AE phenomenon briefly defined as the absence of detectable acoustic
emissions until the previously applied stress level is exceeded. This effect is based on the experimental discovery by Kaiser (1950), that metal materials had the capability to remember the
previous maximum stress level. This capability was also discovered in rock materials (Kurita
and Fujii, 1979) and since then the Kaiser effect is used to detect and assess the amount of damage that has been developed in rocks (Holcomb and Costin, 1986; Holcomb et al., 1990).
The breakdown of the Kaiser effect can be represented quantitatively by the felicity ratio (FR)
that is defined as the ratio of the AE-onset stress to the maximum stress of the previous cycle. It
may be taken as a measure of the quality of the rock. (Li 1993) A high felicity ratio means that
the rock is of good quality.
The aim of this work in to conduct a laboratory experimental investigation and verification of
the Kaiser effect on Dionysos marble specimens. The specimens are subjected to compressive
loading loops where the first loading is in the region where the material leaves the elastic region
and enters the plastic deformation and the stress strain curve deviates linearity.
ISBN: 978-960-98739-8-7
23
Test Facilities and Arrangement
The marble specimen were subjected to three loading unloading loops. The maximum stress
level during the 1st and the 2nd loadings was approximately 60MPa. This stress level, according
to preliminary tests on similar specimens, corresponds to the region that the materials gradually
enters the non-linear region regarding its stress-strain behaviour and initial plasticity effects
take place. During the unloading processes the stress level was maintained at a value of 15MPa
approximately. A third loading was attempted during which the specimen failed at a stress level
of 82.7MPa. The compressive stress was applied following load control at a rate of 320kPa/s.
The specimen (i.e. Dionysos marble) was prismatic with dimensions 35mm×35mm×75mm.
The physical and chemical properties of this kind of marble have already been presented in
bibliography (Kourkoulis et al., 1999; Stavrakas et al., 2003). The strain was measured using
Kyowa strain gauges attached on the Microlink-770, 120Ω resistor bridge.
During the tests the released AE events were monitored using a Physical Acoustic Corporation
(PAC) Mistras Systems. The AEs transducer was attached at the center of the side area of the
specimen and parallel to the stress axis. The AE threshold for detecting acoustic events was set
at 40dB. Further details on the monitoring of the ΑΕ are described in an earlier paper.
Figure 1: Temporal variation of the mechanical
stress (black line), and the corresponding behaviour of the mechanical strain (gray line) during the
complete experimental procedure.
Figure 2: The stress-strain behaviour during all
the three loading / unloading loops.
Figure 1 shows the loading path temporal variation until the failure of the specimen and the corresponding strain behaviour. Figure 2 shows the stress-strain curve during the complete experimental procedure. It must be noted that during the 1st loading the linear region was estimated
to last up to a stress level of 58MPa approximately, a value that corresponds to the 70% of the
ultimate compressive strength of the material and leads to a Young’s modulus of 72GPa.
Verification of the Kaiser Effect
The AE event recordings during the complete experimental procedure are illustrated in Figure
3. Specifically the AE events per second are shown with respect to time. In the same Figure the
temporal variation of the applied mechanical stress is also presented. It may be clearly seen that
AE continuously occurred during the entire process during the first loading loop leading to a total
number of recorded AE events of 978. However, during the second loading loop a significantly
smaller number of AE events were recorded (289 AE events). It is noticeable that practically an
absence of AE events until the load reached the 68% of the compressive strength provided that
only 24 low amplitude events were recorded. It must be stressed that 90% of the detected AE
24
ISBN: 978-960-98739-8-7
events were recorded in the region between 68% and 73% of the compressive strength. This
region corresponds to the upper limit of the linear and the lower limit of the non-linear behaviour of the material regarding its stress-strain curve. During the 3rd loading and before the stress
reached the level of the previously applied stress only 45 AE events were recorded while the 36
of them were recorded when the applied stress was in the region between 71% and 73% of the
compressive strength. When the stress exceeded these values 1237 AE events were recorded
until the failure of the specimen. The AE records of the entire experimental procedure (i.e. the
three loading / unloading loops) are plotted in a cumulative form as shown in Figure 4.
Figure 3: The mechanical stress temporal development during the complete experimental procedure
(gray line) and the corresponding per second AE
event rate. (black bars).
Figure 4: The behaviour of the cumulative number
of the AE events with respect to the normalized value
of the mechanical stress with respect to the compressive strength of the specimen.
Figure 5: The mechanical stress temporal development during the complete experimental procedure
(gray line) and the corresponding per second AE
event rate (black bars).
Figure 6: The behaviour of the cumulative number
of the AE events with respect to the normalized value
of the mechanical stress with respect to the compressive strength of the specimen.
During the 2nd and the 3rd loading unloading loops the Felicity Ratio (FR) can be calculated
to be equal to 0.66 and 0.8 correspondingly. The deviation from the optimum value of FR=1,
may be justified due to the fact that the maximum stress values of the 1st and 2nd loading loops
was selected to lay in the limits of the linear and non-linear region regarding the stress-strain
behaviour of the marble and additionally to the quality of the used marble. It must be noted that
during each loading loop the FR clearly depends on the maximum stress of the previous loop
and decreases as the value of the maximum stress increases (Li and Nordlund, 1993). In order
to verify this observation the same set of specimens was used and the loading protocol that is
ISBN: 978-960-98739-8-7
25
shown in Figure 5 was adopted. The maximum stress of the 1st loading was selected to be at
33MPa and the unloading was conducted as low as 10MPa in order a new loading to follow
until the fracture of the specimen. The same Figure shows the recorded AE events per sec, while
Figure 6 shows the cumulative AE recordings during the complete experimental procedure. For
this experiment the FR reaches 0.9 approximately that is clearly higher from the corresponding
value of the previous experiment.
Concluding Remarks
According to the above, it is experimentally verified that for marble specimens the Felicity
Ratio (FR) lays near 1 1 from the beginning of loading until a certain level of stress where the
material is still in the linear region regarding its stress-strain behaviour. When the stress level
becomes higher and a non-linear behaviour is established regarding the stress-strain curve the
FR starts deviating significantly from the value of 1.
Acknowledgements
This research has been co‐financed by the European Union (European Social Fund – ESF) and
Greek national funds through the Operational Program "Education and Lifelong Learning" of
the National Strategic Reference Framework (NSRF) ‐ Research Funding Program: THALES.
References
1. Holcomb, D.J., Costin, L.S., 1986. Detection damage surfaces in brittle materials using acoustic emissions.
Trans. ASME., 53, 536-544.
2. Holcomb, D.J., Stone, C.M., Costin, L.S., 1990. Combining acoustic emission locations and a microcrack
damage model to study development of damage in brittle materials. In: Hustrulid, W.A., Johnson, G.S. (eds)
Proc., 31st U.S. Symposium, Rock Mechanics Contributions and Challenges, Balkema, Rotterdam, 645-651.
3. Kourkoulis, S. K., Exadaktylos, G. E., Vardoulakis, I., 1999. Unotched Dionysos-Pentelicon marble beams in
three point bending: The effect of nonlinearity, anisotropy and microstructure, Int. J. Fracture, 98, 369–392.
4. Kourkoulis, S. K., Exadaktylos, G. E., Vardoulakis, I., 1999. notched Dionysos-Pentelicon marble beams
in three point bending: The effect of nonlinearity, anisotropy and microstructure, Int. J. Fracture, 98, 369–
392,1999.
5. Kurita, K., Fuji, N., 1979. Stress memory of crystalline rocks in acoustic emission. Geophys. Res. Lett. 6 (1),
9-12.
6. Li, C., Nordlund, E., 1993. Experimental verification of the Kaiser Effect in Rocks. Rock Mech. Engng. 26(4),
333-351.
7. Rao, M.V.M.S., Lakschmi, P.K.J., 2005. Analysis of b-value and improved b-value of acoustic emissions
accompanying rock fracture. Current Science, 89, 1577-1582.
8. Stergiopoulos, C., Stavrakas, I., Hloupis, G., Triantis, D., Vallianatos, F., 2013. Electrical and acoustic
emissions in cement mortar beams subjected to mechanical loading up to fracture, Engineering Failure
Analysis, 35, 454–461.
9. Stavrakas, I., Anastasiadis, C., Triantis, D., Vallianatos, F., 2003. Piezo stimulated currents in marble samples:
precursory and concurrent-with-failure signals, Natural Hazards and Earth System Science 3 (3/4), 243-247.
10. Tonolini, F., Sala, A., Villa, G., 1987. General review of developments in acoustic emission methods,
International Journal of Pressure Vessels and Piping, vol. 28, no. 1, pp. 179–201.
26
ISBN: 978-960-98739-8-7
Experimental evaluation of a retrofitted PV/T collector
Evangelos Sakellariou, Petros Axaopoulos*
Technological Educational Institute (T.E.I.) of Athens, Aegaleo, Greece. *[email protected]
Keywords: retrofitted PV/T, PV/T, Hybrid PV panel, copper-made PV/T absorber.
Abstract
A retrofitted PV/T collector assembled and installed together with a conventional PV panel, at
north hemisphere, facing south at a tilt of 22o. In order to evaluate the performance of the system, outdoor experiments were conducted and a PC-based automatic data acquisition unit has
been used. The results show that the PV/T collector for all the experimentation days managed to
increase its average electrical energy yield by 22.87%. Additionally, the PV/T collector exploits
the panel’s removed heat and generates thermal energy with daily average efficiency of 37%.
Introduction
The solar industry is mainly divided in two dominant technologies, the Photovoltaics (PV) and
the Solar Thermal (T) systems. Photovoltaics use solar energy in order to generate electricity,
while solar thermal systems use the sun with the purpose to generate thermal energy. Nevertheless, PV efficiency is deteriorated by high cell temperature while high collector’s temperature
is welcome for better efficiency in thermal systems. The above controversial operation of those
technologies has driven into a combined collector the PV/T (K.E. Park et al, 2010) and (E.
Skoplaki et al, 2009).
When solar irradiation incidents in a PV panel, only a small fragment of this energy about
5-15% becomes electrical energy, while another portion of 7-10% is reflected and the vast
amount becomes heat. The remaining heat at the PV panel increases the cells’ temperature and
thus decreases the electric efficiency. However, if a heat-absorber is installed on the rear side of
the PV panel, the cells’ temperature will be decreased by removing the cells’ heat and thus the
cells’ efficiency will be increased.
According to the existing literature and especially (H.A. Zondag, 2008) for an unglazed water
based PV/T collector the maximum thermal efficiency can reach 52% and the electrical efficiency at 10%.
Nomenclature
PV/T Photovoltaic and thermal collector
To
Collector’s output temperature (K)
Qu
PV/T’s thermal energy (Wh)
PV
Photovoltaic module
Ta
Ambient temperature (K)
Tcoll
Collector’s mean temperature (K)
Ti
Collector’s entry temperature (K)
TPV
Panel’s mean temperature (K)
The aim of this experiment was to evaluate the performance of a retrofitted PV/T collector
located in the Attica region, Greece. For this experiment, a retrofitted copper-made PV/T collector was built up and installed along with a conventional PV panel in an iron south faced
base. The desired outcomes of the set experiment were firstly to evaluate the PV/T’s thermal
efficiency and capacity during the day and secondly to compare the PV/T’s electrical efficiency
versus the PV’s electrical efficiency as well. Moreover, the results will assist for further investigation on the potential of a retrofitted PV/T system to supply the thermal energy demand in a
typical Greek house.
ISBN: 978-960-98739-8-7
27
Methodology & Experimental set-up
Two identical, commercially available PV panels were utilized and one of those was retrofitted
to a PV/T collector, while the specifications of the panels are shown in table 1. These two panels
were placed in a common south-faced base with tilt angle of 22o on the roof of TEI of Athens/
Renewable Energy Sources laboratory (38° 0’7.11”N/ 23°40’31.01”E).
The collector’s external dimensions were 0.52x1.20m. The absorber type was serpentine-flatwater-based and it was made by copper sheet and tubes as shown in fig.1. The absorber’s configuration was based on (P. G. Charalambous et al, 2011). The serpentine copper collector was
placed and fixed at PV’s rear side with wood reaps. Furthermore, among the collector and the
wood reaps, a layer of 1cm polyethylene was placed in order to mitigate the collector’s rear side
thermal losses. The schematic representation of the experiment is shown in fig. 2.
The collector’s heat removal medium was water with constant flowrate of 0.01kg/s (P. G. Charalambous et al, 2011). A small tank of 20lt was placed below the iron base and was supplied
continuously with fresh tap water in order to maintain a constant temperature. The collector was
supplied by a small pump with water in constant flowrate and the water at collector’s outlet was
rejected to the environment.
Pmax
76.6W
Vmp
17.3V
Imp
4.4A
Voc
21.4V
Isc
5.0A
Table 1. Electrical characteristics of the
PV panel
Fig.1. Collector’s dimensions
Fig.2. Experiment’s schematic
The PV/T and the PV panel were connected each one in a battery charger equipped with MPPT
system in order to earn the maximum power from each panel. The two chargers were responsible to apply battery charging management and to supply, if it was necessary, the load. The load
was supplied with energy via DC to AC power inverter. As load a florescent lamp of 16W was
chosen, while the battery’s capacity was 33Ah at 12V.
During the experiment were measured the input and output collector temperatures, the PV/T’s
and PV’s rear surface temperature as well as the ambient temperature. All the temperature
measurement probes were PT100. In addition, the solar irradiance and the wind speed at the
tilt direction were measured by a pyranometer and a cup-anemometer respectively. The water’s
flowrate was measured once with a volume measurement instrument and it was fixed at 0.01lt/s.
Regarding the electrical measurements, the charger’s input voltage and current were measured
for the PV/T and for the PV independently.
The measurements were sampled and stored in personal computer by an analog multichannel
card. The card scanned all the eleven channels five times during a five-minute period and stored
the average value of each channel. The experimentation time started at 10:30 and ended at
28
ISBN: 978-960-98739-8-7
16:00 every day. Earlier in the morning and later in the evening the results were disturbed due
to neighbor objects to the experimentation PV panels which caused severe shading effect on the
array. The experimentation period was from 26th of August 2015 to 1st of September 2015. The
experimentation days were chosen regarding the weather stability and low irradiance’s fluctuation in order to avoid transition effects at the results.
Results and discussion
At fig. 3 the daily PV/T’s and PV’s thermal and electric energy of seven chosen days is illustrated, while table 2 shows the experimental results. As fig. 3 shows, at all experimentation days
the PV/T has better electrical energy performance than the PV. Furthermore, PV/T’s thermal
energy is not proportional to its electrical energy and that can be easily observed on 31/08/15
and 01/08/15, where for equal electrical energy generation the system generates significantly
different amount of thermal energy.
Fig 3. Daily PV/T’s and PV’s thermal and electrical energy
According to the results, PV/T’s thermal efficiency is extended from 28% early in the morning to 48% at noon. The daily thermal energy generation for seven days was from 1.45kWh
to 1.65kWh. As table 2 illustrates, the average temperature exiting the collector varies from
36.08oC to 38.56oC and the highest To which was measured during the experiment is registered
on the 1st of September 2015 at 41.70oC.
Qu
[Wh]
Eel
[Wh]
Eel
[Wh]
Incident
solar
energy
[Wh]
1468.57
249.98
188.12
4085.98
PV/T
Date
260815
PV
Mean daily value
wind
speed
[Wh]
3.23
Ta
[oC]
Tcoll
[oC]
TPV
[oC]
To
[oC]
Ti
[oC]
31.64
36.89
44.27
36.08
31.16
270815
1467.77
321.09
263.45
4111.51
2.75
31.44
37.00
45.05
36.16
31.22
280815
1585.13
334.78
265.40
4126.50
2.65
33.42
37.83
47.75
37.07
31.96
290815
1473.03
426.72
372.20
4110.80
3.13
33.11
37.63
45.78
36.63
31.98
300815
1451.51
408.31
362.13
4059.91
3.80
34.36
38.24
45.35
37.49
32.76
310815
1497.10
400.62
338.78
4078.61
3.18
34.44
38.23
46.88
37.09
32.50
010915
1644.86
410.49
307.31
4042.15
0.96
35.55
39.62
54.03
38.56
33.53
Table 2. Experimental results
As table 2 shows, the daily solar energy slightly varies from day to day, whilst the Qu is affected
by the wind speed and the ambient temperature. Therefore, at table 2 it can be seen that during
two characteristic days 30/08/15 and 01/09/15 the Ta there is a slight difference of 1.2 oC and
ISBN: 978-960-98739-8-7
29
the wind speed fell from 3.80m/s to 0.96m/s, while the thermal energy yield for a windy day is
193.35Wh less than the day with moderate wind. The collector gets the highest thermal energy
yield for the lower wind speed and the highest Ta. In contrast with the highest yield, the lower
is registered on the day with the highest wind speed and the lower Ta.
Table 3 illustrates the PV/T’s and PV’s average efficiencies and the increased efficiency of
PV/T’s compared to PV’s. By investigating together the tables 2 and 3, the highest increased
electrical efficiency is 2.25% and corresponds to the larger difference between the Tcoll to TPV, as
well as to the day with the highest thermal energy yield. Opposite to the highest increased electrical efficiency, the smallest one appeared on the day with the lowest thermal energy yield.
Date
260815
PV/T mean PV/T mean PV
mean Increased PV/T col- PV/T collector’s effect on
Thermal ef- E l e c t r i c a l E l e c t r i c a l lector’s Electrical ef- Electrical energy generaficiency
efficiency
efficiency
ficiency
tion
35.68 %
6.16 %
4.67 %
1.49%
32.88 %
270815
35.42 %
7.84 %
6.45 %
1.39%
21.88 %
280815
38.53 %
8.27 %
6.62 %
1.65%
26.14 %
290815
35.85 %
10.41 %
9.12 %
1.29%
14.65 %
300815
35.89 %
10.16 %
9.05 %
1.11%
12.75 %
310815
36.39 %
10.84 %
9.61 %
1.23%
18.25 %
010915
41.21 %
11.15 %
8.90 %
2.25%
33.58 %
Average
37.00 %
9.25%
7.77%
1.49%
22.87%
Table 3. PV/T’s and PV’s average efficiencies.
Conclusions
During those seven experimentation days the results show that the PV/T collector outperforms
of PV panel by increasing the average electrical efficiency from 7.77% to 9.25% and the average daily electrical energy yield by 22.87%. As a bonus to the better electrical efficiency, the
PV/T generates additionally thermal energy with an average efficiency of 37%. The collector’s
output temperature can be utilized as preheating source, since it strictly reaches 40oC.
The PV/T collector operation decreases the panel’s temperature and the figures show that the
minimum difference between the Tcoll and TPV is 7.11oC and the larger 14.41oC. The PV/T’s efficiency is deteriorated by the large wind speed and low Ta due to increased heat losses. Among
the two parameters that influence the PV/T’s efficiency, the wind speed has the larger impact.
Regarding the future improvement, it is desirable to emphasize on better PV/T’s thermal insulation and assembling in order to increase the thermal efficiency even in windy and low ambient
temperature conditions.
References
1. K.E. Park. G.H. Kang. H.I. Kim. G.J. Yu. J.T. Kim, 2010, Analysis of thermal and electrical performance of
semi-transparent photovoltaic (PV) module. Energy, 35, (6) 2681–2687.
2. E. Skoplaki. J.A. Palyvos, 2009, On the temperature dependence of photovoltaic module electrical performance:
a review of efficiency/power correlations. Solar Energy, 83, (2009) 614–624.
3. H.A. Zonda,. 2008, Flat-plate PV-Thermal collectors and systems: A review, Renewable and Sustainable
Energy Reviews,12, 891–959
4. P.G. Charalambous. S.A. Kalogirou. G.G. Maidment. K. Yiakoumetti, 2011, Optimization of the photovoltaic
thermal (PV/T) collector absorber. Solar Energy, 85, 871–880
30
ISBN: 978-960-98739-8-7
An Internet-Accessible Experiment Using a Remote Photovoltaic
Laboratory
Emmanouil D. FYLLADITAKIS1, Petros J. AXAOPOULOS2*, Michael
THEODORIDIS1 and Konstantinos MOUTSOPOULOS2
1
2
Brunel University London, Uxbridge, Middlesex, United Kingdom
Technological Educational Institute (T.E.I.) of Athens, Aegaleo, Greece, *[email protected]
Keywords: Remote laboratory, Photovoltaic engineering education, Photovoltaic
panel, Distance education, e-learning.
Abstract
Remote access to a renewable energy lab opens new ways to the education of photovoltaics by
offering the students a feeling of direct experience with actual photovoltaic equipment and is a
process exceptionally useful for part-time learners, distance learners, as well as students with
disabilities. This paper demonstrates the functionality of a remote photovoltaics experimentation system for engineering education purposes. A short description of the system itself and
some of its capabilities are listed and a very important exercise for the thorough education of
energy engineering students is being presented.
Introduction
The sun is an important source of renewable energy, with photovoltaic (PV) panels usable anywhere on the planet, converting solar energy directly to electricity (Meral and Dinçer, 2011).
PV systems adaptation is constantly growing, especially in countries where favourable economic and bureaucratic policies are applied (International Energy Agency, Paris, 2009). That
makes PV panels a major part of any energy education program.
Even though theoretical study is an important aspect of education, engineering education should
be combined with experimentation to aid the student’s comprehension and allow for the easier
adaptation of theory into practice (Leach and Paulsen, 1999). However, this is not always possible, as access to equipment is limited by a great number of factors, such as the location of the
students, the climatic conditions, as well as time and financial constraints. Today, technology
and the Internet can prove to be a very valuable tool for current and future engineers, with the
development of remotely accessible educational laboratories that can provide many advantages
to engineering students and academics alike. Remote laboratory experimentation offers high
quality learning experiences in science and engineering education, while adding flexibility,
especially in terms of time and special needs (Colwell et al., 2002).
The renewable energy laboratory of the Technological Educational Institute (T.E.I.) of Athens,
part of a consortium that was arranged in light of the Erasmus+ Programme, is currently developing a remote photovoltaics experimentation system. This system is going to offer unlimited
remote access to an array of photovoltaics modules which are installed on the roof of the building. The aim of this paper is a basic demonstration of the system through the execution of a
fundamental exercise.
ISBN: 978-960-98739-8-7
31
System Description
The laboratory instruments and equipment used to realize the aforementioned remote lab configuration are:
• Four 20 Wp PV panels.
• A proprietary mount allowing axis adjustments and shading simulation.
• An adjustable electronic load.
• An outdoor web camera.
• Various automation, instrumentation and metering apparatus.
Figure 1 displays the externally mounted PV array, mount and motor.
Figure 1 - Experimental PV Array & Motor
The software is under development and it will be running through a server that, alongside the
access to the remote experimentation system, will be offering a wide selection of open educational resources on PVs. As the system is still under development, a BETA version of the
software is used for the means of this experimental exercise and only one solar panel (Siemens
Solar SM55) will be used.
The most basic function of the experimentation system performs a test on the PV(s) that, once
completed, the software extracts the current-voltage (I-V) and the power-current (P-I) curves of
the PV panel, as well as the following information:
• The panel voltage
• The load’s ohmic resistance
• The output current
• The total solar irradiance on the panel’s plane
• The panel’s temperature
• The test tilt angle and the daily optimal tilt angle
32
ISBN: 978-960-98739-8-7
Application
One of the many laboratory exercises that can be performed with the developed system, is described in this chapter. Even though an engineer would always examine the economic/energetic
performance of a PV system, the technical specifications provided by PV module manufacturers are always presented at standard test conditions (STC) and are considered insufficient to
perform proper modelling (Carrero et al., 2010). It has been proven that the STC curves do
not represent the performance of the PV panel under real-world conditions (Zhou et al., 2007,
Aranda et al., 2009), with scientists reporting an overestimation of the generation by up to 40%
when performing studies by using the STC figures (Durisch et al., 2000). Therefore, solar energy engineers need to be capable of extracting the real V-I characteristics of any PV module
in order to be capable to accurately assess a system’s performance and troubleshoot potential
problems.
The objective of this application is for the students to become capable of extracting the real I-V
curves of a PV panel, as well as the translation of the actual I-V curve to other environmental
conditions. The use of a software capable of developing the theoretical I-V curve using the
manufacturer’s specifications is required for this application. As the real I-V curve is noticeably
lower than the theoretical curve generated by the software using the STC values, it becomes obvious that other factors can reduce the performance of a PV panel. Some of these factors are:
• Cell aging
• Environmental temperature and wind
• Dust and debris
• Cell mismatching
• Cable losses
The next step is to demonstrate how an experimentally extracted I-V curve under actual environmental conditions can be converted to STC conditions. Academic literature proposes a number
of different methods by which the I-V characteristic may be translated (Yuki et al., 2006, Sadok
and Mehdaoui, 2008). However, the most common method for performance verification is the
translation to STC. The measured current and voltage figures can be translated to the module’s
STC values by equations 1 and 2 respectively.
ISTC = Ir×(HSTC / Hr) + a×(Tr - TSTC)
(1)
VSTC = Vr - β×(TSTC-Tr) - Rs×(Ir - ISTC) + Vt×Ln(HSTC / Hr)
(2)
Where ISTC, VSTC, TSTC and HSTC are the current, voltage, temperature and irradiance at STC respectively, Ir, Vr, Tr and Hr are the measured current, voltage, temperature and irradiance respectively, α is the temperature coefficient of the current, β is the temperature coefficient of the voltage and Rs is the series resistance. Vt is the thermal voltage of the module, which is described
by equation 3:
Vt= (A×k×T) / q
(3)
Where A is the area of the module, k is the Boltzmann constant, q is the magnitude of the electrical charge of the electron and T is the module’s temperature.
Given these equations, the students are being taught how to convert the real I-V curve extracted
from the remote experimentation system to STC conditions, compare it to the STC provided
by the manufacturer of the module and then convert it to environmental conditions specified by
the tutor or depending on the local environmental conditions at any application site. After the
completion of this exercise students should be capable of translating the I-V curves of any PV
ISBN: 978-960-98739-8-7
33
panel, as long as a real I-V curve under known environmental conditions has been extracted, to
a second I-V curve displaying the performance of a PV panel under STC conditions.
Concluding Remarks
The developed system gives the ability to perform remotely, through the Internet, and in a very
short time educational experiments using actual solar PV panels, recording data under real
conditions while, at the same time, the student can have a live view of the panel through a web
camera. Thus, students from around the world, whose institutes/schools do not have laboratories with PV equipment, or their countries do not have a high annual number of sunshine hours,
can experiment in their free time and then analyse the results according to the taught theory.
As demonstrated in this paper, one of the many possible applications will allow them to gain
practical experience on the calculation of the real peak power output of a PV panel by extracting
and translating actual I–V curves, comparing them to those supplied by the manufacturer and/
or assessing a panel’s performance under real environmental conditions. This approach opens
new ways in technical and engineering education, giving students the feel of an almost direct
experience with the experimental devices.
Finally, yet importantly, it allows for handicapped students and researchers to access a realworld PV site, making active participation in laboratory coursework and research possible.
Acknowledgements
This project has been funded with support from the European Commission. This publication
reflects the views only of the authors, and the Commission cannot be held responsible for any
use which may be made of the information contained therein.
References
1. Aranda, E. D., Galan, J. A. G., De Cardona, M. S. & Marquez, J. M. A. 2009. Measuring the I-V curve of PV
generators. Industrial Electronics Magazine, IEEE, 3, 4-14.
2. Carrero, C., Rodríguez, J., Ramírez, D. & Platero, C. 2010. Simple estimation of PV modules loss resistances
for low error modelling. Renewable Energy, 35, 1103-1108.
3. Colwell, C., Scanlon, E. & Cooper, M. 2002. Using remote laboratories to extend access to science and
engineering. Computers & Education, 38, 65-76.
4. Durisch, W., Tille, D., Wörz, A. & Plapp, W. 2000. Characterisation of photovoltaic generators. Applied
Energy, 65, 273-284.
5. International Energy Agency Paris, 2009. Photovoltaic power systems programme.
6. Leach, J. & Paulsen, A. 1999. Practical work in science education: Recent research studies. Studies in Science
Education, 33, 168-168.
7. Meral, M. E. & Dinçer, F. 2011. A review of the factors affecting operation and efficiency of photovoltaic
based electricity generation systems. Renewable and Sustainable Energy Reviews, 15, 2176-2184.
8. Sadok, M. & Mehdaoui, A. 2008. Outdoor testing of photovoltaic arrays in the Saharan region. Renewable
Energy, 33, 2516-2524.
9. Yuki, T., Yoshihiro, H. & Kosuke, K. 2006. Translation Equations for Temperature and Irradiance of the I-V
Curves of Various PV Cells and Modules. Conference Record of the 2006 IEEE 4th World Conference on
Photovoltaic Energy Conversion.
10. Zhou, W., Yang, H. & Fang, Z. 2007. A novel model for photovoltaic array performance prediction. Applied
Energy, 84, 1187-1198.
34
ISBN: 978-960-98739-8-7
Determination of Fracture Loads for Mode I Fracture Toughness
Calculations for CCNBD Marble Specimens by Acoustic Emission
Signals
K. Kaklis1*, S. Mavrigiannakis1, V. Saltas2, A. Daskalaki1, F. Vallianatos2,
Z. Agioutantis3
School Mineral Resources Eng., Technical Univ. of Crete, Greece, *[email protected]
Dept Environmental and Natural Res. Eng., Tech. Educational Institute of Crete,
Greece
3
Department of Mining Engineering, University of Kentucky, Lexington, KY
1
2
Keywords: rock fracture toughness, mode I loading, CCNBD, acoustic emissions,
marble.
Abstract
Rock fracture mechanics has been widely applied to blasting, hydraulic fracturing, mechanical
fragmentation, rock slope analysis, geophysics, earthquake mechanics and many other science
and technology fields. Development of failure in brittle materials is associated with microcracks, which release energy in the form of elastic waves called acoustic emissions. This study
presents results from acoustic emission (AE) measurements obtained during cracked chevron
notched Brazilian disc (CCNBD) tests on Nestos marble and the calculation of the fracture
toughness under pure mode I loading.
Introduction
Fracture mechanics can be applied to many engineering ﬁelds including civil and mining engineering, where drilling, excavation, explosion and cutting of rocks are closely related to the
strength, stability and fracture of rock materials and structures. Fracture toughness represents
the ability of a material to resist the propagation of cracks; it is considered to be an inherent
property of the each material and should not be related to testing conditions.
The International Society for Rock Mechanics (ISRM, 1995) suggested the cracked chevron
notched Brazilian disc (CCNBD) for determining the mode I fracture toughness of rock. There
are some unique features characterizing the CCNBD specimen: (a) it is closely related to the
Brazilian test widely used for tensile strength test for rock and concrete (b) it can be easily used
for mode I, mode II and even mixed mode testing (c) it can sustain higher critical load than
other kinds of specimens with comparable size and (d) it also maintains the merit of convenience for specimen preparation from rock cores.
Large numbers of AE signals are generated when a rock specimen is loaded to failure. Since
AE signals are generated by propagation and expansion of microcracks, such signals inherently
include information related to the structural changes taking place within a rock sample.
This study focuses on the correlation between the AE signals and the diametrically applied load
during the CCNBD tests. The acoustic emission activity was monitored using piezoelectric
acoustic emission sensors, and the potential for accurate prediction of the fracture load based
on acoustic emission data was investigated.
ISBN: 978-960-98739-8-7
35
Experiments and Acquisition of Acoustic Emissions
Material
Specimens consisting of Nestos marble were prepared and tested. This marble is quarried by
surface mines in northern Greece and is mainly used as a building material. It is composed of
93.4% calcite, 6% dolomite and 0.6% quartz. Its density is 2.67 g/cm3 and its absorption coefficient by weight is 0.09%. It is of white color with a few thin parallel ash-green colored veins
containing locally silver areas due to the existence of dolomite (Kaklis et al., 2010).
Cracked Chevron Notched Brazilian Disc specimen
The CCNBD specimen (Fig. 1) has the same geometry and shape as the conventional Brazilian
disc used for measuring the indirect tensile strength of rock, except that the CCNBD specimen
has a chevron notch. ISRM (1995) recommends that the following dimensionless parameters
can be used to characterize the geometry of the chevron notch:
(1)
is the initial notch length, is the maximum
where the diameter D is twice the radius R,
notch length, is the intermediate crack length, b is the crack front width, B is the thickness
of the disc, P is the load applied to the disc and Ds is the diameter of the circular cutting blade.
Typical standard dimensions of a CCNBD specimen suggested by ISRM (1995) are given as:
The length of the machine cracks are comparable to the
disc diameter (the actual slit opening is about ¼ D); nevertheless the stress field at the crack tip
is strongly influenced by this geometry which generates higher stress concentrations.
Figure 1.Geometry and loading condition of CCNBD specimen.
The fracture toughness of CCNBD specimens (ISRM, 1995; Wang, 1998, 2010; Chang et al.,
2002) can be calculated by the following equation:
(2)
is the maximum load, and
is the critical dimenwhere is the fracture toughness,
sionless stress intensity value which is determined by the specimen geometry
and
36
ISBN: 978-960-98739-8-7
only, and is given by
only (ISRM, 1995).
where u and v are constants determined by
and
Specimen preparation and experimental setup
Initially, the circular discs of Nestos marble (D = 52 mm) were prepared by coring specimens
out of Nestos cubes with their long axis normal to the plane of transverse isotropy to avoid large
variations in the quality of the stone and ensure similar fracture loads, according to ISRM specifications. Two chevron notches (one from each side) were cut in the center of each disc using a
thin rotary diamond saw with a diameter of Ds = 38 mm and a thickness of t = 1 mm (Fig. 2a).
The indentation depth of the rotary saw on each side of the disc was about 11.7 mm and thus,
two V shape chevron notches were introduced in the circular disc. The diamond saw was cooled
by water during the chevron notch cutting process.
(a)
(b)
(c)
Figure2. (a) The schematic cutting configuration for preparing CCNBD specimens. (b) CCNBD specimen. (c)
Brazilian test apparatus.
The CCNBD specimens (Fig. 2b) were then placed in the Brazilian testing apparatus (Fig. 2c)
and loaded by a diametral compressive force P. For pure mode I loading, the crack direction
should be exactly along the loading diameter. Load was applied using a stiff 1600 kN MTS hydraulic testing machine and a 500 kN load cell, with a loading rate of 200 N/s under load control
mode for all experiments.
Monitoring of Acoustic Emissions
The AE activity is represented in time series of detected signals (hits), counts (signal excursions
over the predefined threshold), amplitude (signal peak in dB) and other acoustic parameters. AE
signals were acquired and analyzed using the AE WIN software package by Enviroacoustics
(2009). Acoustic emissions were detected through 6 miniature piezoelectric sensors (PICO
sensors, 200kHz-1MHz, MISTRAS Group, SA) mounted on the surface of each specimen and
recorded through an integrated multi-channel system of Physical Acoustics Corporation. The
location of the sensors on one side of the specimen is shown in Fig.2c. Usage of such sensors
was dictated due to the small size of the specimens. Their high sensitivity over a broad bandwidth ensures the accurate detection of the AE signals.
Experimental results
Six CCNBD tests were performed, the fracture load was identified using AE records and the
specimen fracture toughness was calculated using equation (2). The average fracture toughness
for these specimens was about 0.75 MPa m1/2.
ISBN: 978-960-98739-8-7
37
Figure 3a shows representative AE results from all six sensors for a typical specimen of Nestos
marble under CCNBD loading. It is observed that the AE counts were relatively low for loads
lower than 4.09 kN or lower than 94% of the failure load (4.33 kN). Fig.3b presents the percentage cumulative AE activity with respect to percent load on the Nestos marble specimen. It is
observed that less than 10% of the total AE activity is recorded for loads up to 95%.
It should be noted that these results are in accordance with previous experimental results where
a different Greek marble (Dionysos marble) shows a higher damage level prior to localization
(Dai and Labuz, 1997; Nomikos et al., 2010; Kaklis et al., 2012).
Conclusions
Experimental results from AE monitoring of Nestos marble under CCNBD test are presented.
For this material significant AE activity is observed at levels higher than 95% of the failure
load. The calculation of the mode I fracture toughness by the relationship (2) is relatively easy
and depends on geometric dimensionless constants and the fracture load. With the aid of the
acoustic emission signals is achieved the accurate determination of the failure load, as shown
in Fig 3a.
(a)
(b)
Figure3. (a) AE results under CCNBD loading for the Nestos marble (b) Cumulative AE counts for the Nestos
marble specimen with respect to percent CCNBD load.
Acknowledgements
This work was supported by the THALES Program of the Ministry of Education of Greece and
the European Union in the framework of the project “Integrated understanding of Seismicity,
using innovative methodologies of Fracture Mechanics along with Earthquake and Non-Extensive Statistical Physics - Application to the geodynamic system of the Hellenic Arc - SEISMO
FEAR HELLARC.”
References
1. Chang S-H, Lee C-I, Jeon S., 2002, Measurement of rock fracture toughness under modes I and II and mixedmode conditions by using disc-type specimens. Eng Geol 66:79–97.
2. Dai, S. Labuz, J., 1997, Damage and failure analysis of brittle materials by acoustic emission. Journal of
Materials in Civil Engineering, 9 (4), 200-205.
3. Enviroacoustics S.A., 2009, Noesis, User’s manual.
4. ISRM Testing Commission (co-ordinator: R.J. Fowell), 1995, Suggested method for determining mode I
fracture toughness using cracked chevron notched Brazilian disc (CCNBD) specimens. Int J Rock Mech Min
Sci Geomech Abstr 32:57–64.
5. Kaklis, K., Mavrigiannakis, S., Agioutantis Z. Bazdanis, G., 2010, An investigation of the mechanical
characteristics of Nestos Marble. In: Volume in Honor of the late Professor K. Kavouridis, Technical University
of Crete Publications, pp. 57-68 (in Greek).
38
ISBN: 978-960-98739-8-7
6. Kaklis K., Mavrigiannakis S. and Agioutantis Z., 2012, Comparison of acoustic signatures of rock specimens
under uniaxial compression. ICCES’12: International Conference on Computational & Experimental
Engineering and Sciences. April 30–May 4, 2012, Crete, Greece.
7. Nomikos, P.P., Katsikogianni, P., Sakkas, K.M., Sofianos, A.I., 2010. Acoustic Emission during Flexural
Loading of Two Greek Marbles. Rock Mechanics in Civil and Environmental Engineering, Zhao, Labiouse,
Dudt and Mathier (eds), Taylor & Francis Group, London, pp. 95-98.
8. Wang, Q.Z., 1998, Stress intensity factors of the ISRM suggested CCNBD specimen used for mode-I fracture
toughness determination. Int J Rock Mech Min Sci 35:977–982.
9. Wang, Q.Z., 2010, Formula for calculating the critical stress intensity factor in rock fracture toughness tests
using cracked chevron notched Brazilian disc (CCNBD) specimens. Int J Rock Mech Min Sci 47:1006–
1011.
ISBN: 978-960-98739-8-7
39
Evaluation of LNG Bunkering Concept for Greek Sea Territory
G.A. Livanos1*, S. Dimitrellou1, E. Strantzali1, G. Theotokatos1,2
1
2
Technological Educational Institute of Athens, Egaleo, Greece, *[email protected]
Currently at University of Strathclyde, Glasgow, UK
Keywords: LNG, bunkering methods, small-scale terminal, fuel consumption.
Abstract
This paper presents an LNG bunkering infrastructure concept in the Greek Sea territory for
the distribution of LNG to Greek islands electricity power plants and for ship bunkering. The
LNG consumption is estimated accordingly to the power plant installed capacities in six Greek
islands and the installed power of five passenger vessels operating at the ports of Piraeus and
Patras. The concept of small-scale LNG terminals is proposed that is a feasible and efficient
solution for countries with areas of little or no pipeline infrastructure.
Introduction
The demand for environmental protection has led International Maritime Organization to set
limits on NOX and SOX emissions from ship exhausts, and prohibit deliberate emissions of
ozone depleting substances (IMO, 2008). Special areas designated as Emission Control Areas
(ECA), where emission limits are more stringent, have been already defined. More areas within
the Mediterranean Sea are planned to become future Emission Control Areas. To prepare for
this possible regulatory development, the use of different fuels with less environmental impact
needs to be studied. Using Liquefied Natural Gas (LNG) as a fuel is a very efficient way to reduce emissions. All SOx emissions are eliminated, the NOx emissions are reduced up to 85%,
whereas the CO2 emissions decrease by 25%-30%.
The significant infrastructure in terms of pipeline projects and LNG terminals that have been
developed, are under development or have been proposed in the region of east Mediterranean,
makes LNG a viable solution for Greek ships and also for other LNG needs. In specific, four
gas transport and trading projects (Nabucco, Interconnector Turkey-Greece-Italy, Trans Adriatic Pipeline, South East Europe Pipeline) are ongoing (Natural Gas Europe, 2013, Gurbanov,
2014). Moreover, one existing Greek import terminal with current LNG storage capacity of
130,000 m3 and two proposed Greek import terminals with LNG storage capacities of 285,000
m3 can support the further distribution of LNG by the common LNG bunkering methods.
LNG Distribution/Infrastructure Concept
Greek Islands Power Plants. The majority of Greek islands, especially those in Aegean Sea,
are not connected to the mainland electricity grid. Their insular electricity system comprises oilfired power plants that cover their demands. The examined Greek islands are: Rhodos, Crete,
Samos, Kalymnos and Lesvos and Kos. Table 1 summarize the total annual consumption of the
HFO needed to cover the energy demands for the years 2013 and 2014 and the total cost of the
HFO needed (making the assumption that the current price of HFO is 600€/mt).
Fuel Consumption [mt]
Fuel Cost [€]
2013
825,868.46
495,521,075.37
2014
848,544.27
509,126,564.37
Table 1: HFO total consumption and cost
40
ISBN: 978-960-98739-8-7
This study investigates the conversion of the existing power plants in Rhodos, Samos, Kalymnos and Lesvos, and the installation of new power plants in Crete, Lesvos and Kos. Table
2 shows the results of the calculation of the LNG consumption for each island power plant
(Johnsson, 2015).
Plant
Rhodos
Crete
Lesvos
Kos
Total
Total Installed Capacity [MWe]
145
207
Samos Kalymnos
24.6
16.4
153
83
629
Average load 80% [MWe]
116
165.6
19.7
13.1
122.4
66.4
503
150.95
1,377.00
740.54
5,749.34
181
121
1,102
592
4,599
66,118
44,079
402,084
216,239
1,678,806
LNG Cons. at 100% load/day [m ] 1,411.36 1,843.04 226.43
3
LNG Cons. at 80% load/day [m ]
3
1,129
1,474
Annual LNG Cons. 80% load
412,118 538,169
[m3]
Table 2: LNG consumption for power plants
According to the energy demand of 2013 and 2014 the LNG needed for the same years was calculated. If the energy demand is higher than the energy produced from the LNG consumption
then it is used HFO as additional fuel. Making the assumption that the price of HFO is 600€/tn
and the price of LNG is 35€/MWh, then the total cost for these six islands had been, also, calculated and it is shown in Table 3. In the last column of Table 3, the fuel cost savings are given,
calculated by the difference of the total costs in Tables 1 and 3.
2013
2014
LNG [tn] HFO [tn] LNG cost [€]
537,300.83 96,270.49 250,740,386.63
536,934.47 119,443.79 250,569,417.33
HFO cost [€]
57,762,295.11
71,666,273.37
TOTAL [€]
308,502,681.73
322,235,690.69
Cost Difference
-187,018,393.64 €
-186,890,873.67 €
Table 3: The total consumption and fuel cost for the six islands
Port LNG nodes. Different types of vessels are sailing at the Greek territory as Ro-Ro, ferries,
bulk carriers, container and general cargo vessels. However, this study is focused on five LNG
fueled passenger vessels. Five of them operate between the main islands at Aegean Sea (sailing
from Piraeus port) and two of them connect Patras port to Italy. The estimation of LNG consumption is shown in Table 4. We assume that vessel main engines operate at 75% load.
Installed Power
Power @ load
Running hours /day
Energy
Fuel consumption
LNG volume
LNG consumption
Hours between bunkering
LNG cons. between bunkering
Bunkering frequency
LNG consumption/ year
LNG consumption/ day
MW
kW
h
kWh
MJ
m3
m3/h
h
m3
days
m3/year
m3/day
Olympic
Spirit
4 x 12,6
37,800
8
302,400
2,531,693
119
15
16
237
2
43,316
119
Piraeus Port
Blue Star
Paros
4 x 4,1
12,300
14
172,200
1,441,658
68
5
14
68
1
24,666
68
Knossos
Palace
4 x 16,8
50,400
10
504,000
4,219,488
198
20
20
396
2
72,193
198
Patras Port
Cruise
Super Fast
Europa
XI
4 x 12,6
4 x 12,6
37,800
37,800
22
22
831,600
831,600
6,962,155
6,962,155
326
326
15
15
44
44
653
653
2
2
119,118
119,118
326
326
Table 4: LNG consumption estimation
ISBN: 978-960-98739-8-7
41
Small Scale Terminal Design. The small scale LNG concept shares much of the technology with
traditional large scale LNG. Large scale is about intercontinental transport of millions of tons
LNG from a LNG production unit to an import terminal where the commodity product is fed
into a national pipeline grid system. Small scale LNG on the other hand is more of a regional
business moving hundreds of thousands of tons from the LNG source, using various modes of
transport ranging from ships to semitrailers and ISO containers, directly to end-users.
The most feasible solution for the Greek Sea case is the installation of small scale LNG terminals in Piraeus Port and Patras port, for the bunkering of the corresponding vessels, and in each
island for the power plant installations.
To determine the storage tank capacity of each small-scale terminal we assume a fill-up interval
of 10 days with a safety inventory of 5 days. The requirement for the heel is 10%. Heel is the
small amount of liquefied natural gas remaining on storage after discharge of the regular LNG
cargo, and is the minimum quantity of LNG necessary to be retained in holding tanks. In each
terminal type C storage tanks to cover the required LNG volume are used, as shown in Table 5.
The tanks shall be filled with LNG up to 85%.
Terminals
Rhodos
Crete
Samos
Kalymnos
Lesvos
Kos
Piraeus Port
Patras Port
Required LNG storage
capacity (m3)
18,630
24,328
2,989
1,993
18,176
9,775
6,353
10,758
LNG tank size (with
25% tank vacuum)
21,918
28,621
3,516
2,345
21,384
11,500
7,474
12,656
No of tanks
9
12
3
2
9
10
8
8
Single tank
capacity (m3)
2,450
2,400
1,200
1,200
2,400
1,150
0,950
1,600
Table 5: Small scale LNG terminals sizing
Fig. 1 illustrates the LNG distribution chain. The import terminal is the existing LNG terminal
of Revithousa Island. For the proposed study a large LNG tank of 100,000 m3 and an export
bunkering reload facility has to be constructed. A LNG carrier of 100,000 m3 capacity will
transfer LNG to Revithousa’s LNG tank once every 10 days.
Figure 1: Small-scale LNG concept and LNG feeder vessel route in Aegean Sea
42
ISBN: 978-960-98739-8-7
To distribute LNG between Revithousa import terminal and the small-scale terminals, two LNG
feeder vessels will be used. The first one, with LNG tank capacity of 82,244 m3, will sail from
the import terminal and upload LNG at the small-scale terminals of Piraeus port and islands.
The second feeder vessel, with LNG tank capacity of 10,758 m3, will sail from the import terminal and upload LNG at the small-scale terminal of Patras Port. Fig. 1 shows the LNG feeder
vessel route from the import terminal to the small-scale terminals (transportation interval of 10
days).
LNG bunkering (DMA, 2012) can then be subdivided into ship-to-ship bunkering (STS), at
quay or at sea, tank truck-to-ship bunkering (TTS) and/or LNG intermediary terminal-to-ship
via pipeline (TPS). The suitability of a bunkering method to a particular vessel depends on a
number of factors such as, port turn-around time, bunkered volume, voyage range, etc.
Conclusions
In this paper an LNG infrastructure concept in Greek territory is presented, focused on the LNG
demand a) for the Greek islands electricity power plants and b) for five LNG fueled passenger
vessels. The small-scale LNG concept is an effective solution for making natural gas available
to energy users not currently connected to pipeline networks. The concept increases the market
for natural gas by distributing LNG from either a LNG plant, LNG import terminal or directly
from a LNG carrier using a combination of both sea and land based transport directly to the
end-user.
Acknowledgements
This work is conducted in the framework of the project LNG COMSHIP (Greek General Secretariat of Research and Technology Code: 12CHN400), which is funded by the European Regional Development Fund (ERDF) and National Resources.
References
1. DMA, 2012. North European LNG Infrastructure Project - A feasibility study for an LNG filling station
infrastructure and test of recommendations, Copenhagen: The Danish Maritime Authority.
2. IMO, 2008, Amendments to the Annex of the Protocol of 1997 to Amend the International Convention for
the Prevention of Pollution from Ships, 1973, as Modified by the Protocol of 1978 Relating thereto(Revised
MARPOL Annex VI), Resolution MEPC.176(58).
3. Johnsson, T., 2015, LNG Action Plan for Europe & Africa 2015, Wärtsilä.
4. Gurbanov, I., 2014. Implications of the demise of South Stream for southern gas corridor, Strategic Outlook,
December 2014.
5. Natural Gas Europe, Greek gas sector eyes flexible LNG infrastructure, Date of access: 10/2013. http://www.
naturalgaseurope.com
ISBN: 978-960-98739-8-7
43
Energy management in the domestic sector - Peculiarities and
chances in the Greek case
John Gelegenis1*, George Mavrotas2 and George Giannakidis3
1
2
3
Energy Technology Dept., T.E.I. of Athens, *[email protected]
School of Chemical Engineering, National Technical University of Athens
Centre for Renewable Energy Sources and Saving (C.R.E.S.)
Keywords: Energy management, dynamic simulation, holistic optimization, Greece.
Abstract
A multi-level analysis (regarding design, technology and regulations) is elaborated to find out,
formulate and propose interventions with a remarkable potential contribution to the improvement of energy performance of Greek dwellings. A holistic approach is followed to this aim,
and the prominent role of energy management is duly revealed.
Introduction
Policy makers in the European Commission have agreed that there is an urgent need to adopt
energy efficiency measures in order to comply with the energy and climate targets set for 2020,
including the commitment to reduce by 20% energy consumption compared to the projected
consumption in that year. Buildings account for 40 % of total energy consumption in the EU,
therefore they are central in energy efficiency programs. The situation is similar in Greece,
where the domestic sector, not including buildings of the tertiary sector, represents 24% of
the total final energy consumption of the country. It should be emphasized that a large part of
the final consumption in the domestic sector is electricity (32%), with the majority of the rest
(45%) being oil products. The renewables share refers mainly to biomass (mostly wood used
in fire-places or ovens) and solar energy used for domestic hot water production (Diakoulaki et
al, 2012, Gelegenis et al, 2014).
In the context of improving the energy performance of buildings (e.g. in the frame of the Energy
Performance of Buildings Directive (EPBD) 31/2010/EC which is valid in the European Union), a stimulus to decrease energy consumption in buildings has been established. Regarding
existing dwellings, the aim is their energy upgrade using conventional technologies and furthermore even going beyond cost-optimal solutions to renovate them into Nearly Zero-Energy
Buildings. Indeed, a variety of energy efficiency measures (EEMs) is applicable in dwellings,
at various cost and effectiveness. Quite usually, the effectiveness of an EEM decreases with its
capacity (e.g. a first layer of insulation is much more effective than an additional one), while at
the same time its cost increases, hence the need for economic optimization. Numerous works in
optimization were published during last ten years. The combined application of several measures stimulates interactions between each other and consequently affects their effectiveness (in
most cases decreasing it). Due to these potential interactions, a holistic optimization seems to
be more suitable, instead of optimizing one EEM at a time. Actually, there are very few works
dealing with optimization through a holistic approach (Gelegenis et al, 2015a), which is the aim
of this work, too.
Methodology
The prevailing regulations were firstly reviewed; more emphasis was given to the recent regulation for the energy performance of buildings, the classification process established and the
experience gained up-to-date; in this framework, thousands of energy performance certificates
44
ISBN: 978-960-98739-8-7
were decoded and analyzed in details (Gelegenis et al, 2014). The best available technologies
were afterwards scrutinized, according to their potential impact and their cost-effectiveness,
without neglecting the uncertainty on their economic evaluation (Diakoulaki et al, 2012). Technologies like “ex post” insulation of the envelope (Axaopoulos et al, 2014), use of condensing
boilers and heat pumps (Bonaros et al, 2013), application of cogeneration (Gelegenis and Mavrotas, 2015a), installation of photovoltaics (Gelegenis et al, 2015b) etc. were analyzed in details
by the use of dynamic simulation. Analytical solutions were attempted, wherever convenient, as
it was the case with the selection of the optimum temperature increment ΔTd,opt of a cogeneration unit (Gelegenis and Mavrotas, 2015b):
(1)
where a and b are respectively the slope and constant term of a linear regression approximation
of the load duration curve, q is the scale exponent of CHP equipment cost and εMIN the minimum
allowed part load operation. Specific emphasis was given on the management of this equipment (Gelegenis et al, 2013a, 2015c). Technical barriers were also considered and wherever
necessary, possible solutions to overcome them were also attempted (Gelegenis, 2015). At last,
holistic optimization was elaborated, based on the most promising technical solutions that had
been revealed by the previous analyses (Gelegenis et al, 2015a). Indicative results for a real
case study were concluded, displaying the most suitable measures to be implemented; the latter
depend on the specific characteristics of the dwelling and the desired energy upgrade, as it is
shown in Fig. 1. In addition, the priorities between the available or alternative measures became
apparent, as it shown in Fig. 2.
Figure 1. Minimum necessary funds to upgrade the apartment of the Case Study, as a function of the targeted
energy-class. Suggested EEMs: Class E=No EEMs, existing situation (ES). Class D=ES + Conventional NG
boiler. Class C=ES + Condensing boiler. Class B=ES + Solar heating/DHW + Insulation of Northern wall. Class
B plus=The same as for class-B, + Insulation of the rest walls + Use of NG condensing boiler. Class A=The
same as for class-B plus + Replacement of windows.
Results and Discussion
The lacking regulations as regards the energy use in dwellings in Greece during the previous
decades, but also the wish to retain low the cost of equipment in dwellings, rises peculiarities in
the Greek domestic sector; these peculiarities is the reason of extraordinary energy consump-
ISBN: 978-960-98739-8-7
45
tion, but at the same time inspire great opportunities to effective energy management.
After the investigation of thousands of Energy Performance Certificates (EPCs) it was apprehended that the certification process deviates from the principal scope of the respective European Directive. In this sense, the spent efforts may achieve a limited only contribution to the energy awareness of people and to the energy upgrade of their dwellings (Gelegenis at al, 2013b).
Proposals to amend this problem and to enhance the effectiveness of the whole endeavor were
developed and documented.
Monitoring of modern central heating systems (via the development and installation of appropriate data-logging apparatus) was conducted for two successive heating periods, revealing in
this way their energy wasteful operation. The reasons of inefficiencies were properly examined
and techniques to overcome them were thoroughly tested and documented.
The use of energy efficient equipment like condensing boilers, heat pumps, cogeneration units,
PVs and several insulating alternatives, were studied in details by applying dynamic simulation; critical technical details (in favor or against a technical solution) appeared in this way,
which are rarely communicated by the suppliers/contractors to the end users however.
At the end it was found that the energy upgrade of dwellings should preferably follow a holistic
optimization approach, to maximize the savings and make the most of the available funds.
A lot of initiatives resulted with this project, which could be either undertaken by the owners
–concerning investments in energy efficiency measures- or by the State- as regards both the
revision of existing regulations and the control mechanisms for their fulfillment. Appropriate
energy education is essential to the efficacious selection and application of potential energy saving interventions but also to propose energy policy measures. In this basis, energy engineering
education should be improved in the Country, with regularly updated curriculum to cover modern areas of energy technology, as the latter presents an impressive progress during the current
years (Gelegenis and Harris, 2014, Gelegenis and Axaopoulos, 2015).
Figure 2. Optimization of applicable EEMs, as a function of the available budget. The limits between the energy
classes from E (upper) to A, are also depicted. Factors A to D are referred to four measures (A=insulation,
B=change of windows, C=change of boiler, D=solar collectors), with their coded values ranging from -1 (the
measure is not applied) to +1 (measure at its maximum assumed capacity).
Concluding Remarks
Energy upgrade of existing low performance dwellings can be achieved at reasonable cost up
to class A, provided that the analysis and holistic optimization that were developed in this work
would be applied. On the other hand, the prevailing practices of applying measures occasion46
ISBN: 978-960-98739-8-7
ally, according to the limited (and usually insufficiently documented) proposals included in the
energy performance certificates, and out of any strategic plan that would target to a specific
energy class, may even lead to the waste of funds and certainly to situations far away from what
could be optimally attained.
Acknowledgements
This work was co-financed by the European Union and the Greek Government under the framework of the Education and Initial Vocational Training Program - ARCHIMEDES III.
References
1. Axaopoulos I., Axaopoulos P., and Gelegenis J., 2014, Optimum insulation thickness for external walls on
different orientations considering the speed and direction of the wind, Applied Energy, 117, 167-175.
2. Bonaros B., Gelegenis J., Harris D., Giannakidis G., and Zervas K., 2013, Analysis of the energy and cost
savings caused by using condensing boilers for heating dwellings in Greece, Proceedings of the 5th International
Conference on Applied Energy ICAE 2013, July, Pretoria, South Africa.
3. Diakoulaki D., Zoupas P., and Gelegenis J., 2012, Treatment of uncertainty in the assessment of the financial
profitability of energy retrofit measures in the Greek building stock, Proceedings of the International Workshop
on Energy Efficiency: for a More Sustainable World, September, University of Azores, Ponta Delgada.
4. Gelegenis J., 2015, Innovative apparatus for compensation in central heating systems with elapsed time meters
(Patent pending, Νο 20150100162)
5. Gelegenis J., Diakoulaki D., Giannakidis G., Lampropoulou H., and Harris D., 2013a, Independent heating in
multi-family dwellings, or how an energy management measure may lead to overconsumption, Proceedings
of the Conference Energy for Sustainability 2013, Sustainable Cities: Designing for People and the Planet,
September, Coimbra, Portugal.
6. Gelegenis J., Diakoulaki D., Lampropoulou H., Giannakidis G., Mavrotas G., Axaopoulos P., Harris D., and
Samarakou M., 2013b, Energy management in buildings; analysis of the market trends and contradictions,
Proceedings of the 9th Panhellenic Conference of Chemical Engineering: The contribution of Chemical
Engineering to Sustainable development, May, Athens.
7. Gelegenis J., and Harris D., 2014, Undergraduate studies in energy education - A comparative study of Greek
and British courses, Renewable Energy, 62, 349-35.
8. Gelegenis J., Diakoulaki D., Lampropoulou H., Giannakidis G., Samarakou M., and Plytas N., 2014,
Perspectives of energy efficient technologies penetration in the Greek domestic sector, through the analysis of
Energy Performance Certificates, Energy Policy, 67, 56-67.
9. Gelegenis J., and Axaopoulos P., 2015, Residential cogeneration of heat and power: a promising way to
sustainability, a challenging field for tutors, International Journal of Higher Education and Sustainability (in
press).
10. Gelegenis J., and Mavrotas G., 2015a, Optimum sizing of residential cogeneration for prefeasibility estimations.
An analytical approach”, Energy Procedia, 75, 993 – 998.
11. Gelegenis J., and Mavrotas G., 2015b, Analytical study on the critical factors in residential cogeneration
optimization, Applied Energy (manuscript under revision).
12. Gelegenis J., Diakoulaki D., Mavrotas G., Axaopoulos P., Samarakou M., Lampropoulou H., Giannakidis G.,
Giantzoudis G., 2015a, Holistic energy upgrade of dwellings. A resource management optimization approach,
International Journal of Renewable Energy Research (accepted for publication).
13. Gelegenis J., Axaopoulos P., Misailidis S., Giannakidis G., Samarakou M., Bonaros B., 2015b, Feasibility for
the use of Flat Booster Reflectors in Various Photovoltaic Installations, International Journal of Renewable
Energy Research, 5(1), 82-98.
14. Gelegenis J., Diakoulaki D., Giannakidis G., Lampropoulou H., and Harris D., 2015c, Determination of
fixed expenses in central heating costs allocation: An arising issue of dispute, Management of Environmental
Quality, 26(6), 810-825.
ISBN: 978-960-98739-8-7
47
Ship Shape Optimization using a BEM-Isogeometric solver for wave
resistance
K.V. Kostas1, A.I. Ginnis2, C.G. Politis1* and P.D. Kaklis3
Naval Architecture Dept., Technological Educational Institute of Athens, [email protected]
School of Naval Architecture & Marine Engineering, National Technical University of Athens
3
Naval Architecture, Ocean and Marine Engineering, University of Strathclyde
1
2
Keywords: T-splines, NURBS, Hull parametric model, Wave-resistance, Optimization.
Abstract
The optimization of a hull form with respect to its hydrodynamic performance is considered
as a prerequisite for designing “green” ships. Shape optimization in naval hydrodynamics is a
demanding and challenging task, for several reasons: firstly, each cost function evaluation (e.g.,
wave resistance) is computationally expensive. Moreover, the output of the numerical solver
could be noisy, due to geometrical and numerical approximations, which generates spurious local minima and may yield the failure of the optimization process. In this context, the proposed
IGA approach is particularly interesting, since the geometry generated by the CAD modeler is
directly used by the solver without any geometrical approximations.
In our previous works (Ginnis et al 2013, Kostas et al 2014), IsoGeometric Analysis (IGA)
has been applied to the solution of the Boundary Integral Equation (BIE) associated with the
Neumann–Kelvin problem in the context of ship hull optimization with respect to the wave
resistance. In Ginnis et al (2013), we presented an optimization process combining modern optimization techniques, a NURBS multi-patch parametric ship-hull model and an isogeometric
NURBS based Boundary Element Method (BEM) solver for the required calculations. In Kostas et al (2014), we presented an alternative ship-hull optimization process combining a T-spline
based parametric ship-hull model and a T-spline based BEM solver (Ginnis et al 2014) for the
calculation of ship wave-resistance. In this work, we are focusing on presenting the optimization framework and demonstrating its application to shape optimizations for both local and
global ship-hull problems. Local problems include the optimization of the bulbous bow shape
of a container ship-hull against the criterion of minimum wave resistance. Global problems
include global hull optimization against multiple criteria and constraints (e.g., minimum wave
resistance constrained by a given deadweight, minimum wave resistance and minimum lightship weight constrained by a given deadweight, etc). Furthermore, the optimization framework
is demonstrated in multistage mode, where at a first stage a global optimization is performed
and at a second stage, further local optimizations are performed in areas of special interest.
Introduction
The optimization of a hull-form with respect to its resistance and resulting fuel consumption
has always been a major task in ship design. Moreover, as the design of the hull form is a prerequisite for the majority of ship design tasks, it is of great importance to complete this task in
the earliest possible time. The present work is focused in presenting an appropriate ship-hull
hydrodynamic optimization process, combining modern optimization techniques, a fully parametric ship-hull model and a BEM hydrodynamic solver. Both the parametric model and hydrodynamic solver are in-house developed. Note that the hydrodynamic solver adopts the concept
of Isogeometric Analysis (IGA), introduced by Hughes et al. (2005), which aims to intrinsically
48
ISBN: 978-960-98739-8-7
integrate CAD with Analysis by communicating the CAD model of the geometry (ship hull) to
the solver without any approximation, e.g., penalization.
The methodology for constructing the parametric ship-hull models is presented in Section SHIP
PARAMETRIC MODELS (SPM). The next section is devoted to the presentation of the CFD
solver’s basic features. The CFD wave resistance solver is based on the Neumann–Kelvin formulation of the problem (Baar & Price 1988). The resulting BIE is numerically solved using
a higher order collocated BEM, which adopts the IGA concept. In The Multi-Objective Optimization Environment section we present the optimization environment integrating the SPM,
the IGA–BEM solver and the employed optimization libraries for designing ship hulls with
minimum resistance. Finally, three optimization cases for a container ship are set-up and presented in the last section. The first case deals with bulbous bow optimization (local optimization
problem) against the criterion of minimum wave resistance under a given displacement. The
second case involves a global ship-hull minimization problem against two objective functions:
total resistance and deviation from a reference ship capacity, i.e., ship’s deadweight. The final
case involves a global and local optimization process in two stages.
Ship Parametric Models (SPM)
NURBS is the de facto standard in representing curves and surfaces in commercial and open
source software packages. Contemporary CAD systems for ship-hull design rely almost exclusively on NURBS and thus its support is imperative. However, the use of NURBS in Isogeometric Analysis exhibits some deficiencies:
1. Due to their tensor-product nature a large number of NURBS control points are not unlikely
to be superfluous, in the sense that they contain no significant geometric information.
2. Refinement (required in the numerical solution of BIE) leads to the insertion of entire rows/
columns of control points, thus increasing complexity and implementation effort.
3. The limitations of rectangular topology lead to multi-patch representations when complex
objects, as e.g., ship hulls, are under consideration.
T-splines constitute a recently developed generalization of NURBS technology that removes
most of the above mentioned NURBS deficiencies. The main advantages of T-splines technology are:
1. T-splines permit representation of complex objects with a single T-spline patch.
2. A T-spline control grid is allowed to have partial rows of control points, terminating in Tjunctions, which allow for local refinement. These T-junctions permit the significant reduction
of superfluous control points.
3. It is possible to merge multiple NURBS patches into a single, gap-free T-spline.
The basic shape characteristics of a typical ship-hull comprise:
• a partition of the ship-hull into three main parts, namely the bow, the midship part and the
stern,
• global dimensions (e.g., length between perpendiculars (Lbp), beam (B), depth (D), draft
(T)) as well as dimensions characterizing each one of the aforementioned ship parts (e.g.,
extent of the midship part),
• a set of control curves that are of boundary (e.g., stern profile, bow profile) or shape-transition character (e.g., FoS (flat-of-side) and FoB (flat-of-bottom) curves) and,
• local geometrical characteristics that serve functional, structural and/or hydrodynamic purposes, e.g., bow-angle of entrance at waterline, bulb-top position, bilge radius, shaft height,
etc.
ISBN: 978-960-98739-8-7
49
On the basis of the above coarse shape description and a set of parent ship-hulls, a list of exposed and internal parameters is devised. The total number of exposed parameters is 30/24 and
are split as follows: 5/3 global, 7/9 for the midship and 8,10/7,5 for the bow and stern parts
respectively; see Figures 1and 2.
Figure 1: Exposed Parameters for the NURBS SPM
l
Lw
T
S_
Fo
L
d
S_w
Fo
_L
id
M
ftL
S_A
Fo
Tran
im
s_D
_R
e
g
ild
B
s
o
_P
id
M
_R
e
g
ild
B
_
B
Fo
Tran
im
s_D
/2
B
ftL
_A
B
Fo
L
d
_w
b
Fo
Figure 2: Exposed Parameters for the T splines SPM
IGA-BEM wave resistance solvers
For the wave resistance solvers, a high-order Boundary Element Method (BEM) based on IsoGeometric Analysis (IGA) is applied for the numerical solution of the Boundary Integral Equation (BIE) (see Equation 1), as described in detail in (Belibassakis et al 2013) and (Ginnis et al
2014).
µ (P)
∂G (P, Q)
1
∂G* (P, Q)
− ∫ µ (Q)
dS (Q) − ∫ µ (Q)
nx (Q)τ y (Q)d (Q) = U·n(P), P, Q ∈ S , S
2
∂n(P)
k 
∂n(P)
(1)
where µ is the density of the Neumann-Kelvin Green function G(P,Q), −U denotes the steady
forward speed of the ship and k = g/||U||2 denotes the characteristic wave number. The IGA philosophy attempts to define the approximate field quantities of the boundary-value problem in
question from the basis that is being used for representing the geometry of the body boundary.
In the case of the boundary integral equation (Equation 1), the dependent variable is the sourcesink density µ, distributed over the body boundary S. The latter is accurately and efficiently
represented as a multi-patch NURBS or T-spline surface as follows:
nq
q
ncp
q =1
i =1
 , q = 1, …, n
S =  S q , S q (î ) = ∑ b iq Riq (î ), î ∈ Ω
q
q
50
(2)
ISBN: 978-960-98739-8-7
where î is the parameters’ vector, nq is the number of patches and q runs over the number of
patches.
In conformity with the IGA concept, the unknown source-sink surface distribution µ is approximated by the very same basis used for the body-boundary representation, see equation (2), that
is:
q
ncp
µ (P ) = ∑ µiq Riq (P ),
q
i =1
P ∈ S , q = 1, …, nq ,
(3)
where Riq (P) = Ri q (î (P)) . Inserting equation (3) into BIE (equation (1)) we obtain a linear
system with respect to the unknown coefficients µiq . From the solution of this linear system,
various quantities, such as velocity, pressure distribution, ship wave pattern and ship wave resistance can be obtained.
The multi-objective optimization environment
Simulation-based optimization is of growing importance in naval engineering, since it allows
improving ship performance for a moderate cost, in comparison with towing tank experiments.
A major difficulty to apply an automated shape optimization procedure is the development of
a fully automated design loop. Indeed, for each set of parameters, a geometric hull model has
to be constructed, allowing the generation of the computational domain used by the solver to
provide the physical response and the performance analysis. All these steps should be fully
automated, without hand-made repairing or arranging process, in order to feed the optimization
algorithm and finalize the design loop. A second obstacle arises from the simulation process: for
complex test-cases, CFD simulations are expensive, in terms of computational time. Moreover,
the numerical solutions obtained can be polluted by errors arising from the discretization and
iterative methods, yielding noisy performance evaluations.
The employed optimizers are the commercial tool modeFrontier® from Esteco along with an
optimization library, equipped with various algorithms ranging from descent methods to semistochastic algorithms, developed by INRIA. A design optimization software environment, depicted schematically in Figure 3, has been setup including the three main components: the
optimizer, the solver and the modeler. Each of these components is wrapped in a corresponding
wrapper that manages the communication and data exchange among the components of the
optimization environment.
Figure 3: Schematic diagram of the optimization environment
Numerical results
Bow-optimization for minimizing wave resistance: The optimization environment is firstly
illustrated for the NURBS setting for a local optimization problem. The bulbous bow area of a
ISBN: 978-960-98739-8-7
51
container ship is optimized against the criterion of minimum wave resistance, under the constraint of given displacement. The main particulars of the ship have as follows: Lpp= 277 m, B
= 32.2 m T = 13.0 m, Cb = 0.6938 and Vs = 26 knots. Since we are interested in optimizing
the bow ship area, the design parameters are chosen to be: B_BulbLength, B_BulbTopPosition,
B_BulbRadius and B_BulbRise; see Figure 1. The employed algorithm is the direct-search optimization strategy from INRIA’s optimization library (Ginnis et al 2013) is carried out to find
optimal values for the design parameters of the bow. The bound constraints of of the design
parameters are implemented using a penalization approach while the displacement constraint is
fulfilled within the SPM construction process.
Figure 4: Evolution of wave-resistance & bow-area shape along with source distribution
Figure 4(b) illustrates the evolution of the wave resistance during the optimization procedure.
More than half of the reduction is obtained during the first 15 simulations, whereas the convergence is then slower. This is due to the fact that the constraints become active after a first
phase of straight descent. Especially, the displacement constraint is highly non-linear and is the
cause of the observed irregular function decrease. Moreover, Figures 4 (a) and (c) depict the
bow-area shape along with the corresponding distribution of µ for two representative instances
of the parametric model.
In a second test of the optimization environment we globally optimize a T-spline ship hull with
respect to the following two criteria: a) minimum total resistance RT and b) minimum deviation from a reference deadweight. The last test of the optimization environment comprises two
stages: Firstly, we globally optimize a T-spline ship hull with respect to the following two criteria: a) minimum total resistance RT and b) minimum deviation from a reference deadweight.
In the second stage we perform a local optimization of the bulbous area using as initial hull, a
suitable hull from the pareto front of the first stage.
References
1. BELIBASSAKIS, K.A., Th.P. GEROSTATHIS, K.V. KOSTAS, C.G. POLITIS, P.D. KAKLIS, A.I. GINNIS
and C. FEURER, 2013, “A BEM-ISOGEOMETRIC method for the ship wave-resistance problem”. Journal
of Ocean Engineering, 60, 53-67.
2. BAAR, J. J. M., W. G. PRICE, 1988, “Developments in the Calculation of the Wavemaking Resistance of
Ships”. Proc. Royal So-ciety of London. Series A, Mathematical and Physical Sciences, 462 (1850), 115–
147.
3. GINNIS, A.I., K.V. KOSTAS, C.G. POLITIS, P.D. KAKLIS, K.A. BELIBASSAKIS, Th.P. GEROSTATHIS,
M.A. SCOTT and T.J.R HUGHES, 2014, “Isogeometric Boundary-Element Analysis for the Wave-Resistance
Problem using T-splines”. Journal of Computer Methods in Applied Mechanics and Engineering, 279, 425439.
52
ISBN: 978-960-98739-8-7
4. GINNIS, A.I., R. DUVIGNEAU, C. POLITIS, K.V. KOSTAS, T. GEROSTATHIS and P.D. KAKLIS, 2013,
“A Multi-Objective Optimization Environment for Ship-Hull Design based on a BEM-Isogeometric Solver”.
The fifth Conference on Computational Methods in Marine Engineering (Marine 2013), Hamburg, Germany.
5. HUGHES, T. J. R., J. A. COTTRELL and Y. BAZILEVS, 2005, “Isogeometric analysis: CAD, finite elements,
NURBS, exact geometry and mesh refinement”. Journal of Computer Methods in Applied Mechanics and
Engineering, 194, 4135–4195.
6. KOSTAS, K.V., A.I. GINNIS, C.G. POLITIS and P.D. KAKLIS, 2014, “Ship-Hull Shape Optimization
with a T-Spline based BEM-Isogeometric Solver”. Journal of Computer Methods in Applied Mechanics and
Engineering, Isogeometric Analysis Special Issue, 284, 611-622.
ISBN: 978-960-98739-8-7
53
Energy harvesting from sea waves and currents using oscillating
hydrofoils
K. Belibassakis1*, T. Gerostathis2, E. Filippas1, J. Touboul3 and V. Rey3
School of Naval Architecture & Marine Engin. National Technical University of Athens,
* [email protected]
2
Dept. of Naval Architecture, Technological Educational Institute of Athens
3
Université de Toulon, CNRS / INSU, IRD, MIO
1
Keywords: renewable energy, oscillating hydrofoils, panel methods
Abstract
In this work the use of oscillating hydrofoils as a renewable energy production system is investigated when operate in the presence of sea waves and currents. In previous works (Belibassakis & Politis 2013, Filippas & Belibassakis 2014) potential based panel methods have been
developed for the prediction of the hydrodynamic behaviour of flapping hydrofoils used as
thrusters on ships for augmenting propulsion. This analysis showed that a significant amount of
energy can be extracted by these kind of hydrofoils, if active pitch control is used. Following
this works, the present model takes into account effects of surface waves and velocities due to
waves and currents on the description of the incident field. The operability characteristics of
the system are investigated through numerical experiments verifying the ability of the system
to produce significant energy output under general operating conditions. The results of these
investigation can be used for the design and optimal control of such systems when operating in
nearshore regions.
Introduction
In this work oscillating hydrofoils are investigated, as energy productions, systems, in the presence of waves and currents. As a starting point, the energy extraction in harmonic waves and
current systems, using active pitch control of hydrofoils, is examined using Boundary Element
Methods. The model takes into account the effect of the wavy free surface and the velocity
component due to waves and currents on the formation of the flow.
Mathematical Formulation
In this problem an area of variable bathymetry is considered, where waves and currents exist,
connecting two regions of constant, but different, depth, see Fig. 1. The depth and the current
functions vary slowly along the horizontal direction, in the variable bathymetry region. In addition, the vertical structure of the current is locally uniform. In this area an oscillating hydrofoil
is operating as a combined wave and current energy harvesting device, of chord c, performing
simultaneously vertical heaving and rotational pitching oscillations, while being connected to
an energy generator system; see Fig. 1.
Harmonic incident waves, of small amplitude, is assumed permitting the linearization of the
free-surface boundary conditions on the mean free surface level; see Fig. 1. The domain of definition of the problem D ⊆ � 2 is open semi-bounded with smooth boundary ∂D (Fig. 1). The
problem is time dependent and the hydrofoil is represented by a moving boundary ∂DB (t ) . The
total wave potential ΦT ( x, y; t ) is considered as the sum of the current potential Ux, the
54
ISBN: 978-960-98739-8-7
Figure 1. Oscillating hydrofoil, of chord c and depth d, under the free surface.
propagating wave field potential Φ I ( x, y; t ) and the disturbance potential Φ ( x, y; t ). The propagating filed potential is calculated following previous works (see, e.g. Belibassakis and Athanassoulis 2014, Belibassakis et al 2014]) and it is the solution of the wave-current propagation
problem in the absence of the oscillating hydrofoil. The disturbance field, which behaves as
outgoing radiating waves, satisfies the 2D Laplace Eq. 1, supplemented by the body boundary
condition, Eq.2, the bottom no-entrance boundary condition, Eq.3, the linearized dynamic and
kinematic boundary conditions, Eq.4 and Eq.5, which are satisfied on the mean free surface,
∆Φ ( x, y; t ) = 0 , ( x, y ) ∈ D,
(1)
∂ (Ux + Φ I ( x, y; t ) )
∂Φ B ( x, y; t )
=b� −
+ VB ⋅ nB , ( x, y ) ∈ ∂DB , ∂nB
∂nB
(2)
∂Φ H ( x, y; t ) dH ∂Φ H ( x, y; t )
+
= 0 , y = − H (x ),
∂y
dx
∂x
∂Φ F ( x, y; t )
∂Φ F ( x, y; t )
= −U
− gη ( x; t ) , on y = 0,
∂t
∂x
∂η ( x; t ) ∂Φ F ( x, y; t )
∂η ( x; t )
=
+U
, y = 0,
∂t
∂n
∂x
(3)
(4)
(5)
In the above system, η ( x; t ) denotes the free surface elevation, associated with the disturbance
field, g the gravitational acceleration, H (x ) the depth function, d the mean depth of the hydrofoil and VB its motion derivative, which consists of vertical oscillations Y(t) simultaneously
with pitching oscillations Θ(t) , with respect to a pivot point selected near the pressure center of
the hydrofoil. The pitching motion is enforced, controlling the angle of attack, and the response
is the power extracted by the vertical oscillations, given by the expression,
2
Po = b Y , where mY + bY + kY = LY , Θ (t ) = θο cos (ωΘt +ψ ) (6)
where m is the mass of the hydrofoil, ωΘ , θο , ψ are the frequency, amplitude and phase of the
pitching motion and b, k are characteristics of the power take off and stiffness of the generator and the elastic connector. With LY (t ) we denote the vertical lift forces obtained through
hydrodynamic pressure integration on the foil. We consider harmonic incident waves of angular
frequency ω equal to the frequency of the controlled pitching motion ω = ωΘ . Also we denote
ISBN: 978-960-98739-8-7
55
with the indices {B, H , F , W } the disturbance field Φ ( x, y; t ) on the body surface, the seabed,
the free surface and the wake of the hydrofoil, respectively. The problem is supplemented by
the Kutta condition at the hydrofoil, necessitating equal pressure at the trailing edge and continuous pressure and normal velocity through the trailing vortex wake,
∂ΦW+ ( x, y; t ) ∂ΦW− ( x, y; t )
p ( x, y; t ) = p ( x, y; t ) , ( x, y ) ∈ ∂DW ,
(7)
=
, ( x, y ) ∈ ∂DW .
∂nw
∂nw
where (± ) indicate the upper and lower sides.
Boundary integral formulation
Applying the representation theorem, Kress (1989), to our problem, we obtain a system of two
+
W
−
W
equations, for Φ F and Φ B , for points at the free surface and the hydrofoil contour, respectively,
as follows
G ( x 0 | x)
G ( x 0 | x)
1
Φ B / F (x 0 ; t ) = ∫∫ b(x; t )G (x 0 | x) − Φ B (x; t )
dS (x) − ∫∫ Φ H (x; t )
dS (x)
2
∂
∂
n
n
∂D (t )
∂D
B
H
∂G (x 0 | x)
∂Φ F (x; t )
(8)
G (x 0 | x) − Φ F (x; t )
dS (x) ∂n
∂n
∂DF
In the above relation we have used the Green’s function consisting of the fundamental solution
of 2D Laplace equation corresponding to a Rankine source
+
∫∫
ln r (x 0 | x)
,
(9)
2π
where x 0 = ( x0 , y0 ) is the field point and x = ( x, y ) is the integration point. Eq. 8 is used to
set-up a Dirichlet-to-Neumann map (DtN) of the boundary values concerning Φ F , Φ B and its
normal derivatives. The latter integral map, after discretization, is applied to the numerical
integration of the free surface boundary conditions (4) and (5), treated as a dynamical system,
providing us with the evolution of the unknown free surface at the specified level of approximation. For more details on the numerical solution see Filippas and Belibassakis (2014).
Numerical results
Assuming relatively large hydrofoil depth, a simplified version of the present method is applied for lifting computations, where the effects of the free-surface have been approximately
omitted in the representation of the disturbance potential, while the effect of time dependent
terms in the foil boundary condition due to propagating waves and foil oscillation is taken into
account. For the numerical experiments, a NACA0012 hydrofoil is considered of chord c=1m
in uniform water layer of depth H / c = 10 , operating in incident waves of amplitude a / c = 0.2
G ( x 0 | x) =
and current, characterised by bathymetric Froude number Fn = U / gH = 0.1 and Strouhal
number St = ω H / g = 0.6 . The shoaling ratio is H / λ = 0.091. The hydrofoil, with d=H/2
is performing simultaneous heaving and pitching oscillations, and the corresponding reduced
frequency is kr = U ω / 2c = 0.3. In this case, the enforced pitching motion has amplitude
Θ0=10deg and its phase lag with respect to the incoming wave is ψ = 0 deg. Indicatively the
angle of attack against the pitching motion is shown in Fig. 2a. In Fig. 2b the resulting heaving
oscillation is plotted. We use b / (2 m ωR ) = 0.5 to model the constants of the PTO, corresponding to resonant frequency ωR = k / m , equal to ωR = 0.71rad / s. The output of
56
ISBN: 978-960-98739-8-7
Figure 2. a) Foil pitching oscillation of amplitude θ0=10deg in incident waves and current of and corresponding
angle of attack, b) Output heaving motion and lift and moment coefficients in the case of oscillating hydrofoil of
(a).
Figure 3. Output power coefficient in the case of Fig. 2.
the considered system operating in the above conditions is plotted in Fig. 3 using the power
coefficient CPb = P / (0.5 ρU 3c). We observe, in this case, that the extracted energy with mean
value CPb =0.027 is significant, especially as compared to the power needed for controlling the
pitching foil motion, indicated in Fig. 3 through the corresponding pitching power coefficient
and shown by using dashed lines. In this work also results from a parametric investigation of
the examined system are presented, concerning the effect of Froude number, Strouhal numbers
and pitching on the average output power.
Acknowledgements
The preparation of this extended abstract have been funded by the Research Funding Program
ΕΠΕΕ ΤΕΙ Α 2015: Researchers Support Internal Program of TEI of Athens for the year 2015.
References
1. Belibassakis K.A., Politis G.K., 2013, Hydrodynamic performance of flapping wings for augmenting ship
propulsion in waves. Ocean Engineering 72, 227-240.
2. Filippas E.S., Belibassakis K.A., 2014, Hydrodynamic analysis of flapping-foil thrusters operating beneath
the free surface and in waves. Engineering Analysis Boundary Elements. 41, 47-59.
3. Belibassakis K.A., Athanassoulis G.A., 2011, A coupled-mode system with application to nonlinear water
waves propagating in finite water depth and in variable bathymetry regions. Coastal Engineering. 58, 337350.
4. Belibassakis K.A., Athanassoulis G.A. and Gerostathis T.P., 2014, Directional wave spectrum transformation
in the presence of strong depth and current inhomogeneities by means of coupled-mode model. Ocean
Engineering. 87, 84–96.
5. Kress, R., 1989, Linear integral equations, Berlin etc., Springer-Verlag.
ISBN: 978-960-98739-8-7
57
Application of a 3D coupled-mode model to the hydroelastic
analysis of very large floating bodies over inhomogeneous seabeds
Th.P. Gerostathis1*, K.A. Belibassakis2 and G.A. Athanassoulis3
Dept. of Naval Architecture, Technological Educational Institute of Athens,
* [email protected]
2
School of Naval Architecture & Marine Engin. National Technical University
of Athens
3
ITMO University, St. Petersburg, Russia
1
Keywords: hydroelastic analysis, very large floating bodies, coupled-mode model,
water waves
Abstract
In the present work, a 3D extension of the coupled-mode model developed by Belibassakis
& Athanassoulis (2005) is applied to the hydroelastic analysis of very large floating bodies.
The bathymetry, in the vicinity of the bodies, is considered variable, in both horizontal directions, joining two regions of constant, but possibly different, depth. The propagation-scattering
problem of monochromatic incident surface waves is considered, with various angles of attack,
under the influence of a variable bathymetry region and a very large floating elastic plate. Linearized water-wave and thin elastic plate theory is applied to the formulation of the hydroelastic
problem, under the assumption of small-amplitude waves and small plate deflections. For the
representation of the wave field under the floating plate, a complete local modal series expansion is used. Emphasis is given to applications of the model, concerning plates with different
aspect ratio of the horizontal dimensions and over flat and inhomogeneous seabeds, in order to
investigate the hydroelastic behaviour.
Introduction
In the present work, the coupled-mode model developed by Belibassakis & Athanassoulis
(2005) is extended and applied to the hydroelastic analysis of 3D large floating bodies, lying
over uneven bottom profiles. The scattering problem of harmonic, obliquely-incident surface
waves is considered, under the combined effects of the bathymetry and a floating elastic plate of
orthogonal shape. Numerical results, concerning various seabeds, are comparatively presented,
and the effects of wave direction, bottom slope and bottom undulations on the hydroelastic
responses are investigated.
Formulation of the problem
The examined physical problem, see Fig. 1, consists of a water layer bounded above by the free
surface and by a very large thin plate or an ice sheet of small thickness, and below by the sea
bottom. It is also assumed that the bottom surface exhibits a variation extending from a deep flat
region (region of incidence) to a shallow one (region of transmission), including into its interior
the bottom inhomogeneity and the floating structure. We consider an incident monochromatic
wave, of angular frequency ω, propagating from the deep area to the shallow one, having an oblique direction with respect to the bottom parallel contours. A function h = h ( x, y ) represents
the depth variation, measured from the mean water level. The shape of the plate is orthogonal
with length L and breadth B, in the variable bathymetry region; see Fig. 1. The wave potential
Φ (x, y, z;t ) and the complex amplitude of the free-surface elevation (η) are expressed in the
following form,
58
ISBN: 978-960-98739-8-7
Figure 1. Thin elastic plate over uneven bottom.
Φ ( x, y , z;t ) = Re {ϕ ( x, y , z )exp (−iωt )},
η ( x, y ) =
(1)
i ∂ϕ ( x, y , z = 0 )
,
ω
∂z
(2)
where g is the acceleration due to gravity. The deflection (w) of the elastic plate is given by a
similar relation derived from the kinematical condition at the water-solid interface,
w ( x, y ) =
i ∂ϕ ( x, y , z = 0 )
ω
∂z
(3)
The differential form of the formulation of the problem consists, in the water later, of the Laplace equation,
(∇
2
+ ∂ 2z )ϕ = 0,
in −h ( x, y ) < z < 0,
( x, y ) ∈ W  E
,
(4a)
(4b)
(4c)
the linearized free-surface boundary condition,
∂ zϕ − µϕ = 0 ,
on
z=0,
( x, y ) ∈ W .
and the no-entrance boundary condition
∂ zϕ + ∇h∇ϕ = 0 ,
on
z = − h ( x, y ) , ( x, y ) ∈ W  E ,
where µ = ω 2 / g the frequency parameter.
For points on the plate the wave potential satisfies the dynamical equation forced by the water
pressure
( )
∇ 2 d ∇ 2 w + (1 − ε ) w =
iµ
ϕ (x ) ,
ω
on
z = 0,
( x, y ) ∈ E
.
(4d)
where the plate edges satisfy the following two conditions,
∂ 3w
∂ 3w
∂ 2w
∂ 2w
+ (2 − ν )
= 0,
+ν
= 0 at
∂n 3
∂n ∂ 2 s
∂ n2
∂ s2
ISBN: 978-960-98739-8-7
(x, y )∈ ∂W
,
(4ef)
59
where n and s denote, respectively, the normal and tangential derivatives. Variables d and ε
are given by the expressions d = D / ρ g and ε = mω 2 / ρ g , where D = Et 3 / 12 (1 − ν 2 ), represents the flexural rigidity of the elastic plate, t the variable thickness and ν the Poisson’s ratio.
Moreover, ρ denotes the fluid density and m the mass per unit area of the plate. Finally, the edge
conditions (4e,f) state that the ends of the plate are free of shear force and moment.
Modal representation of the potential
The problem of water wave propagation and scattering in uneven bottom, have been studied
in various works by the authors, e.g. (Athanassoulis & Belibassakis 1999, Belibassakis et al
2001), where the following local mode series expansion is applied to represent the wave field
in the water region:
∞
ϕ (x, z ) = ϕ −1 (x ) Z −1 (z; x ) + ∑ϕ n (x ) Z n (z; x ) , −h (x ) < z < 0, x = ( x, y ) . (5)
n =0 (5) is enhanced by an appropriate sloping-bottom mode
The modal series expansion
ϕ −1 (x ) Z −1 (z; x ) which enables the consistent satisfaction of the Neumann bottom-boundary
condition on a general topography and accelerates the convergence.
Coupled-mode system
A coupled-mode system (CMS) of differential equations is obtained by using the modal representation of the wave potential (5) to a variational principle of the problem developed by Belibassakis Athanassoulis (2005). This permits the reformulation of Eqs.(4) with respect to the
unknown modal amplitudes ϕ n (x ), n = −1,0,1, 2,....., for x ∈ W  E . The present CMS takes
the following form (Belibassakis & Athanassoulis 2005)
∞
∑ a (x )∇
n =−1
mn
2
ϕ n (x )+ bmn (x )∇ ϕ n + cmn (x ) ϕ n (x ) = iω w (x ) ⋅ χ (Ε ),
m = −1,0,1,.... , (6a)
where χ (E ) = 1, for x ∈ E , and 0 otherwise. The system (6) is supplemented by an equation
completing the coupling between the plate deflection w and the modes ϕ n
iµ
∇ (d ∇ w )+ (1 − ε ) w =
ω
2
2
∞
∑ ϕ (x )
n =−1
n
.
(6b)
Numerical results
The CMS is solved using appropriate boundary conditions specifying the wave incidence (see
Fig.1) and the boundary conditions Eqs.(4e) and (4f) associated with elastic plate deflection
at the edges, enforcing zero shear force and moment. The discrete version of the present hydroelastic CMS is obtained by truncating the local-mode series (5) to a finite number of terms
(modes), and using central, second-order finite differences to approximate the horizontal derivatives.
In order to investigate the effects of the uneven bottom on the hydroelastic behaviour of the
plate, a numerous numerical experiments have being performed in two cases a) a smooth shoal;
Fig. 2, and b) an undulating bottom bathymetry. As an example, in Fig. 2, the geometrical setup
of the first case is presented where a large elastic floating plate having a length L=240m and
breadth B=120m is imposed, with constant characteristics d=105m4, ε=0.005 and Poisson’s
ratio ν=0.3. The floating body is lying over a smooth underwater shoal, characterised by a
depth function smoothly varying from h=15m to h=5m over a distance of 1.5km, as also shown
60
ISBN: 978-960-98739-8-7
in Fig.2. In order to illustrate the 3D effects, in Fig.3, the deflection modulus is plotted along
various sections of the plate (see Fig. 3a), where the elastic deformation is maximized at the
upwave edges of the plate near the corners. Similar results, illustrating the effect of the incident
wave angle on the modulus of the plate deflection are also available. In addition, the effect of
the plate aspect ratio to the hydroelastic wave field is tested, applying the model to a shorter (L/
B=1) and a longer plate (L/B=4). In the case of undulating bottom bathymetry, the effect of the
amplitude of bottom corrugations to the wave field structure and to the hydroelastic behaviour
of the plate is examined.
(a)
(b)
Figure 2. (a) Geometrical setup. (b) Effect of bathymetric variations on the wave field and the elastic plate deflection (wave period T=15s).
(a)
(b)
Figure 3. (a) Distribution of deflection of the elastic plate for the case of Fig. 1. The wave incidence is normal.
(b) Normalized plate deflection along the longitudinal and transverse sections shown in (a).
Acknowledgements
This research has been co-financed by the European Union (European Social Fund – ESF)
and Greek national funds through the Operational Program “Education and Lifelong Learning” of the National Strategic Reference Framework (NSRF) - Research Funding Program:
ARCHIMEDES III. Investing in knowledge society through the European Social Fund.
ISBN: 978-960-98739-8-7
61
References
1. Athanassoulis, G.A., Belibassakis, K.A., 1999, A consistent coupled-mode theory for the propagation of
small-amplitude water waves over variable bathymetry regions. J Fluid Mech. 389, 275-301
2. Belibassakis, K.A. & Athanassoulis, G.A., 2005, A coupled-mode model for the hydroelastic analysis of large
floating bodies over variable bathymetry regions, J. Fluid Mech. 531, 221–249.
3. Belibassakis, K.A., Athanassoulis, G.A., Gerostathis, T. (2001) A coupled-mode system for the refractiondiffraction of linear waves over steep three dimensional topography. Applied Ocean Research. 23, 319-336.
62
ISBN: 978-960-98739-8-7
Degradation effects in field-aged c-Si solar panels of a grid-tie
photovoltaic plant
Konstantinos Gkarakis1, Konstantinos Loukidis2, Petros Axaopoulos1
1
2
RES Lab, Department of Energy Technology, TEI Athens - [email protected]
ENTEKA SA, [email protected]
Keywords: natural ageing, I-V, P-V curve analysis, IR thermography, degradation.
Abstract
The increased interest for solar (photovoltaic - PV) projects worldwide during the recent years
has led to the rapid grown of module production, as a result manufacturers and operators have
been investigating procedures to evaluate the factors affecting the performance and efficiency
of the PV plants. The degradation effects of PV modules are the main reason for significant
performance change in PV plants. This paper presents the performance change, of a 1.54kWp
grid-connected, roof-mounted PV plant located on the RES Lab of TEI Athens- Greece that
has been operating since 2004, due to degradation effects of the panels. Visual inspection, IR
thermography and I-V, P-V measurements has been performed on each of the 28 PV modules.
Visual inspection has showed discoloration of the EVA and degradation of the AR coating and
only a few hot spots have been revealed through IR thermography. I-V, P-V measurements show
performance degradation, higher than the one specified by the manufacturer (1% per year).
Introduction
The global solar photovoltaic (PV) market has experienced a new year of growth in 2014 reaching a cumulative capacity of 178 GW. Due to the massive decline of the PV panels prices
achieved in recent years and continued in 2014, solar power is now broadly recognized as a cost
competitive, reliable and sustainable energy source. The global solar PV cumulative installed
capacity has increased from 25GW at 2008 to 178 GW at 2014. The estimation of SolarPowerEurope (Masson, 2015) is that the 540 GW mark at a global level could be reached in five
years’ time. It is noteworthy that Solar PV is covering more than 7 % of the electricity demand
in three countries in Europe: Italy, Germany and Greece. Greece’s cumulative PV capacity is
2.6GW and was mainly installed recently at the period from 2009 to 2013 (Hellenic Association of Photovoltaic companies, 2015). As a result the majority of PV plants are quite new in
operation (6-7 years) facing the natural ageing in field conditions.
PV modules (and of course cells) ageing is a process that naturally evolves through the years
of operation in field conditions. Ageing effects mainly include discoloration of the encapsulant,
usually ethylene-vinyl-acetate (EVA) (Czanderna and Pern, 1996), degradation of the anti-reflective (AR) coating, the formation of hotspots (King et al, 2000 and Simon, 2010), moisture
intrusion, delamination and corrosion (Quintana et al, 2002) , cracks, tears and bubbles in the
backsheet. Several ageing effects may co-exist even in the same cell and module (Kaplani,
2012a), while optical/ physical, electrical and thermal degradation effects may be linked with,
to a lesser or greater degree, with the power and performance degradation of the module (Chianese et al, 2003 and Paretta et al, 2005) . The progression of ageing is generally measured in
percentage/ per year of power and performance degradation. The PV modules manufacturers
set this value from 0,8-1,5% per year. Also, stronger ageing effects may be introduced by external factors causing continuous or severe damage to the module, including overlaying dust, dirt
and bird droppings, surrounding vegetation or fence, electric pole, causing continuous partial or
total shading and localized temperature increase (Kaplani 2012b), strong winds and hail causISBN: 978-960-98739-8-7
63
ing mechanical shocks, damage of the front glass, back sheet or frame of the module.
This paper presents ageing effects in c-Si PV modules due to natural exposure in field conditions for over 10 years. Using IR thermography and I-V curve analysis, the results of this paper
shows that naturally aged modules (in the specific case) present higher degradation ratio than
the manufacturer’s standard (1% per year) and differentiation between them. Also, temperature
degradation effects are identified through IR thermography in bus bars, contact solder bonds,
blisters, hot spots, and hot areas. I-V curve analysis results showed an agreement between the
source of electrical performance degradation and the degradation effects in the defected cell
identified by the IR thermography.
Experimental procedure
Experiments were performed during a period of several months on SIEMENS SM55 and ISOFOTON I-55 PV modules with 14 and 10 years of field operation, respectively, at the RES Lab
of TEI Athens, Greece. Both digital images and IR images were obtained from cells exhibiting
different degrees and forms of optical degradation. Images of the cell neighborhood were also
obtained for relative comparison reasons.
Digital images were obtained using a conventional 7.1 MP digital camera. The IR images were
obtained using FLIR T335 thermal imaging camera with detector resolution 320 x 240 pixels
and accuracy of ±2◦C. For the I-V curves obtained, an I-V curve analyzer for photovoltaics,
HT IV-400, was used together with reference cell HT340N used as an irradiance sensor and
temperature probe for measuring module temperature. For each measurement the temperature
coefficients of Pm, Isc, and Voc for the PV module were set. The digital and IR images were
processed and analyzed through special software packages.
Figure 1. I-V and P-V measurement devices.
The reduction in power capacity of the 1.54kWp PV plant after 10 years of operation is 13%
with a variance from 6% to 22%. Siemens Solar PV modules with a total age of 14 years were
appeared 12% reduction of capacity and 14% the Isofoton PV modules with total age of 10
years. The measurements of one indicative PV module are presented on table 1 and figure 2.
Optical degradation effects have been observed in many PV modules (figure 3) but thermal
degradation effects are limited (figure 4).
64
ISBN: 978-960-98739-8-7
Table 1. I-V and P-V measurements of a SM55 PV module.
Figure 2. I-V and P-V curves from an aged SM55 PV module.
Figure 3. Various degradation effects observed in PV modules.
Figure 4. IR image of a browned cell which is overheated (hot spot).
ISBN: 978-960-98739-8-7
65
Conclusions
The detection of degradation effects in operational grid connected PV plant through optical
check, IR thermography and I-V, P-V measurements brings several advantages in the area of
defect diagnosis of PV cell and module condition. All these procedures, if applied correctly,
contribute to the early diagnosis of faults, assisting in productive and reliable operation of the
PV plant.
References
1. D. Chianese, A. Realini, N. Cereghetti, et al. “Analysis of weathered c-Si PV modules”. 3rd World
Conference on Photovoltaic Energy Conversion, May 11-18, Osaka, Japan, 2003.
2. A.W. Czanderna, F.J. Pern. “Encapsulation of PV modules using ethylene vinyl acetate copolymer as a
pottant: A critical review”. Solar Energy Materials and Solar Cells, 43: 101-81, 1996.
3. Hellenic Association OF PHOTOVOLTAIC companies, Greek PV Market statistics 2014, Date of
access: 01/09/2015. http://helapco.gr/pdf/pv-stats_greece_2014_Feb2015_eng.pdf
4. E. Kaplani. “Detection of degradation effects in field-aged c-Si solar cells through IR thermography and
digital image processing”. International Journal of Photoenergy, Vol. 2012: Article ID 396792, pp.1-11, 2012.
doi:10.1155/2012/396792
5. E. Kaplani. “Design and performance considerations in stand-alone PV powered telecommunication
systems”. Journal of Engineering Science and Technology Review, 5:1, 1-6, 2012.
6. L. King, J.A. Kratochvil, M.A. Quintana, et al. Applications for infrared imaging equipment in
photovoltaic cell, module, and system testing”. Photovoltaic Specialists Conference, 2000. Conference Record
of the Twenty-Eighth IEEE, pp. 1487-90, 2000. doi: 10.1109/ PVSC.2000.916175.
7. MASSON G. et al, 2015, Global Market Outlook for Solar Power / 2015 – 2019. SolarPowerEurope Report.
8. Parretta, M. Bombace, G. Graditi, et al. “Optical degradation of long-term, field-aged c-Si
photovoltaic modules”. Solar Energy Materials and Solar Cells, 86: 349-64, 2005.
9. M.A. Quintana, D.L. King, T.J. McMahon et al. “Commonly observed degradation in field-aged
photovoltaic modules”. Photovoltaic Specialists Conference, 2002. Conference Record of the Twenty-Ninth
IEEE, pp. 1436-39, 2002. doi: 10.1109/PVSC.2002.1190879
10. M. Simon, E.L. Meyer. “Detection and analysis of hot-spot formation in solar cells”. Solar Energy
Materials and Solar Cells, 94: 106-13, 2010.
66
ISBN: 978-960-98739-8-7
Acoustic and electrical emissions from sandstone under uniaxial
compression
V. Saltas1*, I. Fitilis2, J. P. Makris2 and F. Vallianatos1
Department of Environmental and Natural Resources Engineering,
School of Applied Sciences, Technological Educational Institute of Crete, Greece,
* [email protected]
2
Department of Electronics Engineering, School of Applied Sciences,
Technological Educational Institute of Crete, Greece
1
Keywords: acoustic emissions, pressure stimulated currents, compression, fracture,
sandstone.
Abstract
In the present work, pressure stimulated currents (PSCs) were recorded simultaneously with the
acoustic emission (AE) activity from dry, water- and brine-saturated sandstone specimens of
high porosity, during their uniaxial mechanical loading, up to the ultimate failure. The correlation of the AE episodes with the transient electric signals emitted prior and during rock failure
may give insights to precursory phenomena observed in the meso- and macro-scale, such as the
mechanical damage of materials and the occurrence of pre-seismic electric signals.
Introduction
The application of mechanical stress in rocks and brittle materials induces various fractoemission phenomena such as the emission of transient weak electric currents and elastic waves in the
kHz-MHz range (Cress et al., 1987, Frid et al., 2003, Mori et al., 2009, Vallianatos et al., 2004).
The generation of the weak electric currents (known as pressure stimulated currents, PSCs) has
been observed in many laboratory experiments of rocks fracturing and the underlying physical phenomena has been suggested to be strongly related to the seismic electric signals (SES)
recorded before the occurrence of earthquakes (Stavrakas et al., 2004, Varotsos, 2005, Triantis
et al., 2012)).
Various mechanisms for the PSCs have been proposed involving the piezoelectric effect, the
electrokinetic phenomena in saturated rocks and the motion of charged edge dislocations (Freund, 2002, Yoshida et al., 1998, Vallianatos and Tzanis, 1998). In any case, the emission of weak
currents during mechanical stress of rocks should be correlated with the mechanical behavior
of the materials prior to their failure. In this context, the AE technique, which gives valuable
information about the generation and propagation of micro-cracks inside the material subject to
mechanical stress, may serve as an additional tool for the investigation of the associative emission of the transient currents (Lavrov, 2005).
So, in the present work we attempt to correlate the AE features with the recordings of PSCs
by conducting simultaneous measurements in sandstone samples under uniaxial loading. The
high porosity of the specimens allows us to investigate the influence of the water content and
the kind of the pore fluid (water or brine) to both, AE activity and PSC emission. Experiments
with a numerous set of AE sensors and current electrodes deployed at different directions in the
sandstone specimen were carried out in order to gather information on the spatial distribution
of the generated cracks and the accompanied electric emissions.
ISBN: 978-960-98739-8-7
67
Experimental part
Sandstone specimens of high porosity and prismatic shape (approximately 7cm x 7cm x 3.5cm)
were used for the uniaxial loading tests. Different tests of linear loading with a rate of 0.3kN/s
were carried out in dried (at 105 oC), water- and brine-saturated specimens. In order to measure
the weak electric currents at different directions in the specimens, perpendicular to the direction
of the applied stress, copper electrodes were mounted on different sides of the specimen (Fig.
1a). A conductive silver paste ensured good electric contact of the electrodes with the surface
of the specimen. A high sensitivity electrometer (Keithley 6517A) with a multichannel scanner
card was used to acquire the weak electrical currents.
A set of six miniature piezoelectric sensors (PICO sensors, 200kHz-1MHz, MISTRAS Group,
SA) placed at selected positions of specimen’s surface was used to record the AE activity during
the uniaxial loading of the specimens (Fig. 1a). Data acquisition and digital signal processing)
was achieved through a dedicated AE system (PCI-2 card based AE system, Physical Acoustics
Corporation) using three 2-channel cards. A pre-amplification of 40dB was used in each channel and the threshold of detection was settled at 40dB in order to eliminate the background
noise. The AE activity is presented in time series of detected signals in each sensor (hits), their
amplitudes (signal peak in dB), signal energy and other AE features.
An ALPHA S-3000 (Form+Test GmbH) loading machine equipped with the Digimaxx-21 digital controller was used for the application of the uniaxial stress (Saltas et al., 2014). The test
apparatus was covered with copper sheets for electromagnetic shielding (Fig. 1b).
The signals from the electrometer channels, the load cell and the strain gauges were collected
with a signal-conditioning module (NI-SCXI-1520 National Instruments) connected to a DAQ
card (PCI-6221 of National Instruments), using a program developed in LabVIEW graphical
programming environment.
Figure 1. (a) Side view of the sandstone specimen with AE sensors, electrodes and strain gauges mounted on it.
(b) The specimen inside the shielded test apparatus.
No transient electric signals were detected from dried sandstone specimens during their uniaxial
loading, except at the final stage of their rupture. This observation clearly indicates the crucial
role of the water present either in bound form or in free-state inside the pores of the sandstone.
In our case, the absence of free or loosely bound water should be an excluding condition that
inhibits the generation of the weak currents putting in a second place the possible piezoelectric
effect in quartz-bearing sandstone.
Figure 2 shows results of the AE activity ( in the frame proposed by Turcotte and Shcherbakov, 2006) combined with the pressure stimulated currents from two representative sandstone
specimens. The first specimen, after drying at 105 oC, remained under room environment for 3
days in order to absorb water vapors from air, while the other one was saturated with water in
a chamber under vacuum.
68
ISBN: 978-960-98739-8-7
Figure 2. (a) The recorded PSCs at two directions (see the inset) during linear loading of sandstone specimen as
a function of time, in conjunction with the lateral strain. The specimen remained in air for 3 days to absorb water
vapors. (b) The distribution of the hit rate and the cumulative hits of the 6 AE sensors during the experiment. (c)
The energy rate recorded from all sensors and the linear load, as a function of time. The corresponding graphs
of the water-saturated sandstone specimen are shown in (d), (e) and (f). In this case, PSCs were recorded at 3
different directions (see the inset of (d)).
In the case of the sandstone specimen with low water content (Fig. 2a-c), considerable acoustic emission activity starts at around 75% of the fracture strength, i.e., when material exits its
elastic range and undergoes the brittle deformation through the generation and propagation
of micro-cracks. This results to the initiation of weak currents (in the range of pA) emitted
unevenly in the bulk, as it is indicated from the different intensities recorded in each current
channel. Different recorded polarities in each channel indicate a preferable current direction
which may be associated with the formation of specific shear planes inside the sandstone specimen before the failure. The coalescence of the generated micro-cracks creates structural paths
and the presence of water inside the pores enhances charge transport through these conduction
paths. A critical concentration of cracks should exist at given water content inside the interconnected pore space, in order to enable the electrical transport of the redistributed charges. This
could explain the delay in emission of considerable PSCs which starts at ~390s with respect to
the preceding acoustic emissions at around 350s.
The observation of the different AE features in conjunction with the recorded PSCs reveals that
the rate of the recorded hits or even better the rate of energy (see Fig. 2a,c) is appropriate to
determine the initiation of the emitted PSCs and their possible maximum values.
The situation is quite similar as concern the AE activity during the linear loading of the watersaturated sandstone specimen (Fig. 2d-f) but the intensity of the PSCs is almost 3 orders of
magnitude greater than that produced from the specimen which remained in air. In this case, the
presence of free water into the pore volume affects considerably the conduction paths, resulting
ISBN: 978-960-98739-8-7
69
to higher values of the emitted currents. In the case of the brine-saturated specimen (not shown
here), the intensity of the PSCs is around one order of magnitude higher than that of the watersaturated specimens due to the excess of ionic charge contained in brine while, the AE activity
exhibits similar characteristics.
Concluding Remarks
In the present work, AE and PSC measurements were carried out simultaneously in sandstone
specimens under uniaxial loading, up to the ultimate failure. Specimens with different water
content or pore fluid (water or brine) were subjected to linear loading and PSCs were recorded
at different directions in conjunction with the AE activity from a set of 6 sensors.
Our observations indicate that the presence of water in bound or free-state inside the pore volume of the specimen plays a key role to the generation and the intensity of the transient weak
electric currents. Preferable directions of emitted currents were observed and should be related
to the shear planes originated from the coalescence of the micro-cracks. The concentration of
micro-cracks should exceed a critical value to give onset to measurable macroscopic electric
current.
Acknowledgements
This research is implemented in the framework of the project entitled “Interdisciplinary MultiScale Research of Earthquake Physics and seismo-tectonics at the front of the Hellenic Arc” of
the Archimedes III Call through the Operational Program “Education and Lifelong Learning”
and is co-financed by the European Union (European Social Fund) and Greek national funds.
References
1. Cress, G.O., Brady, B.T., and Rowell, G.A., 1987, Sources of electromagnetic radiation from fracture of rock
samples in the laboratory. Geophysical Research Letters 14, 331-334.
2. Freund, F., 2002, Charge generation and propagation in igneous rocks. Journal of Geodynamics 33, 543-570.
3. Frid, V., Rabinovitch, A., Bahat, D., 2003, Fracture induced electromagnetic radiation. Journal of Physics D:
Applied Physics 36, 1620-1628.
4. Lavrov, A., 2005, Fracture-induced physical phenomena and memory effects in rocks: a review. Strain 41,
135-149.
5. Mori, Y., Obata, Y., Sikula, J., 2009, Acoustic and electromagnetic emission from crack created in rock sample
under deformation. Journal of Acoustic Emission 27, 157-166.
6. Saltas, V., Fitilis, I., Vallianatos, F., 2014, A combined complex electrical impedance and acoustic emission
study in limestone samples under uniaxial loading. Tectonophysics 637, 198-206.
7. Stavrakas, I., Triantis, D., Agioutantis, Z., Maurigiannakis, S., Saltas, V., Vallianatos, F., Clarke, M., 2004,
Pressure stimulated currents in rocks and their correlation with mechanical properties. Natural Hazards and
Earth System Sciences 4, 563-567.
8. Triantis, D., Vallianatos,F., Stavrakas, I., Hloupis, G., 2012, Relaxation phenomena of electrical signal
emissions from rock following application of abrupt mechanical stress, Annals of Geophysics, 55, 207-212.
9. Turcotte, D.L., Shcherbakov, R., 2006, Can damage mechanics explain temporal scaling laws in brittle fracture
and seismicity?, Pure and Applied Geophysics, 163, 1031-1045
10. Vallianatos, F., Tzanis, A., 1998, Electric current generation associated with the deformation rate of a solid:
preseismic and coseismic signals. Physics and Chemistry of the Earth 23, 933-938.
11. Vallianatos, F., Triantis, D., Tzanis, A., Anastasiadis, C., Stavrakas, I., 2004, Electric earthquake precursors:
from laboratory results to field observations. Physics and Chemistry of the Earth 29, 339-351.
12. Varotsos, P. A., 2005, The Physics of Seismic Electric Signals. TERRAPUB, Tokyo.
13. Yoshida, S., Clint, O.C., Sammonds, P.R., 1998, Electric potential changes prior to shear fracture in dry and
saturated rocks. Geophysical Research Letters 25, 1577-1580.
70
ISBN: 978-960-98739-8-7
The displacement field in a circular disc made of either orthotropic
or transversely isotropic material under parabolic radial pressure
Christos F. Markides *
National Technical University of Athens, School of Applied Mathematical and Physical Sciences, Department of Mechanics, 5 Heroes of Polytechnion Avenue, Theocaris
Bld., Zografou Campus, 157 73 Athens, Greece * [email protected]
Keywords: Brazilian-disc test, anisotropy, orthotropy, transverse isotropy, complex
potentials
Abstract
An analytic solution for the displacement field in circular discs, made of non-isotropic brittle
materials, subjected to diametral compression is introduced. To achieve the solution advantage
is taken of Lekhnitskii’s [1,2] complex potential technique for rectilinear anisotropic materials.
The solution introduced is advantageous compared to existing ones since the disc is loaded by a
parabolic (instead of uniform) distribution of radial stresses, an assumption closer to reality.
Introduction
The diametral compression of disc-shaped specimens (known as the Brazilian-disc test) was
introduced as a substitute of the tensile test for materials with increased brittleness, as it is for
example rocks and various cement-based products. Its main advantage is that the specimens are
very easy to be prepared while the experimental procedure is relatively simple. The test is now
adays widely used and its implementation is standardized by both the ASTM and ISRM. The
analytic determination of the stress field developed concerns engineering community almost
immediately after the test was proposed. Today both standards adopt Hondros [3] solution, who
considered an elastic, isotropic disc, compressed diametrally by a pair of uniformly distributed
radial stresses along two ‘small’ arcs of its periphery (symmetric with respect to the disc’s center
and of constant length). Papers relieved from the isotropy assumption are very rare although quite
a few rock like materials, widely used is praxis (like for example Dionysos marble [4, 5]), are of
anisotropic nature. Moreover the distribution of pressure along the loaded rims is cyclic rather
than uniform [6] while the length of these rims is a function of the pressure induced. In this direction an attempt is here described to revisit the problem and cure some of these weaknesses.
Mathematical formulation
The material of the disc is assumed as linearly elastic, homogeneous and either orthotropic or
transversely isotropic. For the latter its planes of isotropy are situated parallel to xz-plane (Fig.
1a) while for the orthotropic disc, its three planes of elastic symmetry coincide with the three
planes formed by the axes of the Cartesian reference (Fig.1b). Clearly, in both cases xy-plane
is a plane of elastic symmetry, thus sufficing pure plane strain conditions (assuming the width
of the disc is well comparable to its radius). The externally applied load acts in the plane of
the cross section of the disc remaining constant all along the generators (the width) of the disc.
The problem is considered as a first fundamental plane strain problem. For the solution, Lekhnitskii’s formalism is employed [1]. For the description of the boundary values of stresses along
the disc’s periphery as well as for the definition of the length of the loaded rim, formulae of
the respective isotropic disc-ISRM’s jaw contact problem [6] are employed with some trivial
modifications adjusting to the anisotropic nature of the disc’s material.
ISBN: 978-960-98739-8-7
71
The disc is considered in the z=x+iy=reiθ complex plane and its center is the origin of the Cartesian reference. R and w denote the disc’s radius and width, respectively. L denotes its outer pe
riphery. The externally applied load, Pframe, forms an arbitrary angle φo with respect to x-axis.
Pframe
Pframe
σr = – P(θ)
σr = – P(θ)
2ωο
y
y
x
�ο
θ
w
z=re
iθ
θ
R
L
z=re
L
r O
y Ε΄, ν΄ ,G΄
iθ
R
r
z
2ωο
x
2ωο
�ο
w
O
z
2ωο
x
z
x
νzx
z
Gyz
Ε, ν, G
Gxy
Εz
z
z
x
Εx
y Εy
y
x
νyx
νyz y
Gxz
Figure 1. The configuration of the problem and definition of symbols
Pframe is parabolically distributed along two finite parts of L (each one of length 2Rωo) according
to a law which is statically equivalent to the actual cyclic one. Concerning ωo, it is not arbitrary
ly prescribed, but it is obtained from a related contact problem [6]. Even more it is expressed as
a function of φo underlying its dependence on the anisotropic nature of the disc’s material.
Following Lekhnitskii’s formalism and considering plane strain conditions, it holds that
u=u(x,y), v=v(x,y) and the set of equations characterizing elastic equilibrium (in the absence
of body forces) are recapitulated as:
∂τ xy ∂σ y
∂σ x ∂τ xy
+
= 0,
+
= 0 ∂x
∂y
∂x
∂y
(1)
∂u
∂v
∂u ∂v
= β11σ x + β12σ y , ε y =
= β12σ x + β22σ y , γ xy =
+
= β66 τ xy ∂x
∂y
∂y ∂x
(2)
∂2
∂2
∂2
β
σ
+
β
σ
+
β
σ
+
β
σ
−
β66 τ xy = 0 (
)
(
)
11 x
12 y
12 x
22 y
∂y 2
∂x 2
∂x∂y
(3)
σ x cos (n, x ) + τ xy cos (n, y ) = X n ,
(4)
εx =
where βij = αij −
α i3α j3
α 33
τ xy cos (n, x ) + σ y cos (n, y ) = Yn (boundary conditions) and aij are the material’s elastic constants [1]. The generalized biharmonic
equation providing the stress function F(x,y) and its characteristic algebraic equation become:
72
ISBN: 978-960-98739-8-7
β22
∂ 4F
∂ 4F
∂ 4F
2
+
β
+
β
+
β
= 0 , β11µ 4 + (2β12 + β66 )µ 2 + β22 = 0 (
)
12
66
11
∂x 4
∂x 2∂y 2
∂y 4
(5)
Its roots (the complex parameters) are pure imaginary quantities, conjugate in pairs, reading
as:
µ1 = iβ1 , µ 2 = iβ2 ,
µ3 = µ1 = − iβ1 , µ 4 = µ 2 = − iβ2 ,
− (2β12 + β66 ) 
β1 
=
β2 
(2β12 + β66 )
2
− 4β11β22
2β11
∈�
(6)
Over bar denotes conjugate complex value. Since it is assumed that μ1≠μ2 F(x,y) is expressed
with the aid of two analytic functions of the complicated complex variables z1,2=x+μ1,2y as:
F (x, y ) = 2ℜ  F11 (z1 ) + F12 (z 2 ) (ℜ denotes the real part) (7)
Introducing Lekhnitskii’s complex potentials, Φ1 (z1 ) := F11′ (z1 ), Φ 2 (z 2 ) := F12′ (z 2 ) (prime de
notes first derivative) stresses and displacements (omitting rigid body motion) are expressed
as:
σ x = 2ℜ µ12Φ1′ (z1 ) + µ 22Φ 2′ (z 2 ) , σ y = 2ℜ Φ1′ (z1 ) + Φ 2′ (z 2 ) , τxy = −2ℜ µ1Φ1′ (z1 ) + µ 2Φ 2′ (z 2 ) (8)






u = 2ℜ  p1Φ1 (z1 ) + p 2Φ 2 (z 2 ) , v = 2ℜ q1Φ1 (z1 ) + q 2Φ 2 (z 2 ) (9)
β22
β
, q 2 = β12µ 2 + 22 µ1
µ2
The boundary conditions on L are written in terms of Φ1 and Φ2 as:
p1 = β11µ12 + β12 , p 2 = β11µ 22 + β12 , q1 = β12µ1 +
S
S
0
0
2ℜ Φ1 (z1 ) + Φ 2 (z 2 ) = − ∫ Yn dS, 2ℜ µ1Φ1 (z1 ) + µ 2Φ 2 (z 2 ) = ∫ X n dS (10)
(11)
According to Lekhnitskii [2] the general solution of the problem is sought in the form:
∞
∞
m=2
m=2
Φ1 (z1 ) = A 0 + A1z1 + ∑ A m P1m (z1 ), Φ 2 (z 2 ) = B0 + B1z 2 + ∑ Bm P2m (z 2 ) (12)
−1
m
m

2
2
2 
2
2
2  

P1m (z1 ) = m
z1 + z1 − R (1 + µ1 ) + z1 − z1 − R (1 + µ1 ) 
m 


 
R (1 − iµ1 )  
−1
m
m

2
2
2 
2
2
2  

P2m (z 2 ) = m
z
+
z
−
R
1
+
µ
+
z
−
z
−
R
1
+
µ
( 2 )  2 2 ( 2 ) 
2
2
m 
R (1 − iµ 2 )  

(13)
The coefficients Aj, Bj of Φ1, Φ2 of Eqs.(12) will be obtained by the fulfillment of the boundary
conditions. In this direction, the right-hand sides of Eqs.(11) are expanded in Fourier series:
S∈ �
−
∫
0
∞
Yn dS = a 0 + ∑ (a ms + a ms
m
m =1
−m
),
S∈ �
∫
0
∞
(
)
X n dS = b 0 + ∑ b ms m + bms − m m =1
(14)
am and bm are, in general, complex constants and s=eiθ. When z is located on L it holds that:
z1 =
R
R
1
(1 − iµ1 )s + (1 + iµ1 ) ,
2
2
s
ISBN: 978-960-98739-8-7
z2 =
R
R
1
(1 − iµ 2 )s + (1 + iµ 2 ) 2
2
s
(15)
73
P1m (s ) = − s m − t1m s − m , P2m (s ) = − s m − t 2m s − m , t1 = (1 + iµ1 ) (1 − iµ1 ), t 2 = (1 + iµ 2 ) (1 − iµ 2 ) (16)
From the above equations one obtains the coefficients of Φ1, Φ2 solving the system:
A m + Bm + A m t1m + Bm t2m = − a m
A mµ1 + Bmµ 2 + A mµ1 t1m + Bmµ 2 t2m = − b m
A m + Bm + A m t1m + Bm t 2m = − a m
A mµ1 + Bmµ 2 + A mµ1t1m + Bmµ 2 t 2m = − bm ,
(m = 2,3,...) (17)
Constants am, bm are obtained by considering that Xn=σrcosθ, Υn=σrsinθ and taking advantage of
the law for the distribution of pressure [6] which for the needs of this study is written as [7]:
P 
2ω − sin 2ωo 
e −i2 φo 
sin 2ωo cos 2ωo   i2 θ
σr = − c 2ωo − o 2
+ sin 2ωo e −i2 φo −
 sin 2ωo − ωo −
 e
2
π
2sin ωo
2sin ωo 
2



ei2 φo 
sin 2ωo cos 2ωo   −i2 θ ∞   sin kωo
+ sin 2ωo ei2 φo −
ω
−
ω
−
−
sin
2
o
o

 e + ∑ 
2
ω
2sin
2
k



=
k
3

o


1
 sin kωo k cos 2ωo sin kωo − 2sin 2ωo cos kωo   −ikφo −ik (φo + π)  ikθ  sin kωo
+e
+

  e
2
 e +  k −
4 − k2
2sin ωo  k


1
 sin kωo k cos 2ωo sin kωo − 2sin 2ωo cos kωo    ikφo
ik (φo + π )
 e − ikθ 
+


  e + e
2
2

2sin ωo  k
4−k


(18)
Given the complex potentials stresses and displacements can be obtained from Eqs.(8,9) by
fullfilling the condition that rigid body translation nd rotation must be zero. The displacement
field components, of interest in the present study, are written as:
u (x, y ) = 2ℜ (p1A 0 + p1A1z1 + p 2 B0 + p 2 B1z 2 ) +
 


  2ℜ p1  A 3P13 (z1 ) + ∑ A m P1m (z1 ) + p 2  B3P23 (z 2 ) + ∑ Bm P2m (z 2 ) 
m = 5,7,9,...
m = 5,7,9,...
 


 
(19)
v (x, y ) = 2ℜ (q1A 0 + q1A1z1 + q 2 B0 + q 2 B1z 2 ) +
 


  2ℜ q1  A 3P13 (z1 ) + ∑ A m P1m (z1 ) + q 2  B3P23 (z 2 ) + ∑ Bm P2m (z 2 ) 
m = 5,7,9,...
m = 5,7,9,...
 


 
(20)
Results and conclusions
Taking advantage of Eqs.(20,21), the deformed shape of a transversely isotropic disc with Es= 3.19 GPa, νstrong=0.36, Eweak=0.80 GPa, νweak=0.42, is plotted in Fig.2 for φο=30o, indicating
trong
the capabilities of the solution introduced.
Exploring the expressions for the displacements and the respective ones for the stresses [7] one
must emphasize their compact form. Moreover the fact that these formulae are valid for either orthotropic or transversely isotropic materials must be stressed out. In conjunction, also, to
the fact that the load considered (parabolic pressure) is very close to the actual one renders the
74
ISBN: 978-960-98739-8-7
Figure 2. Deformed transversely isotropic disc
under parabolic radial pressure
present solution an attractive, flexible and reliable tool in hands of scientists and engineers using the Brazilian disc test.
References
1.
2.
3.
4.
5.
6.
Lekhnitskii S.G., Theory of elasticity of an anisotropic body, Mir, Moscow, (1981).
Lekhnitskii S.G., Anisotropic Plates (English translation by Tsai S. W.), Gordon & Breach, New York. 1968
Hondros G., Australian J. of Applied Sciences, 10, 243-268, 1959.
Kourkoulis S.K., Exadaktylos G.E., Vardoulakis I., Int. J. of Fracture, 98(3-4), 369-392, 1999.
Exadaktylos G.E., Vardoulakis I., Kourkoulis S.K., Int. J. Solids & Structures, 38(22-23), 4119-4145, 2001.
Kourkoulis S.K., Markides Ch.F., Chatzistergos P.E., Int. J Rock Mechanics & Mining Sci, 57,132-141,
2012.
7. Markides Ch.F., Kourkoulis S.K., Stresses and displacements in non-isotropic discs, Submitted, 2015.
ISBN: 978-960-98739-8-7
75
Rejoining fragmented elements of stone monuments: Critical
aspects from the experience of the Athens Acropolis restoration
project *
National Technical University of Athens, Department of Mechanics, Laboratory of
Testing and Materials, Theocaris Building, Zografou Campus, 157 73, Greece,
[email protected]
Keywords: Dionysos marble, Monuments, Restoration, Mechanical behavior,
Fracture, Failure
* The paper is devoted to the memory of late Professor Ioannis Vardoulakis (1949-2009) of the
Department of Mechanics of NTUA, who initiated the author into the subject of Rock Mechanics and its application in the restoration of stone monuments
Abstract
The study of marble and marble structural elements at the Laboratory of Testing and Materials
of the National Technical University of Athens is dated back to the early nineties, under the
supervision of late Professor I. Vardoulakis, and is continued uninterruptedly since then. The
experience gathered enlightened critical aspects of the response of fragmented structural ele
ments restored with the aid of metallic reinforcement. These aspects are recapitulated here in an
effort to assist scientists working for the restoration of unique monuments of Cultural Heritage.
Introduction
Rejoining fractured members of the classical monuments on the Acropolis of Athens is nowadays
achieved according to the pioneering technique proposed almost half a century ago by renowned
scientists [1]. The technique is based on the use of titanium reinforcing elements together with
a suitable cementitious material. The co-existence of brittle marble with the extremely ductile
titanium renders the intermediate material (the cementitious one) the most critical link in the
“chain” of the three materials (marble-cement-titanium) complex. On the other hand the existence of two interfaces (marble to cement and cement to titanium) renders the study (either ex
perimental or numerical) of the mechanical response of the complex a very challenging task.
Additional difficulties appear due to the complicated nature of the specific type of marble (i.e.
Dionysos marble), as it is dictated by anisotropy, size effect, non-linearity, bimodularity etc.
Dionysos marble is the material used almost exclusively for the construction of either copies
of missing structural elements or of patches that complete damaged elements, for the needs of
the restoration project in progress of the Acropolis of Athens, since its mechanical and physical
properties are close to the respective ones of Pentelic marble (that was used by ancient Greeks
to build these monuments). According to the results of a long experimental protocol [2] it is an
orthotropic material which can be approximated as a transversely isotropic one. Very carefully
implemented direct tension (with the aid of specially designed grips and specimens, Fig.1a) and
uniaxial compression tests, with specimens cut along the three anisotropic directions, permitted
drawing of the complete axial stress - axial strain curve, which for the strong anisotropy direction
is shown in Fig.1 (upper left). The non-linearity and bimodularity are clearly shown.
During the same series of tests the pronounced size effect (dependence of strength on the size of
the specimens used) characterizing Dionysos marble was revealed. The specific aspect, which
76
ISBN: 978-960-98739-8-7
dictates the use of specimens of dimensions approaching these of the actual structural elements
(rendering experimental exploration of restored members difficult and expensive), can be seen
for both the compressive and tensile strength in Figs.1 (lower left and right, respectively).
Experimental and numerical study of restored epistyles of the Parthenon temple
The technique used to restore fragmented epistyles of the Parthenon temple (Fig.2, left) was ex
perimentally assessed by subjecting an accurate copy of a multi-fragmented epistyle to six-point
bending (Fig.2, right). The results indicated that well before the maximum load expected was
25
Axial Strain [x 10E-6]
-1000
Ec
-500
-25
-50
0
.
-1500
Et
Axial stress [MPa]
-2000
0
500
-75
-100
10
Tensile Strength [MPa]
Compressive strength [MPa]
.
120
90
60
30
0
8
5
3
0
0
100
200
300
Height [mm]
400
25
50
Diameter, D
75
100
Figure 1. The grips and the specimens used for the direct tension tests (upper left). The axial stress - axial
strain curve along the strong anisotropic direction (upper right). The dependence of the compressive and tensile
strength on the size of the specimen (lower left and right, respectively).
Figure 2. In situ restoring of a huge epistyle of the Propyleae of the Acropolis of Athens (left). Laboratory sixpoint bending test of an accurate copy of a multi-fragmented epistyle of the Parthenon temple (right).
ISBN: 978-960-98739-8-7
77
25000
8
7
6
5
15000
4
10000
3
slip, mm
P(N)
20000
2
5000
1
0
0
0
2
4
6
P-δ
slip-δ
8
Syst.Displ., mm
Figure 3. A typical ‘pull-out’ test (left). The load and the relative slip (pink line) vs. displacement (right)
reached a ‘plateau’ appeared in the load-deflection curve, while upon unloading considerable
permanent deflection characterized the response of the epistyle. Thorough investigation indicated
that at the load of the ‘plateau’ relative slip of the metallic reinforcing bars with respect to the
marble appeared. It became thus necessary to study the ‘pull-out’ phenomenon and the role of
the intermediate layer of cementitious material on the overall behavior of the restored epistyle.
In this direction long series of standardized pull-out tests were carried out (Fig.3) and it was
concluded that the performance of the cement-marble interface was very poor, concerning the
quality of adhesion. Therefore a new project was launched for the development of improved ce
mentitious materials together with an optimized design of the bolts of the reinforcing bars. The
results of this study were then used for the construction of a small number of copies of Parthe
non’s epistyles, consisting of two identical marble pieces restored with a number of bolted
titanium bars which were subjected to either multi-point bending (Fig.4, left), or bending under
dead-loads (additional marble blocks) simulating bending under uniform load (Fig.4, right).
In parallel a numerical model was prepared with the aid of the Finite Element Method (Fig.5)
which was calibrated and validated using data of the bending tests. The model simulated in
realistic manner the boundary conditions (it was considered resting on marble abacuses), all
contact surfaces (marble to marble, marble to cement and cement to titanium) while the loading
considered simulated both the laboratory conditions (multi-point bending) and also the actual
ones (loading by the superimposed structural elements).
The numerical analysis revealed that the axial strains are not distributed linearly along the
height of the epistyle. They follow a sigmoid distribution which is not in accordance to the assumptions of the Bernoulli-Euler technical bending theory. This was to be expected since the
epistyles do not fulfill the requirement of ‘long’ beams. In addition, the neutral axis was found
to be translated towards the bottom side of the epistyle. As a result the theoretical calculations
used to determine the number of titanium bars are rough approximations of the real conditions.
The distribution of equivalent stress along the threads of the reinforcing bars (Fig.5, lower left)
is astonishing: Less than one tenth of the total number of screws carries the load induced on
each bar. It is thus concluded that the empirical formulae used to calculate the anchoring length
of the bars seriously overestimates the actual loading conditions leading to drilling longer holes
78
ISBN: 978-960-98739-8-7
Figure 4. Typical eight-point bending (left) and bending under dead-weight loads (right).
1.00
Root of the thead
Pick of the thread
15
Reduced equivalent stress
Equivalent stress (MPa)
20
10
5
0
0
50
100
150
Number of threads
200
Case a
Case b
Case c
Case d
0.75
0.50
0.25
0.00
0.00
0.25
0.50
x/L
0.75
1.00
Figure 5. The numerical model (upper). The stress level along the reinforcing bar (lower left). The role of the
loading type (a: uniform load all along the epistyle, b: uniform load along the span, c: 8-point bending along the
span, d: 8- point bending all along the epistyle)
ISBN: 978-960-98739-8-7
79
(than the ones required) on the authentic marble elements. This is against the basic principle of
Venice charter which dictates the minimum possible destruction, and should be reconsidered.
The stress level along the length of the epistyle strongly depends on the actual way the load is
applied (Fig.5, lower right): Reproducing actual bending in the laboratory does not give reliable
results. Simulating a uniform load by a multi-point bending test does not lead to comparable
results. Thus the engineer who designs based on an experimentally evaluated and calibrated
model should increase the safety factors of the study introducing a ‘load-correction’ factor that
could enable direct comparison between actual loading states of the epistyle and the results of
the laboratory tests. For the particular case this load correction factor was estimated to about
2.
References
1. Korres M., Toganides N., Zambas C., Skoulikidis Th., Study for the restoration of the Parthenon, Vol.2a,
Athens: Ministry of Culture, Committee for the Preservation of the Acropolis Monuments (in Greek) 1989.
2. Vardoulakis I., Kourkoulis S.K., Exadaktylos G.E., Rosakis A. Mechanical properties and compatibility of
natural building stones of ancient monuments: Dionysos marble. In ‘‘The building stone in monuments”, M.
Varti-Mataranga, Y. Katsikis (eds.), IGME Publishing, Athens 2002.
80
ISBN: 978-960-98739-8-7
An experimental study of the pull-out mechanism of threaded
metallic bars embedded in marble blocks
I. Dakanali1, I. Stavrakas2, D. Triantis2, E.D.Pasiou1, S.K.Kourkoulis1*
1
2
Laboratory of Testing and Materials, Department of Mechanics, School of Applied
Sciences, National Technical University of Athens, * [email protected]
Laboratory of Electronic Devices and Materials, Department of Electronic
Engineering, Technological Educational Institute of Athens, Greece
Keywords: Pullout tests, innovative techniques, Acoustic Emissions, Pressure
Stimulated Currents, marble, cement paste, titanium bar
Abstract
An experimental study is described aiming to assess the bonding quality between the constituent
elements of the marble - cement paste - titanium complex, which appears when marble structur
al members of classic monument are restored. The novelty of the study is the combined and
comparative use of traditional and innovative sensing techniques in an effort to pump data from
the interior of the specimens. The target of this research project is the exploration of the failure
mechanisms leading to debonding and finally to the catastrophic pull-out of the reinforcing
metallic element from the body of the three-materials-complex.
Introduction
Re-joining fractured or fragmented marble epistyles of the Acropolis monuments is nowadays
realized according to a pioneering technique introduced by renowned scientists [1] almost half
a century ago. According to this technique, threaded titanium bars are inserted into holes predrilled in the body of the elements to be restored and filled with a proper cementitious material,
which acts as the matching element between the brittle marble and the ductile titanium.
Previous experimental studies have revealed that the weak link of the three- materials ‘chain’ is
the intermediate layer of cementitious material and more specifically the marble-cement interface [2], especially in case the restored structural element is loaded by tensile or bending loads.
Indeed, while the titanium-cement interface appears to be strong enough (as expected due to
the bolted surface of the bar) and the bar-cement system behaves as a single body, the cementmarble interface is rather weak [3]. As a result the bar-cement block slips as a rigid body with
respect to the marble volume (the familiar pull-out phenomenon) (Fig.1) and the restored member fails although neither the marble nor the bars have reached their critical limits.
In this direction an effort is undertaken to experimentally explore the pull-out phenomenon
using innovative sensing methods as it is Acoustic Emission (AE) and Pressure Stimulated
Currents (PSC). The latter was chosen since strong indications exist that the underlying phenomenon can be considered as a pre-failure indicator in several materials [4-6]. In addition,
traditional sensing methods were also used (electrical strain gauges, LVDTs, dial gauges and
clip-gauges) for comparison reasons and calibration of the innovative techniques.
Experimental set up
The specimens tested were prepared using cubes cut from Dionysos marble blocks. A central
through hole of diameter equal to dhole=14 mm was drilled normally to the upper face of the
cubes and then it was filled with liquid cementitious material. A threaded titanium bar of outer
diameter equal to d=10.5 mm was driven in the hole and it was kept normal to the upper face
of the cube until setting of the cement using a proper device. The dimensions of the cubes used
between broad limits. The ones finally chosen were equal to 20x20x20cm3 as an optimum
ISBN: 978-960-98739-8-7
81
Figure 1. Partly pulled-out reinforcing bar. The failure of the marble-cement interface is clearly visible (Left).
The electrical contacts for recording the PSCs (right).
compromise between economy and the need to expel parasitic phenomena around the critical
region.
For the PSCs recording, a gilded silver needle was embedded in the cement paste before the total
curing of paste. The embedded length of the needle was approximately 1cm. Close to the needle
a hole of 1cm depth was drilled in the marble body and a second electrical contact was inserted
(Figure 1, right) permitting recording of PSCs produced on the marble-paste interface.
The specimens were cured for 28-days. Then the titanium bar was gripped by the frame’s upper
jaw. The marble cube was constrained by a rigid metallic plate with a hole of diameter equal to
100 mm in its center. The plate was supported by 4 threaded bars (Fig. 3). The tests were implemented using a stiff servo-hydraulic INSTRON loading frame of capacity 250 kN. They were
quasi-static under a displacement-control mode at a rate equal to 0.2 cm/min.
During the tests the following quantities were measured and recorded: The load (as obtained
from the calibrated cell of the loading frame), the displacement of the frames traverse, the strain
of the titanium bar (using an Instron-Dynamic Extensometer (2620-602) of 12.5mm gauge
length and a +/-2.5mm travel), the relative displacement of the bar with respect to the cube (using a calibrated LVDT in touch with the bars lowest end through the cube’s bottom face), the
PSC (with the aid of an extremely sensitive KEITHLEY electrometer) and finally the acoustic
events produced (using a system of 6 acoustic sensors (Fig. 2), properly attached on the marble
cube, as close as possible to the area where failure is expected).
Figure 2. The experimental set-up (left) and the system of sensors used to detect acoustic events.
82
ISBN: 978-960-98739-8-7
Depending on the anchoring length and the composition of the cement paste different failure
modes (pull-out, fracture of marble, fracture of the titanium bar and various combinations) were
observed (Fig.3). Taking into account the aim of the study, attention is focused to the tests in
which the marble cube and titanium bar remained intact while the bar was pulled out. In these
tests a small marble cone is systematically fractured remaining in contact with the slipping
bar.
Figure 3. Various failure modes observed during the experimental protocol
A typical load - time curve (which is “equivalent” to a load - displacement curve since the tests
are implemented under displacement control mode) is plotted in Fig.4. In the same figure the
indications of the LVDT used to measure the relative displacement of the bar with respect to the
marble cube are plotted together with the recording os the electrometer for the PSC. It is interesting to observe that the time variation of the LVDT’s indications (red line in Fig.4) consists
of three distinct portions: An initial almost perfectly linear one (up to about 750 sec) follοwed
by a second linear one (up to about 1200 sec). It is assumed that these two regions correspond
to elastic deformations of the supporting system and of the specimen as a stucture rather than
to relative motion of the bar with respect to the marble cube. Then, slightly before the ultimate
load is attained, the indications of the LVDT start increasing rapidly. Clearly this portion of the
graph represents relative motion designating the onset of pull-out. The observation that relative
slip starts slightly before the load’s maximum value supports previous similar conclusions by
Marinelli et al. for reinforcing bars made of steel instead of titanium [7].
Concerning the values of PSC, two peaks of long duration are usually observed: An early one
(usually of relatively low amplitude) and a second one (in most cases of much higher amplitude). A definite explanation of this behavior is not yet available (the specific technique is applied for the first time in pull-out tests) however it could be attributed to changes of the mechanical response of the mechanical system. In any case the time instant at which these two peaks
are observed are closely related to the slope changes of the indications of the LVDT, mentioned
previously, and also are concurrent with increased activity as it is detected by the AE system,
shown in Fig.4 (right), in which the cumulative energy of the acoustic hits recorded by all sensors is plotted vs. time in conjunction to the respective variation of the mechanical load.
ISBN: 978-960-98739-8-7
83
Concluding remarks
Although the research project is at the first steps of its implementation, some qualitative conclusions can be already drawn. Indeed the data gathered from the electrometer (PSC) and
Figure 4. Typical plots of load, PSC and LVDT’s recordings versus time (left). Cumulative energy of hits recorded by all acoustic emissions sensors versus time (right).
the respective ones gathered from the acoustic emission sensors are in very good agreement
with the respective ones provided by traditional sensing techniques. It is thus indicated that in
case these data could be definitely validated and calibrated against data of traditional sensing
techniques (permitting thus quantitative processing) they could provide valuable data about
what (and when) happens at the interior of the specimen during loading. It is also mentioned
that data from another project in progress, dealing with structural tests (bending and shear of
structural members joined together by means of metallic bars and suitable cementitious materials) support the conclusions of the present study [8,9].
Acknowledgements
This research has been co-financed by the European Union (European Social Fund - ESF) and
the National Strategic Reference Framework (NSRF) - Research Funding Program: THALES.
References
1. Korres M., Toganides N., Zambas C., Skoulikidis Th., Study for the restoration of the Parthenon, Vol.2a,
Athens: Ministry of Culture, Committee for the Preservation of the Acropolis Monuments (in Greek) 1989.
2. Kourkoulis S.K., Mentzini M., Ganniari-Papageorgiou E., Restored marble epistyles under bending: A
combined experimental and numerical study. Int. J. Architectural Heritage, 7(1), 89-115, 2013.
3. Kourkoulis S.K, Papanicolopulos S.-A., Marinelli A., Vayas I., Restoration of antique temples: Experimental
investigations on the pull-out behaviour of anchors in marble, Bautechnik, 85(2), pp. 109-119, 2008.
4. Triantis D., Stavrakas I., Kyriazopoulos A., Hloupis G., Agioutantis Z., Pressure Stimulated Electrical Emissions
from cement mortar used as failure predictors, International Journal of Fracture, 175, 53-61, 2012.
5. Kyriazopoulos A., Stavrakas I., Anastasiadis C., Triantis D., Study of weak electric current emission on cement
mortar under uniaxial compressional mechanical stress up to the vicinity of fracture, Journal of Mechanical
Engineering, 57, 235-242, 2011.
6. Triantis D., Stavrakas I., Anastasiadis C., Kyriazopoulos A., Vallianatos F., An analysis of pressure stimulated
currents in marble samples under mechanical stress, Physics & Chemistry of the Earth, 31, 234-239, 2006.
7. Marinelli A., Papanicolopulos S.-A., Kourkoulis S. K., Vayas I., The pull-out problem in restoring marble
fragments: Α design criterion based on experimental results, Strain, 45(5), pp. 433–444, 2009.
84
ISBN: 978-960-98739-8-7
8. Kourkoulis S.K., Pasiou E. D., Triantis D., Stavrakas I., Hloupis G., Innovative experimental techniques in
the service of restoration of stone monuments - Part I: The experimental set up, Procedia Engineering, 109,
pp. 268-275, 2015.
9. Triantis D., Stavrakas I., Pasiou E.D., Hloupis G., Kourkoulis S.K., Innovative experimental techniques in the
service of restoration of stone monuments - Part II: Marble epistyles under shear, Procedia Engineering, 109,
pp. 276-284, 2015.
ISBN: 978-960-98739-8-7
85
Electrical and Acoustic Emissions during Three Point Bending
Tests of Pre-Notched Marble Specimens
Ermioni D. Pasiou1, Ilias Stavrakas2, George Hloupis2,
Stavros K. Kourkoulis1* and Dimos Triantis2
National Technical University of Athens, School of Applied Mathematical and
Physical Sciences, Department of Mechanics, 5 Heroes of Polytechnion Avenue,
Theocaris Bld., Zografou Campus, 157 73 Athens, Greece * [email protected]
2
Technological Educational Institute of Athens, Faculty of Technological Applications,
Department of Electronics, Agiou Spyridonos, 122 10 Aigaleo, Greece
1
Keywords: pressure stimulated currents, acoustic emissions, DIC, Dionysos marble,
3PB.
Abstract
Notched specimens made of Dionysos marble are subjected to three point bending and the
internal mechanisms activated before and during the fracture are studied with the aid the Pressure Stimulated Currents (PSC)- and Acoustic Emissions (AE)-techniques. In addition, Digital
Image Correlation (DIC) and traditional clip-gauge extensometers were used. The experimental
results indicate a strong correlation between the PSC and the acoustic events throughout the
whole loading process while a sudden drop of the PSC-value well before any macroscopically
visible event is observed indicating the onset of failure.
Introduction
A series of experimental techniques for measuring strains and displacements (e.g strain gauges,
dial gauges, clip-gauges, extensometers, photoelasticity, Moiré patterns, caustics etc) are usually
used in the field of Experimental Mechanics. The main limitation of these traditional techniques is
that the data obtained are exclusively related to the external surface of the specimen. Therefore
they detect failure only during its last stage since it is well known that damage mechanisms are
firstly activated in the interior of the specimens. In this context, experimental techniques which
can provide data during the whole loading process are widely used nowadays, for example the
Acoustic Emission (AE)- and the Pressure Stimulated Currents (PSC)-techniques.
The need to estimate the remaining life of a structural member of a monument which is usually
seriously damaged or cracked was the motive of the present study. Therefore, pre-notched marble specimens were subjected to three point bending tests. Both AE and PSC techniques were
used, together with a 3D Digital Image Correlation (DIC) system and traditional clip gauges,
in an effort to correlate the data from the interior of the specimens with the ones obtained from
its exterior. The data drawn will hopefully permit the quantification of a critical parameter (the
critical Notch Mouth Opening Displacement, NMOD) which could play the role of an-easy-touse fracture criterion in case of pre-notched structures.
The material and the specimens
The specimens were made of Dionysos marble, i.e. the material used in the restoration project
of the Acropolis monuments. Its mechanical and physicochemical properties are very similar to
those of the original building stone, i.e. Pentelic marble. Dionysos marble is an extremely finegraded white marble quarried from Mount Dionysos in Attica. It is mainly composed by calcite
(98%) and very small amounts of muscovite, sericite, quartz and chlorite. Its specific
86
ISBN: 978-960-98739-8-7
and apparent densities are equal to 2730 kg/m3 and 2717 kg/m3, respectively. Its extremely low
porosity varies between 0.3% (virgin state) and 0.7% (superficial porosity) [1].
Dionysos marble is an orthotropic material, however it could be characterized as transversely
isotropic since the mechanical properties of the two out of three anisotropic directions are very
close to each other. Dionysos marble was also proved to be slightly non-linear both in the tension and in the compression regime and slightly bimodular (the modulus of elasticity under
compression exceeds that under tension by about 15%) [2-4].
The specimens’ dimensions were 20x20x100 mm3. A notch was mechanically formed at their
mid-span with a ratio of notch’s length-over-specimen’s height equal to 0.20. The load was applied normal to the material layers to avoid scattering due to the transverse isotropy. The speci
mens were placed on two metallic cylinders of radius 15 mm and the load was linearly applied
along the width of the specimens with the aid of a third cylinder of radius 10 mm (Fig.1a).
Experimental results
All specimens were fractured with a crack almost parallel to the load (Fig.1b). Any deviations
(Fig.1c) could be attributed to inevitable local inclinations of the material layers of marble with
respect to the specimens’ longitudinal axis. The fracture stress σf is calculated by the familiar
formula σf=3PL/2bh2 of the classic Bernoulli-Euler technical bending theory, where P is the
load applied, L the specimen’s span and b, h the width and the height of the effective critical
section (i.e. the one without the notch). The stress value obtained was equal to about 14.0 MPa
with very low standard deviation (about 5%). The maximum tensile bending stress versus the
respective deflection of the central cross section is plotted in Fig.2a. Ignoring bedding errors, all
curves are more or less similar to each other with a slope change at a stress level equal to about
60% of the ultimate one. The variation of the Notch Mouth Opening Displacement (NMOD)
obtained by the clip gauge normalized over the initial notch breadth, δo, with respect to the
bending stress (up to the maximum stress level) is plotted in Fig.2b. It is observed that after a
more or less linear portion the curves become strongly non-linear, contrary to what is perhaps
expected for an extremely brittle material like Dionysos marble.
Concerning the PSC, all specimens exhibited similar qualitative behaviour. The variation of
both PSC and load are plotted in Fig.3a for a typical specimen. It is clearly seen that PSC is
almost constant (with negligible changes) during the initial load steps (about 0.1 pA), while at
a load level equal to about 60% of the ultimate stress the PSC starts increasing abruptly. There
fore, a clear correlation exists between the PSC and the characteristic point of the stress- deflection curves where the slope change is observed, indicating the onset of intense micro- cracking
within the volume of the specimen [5-7]. In addition, an abrupt discontinuity of the PSC-time
curve is observed a few seconds before the final microscopic fracture of the specimens (see for
example point S in Fig.4), when the load is almost 95% of the ultimate load.
Similar qualitative results were obtained by the AE technique. The temporal variation of the
“cumulative hits per unit of time”, dN/dt, in juxtaposition to the respective variation of the load
applied is plotted in Fig.3b for a typical specimen. The increase of the quantity dN/dt is very
smooth till the point where the load reaches about 60% of its ultimate value and it gradually
becomes steeper especially while approaching to the final fracture of the specimens.
The NMOD was also determined based on the data from the DIC technique by isolating two
small areas on both sides of the notch at the specimen’s lowest level (after removing any rigid body displacements). The results are in very good agreement with the clip gauge’s values
(Fig.4).
ISBN: 978-960-98739-8-7
87
Discussion and Conclusions
In an effort to understand the mechanisms activated within the volume of the specimen leading finally to macroscopic fracture the PSC, the cumulative hits per unit of time dN/dt and the
NMOD are plotted in Fig.4 versus the maximum stress developed. Despite the completely different nature of the techniques, the three quantities behave similarly. During the first loading
stages the PSC- value exhibits small changes while both dN/dt and NMOD increase smoothly
and almost linearly. At a stress level equal to about 60% of the specimen’s ultimate value (where
the slope change of the stress-deflection curve was detected) the slope of the aforementioned
curves increases faster becoming almost vertical just before fracture. The narrow shaded elliptic area shown in Fig.4 encompasses all three phenomena: (i) The deviation of the PSC-values
from their constant value, represented by line aa΄ (Point Ε). (ii) The deviation of dN/dt curve
from its linear portion, represented by line bb΄ (point A). (iii) The deviation of NMOD curve
from its linear portion, represented by line cc΄ (Point N). Therefore, the mechanical response
of Dionysos marble is strongly related to the respective electric and acoustic processes taking
place within the specimen’s volume. Dionysos marble experiences irreversible failure processes already from the two thirds of the fracture stress for this type of tests since the increase of the
PSC and of the acoustic activity (dN/dt) is directly related to the generation of micro-cracks.
Figure 1. Typical specimen (a) before loaded, (b,c) just after fracture.
Figure 2. (a) Typical stress-deflection curves. (b) The longitudinal nominal stress against the NMOD/δο.
88
ISBN: 978-960-98739-8-7
Figure 3. (a) Typical PSC and load and (b) typical cumulative hits per second and load curves.
Figure 4. The variation of all three quantities (mechanical, electrical and acoustic) versus the maximum stress
developed (according to the Bernoulli –Euler technical theory)
Concerning the PSC, an abrupt drop (point S, in Fig.4) was observed at a stress level equal to
about 95% of the fracture stress indicating probably the generation of fatal crack all along the
active cross section of the specimen. From this point on and despite the fact that the load level
will increase slightly (perhaps due to inertia effects) the specimen should be considered fractured. This sudden drop of the PSC values could be used for the determination of the critical
NMOD.
Concluding it can be said that taking advantage of innovative experimental techniques interesting data can be pumped from the interior of loaded elements uniquely related to failure mechanisms activated well before any macroscopic fracture is observed.
ISBN: 978-960-98739-8-7
89
Acknowledgements
This research has been co-financed by the European Union (European Social Fund-ESF) and
the National Strategic Reference Framework (NSRF) Research Funding Program: THALES:
Reinforcement of the interdisciplinary and/or inter-institutional research and innovation.
References
1. Tassogiannopoulos A.G., 1986, A Contribution to the Study of the Properties of Structural Natural Stones of
Greece (in Greek), Ph.D. Dissertation (National Technical University of Athens, Greece).
2. Zambas C., 2004, Mechanical Properties of Pentelik Marble (in Greek) (Ministry of Culture, Committee for
the Preservation of Parthenon Publications, Athens, Greece).
3. Vardoulakis I., Kourkoulis S.K., 1997, Mechanical properties of Dionysos marble. Final report of the EV5VCT93-0300: Monuments Under Seismic Action, National Technical University of Athens, Athens, Greece.
4. Exadaktylos G.E., Vardoulakis I., Kourkoulis S.K., 2001, Influence of nonlinearity and double elasticity on
flexure of rock beams – II. Characterization of Dionysos marble, Int J Solids Struct, 38(22-23), 4119-4145.
5. Stavrakas I., Anastasiadis C., Triantis D., Vallianatos F., 2003, Piezo stimulated currents in marble samples:
precursory and concurrent-with-failure signals, Nat Hazard Earth Sys, 3, 243-247.
6. Triantis D., Anastasiadis C., Vallianatos F., Kyriazis P., Nover G., 2007, Electric signal emissions during repeated abrupt uniaxial compressional stress steps in amphibolite from KTB drilling, Nat Hazard Earth Sys,
7, 149-154.
7. Kyriazopoulos A., Anastasiadis C., Triantis D., Brown C., 2011, Non destructive evaluation of cement-based
materials from pressure-stimulated electrical emission-Preliminary results, Constr Build Mater, 25(4), 19801990.
90
ISBN: 978-960-98739-8-7
Simple Probabilistic Slope Stability Analysis
Georgios Belokas1*
1
Technological Educational Institute of Athens, *[email protected]
Keywords: slope stability, probabilistic, reliability, Eurocode 7
Abstract
Modern codes of practice for the analysis and the design of geotechnical works give the alternative of probabilistic analyses, which can also be associated to risk and hazard analyses.
This work exhibits how the First Order Reliability Method (FORM) can be applied to the
probabilistic slope stability analysis. The application of the FORM provides a simple but yet
intuitive view on the probabilistic approach for design purposes. It is applied to the simple case
of a planar failure but it can be easily extended to the application of general limit equilibrium
methods.
Introduction
Modern codes of practice for the analysis and the design of geotechnical works (e.g. Eurocode 7
– EC7), give the alternative of preforming probabilistic (non – deterministic) analyses by application of reliability theory principles. One method that can be used in geotechnical engineering
limit states analysis is the first order reliability method (FORM, e.g. Baecher & Christian, 2003,
Nomikos et al 2010, Orr & Breysse, 2008), which has been extensively used on the estimation
of errors, such as the uncertainty of the experimental results from laboratory tests (e.g. GUM:
1995, see ISO/IEC Guide 98-3:2008), and applies directly to cases where a closed form analytical solution exists.
Under the framework of EC7, the reliability analysis (uncertainty calculation) for a limit equilibrium problem can be performed with respect to the safety margin (SM). The safety margin
is expected to have a value of SM≥0.0 for a certain level of confidence, which requires the
knowledge of the uncertainty of the parameters that affect the value of SM. These parameters
normally include the value and the uncertainty of the external and internal loads (permanent
and mobile) and the strength constants, as well as the spatial variability and uncertainty of the
model (Kulhawy, 1992). In the current work the FORM is applied on the analytical solution of
the wedge slope stability.
Probabilistic equations for wedge slope stability
The safety margin is defined according to Eq. 1, where R corresponds to the resisting actions
and E the disturbing (or destabilizing) actions. This is also the definition adopted in EC7 and
should be generally preferred with respect to the factor of safety (FS=R/E), when performing
reliability analyses. This is so, because the disturbing actions (E) in the denominator enhance
any non-linearity, which should generally be avoided in the FORM (see Belokas, 2014).
SM = R – E
(1)
The geometry of the wedge stability problem is given in Fig. 1, in which W is the weight of the
wedge, H the height of the slope, β the angle of the slope (ΑΓ) with respect to the horizontal
and θ a random angle of the wedge slip surface (AB). No external loads or water level have
been considered.
ISBN: 978-960-98739-8-7
91
Figure 1. Geometry of planar two dimensional failure.
Forces equilibrium gives the SM, which is given by Eq. 2 results from the equilibrium equation,
where c is the soil cohesion, φ is soil angle of shearing resistance (c, φ are the Mohr – Coulomb’s failure criterion constants) and γ is the soil unit weight.
SM = c ⋅âH +
ãH
è 12
èsin ( ö −
)(ècos
2
tan
â − sin
è
) sin
 sin (2)
Ignoring the covariance ρR,D and applying the FORM, the variance of the SM is given by Eq.
3, which is accompanied by Eqs 4, 5 and 6. In Eq. 3, uc is the cohesion uncertainty, utanφ is the
the tanφuncertainty and uγ is the unit weight cohesion uncertainty. The probabilistic estimate
of the safety margin is given by SM – k·u(SM), where k a factor that corresponds to a specific
probability of exceedance.
S d , SM = u (SM ) =
(∂SM ∂c ) u
2
(
+ ∂SM
)
2
(
2
∂SM
utan
ϕ +
)u
2
(3)
∂SM
H
=
∂cè
sin
(4)

∂SM
1  1 2 sin ( â − è )
=
cos è   ãH
∂ tan ϕ èsin  2
âsin

(5)
∂SM
1  1 2 sin ( â − è )
=
è
ö
è (cos tan − sin
 H
∂ã
sin è  2
sin â
(6)
2
c
∂ tan ϕ

∂γ
2
γ
) 
Application example
In the following, the solution of the above problem is demonstrated for a geometry with β=40ο
and Η=9m. The estimates of shear strength are taken from Belokas (2014) and are quoted in
Table 1. The corresponding uncertainties are uc=6.1kPa, uφ=0.0179 and γ=2kN/m3. The best
estimate values together with the uncertainties are used for the probabilistic analysis. The design values are used for the deterministic analysis and, according to EC7, they are determined
by the characteristic values. The characteristic values are conservative estimates of the mean
value (best estimate). Greece has chosen the design analysis 3 of EC7 for the total stability of
92
ISBN: 978-960-98739-8-7
natural and cut slopes, which for the drained conditions gives: a) design strength cd=ck/(γΜγm),
tan(φd)=tan(φk)/(γΜγm) with γm=1.0 (independent of type of analysis) and γΜ=1.25, b) permanent
design actions Gd=γGGk, with γG=1.0 and και c) mobile design actions Qd=γQQk, with γQ=0 or 1.3
for favourable or unfavourable respectively.
Best estimates
tanφm
cm
(kPa)
16.4
γm
0.4125
(kN/m )
3
20
Conservative estimates
ck
tanφm
γk
(kPa)
(kN/m3)
6.8
0.4182
20
Design values
cd
tanφm
(kPa)
5.44
0.3346
γd
(kN/m3)
20
Table 1. Parameters adopted for the application example
The design values are used for the deterministic analysis, which by direct application of Eq.
2, give SMmin=5.6kPa and θcr=29o. Note that critical angle is the angle that gives the lowest
SM. For the probabilistic analysis, EC7 (ΕΝ-1997-1, clause 2.4.5.2 §11) defines a probability
of exceedance of p=5%, which for a normal distribution corresponds to k=1.64485 and to an
estimate given by SM – k·u(SM). The best estimates and their uncertainties are applied to Eq.
2 to 6, which result to a minimum value of SM – k·u(SM)=48.0kPa and θcr=32o. The critical
angle for the deterministic and probabilistic analyses is different as they correspond to different
soil properties. Moreover, for the probabilistic analysis the value of the SMmin, also depends on
the uncertainty of the soil constants, which greatly depends on the quality of the geotechnical
investigation. This is exhibited in Fig. 2, which shows the sensitivity of the probabilistic SM on
the magnitude of the uncertainty for the best estimates used above. For this specific problem of
planar failure the SM is mostly dependent on cohesion uncertainty.
Figure 2. Influence of every uncertainty coefficient on the value of SM – ku(SM) (Belokas, 2014).
Conclusions
The FORM was applied for the reliability analysis of a simple limit equilibrium planar failure
problem. This method can be adapted to the more general limit equilibrium method with slices.
When it come to probabilistic analysis it is preferable to examine the safety margin rather than
the factor of safety. It seems that the uncertainty of the cohesion is the most critical to the SM
value. A next step would be to compare the results of this probabilistic method to the result of
the Monte - Carlo method.
ISBN: 978-960-98739-8-7
93
References
1. Baecher G. and Christian J., 2003, Reliability and Statistics in Geotechnical Engineering, Wiley, p. 618.
2. Belokas G., 2014, Probabilistic Stability Analysis with First Order Reliability Method and Estimation of
Strength Parameters According to Eurocode 7, 7th Panhellenic Conference on Geotechnical Engineering.
3. ISO/IEC Guide 98-3:2008 Uncertainty of measurement – Part 3: Guide to the expression of uncertainty in
measurement, (GUM: 1995).
4. Kulhawy H., 1992, On the evaluation of static soil properties, Stability and Performance of Slopes and
Embankments – II, Proceedings of a Specialty Conference, Seed R.B. and Boulanger R.W. (eds), ASCE, pp.
95 – 115.
5. Nomikos P., Christodoulopoulou S. And Sofianos A., 2009, Simplified probabilistic tunnel design, 6th
Panhellenic Conference on Geotechnical and Geoenvironmental Engineering, Volos.
6. Orr T. and Breysse D., 2008, Eurocode 7 and reliability-based design. Reliability-Based Design in Geotechnical
Engineering, Computations and Applications, Kok-Kwang Phoon(edr), Taylor & Francis, pp. 298 – 343.
7. Pohl C., 2011, Determination of characteristic soil values by statistical methods, ISGSR 2011 - Vogt,
Schuppener, Straub & Bräu (eds), pp. 427 – 434.
94
ISBN: 978-960-98739-8-7
Embodied energy and embodied CO2 of building constructive
materials in a typical Hellenic dwelling
Georgios Syngros1and Dimitrios Koubogiannis1*
1
Department of Energy Technology Engineering, Technological Educational Institute
of Athens, Agiou Spyridonos Street, 122 10 Egaleo, Athens, Greece.
e-mails: [email protected], [email protected] (*Corresponding author)
Keywords: Embodied energy, embodied CO2, constructive materials, Hellenic
dwelling.
Abstract
Energy consumption in the building sector corresponds to the 40% of the total energy consumption in Europe and about 48% in USA, while it is responsible for 45% of the CO2 emissions
in the atmosphere. Current research mainly focuses to the reduction of operational energy of a
building, proceeding towards the development of near zero energy buildings. However, the total energy consumed by a building during its life-cycle, contains, apart from its operational energy, its Embodied Energy (EE), which is also related with the so-called Embodied CO2 (ECO2)
emissions. By means of a specific case study based on a typical Hellenic multifamily dwelling,
a methodology to estimate the EE and ECO2 values of the basic Greek building constructive
materials and the corresponding results are presented.
Introduction
According to the European Commission Directive 2003/30/EC and the European Climate Foundation, energy consumption in the building sector corresponds to the 40% of the total energy
consumption in Europe and about 48% in USA. In addition, building energy consumption is
responsible for about the 45% of the CO2 emissions in the atmosphere, which is accountable
for the greenhouse phenomenon. According to the Hastoe Housing Association (2015), the EE
of a building is the total energy required for its construction (raw materials extraction, manufacture, transportation to project site and installation). In particular, this EE is divided in three
categories: (a) Initial EE, i.e. energy required for production of materials and components of a
building, including extraction and manufacture of raw materials (called Indirect EE) and energy
required for transportation, assembly and installation on-site (called Direct EE). (b) Recurrent
E.E. concerning the maintenance of the building (replacement, renovation or demolition of
materials or parts of it). (c) Demolition energy, i.e. energy required for deconstruction of the
building and disposal of its materials.
The main objective of this work is to present methodology and results for the estimation of
the Initial EE and the corresponding ECO2 of common building construction materials used in
a typical Hellenic dwelling. Such a task continues previous research of the authors in which
similar analyses were performed for the electro-mechanical installations of typical Hellenic
dwellings (Koubogiannis et al, 2014a).
Description of case study
According to the Ministerial Decision 3046/304/89 of the Greek Building Regulation (Government Gazette 59/A/03.02.89), buildings and constructions in general, are classified, according
to their use in the following categories: dwellings, temporary accommodation facilities, hotels,
public gathering facilities, education, offices, markets, health and social welfare, industry, manufacturing, storages etc. Concerning the location of a building in the Greek territory, it may be
ISBN: 978-960-98739-8-7
95
located in one of the four distinct climatic zones A, B, C or D. Regarding Hellenic residential
building typologies, according to their construction decade, are categorized in the following
periods: pre-1980, during 1981-2000, during 2001-2010 and after-2010 that marks the EPBD
implementation in Greece (Dascalaki et al. 2011).
In the light of the above, the building case studied herein is described as a Multi-Family Dwelling (MFD). In particular, it is an urban three-floor building, each floor being an apartment, with
ground floor and a basement. It has been constructed in 2008 and it is located in Aigaleo (near
TEI of Athens) at Attica, i.e. at climatic zone A.
The two main sets of materials involved in a building are its construction materials (related to
civil engineering works) and those concerned with its electro-mechanical installations. In previous works of the authors, the methodology and results concerning the material, EE and ECO2
analysis of the electro-mechanical installations of typical Hellenic dwellings have been presented for the MFD mentioned above, as well as for a Single-Family Dwelling (SFD) (Koubogiannis et al, 2013, Koubogiannis et al, 2014b). The present work continues previous research
efforts and adopts a similar methodology for the construction materials of the building under
consideration.
Methodology
A three-step methodology was implemented, consisting of the material analysis, the mass analysis, the EE and the ECO2 analysis (Koubogiannis et al, 2014a).
Material analysis. Material analysis means to analyze a given set of materials and equipment
(called “Set”) into their constitutive materials. At first, the Set is divided in distinct groups of
major components (called “Groups”) in the form of a tree. For each Group, an “Item analysis”
is performed, i.e. the breakdown of the Group to its constitutive main items. In this task, the
major items (called “Items”) of the group are identified and these are continuously split into
sub-items till reaching the lower level of basic items (called “bitems”), i.e. entities that cannot be further split into sub-items. Thus, the material analysis is completed by identifying and
recording the constitutive single materials of the bitems. Summarizing, the material analysis
can be described by the sequence: Set, Groups, Items, sub-items, sub-sub-items, … bitems,
constitutive single materials.
According to the above methodology, the Construction Materials Set of a typical urban Hellenic
building can be divided in materials contained in the following distinct Groups:
• Bearing structure: reinforcement steel (in kg) and concrete (in m3).
• Internal and external masonry: bricks in general (in m3), mortar (cement, sand and lime in
m3), and senaz (concrete, in m3).
• Internal and external coatings: plaster in general (in m3).
• Insulation: for walls, floors and ceilings according to the building insulation datasheet (in
m3), extruded polystyrene or rockwool for waterproofing (depending on the type of wall
insulation) and extruded polystyrene foam board or rockwool (depending on the type of
flooring/ceiling insulation).
• Flooring and ceiling: materials of each level (basement, ground floor, upper floors, roof and
communal stairwell) such as marble, ceramic tiles, lime and concrete (in m3).
• Internal and external paintings: paints (in m2).
• Walls of kitchens and bathrooms: Ceramic tiles (in m2).
• Internal and external joinery: materials (in m2 or kg).
• Kitchens and cabinets: timber (in m3).
96
ISBN: 978-960-98739-8-7
Mass analysis. For each single material, its mass (in kg) is required. Thus, for the materials that
their quantities have been estimated in terms of volume, their mass is obtained multiplying the
volume by their density (found in the literature in kg/m3).
Embodied Energy and Embodied CO2 analysis. Transformation of the mass values (kg)
to EE requires the knowledge and use of the Embodied Energy (MJ/kg) and Embodied CO2
(kgCO2/kg) coefficients. Since these coefficients are nationally dependent parameters and currently there is no complete Hellenic database, values available in the international literature
were used (Hammond et al, 2008).
By means of the methodology described before, some benchmark values for the mass, EE and
ECO2 were obtained for a typical urban Hellenic dwelling. In particular, the total building mass
was estimated to be 1206600.9 kg (or about 1206.6 tones). The total EE and ECO2 values of
the building were estimated to be 3677552.5 MJ (or about 3677.6 TJ) and 320662,2 kgCO2 (or
about 320.6 tnCO2) respectively. The total building floor area was calculated 438 m2. Thus, the
values of mass, EE and ECO2 per unit floor area of the building are 2.75 tn/m2, 8.4 TJ/ m2 and
732 kgCO2/m2, respectively. The utility of the latter values is that they can be compared with
corresponding values from other buildings.
Table 1 presents the results of the methodology applied to the MFD case study under consideration. These results refer to percentage contribution in terms of the various materials in terms of
mass, EE and ECO2. For each value (mass, EE, ECO2), the materials have been placed in the
list according to their contribution to the total building value.
Material
Concrete
Brick Red
Plaster
Steel
Cement - sand
Lime
Ceramic tiles
MDF
Marble
Aluminium
Glass
Cement mortar
Polystyrene
Insulation
Mass [%]
59,8
17,5
9,8
3,4
3,2
2,5
1,7
0,8
0,7
0,2
0,1
0,1
0,1
0,1
Material
Steel
Brick Red
Concrete
Aluminium
Ceramic tiles
Plaster
Lime
Polystyrene
MDF
Cement - sand
Glass
Marble
Insulation
Cement mortar
EE
[%]
32,7
17,2
14,9
9,5
6,6
5,5
4,3
2,9
2,8
1,4
1,2
0,8
0,1
0,1
Material
Steel
Concrete
Brick Red
Lime
Aluminium
Ceramic tiles
Plaster
Cement - sand
MDF
Polystyrene
Glass
Marble
Cement mortar
Insulation
ECO2
[%]
33,2
23,3
15,1
7,1
5,7
4,7
3,8
2,5
2,1
1,1
0,7
0,5
0,1
0,0
Table 1. Percentage contribution in terms of the various materials in terms of mass, EE and ECO2.
With respect to the results of Table 1, some indicative remarks can be stated:
• The materials that dominate in terms of mass, EE and ECO2 become evident, not only qualitatively, but quantitatively with real world values estimated in practice. Even if we had an
idea about some of them from experience, it seems impossible to be able to predict their
comparative contribution to the total values without such an analysis. This demonstrates the
utility of such an approach, apart from deriving benchmark values as stated before.
ISBN: 978-960-98739-8-7
97
• Concrete totally dominates in terms of mass with a percentage of about 60%, while the second material (bricks) contribute 3.4 times less. The situation is similar in the EE and ECO2
lists, where among the various materials steel clearly dominates. However, the contribution
of the second material is 1.9 (for EE) and 1.4 (for ECO2) times less, as well as the contributions are spread in much more materials.
• The sorted list for the three quantities examined (mass, EE, ECO2) is not the same. For
example, while steel is in the fourth position in mass percentage, it is in the first position in
both the EE and ECO2 lists. This is due to the high values of EE and ECO2 per kg of steel.
• In terms of mass, the materials that contribute more than 5% are only three, while in terms
of EE and ECO2, they are six and five respectively. Again, the explanation has to do with
the fact that, despite their small percentage contribution in mass (where concrete thoroughly
dominates), they have high EE and ECO2 coefficients. The case of aluminum is a very characteristic example.
Concluding Remarks
A methodology for calculating mass, EE and ECO2 contribution of common building construction materials was presented and implemented in case of a typical Hellenic dwelling. Benchmark values for these quantities were derived. Dominant materials in terms of either mass, EE
and ECO2 were recorded. This useful information dictates directions in case someone wants
to examine ways to reduce for example the EE of a building. Current research on the topic by
the authors concerns the realization of similar analyses to other buildings, either of the same
typology in order to examine repeatability of the benchmark values and improve their values or
of different typology and compare with corresponding values and search for possible correlations.
Acknowledgement
The authors would like to acknowledge Mr. E. Proestakis, Mechanical Engineer, for providing
the technical drawings and data for the building used as case study in this work.
References
1. European Commission Directive (2003/30/EC) ‘The European Parliament and the Council of 8 May 2003
on the promotion of the use of biofuels or other renewable fuels’, (online), [Accessed 25 February 2015],
Available from: http://ec.europa.eu/index_en.htm.
2. Hastoe, Hastoe Housing Association, (online), [Accessed 25 February 2015]. Available from:http://www.
sustainablehomes.co.uk/upload/publication/Embodied%20Energy.pdf.
3. Koubogiannis, D.G and Balaras C.A. 2014a. Embodied Energy in Electro-Mechanical Installations of Hellenic
Dwellings, EinB2014 - 3rd International Conference “ENERGY in BUILDINGS 2014”, Athens, 4 November,
Greece.
4. Dascalaki, E.G., Droutsa, K.G., Balaras, C.A. and S. Kontoyiannidis. 2011. Building Typologies as a Tool for
Assessing the Energy Performance of Residential Buildings – A Case Study for the Hellenic Building Stock,
Energy & Buildings, 43(12): 3400-3409.
5. Koubogiannis, D.G, Daskalaki A. and Balaras C.A. 2013. A contribution to Building Lifecycle Analysis:
Embodied energy analysis of mechanical installations for a typical urban Greek dwelling. 3rd International
Exergy, Life Cycle Assessment, and Sustainability Workshop & Symposium (ELCAS3), 7-9 July, Nisyros–
Greece.
6. Koubogiannis, D.G., Lavoutas A., Lekkas A. and C.A. Balaras. 2014b. Estimation of Embodied CO2 in
Electro-Mechanical Installations for an Urban Hellenic Dwelling. International Conference on Buildings
Energy Efficiency and Renewable Energy Sources 2014 (BEE RES 2014), 1-3 June, Kozani–Greece.
7. Hammond, G.P and C.I. Jones. 2008. Inventory of Carbon and Energy (ICE) Version 1.6a. Sustainable Energy
Research Team, Department of Mechanical Engineering, University of Bath, [Accessed on 27/03/2014].
Available from: http://perigordvacance.typepad.com/files/inventoryofcarbonandenergy.pdf.
98
ISBN: 978-960-98739-8-7
Comparison of different Air Conditioning Systems implemented in a
typical office building in Greece
Ilia Lampro1and Dimitrios Koubogiannis1*
1
Keywords: Air conditioning, office building, climatic zones, energy consumption,
energy cost.
Abstract
A comparison of seven standard air conditioning systems is conducted by means of an office
building case study considered to be located at the four different climatic zones in Greece. At
each of the four representative geographical locations and for each of these systems, cooling
and heating load assessment is performed, as well as energy analysis and cost analysis in an
annual basis. Representative results are comparatively presented and discussed.
Introduction
According to EIA (Energy Information Administration, 2015), the existing building stock in the
European countries, as well as in the USA, are responsible for over the 40% of the total energy
consumption. Furthermore, the residential sector is responsible for over the 63% of the building energy consumption. In case of Greece, according to the Greek ministry of energy (Ypeka,
2015), buildings are responsible for the 36% of the total energy consumption and, from year
2000 to 2005, building energy increased by 24%, which is of the highest increases in the European Union. In addition, the low energy efficiency of many buildings in Greece is due to use of
insufficient thermal insulation or lack of it, old construction technology of windows and doors,
bad orientation of building, bad use of solar potential and inadequate service of heating and
air conditioning systems. Finally, it should be mentioned that, a significant portion of building
energy is that consumed by air conditioning systems and the various systems exhibit different
performance under different working conditions. In the light of the above, this work makes a
comparative study on the performance of seven standard air conditioning systems implemented
in the same reference building, which is considered to be located in the four different climatic
zones in the Greek territory. The aim is to assess the effect of the geographical location to the
system performance, to compare air conditioning systems in terms of various criteria (energy,
financial, environmental), as well as to draw useful conclusions concerning the implementation
of various such systems in Greece.
Case Study description
A typical office building located in the centre of Athens was selected and the relevant technical
drawings were made available, as well as information for its shell construction. In the present
study, an intermediate floor, namely the third of a total number of five, is studied. It is assumed
that floors above and below it are air conditioned spaces, so no thermal loads occur between
them. The height of the floor is 3.5m (including 0.5m suspended ceiling). Fig.1 depicts the
building orientation, as well as the two comfort zones (I and II) defined on the reference floor.
This floor was considered to comprise the offices of a company. A time schedule was assumed,
namely working hours 800–1800, including lunch break (1300–1400) and meeting hour (1700–1800).
The building was considered to be located in a representative city for each of the four climatic
ISBN: 978-960-98739-8-7
99
zones in Greece, namely at Iraklio for zone A, Athens for B, Thessaloniki for C and Florina for
D. At all locations, the indoor design conditions were set to 22oC dry-bulb (db) for winter and
24oC db for summer, with a relative humidity of 50% for both seasons. Infiltration rate was set
to 1 ach-1. The overall U-value for external walls of the building is 0.584 W/m2K and 3,679 W/
m2K for the windows. Rational assumptions were made for the equipment and number of people in each office according to the floor schedule. Space lighting was set to 15 W/m2 and ventilation rate to 2.5 l/s per person. Seven air conditioning systems were considered: SPLIT type
Units (SPLIUs), Air to Water Heat Pump with Fan Coil Units (AWHP-FCUs), Water to Water
Heat Pump with Fan Coil Units (WWHP-FCUs), Air to Air Variable Refrigerant Flow system
(AAVRF), Water to Air Variable Refrigerant Flow system (WAVRF), central Air Handling Unit
with Air to Water Heat Pump (AHU-AWHP), central Air Handling Unit with Water to Water
Heat Pump (AHU-WWHP).
Figure 1. Reference floorplan of the of the office building under consideration.
Methodology
The methodology consists of the following steps for each system and building location (referring to the same building with the same orientation):
1. Selection of outdoor design criteria, based on local weather data. According to the latter,
August and January are the months to calculate design cooling and heating loads.
2. Calculation of design loads, i.e. cooling (summer) and heating ones (winter). Cooling loads
were carried out by means of the Hourly Analysis Program (HAP) (AHI Carrier LTD, 2015),
software appropriate for designing building HVAC systems.
3. Design of the whole system installation, including relevant drawings and assessment of the
feasibility of such an installation, its operation and thermal comfort it provides.
4. Simulation of the annual system operation based on rational assumptions and using available information for the system in order to calculate its energy consumption, according to
the partial loads involved and the hourly temperature of outdoor air.
5. Economical analysis of the system, consisting installation cost, operational energy cost and
financial cash flow on an annual basis, in order to estimate payback period.
6. Environmental analysis, in which the estimation of carbon dioxide emissions and possible
impact assessment by the system refrigerant are taken into account.
Table 1 presents design and annual cooling and heating loads for the reference building per climatic zone. Table 2 presents the consumption and cost of operational energy per climatic zone
for each system.
100
ISBN: 978-960-98739-8-7
Cooling loads [kW]
Climatic
zone
Sensible
A
B
C
D
25,6
24,3
23,6
22,7
Latent Total
3,1
3,1
3,1
3,1
28,7
27,4
26,7
25,8
Heating
loads
(kW)
11,3
15,0
18,9
23,7
Annual energy load [kWh]
Cooling
Heating
Total
53541,5
51928,8
49849,7
46405,6
4157,5
6267,8
8840,5
12527,6
57699,0
58196,6
58690,2
58933,2
Table 1. Design and annual cooling and heating loads for the reference building.
According to Table 1, the maximum cooling loads occur in zone Α (28,7kW), which is the
southern climatic zone with the higher mean solar gains. However, referring to air source systems (SPLIUs, AWHP-FCUs, AAVRF, AHU-AWHP) and according to Table 2, their maximum
operational energy consumption and costs occur in climatic zone B, because outdoor temperatures in zone B are higher than those of zone A. The numbers validate in practice the fact known
by theory, that performance of air source systems drops whenever outdoor temperature rises. A
similar result is not valid in water source systems, where earth temperature is important and this
temperature is almost constant at any location.
In Table 1, the annual energy load is of about the same value independent the climatic zone, due
to the fact that the greater the cooling loads in a zone, the lower the heating ones and their sum
happens to remain almost constant.
From Table 2, it seems that AHU-AWHP and AWHP-FCUs exhibit same energy consumption.
The same happens with AHU-WWHP and WWHP-FCUs. In fact, the energy consumption of
the system with AHU is a little higher compared to the results shown in Table 2, because zone I
(the northern between the two thermal zones of the floor) requires the use of a reheat coil. The
latter requires an extra heat pump of small cooling capacity to produce hot water at summer, as
well as the use of a steam boiler at winter. However these two components have not been taken
into account in the energy consumption simulation (due to difficulties in their modeling).
Systems containing a WWHP, namely AHU-WWHP and WWHP-FCUs outperform at any zone
the rest of the systems in terms of annual energy consumption and cost, followed by WAVRF
(Table 2 and Fig.2). Concerning the effect of the different location, it seems that zone B requires
the highest consumptions and this is pronounced in the case of systems consuming more energy
(Fig.2 right).
Qualitatively, in terms of moist air processes, the systems containing an AHU are more advantageous (for example they can provide humidification or fresh air to the air conditioned space).
On the other hand, the installation of these systems is more complex, permanent and costly. The
SPLIUs exhibit the simplest installation and of the lowest cost. Finally, water source systems
exhibit large installation costs, due to the high the cost of geothermal heat exchanger. However,
this cost can be diminished whenever the installation can recover heat either from waste water
or from lakes and rivers.
ISBN: 978-960-98739-8-7
101
Figure 2. Annual total energy loads per system in zone B (left), per zone for each system (right).
Conclusions
A parametric comparative study of seven air conditioning systems implemented in a reference
floor of an office building in the four climatic zones of Greece was carried out. Annual energy
building loads are about the same independent the climatic zone. While maximum cooling
loads occur in zone Α, for air source systems maximum energy consumption occurs in zone B.
WWHP systems outperform the rest of systems in terms of annual operational consumption
and cost followed by WAVRF. Climatic zone B requires the highest consumptions and this is
pronounced in the case of systems consuming more energy. Research under progress includes
thorough economical study for the extraction of the payback period times, as well as comparison of the systems in terms of environmental and thermal comfort issues.
Zone
Energy consumption [kWh]
Daily
Μonthly
Daily
Cooling
Μonthly
Heating
A
B
C
D
89,0
97,1
85,7
75,1
2757,4
3010,6
2657,0
2328,1
13,7
19,2
24,8
32,6
425,3
595,7
767,9
1011,9
A
B
C
D
98,0
105,3
93,9
84,0
3038,7
3264,2
2911,7
2603,2
8,2
13,7
21,4
35,2
254,7
424,4
662,7
1091,5
A
B
C
D
51,6
49,9
46,0
39,7
1598,6
1547,2
1427,1
1230,2
6,8
10,1
14,3
20,6
210,1
311,6
443,2
639,7
A
B
C
D
84,7
93,3
81,7
72,3
2625,6
2891,1
2531,7
2240,7
7,6
11,7
16,8
24,8
234,1
361,5
519,3
768,2
102
Energy cost [€]
Annual
Annual
Cooling
Heating
SPLIUs
13384,0
1698,9
14119,7
2455,3
12564,7
3302,6
10893,6
4475,3
AWHP-FCUs
14506,0
964,5
15494,0
1592,1
13800,0
2486,9
12224,0
4074,6
WWHP-FCUs
7705,9
828,7
7420,8
1249,7
6799,8
1832,4
5840,0
2722,8
AAVRF
12491,0
907,3
13660,0
1411,2
11951,0
2064,7
10476,0
3090,5
Total
Cooling
Heating
Total
15082,9
16575,0
15867,3
15368,9
2208,4
2329,8
2073,2
1797,4
280,3
381,0
544,9
738,4
2488,7
2710,8
2618,1
2535,8
15470,0
17086,0
16287,0
16298,0
2424,9
2590,1
2307,1
2043,2
190,9
314,2
483,7
780,3
2615,8
2904,3
2790,8
2823,5
8534,6
8670,5
8632,2
8562,8
1288,3
1240,6
1136,9
976,3
162,2
243,7
350,6
509,4
1450,5
1484,3
1487,5
1485,7
13398,0
15071,0
14016,0
13567,0
2088,0
2283,4
1997,9
1751,1
178,4
276,4
397,6
583,9
2266,4
2559,8
2395,5
2335,0
ISBN: 978-960-98739-8-7
A
B
C
D
45,7
44,3
41,6
38,4
1416,9
1373,5
1288,6
1189,4
5,0
7,4
10,9
16,6
154,9
229,8
336,9
513,5
A
B
C
D
98,0
105,3
93,9
84,0
3038,7
3264,2
2911,7
2603,2
8,2
13,7
21,4
35,2
254,7
424,4
662,7
1091,5
A
B
C
D
51,6
49,9
46,0
39,7
1598,6
1547,2
1427,1
1230,2
6,8
10,1
14,3
20,6
210,1
311,6
443,2
639,7
WAVRF
8534,1
680,2
8205,2
1026,6
7488,3
1552,6
6623,4
2441,4
AHU-AWHP
14506,0
964,5
15494,0
1592,1
13800,0
2486,9
12224,0
4074,6
AHU-WWHP
7705,9
828,7
7420,8
1249,7
6799,8
1832,4
5840,0
2722,8
9214,3
9231,8
9040,9
9064,8
1426,8
1371,8
1252,0
1107,2
133,1
200,2
297,0
456,2
1559,9
1572,0
1549,0
1563,4
15470,0
17086,0
16287,0
16298,0
2424,9
2590,1
2307,1
2043,2
190,9
314,2
483,7
780,3
2615,8
2904,3
2790,8
2823,5
8534,6
8670,5
8632,2
8562,8
1288,3
1240,6
1136,9
976,3
162,2
243,7
350,6
509,4
1450,5
1484,3
1487,5
1485,7
Table 2. Operational energy consumption and cost per climatic zone for each system.
References
1. Energy Information Administration, Energy consumption in USA homes. Date of access: 20/06/2015. http://
www.eia.gov/todayinenergy/detail.cfm?id=10271
2. Ypeka.gov.gr, Ministry of energy in Greece, Building Energy consumption in Greece. Date of access:
26/06/2015. http://exoikonomisi.ypeka.gr/Default.aspx
3. AHI Carrier LTD, Hourly Analysis Program (version 4.51). Date of access: 30/06/2015. http://www.ahicarrier.gr/el/downloads/hourly-analysis-program/
ISBN: 978-960-98739-8-7
103
Optimal Flow Control of Laminar Vortex Shedding around a Cylinder
Ioannis Bonis1, Haralambos Sarimveis2 and Dimitrios Koubogiannis1*
2
School of Chemical Engineering, National Technical University of Athens,
9 Heroon Polytechniou Street, 157 80 Zografou, Athens, Greece.
e-mail: [email protected]
1
Keywords: Dynamic optimization, vortex shedding, cylinder wake, CFD.
Abstract
Vortex shedding often occurs in flows around structures. In applications where flow-structure
interaction may damage the structure, vortex shedding needs to be suppressed. In this work, the
wake control behind a cylinder for laminar flow (of Reynolds number 100) is formulated as a
dynamic optimization problem. The angular speed of the cylinder is the manipulated variable,
adjusted to achieve vortex shedding suppression by means of minimizing the lift coefficient
variation. The optimal angular speed is assumed to be periodic like the wake formation. The
control problem is solved and vortex suppression is checked. Different values for the time horizon tH used to calculate the optimal angular velocity are tested. The impact of tH to control is
evaluated and the need for feedback is assessed.
Introduction
Flow separation is commonly exhibited in nature and the built environment. The phenomenon
gives rise to vortex shedding that can potentially damage the structure behind which vortices
are formed, due to induced vibrations, so it is important to control the phenomenon effectively.
To this end, both active and passive control techniques can be applied. Boundary layer control
by means of rotating cylinders is of the most promising ones (Graham et al, 2000) and this
is followed herein. The proposed method is based on a detailed transient 2D model of a flow
around a cylinder exhibiting laminar vortex shedding that is numerically simulated by Computational Fluid Dynamics (CFD) and has been validated in recent studies (Koubogiannis et al,
2014). The cylinder rotation strategy for a given Reynolds number is calculated by solving an
optimal control problem using the transient CFD model as a constraint. The dynamic optimization problem is solved using a feasible path method. The cylinder angular velocity is considered
to follow a sinusoidal law in time described by three parameters (amplitude, frequency and
phase) which are the independent variables of the optimization problem, the objective of which
is vortex shedding minimization expressed by a suitable measure. The goals of this work are to
formulate the optimal control problem, validate the solution in a benchmark test case, evaluate
the effect of the time horizon tH on which control is based for a given Reynolds number, as well
as assess the need for feedback.
Test case description
The two-dimensional cylinder problem, concerning the flow of a fluid with density ρ past a stationary or rotating cylinder of diameter D [1] serves here as a test case. Governing equations are
the continuity and momentum (Navier-Stokes) equations, written in dimensionless form as:
104
ISBN: 978-960-98739-8-7
∂u
1 2
∇ ⋅u = 0
+ (u ⋅∇)u = −∇p +
∇ u (1)
∂
t
Re
The Reynolds Re number is based on the inflow velocity. Boundary conditions for Eq. 1 are
∂u x
= 0, u y = 0 on Γ 2 , Γ 4 and u = (γ x , γ y ) on Γ c , γ being the
∂y
angular velocity of the cylinder, while initial condition is u t =0 = (0, 0) . Eq. 1 are numerically solved by means of COMSOL Multiphysics software (Multiphysics, 2008) on an unstructured mesh (of 4528 triangular elements and 20766 nodes) by the Finite Element Method using
Lagrange P2-P1 elements. The computational domain is shown in Fig. 1 (left). In what follows,
the discretized Eq. 1 with its boundary conditions will be referred to as fNS-c = 0. Every solution
of this comprises a set of velocity and pressure fields for a given time that satisfy the physics
of the problem expressed by Eq. 1. The parameters of this system are Reynolds number (Re)
and angular velocity of the cylinder (γ). Here, laminar flow of Re=100 with γ=0, which exhibits
vortex shedding and turns to be periodic, was simulated and validated against literature data
u = (1, 0) on Γ1 , p = 0 on Γ3 ,
(Graham et al, 1999). Fig. 2 (right) depicts the time evolution of the cylinder lift ( cL ) and drag (
cD ) coefficients. A period, T, of 5.5 nondimensional time units characterizes the phenomenon.
Figure 1: The computational domain and computational mesh for the benchmark stationary cylinder problem of
Re=100 (left). Time evolution of the lift and drag coefficients computed by the CFD model (right).
Methodology
The aim is to minimize wake formation in the system at hand by means of minimizing the variation of lift coefficient. A dynamic optimization problem is formulated in order to identify the
optimal values of the control parameter z, as follows
zopt = arg min {f (z; t H )}, where f (z; t H ) =
z
tin + t H
∫ [c
L
( γ ( z , t ; t H ), u, t )
] dt
2
tin
subject to γ ( z , t ; t H ) = z1 sin( z2t + z3 ) and
f NS −c (γ ( z , t ; t H ), u , t ) = 0
(2)
The manipulated variable of Eq. 1 is the angular velocity γ. Since wake formation is periodic,
a periodic expression is also considered for the control variable γ which is parameterized by a
sinusoidal expression in time γ ( z , t ; t H ) = z1 sin( z2t + z3 ) , employing vector of three independent variables z = [z1 , z2 , z3 ] (magnitude, frequency and phase of the sinusoidal control action).
The time horizon tH used to solve the optimization problem is treated as a parameter. Here tH is
set to be a multiple of the period T of the uncontrolled stationary cylinder vortex shedding. In
particular, the values T, 2T and 5T are used. The application of a control action to the system
T
ISBN: 978-960-98739-8-7
105
perturbs and alters its periodicity. For the optimization problem examined herein, the cylinder
is initially considered stationary and control is applied after the system reaches a periodic state.
Thus, the system is simulated with γ=0 and zero initial conditions for adequate time and when
periodicity has been established, the control is applied at time instance tin. A value of tin=100
was used here based on Fig.1 (right). Initial conditions for the optimal control problem are
the state variables (velocity and pressure) for γ=0 and t= tin. The objective function f (x; t H )
corresponds to the area below cL2 for time interval tH. By minimizing this minimization of cL
variation is sought and accordingly suppression of vortex shedding. The optimization problem
is solved by the MATLAB active-set implementation of active-set deterministic optimization
(Mathworks, 2007). The independent variables are bounded so that z1 is positive, z2 is allowed
to vary between –π/2 and π/2 and z3 between 0 and T/2. Computing cost is accounted in terms
of function evaluations rather than actual computational time. Most of this time is expended on
the solution of the constraint (CFD problem).
The optimal control problem 2 aims to minimize variation of cL in time tH after enabling the
controller. Three values were used for tH, namely T, 2T and 5T. For each tH, a different strategy γ ( zopt , t ; t H ) = z1,opt sin( z2,opt t + z3,opt ) was found. Three evaluations of the objective function
were performed by f (zopt ; t f )=
tin + t f
∫
tin
2
 cL ( γ ( zopt , t ; t H ), u, t )  dt for tf = T, 2T, 5T), i.e. having
set the control action to a particular γ ( zopt , t ; t H ) , the objective f (zopt ; t f )was computed for different intervals tf under the action of the same controller. Table 1 summarizes the results.
γ=0
γ ( zopt , t ; T )
γ ( zopt , t ; 2T )
γ ( zopt , t ;5T )
reduction in f (zopt ; t H )
f (zopt ; T )
3.22
1.83
2.49
2.84
43.2%
f (zopt ; 2T )
6.64
23.80
6.11
6.26
7.98%
f (zopt ;5T )
16.96
461.40
18.80
16.78
1.06%
(
)
Table 1. Evaluation of objective function f zopt ; t f for different controllers γ ( zopt , t ; t H ) and different time
intervals tf. Reduction of the controlled with respect to the uncontrolled objective function is also included.
At each line of Table 1, the corresponding objective f (zopt ; t f
) becomes minimum for t =t , i.e.
f
H
min f (zopt ; t f )= f (zopt ; t H ) (Table 1, cells in grey). This is logical, since the control horizon
used in the optimization problem’s objective function was tH and suggests that the controller
works correctly, at least qualitatively. Looking at each column of Table 1, we notice that the
106
ISBN: 978-960-98739-8-7
greater the time interval tf , the greater the corresponding value of f (zopt ; t f ), an obvious fact,
since the (positive) integral in equation (3) concerns a greater time of integration. Another remark is that the lower the tH, the greater the f (zopt ; t f ) value for greater tf. This means that the
objective function exhibits a greater deviation from both controlled and uncontrolled cases in
the long-term when it has been “trained” using small time horizon. Finally and most importantly, from the same Table 1 we can see that the percentage reduction in f (zopt ; t H ) becomes significantly greater for smaller values of tH. This reduction is a measure of control benefits, since
it expresses the vortex suppression of the optimally controlled cylinder versus the benchmark,
i.e. the uncontrolled system with stationary cylinder. For larger control horizons, the benefits of
optimal control diminish. Thus, it is more advantageous to perform control using small “bursts”
than relying on control actions of larger duration. Hence, controller formulations with finite
control horizon, such as Model Predictive Control, seem more promising than optimization for
a long, or infinite, time horizon. Shorter control horizons may suggest the need for feedback,
to avoid accumulation of approximation errors, which will lead to insufficient closed-loop performance.
The optimal control variables are zopt,T=[0.791,-0.065,0.207]T, zopt,2T=[1.012,-0.004,0.048]
T
and zopt,5T=[0.967,-0.0006,0.018]T, for which 84, 71 and 76 evaluations were required, respectively. It seems that the values of z1,opt are of the same order of magnitude in all three
cases. On the other hand, z2,opt seems to decrease as the control horizon increases, and thus
to weaken the impact of time t on the variation of γ, since the term z2opt t in the expression
γ ( zopt , t ; t H ) = z1opt sin( z2 opt t + z3opt ) becomes smaller. This is why in Fig. 2 (left) γ is almost
straight line (it holds that for small φ values sin(φ ) ≈ φ , thus γ ( zopt , t ; t H ) ≈ z1opt ( z2 opt t + z3opt ) is
a straight line). In other words the control function has a large period and the larger the “training” horizon tH the larger this period is (Fig.2 right).
Figure 2. Variation of angular velocity γ for various tH at time T (left) and at the long-term (right).
Concluding Remarks
A framework was presented for the optimal control of systems exhibiting vortex shedding.
The case studied concerns laminar wake formation behind a cylinder, the angular speed of
which was the manipulated variable to minimize vortex shedding. Since wake formation is periodic, the optimal angular speed was considered to be sinusoidal. The control method worked
satisfactorily. Parametric studies on the control horizon, showed that it is more advantageous
to perform control using small “bursts” rather than relying on control actions derived from
considering long time horizons. Further work concerns the effect of Reynolds number to the
ISBN: 978-960-98739-8-7
107
optimal control law, the adoption of reduced order modeling techniques or surrogate models for
the substitution of the full CFD model while preserving most of its accuracy to save computing
time and implementation of MPC instead of optimal control.
References
1. Graham, W. R., Peraire, J., Tang, K. Y., 1999, Optimal control of vortex shedding using low-order models.
Part I—open-loop model development. International Journal for Numerical Methods in Engineering, 44(7),
945-972.
2. Koubogiannis, D.G., Sarimveis, H., Bonis, I., 2014, “Development of a reduced order model for the numerical
simulation of the laminar flow around a rotating cylinder”, FLOW 2014: 9th Panhellenic Conference on Fluid
Transport Phenomena.
3. Multiphysics, C. O. M. S. O. L., 2008, 3.5 User’s Guide. Comsol AB.
4. Mathworks, I., 2007, Matlab Optimization Toolbox User’s Guide. MathWorks, Inc. Date of access: 25/09/2015.
http://www. mathworks. com/access/helpdesk/help/pdf_doc/optim/optim_tb. pdf.
108
ISBN: 978-960-98739-8-7
Using b-value analysis to estimate pre-failure characteristics of
cement mortar beams subjected to Three-Point-Bending
I. Dakanali1*, A. Kyriazopoulos2, I. Stavrakas2, D. Triantis2
Laboratory of Testing and Materials, Department of Mechanics, School of Applied
Sciences, National Technical University of Athens, *[email protected]
2
Laboratory of Electronic Devices and Materials, Department of Electronic
Engineering, Technological Educational Institute of Athens, Greece
1
Keywords: Acoustic emissions, cement mortar, b-value, 3 point bending (3PB), failure
Abstract
The aim of this work is to evaluate the use of the Acoustic Emissions (AE) recording technique
to estimate the critical load level during 3 Point Bending (3PB) tests on cement mortar specimens. Under this concept, series of 3PB have been conducted on white cement mortar prismatic
specimens with dimensions 5cm x 5cm x 20cm. The mortar was prepared using coarse and fine
quartz sand mixed with Aalborg white cement. Each specimen was subjected to a constant mechanical load for 200s. The applied load was 0.85Lf, 0.87Lf, 0.91Lf, 0.95Lf, 0.99Lf with respect
to the ultimate flexural load (Lf). When the applied load was in the vicinity of the Lf (0.99Lf)
the specimen collapsed after 84s. While the load was maintained constant, AE recordings were
conducted using the PCI-2 Physical Acoustic Corp. acquisition system. The AE sensor was
mounted near the midspan of the beam since this is the location where the main crack develops
and at the same time it is far from the support points (i.e. 6.5cm) that generate parasitic AE
events. In order to estimate the critical load value that may lead to the failure, the well-known
statistical quantity b-value was calculated using the improved b-value (Ib) formalism.
Introduction
Definitely it is very crucial, potential defects and critical damages on structures to be detected
on time and in their very beginning preventing their abrupt failure. For this reason the Non
Destructive Techniques have been developed. One of the most famous and widely known nondestructive techniques is the Acoustic Emission (AE) technique. AET can be considered as
“passive” non- destructive technique as it recognizes events only when they occur during the
experiment because of a developing fracture. If there is not an external stimulus, a change in
material‘s condition, it is impossible AE to be produced.
The uniqueness of AE technique is the monitoring changes in materials properties over a long
time and detecting crack initiation and propagation occurring not only on the surface but also
deep inside a material. When a material is under stress and the stress field exceeds the material capacity, a spontaneous release of energy occurs due to the initiation of new cracks or the
developing of the existing. These waves propagate spherically outwards from the source and
travel through the material. Finally they are recorded by the acoustic sensors on the material’s
surface. The signals that produced by the aforementioned way are burst and discrete. The range
of AE frequencies can be from 20kHz up to several MHz. The AE detection is used for damage
localization and damage level assessment in brittle materials such as concrete (Aggelis et al
2013, Stergiopoulos et al 2013) and rocks (Lockner 1993, Stanchits 2006). For cement based
materials and concrete structures, AE testing is estimated as the most precise and useful method
to monitor crack growth (Rao, Lakshmi, 2005).
ISBN: 978-960-98739-8-7
109
Specimens- material
Based on AE technique, specific AE signals could be correlated with the mechanical behavior
of the material and give the sign of the upcoming failure. A series of 5 specimens of cement
mortar have been produced in order to be subjected to a 3 point bending test. The specimens
were 20 centimeters long, 5 centimeters wide and 5 centimeters tall.
The mortar was prepared using coarse and fine quartz sand mixed with Aalborg white cement.
The mortar mix that has been followed was the same used in the Acropolis’ restoration works.
The proportion was two parts of coarse quartz sand (grain size 1-2mm), one part fine quartz
sand M32 (grain size 0.1-0.4mm) and one part white Aalborg cement.
The quartz (silica) sand must be perfectly dry, clean and well graded. The proper ratio of grain
sizes can efficiently reduce the gaps between them. In this way, the product attains high friction
strength and consistency. The proper particle size distribution of sand can also minimize the
initial mortar cracks.
White cement of Aalborg consists of pure limestone and fine-ground sand. It is a rapid hardening Portland cement with high early (2 days) and standard (28 days) strengths. The uniqueness
of Aalborg White is its white color, the high consistency, the extraordinarily low content of
alkali and the high sulphate resistance.
Test procedure
After the 28-day curing of mortar, the specimens were subjected to 3PB test. Each specimen
was placed on two supports that were 13cm apart and the actuator applied a force in the exact
middle of the two supports.
Before the final tests, a series of preliminary tests was conducted in order to be determined the
ultimate failure Load (Lf). The applied load on each specimen was different portion of Lf (Figure 1). Specifically, the load for each specimen was respectively 0.85Lf, 0.87Lf, 0.91Lf, 0.95Lf,
0.99Lf. The desirable load reached its final value slowly and gradually as to be considered quasi
static. After that, it was remained stable for 200seconds. This duration was considered adequate
for AE relaxation. It is worth to be mentioned that the last specimen collapsed before the end
of the 200 seconds.
For the AE recording was used two sensors mounted between the support span of the specimen, near to the location that the fracture is expected to be occurred. It is essential, in order to
be achieved the perfect contact between the sensor and the material surface, the space between
them to be filled with a kind of couplant.
b value analysis
One widely used method evaluating AE signals is the b-value analysis (Colombo et al, 2003).
Generally, the high b-values declare the first microcracks that generate in the early stages of
damage (Rao, Lakshmi, 2005). These AE signals are many and have low amplitude. On the
other hand, when b-value takes lower values, this typically means that abrupt events are taking
place. These events show the existence of unstable cracks that leads to the final collapse. They
have high amplitude and low frequency.
In 2001, Shiotani et al. proposed the “improved b- value” (Ib-value). The Ib-value uses statistical values as mean and standard deviation of AE amplitude that varies during the test.
The Ib-value defines as:
Ib =
log N ( µ − α1σ ) − log N ( µ + α 2σ )
(α1 + α 2 ) ⋅ σ
(2)
where σ standard deviation, μ mean value of the amplitude distribution, α1 coefficient related
to the smaller amplitude, α2 coefficient related to the fracture level. The values of α1, α2 coeffi-
110
ISBN: 978-960-98739-8-7
cients vary between 0.5-5.0. Considering that changing the α1 and α2 did not significantly affect
to the Ib-value, their value remained constant and equal to one (α1=α2=1).
The crack onset or the crack propagation causes an acoustic event. Starting the test, definitely
there is parasitic acoustic events mix with those ones caused by the material cracking. In order
to avoid them and obtain pure information, the recorded events during the steady time span of
the 200 seconds were isolated.
improved -b value
2.0
1.5
L/Lf= 0.85
L/Lf= 0.87
L/Lf= 0.91
L/Lf= 0.95
L/Lf= 1
1.0
0.5
0
50
100
150
t (s)
200
Figure 1. Temporal variation of Ib-value for each sample. The values were calculated for an amount of 50 AE
events
During each loading, the Ib value temporal variation was calculated using a 10 event sliding
window of 50 AE events. Figure 1 shows the results of the conducted analysis. The Ib value
gradually increases during the time span of 200s as expected since the b value is related to the
mean AE amplitude variation.
2.0
1.5
60
Ib-value
AE Amplitude (dB)
70
1.0
50
0.5
40
0.0
0
50
100
150
200
250
time (s)
Figure 2. Improved-b values and AE amplitudes (L/Lf=0.9) according to the time
Furthermore, when the applied load is in the vicinity of failure, an indicative characteristic bvalue decrease is observed before the specimen failure. Another interesting observation is that
for higher applied load values the corresponding calculated Ib values are lower. Specifically,
when the L/Lf is greater than 0.9, all the b_values, during the 200s, are less than 1.5.
This observation was confirmed by new similar experiments (figure 2) where the corresponding
b values for L/Lf =0.9 were near to unity in the beginning of the 200sec time span and approxiISBN: 978-960-98739-8-7
111
mately between 1.5-1.7 at the last seconds. This seems to be an indicator that the specimen is
near to its failure.
Concluding Remarks
When the applied bending load is in the vicinity of the ultimate strength of the specimens there
is intense AE activity. The Ib values decrease significantly as the maximum bending load increases respectively to each specimen.
The above observations lead to the conclusion that the AE analysis provides reliable information that may be used as a pre-failure characteristic index under the concept of Non-Destructive
Testing (NDT) techniques.
Using the statistical b-value quantity can provide a pre-failure factor that can forewarn the
specimen’s collapse.
Acknowledgements
This research has been co-financed by the European Union (European Social Fund – ESF)
ARCHIMEDES III. Investing in knowledge society through the European Social Fund.
References
1. Aggelis , D.G. Mpalaskas, A.C. Matikas, T.E. (2013) Investigation of different fracture modes in cementbased materials by acoustic emission, Cement and Concrete Research 48 1–8
2. Colombo S., Main I.G., Forde M.C. (2003), Assessing damage of reinforced concrete beam using ‘‘b-value’’
analysis of acoustic emission signals, Journal of Materials in Civil Engineering, 15, 280-286.
3. Lockner, D. (1993), The role of acoustic emission in the study of rock fracture. Int. J. Rock Mech. Min. Sci.
Geomech. Abstr., 30, 883–899.
4. Rao, M. V. M. S., Lakshmi, Prasanna K. J., (2005), Analysis of b–value and improved b–value of acoustic,
Current Science 89 (9), 1577–1582.
5. Stanchits , S., Dresen , G., Vinciguerra, S. (2006), Ultrasonic velocities, acoustic emission characteristics and
crack damage of basalt and granite, Pure Applied Geophys.163 , 5–6, 975–994
6. Shiotani, T., Yuyuma, S., Li, Z., Ohtsu, M. (2001), Application of AE improved b-value to quantitative
evaluation of fracture process in concrete materials, Journal of Acoustic Emission, 18, 118-133.
7. Stergiopoulos, C. Stavrakas, I. Hloupis, G. Triantis, D. Vallianatos, F. (2013), Electrical and acoustic emissions
in cement mortar beams subjected to mechanical loading up to fracture, Engineering Failure Analysis, Volume
35, pp 454–461.
112
ISBN: 978-960-98739-8-7
Green Islands in Europe and Prospects for Greek Islands. The Tilos
Project
J.K. Kaldellis*, G. Salagiannis, N.C. Ilia, P. Stinis, K. Dimakis
Lab. of Soft Energy Applications and Environmental Protection,
Piraeus University of Applied Sciences, www.sealab.gr, *[email protected]
Keywords: Renewable sources, Autonomous Islands, Green Energy Islands, Hybrid
renewable energy systems
Abstract
There are many islands around the world which are facing energy problems since they are often
not interconnected to mainland΄s electrical networks and have to manage autonomously their
energy needs mainly based on fossil fuels combustion. This results in higher production costs
and greenhouse gas emissions. At the same time, islands are characterized by significant renewable energy potential (especially wind and solar). In the recent years, efforts have been put in
Europe for the sustainable exploitation of the available renewable energy sources (RES) in such
islands in order to develop and implement small scale renewable energy based solutions. These
autonomous energy systems are able to meet the electricity demand with relatively low production cost, providing reliable and environmental solutions. This paper aims to review first the existing cases of European Green islands -El Hierro, Isle of Eigg, Samsoe, Utsira- which adopted
RES-based and storage technologies for energy autonomy purposes, in order to extract results
and gain experience from their operation. Furthermore, the prospects of the smart microgrid of
the Greek island TILOS are presented. This microgrid relies on a hybrid system that combines
renewables (wind and PV) with advanced NaNiCl2 batteries, adopting demand management
strategies and aiming to interact with the existing electricity network.
Introduction
The threat of climate change combined with the continued depletion of fossil fuel reserves has,
in recent decades, raised a demand for a more efficient, sustainable energy supply model which
promotes a less energy-intensive way of life causing less or no environmental impacts. During
the past years, low carbon and renewable energy systems have been successfully deployed at
large scales, but in terms of exploiting their considerable potential and meeting the ambitious
carbon emissions reduction targets, these systems must also be applied at a smaller, local scale
(Chmiel and Bhattacharyya, 2015; Kaldellis et al., 2011).
Locally generated and distributed energy projects are now increasingly seen as a viable and
promising alternative to the traditional model and seem capable of delivering benefits that range
from increased security of supply for stakeholders to local economic benefits as well as reduced
environmental impacts. The provision of electricity to remote, rural communities is a challenging issue. The marginal cost of grid extension is greatly depended on the distance from the
existing electrical network, the size of the community and the electricity consumption profile.
Due to their isolated nature, energy supply in islands is more difficult and expensive (Rae and
Bradley, 2012; Kaldellis et al., 2009). Furthermore, the economies of scale associated with
large generation units are not applicable since even the loss of one unit may destabilize the entire power system. Additionally, the voltage and frequency control problems are usually more
complicated due to the limited interconnection potential with the systems from other islands or
the mainland; consequently, the probability of stability loss is higher. Finally, the overseas fuel
transportation demand raises significantly the cost of energy supply. Costly fossil fuel imports
ISBN: 978-960-98739-8-7
113
from distant locations have serious impacts on every island’s budget and prevent investments
for the socioeconomic development. The exploitation of local RES potential can reduce these
expensive imports and create important business and employment opportunities. Islands facing
similar challenges can benefit by pooling knowledge and optimum practices for successfully
overcoming these problems (Duic et al., 2008).
In recognition of this, Ministers and other participants from 48 countries held a conference in
St. Julian’s, Malta on 6-7 September 2012, and issued the Malta Communique on Accelerating
Renewable Energy Uptake for Islands. They called on IRENA to establish a Global Renewable
Energy Islands Network (GREIN) as a platform for pooling knowledge, sharing best practices,
and seeking innovative solutions for accelerated uptake of clean and cost-effective renewable
energy technologies on islands. This network will help to achieve the aspirations expressed in
the Barbados Declaration on Achieving Sustainable Energy for All in Small Island Developing
States (SIDS) 2012, outcomes of the Rio+20 Conference 2012, and the Abu Dhabi Communique on accelerating renewable energy uptake for the Pacific Islands 2012.
Off-grid renewable energy technologies have been implemented in islands in an attempt of
re-addressing the economic, social and environmental balance, while providing communities
with a regular supply of power as well. Many developing countries have been electrifying rural
areas and islands in this way for decades, acknowledging the advantages that such practices
can have over grid extension: e.g. reduced transmission losses, lower capital requirements,
lower operating and environmental costs, cheaper peak-time generation and more employment
opportunities for the local workforce. With respect to the integration of RES on islands, various technical aspects must be taken into consideration, such as the type of storage system that
should be provided. The system may be the continental power network itself for cases where
such an interconnection is feasible or the implementation of storage means as well. By 2007, a
European project called STORIES was developed with the objective of facilitating the penetration of RES in islands by changing the legislative and regulatory frameworks that were in force
until then, so that technologies related to energy storage can be implemented.
Green Islands in Europe: Best Practices
El Hierro (Canary Islands, Spain)
El Hierro, is the western, southernmost and smallest of the Canary Islands (an Autonomous
Community of Spain), in the Atlantic Ocean off the coast of Africa, with a population of almost
11000 inhabitants. Like all the Canary Islands, El Hierro is a touristic destination. In 2000, El
Hierro was designated by UNESCO as a Biosphere Reserve, with 60% of its territory protected
to preserve its natural and cultural diversity. Like the entire island complex, due to their size and
remoteness, El Hierro had to deal with energy problems, in terms of total dependency on fossil
fuels, high energy costs and often unreliable energy supply.
The energy demand of the island is about 42 GWh annually and it has been met by a 13.3 MW
diesel power plant located at Llanos Blancos on the western side of the Island connected to a
small distribution network of medium voltage. It was estimated that 6000 tn of diesel has been
consumed annually, thus over of 1.8 million Euros has been spent per year to meet the energy
needs and consequently, environmentally-wise, greenhouse gases emissions of 18700 tn of CO2,
100 tn of SO2 and 400 tn of NOx have been produced every year (Godina et al., 2015).
In 1997, El Hierro adopted a sustainable development plan to protect its environmental and
cultural abundance making the island a self-sustained location. The “El Hierro Hydro-Wind
Plant”, approved in 2002 and Gorona del Viento El Hierro S.A., a public-private partnership
was established for the implementation of the hybrid wind-pump hydro energy system (Kaldel-
114
ISBN: 978-960-98739-8-7
lis, 2002). The hybrid system comprises a wind farm of 11.5 MW total installed power, a 6 MW
pumping station, un upper reservoir of 380,000m3 capacity, hydroturbines 11.3MW, the lower
reservoir of 150,000m3 and penstocks.
By the summer of 2014, El Hierro’s wind-pump hydro power plant operated and achieved covering 25% of the electricity demand during peak hours in August. The goal set by Gorona for
the first year was to make sure that 80% of the energy provided annually to the grid would be
renewable (green energy).
Isle of Eigg (Scotland)
Isle of Eigg is the second largest island of the Small Isles Archipelago in the Scottish Inner Hebrides. It is located 20km off Scottish west coast, south of the Isle of Skye, being uneconomical
to connect to the national grid. The island is 9 km long from north to south and 5 km long from
east to west. It has almost 100 inhabitants with 38 households and 5 commercial properties. It
is reached by ferry line from Mallaig.
Since the 1st of February 2008, the island started covering its power demand through a unique
system comprising of RES, backed up by diesel generators. The electricity is distributed around
the island through an underground micro-grid system that supplies energy for 24h a day. Before
that the island did not have electricity supply and most residents used individual diesel generators, while few of them relied on a small hydroelectricity plant. Batteries/inverters have been
commonly used to ensure electricity access. The Isle of Eigg off-grid electrification system
shows that an off-grid system can support the electrical energy needs of a modern life style. The
residents of the island are enjoying a reliable supply of electricity that meets their requirements
effectively but more importantly, their carbon footprint has reduced considerably since about
90% of their electricity comes from RES. It is reported that the CO2 emission per household in
the island is 20% lower than the rest of the UK.
A carefully designed off-grid system can be an effective electrification option for any developing country. The islanders receive 24h supply and have no complains about the supply. This
demonstrates that an off-grid supply needs not to be a temporary or a pre-electrification option.
This is an important example to be used from other countries in their decision making process. From this and several other cases like that around the world, one may notice that off-grid
electrification is not just a temporary solution for the developing world. Furthermore a properly
designed off-grid system can be a viable alternative to grid extension in remote areas. With the
main part of electricity generated from RES and back-up systems being used occasionally, local
grids can successfully reduce carbon emissions from electricity generation. Demand management plays also an important role and requires active participation of the users.
Samsoe (Denmark)
Samsoe is a Danish island located in the centre of the country in the Kattegat Sea which covers a sea area of 30000 km². It is a sparsely populated island covering an area of 114 km² with
almost 3800 inhabitants and with population density of only 34/km². Its total coastline is 120
km with 28 km length and 7 km width.
In 1997, the authorities of the island decided to enter a national competition in order to become
a model renewable energy island. The program must have been implemented and completed
within 10 years and was aiming to reduce the carbon emissions. Before the implementation and
integration of the program, Samsoe electrification has been based on oil which has been brought
in the island through tankers, producing a significant amount of polluting substances, along with
electricity imports from coal-burning power plants via transmission lines from the mainland.
After the completion of the program the island was covering 100% of its electricity demands
ISBN: 978-960-98739-8-7
115
by RES and specifically by wind energy, while 75% of heating demands was generated by solar
and biomass energy. Consequently, the islanders achieved to reduce the carbon emissions and
produce surplus electricity, exporting it to the mainland (Jorgensen, and Nielsen, 2015).
Samsoe can be regarded as a model island which has proven that its project can be examined
and implemented in other off-grid cases, generating clean electrical energy and diminishing the
carbon emissions.
Utsira (Norway)
Utsira is a small Norwegian island located in the North Sea of Europe, in a distance of 18 km
from Haugesunt. It has approximately 200 residents, whose the primary occupation is fishing.
Concerning Utsira electrification, it was interconnected to the Norwegian power supply network through a 22kV cable which could reach a 1MW peak power supply until 2004. By this
year, “Hydro” company developed and established a sophisticated configuration in order to
cover the islands’ power demand through power acquired from two 600 kW wind turbines.
The interesting part of the configuration is the several different storage technologies that were
implemented in order to stabilize the wind turbines’ stochastic power supply. In fact, the establishment contains various technologies, such as hydrogen storage, chemical batteries and a
flywheel unit (Ulleberg et al., 2010).
This project has been installed in order to gain experience from the combination of those different technologies, record the quality of their performance and indicate potential threats and
malfunctions with the final goal of defining how such configurations can operate optimally in
order to support RES power plants. The most significant perhaps module of the project is the
implementation of a hydrogen module, thus an electrolyzer, a compressor, a high pressure storage cylinder, a fuel cell stack and a hydrogen combustion engine have been installed, a configuration that had never previously been evaluated in remote islands for such purposes.
The prospects of T.I.L.O.S. project (Greece)
Figure 1. The load profile of Tilos island where several blackouts have been encountered
Tilos is a small and isolated Greek island, located in the south-eastern Aegean Sea, lying midway between Kos and Rhodes. Up to now, the 780 inhabitants of the island covers their electricity needs through a poor interconnection to the host island of Kos, where a diesel-oil power
station is operating. The electricity network of Tilos faces often several black outs due to the
116
ISBN: 978-960-98739-8-7
undersea interconnection cable faults, Figure 1.
Recently, in the context of Horizon 2020 European Programme the implementation of an innovative project has been approved concerning the development of a smart microgrid, aiming
to meet the electricity demand of the Livadia community of Tilos island, which amounts about
500 residents. Fifteen European partners from seven different countries (Greece, Germany,
France, United Kingdom, Sweden, Italy and Spain) participate for the development, design and
implementation of Technology Innovation for the Local Scale, Optimum Integration of Battery
Energy Storage (T.I.L.O.S) project, under the coordination of the Soft Energy Applications and
Environmental Protection Laboratory of Piraeus University of Applied Sciences.
The smart microgrid is based on a hybrid power system which combines RES, i.e. wind turbines
and PV arrays with an advanced NaNiCl2 battery system, supported by demand management
strategies aiming to interact with the existing Nisyros-Kos electricity grid. The smart microgrid
of Tilos is the first project in the world which involves RES, energy storage and electricity interconnection systems and it can be also applicable to other Aegean or Mediterranean islands.
Conclusions
The present study describes the successful efforts of several island communities to cover their
energy needs on the basis of RES, replacing imported and heavy polluting fossil fuels. According to the official data El Hierro, Isle of Eigg, Samsoe and Utsira are the first European island
communities that have already exploiting the local RES potential contributing on establishing
a Green island model. In this context, there are more than 6,000 Greek islands and islets, 227
of which are inhabited. The majority of these islands are located in the Aegean Sea possessing
excellent solar and wind potential. This is the case of Tilos island, where the first Greek Green
island concept is developing with the financial support of EU. According to the results obtained,
the proposed RES-based solution is able to satisfy the electricity needs of the island exporting
at the same time any energy surplus to the Kos electrical grid. On top of that, a preliminary
economic evaluation undertaken demonstrates that the investment is viable in comparison to
the current situation especially from the local community point of view.
References
1. Chmiel,Z., Bhattacharyya,S.C., (2015). Analysis of off-grid electricity system at Isle of Eigg (Scotland):
Lessons for developing countries. Renewable Energy, 81, pp.578-588.
2. Duic, N., Krajacic, G., Carvalho, M., (2008). RenewIslands methodology for sustainable energy and resource
planning for islands. Renewable and Sustainable Energy Reviews, 12, pp. 1032–1062.
3. Godina, R., Rodrigues, E.M.G., Matias, J.C.O., Catalão, J.P.S., (2015). Sustainable Energy System of El
Hierro Island. International Conference on Renewable Energies and Power Quality (ICREPQ’15) La Coruña
(Spain), 25-27 March, 2015, Renewable Energy and Power Quality Journal (RE&PQJ), ISSN 2172-038 X,
No.13.
4. Jorgensen, S.E., Nielsen, S.N., (2015). A carbon cycling model developed for the renewable Energy Danish
Island, Samsoe. Ecological Modelling, 306, pp. 106-120.
5. Kaldellis J.K. (2002). Parametrical Investigation of the Wind-Hydro Electricity Production Solution for
Aegean Archipelago. Journal of Energy Conversion and Management, 43(16), pp.2097-2113.
6. Kaldellis J.K., Kapsali M., Tiligadas D. (2011. Presentation of a Stochastic Model Estimating the Wind Energy
Contribution in Remote Island Electrical Networks. Applied Energy, 97, pp.68-76.
7. Kaldellis J.K., Zafirakis D., Kaldelli El., Kondili E. (2009). Combined Photovoltaic and Energy Storage
Systems. An Integrated Electrification Solution for Small Islands. International Journal of Environmental
Technology & Management, 10(2), pp.123-149.
8. Rae, C., Bradley, F., (2012). Energy autonomy in sustainable communities - A Review of key issues. Renewable
and Sustainable Energy Reviews, 16, pp. 6497–6506.
9. Ulleberg, O. Nakken, T. Ete, A. (2010). The wind/hydrogen demonstration system at Utsira in Norway:
Evaluation of system performance using operational data and updated hydrogen energy system modelling
tools. Elsevier. International Journal of Hydrogen Energy, 35, pp. 1841-1852.
ISBN: 978-960-98739-8-7
117
Electromobility in EU: Current status and future prospects for the
Greek market
J.K. Kaldellis*, St. Liaros, G. Spyropoulos
Soft Energy Applications & Environmental Protection Laboratory
Piraeus University of Applied Sciences, P.O. Box 41046, Athens, 12201, Greece, http://
www.sealab.gr, *Corresponding author: [email protected]
Keywords: Electric Vehicles, Photovoltaic, Urban Transport, Total Cost of Ownership,
Car Market, Greece.
Abstract
Passenger and light duty vehicles market is currently experiencing a booming growth, primarily
caused by the rapidly rising population and the emerging economies of developing countries as
long as the affordability caused by the technological advancement and modern manufacturing
techniques. At present, the vehicle’s market dominant technology is that of Internal Combustion Engines (ICE) which are related to excessive Greenhouse Gas (GHG) emissions and air
pollution. With over a billion ICE automobiles currently in use globally, the transport sector is
considered to be excessively environmental harmful, being directly linked to pollution, global
warming and millions of premature deaths annually.
As of late, Electric-Vehicles (EV) are promoted around the world as an environmental friendly
mean of transportation and are widely recognized as an approach for reducing the impact of
transport sector to our surroundings.
The present work is bound to contribute in the research for the market development of alternatively fueled vehicles. More specifically by means of providing the total cost of ownership of
coupling electric vehicles with solar assisted chargers in the Greek market, will allow the local
decision makers to better understand and evaluate the present and future market prospect of
such technologies, under the existing economic situation.
Introduction
The continuing and steadily growing dependency of EU from imported fossil fuels and the
ongoing environmental degradation, despite the significant efforts of the last thirty years, accentuate the necessity for additional measures being taken in order to decelerate the depletion
of fossil fuel reserves, secure undisturbed energy supply and safeguard our surroundings from
human induced activities.
With the least renewable energy sources (RES) representation, notable is the contribution of the
transport sector in the overall final energy consumption and GHG emissions of the EU-28 with
values approximating 30% and 20% respectively. In this context, so as to achieve energy supply
security and sustainable development, radical changes are needed to be made in the transport
sector and especially in the road transport sector which contributes by more than 70% of the
entire sector’s carbon dioxide emissions (EC, 2014b).
Since 2001, through various policies and legislative measures, the average carbon dioxide
(CO2) emissions of newly introduced passenger cars in the EU declined by approximately 27%
in 2014 to approximately 123g/km, while targets for further reducing the emissions to 95g/
km till 2020/21 are set (EC, 2009, 2011, 2014a, 2015). Unfortunately, recent studies revealed
that the gap between official and real world emissions increased to approximately 40% in 2014
(Uwe Tietge et al., 2015).
118
ISBN: 978-960-98739-8-7
Currently alternative powered vehicles are gaining much of attention from researchers, policy
makers, manufacturers and the general public, demonstrated by the plethora of newly published
studies and articles. Electric vehicles in particular are expected to significantly contribute in
the decarbonization of the road transportation sector, given the fact that their refueling will be
increasingly dependent on RES.
The economic viability and possible market penetration of electric vehicles is affected by regionally dependent financial, environmental, energy, technological, policy and demographic
factors. Although many of the technological issues are progressively being solved and the alternative fueled vehicles are expended in the long run to be competitive with conventional ones,
their adoptability is still hindered by higher prices and limited refueling infrastructures.
The present work is bound to contribute in the research for the market development of alternatively fueled vehicles. More specifically by means of providing the total cost of ownership of
coupling electric vehicles with solar assisted chargers in the Greek market, will allow the local
decision makers to better understand and evaluate the present and future market prospect of
such technologies, under the existing economic situation.
The European EV market
According to available data by the end of 2014 over 740,000 pure battery (BEV) or Plug-In
Hybrid (PHEV) vehicles were circulating on the roads worldwide, almost doubling their number since the end of 2013 (ZSW, 2015). After the first semester of 2015, during September, the
circulation of just over 1 million BEVs (62%) and PHEVs (38%) were achieved worldwide.
The top 10 countries in registered EVs covered more than 93% of the corresponding market.
In USA there were registered 363,265 units, in China 157,354, in Japan 121,000, in Norway
65,958, in Netherlands 61,025, in France 59,000, in UK 39,616, in Germany 38,154, in Canada
14,429 and in Sweden 12,786 (Ayre, 2015).
The two major obstacles to market expansion of EVs are considered to be the, relatively to
conventional vehicles, high initial and total cost of ownership (TCO) and the limited amount
of charging facilities. The primary cost factor of electric vehicles is that of energy storage. Battery bank cost, in many cases surpasses 50% of the initial cost of the vehicle, while the cost of
replacement due to mismatch of the battery bank’s and vehicle’s operational life needs always
to be considered. According to IEA (2013), in order for the EVs to be marginally competitive
in terms of lifecycle assessment, the battery bank cost needs to approximate values near $300/
kWh. Alongside, recent research data highlighted the fact that within seven years from 2007
to 2014 the battery cost reduced from $1000/kWh to $400/kWh, while the current cost of the
market leading brand model’s are near or below the limit of $300/kWh (Nykvist and Nilsson,
2015).
In a similar manner the work of the investment bank UBS brings the TCO of a Tesla Model S
60kWh is similar to that of an Audi A7 3.0 TFSI, in an economy such as Germany’s with high
priced energy and non-existent purchase incentives, while the battery bank cost approximating
$250/kWh (Parkinson, 2014). Actually, it is not considered unlikely by 2020 the battery cost
to fall below the threshold of $150/kWh which under these conditions will lead to more than 5
million EVs sold by 2025 with leading markets those of Europe and China (Romm, 2015).
The proposed solution
Bearing in mind, the firm decision of EU for sustainable development and energy autonomy,
the support for further development of RES-based applications is a strategic decision towards
2030. Since the transportation sector is by far, the one with the minimum RES participation,
ISBN: 978-960-98739-8-7
119
the significant encouragement of clean EVs is the most promising solution for the private road
transportation.
Hitherto, and accounting for available data (ARF, 2014), in Europe there are more than 15000
non-residential slow EV Charging Stations (EVCS) and nearly 1000 fast ones. Unfortunately,
one cannot get reliable data about residential charging infrastructures. While, in this context,
the representation of solar powered EVCSs is only minimal.
However, for the effective support of the EVs, promotion and expansion of an integrated and reliable charging network is necessary. For this purpose, we propose the adoption of solar-based,
instead of simple grid connected, EVCSs, in order to maximize the environmental benefits and
to minimize fuel imports. Undeniably, the initial cost required is acting as a serious obstacle at
present. However, the gradual cost diminution of the EVCSs’ components and the economies
of scale fashioned by massively installing EVCSs, are expected to reduce the initial installation capital by as much as 50% in the near future. In light of the excellent solar potential of
the Mediterranean region and the volatility of the international oil and gas prices the proposed
solar-based configuration is, expressing the authors’ belief, the optimum solution.
Figure 1. The proposed Solar Electrical Vehicle Charging Station (CARPORT) at the SEALAB of Piraeus
University of Applied Sciences, general view
The Laboratory of Soft Energy Applications and Environmental Protection (SEALAB) of Piraeus University of Applied Sciences (former TEI of Piraeus) has recently undertaken, in the
framework of its innovative activities, the development, construction and operation of the first
standalone Solar EV Charging Station in the country, “CARPORT”, monitoring all energy data
and supporting in this way the country’s effort towards the infrastructure development and
strengthening in the field of electro-mobility, Figure 1.
More specifically this innovative effort described in the present paper is targeting to accelerate the implementation of a European-national electrification action plan for the construction
of EV charging stations based on PV generators. The proposed Solar EV Charging Station is
one of the most environmentally friendly solutions, able to support an EV fleet market all over
Europe.
Methodology
The total cost of ownership (TCO) is a financial estimate that incorporates the direct and indirect costs of product, a choice or an investment. TCO acts as a powerful decision tool mainly in
the professional rather than in scientific context. As for BEVs, it is widely acknowledged that
those outperform the conventional vehicles from an environmental standpoint, but their initial
cost might be discouraging for new potential owners. On the contrary many studies have shown
that BEVs do not fall way behind conventional vehicles in terms of complete direct and indirect
costs for the consumer (Wu, Inderbitzin and Bening, 2015).
120
ISBN: 978-960-98739-8-7
The TCO is calculated as follows:
TCO = Vehicle and infrastructure Initial Cost + Sum of Annual Costs‘ Present Value Present Resale Value – Subsidization (1)
The vehicles that are compared in this study are BEV, diesel and petrol ICEV. The initial cost
of the vehicle and the supporting infrastructure have been estimated via market research and
communication with local retailers. The resale price, due to the fact that there is not yet an active resale market for BEVs, has been taken by bibliographic sources (Wu, Inderbitzin and Bening, 2015). The annual killometers travelled (AKT) have been averaged over various sources
(AFDC, 2015; Eurostat, 2015; FHWA, 2015) of which, arbitrarily chosen, 55% are in stop and
go traffic. The AKT will be subjected to sensitivity analysis. Additionally, supporting infrastructures of the BEV included the installation of a PV array and the interconnection to the grid
via Net-metering, the vehicle charger and spare cables. Arbitrarily, the vehicle holding of first
owner has been chosen to be 10 years.
Annual costs consist of the fuel cost, insurance and maintenance cost, repair and taxation cost,
as well as possible battery replacement cost. Battery and fuel tank size of vehicles as well as
range and consumption have been obtained via market research. The operational and calendar
life of the battery bank was taken to be 160K km and 15 years respectively, while annually a
7% declination of battery pack prices is expected. During the calculation of the present value of
annual costs fuel escalation rates, diminishing factors for the utilization of the vehicle and real
world uplift factors have been considered, while the inflation rate has been taken to be 2.9%,
the market interest rate 7.3% and the corresponding discount rate has been assumed equal to
12.5%, value also subjected to sensitivity analysis.
Following the previously analyzed methodology the TCO and TCO/km of various powertrain
options have been estimated.
As can be seen from Figure 2, although the initial cost of an EV is significantly increased in
comparison with other powertrain options, the MR&O and annual operating costs are considerably lower.
Figure 2. Initial Cost, Annual Maintenance/Repair/Operation and Fuel cost breakdown
The TCO and TCO/km has been found to be 28000€ and 2.14€/km for BEVs, 31000€ and
2.35€/km for compression ignition (CI) ICEVs and 34000€ and 2.58€/km for spark ignition
(SI) ICEVs. By introducing the option of solar assisted charging via a PV system with Netmetering the same values changed to 36500€ or 2.78€/km.
BEVs as an option are becoming the most efficient powertrain technology as long as the first
owner holds the vehicle for at least 5 years. In the case of PV coupling, the BEV option is only
ISBN: 978-960-98739-8-7
121
becoming an efficient choice only in comparison with ICEV-SI and with the vehicle hold by the
first owner for at least 15 years.
AKT has been found to be a crucial forming factor of the TCO and TOC/km, which has also
been suggested by Wu et al (2015). As can be seen from Figure 3, BEVs are becoming the most
attractive option as long as the user travels at least 8000 km. At 5000 km BEVs are already a superior option than ICEV-SI. In the case of coupling the charging facility with a PV Net-metering
system the BEVs are becoming marginally a better option at approximately 30000km driven
annually, while even at 60000km driven the soundest option would be that of an ICEV-CI. In
any case the discount rate does not seem to be of major importance for that particular study.
Figure 3. TCO per km vs. annual km (AKT) traveled by an EV
Concluding Remarks
This study aimed to analyze the present status of BEVs and how those are currently compared
to local market leading technologies, namely diesel and petrol ICEV. Furthermore, the option
of coupling the BEV charging facility to a PV Net-metering system is also studied. In order to
achieve this, the TCO and TCO/km for different vehicles options are calculated. Findings show
that currently BEVs, according to TCO/km, may present competitive advantage in comparison
with other technologies as long as annual kilometers traveled are exceeding 8000. Utilizing a
solar charging system may be also considered a viable option as long as the user is utilizing
the vehicle for more than 15 years, travelling more than 13000 AKT, in comparison with pure
petrol engines. The present study validates the fact that even in an unsteady economy such as
Greece’s, BEVs are already becoming a viable option for the local market. Coupling of BEV’s
with RES is considered to be one of the major options for the decarbonization of the transportation sector. In order the proposed solution to be a viable option for the majority of the local
people, some kind of incentives to support the market are necessary, along with the encouragement of installing more PV-based EV Charging Stations.
References
1. AFDC (2015) Average annual vehicle miles traveled of major vehicle categories, Alternative Fuels Data
Center. Available at: http://www.afdc.energy.gov/data/ (Accessed: 21 October 2015).
2. ARF (2014) Electric vehicles in Europe: Gearing up for a new phase? Amstedam: Amsterdam Roundtable
Foundation. Available at: http://www.mckinsey.com/global_locations/europe_and_middleeast/netherlands/
en/latest_thinking (Accessed: 12 August 2015).
122
ISBN: 978-960-98739-8-7
3. Ayre, J. (2015) One Million EVs Have Been Sold Worldwide To Date! −. Available at: http://evobsession.
com/one-million-evsphevs-sold-worldwide-date/?utm_source=EVO&utm_medium=email&utm_
campaign=371c3042c7-RSS_EMAIL_CAMPAIGN&utm_term=0_a84282aa83-371c3042c7-132335533
(Accessed: 23 September 2015).
4. EC (2009) Regulation (EC) No 443/2009 of the European Parliament and of the Council of 23 April 2009
setting emission performance standards for new passenger cars as part of the Community’s integrated
approach to reduce CO 2 emissions from light-duty vehicles (Text with EEA relevance), OJ L 140. Available
at: http://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex:32009R0443 (Accessed: 26 April 2015).
5. EC (2011) Regulation (EU) No 510/2011 of the European Parliament and of the Council of 11 May 2011
setting emission performance standards for new light commercial vehicles as part of the Union’s integrated
approach to reduce CO 2 emissions from light-duty vehicles (Text with EEA relevance), OJ L 145. Available
at: http://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32011R0510 (Accessed: 26 April 2015).
6. EC (2014a) Commission Delegated Regulation (EU) No 404/2014 of 17 February 2014 amending Annex II
to Regulation (EU) No 510/2011 of the European Parliament and of the Council as regards the monitoring
of CO 2 emissions from new light commercial vehicles type-approved in a multi-stage process Text with EEA
relevance, OJ L 121. Available at: http://eur-lex.europa.eu/legal-content/EN/TXT/?uri=uriserv:OJ.L_.2014.1
21.01.0001.01.ENG (Accessed: 26 April 2015).
7. EC (2014b) EU Energy in Figures - Statistical Pocketbook 2014. European Commission.
8. EC (2015) Reducing CO2 emissions from passenger cars - European Commission. Available at: http://
ec.europa.eu/clima/policies/transport/vehicles/cars/index_en.htm (Accessed: 8 May 2015).
9. Eurostat (2015) Passenger transport statistics - Statistics Explained, Passenger transport statistics. Available
at: http://ec.europa.eu/eurostat/statistics-explained/index.php/Passenger_transport_statistics (Accessed: 21
October 2015).
10. FHWA (2015) Average Annual Miles per Driver by Age Group, U.S. Department of Transportation Federal
Highway Administration. Available at: https://www.fhwa.dot.gov/ohim/onh00/bar8.htm (Accessed: 21
October 2015).
11. IEA (2013) Global EV Outlook. International Energy Agency. Available at: http://www.iea.org/publications/
freepublications/publication/global-ev-outlook.html (Accessed: 26 April 2015).
12. Nykvist, B. and Nilsson, M. (2015) ‘Rapidly falling costs of battery packs for electric vehicles’, Nature
Climate Change, 5(4), pp. 329–332. doi: 10.1038/nclimate2564.
13. Parkinson, G. (2014) ‘Why EVs will make solar viable without subsidies’, Renew Economy, 21 August.
Available at: http://reneweconomy.com.au/2014/why-evs-will-make-solar-viable-without-subsidies-91738
(Accessed: 23 September 2015).
14. Romm, J. (2015) ‘Electric Car Batteries Just Hit A Key Price Point’, ThinkProgress, 13 April. Available at:
http://thinkprogress.org/climate/2015/04/13/3646004/electric-car-batteries-price/ (Accessed: 23 September
2015).
15. Uwe Tietge, Nikiforos Zacharof, Peter Mock, Vicente Franco, John German, Anup Bandivadekar, Norbert
Ligterink and Udo Lambrecht (2015) FROM LABORATORY TO ROAD A 2015 update of official and ‘realworld’ fuel consumption and CO2 values for passenger cars in Europe. The International Council on Clean
Transportation. Available at: http://www.theicct.org/laboratory-road-2015-update (Accessed: 28 September
2015).
16. Wu, G., Inderbitzin, A. and Bening, C. (2015) ‘Total cost of ownership of electric vehicles compared to
conventional vehicles: A probabilistic analysis and projection across market segments’, Energy Policy, 80, pp.
196–214. doi: 10.1016/j.enpol.2015.02.004.
17. ZSW (2015) ZSW: More than 740,000 Cars Worldwide Powered by Electricity, Centre for Solar Energy
and Hydrogen Research. Available at: http://www.zsw-bw.de/en/support/news/news-detail/mehr-als-740000autos-weltweit-fahren-mit-strom.html (Accessed: 12 August 2015).
ISBN: 978-960-98739-8-7
123
Application of recurrence quantification analysis in wind time series
from wind farms
A.Κ. Charakopoulos1, T.E. Karakasidis1*, and I. Sarris2
1
2
Laboratory of Hydromechanics and Environmental Engineering, Department of Civil
Engineering, University of Thessaly, 38334 Volos, Greece, *[email protected]
Department of Energy Technology, Technological Educational Institute of Athens
Keywords: wind farm, wind velocity, recurrence plot, recurrence quantification
analysis.
Abstract
In the present study, we propose a new methodology to study the complex profile of wind data
time series measured at towers by cup anemometers, using Recurrence Plots (RP) and Recurrence Quantification Analysis (RQA). More specifically the RP of wind velocity and angle are
calculated in order to extract statistically significant features of RP patterns. Such measures
include quantities such as recurrence rate, determinism, laminarity and trapping time. The results show that the RP method and the corresponding RQA method allows distinguishing and
identifying dynamical transitions that occur during the system evolution. Thus the proposed
methodology can play a significant complementary role to the existing methods of wind data
analysis and prediction.
Introduction
Renewable energy sources include replenishable natural resources such as solar energy, geothermal, biofuels, ocean and wind power have an increasing interest in recent years (Carrasco,
J. M. et al. 2006). Wind power is an attractive alternative of renewable energy sources due to
its vast potential in power generation, availability, while being quite cost competitive (DíazGonzález, F. et al. 2012). Over the last decades, wind power generation capacity has been increasing due to the development of offshore wind farms (Muyeen, S. M. et al. 2010). The wind
farm site location is very important and it remains a challenging engineering problem based on
wind speed profile in the field of generating wind energy. The majority of research in the literature is focused on predicting and analyzing the wind speed using conventional weather prediction models (AR, ARAM, ARIMA, ANNs, SVM) (Soman, S. S. et al. 2010) which in general
are characterized by the simplicity of model construction, reduced computational requirement,
and accurate results for short-term prediction. However, in general they do not take into account
nonlinearities that may exist in the physical process.
In the present paper we investigate the understanding of the wind profile, velocity and direction,
and to identification of temporal windows (regions) with the same dynamical characteristics
by applying a powerful nonlinear analysis technique, the recurrence plot (RP) (Marwan, N. et
al. 2007), and the statistical features of RP known as recurrence quantification analysis (RQA)
(Webber and Zbilut, 1994). More specifically we concentrate in to investigate if one can identify hidden patterns and dynamical transition during the evolution of the wind time series.
Our analysis is based on, the recurrence plot (RP) technique and the recurrence quantification
analysis (RQA) applied to the time series of wind speed (polar velocity) and direction (polar angle) treated as separated time series. Our findings suggest that recurrence quantification
analysis reveals the underlying time series dynamics and we can exhibit characteristic patterns
which are caused by the typical dynamical behavior.
124
ISBN: 978-960-98739-8-7
The structure of the paper is as follows. In Sec. 2 we describe the data set. In Sec. 3 we present
the recurrence plot and recurrence quantification methodology employed for data analysis. The
results and discussion are presented in Sec. 4. Finally the conclusions are presented in Sec. 5.
Data
The data consist of the average value every ten minutes collected by cup anemometers located
in a measurement tower where the height of the tower is up to the hub height of the planned
wind turbines. The measurements correspond to data from a tower installed in the mountains of
the region Achaia, Peloponnesus, Greece, in the form of polar velocity and angle. Wind speed
is measured by cup anemometers. The data set (Fig.1) consists of 38540 values from months
of March to December 2005 and stored as 10 minute averages. The data recorded are the maximum/minimum wind speed, the mean polar speed, the standard deviation of the mean velocity,
the mean angles and the standard deviation of the angles.
30
m/s
20
10
0
0
0.5
1
1.5
2
time
2.5
2
time
2.5
3
4
3.5
4
x 10
degrees
400
300
200
100
0
0
0.5
1
1.5
3
4
3.5
4
x 10
Figure 1. Time series of a) mean polar velocity in m/s and b) mean angle (degrees)
Recurrence plots and Recurrence quantification analysis
Recurrence Plots is a graphical method to characterize the system behavior in phase space and
introduced in the late 1980’s by Eckmann et al. (1987) in order to detect nonlinearities and
chaotic dynamics in experimental data. Using the RP technique, we first reconstruct the phase
space of the time series x(i) and then we embed the time series into a m-dimensional space with
the time delay (τ). After the state space reconstruction, the main step is the calculation of the
NxN recurrence distance matrix using the following equation
(1)
is the norm (e.g., the Euwhere ε is the cutoff distance, Θ(x) is the Heaviside function and
clidean norm). The elements of distance matrix are the distances between all possible combinations of i-points and j-points. This means that if two phase space vectors xi and xj are sufficiently
close, then Ri,j=1 otherwise Ri,j=0. The RP is obtained by plotting the recurrence matrix. If the
distance between xi and xj is less than ε, then a dot is placed at (i,j) in the RP.
Furthermore, the RP patterns provide evidence for determinism and periodicity in the dynamic
system. As the diagonal lines parallel to the main diagonal line of the RP occur when segments
of the trajectory visit the same region of the phase space at distinct time, the length of these
lines is evidence for periodical signal. While vertical or horizontal lines suggest that the system
remains stationary for a certain time. Also, in the case of isolated recurrence point in the RP,
may indicate that the system are rare with little persistency or large fluctuation.
ISBN: 978-960-98739-8-7
125
The information of the dynamic system can be extracted and quantified using recurrence quantification analysis (RQA) introduced by Zbilut and Webber (1994) and extended by Marwan et
al. (2007). The RQA measures the recurrence point density and the diagonal and vertical line
structures of the RP. We used the Recurrence Rate (RR), the Determinism (DET), the Longest
diagonal line Lmax, the Laminarity (LAM) and the Trapping time (TT). The distribution of
these variables as a function of time is computed considering a moving window approach which
allows detecting temporal changes in the dynamics.
We analyzed the mean polar velocity and the mean polar angle of the wind data. Each time
series consist of 38.540 value recorder by a cup anemometer located at a measurement tower.
The time series recorded at the same time from March to December 2005. In Fig. 2 (a) - (b) one
can see the time series along with the corresponding RPs for the mean velocity and mean angle
respectively. We can see that there are several distinct RP pattern structures.
Figure 2. Time series and global RP of a) mean polar velocity and b) mean angle (degrees)
In Fig. 3(a) we can see that the RP of mean velocity is characterized by successive black and
white rectangles, where some of these regions form some small and some larger square-like
patterns which corresponds to different regions (A through J) and distinguished using different
colors. The same regions are indicated also on the corresponding time series. The RP of mean
angle Fig. 3 (b) presents only small rectangles and is obviously less dense than that of mean
angle indicating more frequent and larger fluctuations in the process.
The RQA allows the quantification of the complex structure of the RP. In Fig. 4 we present the
recurrence quantification analysis (RQA) measures for the mean polar velocity. The distribution
of these variables as a function of time is computed considering a moving window approach
which allows detecting temporal changes in the dynamics.
126
ISBN: 978-960-98739-8-7
a
b
Figure 3. Time series and global RP of a) mean polar velocity and b) mean angle (degrees)
Figure 4. RQA results for mean polar velocity
The results of recurrence features for each epoch reveals information about the changes of
system states it is possible to identify and quantify such transitions. We have demonstrated that
the RP structures contain useful information which can characterize distinct region to the time
series. Such a detailed analysis will be part of a future work.
References
1. Carrasco, J. M., Franquelo, L. G., Bialasiewicz, J. T., Galván, E., Guisado, R. C. P., Prats, M. Á. M., .&
Moreno-Alfonso, N. (2006). Power-electronic systems for the grid integration of renewable energy sources: A
survey.Industrial Electronics, IEEE Transactions on, 53(4), 1002-1016.
2. Díaz-González, F., Sumper, A., Gomis-Bellmunt, O., & Villafáfila-Robles, R. (2012). A review of energy
storage technologies for wind power applications.Renewable and Sustainable Energy Reviews, 16(4), 21542171.
3. Eckmann, J. P., Kamphorst, S. O., & Ruelle, D. (1987). Recurrence plots of dynamical systems. Europhys.
Lett, 4(9), 973-977.
4. Marwan, N., Romano, M. C., Thiel, M., & Kurths, J. (2007). Recurrence plots for the analysis of complex
systems. Physics Reports, 438(5), 237-329.
5. Muyeen, S. M., Takahashi, R., Murata, T., & Tamura, J. (2010). A variable speed wind turbine control strategy
to meet wind farm grid code requirements.Power Systems, IEEE Transactions on, 25(1), 331-340.
6. Soman, S. S., Zareipour, H., Malik, O., & Mandal, P. (2010, September). A review of wind power and wind
speed forecasting methods with different time horizons. In North American Power Symposium (NAPS),
2010 (pp. 1-8). IEEE.
7. Webber, C. L., & Zbilut, J. P. (1994). Dynamical assessment of physiological systems and states using
recurrence plot strategies. Journal of applied physiology, 76(2), 965-973.
ISBN: 978-960-98739-8-7
127
Non-linear Time series Methods Applications on Transport Data
A.D.Fragkou1, T.E. Karakasidis1* and E. Nathanail2
1
2
Laboratory of Hydromechanics and Environmental Engineering, *email: [email protected]
Transportation Engineering Laboratory
Department of Civil Engineering, University of Thessaly, 38334 Volos, Greece.
Keywords: Traffic Data Analysis, Non-linear time series Analysis, Recurrence
Quantification Analysis.
Abstract
In the present study we present results of the application of nonlinear time series analysis on
traffic data. More specifically we analyze records of one day values of Attiki Odos collected
from sensors at several measurement nodes. The data are analyzed using the Recurrence Quantification Analysis (RQA) in order to detect traffic incidents. The analysis indicates that RQA
can contribute in the detection of traffic incidents by locating abrupt changes of system’s dynamics through RQA measures such as recurrence, determinism, maxline, laminarity, and trapping time.
Introduction
Traffic data is analyzed with the nonlinear method of Recurrence plots. Especially along Attiki
Odos we placed sensors at nodes. Every sensor counted every 20 seconds during one day (24
hours) from 00:00 to 24:00. So every sensor took 4320 records of vehicles volume (number of
vehicles passing over every sensor per hour). We consentrate on node with VDS ID 3557 (VDS
name 28.7) which is placed close to the incident (200 meters far) (Table 1).
DATE –
POSITION
(Km)
STARTING
TIME
OF THE
INCIDENT
STARTING
TIME OF
THE LINE
STOPPING
TIME OF THE
INCIDENT
STOPPING
TIME OF
THE LINE
MAXIMUM
LINE
LENGTH
(m)
VDS
NAME
VDS
ID
01/03/2010 A
28,5
12:35
12:38
13:19
13:12
2600
VDS A
28.70
3557
Table 1. Main characteristics (position, date, exact time of starting and stopping of the incident and the line
length) at 3557 measurment node placed on Attiki Odos.
Timeseries (Fig. 1) of 4320 records, firstly embedded with the method of False Nearest Neighbors (Kantz, 2004) after finding the proper time delay with the help of the nonlinear method of
Average Mutual Information (Kennel, 1992). In order to apply the method of Recurrence Plots
(RP) and Recurrence Quantification Analysis (RQA), the embedding parameters were embedding dimension m=9 and time delay τ=1.
128
ISBN: 978-960-98739-8-7
VOLUME
0.5
VOLUME
0.4
0.3
0.2
0.1
0
0
500 1000 1500 2000 2500 3000 3500 4000
TIME
Figure 1. Time series of node 3557 (volume).
The RP method is a graphical tool introduced by Eckmann et al. (Eckmann, 1987) in order to
extract qualitative characteristics of a time series. The method relies on the reconstruction of
the underlying system phase space. For the construction of an RP we first embed the time series
in an m-dimensional space by forming a sequence of vectors si with the appropriate selected
time delay. Then we calculate the distance dij=||si‑sj|| between points i, j in the m-dimensional
reconstructed phase space. As a result, a NxN matrix of distances dij, is obtained, N being the
number of reconstructed state vectors. From this distance matrix a two-dimensional plot is created by darkening the pixel located at (i, j) if the distance between points i and j is lower than a
given cut-off value ri.
Patterns on the RP gives lots of information about the dynamics of the system. Diagonal parallel
lines gives the information of a high deterministic system, while isolated points inform us that
the system acts chaotical. An interesting pattern is white lines vertical and horizontal. This kind
of pattern reveal abrupt changes in the dynamics of the system.
Webber and Zbilut in 1992 quantify RPs to give a more clear view of system dynamics (Webber, 1992). Recurrence Quantification analysis includes parameters which based on the length
of the lines a Recurrence plot contains.
Recurrence Rate (RR) gives the ratio of the number of recurrent points (pixels) to the total
number of points (pixels) of the plot, determinism is the ratio of the number of recurrent points
forming upward diagonal lines to the total number of recurrent points, maxline is the length of
the longest diagonal line segment in the plot, excluding the main diagonal line of identity laminarity is the ratio of the number of recurrent points forming horizontal lines to the total number
of recurrent points and trapping time shows the average length of the vertical lines. High values
of Recurrence, determinism and maxline gives the information of periodic states. System states
are trapped in time, when high values of laminarity and trapping time appears. Values of all
those parameters are become very low when an abrupt change appear in the systems dynamics.
Moreover, systems dynamics may pass over phase transitions. RQA method (Marwan, 2007)
can trace those phase transitions by separate the global analysis into equal time spaces (epoqs).
From this can be separated periodic from laminal from chaotic states which is very useful to
firstly understand the dynamics of the system in order to proceed to more detailed analysis.
ISBN: 978-960-98739-8-7
129
After the proper embedding the recurrence plot of node 3557 is presented (Fig. 1).
Figure 1. a) Recurrence Plot of node 3557, b) Recurrence Plot of node 3557 with phase transitions (red squares)
With a global inspection we can see black patterns at (0-900) and (3400-4320). White bands
make their appearance and also white lines (horizontal and vertical). After this inspection we
can infer that there are changes at system dynamics (white bands) and for some times the system changes abruptly (white lines), althought it is trapped for some period in time (black patterns). Phase transitions are viewed along the main diagonal from Periodic (0-900) to chaotic
(900-1700), to laminar (1700-1900), to chaotic (1900-2600), to laminar (2600-3100), to chaotic
(3100-3400, to periodic (3400-4320). The quantification of the RP with the method of RQA
with epoqs give us more detailed results about phase transitions. We consider 180 epoqs (one
hour each) and recurrence Quantification Analysis applied on each epoq (Fig 2).
Figure 2. a) Recurrence (RR), b) Determinism (DET), are the RQA parameters of node 3557 of Recurrence
Quantification Analysis.
We focus on the analysis of Recurrence and Determinism parameters. From the above diagrams we observe at (0-900) and (3400-4320) high values of RR and DET (Periodic phases).
This is the morning and late at night hours when the traffic is not so dense. This changes at the
rush hour when we get into the chaotic phase (900-1700). At this phase RR and DET fluctu130
ISBN: 978-960-98739-8-7
ate abruptly. After this phase, when the rush hour passes the system seem to balance at region
(1700-1900) where values of RR and DET are big. This stops when RR and DET values again
start to abruptly fluctuate at chaotic region (1900-2600). This behavior (phase transition from
laminar to chaotic) is being repeated at regions (2600-3100) and (3100-3400) where from laminar region goes to chaotic region. This phase transition give us the information that our system
is going through many dynamical changes.
Concluding Remarks
The methods of Recurrence Plot and Recurrence Quantification analysis with epoqs are appropriate methods to capture phase transitions in the dynamics of a system. Specifically the
application of this method to systems such as traffic data gives us a lot of useful information
about the volume of cars during a day at big not urban streets. Especially if those kind of systems come through many phase transitions may be very helpful to localize incidents that may
cause those transitions.
References
1. Eckmann J.-P., Kamphorst S. O., Ruelle D., 1987, Recurrence plots of dynamical systems, Europhysics
Letters, 4(9), 973-977.
2. Kantz H., Schreiber T., 2004, Nonlinear Time Series Analysis (Cambridge University Press, second edition).
3. Kennel M. B., Brown, R., Abarbanel H. D. I., 1992, Determining Embedding Dimension For Phase-Space
Reconstruction Using A Geometrical Construction, Physical Review A, 45(6), 3403-3411.
4. Marwan N, Romano M.C, Thiel M., Kurths J., 2007, Recurrence plots for the analysis of complex systems,
Physics Reports, 438, 237-329,
5. Zbilut J., Webber C. Jr, 1992, Embeddings and delays as derived from quantification of recurrence plots,
Physics Letters A, 171(3-4), 199-203.
ISBN: 978-960-98739-8-7
131
Effect of carbon nanotubes addition on the mechanical performance
of co-polyimide hollow fiber membranes
Filippos D. Gegitsidis1, Nikolaos D. Alexopoulos1*, Evangelos P. Favvas2
and Stavros K. Kourkoulis3
Department of Financial Engineering, University of the Aegean, Chios, Greece,
*Corresponding author’s email address: [email protected]
2
Department of Physical Chemistry, NCSR “Demokritos”, Athens, Greece
3
Laboratory for Testing and Materials, School of Applied Mathematical and Physical
Sciences, Department of Mechanics, National Technical University of Athens, Zografou Campus, Theocaris Building, 157 73 Athens, Greece
1
Keywords: Hollow-fiber membranes, Dry-jet wet spinning, Polyimide, Functionalized
MWCNTs, Pyrolysis, Mechanical properties.
Abstract
The effect of the addition of various MWCNTs concentration as filler material in polyimide
hollow fiber membranes for gas separation is studied in the present work by analysing their
performance and their mechanical properties. Polymeric BTDA-TDI/MDI (P84) co-polyimide hollow-fiber membranes were prepared using the dry/wet spinning technique, which is
based on a phase-inversion method (McKelvey et al, 1997, Favvas et al, 2013). The hollow
fibers after their production stage exhibited selectivity and permeability for gas separation as
well as improved mechanical properties. Direct tension and three point bending tests were
performed. It was found that the addition of MWCNTs essentially increased the tensile and
flexural moduli of elasticity; an enhancement in strength properties was also evident and the
results were discussed by taking into account the gas separation performance of the fibers.
Introduction
In membranes industries, two different configurations exist, namely the sheet and the cylindric
al modules. The widely used membranes in industrial applications are the spiral wound and the
hollow fibers; the gas productivity of the hollow fibers is approximately 10 times higher than
the spiral wound. The focal point of the present work was to produce polymeric hollow fiber
membranes with high gas selectivity and at the same time high permeability. To achieve these
goals, polymeric mixed matrix membranes were produced by using inorganic fillers (MWCNTs) as additives. After the production process, the mechanical properties of the produced
membranes were investigated.
Experimental Procedure
The MWCNTs were prepared in NCSR “Demokritos” Research Center by chemical vapor deposition method and where functionalized by phenol groups on their surface. After the modification of the materials the MWCNTs were dispersed in NMP solvent using a tip-sonicator. When
the solution was ready (more than 12 hours mixing), the solution was forwarded to the spinning
machine to produce the hollow fibers. Typical produced hollow fibers can be seen in Fig. 1.
132
ISBN: 978-960-98739-8-7
Figure 1: Macro-photograph of the produced hollow fibers with different MWCNTs concentration.
The tests were performed in an MTS (Insight Testing Systems) loading frame of capacity 10
kN at the Laboratory of Testing and Materials of the National Technical University of Athens.
The tests were carried under displacement control mode. They were quasi-static at a rate equal
to 0.005 mm/sec. For the three point bending tests the support span was equal to 14 mm. The
experimental setup for the tensile and three point bending tests can be seen in Fig. 2a and b,
respectively. It is mentioned that the three point bending tests were carried out according to the
ASTM D790-10 specification.
(a)
(b)
Figure 2: Experimental setup for (a) tensile and (b) three point bending tests.
Results
Typical curves of the direct tension tests are plotted in Fig. 3a. It is noticed that increasing the
MWCNTs concentration, the slope of the curves increases also and therefore the modulus of
elasticity increases too. Moreover it was observed that the tensile strength increased with increasing filler concentration. On the other hand the elongation at fracture was decreased and
therefore it is concluded that the nano-filler’s addition made the hollow fiber membranes more
brittle. Fig. 3b shows the results of the three point bending tests and their typical stress-strain
curves. Also in this case, the modulus of elasticity increased and the maximum flexural strain
decreased with increasing the MWCNTs concentration.
ISBN: 978-960-98739-8-7
133
Figure 3: Typical stress-strain curves for (a) tensile and (b) three point bending tests.
Concluding Remarks
The MWCNTs addition has an immediate effect on the hollow fiber membranes mechanical
behavior; the moduli of elasticity for both tension and three point bending tests was increased
by almost 15% for the highest investigated nano-filler concentration.
References
1. Favvas, E. P, Nitodas, S.F., Stefopoulos, A. A., Papageorgiou, S. K., Stefanopoulos, K.L. and Mitropoulos, A.
Ch., 2013, High purity multi-walled carbon nanotubes: Preparation, characterization and performance as filler
materials in co-polyimide hollow fiber membranes. Separation and Purification Technology, 122, 262-269.
2. McKelvey, S. A., Clausi, D. T. and Koros, W. J., 1997, A guide to establishing hollow fiber macroscopic
properties for membrane applications. Journal of Membrane Science, 124, 223-232.
134
ISBN: 978-960-98739-8-7
Corrosion-induced fracture toughness degradation of artificially
aged 2024 aluminum alloy specimens
Paraskevi Liberaki1, Nikolaos D. Alexopoulos1*, Constantinos I. Stergiou2
and Stavros K. Kourkoulis3
University of the Aegean, Department of Financial Engineering, 821 00 Chios, Greece,
* Corresponding author’s email address: [email protected]
2
Technological Educational Institute of Piraeus, Department of Mechanical
Engineering, 250 Thivon Ave, 12244 Athens, Greece
3
Laboratory for Testing and Materials, School of Applied Mathematical and
Physical Sciences, Department of Mechanics, National Technical University of Athens, Zografou Campus, Theocaris Building, 157 73 Athens, Greece
1
Keywords: aluminum alloy, fracture toughness, 2024, corrosion, heat treatment
Abstract
The effect of different artificial aging conditions on the fracture toughness of aluminum alloy
2024 is studied in the present work together with the effect of corrosion exposure on the same
artificially aged specimens. Fracture toughness specimens were manufactured according to the
international specifications and then they were artificially aged for three different aging tempe
ratures. The critical stress intensity factor Kcr was found to decrease up to the peak-aging regime
and to increase substantially in the over-aging regime for all the investigated artificial aging
temperatures. Exposure of the already heat-treated specimens to the corrosion solution for the
same duration decreased the critical stress intensity factor for all artificial aging conditions. At
the peak-aging regime, the effect of corrosion seems to have the minimum impact on the critical
stress intensity factor.
Introduction
Since many decades aluminium alloy 2024 is widely used in the aeronautical industry due to
its high specific mechanical properties and damage tolerance capabilities. New versions and
derivatives of AA2024 with lower compositions of Fe and Si have been investigated during the
last decade in an attempt to improve fatigue performance, fracture toughness and resistance to
corrosion. AA2024-T3 is a high strength alloy with a complex microstructure, comprising the
aluminium matrix as well as a number of different intermetallic particles (Starke and Staley,
1996). The main problems during design and inspection phases are the fatigue, corrosion and
impact damage(s) that the fuselage and wing skins are subjected to. The possibility that the
corrosion damage will interact with other forms of damage, e.g. fatigue cracks, can result in loss
of the structural integrity and may lead to fatal consequences.
In aeronautical applications, the most widely observed type of corrosion is exfoliation corrosion
that results mainly from intergranular corrosion. During exfoliation corrosion, the inner material grains are superficially displaced outwardly of the material as a result of adjacent corrosion
products occurring in the metal grain boundary just below the surface. Macroscopically this can
be seen as some grains which are being exfoliated from the material surface. The exfoliation
corrosion is the most typical case in aluminum corrosion products wherein the grain structure of
the material (thin and longitudinal) is oriented in a direction. The technique of exfoliation corrosion causes pitting, intergranular corrosion and embrittlement due to diffusion and trapping
of hydrogen (Kamoutsi et al, 2006; Alexopoulos, 2009).
ISBN: 978-960-98739-8-7
135
The material used in the present study was wrought aluminum alloy 2024-T3 which was received in sheet form with nominal thickness of 3.2 mm. Fracture toughness specimens were
cut from flat sheets of geometrical dimensions 35 x 50 cm (L direction) according to the ASTM
E561standard. Three different aging temperatures, namely 170, 190 and 210 oC were used for
artificial aging, while aging times were chosen to be in the range between 1 and 96 hours and
depending on the artificial aging temperature.
Half of the available specimens were tested according to the ASTM E561 immediately after
the artificial ageing heat treatment in order to assess the effect of artificial ageing on the critical stress intensity factor Kcr. The other half fracture toughness specimens were masked and
exposed for 2 h to the laboratory exfoliation corrosion environment (hereafter called EXCO
solution) according to the ASTM G34 specification.The surfaces of the specimens were cleaned
with acetone and then they were exposed to EXCO solution. The common 2 h exposure time
was selected according to the literature (Alexopoulos et al, 2015) since it was found that for
longer exposure times surface corrosion products are being formed (pits) accompanied by the
respective micro-cracking that further degrade the tensile mechanical performance of AA2024T3. A servo-hydraulic Instron 100 kN testing machine was used for the mechanical tests. An
external extensometer was attached to the specimen surface. A data logger was used to store
the data of axial force, displacement and axial strain from the attached extensometer in a digital
file.
Typical experimental load-CMOD (Crack Mouth Opening Displacement) curves for the different artificial aging times at 190oC of AA2024 can be seen in Fig. 1a. As expected, the resistance curve of the alloy becomes more compliant with the increase of artificial aging time.
This means that the maximum applied force Pmax decreases and the value of the CMOD at the
maximum force decreases with the ageing time increase. Fig. 1b shows the respective curves of
the specimens having the same aging conditions and subsequent 2 h exposure to exfoliation corrosion solution. It is evident that the corrosion exposure affects the resistance curves in terms of
decreasing the maximum applied mechanical load.
7
7
R��
9 ��
6
6 ��
5
L��
L��
6
2 ��
4
3
6 3 ��
6 3 ��
4
2 ��
3
2
2
R��
5
6 ��
9 ��
2 4 ��
1
0
0,0
A �� A �� 2 0 2 4 �T 3
L �� 3 �2 � �
A �� @ 1 9 0 �C
0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0
C �� C O D ��
(a)
1
0
0,0
2 4 ��
A �� A �� 2 0 2 4 �T 3
L �� 3 �2 � �
A �� @ 1 9 0 �C
E X C O �� 2 ��
0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0
C �� C O D ��
(b)
Figure 1: (a) Typical load – crack mouth opening displacement curves for five different artificial aging times at
190oC aging temperature of AA2024 (a) without corrosion (artificially aged) and (b) artificially aged with subsequent 2 hours exposure to exfoliation corrosion solution.
136
ISBN: 978-960-98739-8-7
Elongation at fracture and critical stress intensity factor seems generally to have the same trend
regarding their corrosion-induced decrease and that was absolutely dependant on the formation
of S-phase precipitates. An exception is the under-ageing condition, where the formation of GPB
zones doubled or tripled the Kcr decrease, while the opposite was noticed for the elongation at
fracture Af decrease. The lowest Kcr decrease was noticed for the specimens at the peak-ageing
condition that also exhibited the lowest ductility decrease due to the corrosion exposure.
Concluding Remarks
A short corrosion exposure time was selected to corrode fracture toughness specimens of different tempers, i.e. including under-, peak- and over-ageing conditions. It was found that the
corrosion-induced decrease of the starting point (T3 temper) is almost 7 % and this was attributed to hydrogen embrittlement due to the absence of pits and micro-cracking that act as stress
raisers. The lowest Kcr decrease was noticed for the specimens at the peak-ageing condition.
At the over-ageing condition the corrosion-induced decrease was partially ‘restored’ at the T3
temper. Extreme artificial ageing leads to no significant corrosion-effect on the critical stress
intensity factor.
Acknowledgements
The authors gratefully acknowledge the financial support of the European Union (European
Social Fund – ESF) and Greek national funds through the Operational Program “Education
and Lifelong Learning” of the National Strategic Reference Framework (NSRF) - Research
Funding Program: “Archimedes III – Technological Educational Institute of Piraeus – Expe
rimental and theoretical investigation of mechanical properties degradation of the aeronautical
aluminium alloy 2024 due to corrosion” (MIS 383575).
References
1. Alexopoulos, N. D., 2009, On the corrosion-induced mechanical degradation for different artificial aging
conditions of 2024 aluminum alloy. Materials Science and Engineering A520, 40-48.
2. Alexopoulos, N. D., Velonaki, Z., Stergiou, C., Kourkoulis, S. K., 2015, The effect of artificial ageing heat
treatments on the corrosion-induced hydrogen embrittlement of 2024 aluminium alloy. In press, Corrosion
Science.
3. Kamoutsi, H., Haidemenopoulos, G. N., Bontozoglou, V. and Pantelakis, Sp. G., 2006, Corrosion-induced
hydrogen embrittlement in aluminium alloy 2024. Corrosion Science, 48, 1209.
4. Starke, E. A and Staley, J. T, 1996, Application of Modern Aluminum Alloys to Aircraft. Progress in Aerospace
Science, 32, 131-172.
ISBN: 978-960-98739-8-7
137
The impact of corrosion-induced micro-cracking on the structural
integrity of ultra-thin sheets of aeronautical aluminum alloy 2024
Christina Margarita Charalampidou1, Nikolaos D.Alexopoulos1*,
Panagiotis Skarvelis2, Charis Dalakouras3 and Stavros K. Kourkoulis3
2
National Technical University of Athens, School of Mining and Metallurgical Engineering,
9 Heroon Polytehneiou Avenue, 157 80 Athens, Greece
3
Laboratory for Testing and Materials, School of Applied Mathematical and Physical Sciences,
Department of Mechanics, National Technical University of Athens, Zografou Campus, Theocaris Building, 157 73 Athens, Greece
1
Keywords: aluminum alloy, 2024, corrosion, cracks.
Abstract
Ultra-thin (t ≤ 0.4 mm) 2024-T3 aluminum sheet thickness specimens were exposed to the
exfoliation corrosion solution and corrosion-induced surface cracks were formed. This microcracking network was quantified via metallography and image analysis techniques. The cracking measurements results were used to calculate the distribution of stress intensity factor on the
specimens that were exposed for different corrosion times. Above results were correlated with
the tensile mechanical properties of the corroded specimens (Alexopoulos et al, 2012). The loss
of tensile ductility is discussed and combined along with the presence of the micro-cracking
network as well as the hydrogen embrittlement effect.
Introduction
Aluminum alloy 2024 is the most widely used aluminum alloy in the aircraft industry. For the
case of aircraft structures made of aluminum alloy 2024-T3, several types of damage have been
reported over a large number of repairs (Vogelesang and Vlot, 2000); fatigue cracks was the
most frequent damage observed, having a frequency of more than 57% out of the total repairs
number. Second most frequent observed damage in aircraft structures was corrosion damage;
approximately 30% of total repairs were dealing with the removal of corrosion-induced surface
products and the restoration of structural integrity of the components.
The main damage mechanism on the corroded surface of aluminum alloy 2024 is the formation of pitting. Individual particle-nucleated pits coalesce laterally and in depth in order to form
larger pits and sub-surface micro-cracks that lead to exfoliation of the alloy with increasing
exposure time (Boag et al, 2011). As the specimen is exposed to the corrosive environment, it is
expected that the surface corrosion products will increase their depth with increasing exposure
time. Corrosion strongly affects the crack tip deformation and damage mechanisms and as a
consequence it can modify the crack growth rates as well as the crack path.
Material and experimental set up
The material used was a wrought aluminum alloy 2024-T3 which was received in sheet form of
0.4 mm nominal thickness. Tensile specimens of 2024 were machined according to DIN 50125
specification having 50 mm gauge length and 15.5 mm gauge width at the reduced cross-section
(Alexopoulos et al, 2012). Prior to corrosive solution exposure, all surfaces of the specimens
were cleaned with alcohol according to ASTM G1 specification.
138
ISBN: 978-960-98739-8-7
Tensile and fatigue specimens were afterwards exposed to the laboratory exfoliation corrosion
environment (hereafter called EXCO solution) according to ASTM G34 specification. EXCO
solution consisted of 234 gr NaCl, 50 gr KNO3 and 6.3 mL HNO3 (70% w.t.) diluted in 1l distilled water.
The tests were performed in an MTS (Insight Testing Systems) loading frame of capacity 10
kN at the Laboratory of Testing and Materials of the National Technical University of Athens.
The tensile test results and the respective degradation due to the corrosion exposure time were
reported in (Alexopoulos et al, 2012). Rectangular specimens (30x30 mm2) for metallographic
investigations were also cut and exposed to the same corrosion exposure times as the tensile
specimens. After the corrosion exposure, these specimens were cut in several cross-sections
that were mounted on resin system, grinded, polished and examined in a Leica DM LM optical microscopy. Several cross-sections of the same specimen were examined so as to create
a database (distribution) of the cracking depths for the different exposure times to corrosion
solution.
Experimental results and conclusions
The microstructural analysis of the pre-corroded specimens showed that the corrosion-induced
cracks tend to increase their depth of attack (a) with continuously increasing exposure time.
The stress intensity factor calculation formula can be found in (Dowling, 1993). Eq.1 gives the
K value for the case of double surface notches that have depth a (μm) on the total width w (μm)
when a constant stress σ (MPa) is applied to the specimen.
where Y
(1)
= 1.99 + 0.76
100
Probability [%]
80
Aluminum alloy 2024-T3, L direction, t = 0.4 mm
Both-sided corroded specimens
Exposure for different exfoliation corrosion times
2h
60
0.66 h
1.33 h
4h
6h
40
0.33 h
20
0
0-0,5 0,6-1 1,1-1,5 1,6-2 2,1-2,5 2,6-3 3,1-3,5 3,6-4 4,1-6,5 6,6-8,5
Stress intensity factor (s.i.f) clashes [MPa*m1/2]
Figure 1: Distribution of the calculated stress intensity factor K for the pre-corroded specimens for different
exposure times to exfoliation corrosion solution.
ISBN: 978-960-98739-8-7
139
Taking advantage of the data of the microstructural analysis, the depth of attack of corrosioninduced cracks α was measured and distributed for several depth clashes. For the K calculation
of the corroded specimens, a constant stress σ = 100 MPa was used; this stress level is defin itely
within the elastic regime of the specimen in order to avoid the excess plastic phenomena in
front of the crack tip. On the other hand, this stress level is considered as the maximum allowable stress value for the fuselage design of the aircraft.
Fig.1 shows the distribution of stress intensity factor (K factor) for the different exposure times
to the exfoliation corrosion solution. A crack is being nucleated from a pit during simultaneous
mechanical load (constant refers to stress corrosion cracking and dynamic refers to fatigue)
only if the pit grows to such a level that the K factor reaches a certain threshold value (Rokhlin
et al, 1999). If the value of K factor overcomes the threshold value ( ), then the crack propagates until the specimen’s fractureThe outcome of the exposure to the corrosive environment is
actually a surface layer with corrosion products that cannot withstand the mechanical load.
Fig.2 shows a typical corroded cross-section with corrosion-induced surface notches and the
concept of the ‘effective’ thickness of the specimen that is essentially reduced from the nominal
thickness due to corrosion exposure. Two mechanical models have been devised to calculate the
reduction of the effective thickness of the specimens, subjected to the different exposure times
to EXCO. The first model is the iso-modulus of elasticity (Alexopoulos and Papanikos, 2008)
that lies on the concept that the uncorroded material in the center of the specimen has the same
modulus of elasticity with the reference specimens. Fig.2b shows the results for the effective
thicknesses which is decreasing exponentially along with increasing exposure time. In the same
figure, the experimentally derived effective thickness from metallography and image analysis
can also be seen.
(a)
(b)
Figure 2: (a) Cross-section of the corroded specimen and the concept of true thickness and (b) evaluated and
calculated specimen’s true thickness for different exfoliation corrosion times, by exploiting the concept of isomodulus of elasticity and iso-yield stress.
The second model is the iso-yield stress that is investigated in the present work and lies on the
concept that the uncorroded material in the center of the specimen has the same yield stress Rp
with the reference specimens. This concept uses the mechanical load value Fj (kN) that corresponds to the yield stress
of the specimens. The specimen’s thickness , after j hours to
EXCO solution, can be calculated by means of the respective values of mechanical load F0 and
140
ISBN: 978-960-98739-8-7
initial thickness B0 of the reference specimens along with the mechanical load Fj value of the
corroded specimen for j hours exposure to the corrosion solution as:
.
(2)
It is obvious from Fig.2b that the iso-yield stress concept gives higher effective thickness decrease than the experimentally derived values, while the iso-modulus of elasticity concept quite
underestimates the effective thickness values, for the same exposure time.
References
1. Alexopoulos, N.D. and Papanikos, P., 2008, Experimental and theoretical studies of corrosion-induced
mechanical properties degradation of aircraft 2024 aluminum alloy. Materials Science and Engineering A,
498, 248–257.
2. Alexopoulos, N.D, Dalakouras, C.J, Skarvelis, P and Kourkoulis, S.K, 2012, Accelerated corrosion exposure
in ultra thin sheets of 2024 aircraft aluminium alloy for GLARE applications. Corrosion Science, 55, 289–
300.
3. Boag, A., Hughes, A.E., Glenn, A.M., Muster, T.H. and McCulloch, D., 2011, Corrosion of AA2024-T3 Part
I: Localised corrosion of isolated IM particles. Corrosion Science, 53, 17-26.
4. Dowling, N., 1993, Mechanical behavior of materials: engineering methods for deformation, fracture, and
fatigue (Englewood Cliffs, New Jersey: Prentice hall)
5. Rokhlin, S.I, Kim, J.-Y., Nagy, H. and Zoofan, B., 1999, Effect of pitting corrosion on fatigue crack initiation
and fatigue life. Engineering Fracture Mechanics, 62, 425-444.
6. Vogelesang, L.B and Vlot, A, 2000, Development of fibre metal laminates for advanced aerospace structures.
Journal of Materials Processing Technology, 103, 1–5.
ISBN: 978-960-98739-8-7
141
Fracture toughness of electron beam welded specimens of
aluminum alloy 6156
1
1,*
1
Theano Examilioti , Nikolaos D. Alexopoulos , George Stefanou ,
2
Vasilis Stergiou , Stavros K. Kourkoulis
3
University of the Aegean, Department of Financial and Management Engineering,
Chios Island, Greece
2
Hellenic Aerospace Industry, Special Processes & Laboratories Department, 320 09
Schimatari, Greece
3
Sciences, Department of Mechanics, National Technical University of Athens, Zografou
Campus, Theocaris Building, 157 73 Athens, Greece
1
Keywords: tension, fatigue, fracture toughness, aluminum alloy
Abstract
In the present work, the mechanical behavior of reference and electron beam welded specimens
made of aluminum alloy 6156-T4 was examined. It was found that the decrease in mechanical properties is more pronounced for the tensile ductility, while the respective decrease of the
fracture toughness was milder. A second series of welded specimens were constructed; before
welding, artificial aging was performed under different conditions namely under-, peak- and
over-aging. It was found that the welded specimens with the pre-welding heat treatment exhibits higher decrease in fracture toughness.
Introduction
For many decades now, aluminum 6056 alloy is widely used in fuselage sections where strength
and damage-tolerance properties are required. However, 6056 did not meet the design criteria
due to the low tolerance to damage. An improved version of the alloy should be developed,
which is finally obtained under the commercial code of 6156. The aluminum alloy 6156 is already used during the construction of the bottom of the fuselage shell of different aircraft types.
It is also recommended for shaft parts where high tensile strength and tolerance to damage are
demanded. The 6156 alloy has also excellent formability and corrosion resistance. Finally, in
contrast with the earlier alloy 6056, in 6156 shows higher tenacity. Innovative 6156 fuselage
sheets are particularly recommended for such a use as it is supposed to have higher damage
tolerance capabilities than its predecessor. So far the literature regarding the 6156 aluminum
alloy remains limited. Recently the effect of aging treatments on microstructure and hardness
were reported in the literature (Zhang et al, 2012a) together with the effects of Ag addition on
the long thermal stability of the alloy (Zhang et al, 2012b).
Typical resistance curve of force – Crack Mouth Opening Displacement (CMOD) of the edge
crack for reference AA6156-T4 specimen can be seen in Fig.1a (Stefanou et al, 2014). Typical
curves for unwelded specimens that had been artificially aged for under- (4 hours), peak- (28
hours) and over-aging (96 hours) can also be seen in the same figure.
With artificial aging time increasing, the peak force of the curves increases and the displacement to peak force decreases. For example, after 4 hours of artificial aging the maximum force
142
ISBN: 978-960-98739-8-7
was 7.49 kN while the crack mouth displacement was 3.31 mm. After 28 hours aging, the maximum force was reduced by almost 0.5 kN and the displacement of the edges of the crack was
3.22 mm. However after 96 hours of artificial aging where the alloy is in over-aging condition,
a slight increase of the maximum load is observed.
8
6
7
Reference
5
6
4
4 Hours
4
Load [kN]
Load [kN]
5
28 Hours
3
2
1
0
0,0
0,5
1,0
1,5
2,0
3
96 Hours
28 Hours
2
96 Hours
Resistance curves
Aluminum alloy 6156-T4
L direction, t=3.6 mm
Artificial aging @ 170oC
4 Hours
Reference
Resistance curves
Aluminum alloy 6156-EBW
L direction, t=3.6 mm
Artificial aging @ 170oC
1
2,5
3,0
3,5
4,0
4,5
Crack mouth opening displacement (COD) [mm]
(a)
5,0
5,5
0
0,0
0,5
1,0
1,5
2,0
2,5
3,0
3,5
4,0
4,5
5,0
5,5
Crack mouth opening displacement (COD) [mm]
(b)
Figure. 1 - (a) Typical curves power- displacement of the crack edges for the alloy 6156 with for 3 different time
intervals for temperature artificial aging at 170oC (b) Typical curves power- displacement of the crack edges for
the alloy 6156 EBW with for 3 different time intervals for temperature artificial aging at 170oC
The respective resistance curves after the electron beam welding can be seen in Fig.1b for all
investigated artificial aging prior to welding. The critical stress intensity factor was calculated
according to the compliance method and according to ASTM E561. It was found that the lowest
decrease (~ 5%) was noticed for the case of T4 condition.
Applying artificial aging prior to welding didn’t increase the fracture toughness of the welded
specimens. The highest decrease was noticed for the under-aged specimens while the specimens at the over-aging condition exhibited a decrease around 15% in fracture toughness.
References
1. Stefanou, G., Migklis, E., Kourkoulis, S. and Alexopoulos, N.D., 2014, Mechanical behaviour of aeronautical
aluminum alloy 6156 for different artificial aging conditions. In Proceedings of International Symposium on
Aircraft Materials ACMA2014, Marrakech, Morocco, 23-26 April 2014.
2. Zhang, H., Zheng, Z., Zhong, S., Luo, X. and Zhong, J., 2012a. Effects of two-step aging treatment on
microstructure and properties of 6156 aluminum alloy. Chinese Journal of Nonferrous Metals, 22, 10251032
3. Zhang, H., Zheng, Z., Lin, Y. and Zhong, J., 2012b. Effects of Ag addition on the microstructure and thermal
stability of 6156 alloy. Journal of Materials Science, 47 (9), 4101-4109.
ISBN: 978-960-98739-8-7
143
Polyvinyl alcohol - carbon nanotubes fiber as an embedded sensor
in GFRP under bending loads
Panagiota Lymperta1, Stavros K. Kourkoulis2, Philippe Poulin3 and
Nikolaos D.Alexopoulos1*
1
2
3
Laboratory for Testing and Materials, School of Applied Mathematical and
Physical Sciences, Department of Mechanics, National Technical University of Athens,
Zografou Campus, Theocaris Building, 157 73 Athens, Greece
Université de Bordeaux, Centre de Recherché Paul Pascal - CNRS, Avenue
Schweitzer, 33600, Pessac, France
Keywords: PVA-CNT, fiber, sensor, bending, composites.
Abstract
The evolution of technology is leading to the development of innovative materials that serve
multi - functional purposes, e.g. besides their increased mechanical properties they could serve
additional capabilities such as it is monitoring/sensing. In the present work, a fiber made from
polyvinyl alcohol and carbon nanotubes (hereafter denoted as a PVA-CNT fiber) was embedded
into Glass Fiber Reinforced Polymer (GFRP) coupons for structural health monitoring purposes.
The PVA-CNT fiber is electrically conductive and therefore any strain changes of the coupon
can be monitored by the method of electrical resistance change of the fiber. The fiber was successfully embedded into the non-conductive material GFRP, so the electrical resistance change
of the fiber with increasing mechanical loading gives information related to the surrounding
matrix within the bulk of the material. In the present work, specimens with embedded fiber
under four point bending (4PB) were used. The electrical response of the fiber under uniform
bending loads was recorded and compared with the mechanical response of the coupons so as to
assess the chances of exploiting the electrical resistance change readings of the fiber for strain
sensing and damage assessment purposes.
Introduction
During the last decades, the demand for composite materials in engineering applications is continuously growing, in a broad variety of application fields like aeronautics, automotive, civil
engineering industry etc. Their (quite a few) advantages (high strength, lightweight, the ability
to mold complex shapes etc.) render them an attractive alternative solution. On the contrary
their structural health monitoring is quite difficult. Nowadays traditional sensing methods are
used, such as Fiber Optic sensors, acoustic emission, and piezoelectric sensors. Nevertheless,
these sensors are difficult to be embedded, in most cases they degrade the materials’ mechanical properties and moreover they do not give enough information regarding the component’s
structural health.
In this context Baron and Schulte were the first ones who noticed that sensing in carbon composites can be implemented by taking advantage of the change of their electrical resistance.
This method has been successfully applied to carbon reinforced polymeric materials that exhibit intrinsic electrical resistance. In 1991, the invention of carbon nanotubes (CNTs) was
reported by Iijima (Iijima, 1991) and it was noticed that the same sensing technique could be
used. Indeed CNTs exhibit similar electrical characteristics with carbon fibers and hence, could
144
ISBN: 978-960-98739-8-7
be added in a non-conductive matrix so as to monitor the structural health of a matrix via the
electrical resistance change of the matrix.
In the present work, the PVA-CNT fibers used as embedded sensors were produced at the University of Bordeaux as the coagulation of carbon nanotube suspensions in a PVA solution (Vigolo et al, 2000). This process was in-house developed in 2000 and was since then improved by
Poulin et al (2002) The specific PVA-CNT fibers have extra small diameter and their modulus
of elasticity is similar with that of the composite material, i.e. of GFRP. It is exactly the intrinsic
electrical resistance change of PVA-CNT fiber makes feasible the structural health monitoring
of non-conductive GFRP materials, with the electrical resistance change method.
The present study is focused on the characterization of the piezoelectric behavior of composites
with embedded PVA-CNT fibers under four point bending loads. Correlating the axial strain
from the critical regime of the specimen and the fiber’s electrical resistance change, under the
same bending loads, could provide details for each regime of damage.
Experimental procedure
The specimens were manufactured at the Laboratory of Advanced Materials of the Hellenic
Aerospace Industry (HAI). Typical GFRP material was used to embed the PVA-CNT fiber, with
the matrix being a typical epoxy resin produced with Vacuum Assisted Resin Infusion process
(Alexopoulos et al, 2010). Testing coupons were rectangular and in accordance to the ASTM
D5947 standard. The PVA-CNT fibers were embedded in the composite during manufacturing
and just below the last layer of the fabrics. Four different fiber were used, with different prestretching ratios: More specifically the fibers were: (a) untreated ones, pre-stretching level at (b)
50%, (c) 100% and finally (d) 200%.
An MTS electro mechanical testing machine was used with a 10 kN load cell. An Agilent multimeter was also electrically attached to the fiber to record its electrical resistance change caused
by the applied mechanical loads. In the same time data for time, force, displacement and electrical resistance were monitored and recorded, at a sampling frequency of 1 Hz. In addition, and
in order to determine the axial strain and the deflection at the outer surface of the specimen, a
ten-strain-gauge chain together with suitable LVDTs were also used, respectively.
The experimental set up was arranged in accordance to the ASTM D6272 standard for 4PB
tests.
Experimental results and discussion
Standardized 4PB bending tests were conducted under incremental loading-unloading loops,
both within the elastic regime as well as above it. A typical diagram showing the axial stress,
the respective axial strain and the untreated fiber’s electrical resistance change versus time,
during a typical test with thirteen loading-unloading loops can be seen in Fig. 1a. The same
process was followed for all pre-stretching ratios. According to the experimental results and
recordings it was concluded that higher pre-stretching results to higher sensitivity of sensing in bending. For all fiber’s pre-stretching ratios a linear correlation between the bending
stress, the axial strain and the electrical resistance change in the elastic regime was observed.
Increasing the applied bending load this linear approximation is lost and hysteresis loops are
formed.
Average axial strain in between the loading span was calculated by the readings of the tenstrain-gauge chain and average values were calculated for the different mechanical loads. The
results were compared with the maximum axial strain as calculated by the strain gauge that was
placed in the mid-distance of the loading span. The results can be seen in Fig.1b for the different investigated mechanical loads, where small discrepancies can be noticed for the higher
mechanical loads.
ISBN: 978-960-98739-8-7
145
As a last step, and based on literature and other studies (Alexopoulos et al, 2013), an appro
ximation of the “gauge factor” of the fibers, under four point bending loads, was calculated, as
a correlation of electrical resistance change with axial strain according to Eq.1. CRS parameter
is the acronym of ‘Correlation of Resistance to Strain:
CRS =
1
∆(ERC )
=
CRS
∆å
∆å
or
∆(ERC )
(1)
S urface axial strain
Fiber's E RC
Mechanical stress
12
300N
10
450N
400N
350N
250N
8
200N
6
100N
4
2
0
0
2000
4000
6000
8000
10000
12000
5,0
250
4,5
225
4,0
200
3,5
175
3,0
150
2,5
125
2,0
100
1,5
75
1,0
50
0,5
25
0,0
0
Testing time [sec]
High applied
stress regime
Low applied
12 stress regime
High applied stress regime
(a)
450N
8
6
4
Up to 30% of UTS
100-200N
400N
10
Up to 15% of UTS
25-100N
3
Axial strain at the bottom surface (10
) [ìå]
14
Applied axial stress at fiber's layer [MPa]
14
Electrical resistance change of the embedded fiber
ÄR/R 0 [%]
3
Axial strain at the bottom surface (10
) [ìå ]
Based on early estimates it could be concluded that the gauge factor is constant for bending
loads up to about 15% of UTS while for higher mechanical loads it seems that it is increasing
linearly.
350N
300N
250N
Average axial strain
Max axial strain
2
0
0,0
GFR P materials with embedded PVA-CNT fiber
specimen Pc000_N09, untreated fiber
0,5
1,0
1,5
2,0
2,5
3,0
E lectrical resistance change of the embedded fiber ÄR /R 0 [%]
(b)
Figure 1: (a) Typical variation of engineering stress / strain of the composite as well as of the electrical resistance change of the embedded PVA-CNT fiber with varying testing time and (b) axial strain variation of the
composite against the electrical resistance changes of the embedded fiber.
146
ISBN: 978-960-98739-8-7
Concluding remarks
GFRP specimens with embedded PVA-CNT fiber were subjected to four-point bending, in order to correlate the electrical response of the embedded fiber with the mechanical response of
the coupons for different loading regimes. It was experimentally determined that the correlation
is quite satisfactory. Moreover, a constant gauge factor was calculated for the elastic regime of
the coupons under four-point bending tests.
Acknowledgements
The authors would like to gratefully acknowledge the Hellenic Aerospace Industry for manu
facturing the specimens used in the present study.
References
1. Iijima, S., 1991, Helical microtubules of graphitic carbon. Nature, 354, 56-58.
2. Vigolo, B., Penicaud, A., Coulon, C., Sauder, C. Pailler, R., Journet, C., Bernier, P. and Poulin, P., 2000,
Macroscopic fibers and ribbons of oriented carbon nanotubes. Science, 290, 1331-1334.
3. Poulin, P., Vigolo, B. and Launois, P., 2002, Films and fibers of oriented single wall nanotubes. Carbon, 40,
1741-1749.
4. Alexopoulos, N. D., Bartholome, C., Poulin, P. and Marioli-Riga, Z., 2010, Structural health monitoring of
glass fiber reinforced composites using embedded carbon nanotube (CNT) fibers. Composites Science and
Technology, 70, 260–271.
5. Alexopoulos, N. D., Jaillet, C., Zakri, C., Poulin, P. and Kourkoulis, S.K., 2013, Improved strain sensing per
formance of glass fiber polymer composites with embedded pre-stretched polyvinyl alcohol–carbon nanotube
fibers. Carbon, 59, 65-75.
ISBN: 978-960-98739-8-7
147
The effect of graphene nanoplatelets addition on the mechanical
performance of epoxy resin
Zafeiroula Paragkamian1, Nikolaos D. Alexopoulos1*, Philippe Poulin2
and S. K. Kourkoulis3
1
2
3
University of the Aegean, Department of Financial Engineering, 821 00 Chios, Greece
Université de Bordeaux, Centre de Recherché Paul Pascal – CNRS, Avenue Schweitzer,
33600, Pessac, France
Keywords: graphene nanoplatelets, epoxy, nanocomposite, mechanical properties,
fracture toughness.
Introduction
Graphene, a single layer of graphite, has captured the attention of scientific community due to
recently emerging high performance applications. It is considered to be a unique lightweight
two-dimensional material with exceptional mechanical and electrical properties (modulus of
alacticity ~1 TPa, tensile strength 130 GPa, maximum electrical conductivity over 6000 S/
cm, thermal conductivity 5000 W/mK) which make it extremely powerful with high efficiency
(Prolongo et al, 2013; Naebe et al, 2014). Graphene nanoplatelets (GNPs) is a newly developed
material with low cost (available at low cost of $ 5 / lb) which often increases the modulus of
various matrices, e.g. (King et al, 2013).
In the present work, graphene nanoplatelets (GNPs) were added to a typical epoxy resin matrix. Epoxy/GNPs nano-composites were prepared and mechanically characterized so as to assess the enhancement of tensile mechanical properties and fracture toughness. Different GNPs
concentrations (0, 0.1, 0.25, 1.0, 2.0, 3.0 and 5.0 % wt) as well as two different types of GNPs
(low and high dimensions) were studied and respective specimens for mechanical characterization were prepared. The results indicate that the addition of GNPs in resin results in different
mechanical properties that vary according to the concentrations and the type of the GNPs.
The epoxy resin SR 8100 was used together with the SD 8824 as resin hardener. This resin
system was supplied by SICOMIN (13161 Chateauneuf Martigue Cedex - France) and has very
low viscosity at ambient temperature while high mechanical properties can be achieved. Two
types of GNPs were used in this work, supplied by XG Sciences with the commercial names
xGnP Grade C-750 (flake thickness is 2 nm) and xGnP Grade M-15 (average thickness 6 nm).
To fabricate the epoxy/GNPs composites, the appropriate amount of GNPs were added to resin
and the materials were mixed using a High Speed Shear Mixer Silverson L4RT at 4000 rpm for
30 min. Then hardener (100 g epoxy resin added to 22 g hardener) was added to the GNPs/resin
solution and mixed by hand at room temperature for 3 min. The mixture was poured into the
molds, degassed for about 10 min at room temperature under vacuum until it was completely
bubble free, as described in the neat epoxy specimens. Then it was cured at 60°C for 40 min and
post-cured at 60°C for 4 hours.
148
ISBN: 978-960-98739-8-7
The materials were characterized using optical microscope and scanning electron microscope
(SEM). The analysis was carried out in a Leica DMR Optical Microscopy.
All mechanical tests were performed at room temperature using an MTS Insight electro
mechanical machine with maximum static load of 10 kN and the rate of the displacement of
machine grips was kept constant and equal to 0.016 mm/min for the tensile tests and 0.001 mm/
min for the fracture toughness tests. The tensile tests were carried out according to the standard
ASTM D638 to assess the materials’ tensile properties (i.e., the elastic modulus, ultimate tensile
strength, elongation at fracture) while the fracture toughness tests of the nanocomposites was
evaluated according to ASTM D5045.
Axial nominal stress [MPa]
70
0.25wt%GNPs
Reference
60
50
2wt%GNPs
5wt%GNPs
40
1wt%GNPs
3wt%GNPs
30
20
Epoxy resin SR 8100
Hardener SD 8824
2nm thikness xGnP Grade C-750
10
0
0,00
0,02
0,04
Axial nominal strain [-]
(a)
0,06
0,08
Critical stress intensity factor
Kcr [MPa.m1/2]
Typical tensile curves of the nano-composites can be seen in Fig.1. Fig. 1a shows that with in
corporation of Grade C GNPs and low concentrations (< 1 wt% GNPs) the tensile strength was
increased and the ductility was decreased (~ 32 % and ~ 42 % respectively due to incorporation of 0.25 wt% GNPs). Conversely, for higher concentrations 1, 2, 3 wt% GNPs the tensile
strength was decreased and the ductility was increased.
For the nanocomposites with lower concentrations (> 5 wt% GNPs) there was a decrease for
all mechanical properties. The addition of Grade C GNPs increased the strength properties and
decreased the ductility for the low concentrations (< 1 wt% GNPs) while decreased the strength
properties and increased ductility for even higher concentrations (up to 3 wt% GNPs).
For excess reinforcement (> 5 wt% GNPs) all mechanical properties were decreased.
Fig. 1b shows the results of the critical stress intensity factor Kcr for the investigated GNPs concentrations. The critical stress intensity factor Κcr is increasing up to 1 wt% GNPs concentration
while for higher GNPs concentration, Kcr decreases, as this is the case for our experimental
results. Higher Kcr decrease is noticed for the higher GNPs thickness.
1,6
1,4
GNPs Grade C-750 - SR 8100 (experiment)
GNPs Grade M-15 - SR 8100 (experiment)
Jun Ma et al - DGEBA Araldite-F
S. Chatterjee et al - Epikote 8282VEL
S. Chandrasekaran et al - Araldite LY556
Yanli Zhang et al - Epon44
M. M. Shokrieh et al - ML 526 Bisphenol-A
1,2
1,0
0,8
0,6
0,4
0,2
0
1
2
3
4
5
Concentration of GNPs [wt%]
(b)
Figure 1 (a) Typical tensile curves for GNPs-Grade C and resin nano-composites and (b) plot of the evaluated
critical stress intensity factor Kcr for different GNPs concentrations.
References
1. King, J., Klimek, D.R, Miskioglu, I. and Odegard, G.M., 2013, Mechanical Properties of Graphene
Nanoplatelet/Epoxy Composites. Journal of Applied Polymer Science, 128, 4217–4223.
2. Naebe, M., Wang, J., Amini, A., Khayyam, H., Hameed, N., Li, L.H., Chen, Y. and Fox, B., 2014, Mechanical
Property and Structure of Covalent Functionalised Graphene/Epoxy Nanocomposites. Scientific Reports, 4,
4375.
3. Prolongo, S.G., Jimenez-Suarez, A., Moriche, R. and Urena, A, 2013, In situ processing of epoxy composites
reinforced with graphene nanoplatelets. Composites Science and Technology, 86, 185-191.
ISBN: 978-960-98739-8-7
149
Effect of prior solid solution heat treatment on the mechanical
behavior of artificially aged aluminum alloy 2024 specimens
Antonios I.Pasoudis1, Paraskevi Liberaki1, Alexis Kermanidis2 and
Nikolaos D.Alexopoulos1*
2
Department of Mechanical Engineering, University of Thessaly, Volos, Greece
1
Keywords: aluminum alloy, fracture toughness, 2024, corrosion, heat treatment.
Abstract
This paper investigates the effect of solid solution heat treatment on the mechanical behavior
of 2024 aluminum alloy. Tensile and fracture toughness specimens were machined according to
international specifications. Half of the specimens were subjected to solid solution heat treatment and subsequent to artificial aging while the other half were subjected to the same artificial
aging conditions only. A direct comparison of the quasi-static properties and fracture toughness is attempted for the different aging conditions, including under-, peak- and over-aging
regimes.
Introduction
Aluminum alloy 2024 is a precipitation-hardenable Al-Cu aluminum alloy that presents high
damage tolerance capabilities and is widely used to produce aeronautical components. The better understanding of the physical metallurgy background of the age-hardened cast aluminum alloys indicated that chemical composition (Alexopoulos and Pantelakis, 2003), solidification rate
(Alexopoulos and Tiryakioglu, 2009) and heat treatment (Alexopoulos and Pantelakis, 2004)
are the key parameters to improve the mechanical properties. Aluminum alloy 2024 are precipitation hardened alloy by the S-type particles (Al2CuMg) reinforcing the aluminum matrix.
Although deformation and fracture of these alloys are strongly affected by the size, distribution,
and morphology of Al-grains and possible defects, e.g. voids, porosity etc., one of the main
factor that contribute to the mechanical behavior and fracture of these alloys is the presence of
the secondary precipitation-hardening particles (Polmear, 1989). Thus, the heat treatments that
controls the formation, size, distribution and morphology of the precipitates are important and
their effects on mechanical properties are studied in the present work.
It is well known that AA2024 is strengthened by microstructure evolution (precipitates) during
ageing. In the 1950s, Bagaryatsky (Bagaryatsky, 1952) proposed the following precipitation
sequence:
SSS → GPB zone → S’’ → S’ → S,
where SSS stands for supersaturated solid solution and GPB stands for Guinier-Preston- Bagaryatsky; it is considered to be a short range ordering of Cu and Mg solute atoms while S’’ are
very small precipitates fully coherent with the Al matrix. The S phase is an equilibrium phase
and is incoherent with the Al matrix. These S-phase particles are small round particles and 60%
of the particles are larger than 0.5 μm (Bagaryatsky, 1952). In contrast to the S-phase particles,
the Al(Cu,Mn,Fe,Si) intermetallics present irregular shapes and size typically larger than 5 μm.
The S’ phase has generally been considered as semi-coherent with the matrix, having the same
structure as the S-phase but with slightly different lattice parameters (Mondolfo, 1976). Ringer
150
ISBN: 978-960-98739-8-7
et al confirmed that GPB zones and other precipitate structures prior to the S-phase formation are the dominant precipitates at the strengthening regime of the ageing curve, while the
S-phase appears in the softening regime (over-ageing condition) (Ringer et al, 1996, Ringer et
al, 1998).
The material used was a wrought aluminum alloy 2024-T3 which was received in sheet form
with nominal thickness of 3.2 mm. Tensile and fracture toughness specimens were cut from
flat sheets of geometrical dimensions (L direction) according to standards ASTM E8 and E561,
respectively. The solid solution temperature was 495oC, while the duration was 30 minutes
(Kamoutsi et al, 2014). Three different aging temperatures, namely 170, 190 and 210oC were
used for artificial aging, while the artificial aging times were chosen to be in the range between
1 and 96 hours. Half of the available specimens were tested according to the specifications im
mediately after the artificial ageing heat treatment. The other half specimens were first subjected to solid solution heat treatment (heated and subsequently immediate quenched in < 10oC in
water with ice) and then artificially aged to the same aging times with the previous specimens. A
servo-hydraulic Instron 100 kN testing machine was used for the mechanical tests. An external
extensometer was attached to the specimen surface. A data logger was used to store the data of
axial force, displacement and axial strain from the attached extensometer in a digital file.
All tensile mechanical properties (yield stress Rp0.2%, tensile strength Rm, elongation at fracture
Af, strain energy density W) and fracture toughness Kcr (critical stress intensity factor for 3.2 mm
thickness) of the investigated specimens have been evaluated and compared with each other in
order to draw conclusions how solid solution and artificial ageing affects the above mentioned
mechanical properties. Figure 1 shows both the effect of solid solution and artificial aging heat
treatments on (a) yield stress and (b) elongation at fracture. It can be noticed that the solid solution treated specimens presented lower yield stress values than the non-solid solution treated
specimens. This can be explained by the dissolution of the possible S-type particles that were
formed during the stretching phase of the sheets.
(a)
(b)
Figure 1: The effect of solid solution and artificial aging on the (a) yield stress Rp0.2% and (b) elongation at fracture Af of aluminum alloy 2024-T3.
ISBN: 978-960-98739-8-7
151
With increasing artificial ageing time the S-type particles begin to precipitate from the nucleation stage for the solid-soluted specimens, while for the specimens without the solid solution,
the aging process starts from the growth stage as some precipitates were already formed, preferably at the grain boundaries.
For the case of the elongation at fracture, higher values were noticed for the solid-solution
specimens when compared with the respective artificially aged specimens. This is the case up
till the peak-aging regime (that presented the lowest ductility value at 48 hours artificial aging). Further aging the specimens (over-aging regime) showed that both heat treatment cases
presented almost the same elongation at fracture values, e.g. at 96 hours aging. Similar results
were noticed for fracture toughness; higher values were noticed in the under- and peak-aging
regimes while in the over-aging condition the solid solution and subsequent aged specimens
presented slightly higher fracture toughness values.
Concluding remarks
From this experimental work, it was shown that the mechanical properties of the precipitation
hardened aluminum alloy 2024 is sensitive to the microstructural changes due to heat treatment.
Solid solution and artificial aging have a profound effect on both strength and ductility tensile
properties as well as fracture toughness. Solid-solution heat treatment gives higher ductility
and fracture toughness values up till the peak-aging regime on the expense of lower strength
properties.
Acknowledgements
The authors gratefully acknowledge the financial support of the European Union (European
Social Fund – ESF) and Greek national funds through the Operational Program “Education and
Lifelong Learning” of the National Strategic Reference Framework (NSRF) - Research Funding
Program: “Archimedes III – Technological Educational Institute of Piraeus – Experimental and
theoretical investigation of mechanical properties degradation of the aeronautical aluminium
alloy 2024 due to corrosion” (MIS 383575).
References
1. Alexopoulos, N. D. and Pantelakis, Sp. G., 2003, Evaluation of the effects of variations in chemical composition
on the quality of Al-Si-Mg, Al-Cu, and Al-Zn-Mg cast aluminum alloys. Journal of Materials Engineering and
Performance 12 (2), 196–205.
2. Alexopoulos, N. D. and Pantelakis, Sp. G., 2004, Quality evaluation of A357 cast aluminum alloy specimens
subjected to different artificial aging treatment. Materials and Design, 25 (5), 419–430.
3. Alexopoulos, N. D.and Tiryakioglu, M., 2009, Relationship between fracture toughness and tensile properties
of A357 cast aluminum alloy. Metallurgical and Materials Transactions A, 40 (3), 702–715.
4. Bagaryatsky, Y.A., 1952, Structural changes on Aging Al-Cu-Mg alloys, Dokl Akad SSSR 87, 397-559.
5. Kamoutsi H., Haidemenopoulos G. N., Bontozoglou V., Petroyiannis P. V. and Pantelakis S., 2014, Effect of
prior deformation and heat treatment on the corrosion-induced hydrogen trapping in aluminium alloy 2024.
Corrosion Science 80(3), 139-142.
6. Mondolfo, L.F, 1976, Aluminum alloys – structure and properties (London: Butterworth).
7. Polmear, J., 1989, Light Alloys: Metallurgy of the Light Metals (Edward Arnold).
8. Ringer, S. P., Caraher, S. K. and Polmear I. J., 1998, Response to comments on cluster hardening in an aged
Al-Cu-Mg alloy. Scripta Materialia ,39, 1559-1567.
9. Ringer, S.P., Hono, K., Polmear, I. J. and Sakurai, T., 1996, Nucleation of precipitates in aged Al-Cu-Mg-(Ag)
alloys with high Cu:Mg ratios. Acta Materialia, 44 (5), 1883-1898.
152
ISBN: 978-960-98739-8-7
Nano-additions of multiwall carbon nanotubes in white cement for
restoration of monuments of Cultural Heritage
Zoi Metaxa1(*), Spyridoula Boutsioukou2, Stefanos Nitodas3 and
Stavros K. Kourkoulis1
(*)
E-mail: [email protected]
2
Department of Financial Engineering, University of the Aegean, Chios, Greece
3
Glonatech S.A. TE.S.P.A “Lefkippos” Ag. Paraskevi, Attica, Greece
1
Keywords: carbon nanotubes, cementitious materials, mechanical properties.
Abstract
Multi-wall carbon nanotubes (MWCNTs) were added as a reinforcement in cementitious materials which are currently used for the restoration of monuments of cultural heritage. The target
of the study is the quantification of the influence of different types of surfactants and different
concentrations of the surfactants as well as of MWCNTs on the mechanical performance of cement paste and mortar matrices, respectively. The abovementioned parameters were examined
with regard to the materials’ compressive and flexural strength. In addition the electrical resistance of the specimens was measured. It was concluded that optimum dispersion of the nanoreinforcement is directly related to the observed enhancement of the modulus of elasticity as well
as to the increased strength properties.
Introduction
It is well known that by adding carbon nanotubes in cementitious materials the mechanical
performance of the matrix is improved (Konsta-Gdoutos et al, 2010, Metaxa et al, 2012). This
increase depends on the degree of the nanomaterials’ dispersion. The addition of these nanoreinforcements at concentrations above the percolation threshold, besides the improvement of
the mechanical performance, might also be used to enhance the material’s monitoring ability
via the electrical resistance change method. This is a key concept for their wide-spread use, as
currently the traditional approaches for structural health monitoring of Historical Monument
components is largely based on strain gauges that are applied on the outer material’s surface
and therefore cannot detect any strain changes or induced damage inside the material / structure (Kourkoulis et al, 2013). The incorporation of electrically conductive carbon nanotubes,
is expected to improve the mechanical performance of the cementitious materials used for the
restoration of cultural heritage monuments and supply them with improved electrical properties.
Aalborg white CEM I 52,5R cement was used in the present study. Three different types of surfactants were used, namely sodium dodecyl benzene sulfonate (SDBS), Triton and Ceresit. The
MWCNTs/surfactant solutions were sonicated using a titanium probe. The materials were mixed
using a standard mixer, following the ASTM specifications.
Compression and three-point bending tests were performed on intact and notched specimens,
respectively. During the mechanical tests, time, force, and displacement were continuously monitored and recorded. The results indicated that the SDBS and Triton surfactants are possibly not
ISBN: 978-960-98739-8-7
153
suitable for use in cementitious materials as they both degraded the material’s mechanical properties.
Ceresit superplasticizer gave the best results in terms of electrical resistance and mechanical
properties. The superplasticizer concentration was found to play also an important role on the
mechanical behaviour of the matrix. A very small deviation of the evaluated test samples was
noticed as well as a local peak in 0.8 %wt. concentration that gave the best results also in terms
of compressive and tensile strength.
(a)
(b)
Figure 1: Effect of MWCNTs concentration on (a) electrical resistivity and (b) on flexural strength of investigated nano-reinforced cementitious materials.
In addition, the results have showed that the MWCNT concentration plays an important role on
the matrix reinforcement capability, Fig.1. Around 0.2 %wt. of MWCNTs is needed to better
reinforce the matrix in terms of decreased electrical resistivity, increased flexural strength and
to achieve optimum properties in terms of piezoresistivity.
The results of the small coupons were also exploited for extrapolation purposes. The above
MWCNTs concentration was used to construct a typical joint currently used in the Restoration
of Parthenon in Acropolis. A titanium element and nano-reinforced mortar was used to construct the joint of Fig.2 that was subsequently tested under shear loading. Preliminary results
show that the stress/strain variations in the gauge area of the nano-reinforced mortar can be
detected with the electrical change method.
154
ISBN: 978-960-98739-8-7
(a)
(b)
Figure 2: Photograph of the joint when (a) being constructed and (b) mechanically tested under shear.
Acknowledgements
The authors acknowledge the financial support of the EU (European Social Fund – ESF) and
Greek national funds through the Program “Support newly established firms in their research
and development activities”, Research Funding Program: “Monitoring of the structural integrity
of restored structural parts in ancient monuments of cultural heritage by employing hybrid materials reinforced with CNTs”.
References
1. Konsta-Gdoutos, M.S., Metaxa and Z.S. and Shah, S.P., 2010, Multi-scale mechanical and fracture char
acteristics and early-age strain capacity of high performance carbon nanotube/cement nanocomposites.
Cement and Concrete Composites, 32 (2), 110-115.
2. Kourkoulis, S.K., Mentzini, M. and Ganniari-Papageorgiou, E., 2013, Restored marble epistyles under bending:
A combined experimental and numerical study. Int. J. Architectural Heritage, 7(1), 89-115.
3. Metaxa, Z.S., Seo, J.W.T, Konsta-Gdoutos, M.S, Hersam, M.C. and Shah, S.P., 2012, Highly concentrated carbon
nanotube admixture for nano-fiber reinforced cementitious materials. Cement and Concrete Composites, 34,
612-617.
ISBN: 978-960-98739-8-7
155
Hydroelastic analysis of VLFS elastically connected to the seabed
in shallow wave conditions
E. Karperaki1*, K. A. Belibassakis1 , T. K. Papathanasiou2 and S. I.
Markolefas3
1
2
3
School of Naval Architecture and Marine Engineering, National Technical University
of Athens, Greece *[email protected]
DICAM, University of Trento, Italy
Department of Mechanical Engineering, Technological Educational Institute
of Central Greece
Keywords: VLFS, shallow water, FEM, transient hydroelastic analysis.
Abstract
The anti-motion effect of elastic connectors on the hydroelastic response of a thin, elastic floating plate over shallow bathymetry is studied. The classical Euler-Bernoulli strip, featuring a series of elastic connectors with the seabed along its length, is coupled with the linearized shallow
water equations. A higher order finite element scheme is employed for the numerical solution of
the equivalent variational problem. Numerical results exhibit the potential of the methodology
as a simple tool to assess the response mitigating effect of strip –connector configurations.
Introduction
Recent advancements in marine technology along with the increasing need for commercial
space, especially near harbors, has led to the development and Very Large Floating Structures
or VLFS. In turn, the rise of VLFS created the need for versatile computational tools that will
aid in their robust design. The present work focuses on pontoon-type VLFS which are essentially floating, thin plates resting on the water surface (Wang et al, 2008). The large horizontal
dimensions of such structures provide the much desirable commercial space away from the
condensed shore. The small thickness to length ratio inherent in pontoon-type VLFS renders
hydroelastic effects dominant of rigid body motions. A combination of mooring, breakwater
and anti-motion devices ensure the safe operation of the floating facilities, avoiding drift and
extreme vibration. In Khabakhpasheva and Korobkin (2002) the response mitigating effect of
an elastic spring connecting the structure with the seabed is examined while Karmakar and
Guedes Soares (2012) study the time harmonic scattering of gravity waves by a moored strip,
floating over shallow water.
In the present contribution, the transient hydroelastic response of a floating, elastic strip elastically connected to the seabed will be examined in shallow wave conditions. The plate is
modelled as an Euler-Bernoulli thin strip while the linearised shallow water are used for the
hydrodynamic modelling.
Governing Equations
In the present section, a floating elastically connected with the sea bottom boundary is considered (see Fig. 1). The strip is assumed to be connected at both free edges, while a series of
elastic connectors (N − 1) are distributed along its length. The connectors resemble mooring
lines with stiffness and damping coefficients kn and cn respectively.
156
ISBN: 978-960-98739-8-7
Figure 1 Elastic strip featuring a series of elastic connectors with the seabed.
Moreover, the plate extends infinitely in the direction vertical to the page, hence allowing the
treatment of the floating body configuration in the xz plane. The middle hydroelasticity dominated region S0 and the free-surface regions S1 and S 2 . Notably, the water surface elevation and
the plate deflection coincide in S0 , henceforth referred to as upper surface elevation and denoted as η (x, t ). The plate with density ρ p floats over a layer of inviscid and irrotational fluid with
density ρ w . The previous assumptions allow for the definition of velocity potential functions in
each sub region, namely ϕi , i = 0,1, 2 . The depth function is given by b (x ) = h (x ) − d (x ) , with
h (x ) being the variable depth function and d (x ) the strip draft. The domain is subdivided into
three regions. Finally, the edges of the strip are elastically connected to the seabed
Using
the
non-dimensional
variables
x = x / L
,η = η / L ,
t = g 1/2 L−1/2t
and
ϕi = g −1/2 L−3/2ϕi , for i = 0,1, 2 and dropping tildesm, the initial boundary value problem is
written as,
N


M (x )η + (K (x )η xx )xx + η + ϕ0 = ∑ δ ( x − xn ) knη + cnη , x ∈ S0 ,
η + (B (x )ϕ0 x )x = 0, x ∈ S0 ,
n=2
(
)
ϕ1 − (B (x )ϕ1x )x = 0, x ∈ S1 ,
ϕ2 − (B (x )ϕ2 x )x = 0, x ∈ S 2 ,
(1)
(2)
(3)
(4)
where M ( x) = m( x) ρ w−1 L−1 with m( x) being mass per unit length of the strip, K (x ) = D ρ w−1 g −1 L−4
where D is the flexural rigidity of the floating plate and B ( x) = b( x) L−1 . The far field conditions are given as,
ϕ1x ( x → −∞, t ) = 0
and
ϕ2 x ( x → ∞, t ) = 0.
(5)
At the interfaces between subregions, mass and energy conservation principles yield the following matching conditions,
B(0− )ϕ1x (0− , t ) = B (0+ )ϕ0 x (0+ , t ) , B(1− )ϕ0 x (1− , t ) = B(1+ )ϕ2 x (1+ , t )
(6)
ϕ1 (1− , t ) = ϕ0 (1+ , t ) , ϕ0 (1− , t ) = ϕ2 (1+ , t )
(7)
Since elastic connectors are placed at the free edges of the strip, the local bending moment and
shear force are given by
ISBN: 978-960-98739-8-7
157
Kη xx
x =0
= 0 , Kη xx x
Kη xx
x =1
= 0 , Kη xxx



=
−
k
1η (0, t ) − c1η (0, t )
x =0


= k2η (L, t ) + c2η (L, t ),
x =1
(8)
(9)
At the beginning of time, the fluid and body are at rest while a free surface disturbance G ( x)
begins to propagate in S 2 . The initial-boundary value problem is supplemented by the following initial conditions,
η ( x, 0) = η ( x, 0) = ϕ0 x = 0, x ∈ S0 and ϕ1 = ϕ1 = 0, x ∈ S1 , ϕ2 = 0, ϕ2 = −G ( x), x ∈ S 2
(10)
Variational formulation. In order to derive the variational formulation Eqs. (1)-(4) are multiplied by the weight functions v ∈ H 2 ( S0 ) , − w0 ∈ H 1 ( S0 ) , w1 ∈ H 1 ( S1 ) , and w2 ∈ H 1 ( S 2 ) ,
respectively (where H denotes the Sobolev spaces in the corresponding intervals). Performing
integration by parts and imposing the farfield, boundary and interface conditions the equivalent
variational problem is formed
1
1
0
0
1
∫ Mvηdx + ∫ vϕ dx − ∫ w ηdx + ∫
0
0
0
0
−∞
∞
w1ϕ1dx + ∫ w2ϕ2 dx +
1
+ a (η , v) + b0 (ϕ0 , w0 ) + b1 (ϕ1 , w1 ) + b2 (ϕ2 , w2 ) + q (η , v) + c(η , v) = 0,
where the bilinear functionals in Eq. 10 are given by,
1
1
(10)
0
a (η , v) = ∫ (Kvxxη xx + vη )dx b0 (ϕ0 , w0 ) = ∫ w0 x Bϕ0 x dx b1 (ϕ1 , w1 ) = ∫ w1x Bϕ1x dx
0
0
−∞
,
,
b2 (ϕ2 , w2 ) = ∫
∞
1
N +1
N +1
c(η , v) = ∑ v( xn )cnη ( xn , t ).
w2 x Bϕ2 x dx q (η , v) = ∑ v( xn )knη ( xn , t )
n
=
1
n =1
,
and
Finite element formulation. For the numerical solution of Eq. 10 a higher order finite element scheme is employed. A special hydroelastic element (Papathanasiou et al, 2014), featuring
quantic Hermite polynomials for the interpolation of η ( x) in S0 and 4th order Lagrange shape
functions for ϕ ( x) , in Si , i = 0,1, 2.
Numerical Results and Discussion
In the present section, the mitigating effect of the employed elastic connector will be investigated for a flat bottom profile ( kn = k , cn = c ). The depth is taken as 8 m , the length and
thickness of the strip are assumed to be 200 m and 2 m respectively. The initial excitation is
provided by a wave packet of the form η ( x, 0) = A exp[−α ( z − z0 ) 2 ]cos(k0 z ) , with A = 0.4 m
, α = 0.0013m −2 and k0 = 0.125 m −1. A parametric study was conducted for a three-connector
configuration (where the third connector was placed at the middle of the strip). Fig. 3 shows
the maximum absolute upper surface elevation obtained over the course of time for a range of
nondimensional stiffness and damping coefficients. Maximum deflection is observed when the
freely floating case is approximated (negligible k and c ) observed that response mitigation is
best achieved by the three-connector configuration for c = 0.01 and k = 0.05 . In Fig. 6 the
same analysis is carried out for the maximum absolute bending moment. Minimum values are
obtained for c = 0.01 , stressing the importance of connector parameter optimization. Finally,
it is noted that when the system becomes over stiffened (k>10) the examined damping coefficients yield the same effect.
158
ISBN: 978-960-98739-8-7
Concluding Remarks
In the present contribution, the transient hydroelastic response of a floating elastic, thin strip
elastically connected to the seabed is studied by means of the higher-order finite element
Figure 2 Semi-log plot of the maximum absolute elevation for a three connector configuration
Figure 3 Semi-log plot of the maximum absolute bending moment for a three connector configuration
scheme. The anti-motion effect of a three connector system was examined in a numerical example, stressing the importance of connector parameter optimisation. It is finally noted that, the
present methodology provides a simple computational tool, able to assess the response mitigating capabilities of elastic strip- connector configurations.
Acknowledgements
The present work has been supported by the project HYDELFS funded by the Operational
Program “Education and Lifelong Learning” of the National Strategic Reference Framework
(NSRF 2007-2013) - Research Funding Program ARHIMEDES-III: investing in knowledge
society through the European Social Fund.
References
1. Karmakar, D. and Guedes Soares , C, 2012, Scattering of gravity waves by a moored finite floating elastic
plate. Applied Ocean Research, 34, 135-149.
2. Khabakhpasheva, T. I, and Korobkin, A. A., 2002, Hydroelastic behaviour of compound floating plate in
waves. Journal of Engineering Mathematics, 44, 21–40.
3. Papathanasiou, T. K., Karperaki A., Theotokoglou, E. E. and Belibassakis, K., 2014, A higher order FEM for
time-domain hydroelastic analysis of large floating bodies in an inhomogeneous shallow water environment.
Proceedings of the Royal Society A, 471, 20140643.
4. Wang, C. M., Watanabe, E. and Utsunomiya, T., 2008, Very large floating Structures (London, U.K.: Taylor
and Francis)
ISBN: 978-960-98739-8-7
159
160
ISBN: 978-960-98739-8-7
Volume 1 / Topic B: Earth & Environmental Sciences
Topic B
Earth & Environmental Sciences
ISBN: 978-960-98739-8-7
161
162
ISBN: 978-960-98739-8-7
Coastal Transport Integrated System (Co.Tr.I.S)
Dimos N. PANTAZIS1*, Panagiotis STRATAKIS1, Vassilis MOUSSAS1,
Elias LAZAROU1, Dimitris STATHAKIS1,2, Eleni GKADOLOU1,
Charalampos KARATHANASIS1, Costas POLITIS, Anna Christina
DAVERONA1, Vassiliki KYRIAKOPOULOU1, Eleni BABALONA1, John
CHRYSOULAKIS1, Maria PARALIKA1, John TYRINOPOULOS1
1
2
S.O.C.R.A.T.E.S. (Society for Organizations, Cartography, Remote sensing/Road Design
and Applications using Technology/Transport Engineering on Earth and Space) Research
laboratory, Department of Civil & Survey Engineering, Technological Educational Institution
(TEI) of Athens, Greece,
University of Thessaly
*[email protected]
Keywords: Coastal transport modelling, spatial information systems, coastal lines
design
Abstract
Costal Transport Integrated System (Co.Tr.I.S) is an integrated spatial information system for
the optimal design of coastal transport lines. The main objective of this extended abstract is
to present a specific case of “Land”-“Sea” GIS convergence, in the frame of an innovative research project developed for the maritime environment.
Introduction
Co.Tr.I.S combines the graph theory and the sale’s man problem, as well as optimization tools
and evolutionary methods like Genetic Algorithms (GAs) and game theory techniques such as
the Nash Equilibrium. The final aim is to support the Co.Tr.I.S users with an arsenal of decision
support tools that will help any decision maker to design improved and optimal maritime networks in terms of distance, cost, time, safety or any other goal of interest (Lekakou, 2007). In
the next section, a summary of Co.Tr.I.S is presented. The third section discusses the system’s
functions development and implementation. The fourth section analyses the optimization tools.
The last section discusses the future steps and the research perspectives.
Co.Tr.I.S description
Co.Tr.I.S. is an integrated spatial information system which will to contribute to a more effective design of coastal transport lines in Aegean Sea, Greece. Nevertheless, it may be applied
to any maritime environment and for any type of maritime transport. The system is based on a
Geographic Information System (GIS) software, adds-on traffic management and other applications (mapping, statistics, design of new costal lines etc.) (Figure 1).
The system design is based on MECOSIG method (Pantazis and Donney, 1996) and an innovative approach combining network (vector) analysis and spatial analysis of raster surfaces (e.g.
representing the cost), game and graph theory including Nash equilibrium and salesman problem. Coastal transport network is stored in a geodatabase using a custom transportation data
model (Pantazis et al., 2013). The data model allow maintaining all aspects of the coastal lines
transport network such as one way-two way rules, inter-island costal lines configurations, intersection attributes and geometry, passengers, cars, trucks or goods transport and other modes
and schedules (Pantazis et al., 2014; Sturmey at al., 1994).
ISBN: 978-960-98739-8-7
163
The system’s users are divided into five different categories (groups). Each group will use a
slightly different version of application that is adjusted to their needs.
REQUIREMENTS
New coastal lines
Modifications of existing
coastal lines
Proposals for new coastal
lines
Proposals for
modifications of existing
coastal lines
FUNCTIONS
DATA
Ports
Nodes
Coastal lines
Passengers
Airports
Ships
Population
Coastal lines design,
modifications,
proposals
Information / Data /
Queries
Spatial dimension of
coastal lines network
Statistics / Thematic
maps / Graphs
Sale's man problem
Nash equilibrium
Games theory
Figure 1: Co.Tr.I.S. requirements data and functions
The users are: Local Authorities, Ship-owners union, Coastal lines companies, Ministry of Marine and Aegean (Policy makers) and Passengers (Table 1). For the moment we focus in the
three potential users which are: Local Authorities, Coastal lines companies, Ministry of Marine
and Aegean (Policy makers).
Ministry
Local Authorities
Maritime Companies
Ship-owners Union
Population Passengers Tourists
Observations /Actions
Observations
Actions
•Lack of coastal lines between islands-mainland and
islands-islands
•Lack of frequent itineraries
•Call for new coastal lines with financial support
•Insufficient and long itineraries
•"Must" : − Coastal lines for all islands
•Problems with patients transport
− Ports infrastructure improvement
•Lack of optimal connections
•Proposals for new coastal lines creation /modification
•Lack of connection with neighbor islands /ports in
of existed coastal lines
mainland
•"More" + "Must": − More frequent itineraries
− More direct Lines
− More connections
•Lack of specific coastal lines analysis tool
•Weak sustainability of coastal lines
•Conflicts
•No sustainable coastal lines
•Lack of coastal lines between islands-mainland and
islands-islands
•Lack of frequent itineraries
•Insufficient and long itineraries
•Problems with patients transport
•New profitable coastal lines design
•"Competition", profit, cost-benefits for each coastal line
•Consultation - Information - Nash Equilibrium - Cartel
•Decrease of passengers number
Table 1: Users and Actions
164
ISBN: 978-960-98739-8-7
System’s Functions Development and Implementation
The initial analysis of the system, establishes the methodological framework on which the
spatial and alphanumerical database schema was based. Using this framework we developed
an original set of system’s rules concerning the coastal transportation network and its operation. These rules have been transformed into relations between the spatial and alphanumerical
entities of the database (Pantazis et al, 2014). E.g. Rule 19: One shipping line can include 2 to
N nodes. Rule 20: A node may be the start (Initial departure node or Intermediate destination
node) or the end (Final destination node, Destination node, Intermediate destination node) of 1
to N partial shipping lines.
The Co.Tr.I.S. system includes eight subsystems (see Figure 2) and is based on the ArcGIS platform, and this actually means that the data, geographic or not, will be hosted in a Geodatabase.
Through this environment, the system has the ability to store any kind of information that it
needs. Along with the geographic information, there are multiple tables that contain statistical
information. This information may refer to the passengers that embark or disembark on each
port, the capacity of each ship, the ticket fare etc. All these datasets can be used in conjunction
with the geometries (spatial representations) in order to enrich the system’s calculation efficiency but also its prediction capabilities.
Figure 2: Co.Tr.I.S. subsystems
Optimization Tools
In order to support user decisions (Giziakis at al., 2006), a set of optimization tools are included
in Co.Tr.I.S. The high design detail of the Coastal Transport System, the complex relationships and the large number of realistic parameters and actors used in this work, create a Nondeterministic Polynomial (NP-hard) optimization problem that cannot always be solved within
ISBN: 978-960-98739-8-7
165
acceptable computer time by exact algorithms. Therefore, the main Optimization Module, is
based on evolutionary techniques that search only a part of the vast solution space and converge
near the optimal solution much faster. In addition to the Evolutionary techniques, more classic
methods such as ‘shortest path’ or ‘travelling salesman’ are also available to the user, during the
complex process of coastal network design.
Concluding Remarks
The development of Co.Tr.I.S. prototype system was summarily presented including the system’s spatial databases, maps, system’s functions, the different modules and the system interface. The interoperability with other Information Systems is still under investigation. Information concerning the itinerary of a specific ship on a selected day of the week from the beginning
to the end of the season is currently under development.
Although some specific modules of the system are designated for specific potential users, a
number of modifications are scheduled in order to operate Co.Tr.I.S. as an integrated platform
enabling greater collaboration between the potential users (local authorities, Ministry, maritime companies). Two specific transportation modeling software adds-ons (Sugar and Cube) are
still under investigation concerning their potential utility in Co.Tr.I.S. The technical problems
concerning the full integration of the statistical treatments and the optimization module can be
resolved using a number of strategies.
Acknowledgements
This research has been co-funded by the European Union (European Social Fund) and Greek
national resources under the framework of the “Archimedes III: Funding of Research Groups in
TEI of Athens” project of the “Education and Lifelong Learning” Operational Programme.
References
1. Giziakis K., Paravantis J. A., Michalochrista M. and Tsapara A., 2006, Optimal Operation of Passenger
Shipping in The Aegean, International Conference “Shipping in the era of Social Responsibility” Argostoli,
Cephalonia, Greece, 14-16 September 2006.
2. Lekakou, M. B., 2007, The eternal conundrum of Greek coastal shipping. In: Maritime Transport: The Greek
Paradigm, edited by A. A. Pallis, Oxford: Elsevier, pp. 257-296.
3. Pantazis, D. and Donnay, J.-P. Conception des SIG Méthode et formalisme, Editions Hérmès, Paris, 1996.
4. Pantazis, D.N., Stratakis, P., Daverona, A.C., Gkadolou, E., Lazarou, E. and Babalona, E., 2014, Coastal
Transport Integrated System: Spatial Database Schema, Metadata and Data Dictionary, 2nd International
Conference On Advances in Computing, Electronics and Electrical Technology - CEET 2014, December 2021, 2014, Malaysia.
5. Pantazis, D.N., Stratakis, P., Karathanasis, C. and Gkadolou, E., 2013, Design of a Coastal Transport Integrated
System: Preliminary System Specifications and Data Collection for the Aegean Sea Islands. 13th International
Conference, Ho Chi Minh City, Vietnam, June 24-27, 2013, Proceedings, Part IV, Lecture Notes in Computer
Science Volume 7974, pp. 268-283.
6. Sturmey, S.G., Panagakos, G. and Psaraftis H.N., 1994, Institutional and Socio-economic Issues in Greek
Ferry Services. Second European Research Roundtable Conference on Short sea Shipping, Athens.
166
ISBN: 978-960-98739-8-7
3D Heritage Recording using Terrestrial Laser Scanning Techniques
Vassilios D. ANDRITSANOS1, Michail GIANNIOU1, Vassilios
PAGOUNIS1*, Maria TSAKIRI2
1
2
Department of Surveying Engineering, Technological Educational Institute, Athens, Greece,
*[email protected]
School of Rural & Surveying Engineering, National Technical University of Athens, Greece
Keywords: Heritage recording, Terrestrial laser scanning (TLS), Geodetic techniques,
3D modeling, Documentation.
Abstract
The importance of heritage recording and documentation with optical ranging sensors is well
recognized at international level. The continuous development of new sensors such as terrestrial laser scanning (TLS), data capture methodologies and multi-resolution 3D representations, contributes significantly to the digital 3D documentation, mapping, conservation and
representation of heritages. This article discusses the use of TLS for 3D data capture and 3D
modeling techniques in the documentation and modeling of two significant heritage sites and
monuments.
Introduction
The creation of 3D models of heritage and archaeological objects and sites in their current state
requires powerful techniques capable of capturing large volumes of data which are then used to
digitally model the fine geometric and appearance details of such sites. Digital recording is demanded as the heritages are globally experiencing natural or human driven deterioration requiring preservation. Furthermore, the digital documentation record is important so that it can be
transferred to future generations. Nowadays, remote sensing technologies and methodologies
for cultural heritage 3D documentation and modeling allow the generation of very realistic 3D
products (in terms of geometric and radiometric accuracy) that can be used for many purposes,
such as historical documentation, digital preservation and conservation, cross-comparisons,
monitoring of shape and colours, simulation of aging and deterioration, virtual computer-aided
restoration, multimedia museum exhibitions, visualization and so on (e.g. Bernadini et al., 2002;
Gruen et al., 2004; Doneus and Neubauer, 2005; El-Hakim et al., 2008; Bruno et al., 2010).
Today, state-of-the-art remote, non-invasive optical techniques are available for mapping purposes and digital recording of cultural heritage. Generally, non-invasive optical recording sensors are divided into passive and active systems. Passive sensors (e.g., digital cameras) deliver
image data which are then processed with some mathematical formulations to infer 3D information from the 2D image measurements (e.g. Sturm et al., 2011). On the other hand, active
sensors (e.g., laser scanner or radar) can provide data directly for 3D information or ranges.
Terrestrial active and passive sensors employed to derive 3D shapes are often referred to 3D
imaging techniques (e.g. Sansoni et al., 2009).
Almost fifteen years after the introduction into the market of the first TLS, nowadays a great
variety of sensors is available. TLSs work from very short ranges (few cm up to a few km) in
accordance with surface properties and environment characteristics, delivering 3D data with
accuracy from some mm to few cm.
ISBN: 978-960-98739-8-7
167
The main challenges in 3D surveying and modeling of large sites or complex objects using TLS
arise in every phase, from the data acquisition to the visualization of the achieved 3D results.
3D modeling should be intended as the generation of structured 3D data from the surveyed unstructured data and it consists of geometric and appearance modeling. The geometric modeling
deals with data registration and processing (editing, cleaning, meshing) while the appearance
modeling deals with texturing, blending and rendering simplification. In general, the 3D modeling pipeline involves the following steps: (i) selecting the appropriate methodology (sensor, hardware, software) and data processing procedure; (ii) designing the proper production
workflow, guaranteeing that the final result is in accordance with all the given technical specifications; (iii) speeding up the data processing time with as much automation as possible but
always with the accuracy as primary goal; (iv) being able to fluently display and interact with
the achieved 3D model. This paper deals with the 3D modeling pipeline as discussed above and
gives two examples of 3D surveying and modeling of heritage sites and monuments.
Case Studies
The reported examples show the potentialities of the modern surveying technologies to digitally
document and preserve landscape and heritage sites as well as share and manage them. In both
case studies TLS systems is the main source for acquiring 3D data.
a) The ancient theatre of Thebes, Fthiotidas, Greece
The ancient theatre of Thebes Fthiotidas, the excavation of which is not yet completed, is located on the east side of the ancient city of Thebes Fthiotidas, on the natural slope of the hill
“Castle”, east of the village Mikrothives. The ancient theatre is Hellenistic but there is also the
Roman phase, which converted it to Arena Theatre. With a maximum capacity of 3000 spectators, the theatre housed ancient drama, music competitions, and during the Roman Era gladiator
fights. The theatre was investigated in 1992-93 under a project of the Hellenic Military Aviation. Research in the theatre has not been completed and of the main aims is the 3D documentation of this important monument, which subsequently needs restoration works, in order to make
it available to visitors.
Fig. 1a. The ancient theatre of Thebes Fthiotidas
with the positions of the geodetic network (red) and
TLS (orange).
Fig. 1b. Point cloud data of the theatre.
The documentation phase included a combination of geodetic and TLS measurements. Initially,
a control survey network was established around the site to provide the reference coordinate
system and the necessary geodetic information with an accuracy of 0.7cm in the horizontal and
0.9cm in the vertical component (Fig. 1a). The network establishment was based on GNSS
168
ISBN: 978-960-98739-8-7
measurements using geodetic receivers and referenced to the national geodetic system (Greek
Geodetic Reference System 1987 - GGRS87). Also, a detailed total station surveying data collection was performed for specific sections of the theatre which were compared with the TLS
data (Fig. 1b). The TLS documentation required 28 scans and the georeference was achieved
using reflective spherical targets at an accuracy of 1cm. The accuracy of the final registered
cloud is 1.4cm. Using a CAD environment, the TLS data were imported in order to create a 2D
detailed plan of the theatre (Fig.2a) and sections at selected parts (Fig. 2b).
Fig. 2a. The 2D topographic plan using TLS data.
Fig. 2b. Detail of a 2D plan derived from TLS data.
b) The ancient tower of Agia Marina, Kea, Greece
The ancient monument of Agia Marina is a five-storeyed, square tower around 20m high and is
a UNESCO protected heritage. It is made of local schist and marble in the Hellenistic period.
The north side of the monument is preserved in very good condition but the rest has collapsed
and required restoration. For this reason, a detailed 3D model was created using TLS data.
Fig. 3a. The TLS data of the monument (with
scaffoldings).
Fig. 3b. 3D model derived from TLS data.
Prior the TLS data acquisition, a control survey network was established using GNSS techniques. A total number of 16 scans were acquired due to the fact that the monument was surrounded by scaffoldings (Fig 3a). Using the Cyclone software the data registration was achieved
at 1cm (www.leica-geosystems.com). The data were cleaned from the noise with the point cloud
having about 95 million points (Fig. 3b). Using the 3D Reshaper software (www.3dreshaper.
ISBN: 978-960-98739-8-7
169
com) the cleaned registered point cloud was then used to create a 3D model (Fig 3b). During the
TLS data acquisition, digital photos were also taken which were used to create a rendered 3D
model (Fig 4a). This was performed in the 3D Reshaper software using the relevant texture map
facilities. The final product was an orthophoto mosaic which gives not only the true geometric
description but also provides in detail the current appearance of the monument so it can be used
for restoration and preservation purposes (Fig. 4b). Finally, a virtual walking through video was
also created which can be used for visualization purposes.
Fig. 4a. View of the rendered model.
Fig. 4b. Detail of the orthophoto.
Concluding Remarks
This article discussed the use of 3D measurements sensors and specifically TLS techniques
used for surveying, mapping, digital documentation and 3D modeling applications in the heritage field. The 3D modeling pipeline was also reported through two examples from heritage
sites. Despite the fact that the 3D documentation is not yet widely used in the heritage field,
the reported examples show the potentialities of the modern surveying technologies to digitally
document and preserve sites and monuments as well as share and manage them. Clearly, the
wide access of archaeologists to geospatial technologies can aid for more effective decisions.
References
1. Bernardini, F.; Rushmeier, H.; Martin, I.M.; Mittleman, J.; Taubin, 2002, Building a digital model of
Michelangelo’s Florentine Pieta. IEEE Comput. Graph. Appl. 22, 59-67.
2. Bruno, F.; Bruno, S.; De Sensi, G.; Luchi, M.L.; Mancuso, S.; Muzzupappa, M., 2010, From 3D reconstruction
to virtual reality: A complete methodology for digital arch. exhibition. J. Cult. Herit. 11(42-49).
3. Doneus, M.; Neubauer, W., 2005, 3D Laser Scanners on Archaeological Excavations. Proc.CIPA 2005 XX
International Symposium, Torino, Italy, 26 Sept.–1 Oct., Volume 34(5/C34/1), pp. 226-231.
4. El-Hakim, S.; Beraldin, J.; Remondino, F.; Picard, M.; Cournoyer, L.; Baltsavias, E., 2008, Using Terrestrial
Laser Scanning and Digital Images for the 3D Modeling of the Erechteion, Acropolis of Athens. Proc. DMACH
Conf. on Digital Media and Applications in Cultural Heritage, Amman, Jordan, 3–6 Nov, pp. 3-16.
5. Gruen, A.; Remondino, F.; Zhang, L., 2004, Photogrammetric reconstruction of the Great Buddha of Bamiyan.
The Photogrammetric Record 19, 177-199.
6. Sansoni, G.; Trebeschi, M.; Docchio F., 2009, State-of-the-art and applications of 3D imaging sensors in
industry, cultural heritage, medicine, and criminal investigation. Sensors 9, 568-601.
7. Sturm, P.; Ramalingam, S.; Tardif, J.-P.; Gasparini, S.; Barreto, J., 2011, Camera models and fundamental
concepts used in geometric Computer Vision. Found. Trend. Comput. Graph. Vis., 6, 1-183.
170
ISBN: 978-960-98739-8-7
A Shale Brittleness Index Based on Shale Peak Strength and Strain
Yanhu Tan1, Fenglin Xu1,2, Lin Chen3, Qiao Chen1, Linhong Zhu1,
Lisha Wang1 and Wen Nie4*
Chongqing Institute of Green and Intelligent Technology, Chinese Academy of Sciences,
Chongqing, 400714, China
2
State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation, Southwest
Petroleum University, Chengdu, 610500, China
3
School of Sciences, Southwest Petroleum University, Chengdu, 610500, China
4
State Key Laboratory of Geo-Hazard Prevention and Geo-Environment Protection,
Chengdu University of Technology, Chengdu, 610059, China.
1
Keywords: shale brittleness index, stress-strain curve, rock mechanics, evaluation
Abstract
In this study, a shale brittleness index for evaluating the brittleness of shale is put forward. This
brittleness index, which is easily acquired by uniaxial (triaxial) compression testing, is determined by the shale peak strength and strain in the meantime. The shale samples from Maxi,
Sichuan Basin, China are used to validate our shale brittleness index by triaxial compression
testing. The results indicate that the degree of shale damage is in agreement with the evaluation
provided by our shale brittleness index.
Introduction
Shale gas is the second largest unconventional energy resource after heavy oil. The production
of shale gas involves drilling, fracturing, and gathering by injected fluids. The brittleness of the
shale plays an important role, especially in the drilling and fracturing of shale. Unfortunately, a
universally accepted brittleness concept and a brittleness measurement method are still lacking.
A brittleness concept that is commonly used in rock mechanics is the ratio of rock compressive strength to tensile strength (Hucka & Das, 1974). Hajiabdolmajid et al. (2003) relate the
brittleness of failing hard rocks to a strain-dependent brittleness index which characterizes the
entire failure process of rock and accounts for the mechanisms involved in inducing inelastic
strains (damage) inside the failing rock. Another rock brittleness index (BI) considered changes
in hardness dependence and stress (Honda & Sanada, 1956; Copur et al., 2003). In our study, a
peak strength and strain-dependent shale brittleness index is developed to evaluate the brittleness of shale. This index considers both peak strength and strain factors and the parameters are
more easily acquired compared to the hardness and variable stress indexes. The shale samples
from Maxi, Sichuan Basin, China are used to validate our shale brittleness index by triaxial
compression testing.
Peak strength and strain-dependent shale brittleness index
The shale brittleness index is as shown in Eq. 1.
εp
BI = 100ke
σp
(1)
where ε p and σ p are the peak strain and peak strength (see Fig. 2(a)); k is an adjustment coefficient which is used to adjust the BI distribution range; 100 is to constrain the range of BI from
0 to 100.
ISBN: 978-960-98739-8-7
171
This brittleness index can explain the fact that under similar stress loads a higher peak strain
usually means there is a higher ductility but lower brittleness of the material, while a higher
peak strength of the material may mean that more energy is required to damage the material,
more fragments are produced, and therefore the brittleness is higher.
Validation of shale brittleness index
Some random shale samples from Maxi, Sichuan Basin, China are used to validate our shale
brittleness index by triaxial compression testing. Figure 1 shows the location information of the
production of the shale sample. Figure 2 indicates the results of triaxial compression testing and
damage to shale samples. Since the higher the magnitude of BI, the more brittle the rock, the
results indicate that the degree of shale damage is in agreement with the evaluation provided by
our shale brittleness index.
Figure 1. Shale sample location map of Maxi, Sichuan Basin, China: (a) location map of shale sample;
(b) tectonic map
172
ISBN: 978-960-98739-8-7
Figure 2. Fragmentation of shale and stress-strain curve: (a) BI = 86.94; (b) BI = 88.86; (c) BI = 77.70; (d) BI
= 98.11; (e) BI = 81.78; (f) BI = 52.75
Outlook
In the future, more uniaxial (triaxial) compression tests of shale samples will be carried out to
increase the reliability of the brittleness index. More factors including pore size, geometry of
the sample, and direction of bedding planes are considered to how they affect our brittleness
index.
Concluding Remarks
The peak strength and strain-dependent shale brittleness index can describe the brittleness of
shale. Aspects of this index that can be highlighted include the following:
(1) The brittleness index combines the peak strength and strain to evaluate
the brittleness of shale;
(2) The parameters of this brittleness index can be easily acquired just by uniaxial
or triaxial compression tests.
Acknowledgments
The research is supported by the Chinese National Natural Science Foundation (Grant No.
41502287), Chongqing Basic and Frontier Research Projects (Grant No. cstc2014jcyjA90008),
and Chongqing Basic and Frontier Research Projects (Grant No. cstc2014jcyjA90008).
References
1. Copur, H., Bilgin, N., Tuncdemir, H., & Balci, C. (2003) A set of indices based on indentation tests for
assessment of rock cutting performance and rock properties. Journal of the South African Institute of Mining
and Metallurgy, 103 (9), 589–599.
2. Hajiabdolmajid, V., & Kaiser, P. (2003) Brittleness of rock and stability assessment in hard rock
tunneling. Tunnelling and Underground Space Technology, 18 (1), 35–48.
3. Honda, H., & Sanada, Y. (1956) Hardness of coal. Fuel, 35 (4), 451–461.
4. Hucka, V., & Das, B. 1974. Brittleness determination of rocks by different methods. International Journal of
Rock Mechanics and Mining Sciences & Geomechanics Abstracts, 11 (10), 389–392.
ISBN: 978-960-98739-8-7
173
Urban run-off management in a Greek coastal city: Citizens’
awareness, attitudes and proposed solutions
Kalderis D.1*, Stavroulakis G.1 and Diamadopoulos E.2
Technological and Educational Institute of Crete, Department of Environmental and
Natural Resources Engineering, *[email protected]
2
Technical University of Crete, Department of Environmental Engineering
1
Keywords: urban run-off, storm water management, coastal city, water pollution
Abstract
The storm water quality of the coastal city of Chania, Greece was studied in the period 20122015 in order to determine potential drawbacks of the urban run-off management system and
develop solutions for sustainable protection of the main receiving body, the Venetian harbour.
During this study, a citizen’s survey was performed that highlighted the lack of knowledge with
respect to the environmental threats posed by storm water that ends up in the sea. In this work,
six important solutions for the sustainable management of urban run-off of Chania (applicable
also to other coastal cities) are proposed and discussed. Additionally, the initial results of the
citizen’s survey are provided, pinpointing the need for a comprehensive education and information strategy on urban run-off management issues.
Introduction
The sustainable environmental management of the coastal zone is a multi-component issue,
common in many coastal cities worldwide. Sea water quality can be affected by inland practices as well as intensive sea water use. In the framework of Archimedes III (Research Program
EPEAEK, financed by The European Union & Ministry of Education, Lifelong Learning and
Religious Affairs of Greece) the following project was carried out in the years 2012-2015: Decision support tool for the management of urban run-off in coastal cities (COASTSURF). The
project’s main objectives were a) to review of the current literature on the polluting nature of
sewer and storm water run-off, b) the qualitative assessment of the human activities and potential polluting sources and determination of sampling points within the sewer infrastructure of
the city of Chania, c) the quantitative assessment of the polluting load, throughout the sewer infrastructure as well as the sea outflow points, d) the development of a mathematical model that
would include the hydrodynamic behavior of the sea and will relate the quality of the coastal
waters to the urban run-off of the city and the climatic conditions and e) the determination of
the Best Available Practices (BAPs) for sustainable management of urban run-off (including
a citizen’s questionnaire reflecting the public opinion and trends on storm water threats and
potential solutions).
The urban runoff infrastructure in the city of Chania (Crete, Greece) has several outlets in the
Venetian harbor, which is the most important touristic attraction of the city and one of the most
significant of Crete. Storm water flow can carry dust, road depositions and other pollutants and
affect the sea water quality of the harbor. Leaks from the wastewater infrastructure or illegal
connections to the sewer further increase the polluting load reaching the sea. Additionally, the
heavy use of the harbor’s marina for recreational purposes (especially the summer months)
often leads to pollution incidents.
174
ISBN: 978-960-98739-8-7
After extensive sampling of the sea water from the Venetian port, it was determined that in
many cases the polluting load –especially microbiological indicators such as fecal coliforms
and enterococci - was considerably higher compared to guidelines values (Table 1). Therefore,
it was established that there is an environmental risk involved in the management of the Venetian harbor and actions for the protection of the sea water quality and marine life are required.
The necessary steps for sustainable protection of the Venetian harbor are the following:
1. Development of a digital map of the sewer infrastructure - More than just digital replicas,
these maps are linked to millions of bits of information, or attributes — the size of the pipes,
the dates they were built and repaired, what they were made of — that can be called up and
sorted with the click of a mouse. The maps can be updated instantly when water mains and
sewers are installed, removed or repaired.
2. Illicit discharge and connection detection program - Studies have shown that dry weather
flows from the storm drain system may contribute a larger annual discharge mass for some
pollutants than wet weather storm water flows. Detecting and eliminating these illicit discharges and connections (e.g. old sanitary tanks) involves complex detective work and close
cooperation among several parties.
3. Establishment of the total maximum daily loads (TMDLs) for the Venetian harbor and adjacent areas - TMDL is a calculation of the maximum amount of a pollutant that a waterbody
can receive and still meet water quality standards, and an allocation of that load among the
various sources of that pollutant. The TMDL process is important for improving water quality because it links the development and implementation of control actions to the attainment
of water quality standards. TMDLs are developed using a range of techniques, from simple
mass balance calculations to complex water quality modeling approaches. The degree of
analysis varies based on a variety of factors including, the waterbody type, complexity of
flow conditions, and pollutant causing the impairment.
4. Development of national regulation where special measures would have to be taken for
building permits that exceed 1000 m2. It is known that impermeable surfaces (concrete,
pavement) largely reduce the storm water that infiltrates to the ground, thus leading to larger
volumes of run-off. Therefore, urban run-off measures (such as rain gardens, “green” roofs
and others) should become a necessary component of building permits in order to reduce
the environmental impact of residential/commercial development.
Table 1. Polluting load in the seawater of the Venetian harbora [1]
pH
7.3-7.9
BOD5
(mg/L)
1-10
COD
DO
(mg/L) (mg/L)
3.4-31.8 4-8.5
NO3(mg/L)
2.6-6.6
NH4+
(mg/L)
0.08-0.4
PO43(mg/L)
0.03-1.2
FC
Enterococci
(cfu/100ml) (cfu/100ml)
2-4.200
2-530
Minimum and maximum values reported
5. Registration of the Venetian harbor to the European Sea Ports Organisation (ESPO) and
EcoPorts Network. Serving the principle of “ports-helping-ports” EcoPorts brought together a network of port professionals from several European ports committed to exchange
views and practices and to commonly work towards the improvement of the sector’s environmental performance in line with the principles of voluntary self-regulation. Membership
to the Organization offers the opportunity to use the well-established EcoPorts tools, SelfDiagnosis Method (SDM) and Port Environmental Review System (PERS)
ISBN: 978-960-98739-8-7
175
6. Information and education events for citizens and professionals - understanding how urban
run-off is created and the threats it poses, is fundamental for the successful implementation
of any hard or soft protective action. Press-conferences, interactive web pages, seminars
and workshops in schools are only a few of the dissemination tools available.
As part of the information and education strategy, a questionnaire was developed by the project
team and was delivered to 400+ citizens of the city. The objective of this questionnaire was to
assess their knowledge on storm water management and eagerness to contribute to potential
future actions towards protection of the Venetian harbor. The survey consisted of 14 questions.
Selected results can be seen in Figure 1a-f. Figure 1a shows that approximately 1/3 of the citizens did not know what urban run-off is. This is a clear indication that there is a knowledge gap
in this subject. This gap was confirmed in Figure 1b – only 55% of the citizens knew that storm
water ends up in the sea. It is interesting to note that about 1/3 of the citizens (34%) thought that
urban run-off is collected and treated by the wastewater treatment plant of Chania. Only 64%
of the citizens were aware of the fact that paints and solvents should be disposed of in the toilet
or kitchen sink and not in an open permeable or impermeable surface (Figure 1c). A green roof
and a rain garden are two widely available technologies for collection and partial treatment of
urban run-off. In recent years, green roofs have been increasingly popular in Greece, especially
in new residential and commercial buildings. This reflects in the results shown in Figure 1d,
where 83% of the citizens knew what a green roof is. However, this trend was reversed for the
rain gardens, where the same percentage were not aware of this technology. This result may be
explained by the high population density in all major Greek cities and the fact that most people
live in apartments and building blocks, where gardens and green spaces are almost completely
absent. Finally, Figure 1f provides some encouraging results with respect to future actions.
66% agreed to a 20 euro yearly charge in their water bill in order to maintain the sewer infrastructure quality.
Concluding remarks
The results obtained from the 3-year study of the urban run-off and sea water quality of the
coastal city of Chania, indicate the need for further action, including a comprehensive information and education strategy. The six storm water management actions proposed earlier are
applicable to any coastal city but their success will be limited if they are not accompanied by
educational initiatives that will target the citizens, professionals and non-professionals alike.
The survey results clearly indicate the need for these initiatives, most of which can be applied
at a voluntary basis at a low cost. Protecting coastal waters is not only a scientific issue, is more
of a complex political issue that requires the cooperation of people from different backgrounds
in order to be dealt with.
176
ISBN: 978-960-98739-8-7
a
b
c
d
e
f
Fig. 1a-f. Selected results of the citizens’ questionnaire on urban run-off management
Acknowledgements
This project is implemented through the Operational Program “Education and Lifelong Learning” action Archimedes III and is co-financed by the European Union (European Social Fund)
and Greek national funds (National Strategic Reference Framework 2007 - 2013). Project title:
Decision support tool for the management of urban run-off in coastal cities –COASTSURF.
References
1. Stavroulakis G., Kalderis D., Papafillipaki A. and Minou A., Distribution of sea water pollution in the Venetian
harbor, Chania, Greece. Proceedings of the 14th International Conference on Environmental Science and
Technology, Rhodes, Greece, 3-5 September 2015.
ISBN: 978-960-98739-8-7
177
Forest fires: a phenomenon with serious environmental
and other impacts
Georgia Tsakni
Economist, PhD Candidate, MA «Public Management» Panteion University,
ΜΑ «Environment and Health» National and Kapodistrian University of Athens,
BA in «Economics» University of Piraeus
Tel. 6938750794, E-mail: [email protected]
Abstract
Climate change is a multifaceted phenomenon of paramount importance. Forest fires are considered one of the most important environmental problems having both natural and manmade
causes. Both the environment and human health are affected by the occurrence of fires, mainly
due to the harmful effects of smoke.
Forest and ecosystems
The definition as established by the Greek Supreme Court (A.E.D.) and included as interpretative statement, almost intact in the Constitution is that:
“As a forest or forest ecosystem is defined the organic whole of wild plants with woody trunk
on the necessary area of ground which, along with their coexisting flora and fauna, are through
mutual interdependence and interaction, a particular biocommunity (forest biocommunity) and
special natural environment (forest derived). Forest cover exists when all the wild woody vegetation, either high or shrubbery, is sparse.”
On the planet there are two types of ecosystems: terrestrial and aquatic. Regarding terrestrial
ecosystems, their basis of classification is their vegetation, which characterizes them and depends on the climatic conditions1. Forest ecosystems are a particularly important part of the
2
natural wealth of our country, covering 49.3% of the total area of Greece

.
Sustainable management is based on principles of social justice between the generations to
come, in order to maintain the productivity of natural resources3.
Forest fire smoke
Wildfires are a controversial issue worldwide. The extent of forest fires has greatly influenced
the shaping of the environment and has disturbed biodiversity on a global scale.
Nevertheless, there are certain ecosystems, in which the fire is considered a natural process, as
the fire has been used by people as a natural land management tool, which in turn may have
cultural significance.
The lack of leadership policy measures is considered a problem, as are poor forest management
policies. All these contribute to the degradation of ecological sustainability and the increase in
the number of forest fires4.
1
Karvounis, S. (1995), «Environmental Management», Stamoulis, Thessaloniki
2
Kodosakis, D. (1994), «Natural Resources and Energy Management», Stamoulis, Thessaloniki
3
Kasouris, Th. And Athanasakis, A. (1996), «Environment, Ecology, Education» Publications Savalas,
Athens
4
Schmidt, D, Kaufmann, M, Myers, R, Alencar, A, Kearns, F, Johnson, D., Smith, J., Zollner, D., (2007),
«Fire, ecosystems and people: Threats and strategies for global biodiversity conservation », Sevilla - Espana Wildfire
178
ISBN: 978-960-98739-8-7
Environmental, social and economic impacts of forest fires
The impact of forest fires are a multifaceted phenomenon, which impacts both on human health,
the environment and on society. The multiple benefits role of vegetation in the urban environment have been repeatedly demonstrated.
Socially, the fires may affect tourism due to reduced traffic. The degradation of environmental
quality can also leads to an increase in unemployment due to the desertification of affected areas and the disappearance of agricultural occupations.
Finally, we should not overlook psychological impacts, as of people who lose their house and
their work may suffer psychological harm, including depression1.
In conclusion, among the environmental, economic and social impacts of forest fires are2:
• Evacuation of adjacent communities
• Destruction of cultural and archaeological sites
• Losses in soil, watersheds and water sources
• Air pollution and public health implications
Health Issues
Long-term exposure to particulate matter may increase susceptibility to infections and the emergence of respiratory symptoms, such as bronchiolitis and pneumonia3.
There are also several incidences of psychological problems in people who were close to a forest fire. The psychological condition of the patients inevitably led to increased consumption of
anti-anxiety pills and increased counseling4.
Also, there are several known incidents of poisoning from contaminated fish in forested areas.
Specifically, in the Amazon high levels of mercury have been reported. Exposure to mercury
can lead to mental retardation and insanity, and for this reason research has been intensified in
recent years into reducing exposure to this harmful metal5.
Additionally, firewood collected in the forest can present serious risks to the human respiratory
system. Therefore, the importance of forests varies for different groups of people living near
forest areas whether farmers, migrants living for long periods in forests, hunter - gatherers and
urban residents6.
Conclusions
The impacts of forest fires on the environment and human health are enormous. We must all
realize that we are part of an ecosystem and try to follow specific actions which will promote
viable and sustainable development. The environment in which the future generations will live
is a crucial issue and forest protection is our responsibility!
1
www.chiospress.gr
2
Birot. Y. (2009), EFI Discussion Paper, «Our Life with Wildfires: The view of science - policy», European Forest Institute
3
http://currents.plos.org
4
http://currents.plos.org
5
C.J.P. Colfer, D. Sheil, D. Kaimowitz and M. Kishi, 2006, «Forests and human health in the tropics: some
important connections», FAO Corporate Document Repository
6
C.J.P. Colfer, D. Sheil, D. Kaimowitz and M. Kishi, 2006, «Forests and human health in the tropics: some
ISBN: 978-960-98739-8-7
179
REFERENCES
1. Karvounis, S. (1995), «Environmental Management», Stamoulis, Thessaloniki
2. Kasouris, Th. And ATHANASAKIS, A. (1996), «Environment, Ecology, Education» Publications Savalas,
Athens
3. Kodosakis, D. (1994), «Natural Resources and Energy Management», Stamoulis, Thessaloniki
4. Birot. Y. (2009), EFI Discussion Paper, «Our Life with Wildfires: The view of science - policy», European
Forest Institute
5. C.J.P. Colfer, D. Sheil, D. Kaimowitz and M. Kishi, 2006, «Forests and human health in the tropics: some
6. Schmidt, D, Kaufmann, M, Myers, R, Alencar, A, Kearns, F, Johnson, D., Smith, J., Zollner, D., (2007),
«Fire, ecosystems and people: Threats and strategies for global biodiversity conservation », Sevilla - Espana
Wildfire
7. www.chiospress.gr
8. http://currents.plos.org
180
ISBN: 978-960-98739-8-7
Spatio-temporal cluster analysis of seismicity using a modified
density-based clustering algorithm
Dionysios MOUNTAKIS1*
1
University of Peloponnese, Dpt. of Computer Science and Technology, Tripoli 22100
*[email protected]
Keywords: cluster-analysis, dbscan, seismicity, evaluation, validation
Abstract
Statistical pattern analysis techniques is a common approach in modern seismicity. Our main
goal is to indentify natural underlying structural patterns in seismicity, with the use of densitybased clustering analysis. The main issue with seismic clustering is the evaluation of densitybased arbitrarily shaped clusters, since there is almost none validation criterion in literature
designed for density cluster analysis or to distinguish the presence of noise. A second problem,
that also arises, is the fact that different seismic clusters can very well be spatially overlapping,
appearing as a single cluster. However, that cluster usually encompasses families of events with
distinguished characteristics i.e. “classification of detected clusters into several major types,
generally corresponding to singles, burst-like and swarm-like sequences” (Zaliapin and BenZion, 2013). This report will examine the behavior of Density Based Spatial Clustering of Applications with Noise algorithm (DBSCAN) (Daszykowski et al, 2001) with the performance
of a modified DBSCAN, within the vicinity of the Hellenic seismic arc and the surrounding
Hellenic area. The modified DBSCAN utilizes weighing parameters that weight seismic events
depending on energy emission. The algorithm has been implemented on MatlabTM suite. Results will be compared and discussed in order to examine the substantial degree of their alleged
benefits.
Introduction
Recent results, even they are in question, have provided us with the ability to delineate the behavior of seismic activity within the Hellenic seismic arc, which up to this day, remains a challenge for Seismology. Until recent years, earthquakes were believed to be randomly occurring
events depending on the movement among the colliding continental plates. Earthquakes occur
when tectonic plates collide or when an accumulated amount of elastic energy at a specific area
along a regional fault exceeds a threshold causing a rupture. The sensible question that arises is
if these incidents can be predicted in space and time domains and what the approximate magnitude of the occurring events would be.
Recent studies have come to the conclusion that earthquakes are not a randomly occurring
event but follow a certain pattern (Vallianatos et al, 2013). The identification and analysis of
the spatiotemporal characteristics of such a pattern would provide better understanding of how
the mechanics and underlying physics of the earthquake phenomenon work. Use of modern
computational techniques, such as neural networks and clustering algorithms, will assist in
decoding such patterns.
DBSCAN has the advantage of identifying clusters of arbitrary shape, improving cluster scalability and efficiency. DBSCAN’s algorithm source code has been written by Daszykowski
(Daszykowski et al, 2001).
ISBN: 978-960-98739-8-7
181
Methodology
A brief reference to the DBSCAN algorithm has already been made. The classic DBSCAN
requires two input parameters, the Epsilon radius (Eps) of the Eps-neighborhood and the Minimum Number of Points (MinPts) that lie in this Eps-Neighborhood. Points belonging to different neighborhoods can either be density-connected, density-reachable or direct density-reachable. The summation of these formations creates a cluster, arbitrarily shaped or not (Ester et al,
1995, Ester et al, 1996, Daszykowski et al, 2001).
One of DBSCAN’s major disadvantages is its deficiency when it comes to data of different
density areas i.e. some areas of ‘thicker’ data than others (Drakatos and Latoussakis, 2001). Ideally, new sets of input parameters should be selected each time data density changes, instead of
globular variables. Since MinPts and Eps-radius combination cannot be chosen for the various
density formations independently, a number of separate clusters cannot be individually identified, resulting to spatial overlapping and cluster within cluster.
In an effort to provide a solution to the aforementioned problem, our approach encompasses
the relations between the aftershock duration and the magnitude of the main shock (Eq. 1) and
between the subsurface rupture length and the magnitude of the main shock (Eq. 2). To evaluate
the aforementioned parameters the following expressions are used (Drakatos and Latoussakis,
2001):
log(T) = 0.51M – 1.15 (1)
log(L) = 0.35M – 0.62 (2)
The key part of the proposed approach is that the parameter of Minimum Points (MinPts) has
been replaced by the earthquakes’ magnitudes and time of occurence, both normalized on the
spatial and temporal planes, i.e the algorithm calculates the normalized dimensionless values,
which lie inside a predefined Eps-radius. If the result exceeds a specified threshold, then an
Eps-neighborhood is formed, if not that point is either border or noise point. Then the cluster’s
expansion follows the same rules as the traditional DBSCAN.
Our model is inextricably associated with the emitted energies of main events and their aftershock sequences (Yang and Lee, 2004, Petersen et al, 2008, Vallianatos et al, 2013, Yeck et al,
2015). The classic DBSCAN algorithm has been modified in order to fulfill those criteria and
the input data have been normalized in a manner that the algorithm comprehends spatial and
temporal data as well.
The most crucial part however, of this venture is the part of clustering evaluation and validation of the clustering scheme. We need to answer the questions: “How many clusters? How are
they placed and distinguished in the spatial plane? Is the clustering reasonable?” Although, a
lot of literature has been written about validation indexes for distance based clustering, there
are no appropriate criteria for density-based clustering validation. Most well-known classifiers have major drawbacks when it comes to arbitrarily-shaped non-convex clusters and noise
i.e. k-means cannot properly identify non-circular clusters nor classify noise as outlier. Such
measures compute the within-cluster dispersion to between-cluster separation and results vary
depending on different formulations (Cesca et al, 2014).
Gaussian Mixture models are efficient regarding the overlapping issue unlike k-means, but still
are not ideal for density clusters. Expectation-Maximization algorithm assigns points to clusters by some probability density estimation and not strictly like k-means. Other measures for
arbitrary shaped clusters are Minimum Spanning Tree and Dunn index (Dunn, 1974), Proximity
Graph by Yang and Lee (2004) etc.
182
ISBN: 978-960-98739-8-7
Searching in literature three measures seems to be distinguishing: CDbw (Halkidi and Vazirgiannis, 2001), DBCV (Moulavi et al, 2014) Density-Based Clustering Validation index and
Density-Based Silhouette diagnostics (Menardi, 2011, Contreras-Reyes, 2013). “Seems to be”
translates that we didn’t test them properly: DBCV output is between -1 and 1, most positive
value means good clustering structures; however, there is no output regarding a cluster assigning
row vector. Regarding CDbw index we were not able to find source code. The Density-Based
Silhouette is implemented in R. Evaluation scheme has taken place using the GAP statistic
(Tibshirani et al, 2001) with Linkage, Kmeans and Gaussian Mixture Distribution algorithms
(Yeck et al, 2015).
The seismic catalogue used extracted from the National Observatory of Athens national catalogue, over an eleven year period 2000 to 2010, with completeness magnitude of M 3.1 Richter.
The catalogue has been declustered using Reasenberg and Urhammer methods.
Applied the aforementioned techniques, we lead to an optimal solution of 73 clusters, using
the Gap criterion with Gaussian Mixture Distribution, Kmeans and Linkage (Ward’s Method)
algorithms. However, the solution failed to converge in the 100 iteration mark producing empty
clusters. The classifier row vector however, holds the correct solution of 53 clusters. Similarly,
the Reasenberg declustered catalogue identified 46 clusters, while the Urhammer came up with
50. The Cophenetic Correlation coefficient values and DBCV validity indices are rather low i.e.
our identified structures are accurate, but they could have been even more solid.
Figure 1. Density-based clustering results from NOA catalogue for the period 2000 - 2010, optimal solution 53
clusters, with (a) Gaussian Mixture Distribution, (b) Linkage (Ward’s method) and (c) Kmeans algorithms.
ISBN: 978-960-98739-8-7
183
Concluding Remarks
As it can be easily concluded, neither of the evaluation methods used identified solid cluster
underlying structures. That was expected however, since very few classifiers in literature are
designed for density-based clustering: CDbw (Halkidi and Vazirgiannis, 2001), DBCV (Moulavi et al, 2014) and Density-Based Silhouette diagnostics (Menardi, 2011, Contreras-Reyes,
2013). They have to be properly implemented and tested thoroughly. Thus, we could have a
clarified view, whether our approach solves, at some extent, the density-based clustering disadvantages or not.
Acknowledgements
The work was supported by the THALES Program of the Ministry of Education of Greece
and the European Union in the framework of the project entitled ‘‘Integrated understanding of Seismicity, using innovative Methodologies of Fracture mechanics along with
Earthquake and non-extensive statistical physics-Application to the geodynamic system
of the Hellenic Arc. “SEISMO FEAR HELLARC”, (MIS 380208)”.
References
1. CESCA, S., et al., 2014, Seismicity monitoring by cluster analysis of moment tensors, Geophys. J. Int. (2014)
196, pp. 1813–1826, doi: 10.1093/gji/ggt492
2. CONTRERAS-REYES, E., J., 2013, Nonparametric Assessment of Aftershock Clusters of the Maule
Earthquake Mw = 8.8, Journal of Data Science 11(2013), pp. 623-638.
3. DASZYKOWSKI, M., et al., 2001, Looking for Natural Patterns in Data. Part 1: Density Based Approach,
Chemometrics and Intelligent Laboratory Systems, Volume 56, Issue 2, pp. 83-92. 4. DRAKATOS, G., and LATOUSSAKIS, J., 2001, A catalog of aftershock sequences in Greece (1971–1997):
Their spatial and temporal characteristics, Journal of Seismology, Volume 5, pp. 137–145.
5. DUNN, J., C., 1974, Well separated clusters and optimal fuzzy partitions. Journal of Cybernetics, Volume 4,
pp. 95–104.
6. ESTER, M., et al., 1995. A Database Interface for Clustering in Large Spatial Databases, Proc. 1st Int. Conf.
on Knowledge Discovery and Data Mining, Montreal, Canada, AAAI Press.
7. ESTER, M., et al., 1996, A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases
with Noise, KDD-96 Proceedings. Copyright © 1996, AAAI (www.aaai.org).
8. HALKIDI, M., and VAZIRGIANNIS, M., 2001, Clustering Validity Assessment: Finding the optimal
partitioning of a data set, Data Mining, ICDM 2001, Proceedings IEEE International Conference on, San
Jose CA, pp. 187-194.
9. MENARDI, G., 2011, Density-based Silhouette diagnostics for clustering methods, Stat Comput (2011) 21,
Springer Science and Business Media, LLC 2010, pp. 295–308, doi: 10.1007/s11222-010-9169-0.
10. MOULAVI, D., et al., 2014, Density-Based Clustering Validation, Proceedings of the 14th SIAM International
Conference on Data Mining (SDM), Philadelphia, PA, 2014.
11. PETERSEN D., M., et al., 2008, Appendix J: Spatial Seismicity Rates and Maximum Magnitudes for
Background Earthquakes, USGS Open File Report 2007-1437J, CGS Special Report 203J, SCEC Contribution
#1138J, Version 1.0.
12. TIBSHIRANI, R., et al., 2001, Estimating the number of clusters in a data set via the gap statistic, J. Royal
Statistical Society B, 63, Part 2, pp. 411-423.
13. VALLIANATOS, F., et al., 2013, A Non-Extensive Statistical Physics View in the Spatiotemporal Properties
of the 2003 (Mw6.2) Lefkada, Ionian Island Greece, Aftershock Sequence, Pure Appl. Geophys. 171 (2014),
pp. 1343–1354, doi: 10.1007/s00024-013-0706-6.
14. YANG, J., and LEE, I., 2004, Cluster validity through graphbased boundary analysis. In IKE, pp. 204–210.
15. YECK L., W., et al., 2015, Maximum magnitude estimations of induced earthquakes at Paradox Valley,
Colorado, from cumulative injection volume and geometry of seismicity clusters, Geophys. J. Int. (2015) 200,
pp. 322–336, doi: 10.1093/gji/ggu394.
16. ZALIAPIN, I., and BEN-ZION, Y., 2013, Earthquake clusters in southern California I: Identification
and stability, Journal of Geophysical Research: Solid Earth, Volume 118, pp. 2847-2864, doi:10.1002/
jgrb.50179.
184
ISBN: 978-960-98739-8-7
Advances in Satellite Remote Sensing of major Natural,
Environmental and Industrial Hazards
V. Tramutoli1,2 and N. Pergola2,1
1. University of Basilicata, Potenza, Italy, [email protected]
2. CNR-IMAA, Tito Scalo (Pz), Italy, [email protected]
KEYWORDS: Robust Satellite Techniques (RST), Forest Fires, Floods, Volcanic
Eruptions, Oil Spills, Pipelines and Infrastructure Monitoring, Earthquake Precursors,
Floods, Soil moisture
Satellite change detection algorithms devoted to identify anomalous space-time transients related to potentially dangerous (for the environment, for the population safety, etc.) events, are
usually fixed-threshold based, single-image oriented, rarely able to maintain the same performance when applied in different geographical areas or to different satellite sensors. Performance, even when contextual spatial tests are introduced, is generally poor because when high
thresholds are chosen, only very big events are detected (low sensitivity), when they are chosen
too low, false positives proliferation occurs (low reliability). Moreover, moving from place
to place, from season to season, different thresholds have to be selected and the same occurs
moving from one satellite sensor to a new one (low exportability). However experience and
common sense tell us that no signal can be interpreted as anomalous „in se“ but only by comparison with a normality which must be preliminarily defined. In fact it is quite obvious that the
same signal, which is normally observable at a specific time and place, could result anomalous
when observed in a different time and place. On these quite obvious concepts is based the RST
(Robust Satellite Technique) change detection approach: for each location of the scene, for each
observation period, a preliminary characterization of the expected signal is obtained by analyzing a multi-years time-series of satellite records collected in similar observational conditions
(e.g. time of the day, month of the year, etc.). An unsupervised, automatic change detection is
then possible, measuring the differences between observed signal and its expected value (e.g.
temporal mean) in units of the natural variability (e.g. standard deviation) of the signal historically observed in the same place in similar observational conditions in the past. Based on the
general RST concepts, several change detection algorithms relevant to African ecosystems, at
the same time effective, reliable and exportable, have been developed which will be described
in this paper. Results so far achieved, by applying RST approach to optical and microwaves
passive sensors, will be presented also by comparison with the most quoted traditional methods.
In particular the possibility offered by RST-FIRES to give timely forest fire alarms, detecting
within 15 minutes, fires as small as a ping-pong table (often anticipating by hours the fire alarms
given by traditional techniques) will be described together with its unique robustness against
false alarm proliferation. Similar performance will be shown in the case of oil spills which have
been automatically detected by RST-OIL with false alarm rates close to zero. Results of the attempts, performed during national (the Italian RITMARE) and European (IOSMOS) projects,
to increase the RST-OIL sensitivity in order to give early warning even in case of very small
discharge will be also described. The possibility to provide (by applying RST approach to passive microwaves sensors) systematic and reliable maps of soil moisture variations, will be also
discussed in the context of a dynamic flood risk evaluation system. Finally the performances
of RST based volcanic eruption (RST-ASH and RST-VOLC) and dust clouds (RST-DUST)
ISBN: 978-960-98739-8-7
185
monitoring systems will be presented, together with RST applications to energy infrastructures
monitoring, flooded area mapping, and earthquake precursors studies that validated during the
EU Projects GMOSS, G-MOSAIC and PRE.EARTHQUAKES.
186
ISBN: 978-960-98739-8-7
A non-extensive statistical physics view in Earth Physics from
laboratory to planetary scale: A review.
Filippos Vallianatos 1,2
Technological Educational Institute of Crete, Laboratory of Geophysics & Seismology,
Chania, GR 73100, Crete, Greece
2
UNESCO Chair on Solid Earth Physics & Geohazards Risk Reduction
1
Keywords: Non extensive statistical physics, Earth Physics, Tsallis Entropy.
Abstract
A non-extensive statistical physics view in Earth Physics from laboratory to planetary scale is
presented. An extented topic of applications is presented from laboratory to planetary scale. The
results of the analysis indicate that the ideas of non-extensive statistical physics can be used
to express the non-linear dynamics that control the evolution of Earth’s dynamics at different
scales and could be used as the key scientific methodology to understand it in a unified way,
using Tsallis entropy principles.
Introduction
Boltzmann-Gibbs (BG) statistical physics is one of the cornerstones of contemporary physics.
It establishes a remarkably useful bridge between the mechanical microscopic laws and macroscopic description using classical thermodynamics. Although BG entropy seems the correct
one to be used in a large and important class of physical systems with strongly chaotic dynamics
(positive maximal Lyapunov exponent), an important class of weakly chaotic systems (where
the maximal Lyapunov exponent vanishes) violates this hypothesis. On the other hand, if longrange interactions, non-markovian microscopic memory, multifractal boundary conditions and
multifractal structures are present then another type of statistical mechanics seems appropriate
to describe nature (Tsallis, 2001).
To overcome at least some of these anomalies that seem to violate BG statistical mechanics,
non-extensive statistical physics (NESP) was proposed by Tsallis in 1988 (Tsallis, 1988) that
recovers the extensive BG as a particular case. The associated generalized entropic form for the
discrete case is
W
Sq = kB
1 − ∑ piq
i =1
q −1
,q∈ R
W
with
∑p
i =1
i
=1
(1)
where Sq is Tsallis entropy and q is the entropic index that represents a measure of the nonextensivity of a system. Sq recovers SBG in the limit q → 1.
Optimizing Tsallis entropy
Assuming a continuous variable X with a probability distribution p(X). In Earth physics, this
variable can be for instance seismic moment (Mo), inter-event times (τ) or distances (r) between
the successive earthquakes or the length of faults (L) in a given region.
We require optimizing Tsallis entropy Sq under the appropriate constraints. The first constraint
refers to the normalization condition of p(X) and the second one is the condition about the generalized expectation value (q-expectation value), Xq defined as:
ISBN: 978-960-98739-8-7
187
where Pq(X) is the escort probability given (Tsallis, 2009) as follows:
.
Using the standard technique of Lagrange multipliers, we obtain the physical probability:
(2) where the q-exponential function defined in Tsallis,
. Typical examples of q-exponen(2009) as
tial are plotted in Fig 1 for various values of q. In non-extensive statistical physics it has been
proposed that the quantity to be compared with the observed distribution is not the physical probability p(X) but its associated escort distribution (see Tsallis, 2009;Vallianatos, 2013).
Following the latter approach, the cumulative distribution function is given by the expression
(3).
Figure 1:The q-exponential distribution for various values of q and for X0=1 in log-linear and log-log scales.
The distribution is convex for q> 1 and concave for q< 1. For q< 1, it has a vertical asymptote at x = (1-q)-1 and
for q> 1 an asymptotic slope -1/(q-1). For q=1 the standard exponential distribution is recovered.
Figure 2: The q-Gaussian distribution for various values of q and for p0=0.5 and X0=1 in linear and log-linear
scales. For q=1 the normal Gaussian distribution is recovered.
Another type of distributions that are deeply connected to statistical physics is that of the
squared variable X2. In BG statistical physics, the distribution of X2 corresponds to the wellknown Gaussian distribution. If we optimize Sq for X2, we obtain a generalization of the
normal Gaussian that is known as q-Gaussian distribution (Tsallis, 2009) and has the form:
188
ISBN: 978-960-98739-8-7
(4) . In the limit q→1, Eq. (2.1.14) recovers the normal
Gaussian distribution. For q> 1, the q-Gaussian distribution has power-law tails with slope -2/
(q-1), thus enhancing the probability of the extreme values. Typical examples of q-Gaussians
are plotted in Fig 2 for various values of q.
In the present work we review a collection of Earth physics problems such as a) Non-extensive
pathways in earthquake size distribution b) The fragment-asperity model, c) Global earthquake size distribution. The effect of mega-earthquakes, d) Increments of earthquake energies
e) Spatiotemporal description of Seismicity, f) Fault networks, g) the plate tectonics as a case
of non-extensive thermodynamics, g) laboratory seismology and fracture and h) stress in Arctic
sea ice and climate change.
The aforementioned cases cover the most of the problems in Earth Physics indicated that non
extensive statistical physics could be the underline interpretation tool to understand earth’s
evolution and dynamics.
As a typical example we present the probability distribution in the incremental earthquake
energies is referred to the probability that an earthquake of energy S(i) will be followed by
one with energy S(i+1) with difference R, expressed as R=S(i+1)-S(i) (i=1,2,...,N-1 where N
the total number). Here we considering the global earthquakes with magnitude M≥7that occurred during the period 1900-2012, as these are reported in the latest version of the Centennial
earthquake catalog (available at http://earthquake.usgs.gov/data/centennial/) and supplemented
by the ANSS earthquake catalog (http://www.ncedc.org/anss/) for the period 2007-2012. The
probability density of the incremental earthquake energies exhibits fat tails and deviates from
the normal Gaussian distribution (Fig 3). A q-Gaussian distribution with q=1.85 ±0.1 can well
describe the observed distribution, thus enhancing the probability of large differences in the
energies of successive earthquakes in global scale.
Figure 1: Normalized Probability density function P(X) (solid circles) for the 1900-2012 global seismicity with
magnitudeM≥7 on a semi-log plot, where S=exp(M), R=S(i+1)-S(i), X =(R-<R>)/σR and <R> the mean and σR
the standard deviation. The dashed curve represents the standard Gaussian shape and the solid line the q-Gaussian distribution for the value of q=1.85 ±0.1.
Similar results obtained analyzing time series of principal stress values σ1 ,σ2 along with the
direction of principal stress θs recorded during the CEAREX Arctic field campaign . The Probability density function of normalised fluctuations, of the principal stress σ1(t) for the Arctic sea
ice exhibits fat tails and deviates from the normal Gaussian distribution (Fig. 4). A q-Gaussian
distribution with q=1.82 can well describe the observed distribution.
ISBN: 978-960-98739-8-7
189
Figure 2. Probability density function of normalised fluctuations, of the principal stress σ1(t) (black circles) with
q-Gaussian fit (solid red line) and Gaussian fit (dashed black line) for the Arctic sea ice with q=1.82 .
Summarizing we can state that the study of the non-extensive statistical physics of earth dynamics remains wide-open with many significant discoveries to be made. The results of the analysis
in the cases described previously indicate that the ideas of non-extensive statistical physics can
be used to express the non-linear dynamics that control the evolution of the earth dynamics
at different scales. The key scientific challenge is to understand in a unified way, using NESP
principles, the physical mechanisms that drive the evolution of fractures ensembles in laboratory and global scale and how we can use measures of evolution that will forecast the extreme
fracture event rigorously and with consistency.
Acknowledgments
The work was supported by the THALES Program of the Ministry of Education of Greece and
the European Union in the framework of the project “SEISMO FEAR HELLARC’’, (MIS
380208).
References
1. Michas, G., Vallianatos, F., Sammonds, P., 2013.Non-extensivity and long-range correlations in the earthquake
activity at the West Corinth rift (Greece). Nonlinear Processes in Geophysics 20, 713-724.
2. Papadakis, G., Vallianatos, F., Sammonds, P., 2013.Evidence of nonextensive statistical physics behavior of
the Hellenic Subduction Zone seismicity. Tectonophysics 608, 1037-1048.
3. Papadakis, G., Vallianatos, F., Sammonds, P., 2014.A nonextensive statistical physics analysis of the 1995
Kobe earthquake, Japan. Pure and Applied Geophysics (accepted).
4. Tsallis, C., 1988. Possible generalization of Boltzmann-Gibbs Statistics. Journal of Statistical Physics 52,
479-487.
5. Tsallis, C., 2009. Introduction to nonextensive statistical mechanics: Approaching a complex world, Springer,
Berlin.
6. Vallianatos, F., 2013. On the non-extensivity in Mars geological faults. Europhysics Letters 102. 28006.
7. Vallianatos, F., Sammonds, P., 2010. Is plate tectonics a case of non-extensive thermodynamics? Physica A
389, 4989-4993.
8. Vallianatos, F., Triantis, D., 2012. Is pressure stimulated current relaxation in amphibolite a case of nonextensivity? Europhysics Letters99, 18006.
9. Vallianatos, F., Sammonds, P., 2013. Evidence of non-extensive statistical physics of the lithospheric instability
approaching the 2004 Sumatran- Andaman and 2011 Honsu mega-earthquakes. Tectonophysics 590, 52-58.
10. Vallianatos, F., Benson, P., Meredith, P., Sammonds, P., 2012. Experimental evidence of a non-extensive
statistical physics behaviour of fracture in triaxially deformed Etna basalt using acoustic emissions. Europhysics
Letters , 97, 58002.
190
ISBN: 978-960-98739-8-7
A Non Extensive Statistical Physics view on the spatiotemporal
distribution of Earth‘s seismicity
Kalliopi Chochlaki1,2*, Filippos Valianatos2 and George Michas3
University of Crete, Department of Physics, Heraklion, Crete, Greece *kchochlaki@gmail.
com, (2) Technological Educational Institute of Crete, Laboratory of Geophysics and Seismology, Crete, Greece ([email protected]) , (3) Institute for Risk and Disaster Reduction,
University College London, Gower Street, London, WC1E 6BT, UK
Keywords: Non Extensive Statistical Physics, Tsallis entropy, regionalization
Abstract
In this research we use the concept of Non Extensive Statistical Physics, introduced by Tsallis
in 2009 (Tsallis, 2009) in order to study the distribution of worldwide seismicity and specifically the collective properties of large earthquakes. Taking into consideration the complexity
and the heterogeneity of the Earth’s seismicity, we used the Flinn-Engdahl regionalization technique ( Flinn and Engdahl, 1965) along with the merging of some regions, made by Lombardi
and Marzocchi ( Lombardi and Marzocchi, 2007). We applied some basic methodological tools
of the NESP approach, studying the interevent times distribution between successive earthquakes, using a q- exponential function, which satisfies the main principle of maximization of
the Tsallis entropy. Furthermore, we analyzed the frequency magnitude distribution according
to the NESP concept, which seems to vary significantly from the classic scheme. Our results indicate that the non extensive statistical physics approach offers us the possibility to investigate
and explore complex systems, such as the Earth’s seismicity, where long-range interactions are
important and it is able to give us important information about the behavior of each seismic
region. The aforementioned techniques were applied to worldwide catalogues with magnitude
greater than Mw=5.0, 5.5 and 6.0.
Introduction
The term “entropy” is used in statistical mechanics in order to describe the measure of disorder
of a complex system. In non extensive statistical physics we consider that the entropy, known
as Tsallis entropy Sq (Tsallis, 2009), is always positive and expansible but most importantly it
is non-additive and this is the main difference of the classic term of entropy, defined by Boltzmann- Gibbs. This means that for two probabilistic events A and B the overall entropy will be:
S q ( A + B)
KB
=
S q ( A)
KB
+
S q ( B)
KB
+ (1 − q )
S q ( A) S q ( B)
KB
K B (1)
The factor q is the q-exponential function that maximizes the Tsallis entropy. In the limit q→1
the q-exponential and q-logarithmic functions lead to the ordinary exponential and logarithmic
functions respectively. If q>1 asymptotic power-law behavior is observed with slope -1/(q-1).
In contrast, for 0<q<1 a cut-off appears (Abe and Suzuki 2003, 2005).
The interevent times distribution
The probability distribution of the interevent time is given by the expression
ISBN: 978-960-98739-8-7
191
1
[
1 + (1 − q )T ] 1− q
P( X ) =
Zq
=
exp(− BqT )
Zq
1
ln(q ) =
(T 1−q − 1)
1− q
while the q- logarithmic function is defined as:
(2)
(3)
Following the latter approach we applied the non extensive statistical physics model inside the
39 regions from the final regionalization made by Lombardi and Marzocchi and, as a result,
the q function has an average of q=1.160 for the catalogue that consists of earthquakes with
magnitude Mw≥5.0, q=1.118 for the catalogue with events that have a magnitude Mw≥5.5 and
q=1.090 for earthquakes with magnitude Mw≥6.0.
Fig. 1: The cumulative distribution for the interevent
times for the data set of period 2005-2009 with earthquakes with magnitude equal or greater than the threshold magnitude Mw=5.0 and depth H≤150 km.The value
of qT is 1.063.
Fig. 2: The semi-q-plot of the cumulative distribution for
the interevent times for the data set of period 2005-2009
with earthquakes with magnitude equal or greater than
the threshold magnitude Mw=5.0 and depth H≤150km.
The straight line represents the q-logarithmic function
with value qT=1.063
The frequency-magnitude distribution according to NESP
The frequency-magnitude distribution has been analyzed for many years using the known
Gutenberg-Richter’s law, which log-linearly relates the total number of earthquakes with the
magnitude as shown in equation 4.
(4)
where N stands for the total number of earthquakes, M for the magnitude and a,b are constants
for a dataset of earthquakes. The G-R law predicts that the energy of earthquakes exhibit power
law decay. In 2004 a new model for earthquake’s dynamics was proposed by Sotolongo and
192
ISBN: 978-960-98739-8-7
Posadas ( Sotolongo and Posadas, 2004 ) . Starting from the main principle of maximization
of the Tsallis entropy they derived equation 5, which seems to describe more accurately the
frequency-magnitude distribution, when long range interactions are important.
(5)
The constant α expresses the proportionality between the released energy E and the fragment
size r, while q is the entropic index that from now on we will refer to as “the entropic index q”.
The above model, seems to describe better the energy distribution in a wide range of magnitudes than the G-R law, which above some threshold magnitude can be considered as a particular case (Telesca 2012).
Calculating the b-value from the G-R law for the whole unregionalized catalogue we take an
average value close to 1 that tends to increase as we change the threshold magnitude from 5
to 5.5 and then to 6, while inside the 39 regions the b-value seems to differ significantly from
region to region. The calculation of the entropic index q has an average of 1.459 for the whole
catalogue and an average of 1.419 inside the seismic regions, proving that the NESP approach
seems to be the most adequate tool to analyze the frequency-magnitude distribution both inside
and outside of the seismic regions.
Conclusions
It is now becoming evidential that the behavior of such complex systems cannot be fully described from the classical principles of statistical physics, while the model of non extensive statistical physics along with regionalization techniques can give us important information about
the seismic activity and the geological characteristics of each region. The model used fits rather
well to the observed distributions, implying the complexity of the spatiotemporal properties of
seismicity and the usefulness of NESP in investigating such phenomena, exhibiting scale-free
nature and long-range memory effects. Furthermore, we notice that the parameter q of the interevent times tends to decrease as we increase the magnitude threshold leading to the assumption that the correlation between them weakens when we cut out the smaller earthquakes and
the conclusion that these small and moderate events take place in the generating process of the
earthquakes. In addition, the NESP approach for the frequency-magnitude distribution works
rather well inside and outside of the 39 regions.
Acknowledgments
This work was implemented through the project entitled “IMPACT-ARC” in the framework of
action “ARCHIMEDES III—Support of Research Teams at TEI of Crete” (MIS380353) of the
Operational Program “Education and Lifelong Learning” and is co-financed by the European
Union (European Social Fund) and Greek national funds
References
1. Abe, S., and N. Suzuki, Law for the distance between successive earthquakes, J. Geophys. Res., 108(B2),
2113, doi:10.1029/2002JB002220, 2003.
2. Brouers F., O. Sotolongo-Costa, K. Weron, Burr,Levy,Tsallis, (2004) , Physica A, doi:10.1016/j.
physa.2004.06.008
3. Flinn E. A., Engdahl E. R., 1965. A proposed basis for geographical and seismic regionalization, Rev.
Geophys.,3,123–149.
4. Flinn E. A., Engdahl E. R., Hill A. R., 1974. Seismic and geographical regionalization, Bull, seism. Soc. Am.,
64, 771–993.
ISBN: 978-960-98739-8-7
193
5. Lombardi, A. M., and W. Marzocchi (2007), Evidence of clustering and nonstationarity in the time distribution
of large worldwide earthquakes, J. Geophys. Res., 112, B02303, doi:10.1029/2006JB004568.
6. Michas G.,F. Vallianatos and P. Sammonds, Non-extensivity and long-range correlations in the earthquake
activity at the West Corinth rift (Greece), (2013), Nonlin. Processes Geophys., 20, 713–724, 2013, www.
nonlin-processes-geophys.net/20/713/2013/doi:10.5194/npg-20-713-2013
7. Papadakis G., F. Vallianatos, P. Sammonds, Evidence of No Extensive Statistical Physics behavior of the
Hellenic subduction Zone seismicity,Tectonopfysics, http://dx.doi.org/10.1016/j.tecto.2013.07.009
8. Telesca L.,(2011), Tsallis-Based Nonextensive Analysis of the Southern California Seismicity, Entropy 2011,
13, 1267-1280; doi:10.3390/e13071267
9. Telesca L., (2012) , Maximum Likelihood Estimation of the Nonextensive Parameters of the Earthquake
Cumulative Magnitude Distribution, Bulletin of the Seismological Society of America, Vol. 102, No. 2, pp.
886–891, April 2012, doi: 10.1785/0120110093
10. Tsallis S., Nonadditive Entropy Sq and Nonextensive Statistical Mechanics: Applications in Geophysics and
Elsewhere, (2012) , Acta Geophysica, 60, 3, 502-525, DOI: 10.2478/s11600-012-0005-0
11. Vallianatos F., G. Michas, G. Papadakis and P. Sammonds,(2012), A Non-Extensive Statistical Physics View
to the Spatiotemporal Properties of the June 1995, Aigion Earthquake (M6.2) Aftershock Sequence (West
Corinth Rift, Greece), Acta Geophysica, 60, 3, 758-768 DOI: 10.2478/s11600-012-0011-2
12. Vallianatos F., P. Sammonds, Is plate tectonics a case of non-extensive thermodynamics?, (2010), Physica
A:Statistical Mechanics and its Applications, 389, 21, 1, 4989-4993, DOI: 10.1016/j.physa.2010.06.056
194
ISBN: 978-960-98739-8-7
Paikon and Tzena terranes (Axios zone, Greece): Can they be
correlated?
Katrivanos E.1*, Kilias A.1 and Mountrakis D.1
Aristotle University of Thessaloniki, Department of Geology, 54124, Thessaloniki, Greece,
* [email protected]
1
Keywords: accretion, deformation, nappe, obduction, tectonic window.
Abstract
Paikon and Tzena terranes are situated in the center part of the Axios zone, between Almopia
and Paionia ophiolitic belts. Tectonostratigraphic data reveal that they consist of similar tectonostratigraphic units and both have affected by the same polyphase deformation and metamorphism from Mid-Jurassic to recent times (D1 - D6). Deformation started in Mid-Late Jurassic
time with an intra-oceanic subduction in Axios ocean (=Neotethys) and continued progressively
until today. Shortening alternated with extension and the deformation conditions evolved from
ductile to brittle. Compressional events are related to ophiolite obduction, nappe stacking, terrane accretion and crustal thickening, while extensional ones are related to orogenic collapse,
unroofing, crustal thinning, terrane dispersion and basin formation. The lower tectonostratigraphic units of both terranes, i.e. gneisses and marbles, may be of Pelagonian origin, exhumed
as a multiple tectonic window under the overthrusted nappe pile.
Introduction
Paikon and Tzena terranes are situated in the center part of the Axios zone, between the Pelagonian nappe and the Serbomacedonian/Rhodope metamorphic province. Axios zone comprises
the ophiolite-bearing Almopia Subzone in the West, the Paikon and Tzena terranes in the Middle, and the also ophiolite - bearing Paionia Subzone in the East (Mercier, 1968, Ricou and
Godfriaux, 1995, Brown and Robertson, 2003).
It is clear that the structural evolution and geotectonic position of Paikon and Tzena terranes
have long been a matter of controversy between researchers, with several questions remaining
unanswered. We try here to correlate Paikon and Tzena terranes, reconstructing the evolution
of deformation from Jurassic till recent times, and for this purpose we carried out geological
mapping and detailed structural investigations, combined with all available geochronological
and stratigraphic data.
Tectonostratigraphic results
Paikon & Tzena terranes are two identical tectonostratigraphic terranes, constituting a complicated tectonic nappe pile with several tectonostratigraphic units (Fig. 1).
Paleozoic gneisses and Triassic marbles intercalated with schists (Lower Units) are tectonically
overlain by volcanoclastic and carbonate rocks of Jurassic to Early Cretaceous age (Medium
Units), as well as the obducted Neotethyan ophiolites far-traveled westwards from their initial
place. Transgressive Upper Jurassic to Lower Cretaceous sediments, as well as Upper Cretaceous limestone and flysch (Upper Units) overlie discordantly all the above mentioned units.
The Paleozoic basement rocks do not exhumed in the Paikon terrane.
Both terranes have undergone the same polyphase deformation and metamorphism (D1 - D6),
from Mid-Jurassic to recent times (Katrivanos et al 2001, Katrivanos et al, 2013). Shortening
alternated with extension and the deformation conditions evolved from ductile to brittle. Compressional events are related to nappe stacking, terrane accretion and crustal thickening, while
ISBN: 978-960-98739-8-7
195
extensional ones are related to orogenic collapse, terrane dispersion, isostatic rebound, unroofing and crustal thinning.
Figure 1. Tectonostratigraphic columns of Paikon and Tzena terranes.
Ophiolite obduction and nappe stacking started during the Mid- to Upper Jurassic (D1). The
structures of D1 event are penetrative (synmetamorphic foliation, S1 and mineral stretching lineation, L1) and affect clearly the lower units of Triassic age, as well as the Mid- to Late Jurassic
volcano-sedimentary series. We assume that D1 is responsible for the strong synmetamorphic
deformation and internal imbrication of the Triassic basement units and the Mid-Late Jurassic
volcano-sedimentary series. A HP-LT metamorphism of Mid– to Late Jurassic age, predates D1
event and is possibly related to a subduction zone, as referred by Baroz et al. (1987).
During D2 in the Early Cretaceous, W- to SW-vergent imbrication and folding took place again,
now affecting the Upper Jurassic - Lower Cretaceous sedimentary series, the Jurassic volcanosedimentary series, the Triassic ones, as well as the ophiolitic bodies. The ophiolites continued
to advance westwards during that time, overthrusting the Upper Jurassic - Lower Cretaceous
sediments.
Upper Cretaceous extension (D3 event) and exhumation of deeper crustal rocks is associated
with basin subsidence and deposition of the Upper Cretaceous limestones and flysch (Cenomanian to Maastrichtian), outcropping on the western flanks of Paikon and Pinovon Mts.
D4 structures form a well developed, NW-SE trending fold and thrust belt, active from Paleocene to Eocene and resulting in renewed SW-directed imbrication of all units. In the western Paikon, Almopia ophiolites overthrust the Theodoraki Carbonate Formation and Tchouka
flysch, along a NEward D4 back-thrust (Mercier and Vergely, 2001, Brown and Robertson,
2003).
196
ISBN: 978-960-98739-8-7
The final exhumation of Paikon & Tzena tectonostratigraphic units, as well as crustal thinning
occurred during the Oligocene-Miocene extensional event D5 under brittle conditions. Simultaneously, unroofing and exhumation of the External Hellenide units, took place under ductile
conditions at a deeper structural level of the Hellenic orogen (Kilias, 1995).
During Miocene-Pliocene, faulting overprinted all previously described structures. D6 structures include high-angle normal, dip-slip to oblique faults, as well as strike-slip faults. A very
important D6 fault is the ENE - WSW striking and SSE - dipping Aridea fault zone, which is
considered to be originally a dextral strike-slip fault, reactivated during the Pliocene - Quaternary until recent as a normal dip-slip fault (Pavlides et al 1990). In the study area, this fault
separates Tzena from Paikon terrane, forming the Neogene to Quaternary Notia - Periklia depression and leading to post-accretional terrane dispersion.
Discussion & Conclusion
In our view, the primary emplacement of ophiolites upon the Triassic marbles was possibly
associated with the overthrusting of the volcanoclastic series. Brown and Robertson (2003) assumed Jurassic, northeastward subduction that created an island arc within the Paikon Massif,
coupled with back-arc rifting, generating the Guevgeli Ophiolites and related units behind the
Paikon arc. They also propose that overthrusting of those ophiolites upon the Paikon Massif
took place firstly during the Late Cretaceous to Early Tertiary.
Nevertheless, in the Lower Units of both terranes, feeds dykes of the overlying volcanic formations are not found. Therefore, we suggest that these volcanic rocks of calc-alcaline composition formed somewhere else, during an initial Jurassic intra-oceanic subduction in the Neotethys
Ocean, as also assumed by Michard et al (1998). Furthermore, Jurassic magmatic series (Chortiatis series; Mussalam 1991) including reefal limestones of Late Jurassic age (KimmeridgianTithonian) and clastic deposits with spilite, keratophyre, andesite and granodiorite elements
(Mussalam 1991, Michard et al 1998), exposed along the western margin of the Serbomacedonian massif, can be correlated in lithology and age with the Paikon volcano-sedimentary
formations. This Chortiatis magmatic series and the volcanoclastic series along the western
Serbo-Macedonian margin was interpreted by Michard et al (1998) as having developed in an
intra-oceanic island arc setting related to an intra-oceanic subduction predating the obduction
of the Axios (=Neotethyan) ophiolites. Such a tectonic setting and place may be assumed also
for the Paikon & Tzena volcano-sedimentary units, prior their westwards emplacement onto the
Triassic units, owing to the arc - continent collision.
Therefore, we suggest that ophiolite obduction upon the continental series of Paikon & Tzena terranes took place simultaneously with the ophiolite tectonic emplacement towards west
upon the Triassic marbles of the eastern Pelagonian margin, as described by several researches
(Brown and Robertson 2003, Kilias et al 2010).
Finally, the nappe pile evolved into a multiple tectonic window, where the Lower Units, possible of Pelagonian origin (Gandatch Unit & Tzena marbles) emerged under the Jurassic volcano-sedimentary series and the Upper Jurassic - Lower Cretaceous carbonate series (Medium
Units) and these again under the duplicated, ophiolitic nappe (Almopia and Garefi ophiolites to
the west and Guevgeli Ophiolites to the east). The tectonic window formed due to successive
compressional and extensional events from Jurassic to Tertiary time. The final configuration of
the tectonic window resulted from the Oligocene - Miocene extension (D5).
An overall westward overthrusting of Paikon units, but of Tertiary age, was also proposed by
Godfriaux and Ricou (1991), Ricou and Godfriaux (1995). In this approach, Paikon Massif
is considered as a multiple tectonic window of Tertiary age beneath the thrusted slices of the
Axios Zone ophiolites and the Pelagonian Nappe of metamorphic rocks. Additionally, Tzena
Massif is considered as a Tertiary metamorphic nappe with an origin from the east (Rhodopic
ISBN: 978-960-98739-8-7
197
hinterland), that was placed tectonically upon both the ophiolites and the Paikon units. The
lower unit of the Paikon Massif is correlated with the carbonate unit of Olympos-Ossa, while
the overthrusted volcanic rocks with the Ambelakia HP/LT unit.
However, we do not share this view, because those two windows show a different structural
evolution and tectonic position. The Olympos - Ossa window represents a deeper tectonic unit
of Hellenides belonging to the External Hellenides below the Pelagonian nappe pile, exhumed
first during the Tertiary.
In conclusion, we suggest that Paikon & Tzena terranes are tectonostratigraphically identical
and each one of them is a composite terrane that formed by terrane accretion, i.e. a complicated
tectonic nappe pile, including obducted ophiolites rooted in the Axios zone (=Neotethys), fartravelled westward from their initial place. The core of both terranes, probably of Pelagonian
origin, is exhumed as a multiple tectonic window under the overthrusted nappe pile.
References
1. Baroz, F., Bebien, J., and Ikenne, M., 1987. An example of HP-LT metamorfic rocks from an island-arc: the
Paikon series (Innermost Hellenides, Greece). -Journal of Metamorphic Geology, 5, 509-527.
2. Brown, S., and Robertson, A., 2003. Sedimentary geology as a key to understanding the tectonic evolution
of the Mesozoic – Early Tertiary Paikon Massif, Vardar suture zone, N. Greece. -Sedimentary Geology, 160,
179-212.
3. Godfriaux, I., and Ricou, L. E., 1991. Le Paicon, une fenetre tectonique dans les Hellenides Internes
(Macedoine, Grece). - C. R. Acad. Sci. Paris, 313/2,1479-1484.
4. Katrivanos, E., Mountrakis, D., Kilias, A. & Pavlides, S., 2001. Preliminary results of the geological structure
and kinematics of deformation in Tzena Mt. (Paikon Subzone, Central Macedonia, Greece). - Bulletin
Geological Society of Greece, 34/1, 137-147 (in Greek).
5. Katrivanos, E., Kilias, A. and Mountrakis, D., 2013. Kinematics of deformation and structural evolution of
the Paikon Massif (Central Macedonia, Greece): A Pelagonian tectonic window? N. Jb. Geol. Palaeont. Abh.
269/2, 149-171.
6. Kilias, A.,1995. Tectonic evolution of the Olympus-Ossa Mt.: Emplacement of the blueschists unit in eastern
Thessaly and exhumation of Olympus-Ossa carbonate dome as a result of tertiary extension (Central greece).
- Mineral Wealth, 96, 20-42.
7. Kilias, A., Frisch, W., Avgerinas, A., Dunkl, I., Falalakis, G. and Gawlick, H.J., 2010. Alpine architecture and
kinematics of deformation of the northern Pelagonian nappe pile in the Hellenides. -Austrian Journal of Earth
Sciences, 103/1, 4-28.
8. Mercier, J., 1968. Etude geologique des zones Internes des Hellenides en Macedoine centralle. Contribution
a l’ etude du metamorphisme et de l’ evolution magmatique des zones internes des Hellenides. –Annales
Geologique Pays Hellenic, 20 (1968), 1 - 739.
9. Mercier, J., and Vergely, P., 2001. The Paikon Massif revisited comments on the Late Cretaceous – Paleogene
geodynamics of the Axios-Vardar zone. How many Jurassic ophiolitic basins?
10. Michard, A., Feinberg, H. and Montigny, R., 1998. Supra-ophiolite formation from the Thessaloniki mappes
(Greece) and associated magmatism: an intra-oceanic subduction predating the Vardar obduction. –C. R.
Acad. Sciences, Paris, 327, 493-499.
11. Mussalam, K,. 1991. Geology, geochemistry and the evolution of an oceanic crustal rift at Sithonia. NE
Greece. – In: Peters T., Nikolas A., Coleman R.G. (Eds), Ophiolite Genesis and the Evolution of the Oceanic
Lithosphere. Kluver Academic Publishing, London, 685-704.
12. Pavlides, S., Mountrakis, D., Kilias, A., and Tranos, M., 1990. The role of strikeslip movements in the
extensional area of the northern Aegean (Greece). - Annale Tectonique, 4, 196-211.
13. Ricou, L. E., and Godfriaux, I., 1995. Mise au point sur la fenetre multiple du Paicon et la structure du Vardar
en Grece. - C. R. Acad. Sci. Paris, 321/ 2, 601-608.
198
ISBN: 978-960-98739-8-7
Temporal Properties of Seismicity in the Corinth Rift: Are
Earthquakes Random?
Georgios Michas1*, Filippos Vallianatos2,1 and Peter Sammonds1
Institute for Risk and Disaster Reduction, University College London, Gower Street,
London, WC1E 6BT, UK, * [email protected]
2
Technological Educational Institute of Crete, Laboratory of Geophysics and Seismology,
Crete, Greece.
1
Keywords: Earthquakes, Corinth Rift, Multifractality, Long-term correlations.
Abstract
One of the main research questions that concern the earthquake generation process in a region is
whether earthquakes occur randomly in time or there is some kind of “memory” in the process,
where the time of the next earthquake is related to the time of the previous ones. The temporal
properties of seismicity have important implications about earthquake physics and probabilistic
earthquake hazard assessments. We study the temporal properties of seismicity in one of the
most seismically active areas in Europe, the Corinth Rift. The high earthquake activity in the
area is revealed from both historic and instrumental records, where several earthquakes of magnitude greater than 6 have occurred in the past. The analysis for various earthquake catalogues
indicates multifractal temporal structure and clustering effects at all timescales. Such properties
can well be approximated by the q-generalized gamma distribution that is associated with the
non-stationary character of the earthquake activity and memory in the seismogenic process.
Such type of modeling helps us understand better the physical mechanism of seismogenesis and
improve the efficiency of probabilistic earthquake hazard assessments.
Introduction
Earthquake time series are typically characterized by non-stationarities, intermittency and clustering variability, highlighting the complex nature of the earthquake generation process. The
temporal properties of seismicity have been the subject of intensive study during the past decade. Research focused on whether inter-event times, i.e. the time intervals between successive
earthquakes, for various threshold magnitudes and spatial scales follow particular distributions,
the form and the physical mechanism of which is of great importance in earthquake modeling
and hazard assessments. One of the main research questions related to the topic is whether
earthquakes occur randomly in time or they possess some kind of “memory”, where the time of
the next earthquake is related to the time of the previous ones.
We study the temporal properties of seismicity in one of the most seismically active areas in
Europe, the Corinth Rift (central Greece). The area is characterized by rapid continental extension and high earthquake activity, with six earthquakes of magnitude greater than 6 occurring in
the last 50 years. Earthquake activity in the Corinth rift is typically characterized by fluctuating
behavior, where periods of low to moderate activity are interspersed by sudden seismic bursts,
which are related to frequent earthquake swarms and the occurrence of stronger events, followed by aftershock sequences. In such seismically active areas it is crucial to better understand
earthquake mechanics so as to increase resilience for the society and infrastructures.
ISBN: 978-960-98739-8-7
199
We perform multifractal analysis, in order to study the local fluctuations, the range of correlations and the degree of heterogeneous clustering in the earthquake activity in the Corinth Rift.
The multifractal detrended fluctuation analysis (MF-DFA) (Kantelhardt et al, 2002) is applied to
the inter-event time series, according to the National Observatory of Athens (NOA) earthquake
catalogue (available at http://www.gein.noa.gr/en/), for 1964-2011 and for the threshold magnitude of Mc=3, the revised catalogue of the Hellenic Unified Seismological Network (HUSN)
(available at http://bbnet.gein.noa.gr/HL/), for 2008-2013 and for Mc=2 and the Corinth Rift
Laboratory (CRL) earthquake catalogue that reports earthquakes at the west part of the Corinth
Rift (available at http://crlab.eu/), for 2002-2008 and for Mc=1.2.
To perform the analysis, initially we integrate and divide the series into non-overlapping segments of equal size n. In each segment a polynomial function is fitted and subtracted from the
integrated signal to estimate the root mean-square fluctuations. Then the qth order fluctuation
function Fq(n) is obtained after averaging over all segments. For various values of n, Fq(n) will
increase as a power-law with exponent h(q), Fq(n)~nh(q), if the series are long-range power-law
correlated (see Michas et al, 2015). For multifractal series, the exponent h(q) will depend on
the various values of q (Fig.1a). Another way to characterize multifractal series is the singularity spectrum f(a), where a is the singularity strength or Hölder exponent (Fig.1b). The singularity spectrum f(a) indicates the fractal dimensions of the subsets that have the same singularity
strength a and gives information about the relative importance of each fractal dimension. In
monofractal series the singularity strength is similar in the entire range of the set so that the singularity spectrum collapses into a single point. MF-DFA analysis on the three datasets reveals
the multifractal structure of the time series and the degree of heterogeneous clustering (Fig.1).
a)
b)
Figure 1. a) Range of generalized Hurst exponents h(q) for q ∈ [−5,5] and step size 0.2 for various earthquake
catalogues. The corresponding confidence intervals are plotted as error bars. b) The singularity spectrum f(α)
versus the Hölder exponent α for the same earthquake catalogues.
Next we study the histogram of inter-event times for the three earthquake catalogues and for
various time periods and threshold magnitudes. We rescale inter-event times τ with the mean
seismic rate R=1/<τ> as T=R∙τ and estimate the probability distribution functions by counting
the number of T that fall into each logarithmically spaced bin and then normalize this number
by dividing it to the bin width (e.g., Corral, 2004). To obtain the probability density of the
series, the counts are further divided by the total number of counts so that the probabilities of
occupation sum to one.
200
ISBN: 978-960-98739-8-7
In Fig.2 the probability densities p(T) of the rescaled inter-event times T are plotted. The various
p(T) exhibit similar scaling behavior and approximately fall onto a unique curve, indicating the
self-similarity of the earthquake generation process in the Corinth Rift. To model this behavior
we use a stochastic dynamical mechanism with memory effects. This can be expressed mathematically in terms of a Feller process and its corresponding Fokker-Planck equation (Queiros,
2005). For stationary periods, where the mean inter-event time <τ> does not fluctuate, the solution is the gamma distribution:
f (T ) = CT γ −1 exp(− T β ) ,
(1)
which has frequently been used to describe the probability densities of inter-event times (e.g.
Corral, 2004). Regarding the physical implementations, this function exhibits two regions,
where short inter-event times scale as a power law with exponent 1-γ (correlated earthquakes,
e.g. aftershocks) and large inter-event times decay exponentially (uncorrelated earthquakes,
e.g. background activity).
Figure 2. Probability density p(T) of the rescaled inter-event times T for the HUSN, CRL and NOA datasets and
for various time periods and threshold magnitudes. The solid line represents the corresponding fit according
to the q-generalized gamma distribution (Eq.2) for the values of C = 0.4, β = 1.65, γ = 0.38 and q = 1.23. The
dashed line represents the corresponding fit according to the gamma distribution (Eq.1) for the values of C =
0.4, β = 1.65 and γ = 0.38.
For non-stationary periods, where local fluctuations of mean <τ> are produced, the solution is
a q-generalized gamma distribution (Queirós,2005) that has the form:
f (T ) = CT γ −1 exp q (− T β ) .
(2)
1 (1− q )
exp q (x ) = 1 + (1 − q ) x 
The last term in this equation is the q-exponential function
that induces the power-law asymptotic behavior in the distribution (Michas et al, 2013).
The q-generalized gamma distribution provides an excellent fit to the full range of the observed
p(T) indicating clustering effects at all timescales. According to Eq.2, for small values of T,
ISBN: 978-960-98739-8-7
201
p(T) scales as a power-law, p (T ) � T γ −1 , and for large values of T, p(T) decays asymptotically
as another power-law, p (T ) � T (1−γ ) (1− q ) . In the limit of q→1 the q-generalized gamma distribution (Eq.2) recovers the ordinary gamma distribution (Eq.1). The bimodality in the probability
density of inter-events times and the gradual crossover between two power-law regimes for
short and long inter-events times respectively suggest that the earthquake activity evolves as
two processes. The first one is related to clustering effects at short timescales, induced by aftershock sequences and earthquake swarms and the second one to long-term clustering effects,
related to the background activity.
Concluding Remarks
The temporal properties of the earthquake activity in the Corinth Rift were studied by looking
into the inter-event times between successive earthquakes for various earthquake catalogues.
The inter-event time series exhibit a range of fractal dimensions and multifractality, indicating
the heterogeneous degree of clustering in the evolution of the earthquake activity. The rescaled
probability density function of inter-event times exhibits bimodality and a gradual crossover
between two power-law regimes, suggesting that the distribution is a mixture of correlated
events at short timescales, induced by aftershock sequences and earthquake swarms, and correlated background activity at long timescales. Such properties indicate clustering effects at all
timescales and memory in the seismogenic process. In addition, the almost perfect collapse of
the rescaled probability densities, for various time periods, threshold magnitudes and spatial
scales, onto a unique curve that is well approximated by the q-generalized gamma distribution,
indicates the self-similarity of the earthquake generation process in the Rift.
Acknowledgements
This work was implemented through the project entitled “IMPACT-ARC” in the framework of
action “ARCHIMEDES III—Support of Research Teams at TEI of Crete” (MIS380353) of the
Operational Program “Education and Lifelong Learning” and is co-financed by the European
Union (European Social Fund) and Greek national funds
References
1. Corral, A., 2004, Long-term clustering, scaling, and universality in the temporal occurrence of earthquakes.
Physical Review Letters, 92, 108501.
2. Kantelhardt, J. W., Zschiegner, S. A., Koscielny-Bunde, E., Havlin, S., Bunde, A., Stanley, H. E., 2002,
Multifractal detrended fluctuation analysis of nonstationary time series. Physica A, 316, 87-114.
3. Michas, G., Vallianatos, F., Sammonds, P., 2013, Non-extensivity and long-range correlations in the earthquake
activity at the West Corinth rift (Greece). Nonlinear Processes in Geophysics, 20, 713–724.
4. Michas, G., Sammonds, P., Vallianatos, F., 2015, Dynamic multifractality in earthquake time series: Insights
from the Corinth rift, Greece. Pure and Applied Geophysics, 172, 1909-1921.
5. Queirós, S. M. D., 2005, On the emergence of a generalized gamma distribution. Application to traded volume
in financial markets. Europhysics Letters, 71, 339-345.
202
ISBN: 978-960-98739-8-7
The Hellenic Seismological Network of Crete (HSNC): Monitoring
results and the new strong motion network
Georgios Chatzopoulos1, Ilias Papadopoulos1 and Filippos Vallianatos1
1
Technological Educational Institute of Crete, School of Applied Sciences, Laboratory
of Geophysics & Seismology
email: [email protected]
Keywords: Seismic network, seismic acceleration, strong motion network.
Abstract
The Crete Island is located in the front of the Hellenic Arc, where the subduction of African
plate under the Eurasian takes place resulting in a very seismic active environment. The constant need for monitoring such a complex geotectonic setting is the main reason for establishing the Hellenic Seismological Network of Crete. Since the HSNC started its operation, a large
amount of data has been recorded in the vicinity of the front Hellenic Arc. Aside from the
continuous monitoring of seismic activity, HSNC is also interested in the seismic hazard in
the area. An area of great interest is the Western Crete, where the city of Chania resides, built
in a sedimentary basin with different geological formations (Mountrakis et al, 2012). A strong
motion network is established to monitor and measure the Peak Ground Velocity (PGV) and
Acceleration (PGA) of the different geological formations. The monitoring results after a shock
occurred are published in G.I.S. maps which are available for public, authorities and scientists
in SAFER CHANIA webpage.
Introduction
The Hellenic arc is created by the subduction of the Mediterranean ocean plate under the Aegean
microplate (McKenzie, 1972). The convergence of these plates has a velocity rate about 4cm/y
which makes the vicinity of Crete a very seismic active environment (Pichon and Angelier,
1979). It is estimated that more than 60% of the Europe’s total amount of seismic energy is
being released to South Aegean area (Papazachos, 1990). The monitoring of HSNC started
working in 2004 with few short period seismometers and nowadays it has 13 broadband seismometers. The stations are equipped with Reftek 130-01 digitizers (www.reftek.com) recording data at 100 Hz sps coupled with Guralp CMG ESPC 60 and 120 sec sensors s. The HSNC
collaborates with three neighboring seismic networks and exchanges data in real time. These
networks are the GEIN-NOA (Athens), AUTH (Thessaloniki) and KOERI (Istanbul) : (Fig. 1).
The seismic signals of more than 30 stations are checked by three Earthquake Monitoring Systems (SeisComp3, Earthworm and SNDA). Since 2011, near the network operation area, HSNC
has recorded 8 large earthquakes M≥6.0. One of these earthquakes with magnitude M=6.4, occurred in 12/10/2013 shook the west part of Crete. This event gave origin to the idea of installing a strong motion seismic network (Fig. 2). We installed seven Reftek 130-SM instruments
in different geological formations in the general area of Chania basin. An automated system
calculates ground shaking values (Peak Ground Velocity and Acceleration) and creates GIS
maps using prior information regarding the geological formations, such as the shear velocity of
surface layers and topography. Interpolation of data values fills the gaps between stations.
ISBN: 978-960-98739-8-7
203
Figure 1. The HSNC and neighbor networks
Figure 2. The SAFER Chania network. Geology of the general area is after Mountrakis et al. (2012)
Monitoring results
204
ISBN: 978-960-98739-8-7
The HSNC
The Hellenic Seismological Network of Crete (HSNC) is a continuous expanding local seismological network with main purpose to monitor the seismicity in the front of Hellenic Arc. The
last 5 years the complex regime of Crete region provided 8 large earthquakes (M≥6.0) and an
enormous amount of events with M≥5.0. The network provided detailed results for the large
earthquakes and their aftershock sequences. A good example of HSNC analyses are the two
large events occurred in 2013 in southern (15/06) and western (12/10) part of Crete region.
Another example is the 16/04/2015 event in eastern area of Crete. The location of main shocks
and their aftershocks are products of Hypoinverse relocation software (Klein, 2002).
Figure 3. The 12/10/2013 Mainshock and its aftershocks
The SAFER CHANIA strong motion network
The SAFER Chania strong motion network focuses to monitor the city of Chania and its
Southern basin - a cultural monument in Mediterranean area - and provide all the useful information to Civil Protection authorities. The ground shaking recording in terms of pick ground
acceleration and velocity (PGA and PGV) is calculated for every seismic event recorded by
sensors. The response spectrum of the sensors is also calculated and all these information are
illustrated at the SAFER Chania website. In 24/06/2015 a small earthquake is recorded from
HSNC’s network West of Chania city, with Magnitude ML=2.3 (Fig. 4). The distance from the
city is ~50 Km. The SAFER Chania strong motion network recorded the event. After processing of data, the Peak Ground Acceleration is calculated (PGA) and placed on the map below.
Interpolation is done by using an Inverse Distance Weighting algorithm.
ISBN: 978-960-98739-8-7
205
Figure 4. A small earthquake example
Conclusions
The HSNC operation covers the monitoring need of the Hellenic Arc front. It has a dense network of sensors and in the future new stations will be installed to strategic places. The new
strong motion network called SAFER CHANIA provides some good results but it is still under
development. The plan is to expand the strong motion network by installing a few more sensors
to have better resolution of ground shaking.
Acknowledgements
This work was implemented through the project IMPACT-ARC in the framework of action
“ARCHIMEDES III–Support of Research Teams at TEI of Crete” (MIS380353) of the Operational Program “Education and Lifelong Learning” and is co-financed by the European Union
(European Social Fund) and Greek national funds.
References
1. Klein F.W., 2002 User’s Guide to HYPOINVERSE-2000, a Fortran program to solve for earthquake locations
and magnitudes, Open File Report 02-171, u.s. Geological Survey.
2. Le Pichon X., and Angelier J., 1979, The Hellenic Are and Trench System: a key to the neotectonic evolution
of the eastern Mediterranean area. Tectonophysics, 60, pp. 1–42.
3. Mountrakis D., Killias A., Pavlaki A., Fassoulas C., Thomaidou E., Papazachos C., Papaioannou C., Roumeliozi
Z., Benetatos C. and Vamvakaris D., 2012, Neotectonic study of Western Crete and implications for seismic
hazard assessment, Journal of the Virtual Explorer, 42.
4. McKenzie D., 1972, Active Tectonics of the Mediterranean Region. The Geophysical Journal of the Royal
Astronomical Society 30: 109-185.
5. Papazachos B.C., 1990, Seismicity of the Aegean and surrounding area, Tectonophysics, 178, 287-308.
206
ISBN: 978-960-98739-8-7
Studying Lithosphere-Atmosphere-Ionosphere Coupling at the
south front of the Hellenic Arc
John P. Makris1*, Bruno Zolesi2, Massimo Chiappini2, Filippos
Vallianatos3, Hercules Rigakis1, Maxim Smirnov4, Andreas Tzanis5,
Despina Kalisperi3, Fragkiskos Pentaris1,Vassileios Saltas3, Ioannis
Barbounakis1, Ilias Papadopoulos3, George Hloupis6, Paolo Spalla7,
Adriano Nardi2, Luca Spogli2, Roberto Carluccio2, Enrico Zuccheretti2,
Lucilla Alfonsi2, Pantelis Soupios3 and Eleni Kokinou3
1
2
3
4
5
6
7
Dept of Electronic Engineering, Technological Educational Institute of Crete, Greece,
*[email protected]
Dept of Geomagnetism, Aeronomy & Environmental Geophysics, Istituto Nazionale di
Geofisica e Vulcanologia, Italy
Dept of Environmental and Natural Resources Engineering Technological Educational
Institute of Crete, Greece
Dept of Physical Sciences, University of Oulu, Finland
Dept of Geology and Geoenvironment, National & Capodistrian University of Athens,
Greece
Dept of Electronic Engineering, Technological Educational Institute of Athens, Greece
Consiglio Nazionale delle Ricerche, Institute of Applied Physics, Italy
Keywords: seismoelectromagnetic, lithosphere, ionosphere, coupling, seismic
precursor, Hellenic Arc.
Abstract
Historically, the predominant view of the scientific community seems to lean against earthquake
prediction, especially the short-term one. Taking into account the complex nature and rarity of
strong seismic events, as well as the form multiplicity and timing variety of possible preseismic
signatures, someone easily understands the grounds of the aforementioned views.
Although earthquakes are considered chaotic events due to the heterogeneity of the Earth’s
crust, science and technology are challenged to come up with prognostic power on earthquakes
based on the understanding of the underlying physical mechanisms.
Project No 23 TeCH-SEM “Technologies Coalescence for Holistic SeismoElectroMagnetic Research (Lithosphere-Atmosphere-Ionosphere Coupling)” of the Archimedes III Call “Support
Research Groups at TEI of Crete” of EPEDBM is establishing a multidisciplinary network of
telemetric stations to study in long term the lithosphere-atmosphere-ionosphere coupling at the
seismically active south part of the Hellenic Arc.
The present work reports the technologies being implemented for a broadband seismo-electromagnetic (SEM) investigation of the seismic region of South Aegean, Greece and with emphasis for the study of lithosphere-ionosphere coupling. Furthermore, we discuss candidate electric
field precursory signals recorded from single stations prior major seismic events.
ISBN: 978-960-98739-8-7
207
Introduction
Earthquakes are natural disasters with ever increasing catastrophic effects, mainly due to the
urban and industrial development of seismically active regions and despite the substantial increase of our technological capacity to build robust and resistant structures.
At present only two approaches exist in order to defend the society from the destructive effects
of earthquakes: prevention and prediction. In our opinion, it is necessary to develop and apply
both these complementary approaches, also taking into account that they are at a very different
level of maturity and applicability.
Despite some claims in the seismological community, the feasibility of earthquake prediction
has not been refuted scientifically (Gerstenberger et al, 2005). On the other hand, the slow
progress hitherto achieved in earthquake forecasting, emphasizes the difficulty of the effort
(which in part is due to the absence of multidisciplinary interaction).
In this context, seismoelectromagnetic (SEM) research appears to be a promising approach for
earthquake prediction problem, as well as an innovative one, because it requires the convergence of many different disciplines which extend from the geosciences (geology, seismology,
volcanology, geoelectromagnetism, geophysics) to aeronomy, the physics of ionosphere and
radio-waves science.
Project No 23 TeCH-SEM “Technologies Coalescence for Holistic SeismoElectroMagnetic Research (Lithosphere-Atmosphere-Ionosphere Coupling)” of the Archimedes III Call “Support
Research Groups at TEI of Crete” of EPEDBM adopts exactly this multidisciplinary approach
to improve the odds of detecting and identifying the elusive electric, magnetic and electromagnetic precursory signals in the ULF to HF range.
Seismoelectromagnetic Network at the south front of the Hellenic Arc
Literature is abundant with reports of pre-earthquake phenomena, mainly of electric, magnetic
or electromagnetic nature (Park et al, 1993, Teisseyre, 1997, Tzanis and Vallianatos, 2001, Hayakawa and Molchanov, 2002, Varotsos, 2005, Sgrigna et al, 2007, Pulinets and Ouzounov 2011
etc). During the last two decades, understanding of seismoelectromagnetic signals is cuttingedge research which integrates laboratory experiments with theoretical advances (models) and
field observations.
To this direction, seismoelectromagnetic stations of novel design are being installed and operated in the south front of the Hellenic Arc, featuring near real-time telemetry, extended autonomy (using solar panel), light-weight and small-size but robust and powerful datalogging,
self-diagnostics and smart power control for low-noise, reliable, long-term operation. Since
June 2013 a (SEM) station (Fig. 1) is operating at Omalos plateau, W. Crete (35.33 Ν, 23.89 Ε).
A second one was installed on April 2014 in Rhodes Island (36.17 Ν, 27.97 Ε).
Figure 1. The SEM station located at Omalos plateau, W.Crete, featuring 3G telemetry and solar energy harvesting.
208
ISBN: 978-960-98739-8-7
Despite the short recording history of these stations, there are significant indications that probably both stations are sensitive to the detection of preseismic signatures. Figure 2 maps (Google
Earth product) the epicentres for the earthquakes with M ≥ 4.5 as reported by EMSC. The
geographical window is: 34.5 – 37.0 N, 22.5 – 29.0 E and the time period is from July 2013 till
April 2015. The two continuously operating seismoelectromagnetic stations of Omalos (Crete)
and Rhodes stations are also depicted. Green dot line ellipses indicate the cases of seismic
activity for which it is investigated the manifestation of precursory seismoelectric signatures
observed at either station.
Figure 2. Map (34.5 – 37.0 N, 22.5 – 29.0 E) of the epicentres for the earthquakes with M ≥ 4.5 as reported by
European Mediterranean Seismological Centre (EMSC). The colours of epicentre circle denote shallow (red),
intermediate depth (yellow) and deep (green) earthquakes. The SEM stations of Omalos (Crete) and Rhodes are
also depicted.
The case of the strong M6.4 earthquake of October 12, 2013
On October 12, 2013, at 13:11:54 UTC, occurred a very strong earthquake (M6.4) with epicentre (35.56 N, 23.31 E) approximately 60km WNW from Omalos’ SEM-station with depth 47km
(Fig. 3 left). This earthquake was the strongest event that took place in Greece the last three
years, a region that concentrates more than 80% of the seismic activity in Europe.
Omalos SEM
station
Figure 3. left: Epicentre map of the strong earthquake (M6.4) of October 12, 2013, at 13:11:54 UTC, occurred
~60km WNW from Omalos’ SEM-station with depth 47km. The aftershocks of first 24h are also depicted. right:
Swarm of electrical anomalies recorded at Omalos’ SEM station on July 21, 2013 mainly in NS component but
not in the magnetics (see the respective magnetic component).
ISBN: 978-960-98739-8-7
209
Careful retrospective examination of the history records of the Omalos’ SEM-station revealed
the existence of transient electrical anomalies that appeared mainly on ENS channel (very weak
on EEW) and do not accompanied by magnetic anomalies. These anomalies appeared for first
time on July 21, 2013 (81 days before the earthquake) lasting from 09:04 to 14:25 UTC (Fig.
3 right). Same type of anomalies appeared again on August 25, on October 11 (one day before
the earthquake) and hours before the earthquake and possibly associated with the M6.4 major
seismic event and its aftershocks.
Lithosphere-Ionosphere Coupling: Experimental Observations in the Mediterranean
Area
The study of the geophysical phenomena that appear prior, during, and after a seismic event
has increased considerably over recent decades. This has been the object of research of numerous scientists from different disciplines, including, in addition obviously to seismology, also
hydrology, geomagnetism, atmospheric physics, geochemistry, as well as ionospheric radio
propagation (Zolesi and Cander, 2014).
Experimental observations both from ground based measurements and from space satellites
have been performed or are in progress to study coupling of the lithosphere and ionosphere,
with the possibility of determining an ionospheric precursor to earthquakes (Parrot et al, 1994,
Sgrigna et al, 2007, Pulinets and Ouzounov 2011).
In fact ionospheric effects of lithospheric phenomena could be considered together with a larger
class of studies defined as electromagnetic effects of seismic events. A short review on the
important facts in the last 30 years about lithosphere-ionosphere coupling is reported taking
into account the numerous list of papers recently published and presented to the European
Geophysical Union General Assembly. An overview of the principal ionospheric experimental
observations active in the Mediterranean area is also reported as well as the basic physical principles of these measurements.
Ionospheric Scintillation and TEC Monitoring
Space weather conditions have a great influence on GNSS integrity, accuracy and availability as
the ionosphere is the largest contributor to the error budget for GNSS receivers. Moreover, the
inhomogeneity of ionospheric electron distribution (i.e. ionospheric irregularities) can cause
sudden, rapid and irregular fluctuations of the amplitude and phase of the received signals,
termed “ionospheric scintillation”.
Given the morphology of the Earth’s magnetic field, the geographic regions in which scintillations are more likely to occur are the polar and the equatorial areas, exacerbating in the low
latitude regions. Scintillation can cause degradation on GNSS measurements and, in the worst
case, can lead to the loss of lock with the satellite, affecting the availability of the GNSS based
services. Amplitude scintillation is traditionally monitored by means of the S4 index, which
is the standard deviation of the received power normalized by its mean value, whereas phase
scintillation is monitored by the σϕ index, which is the standard deviation of the detrended carrier phase. In modern high sampling frequency (50 Hz) receivers for scintillation monitoring,
the time intervals in which such indices are calculated are of 60 seconds for S4 and 1, 3, 10, 30
and 60 seconds for σϕ.
210
ISBN: 978-960-98739-8-7
Figure 4. The GNSS receiver installed and operating at TEI of Crete premises in Chania.
INGV and TEI of Crete manage a network of two high rate (50 Hz) sampling GNSS receivers
at low Mediterranean sites: Lampedusa, Sicily and Chania, Crete (see Fig. 4). The significant
amount of data collected since 2008 has been structured and made accessible to registered users
via the eSWua service: www.eswua.ingv.it. The research activities focus on the morphology
and dynamics of the ionospheric plasma to improve the capabilities of TEC and scintillation
mapping and modelling that may reveal ionospheric variations as earthquake forerunners (Alfonsi et al, 2007).
Acknowledgements
This research is implemented in the framework of the project entitled “Technologies Coalescence for Holistic Seismoelectromagnetic Research (Lithosphere-Atmosphere-Ionosphere
Coupling)” of the Archimedes III Call through the Operational Program “Education and Lifelong Learning” and is co-financed by the European Union (European Social Fund) and Greek
national funds.
References
1. Alfonsi, L., Mitchell, N., Romano, V., Spalla, P., 2007, MIRTO: a prototype for real time ionospheric imaging
over Mediterranean area, Ann. Geophys., 50(3), 447-452.
2. Gerstenberger, M., Wiemer, S., Jones, L. M., and Reasenberg, P. A., 2005, Real-time forecasts of tomorrow’s
earthquakes in California, Nature, 435, 328-331.
3. Hayakawa, M., and Molchanov, O. A., (eds), 2002, Seismo Electromagnetics: Lithosphere-AtmosphereIonosphere Coupling, (Tokyo: Terra Sci. Pub).
4. Park, S., Johnston, M., Madden, T., Morgan, D., and Morrison, F., 1993, Electromagnetic precursors to
earthquakes in the VLF band: A review of observations and mechanisms, Rev. Geophys., 31, 117-132.
5. Parrot, M., 1994, Statistical study of ELF/VLF emissions recorded by a low-altitude satellite during seismic
events, J. Geophys. Res., 99, 23339–23347.
6. Pulinets, S., Ouzounov, D., 2011, Lithosphere–Atmosphere–Ionosphere Coupling (LAIC) model – An unified
concept for earthquake precursors validation, J. Asian Earth Sci., 41(4/5), 371–382.
7. Sgrigna, V., Buzzi, A., Conti, L., Picozza, P., Stagni, C., and Zilpimiani, D., 2007, Seismo-induced effects in
the near-earth space: Combined ground and space investigations as a contribution to earthquake prediction,
Tectonophysics, 431, 153–171.
8. Teisseyre, R., 1997, Generation of electric field in an earthquake preparation zone, Ann. Geofis., 40, 297–
304.
9. Tzanis, A., and Vallianatos, F., 2001, A Critical Review of Electric Earthquake Precursors, Annali di Geofisica,
44(2), 429-460.
10. Varotsos, P. A., 2005, The Physics of Seismic Electric Signals (Tokyo: Terra Sci. Pub).
11. Zolesi, B., and Cander, L., 2014, Ionospheric Prediction and Forecasting, (London: Springer).
ISBN: 978-960-98739-8-7
211
Spectral fractal analysis of MHz electromagnetic signals
Dimitrios Nikolopoulos1*, Petraki Ermioni1, Cantzos Demetrios1 , Grigorios
Coulouras2 and Constantinos Nomicos2
Piraeus University of Applied Science ,Petrou Ralli & Thivon 250, GR122 44, Aigaleo, Greece
*[email protected]
2
TEI of Athens, Agiou Spyridons, Aigaleo, GR 122 44 Greece
1
Keywords: earthquakes, fractal analysis, long memory.
Abstract
In the following, results will be presented from a wavelet-based power spectral analysis of
fractals applied in electromagnetic time-series of the MHz range. By tracing pre-earthquake
long-memory patterns hidden in the investigated time-series, the analysis attempts to investigate possible associations of the results with thirty seven seismic events. The events occurred
in Greece from 2007 to 2015 with local magnitudes M L ≥ 5.0 . The majority of the investigated
time-series exhibited numerous parts associated with long-memory of the earthquake generation system. The findings were considered also indicative of the self-organised critical states of
the last stages of preparation of the investigated earthquakes. The results showed that spectral
fractal analysis achieves to detect successfully if the signal follows the most popular Gaussian
1 / f processes: fractional Brownian motion (fBm) or fractional Gaussian noise (fGn) models.
The precursory value of the signals was also discussed.
Introduction
1. Fractal analysis from wavelet-based power spectral density
During the complex process of earthquake preparation, linkages between space and time produce characteristic fractal structures (e.g. Eftaxias, 2010; Petraki, et al., 2013a,b and references
cited in these publications). It is expected that these fractal structures affect signals rooted in
the earthquake generation process. The power spectral density, S ( f ), is probably the most commonly used technique to provide useful information about the inherent memory of the system
(Eftaxias, 2010; Petraki et al., 2013a,b). Although the power spectrum is only the lowest order
statistical measure of the deviations of the random density field from homogeneity, it directly
reflects the physical scales of the processes that affect structure formation (Eftaxias, 2010).
If a recorded time-series, A(ti ), is a temporal fractal, then a power-law spectrum is expected,
i.e., S ( f )= a ⋅ f − b ,where f is the frequency of the transform. In the log(S ( f ))− log( f ) representation, the power spectrum is a straight line, with linear spectral slope b . The spectral
amplification quantifies the power of the spectral components following the power spectral
density law. The spectral scaling exponent is a measure of the strength of time correlations. The
goodness of the power-law fit to a time-series is represented by the Square of the Spearman‘s
correlation coefficient (Petraki, et al., 2013a). Attention is paid to whether distinct changes in
the scaling exponent emerge before or during any detected bursts or anomalies. In this study, the
continuous wavelet transform was used with the Morlet wavelet base function.
For the application of the fractal analysis from wavelet-based power spectral density (hereafter
abbreviated as spectral fractal analysis) the following steps were followed: (i) The MHz electromagnetic signals were divided in segments (windows): 1024-2048 samples per segment. These
212
ISBN: 978-960-98739-8-7
segmentations were expected to reveal the fractal characteristics of the signals (e.g. Eftaxias, 2010; Petraki et al., 2013a,b); (ii) In each segment the PSD of the signal was calculated.
As aforementioned, for the PSD calculation, the DWT using the Morlet wavelet was employed; (iii) In each segment the existence of a power-law of the form of S ( f )= a ⋅ f − b was
investigated. In DWT PSD calculation, the employed frequency f was the central frequency
of the Fourier transform of the Morlet scale; (iv) The least square method was applied to the
log(S ( f ))− log( f ) linear representation. Successive representations were considered those that
exhibited squares of the Spearman’s correlation coefficient above 0.95, i.e., 95% confidence
interval.
2. Earthquakes analysed & criteria for selection of earthquakes
Greece is a country prone to earthquakes because it is bounded by the regions of convergence of
the Eurasian and African plates and the western termination zone of the North Anatolian Fault
(Figure 3.3). The country is dominated by extensional seismicity structures and numerous active faults of complex active stress field. The seismicity structures evoked several earthquakes
of M L ≥ 5.0 and M w ≥ 5.0 during the last century. During the period of the analysis of this research, i.e., between January 2007 and May 2015, several earthquakes occurred in Greece and
near with local magnitudes M L ≥ 5.0 and depths from 2 to 165 km. Thirty seven of these events
were analysed in this thesis. The total number of the events with M L ≥ 5.0 between 2007 and
2015, the number of the analysed events through spectral fractal analysis and the succeeded rate
of analysed events are presented in Table 1.
Table 1: Total number of earthquakes with M L ≥ 5.0 between 2007 and 2015, number of the
analysed events through spectral fractal analysis and the succeeded rates.
Year
Number of events Analysed events
2007
2008
2009
2013
2014
2015
with M L ≥ 5.0
7
20
10
8
12
3
4
12
9
3
6
3
Rate of the analysed
events
57.0%
60.0%
90.0%
37.5%
50.0%
100%
It can be observed from Table 1 that the selected seismic events analysed with fractal methods,
have local magnitudes M L ≥ 5.0 . The criterion for choosing these earthquakes was the national
risk that these may cause because of their relatively high magnitudes. A percentage of 67.5% of
the studied events with M L ≥ 5.0 in the years 2007 to 2015, were seismic events that occurred
in the Hellenic Trench. The Hellenic Trench is developed in the western part of the Hellenic
Arc and for this reason it is an area of great seismicity and interest. Hence, it is significant that
the majority of the studied events corresponded to earthquakes occurred in the Hellenic Trench.
Additionally, a percentage of 5.5% of the analysed earthquakes, were seismic events that occurred near the Anatolian Plate. This is also of significance. The remaining 27% of the studied
events occurred in scattered areas of the Aegean Sea Plate. From Table 1 it is observed that the
percentages of the implemented analysis was above 50% for all years except 2013. This marks
ISBN: 978-960-98739-8-7
213
up a fairly good proportion for the analysis implemented with PSD based fractals. The overall
analysis is significant and towards this point certain facts are outlined below.
A set of characteristic results based on the spectral fractal analysis are presented in Figs 1 and
2 (U: underwater, S:Shallow). The signals are electromagnetic of the MHz-range and extended
approximately one month before earthquakes. The top sub-figures present the time evolution
of the fractal power-law b - values. Each value correspond to the log − log power-law fit in
a window of 1024 samples. To obtain each power-law b - value, the window was slid one
sample forward. This technique is more difficult for real time processing, as long-memory
analysis methods operate on successive, fixed length segments (e.g. 1024 samples), advancing
one sample ahead each time i.e. the sliding window step is one sample (Cantzos et al., 2015).
With these parameters, the analysis captures very fine EM variations in time at the expense of
high computational cost as any given fractal analysis window almost completely overlaps the
preceding one given the same window length and step size (Cantzos et al., 2015). The middle
sub-figures of present the average (in each window) r 2 , as well as the number of successive (
r 2 ≥ 0.95 ) values. The blue points indicate the successive ( r 2 ≥ 0.95 ) segments. Note that, as
aforementioned, the fBm class is associated with power-law b - values within 1 < b < 3 and the
fGn class with values in the range − 1 < b < 1 . Red points refer to the remaining segments (nonsuccessive fBm & non-successive fGn classes). The bottom sub-figures present the detected
signal. In the following figures, the term (EQ) refers to earthquake, the term (JD) refers to day
of occurrence of each earthquake in the Julian‘s calendar as well as to the range of days of the
fractal analysis. Finally, the term (S) refers to shallow earthquakes (epicentre‘s depth<10 km)
and (U) to undersea earthquakes (depth>30 km, epicentre under the surface of the sea). Since
there exist no one to one correspondence between earthquake events and detected disturbances
(e.g. Eftaxias 2010), the earthquakes in the captions of are indicated arbitrarily and only for the
purpose of referencing. The indicated earthquakes are those spatially closest to the recording
station and temporarily nearest to the presented day-range of fractal analysis.
Figure 1: (a) EQ of JD 257, 2009, Corfu station, Signal JD 227-257, 2009, 46 MHz (S); (b) EQ of JD 182,
2009, Vamos station, Signal JD 152-182, 2009, 41 MHz (U).
It can be supported that: (a) Significant number of time-series parts exhibited successive (
r 2 ≥ 0.95 ) power-law b - values between 1.5 and 2.0; (b)Numerous parts presented b - values
greater than 2.0. As aforementioned, b - values between 2.0 and 3.0 suggest persistent profiles
qualitatively analogous to the fBm model; (c) Some b - values were above 3.0. This is very
214
ISBN: 978-960-98739-8-7
peculiar since, according to the theory, the power-law b - values are normally in the range
− 1 < b < 3 ; (d) The loga - values associated with (a) and (b) were between 1.0 and 5.0. This
finding was also addressed in several publications (e.g.Eftaxias, 2010 and references therein);
(e) Periods with significant electromagnetic disturbances in the time-series are not associated
a-priori or de-facto with observations such as those of (a)-(d). All these are significant scientific
evidence of strong system‘s memory and are considered as footprints of earthquake occurrence.
Figure 2: (a) EQ of JD 47, 2009, Ileia station, signal JD 17-47, 46 MHz (U); (b) EQ of JD 170, 2009, Rhodes
station, signal JD 140-170, 2009 46 MHz (U).
Concluding Remarks
Fractal Analysis is a significant method to identify hidden long-memory trends in MHz preearthquake signals.
Acknowledgements
This work was co-financed by Greece and the European Union under the European Social Fund
NSRF 2007À2013 (Archimedes III). Managing Authority, Greek Ministry Of Education and
Religious Affairs, Culture and Sports.
References
1. Cantzos, D., Nikolopoulos, D., Petraki, E., Nomicos, C., Yannakopoulos, P.H., & Kottou, S. (2015). Identifying
Long-Memory Trends in Pre-Seismic MHz Disturbances through Support Vector Machines. Journal of Earth
Science & Climatic Change 6(3):1-9.
2. Eftaxias, K. (2010). Footprints of non-extensive Tsallis statistics, self-affinity and universality in the preparation
of the L’Aquila earthquake hidden in a pre-seismic EM emission. Physica A 389(1):133-140.
3. Petraki, E., Nikolopoulos, D., Fotopoulos, A., Panagiotaras, D., Koulouras, G., Zisos, A., Nomicos, C., Louizi,
A., & Stonham, J. (2013a). Self-organised critical features in soil radon and MHz electromagnetic disturbances:
Results from environmental monitoring in Greece. Applied Radiation and Isotopes 72:39–53.
4. Petraki, E., Nikolopoulos, D., Fotopoulos, A., Panagiotaras, D., Nomicos, C., Yannakopoulos, P., Kottou, S.,
Zisos, A., Louizi, A., & Stonham, J. (2013b). Long-range memory patterns in variations of environmental
radon in soil. Analytical Methods 5(16):4010-4020.
ISBN: 978-960-98739-8-7
215
ImproDeProF Project: Recent Advances and New Challenges in the
development of the DeProF tentative theory for steady-state twophase flow in porous media
Marios S. Valavanides
Lab. of Applied Mechanics, Dept. of Civil Eng., TEI Athens, [email protected]
Keywords: porous media, multi-phase flow, energy efficiency, flow characterization,
entropy, DeProF theory.
Abstract
The majority of industrial applications of two-phase flow in macroscopically heterogeneous
porous media are based on inherently transient processes. Nevertheless, to understand the
physics of such processes in a deeper context, we need first to understand the stationary case,
steady-state flow in macroscopically homogeneous media. To this play, the DeProF theory
implements hierarchical mechanistic modeling, i.e. scaling-up from pore- to core- to fracture
scales, to predict relative permeabilities in terms of the values of the operational parameters
(capillary number and flowrate ratio), given the values of the system parameters. This synopsis
focuses on latest research advances within the development of the DeProF tentative theory,
with scope in disseminating a portfolio of emerging research challenges and potential applications.
Introduction
To date, the characterization and modeling of multiphase flows in porous media has delivered
promising theoretical and experimental results over a hierarchy of scales (pore-to-production).
Nevertheless, integration across those scales remains an outstanding problem. Likewise, technological progress has enabled laboratory studies to expose latent flow mechanisms; pore-scale
phenomena and critical interstitial physical quantities can now be identified and assessed digitally, computationally or experimentally. At the same time, pragmatic sustainability issues on
energy production/management shifted “recovery increase” trends into “process efficiency
optimization” scopes and targets. As a consequence, new challenges emerge within a wide
spectrum of technological problems, extending from laboratory scales, e.g. design systematic
protocols for regular/special core analysis (R/SCAL) for data collection/interpretation, to industrial scales, e.g. unconventional/ enhanced oil recovery (EOR) /carbon capture & sequestration
(CCS), soil & aquifer pollution & remediation or operation of trickle-bed reactors (Valavanides
et al., 2015a)
To address these issues we need to think “out of the box” and decide to explore the untapped
potential of theoretical tools and methodologies translated from other scientific fields. Some of
the current approaches in modeling multi-phase flow propose a number of disruptive directions,
such as:
• Disconnected flow - a substantial and many times prevailing flow pattern (Georgiadis et al.,
2013, Youssef et al., 2014, Tsakiroglou et al., 2007).
• Inherent degrees of freedom - nested within the physics of multi-phase flows, they are expressed on the process stationary character even under steady-state flow conditions (Van de
Merwe and Nicol, 2009, Valavanides, 2012).
• Independent variables - reappraise the set without surrendering a specific-enough representation of the process phenomenology by oversimplification (Valavanides, 2012).
216
ISBN: 978-960-98739-8-7
Without neglecting the fact that the majority of industrial applications of two-phase flow in
macroscopically heterogeneous porous media are based on inherently transient processes, to
understand the physics of such processes in a deeper context, we need first to understand the
stationary case, steady-state flow in macroscopically homogeneous media.
To this play, the DeProF theory (Valavanides, 2012) implements hierarchical mechanistic modeling, i.e. scaling-up from pore- to core- to fracture scales, to predict relative permeabilities in
terms of the values of the operational parameters (capillary number and flowrate ratio), given
the values of the system parameters.
This research synopsis will focus on latest advances, indicative applications and future challenges in the development of a tentative theory for two-phase flow in porous media, namely the
DeProF theory.
Results
The research was carried during the past 5 years, mainly under the ImproDeProF project (2015).
It is the result of collaboration of a research group spanning 5 research organizations (see Acknowledgment). The results of the ImproDeProF project are as follows:
Theoretical developments
The delivery and proof-of-concept on the existence of a universal map that describes process
performance and delineates its operational efficiency was furnished by Valavanides (2014b),
including the delivery of a normative methodology for characterizing a process as to the predominance of capillarity or viscosity as well as its untapped efficiency margin.
Synergy with statistical thermodynamics will eventually correlate the multiplicity of physically
admissible, internal flow arrangements with the existence of optimum operating conditions
(Valavanides, 2010). In this context, Daras and Valavanides (2015) identified the ensemble of
physically admissible microstates for the sought process and furnished an analytical methodology –based on combinatorics- to estimate their number. This is a prerequisite step taken towards
developing a robust theoretical justification of the DeProF model predictions.
Laboratory work - experimental developments
Tsakiroglou et al., 2015, have performed laboratory experiments of steady-state two-phase
flows in packed sand columns with equal viscosity fluids to quantify the dependence of oil and
water relative permeability on capillary numbers, and correlate the estimated parameters of
power functions with the viscosity ratio. They provided new explicit relationships of relative
permeabilities and water saturation with oil and water capillary numbers to set the bases for a
new conceptualization of the two-phase flow at reservoir-scale where the mobility of the fluids
is decoupled from saturation and become non-linear functions of the local flow rates.
Valavanides et al., 2015b have provided a major re-examination of available experimental data
on the phenomenology of steady-state two-phase flow in porous media processes, recorded in
the conventional relative permeability diagrams. The objective was to test the DeProF model
predictions (hypothesis) on the existence of steady-state flow conditions, for which the energy
efficiency of two-phase flow in porous media processes attains a maximum value. The acquired
data were transformed into energy efficiency data sets for the corresponding system and flow
settings. This re-examination of relative permeability data sets from a total of 179 relative
permeability diagrams in 35 published laboratory studies, pertaining to a variety of steadystate two-phase flow conditions and types of porous media, provided extensive experimental
evidence on the existence of optimum operating conditions as well as on distinct trends of the
energy efficiency over the pertinent flow regimes and system configurations.
ISBN: 978-960-98739-8-7
217
Application developments
Within the ImproDeProF project, potential applications of the DeProF theory in process design and optimization have been furnished. The applications are focused on the exploitation
of the true-to-mechanism description of the sought process in terms of the actual independent
variables, i.e. the capillary number, Ca, and the flowrate ratio, r, as well as the incorporation
of multi-physics modeling, in improving process efficiency (Valavanides and Skouras, 2014),
and in providing more reliable simulations of industrial processes of two-phase flows in porous
media (Valavanides et al., 2015c, Skouras et al., 2015a,b).
Other (translational) developments
The problem of saturated flow within a homogeneous and isotropic pore formation, confined
between two horizontal impermeable planes, under 7-spot injection-extraction well pattern,
has been considered by Kamvyssas and Valavanides, 2015. To deal with this unconventional
geometry, a new method, for boundary value problems for the Laplace, the Helmholtz and the
modified Helmholtz equations in the interior of an equilateral triangle, was implemented.
In an effort to examine the transfer of project results into other scientific areas, implementation
of the maximum entropy production principle has been shown to provide a potential for optimizing project portfolio management (Valavanides, 2014a).
Concluding Remarks
The DeProF model has been developed on the purpose of providing a mechanistic modeling
tool that could explain (on physical principles) the complicated phenomenology observed in
two-phase flows in porous media (Valavanides, 2012). It turned out that the DeProF model was
self-contained and rigorous enough to be further exploited as a simulating tool, with the capability to reveal latent process characteristics. Further downstream the evolution path, the DeProF model, harnessed with a rational theoretical framework (namely, the maximum entropy
principle), eventually evolves into a new theory. The DeProF theory for steady state two-phase
flow in porous media is consistent with the pre-existing theory (Darcy’s fractional flow formulation) (nevertheless, it shows conventional wisdom to be not complete), shows remarkable
specificity, is tentative and dynamic in allowing for changes as new facts are discovered, has
the potential to serve as the current (for-the-time-being) theory for steady-state two-phase flow
in porous media.
The proof of progress and the results accomplished to date, show that there are many potential applications of this new theory in numerous theoretical and application oriented developments.
Acknowledgements
Research work has been co-funded by the European Union (European Social Fund) and Greek
national resources in the frameworks of the “Archimedes III: Funding of Research Groups in
TEI of Athens” (MlS 379389), of the “Education & Lifelong Learning” Operational Program.
The contribution of the ImproDeProF Project Team of fellow researchers, T. Daras (T.U. Crete),
V.N. Burganos, C.D. Tsakiroglou, C.A. Aggelopoulos, D.G. Avraam, E.D. Skouras, (FORTH/
ICE-HT), C.A. Paraskeva (U. Patras), A.N. Kalarakis (TEI West. Greece), T.T. Kokkinos (TEIA
Athens), is greatly acknowledged.
E. Totaj, and M. Tsokopoulos, graduate students, TEI Athens, and K. Terzi postgraduate researcher (FORTH/ICE-HT) have also contributed with their enthusiasm and hard work.
218
ISBN: 978-960-98739-8-7
References
1. Daras, T., Valavanides, M.S., 2015, Number of Microstates and Configurational Entropy for Steady-State
Two-Phase Flows in Pore Networks, AIP Conf. Proc., 1641, 147-154, http://dx.doi.org/10.1063/1.4905973
2. Georgiadis, A., Berg, S., Makurat, A., Maitland, G., Ott, H., 2013, Pore-scale microcomputed-tomography
imaging: Nonwetting-phase cluster-size distribution during drainage and imbibitions, Phys. Rev. E 88, 033002
DOI: 10.1103/PhysRevE.88.033002
3. ImproDeProF, 2015, Two-phase flow in porous media: Improvement of the Mechanistic Model DeProF and
implementation in practical applications, research project, “Archimedes III: Funding of Research Groups in
TEI of Athens” Contr. No MlS 379389, http://users.teiath.gr/marval/ArchIII/ImproDeProF.html
4. Kamvyssas, G. Valavanides, M.S., 2015, Analytical solution of the saturated flow problem in 7-spot, 2D
geometries, 5th Intern. Conf. on Environmental Management, Engineering, Planning and Economics, Mykonos,
Greece, July 14-18, http://users.teiath.gr/marval/publ/Kamvyssas_Valavanides_CEMEPE5_2015.pdf
5. Skouras, E.D., Valavanides, M.S., Burganos, V.N., 2015, A Model for Spatiotemporary Varying Mass Transfer
Problems During Two-Phase Flow Within Pore Networks, Based on the DeProF Model Description of the
Flow Patterns, 7th Intern. Conference on Porous Media, Padova, Italy, May 18-21, http://users.teiath.gr/marval/
ArchIII/Skouras_etal_IPOR2015-749.pdf
6. Skouras, E.D., Kalarakis, A.N., Valavanides, M.S., Burganos, V.N., 2015, Two-Phase Flow Calculations
in Pore Unit-Cells Implementing Mixed FEM/Lattice-Boltzmann Simulators, COMSOL 2015 Conference,
Grenoble, France, Oct. 14-16, http://users.teiath.gr/marval/publ/Skouras_etal_FEMLB_COMSOL2015.pdf
7. Tsakiroglou, C.D., Avraam, D.G., Payatakes, A.C., 2007, Transient and steady-state relative permeabilities
from two-phase flow experiments in planar pore networks, Advances in Water Resources 30, 1981–1992,
doi:10.1016/j.advwatres.2007.04.0029
8. Tsakiroglou, C.D., Aggelopoulos, C.A., Terzi, K., Avraam, D.G., Valavanides, M.S., 2015, Steady-state twophase relative permeability functions of porous media: A revisit, International Journal of Multiphase Flow, 73,
34–42, http://dx.doi.org/10.1016/j.ijmultiphaseflow.2015.03.001
9. Valavanides, M.S., 2012, Steady-State Two-Phase Flow in porous Media: Review of Progress in the
Development of the DeProF Theory Bridging Pore- to Statistical Thermodynamics- Scales, Oil & Gas Science
and Technology, 67(5), 787-804, DOI: 10.2516/ogst/2012056
10. Valavanides, M.S., 2014(a), Portfolios as off-equilibrium processes: similarities and affinities, Procedia Social and Behavioral Sciences, 119, 539-548, http://dx.doi.org/10.1016/j.sbspro.2014.03.060
11. Valavanides, M.S., 2014(b), Operational Efficiency Map and Flow Characterization for Steady-State TwoPhase Flows in Porous Media, Society of Core Analysts Symposium - SCA2014, Avignon, France, Sept. 8-14,
http://users.teiath.gr/marval/publ/Valavanides_SCA2014-047.pdf
12. Valavanides, M.S., Skouras, E.D., 2014, Rational solitary well spacing in soil remediation processes, Fresenius
Environmental Bulletin, 23(11), 2847-2851, http://users.teiath.gr/marval/publ/Valavanides_Skouras_
FEB_23_11_2014.pdf
13. Valavanides, M.S., Tsakiroglou, C.D., Ioannidis, M.A., Vizika, O. 2015, Unconventional modeling of multiphase flows in porous media. In Minisymposium 1.03, 7th Intern. Conference on Porous Media, Padova,
Italy, May 18-21, https://www.interpore.org/images/conferences/15Padova/minisymp_abstracts/MS_1_3.pdf
(accessed 09/2015)
14. Valavanides, M.S., Totaj, E., Tsokopoulos, M., 2015, Energy Efficiency Characteristics in Steady-State Relative
Permeability Diagrams of Two-Phase Flow in Porous Media, Journal of Petroleum Science and Engineering,
PETROL6936R1, 1-34 (accepted), http://users.teiath.gr/marval/publ/Valavanides_etal_JPSE_2015.pdf
15. Valavanides, M.S., Skouras, E.D., Kalarakis, A.N., Burganos, V.N., 2015, Integration of relative permeability
maps for two-phase flow in porous media into the COMSOL MultiphysicsTM Earth Science Module,
COMSOL Conference 2015, Grenoble, France, October 14-16, http://users.teiath.gr/marval/publ/Valavanides_
etal_COMSOL2015.pdf
16. Van de Merwe, W., Nicol, W., 2009, Trickle flow hydrodynamic multiplicity: Experimental observations
and pore-scale capillary mechanism, Chemical Engineering Science 64, 1267–1284, doi:10.1016/j.
ces.2008.10.069
17. Youssef, S., Rosenberg, E., Deschamps, H., Oughanem, R., Maire, E., Mokso, R., 2014, Oil ganglia dynamics
in natural porous media during surfactant flooding captured by ultra-fast x-ray microtomography, International
Symposium of the Society of Core Analysts, France, 11-18 Sept, 2014
ISBN: 978-960-98739-8-7
219
Microseismicity study in the area of Florina (Western Macedonia,
Greece)
Mesimeri M.1, Karakostas V.2, Papadimitriou E.3 and Tsapanos T.4
Geophysics Department, School of Geology, Aristotle University of Thessaloniki,
GR54124 Thesaloniki, Greece [email protected], [email protected], ritsa@
geo.auth.gr3, [email protected]
Keywords: local network, cross correlation, crustal model, double difference
Abstract
A local seismic network was installed and operated in the area of Florina during July 2013 –
January 2014 for identifying seismotectonic features of the active structure associated with an
earthquake of M=4.0 which occurred on February 17, 2013 and the ensuing high seismic activity. The recordings of the stations of the Hellenic Unified Seismic Network (HUSN) were additionally used and a new velocity model is defined for the broader area of Western Macedonia.
All the earthquakes were accurately located using the double difference technique and differential times from both catalog data and cross correlation measurements. The resulted locations
revealed in detailed the existence of different structures, responsible for the increased seismicity in the area for several months. Comparison of the focal coordinates of the common events
recorded in both networks indicates that the recordings of local networks could be used for the
correction and the significant improvement of the locations obtained by the regional network.
Introduction
The seismicity in the study area (Fig.1) is rather low with lack of major destructive earthquakes
and only five earthquakes with M≥5.0 (stars Fig.1). The largest one occurred in 1709 (M=6.0)
about 20 km south of Florina with limited damages to few temples according to Papazachos
and Papazachou (2003). Three earthquakes with M=5.4 occurred north of Florina in epicentral
distances of 20-30 km. There is one more earthquake occurred in 1984 with M=5.2 in a distance
of 40 km east-south-east of Florina. On 17 February 2013 an earthquake with M=4.0 occurred
near the city of Florina, followed by a large number of aftershocks. Several months later the
seismic activity of low magnitude earthquakes was very high. Since the permanent network in
the broader area is not adequately dense for microseismicity monitoring, a local seismological
network was employed and operated for six months in the activated area.
Data and Processing
The local seismic network which installed at the end of July 2013 and operated until the end of
January 2014, was consisted of 6 stations equipped with three component broad-band seismometers (Guralp CMG40T) and high resolution (24 bits) digitizers (Reftek 130-01). The recordings are digitized at a rate of 125 samples/second. Time and station location have been accurately evaluated by GPS receivers. Approximately 1,500 earthquakes were detected and P and S
-phases were manually picked. In addition, the recordings of 13 stations of the Hellenic Unified
Seismic Network (HUSN) located at epicentral distances of less than 150 km (hexagons Fig.
1) from the epicentral area were used. All the available phases of the earthquakes occurred in
the region in the period 2012 – 2014 which have been published in the monthly bulletins of
National Observatory of Athens (NOA) and the Geophysics Department of Aristotle University
of Thessaloniki were collected, and resulted to the locations of 521 events with magnitudes
ranging between 0.6 and 4.0.
220
ISBN: 978-960-98739-8-7
Figure 1. Morphological map of Western Macedonia along with the relocated epicenters of earthquakes recorded by the permanent network that occurred during 2012-2014 (blue circles) and these from the temporal
network during July 2013 – January 2014 (red circles). The historical and instrumental earthquakes with M≥5
are depicted by stars. The seismological stations of the HUSN are shown with hexagons and these from the local
network with triangles. Inset Map. The back-arc Aegean sea and the surrounding area, with the dominant seismotectonic features, including the Hellenic Arc where the subduction of the East Mediterranean lithosphere under the Aegean takes place, and the North Aegean Trough (NAT) which accommodates the westward migration
of the Anatolian plate into the Aegean. The tectonic setting is supplemented with the existence of the Cephalonia
(CTF) and Rhodes (RTF) transform faults. The study area is enclosed in the rectangle.
A new 1-D velocity model was obtained using the VELEST software (Kissling et al., 1994) and
the phases from the permanent network (HUSN) for the region. The selected data used for the
inversion were the ones with 4 or more P phases (425 events). As an a-priori velocity model was
used a combination of the ones suggested by Rigo et al. (1996) and Panagiotopoulos and Papazachos (1985). After repeating the inversion several times a new crustal model was obtained
along with its station delays for the permanent network. The obtained crustal model is similar
with the one proposed by Hatzfeld et al. (1997) for the area of Kozani.
The two datasets were relocated separately following the same procedure. Firstly, Wadati plots
were applied and a Vp/Vs ratio was calculated. Then using the new crustal model and the HYPOINVERSE software (Klein, 2002) the earthquakes were located. Station delays were calculated only for the stations of the local network following a procedure described by Karakostas
et al. (2014). The results were used as input in the double difference technique (Waldhauser,
2001). Then, waveforms were selected from both networks for the cross correlation processing.
The method described by Schaff et al. (2004) and Schaff and Waldhauser (2005) was followed
for computing differential times derived from waveform cross correlation measurements. Proper windows were selected for both datasets in order to get the best results.
There was a period that both networks were simultaneously in operation and there were about
35 events with solutions based on recordings from both of them. In order to test the robustness
of our results and the solutions stability of the permanent network, the common events were
located using separately the recordings of the local network and the HUSN. The results show
ISBN: 978-960-98739-8-7
221
that the depths of the earthquakes recorded by HUSN were displaced by approximately 2 km
deeper in comparison to these from the local network. Considering the locations of the local
network as the most accurate ones, we shifted these from the HUSN according to differences of
the common events of both networks. The final catalog includes 289 earthquakes recorded by
the permanent network and 1333 earthquakes from the local network.
Figure 2a shows the epicentral distribution of the relocated earthquakes. Red circles depict the
earthquakes recorded from the local network, blue circles depict the ones recorded from the
permanent network and with yellow star the largest earthquake during this study with M=4.0 is
shown. The epicenters of the earthquakes located using the recordings of the local network are
distributed at least in three different clusters. A cross section, normal to the trend of each one
of the clusters (Fig. 2b) along the line NN’, shows that the main cluster, located at the southern
part of the seismic zone is very well defined. The other clusters to the north could not be
connected clearly with individual structures. The foci of the earthquakes using the recordings
of the permanent network are rather sparse with most of them comprised in the middle cluster.
The major earthquake is located at the deeper part of the seismic zone. The cross section reveals
a north dipping zone starting from the southern cluster and ending to the major earthquake (star
Fig. 2b). This is in agreement with the focal mechanism of the largest earthquake available from
NOA (228/49/-122) and from the present study (225/53/-124).
Figure 2. (a) Epicentral distribution of the relocated catalog. The locations of HUSN are shown with blue
circles, with red circles these from the local network and with star is shown the 17/02/2013, M=4.0 earthquake.
(b) Cross section of the relocated earthquakes along the line of N-N’.
Concluding Remarks
In the present study a crustal model was proposed for the region of Western Macedonia, active
structures were identified using highly accurate locations and the locations of the permanent
networks were improved based on these from the local network.
Furthermore, perspective work including computation of fault plane solutions, definition of
source parameters with spectra analysis and the possible role of stress transfer would give a
clear interpretation of the seismotectonic setting in the area.
222
ISBN: 978-960-98739-8-7
Acknowledgements
The software Generic Mapping Tools was used to plot the figures (Wessel and Smith, 1998).
The authors express their sincere thanks to the Region of Western Macedonia and Regional
Unity of Florina for the financial support of the present study through the research project
‘Study of seismic activity in the area of Florina’.
References
1. Hatzfeld, D., Karakostas, V., Ziazia, M., Selvaggi, G., Leborgne, S., Berge, C., Guiguet, R., Paul., A., Voidomatis,
P., Diagourtas, I., Kassaras, I., Koutsikos, I., Makropoulos, K., Azzara, R., Di Bona, M., Baccheschi, S.,
Bernard, P. and Papaioannou C., 1997, The Kozani-Grevena (Greece) Earthquake of 13 May 1995 Reviseted
from a Detailed Seismological Study. Bulletin of Seismological Society of America, 87(2),463–473.
2. Karakostas, V., Papadimitriou, E. and Gospodinov, D., 2014, Modelling the 2013 North Aegean (Greece)
seismic sequence: Geometrical and frictional constraints, and aftershock probabilities. Geophysical Journal
International, 197(1), 525–541.
3. Kissling, E., Ellsworth, W.L. and Kradolfer, U., 1994, Initial reference models in local earthquake tomography.
Journal of Geophysical Research, 99(B10),19,635 – 19,646.
4. Klein, F.W., 2002, User’s guide to HYPOINVERSE-2000, a Fortran proram to solve earthquakes locations
and magnitudes. U.S. Geological Survey Open File Report 02 - 171, version 1.0.
5. Panagiotopoulos, D.G. and Papazachos, B.C., 1985, Travel times of Pn -waves in the Aegean and surrounding
area. Geophysical Journal of the Royal Astronomical Society, 80, 165–176.
6. Papazachos, B.C. and Papazachou, C.C., 2003, The earthquakes of Greece (Thessaloniki, Greece, Ziti
publications).
7. Rigo, A., Lyon-Caen, H., Armijo, R., Deschamps, A., Hatzfeld D., Makropoulos, K., Papadimitriou, P. and
Kassaras, I., 1996, A micro-seismicity study in the western part of the Gulf of Corinth (Greece): implications
for large-scale normal faulting mechanisms. Geophysical Journal International, 126, 663–688.
8. Schaff, D.P., Bokelmann, G.H.R., Ellsworth, W. L., Zanzerkia, E., Waldhauser, F. and Beroza, G.C., 2004,
Optimizing correlation techniques for improved earthquake location. Bulletin of the Seismological Society of
America, 94(2), 705–721.
9. Schaff, D.P. and Waldhauser, F., 2005, Waveform cross-correlation-based differential travel-time measurements
at the northern California seismic network. Bulletin of the Seismological Society of America, 95(6), 2446–
2461.
10. Waldhauser, F., 2001, HypoDD: A program to compute double-difference hypocenter locations.U.S. Geological
Survey Open File Report, 01-113.
11. Wessel, P. and Smith, W.H.F., 1998, New,improved version of the generic mapping tools released. Transactions
- American Geophysical Union, 79(47), 579.
ISBN: 978-960-98739-8-7
223
Evidence of accelerating deformation prior to large earthquakes: a
case study on the southern Hellenic Arc of Greece
G. Minadakis1*,G. Chatzopoulos1 and F. Vallianatos1
1
[email protected]
Keywords: critical earthquake model, accelerating deformation, power-law.
Abstract
The present study focuses on three cases of large earthquakes occurred at the Southern Hellenic
Arc and Trench system of Greece. We employ a critical earthquake model based on principles
of statistical physics, able to identify critical areas exhibiting accelerating intermediate magnitude seismicity. Using two different catalogues, we scan the southern Greece in terms of a
parametric modelling, for specific time periods before the occurrence of the main events under
study. Analysis reveals critical areas preceding accelerating behavior prior to the mainshock.
We further apply selective estimations on the deriving areas showing critical exponents that
indicate a terminal-stage critical point behavior, stable for different magnitude thresholds. This
work provides significant insights to an intermediate term prediction.
Introduction
A critical earthquake model has been proposed, able to identify critical areas that exhibit accelerating intermediate magnitude seismicity, and further provide potential estimations of the
time and location of ensuing mainshocks (Allegre and Le Mouel, 1994; Sornette and Sammis,
1995). According to this model, this accelerated moderate seismicity is a critical phenomenon
that potentially leads to the occurrence of ensuing large event, and could be considered as a
critical point for the prediction of earthquakes (Vanneste and Sornette, 1992; Lamaignereet al.,
1996). A large number of studies in the field, have provided evidence of strong earthquakes
which are preceded by accelerating generation of intermediate-magnitude events, considered as
preshocks of the main event (Jaumè and Sykès, 1999; Tzanis and Vallianatos, 2003;Papazachos
et al., 2005;). Expanding on this type of analysis the present study focuses on three cases of
large earthquakes occurred at the Southern Hellenic Arc and Trench system of Greece, in the
last decade. Firstly, we use a parametric modelling approach to scan the southern territory of
Greece using a shift of 0.1° in longitude and latitude. Analysis in terms of this modelling reveals critical areas preceding accelerating behavior prior to the mainshock. Secondly we apply
selective estimations on the deriving critical areas in order to show that the deriving parameters
of the model follow the theoretical ones, indicating a terminal-stage critical point behavior.
Finally we further verify the stability of the results in terms different thresholds of magnitudes.
This spatiotemporal analysis, provides significant information, contributing to the research for
earthquake prediction.
The critical earthquake model
According to the critical earthquake (CP) model, a strong earthquake may preceded by intermediate-magnitude precooks in which the time-evolution of the cumulative Benioff strain follows
a Power-Law behavior, expressed by means of the following relation:
224
ISBN: 978-960-98739-8-7
,
(1)
is the energy of the ith event and
is the total number of events at time . The enwhere
ergy
required by Eq. (2) can be estimated according to the Gutenberg and Richter (1956)
formula as:
.
Recent investigations on the CP model have empirically shown that the exponent typically is
close to
. Moreover the performance of the power law fit against the null hypothesis of
constant seismic release rate, is estimated by the curvature parameter given by:
,
(2)
In the following sections we further present observational evidence
where is typically
from the South Hellenic Arc.
Data analysis and results: observational evidence from the South Hellenic Arc
The seismicity data used in this study are taken from two different available catalogues: (i) the
bulleting of the Geodynamic Institute of the National Observatory of Athens, Greece (GI-NOA,
http://www.gein.noa.gr/services/cat.html), and (ii) the THALES relocated catalogue which
contains EQs up to 31-Dec-2013, and has been developed in the context the THALES project
``SEISMO FEAR HELLARC’’, funded by European Union and national sources. Specifically
we focus on three major EQs occurred in the southern territory of the well-known Hellenic Arc
of Greece. Table 1 represents the list of earthquakes and their additional parameter information.
1
Date From
01-Jan-2012
List of earthquakes under study
Date To
Lat.
Lon.
Mag.
15-Jun-2013 16:10:59
34.4488
25.0435 6.2
2
01-Mar-2011
12-Oct-2013 13:11:52
35.5142
23.2523
6.6
3
01-Mar-2014
16-Apr-2015 18:07:43
35.23
26.82
6.1
NUM
Catalogue
Location
THALES Tympaki
THALES Chania
GI-NOA Kassos
Table 1: List of examined earthquakes along with their additional parameters
A time-to-failure modelling is used to examine the spatial distribution of the parameters and
, deriving from Eq. (1) and (2) respectively. The south territory of Greece is scanned using a
step of 0.1o square grid over the area 19o E -30o E and 34o N - 37o N. Each step calculates the
curvature
for different radius (
) around each epicenter as
well as the relevant parameter . The optimal radius is considered that with the lowest curvature value and accurate fitting (
). For each calculation, the parameters obtained are
graphically placed at the center of each scanned area with different colors. A minimum number
of 20 events was used as a criterion for the statistical completeness of each calculation. Herein
is should be stressed that CP theory does not make any particular predictions about the configuration or shape of the critical region around the ensuing mainshock. Thus in order to ensure
that there will be not any conflict from EQs located in deeper layers, which in turn may affect
the results of the analysis, here we only examine the relative shallow EQs that do not exceed a
depth of
. The main event has been excluded from the analysis. Firstly we examine
whether intermediate magnitudes of the surrounding areas around the EQs 1 and 2 of Table 1,
ISBN: 978-960-98739-8-7
225
exhibit critical-point behavior before their occurrence. Indeed, Fig. 1 depicts the results of the
analysis for the periods included in Table1. The top chart of each figure, depicts the contour
map of the deriving curvature
for the optimal radius, without applying any filter. The
bottom charts depict the colored map for
,
and
.
The red-star-marks indicate the location of the EQ.
Figure 1: Time-to-failure modelling for the periods 1, 2 and 3 of Table 1, respectively.
From the top charts of Fig. 2, it is possible to distinguish between areas that exhibit accelerated
behavior and areas that are not, where the lowest curvature values are observed close to the
epicenter of the earthquakes under study. This is even clearer on the bottom charts where the
examined parameters have been filtered accordingly, and the remaining points are located very
close to the locations of the main events, with critical exponents very close to the typical ones
(
). In order to further examine the validity of the above mentioned results, we further
apply selective estimations of the areas defined according to previous modelling. Fig. 3 shows
both the linear and power-law fitting for the estimation of the parameters and , applied on
the cumulative Benioff stain release of the Eqs included in the examined areas.
Figure 3: Analysis in terms of critical earthquake model, for the periods 1, 2 and 3 of Table 1.
226
ISBN: 978-960-98739-8-7
The deriving critical exponents indicate a terminal-stage critical point behavior with exponents very close to theoretical ones: (i)
, (ii), and (iii)
respectively. On the contrary the deviation the original an estimated dates of occurrence is relative high, ranging between
and
months. In the prospect to show the consistency and stability of CP method used in this study, in Fig. 4 we show
the variation of estimated parameters
for different magnitude thresholds for each period under study. It is observed that both parameters conform to the typical limits and do not
exhibit significant variation for different magnitude thresholds, indicating the consistency of
the results.
Figure 4: CP analysis, for different magnitude thresholds for the periods 1, 2 and 3 of Table 1.
Discussion &Conclusions
Focusing on three cases of EQs located at the Southern Hellenic Arc and Trench system of
Greece, we examined the shallow seismicity of the broader surrounding area of the EQ epicenters, in terms of the CP model. Analysis identified specific areas very close to the main events,
preceding a critical point accelerating behavior prior to their occurrence. This in effect further
confirms the suggestion that many-fault interactions may account for the power-law behaviour
of the accelerating seismic release (Sornette and Sammis, 1995) with a critical exponent very
close to
and curvature lower that
, as shown in this study. Focusing on the date prediction
( in Fig. 3) in terms of the CP model, in this study the deviation of from the original dates of
occurrence is relative high, ranging between
and
months. In conclusion our analysis
has provided further evidence that the CP model provides reasonable results for the estimation
of the location of large and strong EQs, but the date prediction still remains an open unsolved
issue.
Acknowledgements
The work was supported by the THALES Program of the Ministry of Education of Greece
and the European Union in the framework of the project “SEISMO FEAR HELLARC’’, (MIS
380208).
References
1. Allègre, C. J., & Le Mouël, J. L. ,1994, Introduction of scaling techniques in brittle fracture of rocks. Physics
of the Earth and Planetary Interiors, 87(1), 85-93.
2. Lyakhovsky, V., Ben-Zion, Y., & Agnon, A. ,1997, Distributed damage, faulting, and friction. JOURNAL OF
GEOPHYSICAL RESEARCH-ALL SERIES-, 102, 27-635.
3. Jaumé, S. C., & Sykes, L. R. ,1999, Evolving towards a critical point: A review of accelerating seismic moment/
energy release prior to large and great earthquakes. In Seismicity Patterns, their Statistical Significance and
Physical Meaning (pp. 279-305). Birkhäuser Basel.
4. Papazachos, C. B., Karakaisis, G. F., Scordilis, E. M., & Papazachos, B. C. ,2005,. Global observational
properties of the critical earthquake model. Bulletin of the Seismological Society of America, 95(5), 18411855.
ISBN: 978-960-98739-8-7
227
5. Sornette, D., & Sammis, C. G. ,1995, Complex critical exponents from renormalization group theory of
earthquakes: Implications for earthquake predictions. Journal de Physique I, 5(5), 607-619.
6. Tzanis, A., & Vallianatos, F. ,2003, Distributed power-law seismicity changes and crustal deformation in the
SW Hellenic ARC. Natural Hazards and Earth System Science, 3(3/4), 179-195.
7. Vanneste, C., & Sornette, D. ,1992, The dynamical thermal fuse model.Journal de Physique I, 2(8), 16211644.
228
ISBN: 978-960-98739-8-7
Aftershock statistics of large earthquakes in Greece
G. Minadakis1, F. Vallianatos1 and I. Baskoutas2
1
2
Institute of Geodynamics, National Observatory of Athens, Athens, Greece
[email protected]
Keywords: aftershocks, seismicity, Gutenberg & Richter, Omori-Law, modified Omori
Law.
Abstract
This work examines 36 of aftershock sequences, most of them deriving from the Hellenic Arc
and Trench system of Greece, for the period between 1980 and 2015. Analysis is first carried
out by means of the: (i) Gutenberg & Richter Law to estimate the parameters of the magnitudefrequency relation, and (ii) the modified Omori’s Law to estimate the fitting parameter p-value.
We further apply a windowing method to examine the temporal evolution of aftershock sequences in terms of the p-value. This work provides significant information to the study of
aftershock sequences in Greece.
Introduction
The analysis of temporal properties of aftershock sequences has been an evolving field of research towards understanding the natural mechanisms behind the space and time domain distributions of earthquakes (EQs). In the last decades, several studies have been carried out focusing
on the statistical properties of aftershock sequences in terms of the Gutenberg & Richter (G-R)
magnitude-frequency relation, and the traditional Omori’s Law. Although the relation between
these two parameters has been extensively examined, there are both negative and positive conclusions indicating more complex mechanisms involved during the aftershock decay. On these
grounds, the present study focuses on 36 cases of large and relative shallow EQs, most of them
occurred on the Hellenic Arc and Trench system of Greece. Herein, analysis is carried out by
means of the G-R Law in order to estimate the b-value of the magnitude-frequency relation, and
the modified Omori’s Law in order to examine the behavior of the fitting parameter p-value,
respectively. First, we examine whether the two underlying metrics have common properties
for different types of aftershock sequences. Second, we use a windowing method to examine
the time evolution of p-value in order to find additional common characteristics, during the aftershock decay. This work provides useful information to the dynamical features of aftershock
sequences in Greece.
Background: setting the context for the analysis
The Gutenberg & Richter (Gutenberg, 1954) magnitude-frequency relation, reads as follows:
(1)
is the cumulative number of EQs with a magnitude greater than occurring in
where
a specified area and time. Parameters and are constants. In the last decades the underlying
formula has been mostly used to study the foreshock seismicity rather than the aftershock series
since it has been shown that the frequency of occurrence of EQs in aftershock series rapidly
follows the Omori’s law, where, the rate (n(t)) decay of aftershock time series is proportional to
, where is the time elapsed from the main event. Later on, Utsu et al. (1995) used a modiISBN: 978-960-98739-8-7
229
fied Omori’s law that reads as follows:
, (2)
where, and are and is near unity. In this work the estimation of the p-value, has been obtained by applying a nonlinear fitting procedure to the cumulative number of events which in
turn represents the integral of the modified Omori’s law (Utsu et al., 1995), as follows:
(3)
Data analysis and results: observational evidence from the South Hellenic Arc
The earthquake catalogue used in this study mainly derives from the data published in the bulletins of the Institute of Geodynamics of the National Observatory of Athens (NOAGI). Specifically we focus on 10 clusters of aftershock sequences contained in the wider Hellenic Arc
of Greece and 7 clusters of aftershock sequences contained in the rest of Greece as shown in
Fig. 1, for the specified periods under study. The star-marks indicate the location of the mainshocks.
Figure 1: Spatial distribution of aftershock sequencies included in Table 1. The star-marks indicate the location
of the mainshocks .
Table 1, shows a detailed list of the earthquakes under study along with their additional parameters. For reasons of clarity, the dataset includes six aftershock sequences occurred in Greece
from 1980 to 1997, that have been deeply analyzed in (Telesca et al., 2001), in terms of the
modified Omori Law (Utsu, 1995) and a set of fractal tools. The completeness of each series
under study has been estimated in terms of the G-R magnitude-frequency relationship. All aftershocks series have been sampled thoroughly from the whole source region of the mainshock, in
230
ISBN: 978-960-98739-8-7
comparison to the rate of the background seismicity. Please also note that some of the sequences
included in the table contain large and strong aftershocks that have been examined separately.
Num
Date of Ocurrence
Lon.
Lat.
Depth
Magn.
Mc
N
Days
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
09-Jul-1980 02:11:57
24-Feb-1981 20:53:37
25-Feb-1981 05:08:16
04-Mar-1981 21:58:07
19-Dec-1981 14:10:49
27-Dec-1981 17:39:12
17-Jan-1983 12:41:30
23-Mar-1983 23:51:07
13-May-1995 08:47:17
20-Jul-1996 00:00:43
18-Nov-1997 13:07:36
07-Sep-1999 11:56:50
24-May-2000 05:40:37
02-Dec-2002 04:58:56
14-Aug-2003 05:14:53
06-Aug-2006 07:49:43
14-Feb-2008 10:09:23
26-Feb-2008 10:46:05
08-Jun-2008 12:25:27
21-Jun-2008 11:36:22
14-Oct-2008 02:06:34
16-Feb-2009 23:16:38
24-May-2009 16:17:50
01-Jul-2009 09:30:13
18-Jan-2010 15:56:09
23-Nov-2011 12:17:48
26-Jan-2012 04:24:58
06-Apr-2013 11:26:06
15-Jun-2013 16:11:01
16-Jun-2013 21:39:04
07-Aug-2013 09:06:51
12-Oct-2013 13:11:53
26-Jan-2014 13:55:42
03-Feb-2014 03:08:44
08-Nov-2014 23:15:42
16-Apr-2015 18:07:44
22.910
22.970
23.200
23.260
25.300
24.900
20.250
20.400
21.710
27.140
20.490
23.620
22.010
21.150
20.560
19.490
21.780
21.700
21.510
21.830
23.620
20.780
22.740
25.400
21.950
25.030
25.070
24.110
25.060
25.130
22.680
23.280
20.530
20.390
20.440
26.820
39.290
38.230
38.190
38.240
39.200
38.900
37.970
38.190
40.180
36.210
37.260
38.150
36.000
37.800
38.790
40.210
36.500
35.960
37.980
36.030
38.850
37.130
41.300
34.350
38.410
34.120
36.060
34.820
34.340
34.280
38.700
35.500
38.220
38.250
38.100
35.230
47.5
17.6
33.0
31.9
10.0
10.0
14.0
12.6
39.0
26.8
10.0
30.0
5.0
18.7
12.0
14.0
31.5
5.0
15.8
12.0
24.0
13.2
23.0
13.7
20.0
10.0
29.0
24.0
32.0
28.0
8.0
65.0
21.0
10.0
18.0
37.0
6.0
6.3
5.1
5.8
6.3
6.0
6.2
5.7
6.1
5.9
6.1
5.4
5.4
5.3
5.9
5.1
6.2
5.2
6.5
5.5
5.6
5.5
5.1
5.8
5.2
5.1
5.3
5.1
5.8
5.6
5.2
6.2
5.8
5.7
5.0
6.1
3.3
3.2
3.2
3.2
3.1
3.1
3.7
3.7
3.0
3.6
3.0
3.1
3.0
3.0
2.9
3.2
3.0
3.0
3.0
3.0
2.1
3.1
3.0
3.3
2.5
2.0
2.5
2.3
3.0
2.4
1.4
1.7
2.5
2.0
1.4
2.6
173
149
135
110
123
71
230
70
404
332
819
165
109
44
255
43
121
1214
310
544
80
92
62
195
711
52
118
111
88
102
1426
214
773
4822
925
294
40.3
1.7
6.1
2.1
4.2
17.9
28.6
5.7
14.9
37.9
47.3
37.3
10.0
3.9
12.6
15.4
5.2
71.7
10.7
161.2
3.8
8.0
3.0
13.6
13.5
7.1
4.1
87.0
5.1
7.3
38.2
38.8
7.5
129.9
52.8
9.3
Table 1: List of examined earthquakes
Figs 2a and 2b, indicatively depict the analysis for the 33, 34, 35 and 36 samples included in
Table 1. The main event has been excluded.
ISBN: 978-960-98739-8-7
231
Figure 2a: The G-R relation and the Omori Law for the cases 33, 34 of Table 1.
Figure 2b: The G-R relation and the Omori Law for the cases 35 and 36 of Table 1.
It is clearly observed that the quality of fitting process for both metrics is fairly satisfactory.
In Fig. 3 we show a comparative analysis between the estimated p-value (depicted with blue
bars) and the b-value (depicted with blue bars) for all the examined time series of Table 1. It is
observed that there is not any common features between the two metrics since there is a variety
of combinations. The p-values range between 0.53 - 2.46, and b-values between 0.61 - 2.81.
232
ISBN: 978-960-98739-8-7
Figure 3: Comparative analysis between b-value and p-value for the list of earthquakes included in Table 1.
We further apply a windowing method by means of Eq. (3), in order to examine potential
variations that may lead to infer numerical investigation. Starting from an initial sample of 50
events (for statistical completeness), we use a custom increasing step of 10 events, and we calculate the p-value of the deriving series. Fig. 4 depicts the result of this approach cases 33 and
36 of Table 1, where an abrupt onset is observed right after the mainshock.
Figure 4: Windowing analysis in terms of Eq. (3) for the 33 and 36 samples included in Table 1.
Discussion & Conclusions
We studied 36 aftershock sequencies in the whider region of Greece in terms of the G-R
magnitude frequency relation and the intergration of the modified Omori Law. The results
reveal that there is not any significant correlation between the two metrics when studied for
spesific time periods. On the contrary, the time evolution of aftershocks reveals that in several
cases under study there is a significant onset of the p-value right after the occurence of the main
shock, gradually stabilised at lower values as the sequence decays. This in effect indicates that
the system does not always release the stress completely after the mainshock, and some patches
of the fault still remain unruptured.
Acknowledgements
The work was supported by the THALES Program of the Ministry of Education of Greece and
the European Union in the framework of the project “SEISMO FEAR HELLARC’’, (MIS
380208).
References
1. Gutenberg, B., Richter, C.F., 1944. Frequency of earthquakes in California. Bull. Seismol. Soc. Am. 34, 185– 188.
2. Telesca, L. Cuomo, V. Lapenna, V. Vallianatos, F. Drakatos, G., 2001. Analysis of the temporal properties of
greek aftershock sequences, Tectonophysics 341 (1) , 163-178.
3. Telesca, L. Lapenna, V. Macchiato, M., 2005. Multifractal uctuations in seismic interspike series, Physica A:
Statistical Mechanics and its Applications 354, 629-640.
4. Utsu, T., Ogata, Y., 1995. The centenary of the omori formula for a decay law of aftershock activity., Journal
of Physics of the Earth 43 (1), 1-33.
ISBN: 978-960-98739-8-7
233
Automatic Image Orientation for Accurate Texture Mapping of 3D
City Models
L. Grammatikopoulos1*, I. Kalisperakis1and E. Petsa1
1
Department of Civil Engineering and Surveying & Geoinformatics Engineering,
Technological Educational Institute of Athens, Greece, *[email protected]
Keywords: Texture Mapping, Lidar, 3D city models, Image orientation
Abstract
In the last decade, 3D city model generation is proved to be an indispensable task for a variety of
applications related to tourism, real-estate management, 3D GIS, cadastre, extending common
geospatial products such as maps and orthophotos to more realistic representations. Besides the
acquisition of high fidelity 3D data, which is usually performed by airborne Lidar scanners or
automated photogrammetric procedures, a crucial step for the generation of visually detailed
3D models is texture mapping from aerial imagery. In this contribution we propose a framework for accurate texture mapping of 3D models with the integration of vertical and oblique
aerial images, which consists of three main steps: In the beginning, the image set is oriented
through a Hierarchical Structure from Motion scheme. Afterwards, a registration is performed
with respect to the existing 3D city model and finally, texture mapping is applied using a multiimage blending technique. The proposed approach has been evaluated on a dataset covering a
densely build neighbourhood of Athens in Greece which includes a DSM generated from Lidar
an orthomosaic and a set of unoriented overlapping vertical and oblique aerial images.
Introduction
The generation of photorealistic 3D city models is a task of great interest for many researchers
in the fields of Photogrammetry and Computer Vision. The two main issues that are involved
and studied are image to model registration and texture mapping. Concerning the registration
stage, Delara et al, 2004 and Habib et al, 2005 extract manually Ground Control Points required
for the registration phase and apply a bundle adjustment solution. Armenakis et al, 2012 use
a semi-automatic approach for the alignment of photogrammetric and Lidar data based on the
identification of corresponding buildings in the two datasets. Gneeniss et al, 2014 optimize
their registration by extracting automatically Lidar control points and integrating them into a
final self-calibrating aerotriangulation. On the other hand, Yang and Chen, 2015 combine miniUAV images with Lidar data. In their approach images are first oriented via a Structure from
Motion (SfM) algorithm and the obtained point cloud is georeferenced based on a building
matching scheme and a subsequent ICP procedure. Concerning texture mapping, it is advisable
to blend colour from more images in a weighted average of corresponding values (Baumberg,
2002, Orzan & Hasenfratz, 2005). In this contribution, we extend our previously presented approach for automatic orthophotos generation (Karras et al, 2007) to produce photorealistic 3D
models from Lidar and image data.
Automatic orientation of aerial imagery
The first step of our approach is to automatically compute the relative orientation of all available aerial images. This is performed at different image scales in order to handle high resolution
imagery. We start at a low resolution where 2D SIFT (Lowe, 2004) and SURF (Bay et al, 2008)
features with descriptors are extracted on all images. Afterwards, sparse matching is performed
and outliers are identified through RANSAC on the fundamental matrix computation. Combin234
ISBN: 978-960-98739-8-7
ing the inlier matches across different stereopairs we obtain multi image point correspondences.
A bucketing algorithm is also applied to reduce the number of tie points without affecting their
distribution. Then, by means of closed-form algorithms and successive small bundle adjustment
solutions we estimate initial image orientations which are refined through a typical self-calibrating bundle adjustment solution (Fig. 1). This procedure is repeated at higher image resolutions but this time the matching is restricted by the estimated epipolar geometry and a rough
3D reconstruction from the tie points of the previous resolution. The final interior and exterior
orientation of all images is calculated again by a bundle adjustment solution.
Figure 1. Automatic Image Orientation Scheme using SIFT and SURF features
From the evaluation with our dataset, the standard error of the relative orientation using hierarchical SFM was about 0.5 pixel.
3D model to image registration
In order to apply texture mapping to the existing 3d city model, image orientations acquired
from previous step, need to be transformed to the reference projection system of the DSM. For
this purpose different approaches can be followed. Conventionally, this is performed by the use
of Ground Control Points measured in the field with typical surveying techniques such as GPS
which are manually identified on the available images. However, several approaches have been
proposed for the automation of this step. Here we apply a framework suggested by our team
(Grammatikopoulos et al., 2015) that uses an available orthophoto as a reference. Based on this,
a DSM is generated from the relative oriented image set using dense stereo algorithms (Stentoumis et al., 2015) and a new orthomosaic is computed (Karras et al., 2007). Then, Ground
Control Points are extracted by matching 2D features between the newly created and the reference orthophotos and 3D coordinates are assigned to them via interpolation on the reference
DSM. These Control Points are automatically identified on the aerial images and a final bundle
ISBN: 978-960-98739-8-7
235
adjustment solution allows the optimal registration among 3D model and aerial images.
Applying this approach to our dataset, we had a final RMS error for the GCPs of about 40 cm,
which is about two times the size of the reference orthomosaic pixel and half the 3D model
resolution.
Texture mapping
Once all available data (3d surface model and aerial images) have been registered together, an
accurately textured 3d city model can be acquired by employing texture mapping algorithms.
The simplest approach is to attribute RGB colour values to every polygon vertex of the 3D
model through back-projection on the aerial images. However, this leads to a less detailed representation, where the texture of each surface triangle is simply generated from the interpolation of the corresponding vertices’ colours. Hence, details smaller than the size of a triangle are
not visible.
Figure 2. Overall and detailed view of 3D City Model with Multi-image Texture Mapping
This drawback can be overcome by applying texture instead of colour to each triangle. A
straightforward approach is to compute for all vertices of the model u,v coordinates on the
original images. In this way, texture from a suitably chosen image (most nadir, closest distance
etc.) is applied to every model triangle. The main disadvantages of this approach is memory
inefficiency since all original images are required to display the model and colour discontinuities for neighbouring triangles textured from different images.
A more accurate approach is ensured by blending texture from multiple images. A new image
(texture atlas) is generated that stores texture for the whole model. In a first step u, v coordinates
on the texture atlas are estimated for each triangle via a surface parameterization algorithm and
the final texture is generated by combining texture from corresponding image regions using a
weighted blending scheme that can compensate for different orientations, scale and resolution
of the images involved (Fig. 2). Occlusion handling is also incorporated along with means for
236
ISBN: 978-960-98739-8-7
automatically eliminating colour outliers from individual images, especially near image occlusion borders.
Concluding Remarks
The work presented here is part of an ongoing research project on 3D city models that focuses
on multimodal representations of urban areas. A framework was proposed for the automatic orientation of aerial imagery with respect to an existing 3D model obtained from Lidar scanning.
First results have shown the effectiveness of the suggested approach.
Acknowledgements
national resources under the framework of the Archimedes III: Funding of Research Groups
in T.E.I. of Athens project of the Education & Lifelong Learning Operational Programme.
We would also like to thank GeoIntelligence S.A. for providing the DSM elevation data as well
as the National Cadastre & Mapping Agency of Greece for supplying us with the high resolution aerial photographs.
References
1. Armenakis, C., Gao, Y. and Sohn, G., 2012. Semi-Automatic Co-Registration of Photogrammetric and Lidar
Data Using Buildings. ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences
I-3, 13–18.
2. Baumberg A., 2002. Blending images for texturing 3D models. British Machine Vision Conference, pp. 404413.
3. Bay, H., Ess, A., Tuytelaars, T. and Van Gool, L., 2008. SURF: Speeded Up Robust Features. Computer Vision
and Image Understanding, Vol. 110, No. 3, pp. 346–359.
4. Delara, Jr. R., Mitishita, E. A. and Habib, A., 2004. Bundle Adjustment of Images From Non-Metric CCD
Camera Using Lidar Data As Control Points. International Archives of XXth ISPRS Congress, Commission
III, 13–19.
5. Gneeniss, A. S., Mills J. P. and Miller, P. E., 2014. In-Flight Photogrammetric Camera Calibration and
Validation via Complementary Lidar. ISPRS Journal of Photogrammetry and Remote Sensing 100, 3–13.
6. Grammatikopoulos L., Kalisperakis I., Petsa E., Stentoumis C., 2015. 3D city models completion by fusing
lidar and image data. Videometrics, Range Imaging, and Applications XIII .Proc. SPIE, vol. 9528.
7. Habib, A., Ghanma, M., Morgan, M. and Al-Ruzouq, R., 2005. Photogrammetric and Lidar data registration
using linear features. Photogramm. Eng. Remote Sensing 71 (6), 699–707.
8. Karras, G., Grammatikopoulos, L., Kalisperakis, I. and Petsa, E., 2007. Generation of orthoimages and
perspective views with automatic visibility checking and texture blending. Photogrammetric Engi¬neering
and Remote Sensing, 73(4), 403-411.
9. Lowe, D., 2004. Distinctive image features from scale-invariant keypoints. Int. J. Comput. Vision 60 (2),
91–110.
10. Orzan, A., Hasenfratz, J.-M., 2005. Omnidirectional texturing of human actors from multiple view video
sequences. IEEE Int. Workshop on Human Computer Interaction, Beijing.
11. Stentoumis, C., Grammatikopoulos, L., Kalisperakis, I. and Karras, G., 2014. On accurate dense stereomatching using a local adaptive multi-cost approach. ISPRS Journal of Photogrammetry and Remote Sensing,
Volume 91, May 2014, Pages 29-49.
12. Yang, B. and Chen, C., 2015. Automatic registration of UAV-borne sequent images and Lidar data. ISPRS J.
Photogramm. Remote Sens., 101, 262–274.
ISBN: 978-960-98739-8-7
237
Exploration of spatial variability of cycling in a British Urban
Environment - Evidence from North East England
Seraphim Alvanides1*, Godwin Yeboah2
Engineering & Environment, Northumbria University at Newcastle, NE1 8ST,
United Kingdom
*[email protected]
2
Centre for Transport Research, University of Aberdeen, AB24 3UH, United Kingdom
1
Keywords: cycling prevalence, spatial variability mapping, visual analytics, spatial
statistics.
Abstract
This paper contributes to the evidence base of spatial variability of cycling prevalence in British urban environment and at sub-regional levels. Very little is known about the visualisation
of spatial variability of cycling prevalence in available literature; particularly, when it comes
to our understanding of behaviours in the British geographical context. Providing evidence at
the local level is often challenging as spatial datasets at the local level for cycling is rare and
those that are available are usually unexplored and analysed a-spatially. Spatial analysis techniques comprising flow mapping kernel density estimation, dissimilarity indexing and standard
directional distribution are employed to examine spatial variability of cycling prevalence across
the study area. The outcomes of the kernel density together with the analyses of the Tyne and
Wear flows suggest that the denseness/concentration of cycling uptake is more visible around
the Newcastle city centre and parts of South and North Tyneside. A Dissimilarity Index of 11
indicates that the spatial distribution of travel to work by bike across North East England is not
evenly distributed. Compared to a previous (2001) Dissimilarity Index of 5 it is evident that dissimilarity is increasing across the region, which is likely to have cycling policy implications.
Introduction
The selection of the study area was informed by the findings from an exploratory analysis
which is not presented here but can be found in Yeboah et al. (2015). The choice of study area
was based on assumptions focusing on: the fact that cycling appears to be prevalent around the
central part of the Tyneside conurbation; practicality; and convenience. The Tyneside conurbation comprises four local authorities: Newcastle upon Tyne, Gateshead, North Tyneside and
South Tyneside. Newcastle upon Tyne, North Tyneside and South Tyneside were found to have
relative prevalence of cycling activity with the area around the central station of Newcastle
upon Tyne being more visible. This area constitutes the central part of the conurbation although
no exact boundary definitions exist. That notwithstanding, Newcastle upon Tyne has long been
acknowledged as the central place for the conurbation (Freeman and Snodgrass 1966, pp. 180205). Empirical study by Yeboah (2014) suggests that the central part of the conurbation would
be ideal to focus in terms of strategy for data collection. Since the intent is to use GPS technology as part of the data collection and it was not explicitly evident as to which routes potential
participants could cycle on, a selection criteria could provide the means to capture cycling
activity which reflects the pattern identified.
Furthermore, the central part of the Tyneside conurbation also had a high potential of traffic
congestion as reported in the Local Transport Plan 3 (LTP3 2011). Traffic congestion increases
the operation costs of employers in situations where travel demand increases (Aditjandra et al.
2013, p. 55). According to the area’s next decade strategy for local transport plan 3 (LTP3), the
238
ISBN: 978-960-98739-8-7
promotion of sustainable and safe communities has been considered as one of the challenges
to address. Given that the least prosperous areas in Tyne and Wear are reported to be the most
deprived in England, there are motivations to improve active travel modes such as cycling
while promoting healthy living lifestyles (LTP3 2011, p. ii). Rogers (2011) report, based on the
2010 Index of Multiple Deprivation data and summary measures constructed by Social Disadvantage Research Centre at Oxford University, ranks (average LSOA scores) districts in Tyne
and Wear as Newcastle upon Tyne (40), Gateshead (43), Sunderland (44), South Tyneside (52)
and North Tyneside (113) in increasing order respectively. Moreover, understanding cycling
behaviours can help transportation engineers to devise strategies to control traffic flow in and
around the area. Given the design of the research, the criteria for the sample for data collection
was around the central part of the Tyneside conurbation whereas the observed route choices of
cyclists were more flexible but generally constrained to the North East region of England. Taking this approach allowed the spatial characteristics of the primary sample to reflect the spatial
dynamics identified prior to the primary data collection. Figure 1 shows the study area depicting only home and work/school geographic locations of respondents with Google map as the
background; a fuzzy red boundary is also shown on the right map. The map on the left shows
the major transport networks with the major hub around Newcastle Central Rail Station.
Tyne and Wear
Study area with Google basemap
Figure 1. Maps of study area with major transport network
Five specialised variants of exploratory data analysis techniques used with the aim of assessing the distribution of activities across geographical space. Outputs from such computations
help in the triangulation and validation of results from exploratory analysis. The methods used
comprise:
ISBN: 978-960-98739-8-7
239
1. Flow mapping to visually present cycling flows across Tyne and Wear;
2. Kernel density distribution to assess the concentration/denseness of cycling activity across
Tyne and Wear;
3. Location quotient calculation to assess inequalities at local authority district level;
4. Lorenz curve graph together with Dissimilarity Index to assess cycling differences in the
North East of England;
5. Standard directional distribution to describe the spatial extent of cycling across the study
area using the primary data.
The exploration of the available secondary data on cycling was divided into three main phases.
The first phase explored data available only within the Tyne and Wear area. In this respect, the
TWHTS data, which is a cross-sectional survey, was used to identify more detailed travel patterns. Despite its usefulness, it lacks actual route preferences of respondents since it was a stated preference survey. In the case of using the origin-destination information from the TWHTS
data, Cycling Prevalence (CP) was defined as the degree of occurrence of cycling trips around
origin or destination locations (i.e. postcodes locations). In all other cases in the thesis, CP is
referred to as degree of cycling uptake. In addition, the 2012 National Travel Survey along with
data collected by the Newcastle Cycling Campaign was used for this phase. The second phase
explored cycling traffic flows from automatic cycle-counters which were obtained from the
Tyne and Wear Accident Data Unit (TADU). The third phase explored the available census data
for the whole of North East of England to examine the dissimilarities of travel to work by bike
across the region.
Further analysis was carried out to find trends between the nine years (2003 to 2011) of collected data. Given the increasing interest in cycling in the region, it is proper that the data is
explored to deepen our understanding of cycling with the hope of informing policy strategies.
Different graphs were generated based on some age-groupings and superimposed to show how
various age groupings made various kinds of trips in their method of travel categories. Cycling
uptake was relatively low among all age groups when compared to other modes of transport.
The uptake tends to increase gradually from adolescents till about 25 to 34 years and then it
decreases again after a decade. The uptake increases again from (35-44) years to (45-49) years,
but slightly less than the uptake volume during (25-34) years. There is continuous decrease
from the (45-49) year age group onwards. As expected, cycling uptake decreases with age due
to the energy requirement to perform a motile activity. Most of the travellers fall between 35 to
44 years old with combined trips over 10%. However, their cycle trip share is far below 0.1%
of all trips. This age group tends to use car/van and also walk more than cycle.
We now examine the spread of cycle commuting at the sub-district level by using the LSOA
statistics. The picture is completely different when dissimilarity assessment was undertaken at
the Lower layer Super Output Areas (LSOAs). This assessment was done using both the 2001
and 2011Census data. Using the extracted cycle commuting numbers out of the 1,656 LSOAs
covering North East of England, taking from 2001 Census, the computed Dissimilarity Index was 26% indicating a measure inequality and the relative distribution of cycle commuting
across the North East of England. Similarly, using the extracted cycle commuting numbers out
of the 1,657 LSOAs covering North East of England, taking from 2011 Census, the computed
Dissimilarity Index was 25% indicating a measure inequality and the relative distribution of
cycle commuting across the North East of England.
240
ISBN: 978-960-98739-8-7
Concluding Remarks
This paper examines extensively several secondary datasets related to cycling around the study
area to further deepen our understanding of cycling in a British urban environment; particularly
Tyneside conurbation in North East England. The paper analyses and discusses stated preferences of cyclists since it was impossible to objectively measure their “subjective” cycling
behaviour. The findings suggest that objectively measured distance and time expenditures are
significantly different from corresponding subjective measurements. This confirms the reported
case that subjective and objective measurements of the environment are not the same; therefore
a combination of the two strands is recommended when researching perceptions of the environment. The outcomes of the kernel density together with the flow mapping analyses of the Tyne
and Wear and the AAWT flow suggest that the denseness/concentration of cycling uptake is
more visible around the Newcastle city centre and parts of South and North Tyneside.
A Dissimilarity Index of 11 (from the 2011 Census data) indicates that the spatial distribution
of travel to work by bike across North East England is not evenly distributed. As a result, the
Lorenz curve reveals relatively higher concentrations of travel to work by bike in North Tyneside, Newcastle upon Tyne, and South Tyneside according to the location quotient values
derived from the Travel to Work data (2011 Census). Compared to the computed 2001 Dissimilarity Index of 5 it is evident that dissimilarity is increasing across the region. The collected
route choice preference information (from primary GPS tracks) suggests that many cyclists
cycle across the whole study area, regularly crossing local authority boundaries; therefore, the
different Councils need to work closer when planning cycling infrastructure and interventions.
References
1. Aditjandra, P. T., Mulley, C., & Nelson, J. D. (2013). The influence of neighbourhood design on travel
behaviour: Empirical evidence from North East England. Transport Policy, 26(0), 54-65, doi:http://dx.doi.
org/10.1016/j.tranpol.2012.05.011.
2. Andrienko, N., & Andrienko, G. (2005). Exploratory Analysis of Spatial and Temporal Data: A Systematic
Approach (2006 ed.).
3. Freeman, T. W., & Snodgrass, C. P. (1966). Tyneside. In The Conurbations of Great Britain (2 ed.): Manchester
University Press.
4. Indices of multiple deprivation: find the poorest places in England. (2011, Tuesday 29 March 2011 10.00
BST). TheGuardian.
5. LTP3 (2011). LTP3: The Third Local Transport Plan for Tyne and Wear Strategy 2011 - 2021. Tyne and Wear:
Tyne and Wear Integrated Transport Authority.
6. Yeboah, G. (2014). Understanding Urban Cycling Behaviours in Space and Time (Unpublished Doctoral
Thesis). Northumbria University at Newcastle, Newcastle upon Tyne, United Kingdom.
7. Yeboah G., Alvanides S., Thompson E. M., (2015). Everyday cycling in urban environments: Understanding
behaviours and constraints in space-time. In: Helbich M., Arsanjani J.J., Leitnez M. (ed) Computational
Approaches for Urban Environments, Geotechnologies and the Environment Series. ISBN 978-3-319-114682, pp. 185-210, Springer Press. New York.
ISBN: 978-960-98739-8-7
241
Stress and Moment Release Models for Seismic Hazard Assessment
in Corinth Gulf
Mangira Ourania1*, Tsaklidis Georgios2 and Papadimitriou Eleftheria3
Mathematics Department, Aristotle University of Thessaloniki,*[email protected]
Mathematics Department, Aristotle University of Thessaloniki, [email protected]
3
Geophysics Department, Aristotle University of Thessaloniki, [email protected]
1
2
Keywords: stress, moment, conditional intensity function, Corinth
Abstract
The aim of this study is the application of stress and moment release models in Corinth Gulf,
an area that accommodates high seismicity. Given that earthquake generation induces spatiotemporal stress changes on adjacent fault segments that can alter the occurrence probability
of subsequent earthquakes on these faults, Linked Stress and Moment Release Models are applied in two subregions of the Corinth Gulf, a division based on seismotectonic features. Point
process theory is applied by means of the conditional intensity function, and the conditional
intensity function has the form of the exponential distribution. Applications of both models indicate that interaction is actually observed between the two subregions. Specifically, earthquake
occurrence in the western part provokes damping in the eastern part. The reverse process also
takes place, yet it not that strong.
Introduction
Considerable efforts have been exerted during last decades evaluating the seismic hazard of a
region by means of the stress release models (e.g. Zheng & Vere-Jones, 1994; Lu et al., 1999;
Bebbington & Harte, 2003; Votsi et al., 2011). More recently, information gains and entropy
scores have been used for scoring probability forecasts (Bebbington, 2005; Harte & Vere-Jones,
2004), whereas Rotondi & Varini (2007) analyse stress release models from the Bayesian viewpoint.
In this paper, we model the earthquakes occurrence in the Gulf of Corinth as a stochastic point
process. Since we are interested in the interaction between adjacent subregions, we concentrate
on the Linked Stress Model. The region which was chosen to be studied is the Corinth Gulf
since it is one of the most prominent active structures in the Aegean region, which separates
continental Greece from the Peloponnese (McKenzie, 1978). In addition, we transform the
Linked Stress Release Model by considering that, instead of stress transfer, the interaction between the different subregions is measured via moment transfer. Moment release is used, since
it is calculated for every earthquake via its magnitude and can be easily transformed by means
of known relationships.
Data selection
The data used for the current study are taken from the catalog compiled from the Geophysics
Department of Aristotle University of Thessaloniki (http://geophysics.geo.auth.gr/ss/). Due to
the large number of parameters of the models used, a large data sample is required. We need
though to compromise the fact that we are interested in the strong earthquakes that occurred
in the Corinth Gulf. Thus, an effort is made to obtain the largest and longest possible complete earthquake catalog, which in our case is the one comprising earthquakes with magnitude
M ≥ 5.0 that occurred in the area since 1911.
242
ISBN: 978-960-98739-8-7
Application of the Linked Stress Release Model
In the Linked Stress Release Model, the key variable is the stress level in a region, which determines the probability of an earthquake to occur (Vere-Jones & Deng, 1988). The evolution of
stress X i (t ) in the i-th region with respect to time, could be written as (Eq. 1)
X i (t ) = X i (0) + ρi t − ∑ θij S (t , j ) ,
(1)
j
where S (t, j) stands for the accumulated stress release in the subregion j over the period (0,t)
and the coefficient θij measures the fixed proportion of stress drop, which is transferred from
subregion j to subregion i. It is plausible to set θii=1. Positive values of θij indicate damping,
whereas negative values indicate excitation.
The amount of stress released during an earthquake can be estimated from its magnitude (Kanamori & Anderson). The relation that has been found to hold for shallow earthquakes in the
Aegean area is given by Eq. 2
S = 100.75( M − M 0 )102.35 ,
(2)
where 𝑀 denotes the earthquake magnitude and 𝑀0 the normalized magnitude (𝑀0 = 5.0 in our
study) (Papazachos & Papazachos, 2000).
The stochastic behavior of the model is characterized by the conditional intensity function,
which is assumed to have the exponential form (Eq. 3)


 
λi* (t ) = Ψ ( X i (t )) = exp α i +ν i  ρi t − ∑ θij S (t , j )   ,
j

 

(3)
for each region i, where αi, νi , ρi and θij are the parameters to be fitted.
For the estimation of the parameters of the model, we use the method of maximum likelihood,
through a Newton-type algorithm. Fitting the LSRM to the dataset, the parameters 𝜃𝑖𝑗 concerning interactions between the subregions take the values θ12 =0.018 and θ21=1.590. This indicates
that earthquake occurrence in the western part of the Corinth Gulf provokes damping, i.e., decrease of occurrence rate in the eastern part. Moreover, possible slight damping of the western
part of the Corinth Gulf due to seismicity of eastern part is observed.
Application of the Simple Stress Release Model
Due to the fact that the coefficient θ12, concerning the stress transfer from the eastern to the
western part of the Corinth Gulf, is almost 0, we apply the SRM to each subregion separately.
In that case, the conditional intensity function takes the form (Eq. 4)
λi* (t ) = exp{ai + bi [t − ci Si (t )]} .
(4)
By applying the SRM to the western and eastern part of the Corinth Gulf, it emerges that the
parameters ai, bi and ci differ for each subregion. That means the two subregions can be studied
separately and not as a whole.
Application of the Linked Moment Release Model
Seismic moment release is acquired for further investigating interaction between the two subareas. Earthquake magnitudes were transformed to seismic moments based on well-known scaling laws (Hanks & Kanamori, 1979; Papazachos et al., 1997). The stochastic behavior of the
model is characterized, as in the LSRM, by the hazard function (Eq. 5)
ISBN: 978-960-98739-8-7
243


λi* (t ) = exp  ai + ν i





 mi t − ∑ θ ij M (t , j )  
j
 ,


(5)
where mi is the moment rate and M(t, j) is the moment release in the subregion j over the
period (0, t).
By fitting the Linked Moment Release Model to the same dataset, it emerges that θ12 = 0.189
and θ21=0.715. Again, the positive values of 𝜃𝑖𝑗 indicate damping. Particularly, the larger positive value of 𝜃21 indicates that -especially- large earthquakes in the western part of the Corinth
Gulf, will cause large damping in the eastern part and consequently seismic risk reduction.
Discussion and Concluding Remarks
The Linked Stress Release Model was at first applied to a complete catalog of earthquakes with
M ≥ 5.0 that occurred in the Corinth Gulf since 1911 in order to identify possible interactions
between the western and eastern part of the Gulf. The results indicate some kind of interaction
between the two subregions, and particularly damping. i.e., discouragement of earthquake occurrence in the eastern part of the study area due to earthquake occurrence in the western part
of the Gulf. The reverse process is not that strong. For further investigation of the interactions
we confirmed that the two subregions are distinct through the application of the Simple Stress
Release Model. We also applied a new model, called Linked Moment Release Model, which
demonstrates the same results concerning the interaction between the two subregions.
Acknowledgements
This work was supported by the THALES Program of the Ministry of Education of Greece
and the European Union in the framework of the project entitled «Integrated understanding of
Seismicity, using innovative Methodologies of Fracture mechanics along with Earthquake and
non-extensive statistical physics – Application to the geodynamic system of the Hellenic Arc.
SEISMO FEAR HELLARC».
References
1. Bebbington, M. and Harte, D., 2003, The Linked Stress Release Model for Spatio-Temporal Seismicity:
Formulations, Procedures and Applications. Geophys. J. Int, 154, 925-946.
2. Bebbington, M., 2005, Information gains for stress release models, Pure Applied Geophys., 162, 2229-2319.
3. Daley, D., Vere-Jones, D., 2003,An Introduction to the Theory of Point Processes, Vol. 1, second edition,
Springer, New York.
4. Hanks, T. C., Kanamori H., 1979, A Moment – Magnitude Scale, Geophys. Res., 84, 2348-2350.
5. Harte, D.S. and Vere-Jones, D. 2005, The Entropy Score and its Uses in Earthquake Forecasting, Pure Appl.
Geophys., 162, 1229–1253.
6. Kanamori, H., Anderson, D. L., 1975, Theoretical basis of some empirical relations in seismology, Bull.
Seismol. Soc. Am., 65, 1073–1095.
7. Lu, C., Harte, D. and Bebbington, M., 1999, A Linked Stress Release Model for Ηistorical Japanese
Earthquakes: Coupling among Major Seismic Regions,Earth Planets Space, 51, 907-916.
8. McKenzie, D., 1978, Active tectonics of the Alpine-Himalayan Belt: the Aegean Sea and surrounding regions.
Geophys. J.R. astr. SOC. 55,217-254.
9. Papazachos, B. C., Kiratzi, A.A., Karakostas, B., G., 1997, Towards a Homogeneous Moment- Magnitude
Determination for Earthquakes in Greece and the Surrounding Area, Bull. Seismol. Soc. Am., 87, 474-483.
10. Papazachos, B.C., Karakaisis, G.F., Papadimitriou, E.E., Papaioannou, Ch.A. , 1997, Time dependent
seismicity in the Alpine-Himalayan Belt, Tectonophysics, 271, 295-324.
11. Reid, H.F., 1910, The mechanism of the earthquake, in The California Earthquake of April 18,1906, Report of
the State Earthquake Investigation Commission, Carnegie Institute of Washington, Washington, DC, 16-28.
12. Rotondi, R., Varini, E., 2007, Bayesian inference of stress release models applied to some Italian seismogenic
zones, Geophys. J. Int., 169,301-314.
13. Vere-Jones, D., Deng, Y.L., 1988, A point process analysis of historical earthquakes from North China, Earthq.
Res. China, 2,165-181.
244
ISBN: 978-960-98739-8-7
14. Votsi, I., Tsaklidis, G., Papadimitriou, E., 2011, Seismic Hazard Assessment in Central Ionian Islands Area
Based on Stress Release Models, Acta Geophys., 59, 701-727.
15. Zheng, X., Vere-Jones, D., 1994, Further applications of stress release models to historical earthquake data,
Tectonophysics, 229, 101-121.
ISBN: 978-960-98739-8-7
245
A MATLAB routine for earthquake epicentre determination using
macroseismic data
Vasiliki Kouskouna1* and Georgios Sakkas1
Department of Geophysics, National and Kapodistrian University of Athens,
* [email protected]
1
Keywords: macroseismic epicentre, macroseismic intensity, direction of earthquake
shaking, polygon centroid.
Abstract
A MATLAB routine that calculates the macroseismic epicentre using the location of maximum
intensity and the distribution of directions of shaking is developed. For each site, the opposite
half-lines are calculated, resulting to a pattern of intersecting lines, which form an outer polygon. The centroid of the polygon is the calculated macroseismic epicentre. The method is tested
for two earthquakes, one instrumental and one historical with promising results, taking into
consideration the uncertainties introduced by macroseismic data.
Introduction
The determination of the location and size of earthquakes from the pre-instrumental period (historical) is a crucial problem for seismologists, especially for those dealing with the compilation
of catalogues. The availability of macroseismic intensities from instrumental earthquakes allows for the calibration of macroseismic parameters determination. In Greece, the instrumental
period officially starts in the year 1901, but this is probably shifted 2 or 3 decades later, as the
earthquake parameters of the early 20th century are mostly based on macroseismic data (e.g.
Makropoulos et al 2012, Shebalin 1974).
At regional level, problems of inconsistency are revealed, taking into account that regional
catalogues are based on national catalogues, each one using its own method for parameters
determination. In the framework of the European projects NERIES and SHARE, a catalogue
of European earthquakes before 1900 was produced (Stucchi et al 2012), in which the greatest
possible level of internal consistency in the determination of earthquake parameters was accomplished, i.e. the use of uniform procedures for determining earthquake parameters over the
whole of Europe.
These methods are based on the available macroseismic intensity data points (MDPs), i.e. on
the distribution of the various macroseismic intensity values attributed to the corresponding
localities. Obviously, most of the methods rely on the number of MDPs of each earthquake.
However, macroseismic information, that is the description of earthquake effects by contemporary local observers, does not always lead to an intensity value. If the earthquake was damaging
at a specific locality, intensity assessment is a straightforward procedure. When the earthquake
was simply felt, the information is often limited to the announcement of shaking. Intensity
scales provide several diagnostics for degree up to 5, the limit over which damage is observed.
For lower intensities the information is usually more general, including also observations useless for intensity estimation, such as “an earthquake was felt in the direction N-S” or “the shaking was coming from SE” etc.
The method
In an attempt to exploit such additional information, this study presents a routine designed to
use the direction of shaking from various localities for the determination of the epicentre, tak246
ISBN: 978-960-98739-8-7
ing into consideration the essential fact that the felt shaking (i.e. peak ground motion) is due to
the S waves, which oscillate on a plane perpendicular to the direction of propagation from the
source to the reporting site. For example, if an observer feels the shaking in the N-S direction,
the epicentre should lie in the direction E-W.
The routine inputs are: a. the coordinates of localities and the corresponding reported directions
in terms of azimuth angles, e.g. E-W corresponds to 90˚ and NW to 315˚, b. a trial epicentre,
which is the coordinates of the maximum intensity locality or the centroid of all such localities
and c. an earthquake catalogue of the region.
The routine calculates opposite half-lines with the maximum intensity coordinates as a starting
point. This procedure is carried out for all localities of the input file, resulting in a pattern of
intersecting lines. The points where lines intersect form the outer polygon. Points are sorted according to azimuth in order to create the polygon. The centroid of polygon is calculated. Polygon geometry is important for the validity of the centroid. For this reason a built-in MATLAB
indicator that checks the polygon geometry is provided. In case the polygon is self-intersecting,
the validity of the centroid is biased due to the polygon’s geometry. The calculated centroid is
the epicentre. In addition, the distance (in km) between the trial and the calculated epicentre is
presented to the user. Finally, maps of the calculated epicentre compared to the trial epicentre
and the regional seismicity (radius of 40 km around the calculated epicentre) are created, along
with data files of the epicentre and the points constituting the polygon. These steps are presented in the diagram of Fig. 1.
Figure 1: Routine calculation steps
Application
For testing the validity of the routine, two earthquakes in Greece are selected for processing.
The 1930 Keramidi event (Fig. 2), for which a comparison between instrumental and calculated
macroseismic epicentre is possible and the unpublished historical 1898 Argos event (Fig. 3).
1. 1930 February 23, Keramidi (I=8) Mw=6.0 earthquake
For this earthquake, 10 localities with information on the direction of shaking where identified.
ISBN: 978-960-98739-8-7
247
The trial epicentre is at Keramidi, where maximum intensity was observed and the calculated
polygon is self-intersecting, consisting of 31 vertices. The estimated epicentre is located at 31.6
km offshore Keramidi and 29.2 km from the instrumental epicentre (Makropoulos et al 2012).
Results are depicted in Fig. 2.
Figure 2: Left: Estimated epicentre of the 1930 earthquake, trial epicentre and created polygon.
Right: The estimated epicentre and instrumental events at 40 km radius.
2. 1898 August 5(O.S.)/17(N.S.) Argos earthquake
For this unpublished earthquake macroseismic information is available from the Annales de
l’ Observatoire National d’ Athenes (AOA, 1898). The earthquake caused some damage and
frightened the inhabitants of Argos, implying intensity IEMS=6-7. In this case, the coordinates of
Argos is the trial epicentre; 9 localities with information on the direction of shaking where identified. The calculated polygon is non-self-intersecting and consists of 19 vertices. The estimated
epicentre is located at 32.5 km ESE of Argos. Results are depicted in Fig. 3.
Figure 3: Estimated epicentre of the 1898 earthquake, trial epicentre and created polygon.
Right: The estimated epicentre and instrumental events at 40 km radius.
248
ISBN: 978-960-98739-8-7
Concluding Remarks
The MATLAB routine developed for the calculation of the earthquake epicentre using macroseismic data provided promising results in the two above-mentioned cases. The 1930 earthquake allowed for comparison between instrumental and macroseismic data. The method applied to the historical 1898 event implies that a considerable number of similar unpublished
events may be tested, along with the established methods of Boxer (Gasperini et al 2010),
MEEP (Musson and Jimenez 2008), Bakun and Wentworth (1997) and introduced to the earthquake catalogues.
However, care should be taken on the input data used: at larger distances the direction of feeling is influenced by a number of factors that are not easily identified. It is therefore suggested
that the method is tested for calibration with instrumental earthquakes in order to identify the
distance threshold, over which the direction data should be omitted.
References
1. Bakun, W. H., and Wentworth, C. M., 1997, Estimating Earthquake Location and Magnitude from Seismic
Intensity Data. Bulletin of the Seismological Society of America, 87(6), 1502-1521.
2. Gasperini, P., Vannucci, G., Tripone, D., and Boschi, E., 2010, The location and sizing of historical earthquakes
using the attenuation of macroseismic intensity with distance. Bulletin of the Seismological Society of
America, 100(5A), 2035-2066.
3. Makropoulos, K., Kaviris, G., and Kouskouna, V., 2012, An updated and extended earthquake catalogue for
Greece and adjacent areas since 1900. Nat. Hazards Earth Syst. Sci., 12, 1425-1430.
4. Musson. R. M. W., and Jimenez, M. J., 2008, Macroseismic estimation of earthquake parameters. Network of
Research Infrastructures for European Seismology, November 2008.
5. Shebalin, N. V., (ed), 1974, Catalogue of Earthquakes, Part I, 1901-1970. UNDP/UNESCO Survey of the
seismicity of the Balkan region”, Skopje.
6. Stucchi, M., Rovida, A., Gomez Capera, A. A., Alexandre, P., Camelbeeck, T., Demircioglu, M. B., Gasperini,
P., Kouskouna, V., Musson, R. M. W., Radulian, M., Sesetyan, K., Vilanova, S., Baumont, D., Bungum, H.,
Fäh, D., Lenhardt, W., Makropoulos, K., Martinez Solares, J. M., Scotti, O., Živčić, M., Albini, P., Batllo, J.,
Papaioannou, C., Tatevossian, R., Locati, M., Meletti, C., Viganò D., and Giardini, D., 2013, The SHARE
European Earthquake Catalogue (SHEEC) 1000–1899. Journal of Seismology, 17(2), 523-544.
ISBN: 978-960-98739-8-7
249
A homogeneous Mw equivalent earthquake catalog for Greece 19902015
Karakostas Vasileios1, Papadimitriou Eleftheria2 and Mesimeri Maria3
Geophysics Department, Aristotle University of Thessaloniki, GR 541 24 Thessaloniki, Greece
[email protected] 1, [email protected] 2, [email protected] 3
Keywords: magnitude scales, moment magnitude, seismicity, seismic network
Abstract
The earthquake magnitudes published in the monthly bulletins of the International Seismological Center (ISC) were used for the correction of the magnitudes estimated by Geophysics
Department of the Aristotle University of Thessaloniki (AUTH) and the Geodynamic Institute
of the National Observatory of Athens (NOA) for the period 1990 – 2015. Data were divided in
spatial subsets because it came out that the necessary corrections were not unique throughout
the study area. Then, knowing the dates of major changes in the estimation of magnitudes in
AUTH and NOA, and after thorough checking for any other notable change shown by the data,
they were grouped into six subsets temporally divided. First corrections were applied by comparison of the AUTH and NOA magnitudes with the body wave magnitudes estimated by ISC.
Using these corrections the magnitudes were reduced to the magnitudes estimated by AUTH
after 2011. Corrections were continued by correlation in pairs between AUTH magnitudes, and
ML and MD magnitudes calculated by NOA. Then all the data were transformed into AUTH
magnitudes as previously. All the available moment magnitudes estimated by AUTH and NOA
were correlated with the ones provided from the GCMT (Global Centroid Moment Solutions)
solutions (Ekstrom et al., 2012). This latter correction along with the earlier compiled catalog,
resulted to the final catalog with equivalent moment magnitudes was compiled.
Introduction
Magnitude estimation in different scales and in different Institutes is the initial source of differences in the estimated values. However, more significant differences are observed after changes
in the instrumentation and the software of the seismological networks. Transition of the analog
to digital era is among the main factors of errors in the magnitude estimation. Other sources
of errors are the accuracy of the amplitude measurements and the duration of the earthquake
signals. Stations in very close epicentral distances usually overestimate the magnitude of the
earthquakes. Magnitude estimations are also influenced by the azimuthal station distribution
since the attenuation of the seismic energy depends upon the faulting properties and directivity effects. The requirement of homogeneous catalogs in seismological studies necessitates the
magnitude correction and their unification to a common magnitude scale, which is broadly accepted to be the moment magnitude.
Data
Earthquake catalogs in Greece are continuously compiled during the instrumental era, as the
one compiled by Papazachos et al. (2000) and its continuation using the data from Hellenic Unified Seismic Network (HUSN), with magnitudes converted to equivalent moment magnitudes
based on relevant studies (Papazachos et al., 1997, Scordilis, 2006). Plot of the estimated magnitudes from both Institutes, reveals significant discrepancies at certain dates, which coincide
in time with changes in the magnitudes calculation. Figure 1 shows the
magnitudes of earthquakes that occurred during January 1990 – January 2015 in central Ionian
250
ISBN: 978-960-98739-8-7
Islands. Two abrupt changes in the magnitude distribution are observed for AUTH magnitudes
(Fig. 1 top) in February 2007 and August 2008. The most significant changes in the distribution
of NOA magnitudes (Fig. 1 bottom) are observed in August 2008 and February 2011, respectively. All these changes are connected with the transition period from analog to digital data
and the integration of all the seismological stations in Greece into the Hellenic Unified Seismic
Network (HUSN).
Figure 1. Earthquake magnitudes estimated in the Geophysics Department of the Aristotle University of Thessaloniki, M(THE), and the Geodynamic Institute of the National Observatory of Athens, M(ATH). The original
magnitudes were converted into equivalent moment magnitudes according to Papazachos et al. (1997).
Data Processing and Concluding Remarks
For correcting the magnitudes from AUTH and NOA, they were compared with the body wave
magnitude, mb, published in the online catalog of ISC (International Seismological Center
2013). Although this magnitude scale presents several disadvantages it accounts the longest
record and for this reason is selected in relevant investigations.
The two left columns of Figure 2 depict the correlation of the AUTH and mb magnitudes of the
earthquakes in central Ionian Islands in six different time intervals. It is observed that the differences between these two magnitude scales are not the same. The AUTH magnitudes of the
first five time intervals were transformed to the ones of the last period and were plotted on the
right part of Figure2.
All the available local (ML) and duration (MD) magnitudes estimated by NOA were transformed to AUTH magnitudes, as they are calculated for the last time interval (February 2011
- 2015), after correlation between them.
ISBN: 978-960-98739-8-7
251
Figure 2. Correlation between magnitudes estimated by AUTH (M(THE)) and the ISC body wave magnitudes (M(ISC)) for six different periods for the earthquakes occurred in central Ionian Islands. On
the left part the correlation of the initial data is shown. On the right part the AUTH data of the periods
January 1990 – January 2011 were transformed to those of the last period.
Figure 3. Magnitudes of the earthquakes in central Ionian Islands plotted against their occurrence time, after the
correction of different magnitude scales and their conversion to equivalent moment magnitudes.
252
ISBN: 978-960-98739-8-7
Moment magnitudes estimated by both AUTH and NOA are available for a number of events,
and found to be almost identical. After their conversion to GCMT moment magnitudes they
were correlated with the already corrected magnitudes. Using these results, the final catalog
was compiled and includes all the equivalent magnitudes are derived after the procedure applied here. All the original moment magnitudes were also added in this catalog. Figure 3 reveals
that until 2003 the earthquake detectability threshold in central Ionian is rather stable, varying
between 2.5 – 3.0. Later on, this threshold started to decrease gradually with the installation of
new stations. Since 2007, after the HUSN was set up, this threshold was decreased to 1.5 – 2.0.
After the occurrence of the M6.1 earthquake in Kefalonia (Jan. 26, 2014) and the installations
of additional new seismological stations in the area, the detectability threshold was further decreased up to M=1.0.
References
1. Ekström, G., Nettles M. and Dziewonski A. M., 2012, The global CMT project 2004-2010: Centroid-moment
tensors for 13,017 earthquakes. Physics of the Earth and Planetary Interiors, 200-201, 1-9, doi:10.1016/j.
pepi.2012.04.002.
2. International Seismological Centre, On-line Bulletin, http://www.isc.ac.uk, Internatl. Seis. Cent., Thatcham,
United Kingdom, 2013.
3. Papazachos, B. C., Kiratzi, A. A. and Karakostas B. G., 1997, Toward a homogeneous moment-magnitude
determination for earthquakes in Greece and surrounding area. Bulletin of Seismological Society of America,
87, 474–483.
4. Papazachos, B. C., Comninakis, P. E., Karakaisis, G. F., Karakostas, B. G., Papaioannou, Ch. A., Papazachos,
C. B., and Scordilis, E. M., 2000, A Catalogue of Earthquakes in Greece and Surrounding Area for the Period
550 BC–1999, http://geophysics.geo.auth.gr/ss.
5. Scordilis, E. M., 2006, Empirical global relations converting Ms and mb to moment magnitude. Journal of
Seismology, 10, 225–236.
ISBN: 978-960-98739-8-7
253
“Earthshield”: An implementation for automatic macro-seismic
calculations
P. Loumpardias1, K. Chimos2 , T. Karvounidis2, A. Ganas4, .C. Douligeris3
1
2
3
4
Computer Science Department (BSc), University of Piraeus,
[email protected]
Computer Science Department (Ph.D.), University of Piraeus,
[email protected], [email protected]
Computer Science Department (Prof.) University of Piraeus,
Institute of Geodynamics, National Observatory of Athens, Athens, Greece
Keywords: web based application, macro-seismic observations, automation.
Abstract
The application helps the researchers to select more information for an earthquake. It automates
the import of earthquakes for evaluation, the checking of the earthquake age for selection and
data processing regardless the language.
Introduction
Within the scope of the ASPIDA project we undertook the creation from the beginning of the
macro-seismic observations web page. The task has been implemented in two phases:
The first phase included the creation of a multilingual page of an existing questionnaire connected with the evaluation of seismic activity.
The second phase included the implementation (programming) of an application of automatically updating the list of earthquakes that are available for evaluation.
The application
The application consists of the following 3 main parts:
• The reader of the feeding which performs the insertion to the database.
• The page at which the earthquakes are being show in order the visitors to evaluate them.
• The questionnaire at which the visitors record their remarks and observations.
Initially a check of the RSS feed is performed for new earthquakes (figure 1).
254
ISBN: 978-960-98739-8-7
Figure 1. Checking for new earthquakes
Next to be done is redirection to the page of earthquake selection for evaluation (figure 2)
Figure 2. Selection of earthquake for evaluation
Next to earthquake selection is the visitor to be lead to the page of the questionnaire. The visitor
has the option to choose the language. This option is shown at the top left corner of the panel
(see figure 3) where the relevant flags are shown.
ISBN: 978-960-98739-8-7
255
Figure 3. Questionnaire
Upon completion of the questionnaire by a visitor the data are being sent to the database in
order to be evaluated by the researchers.
Concluding Remarks
The application helps the researchers to gather more information for an earthquake as the questionnaire can be filled from people speaking different languages. It automates the import of
earthquakes for evaluation leaving researchers more time for substantial scientific study of the
earthquakes and collected answers instead of data entry. Also by automatically checking the age
of the earthquakes that are available for evaluation and keeping only those that are more recent
it makes sure that the data submitted by the visitors are not outdated and inaccurate.
Future Work
Further improvement could be the upload of photos by the visitors that have embedded geodata
and provide more accurate information about the immediate effects of the earthquakes. Also
the application can be designed with a responsive interface so as to adapt to different screen
256
ISBN: 978-960-98739-8-7
sizes, making easier to answer the questionnaire from various devices resulting in even more
replies.
Acknowledgements
This study has been partially supported by the Greek Secretariat for Research and Technology and the European Regional Development Fund of the European Union under the research
project “Infrastructure Upgrade for Seismic Protection of the Country and Strengthen Service
Excellence through Action”, project MIS-448326, implemented under the Action “Development Proposals for Research Bodies-KRIPIS.”
References
1. RSS version 2.0 Specifications, http://blogs.law.harvard.edu/tech/rss.
2. Blekas, Alexander, John Garofalakis, and Vasilios Stefanis. “Use of RSS feeds for content adaptation in mobile
web browsing.” Proceedings of the 2006 international cross-disciplinary workshop on Web accessibility
(W4A): Building the mobile web: rediscovering accessibility?. ACM, 2006.
3. Gilmore, W. Jason. Beginning PHP and MySQL: from novice to professional. Apress, 2010.
ISBN: 978-960-98739-8-7
257
Small-world property and distinct evolution of complex networks
at periods of main shocks from historical earthquake records in
Greece
D. Chorozoglou1, D. Kugiumtzis2 and E. Papadimitriou1
1
2
Department of Geophysics, Aristotle University of Thessaloniki,
Department of Electrical and Computer Engineering, Aristotle University of Thessaloniki,
[email protected], [email protected], [email protected]
Keywords: Nodes, Successive earthquakes, Connections, Network measure, Main
shock.
Abstract
Recent works have shown that disparate systems can be described as complex networks. Earthquake networks are constructed from successive earthquakes occurred in the Greek area during
the period 1981–2014. The network nodes represent 2–D cells and the connections between
them are drawn whenever there is succession of earthquakes from then. The earthquake networks constructed from full historical record for two different cell sizes and two magnitude
thresholds are all found to have the small–world property with distinct network characteristics,
i.e. large clustering coefficient and small characteristic path length. By studying the complex
network at sliding windows throughout the historical record exhibited a distinct dynamic evolution at the time of main shocks (strong earthquakes). Especially, the clustering coefficient tends
to be stable with small variations before a main shock, it has an abrupt jump at the time of main
shock occurrence, and then slowly decreases and becomes stable again.
Introduction
Complex network analysis is an emerging field that was introduced recently also in seismology
by Abe and Suzuki (2004) in order to study seismicity as a spatiotemporal complex system.
Historically, random networks were first introduced by Erdős and Rényi (Erdős and Rényi,
1959). A random network
, where is the set of nodes and the set of connections,
is formed by a probability distribution, i.e. the connections are drawn with a predefined probability. A second network type is the small–world network (Watts and Strogatz, 1998), having
the characteristic that most nodes can be reached from every other by a small number of steps,
and thus the typical distance between two randomly chosen nodes grows proportionally to
(the number of nodes). A small–world network has a high clustering coefficient compared to the
random network of same
and
. The third main network type is the scale–free network
(Albert and Barabasi, 2002), where its name is due to the power–law distribution of the number
of connections in the network, meaning that some nodes play a central role because they have
many connections (hubs).
The aim of this study is the investigation of which network type the earthquake networks constructed for the area of Greece, are more likely to be. Here the earthquake networks are defined
to have nodes the cells containing earthquakes and the connections are given by the time succession of two earthquakes occurred in any cell. The network type is investigated for the full
historical earthquake record and two different cell sizes, as well as two different magnitude
thresholds. Then, after constructing networks in the same way but on the basis of sliding time
windows by one month time step, the network evolution is examined by using network measures, focusing in particular on periods of strong earthquakes.
258
ISBN: 978-960-98739-8-7
Data and Methodology
The observation data are obtained from the earthquake catalog compiled in the Geophysics
Department of the Aristotle University of Thessaloniki (http://geophysics.geo.auth.gr/ss/). The
data used in this study comprise crustal earthquakes (focal depth less than 40 Km) of magnitude
(93.158 events) and
(18.029 events) that occurred in the period 1981–2014
(Fig. 1).
Figure 1. Epicentral distribution of
that occurred in 1981–2014 in the broader area of Greece. The
diameter of the cycles is proportional to the earthquake magnitude and the color proportional to the focal depth
as given by a color scale.
For the construction of networks, firstly the study area is divided into cells, being the nodes of
the networks (if at least one earthquake over the considered magnitude threshold occurs inside
the cell). Two successive earthquakes define a connection between the respective nodes. If two
successive earthquakes occur inside the same cell, they form a loop. The
network connections of the
nodes can be directed or undirected, weighted or binary. Here, we consider the
simplest form of undirected binary connections without loops, such that a connection is drawn
whenever a pair of cells of successive earthquakes occurs, regardless of the number of times the
same succession occurs and independently of the order of the cells in the succession. The network is fully described by the adjacency symmetric matrix
of size
, having
entries
. It takes a values of one (1) if there is a connection between nodes and , and then
the two nodes are called neighbors, and zero (0) otherwise. The network properties are quantified with a number of characteristics (indices) computed on . The simplest and most known
characteristic is the node degree ,
,
which measures the connections at each node, and the average degree
(1)
is given by:
(2)
The average degree is the average number of connections in the network. Another well–known
characteristic is the clustering coefficient of the node, quantifying the tendency of any node to
group in triads and defined as:
ISBN: 978-960-98739-8-7
(3)
259
The average clustering coefficient is given by:
For a random network of Erdős–Rényi the average clustering coefficient is given by:
(4)
,
(5)
where
is the average degree of the random network.
For any two nodes and the distance
between them is defined as the length of the shortest path from to , if the nodes are connected, whereas
otherwise. The shortest
path length of a node defined as:
The average shortest path length or characteristic path length is given by:
(6)
(7)
The small–world property of a network is quantified by the so–called small–worldness measure
, defined as:
,
(8)
where
and
are the average clustering coefficient and average shortest path length
of the random network that has the same number of nodes and connections as the examined
network.
A value of much larger than one (1) suggests that the network is of the small–world type.
We use the above mentioned network measures in order to characterize the networks of binary
undirected connections formed from the historical record for the four settings (combining two
cell sizes and two magnitude thresholds). To study the evolution of the earthquake network
structure at the time of main shock, we form the earthquake network and compute the average
clustering coefficient on non–overlapping time windows of one month sliding across a time
period that spans from one year before to one year after a main shock. We consider two main
shocks, the earthquake of
near Zante in 1997 and the earthquake of
near
Crete in 2009.
Firstly we consider the four networks derived from the entire historical record. The degree distribution of any of the networks does not exhibit distinct power law behavior. Though the occurrence frequency is high for small degrees (number of connections of a node) and decreases
as the degree increases, the slope of the down trend is not constant if both frequency and degree
are seen in a logarithmic scale. Thus, the networks cannot be regarded as scale–free. To the
contrary, all four networks can be characterized as small–world networks, as tends to be much
larger than one (1) (Table 1). In the two settings of earthquakes with magnitude
(small
and large cell size) the evidence is stronger as is as large as double the value obtained for the
two settings of earthquakes with
.
260
ISBN: 978-960-98739-8-7
M
≥3
≥3
≥4
≥4
Cell size
44x44(km)
55x55(km)
44x44(km)
55x55(km)
Nodes
666
436
575
395
6,5
6,07
6,35
6
0,57
0,7
0,31
0,46
0,0098
0,014
0,011
0,015
1,96
1,81
2,27
2,11
3,68
3,56
3,63
3,52
109,73
100
45,61
51,1
Table 1. Estimated network measures for the four different settings (as indicated in the first two columns for the
earthquake magnitude and cell size).
Regarding the evolution of networks at the occurrence time of main shocks, the profiles of the
average clustering coefficient are shown in Fig. 2 for the occurrence time of the two main
shocks. For both magnitude thresholds,
and
, is quite stable with small fluctuations before the main shock, it rises abruptly at the time of its occurrence, and then decreases
and stays at the same level as before the main shock. This pattern is rather consistent with respect to the main shock occurrence time, the magnitude threshold, as well as the cell size (the
latter not shown). It is noteworthy that tends to be larger for
than for
. This
happens because the aftershocks associated with a main shock tend to occur close to the main
shock location, making the node of the main shock a hub, and thus there is a larger number of
than
aftershocks after a strong earthquake that are best accounted for in the setting of
.
0.8
M>3
M>4
Clustering coefficient
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
01/97 3/97 5/97 7/97 9/97 11/97 01/98 03/98 05/98 07/98 09/98 11/98
Time
0.4
Clustering coefficient
0.35
M>3
M>4
0.3
0.25
0.2
0.15
0.1
0.05
0
9/08 11/08 1/09 3/09 5/09 7/09 9/09 11/09 1/10 3/10 5/10 7/10
Time
Figure 2. The average clustering coefficient for
and
, as a function of time, computed on sliding time windows as denoted in the
. The upper panel is for the earthquake of
occurred on
November 18, 1997, and the lower one for the earthquake of
occurred on July 1, 2009, both indicated
with a vertical line. The cell size is 55x55(km).
Concluding Remarks
The earthquake networks formed from the succession of earthquakes over a long period in the
area of Greece are found to be small–world networks, regardless of the data setting, i.e. cells
size and magnitude thresholds. It was found that the average clustering coefficient can track
changes in the evolution of networks, computed on sliding time windows of one month step, asISBN: 978-960-98739-8-7
261
sociated with main shocks. This network index can be regarded as an index of the level of seismicity and may be a useful tool in the study of earthquake networks. These first results suggest
the usefulness of network analysis in the study of seismicity in a broader area, such as Greece
and its surroundings. More evidence needs to be accumulated from other network measures,
regarding the type of the network, as well as the network evolution at the occurrence time of
main shocks. For this scope, the analysis is planned to be extended to networks with weighted
as well as directed connections.
References
1. Abe, S. and Suzuki, N., 2004, Scale-free network of earthquakes, Europhys Letters, 65, 581-586.
2. Albert, R. and Barabasi, A. L., 2002, Statistical mechanics of complex networks, Rev. Mod.Physics., 74,
47–97.
3. Erdős, P. and Rényi, A., 1959, On random graphs, Publ. Math. (Debrecen), 6, 290–297.
4. Watts, D. J. and Strogatz, S. H., 1998, Collective dynamics of small-world networks, Nature, 393, 440–442.
262
ISBN: 978-960-98739-8-7
Design and Development of Embedded Systems for Precision
Agriculture Applications: The Case of Monitoring Dacus Oleae
Population and Bait Sprays
George N. Fouskitakis1*, Lefteris D. Doitsidis1, Hercules Rigakis1, Kyriaki
Varikou2, Ioannis Sarantopoulos1
Department of Electronic Engineering, Technological Educational Institute of Crete,
Chania 73100, Greece, * [email protected]
2
Institute for Olive Oil Tree & Subtropical Plants, Hellenic Agricultural Organization
“DIMITRA”, Chania 73100, Greece
1
Keywords: Embedded Systems, Olive Oil, Dacus Oleae, Bait Spray, Precision Agriculture.
Abstract
Olive fruit production is of crucial importance for Greece as it consists the third biggest oliveoil producer worldwide, thus offering more than 1.6 billion euros of incomes per year to the
Greek economy. Monitoring and controlling Bactrocera oleae (olive fruit fly) is of major importance for olive growers since it causes high damages to their olive fruit production. The
most frequently used method for protecting olive trees is that of bait sprays implemented from
the ground. To automate, optimize and assure the proper application of bait sprays, a prototype
integrated system was developed which is expected to offer a number of benefits such as: (a)
improvement of the produced olive-oil in terms of both quantity and quality, (b) environmental
protection, and (c) human protection. The system consists of a device for recording the bait
spray procedure, a gun sprayer and its nozzle providing precise pressure, flow rate and spraying
duration adjustment, a prototype “smart” McPhail trap offering wireless monitoring of the pest
population and a novel web-based platform for monitoring and control of both the pest population within the traps and of the bait sprays, thus providing aid to real-time decision making. The
system has been tested in the field and the presented results, though at a preliminary stage, are
encouraging.
Introduction
Monitoring and control of Bactrocera oleae (Gmelin) (Diptera: Tephritidae) is a major parameter for olive growers. Bactrocera oleae (Gmelin) or Dacus oleae (Diptera: Tephritidae) causes
high damages to olive fruit production of all countries around Mediterranean. Currently, several control measures against B. oleae are applied – including but not limited to – cover sprays
from the ground (Broumas 1994), bait sprays from air (Broumas 1995), steril insect technique
(Economopoulos et al. 1977), mass trapping (Varikou et al. 2004a; Varikou et al. 2004b).
The most frequently used method for protecting olive trees against it, is that of bait sprays implemented from the ground. It is characterized by a number of advantages, but its effectiveness
highly depends upon its proper application (Varikou et al. 2015). GIS techniques were used in
the past (Papafilippaki et al. 2007). Modern spectral techniques for recognizing the pest’s entrance in the trap have also been reported (Potamitis et al. 2014).
The aforementioned procedure is characterized by a number of disadvantages, mainly due to
human intervention. Τhe weekly count of insects in the field by human experts, instead of an
automated-aided pest population count, sometimes leads to delayed decisions for implementing
bait sprays. The current way of their applications is characterized by a number of drawbacks
such as: (a) incorrect sprayed quantity, (b) spraying of a rather larger - than required - part of the
tree canopy, (c) chemical burden on the ecosystem, and, (d) inadequate work/safety measures.
ISBN: 978-960-98739-8-7
263
There is a profound need for the optimization and certification of the bait spray process as well
as the minimization of human intervention, in order to achieve: (i) improvement of the health
and safety at work, (ii) environmental protection, (iii) effective protection from Dacus oleae,
(iv) improvement of olive-oil productivity. To meet the aforementioned requirements, a prototype-integrated system has been developed for the automation, optimization, and the assurance
of proper bait sprays application. The system consists of an embedded device for recording the
tractor’s route within the olive grove (geographical coordinates) and quantity of the sprayed
substance, a gun sprayer and its nozzle allowing for spraying the recommended chemical quantity, a prototype McPhail trap equipped with an embedded system providing wireless monitor of
the pest population and a web-based platform for remote monitoring and control of bait sprays,
thus allowing for real-time decision making.
The proposed System
The proposed project was focused in the design and development on an integrated automated system for the proper and controlled implementation of bait sprays. Initially, a network of
“smart” McPhail traps was developed, providing wireless (using the GSM network) transmition
of the related information. By using a camera in every smart trap, a series of photos of their
content were regularly taken and transferred to a remote server. The information was accessible
thru a web based system. Thus, olive fruit fly population was remotely monitored allowing the
expert scientists to plan and schedule bait sprays. These were performed by a modified tractor, equipped with a custom spraying system, able to completely control the spraying process
and simultaneously record the tractor’s path and each single spray (quantity). The procedure is
schematically presented in Fig. 1.
Figure 1. Schematic representaion of the proposed system.
The proposed system is under evaluation since last July (2015). For this purpose, a test olive
grove comprising of eight (8) sectors of about a thousand (1000) olive trees each, is used (Fig.
2(b)). The attractiveness of the “smart” automated trap is evaluated by simultaneously placing
264
ISBN: 978-960-98739-8-7
three McPhail traps in each sector: a plastic, a glass, and a “smart” one, with the first two kinds
already known to be certified, Fig. 2(a, b). The images taken from the “smart” McPhail traps are
found to provide adequate information for accurate pest population estimation, Fig 2(c).
Figure 2. (a) The “smart” McPhail trap; (b) the test olive grove; (c) McPhail content image.
Bait sprays are then performed using the developed spraying device. The spraying procedure is
continuously recorded/controlled via the embedded system in terms of both space (geographical position) and quantity. The recodred information is subsequntly depicted on a properly
modified digital map allowing for monitoring the tractor’s route coverage within each sector as
well as the details of each single spray, Fig. 3.
Concluding Remarks
In this study, embedded systems were developed for the use on a prototype-integrated system
which automates, optimizes and assures the proper application of bait sprays against Dacus
oleae. Its main goal was to minimize human intervention in order to achieve: (i) improvement
of the health and safety at work, (ii) environmental protection, (iii) effective protection against
Dacus oleae, (iv) improvement of olive-oil productivity. The results of the study, though at a
preliminary stage, indicated that the proposed system was fully operational, thus achieving its
main goals.
Acknowledgements
This work is implemented within the context of the Operational Program “Education and Lifelong Learning” action Archimedes III, Project 25, and is co-financed by the European Union
(European Social Fund) and Greek national funds (National Strategic Reference Framework
2007 - 2013). The authors would like to thank BASF for providing specific chemicals and Vodafone for providing GSM SIM cards.
ISBN: 978-960-98739-8-7
265
Figure 3. Detailed resperentation of bait sprays on a digital map.
References
1. Broumas, T., 1994, Olive fruit fly. Review of its biology and its chemical control. Agrotypos, 8, 26-31.
2. Broumas, T., 1995, Olive fruit fly. Βiological and Biotechnical control methods. Agrotypos, 2, 44-54.
3. Economopoulos, A., Avtzis, N. et al., 1977, Experiments on the control of the olive fly, Dacus oleae (Gmel.),
by the combined effect of insecticides and releases of gamma-ray sterilized insects. Z. Ang. Ent, 83, 201215.
4. Papafilippaki, A., Nikiforakis A., Stavroulakis, G., 2007. Optimization of protection against bactrocera oleae
by using geographical information systems, Proceedings of the 12th Greek Entomological Conference,
Cyprus, November.
5. Potamitis, I., Rigakis, I., Fysarakis, K., 2014, The Electronic McPhail Trap. Sensors, 14, 22285-22299.
6. Varikou, K., Alexandrakis, V., et al., 2004a. Study of the effectctiveness of mass trapping and new products
for the control of the olive fruit fly Bactrocera oleae (Gmelin) (Diptera: Tephritidae) in organic olive groves.
Proceedings of the Premier Seminaire International sur les Biotechnologies et Qualite des produits de l’Olivier
dans le basin Mediterraneen.
7. Varikou, K., Alexandrakis, V., et al., 2004b. New records for control of Olive fruit fly Bactrocera oleae Gmelin.
Proceedings of the 5th International Symposium on Olive Growing.
8. Varikou, K., Garantonakis, N., Birouraki Ath., 2015, Residual attractiveness of various bait spray solutions to
Bactrocera oleae. Crop Protection, 68, 60-66.
266
ISBN: 978-960-98739-8-7
Seismicity Memory Properties in Corinth Gulf (Greece)
Gkarlaouni Charikleia1*, Lasocki Stanislaw2 and Papadimitriou Eleftheria1
Department of Geophysics, School of Geology, Aristotle University of Thessaloniki,
* [email protected]
2
Polish Academy of Sciences, Institute of Geophysics
1
Keywords: stochastic analysis, Hurst exponent, seismicity, Greece
Abstract
The investigation of seismicity interrelations which are expressed by long memory dependence
and clustering is achieved with the use of stochastic means. Rescaled Range (R/S) analysis is
performed for recent seismicity in an extensional back arc basin of the Aegean region, the Corinth rift. The description of the collective properties of magnitudes (M), interevent times (IET)
and epicentral distances (IED) between consecutive events is examined in the view of the Hurst
exponent. The spatial distribution of Hurst value is used as an indicator of the active zones
exhibiting the highest degree of interaction. The analysis revealed a significant long memory
content in the seismic process, especially for the interevent time of recent micro seismicity indicating that weaker events may be primary carriers of stochasticity in the earthquake process.
Introduction
The stochastic investigation of seismicity provides an important tool towards understanding the
evolution and behavior of seismogenic fault populations and the assessment of seismic hazard.
One of the pronounced features of the earthquake occurrence, is the interaction among earthquakes and the clustering of events, expressed in time, space and size. In an attempt to seek for
a process where earthquakes depend on their preceding, or influence the time and the location
of the following events, long memory characteristics are investigated. The term “memory” or
“persistency” in a given stochastic process is used to express the elements dependency to one
or more elements, over a specific lag, evidencing the existence or lack of clustering and time
persistence. Rescaled Range (R/S) analysis for the estimation of Hurst coefficient, was firstly
implemented in seismology by Lomnitz (1994) who analyzed earthquake slip data. Moderate–
strong earthquakes persistency was later identified in the Mediterranean area by Xu and Burton
(2001) and in Maramaras sea, by Cisternas et al. (2004).
In the current study the analysis is targeted to a specific fault zone with well-known seismotectonic properties, where the associated seismicity may be considered more or less independent
from the one of the neighboring fault zones. The new concept in this approach comprises that
seismicity properties are sought in a smaller scale thus avoiding oversmoothing in the analysis
or identification of gross characteristics as large areas are examined. Intense seismicity is a
characteristic feature of the Corinth Gulf, which exhibits a constantly intense seismicity with
frequent strong earthquakes (documented in historical records) and also accompanied with seismic excitations of numerous aftershocks especially in the western part of the gulf. It is expected
that the new findings will be proven effective and will allow a further contribution to the future
seismic hazard assessment in the study area.
ISBN: 978-960-98739-8-7
267
Figure 1. The seismotectonic pattern of the Corinth Gulf.
Data and Methodology
The seismicity parameters to be investigated for their persistency or antipersistency under the
framework of R/S analysis are the time series of magnitude, (M), interevent time (IET, dt = t i +1 − t i
) which is the lapse time between two successive events and interevent spatial distance (IED,
dr = ri +1 − ri ), between epicenters of successive events. Earthquake information is obtained from
the Department of Geophysics, Aristotle University of Thessaloniki (http://geophysics.geo.
auth.gr/ss/). The completeness (MC) of the catalog for the study period 2008–2014 was defined
equal to 2.4 according to the proposed methodology of Wiemer and Wyss (2000). The catalog
was also de-clustered with the tool of Reasenberg (1985) for the removal of the most dependent
events (aftershocks) which bias the results. Long memory effects are measured by Hurst coefficient, H, ranging between 0 and 1 (Hurst, 1951) which quantifies the stochastic memory of a
process. According to Hurst values there is an indication for Brownian motion (H=0.5) or for
persistence or long-range dependence (0.5<H<1.0). Among the suggested methodologies for
the Hurst exponent, the most commonly used method is the R/S analysis. A set of observations,
X ∈ {X 1 , X 2 ,..., X n }, is subdivided into d non overlapping sub-series with sub-series length equal
to s = n d . The rescaled range ratio (R/S) is then qiven by Eq.1:
(R / S )s =
1
d
d
∑ (R
k =1
k ,s
/ S k ,s ) s
(1)
Where Rk , s is the largest measured difference between the maximum positive and maximum
negative cumulative deviation of X from its mean and S k ,s refers to the standard deviation of
the same value. Finally, R/S is connected with Hurst coefficient according to Eq.2, where Hobs
expresses the Hurst coefficient estimate (according to Weglarczyk and Lasocki, 2009):
log( R / S ) s = const + H obs log s (2)
Eq. 2 provides an estimate of the true H, which hereafter will be mentioned as Hobs. Hurst exponent is in all cases estimated directly from the observations with the application of the equations on the original data sets (Hobs). However, in order to determine the significance of Hobs
estimate deviations from the value 0.5, which indicates randomness, a null hypothesis stating
268
ISBN: 978-960-98739-8-7
that this population is long memory free is set and the H estimate is compared to 1000 estimates
obtained for random permutations of the original time series. Random permutations of a time
series destroy causality in the series although the values describing the effect remain the same.
If this hypothesis is true then Hobs will not differ significantly from H-estimates obtained from
actual memory-free samples.
Application
Results from the analysis of the data set are summarized in Table 1 and depicted in Fig. 2. For
the interevent times, Hobs takes the largest value among all cases (Hobs=0.806) with a probability
(p) much less than 1% for the process to exhibit no memory, meaning that it is highly probable
that consecutive interevent times tend to retain similar values. Distances between consecutive
events show also strong interrelation (Hobs=0.605), and the same for magnitudes (Hobs=0.774),
implying persistence characteristics.
IET
N
2106
IED
Real
Data
Random
Data
p
Real
Data
Hobs
0.806
<H>
0.533
≤
0.001
Hobs
0.605
Random
Data
<H>
0.531
M
p
Real
Data
0.031
Hobs
0.774
Random
Data
<H>
0.530
p
0.022
Table 1. Rescaled range analysis results for microseismicity of Corinth Gulf. Hobs denotes the estimated Hurst
exponents for the real catalog, whereas <H> is the mean value of H for N=1000 random permutations of the
original data series.
Figure 2. Rescale Range analysis results for microseismicity of Corinth Gulf. Upper row: Logarithmic values for
the rescaled ranges are plotted against the logarithmized number of the subseries. The regression line (red line)
for real observations, along with the estimated Hobs are presented. Lower row: frequency histograms for 1000 H
values calculated from memoryless samples. Red line is an approximation of the histogram by the normal distribution, red dot indicates the estimated Hobs, black dot indicates the mean H, <H>, from memoryless samples. The
horizontal dashed line indicates the upper limit of 99% confidence interval of H-estimate for memoryless sample.
The first picture column is related to IET, the second to IED, the third to M.
ISBN: 978-960-98739-8-7
269
Spatial distribution of Hurst values
R/S analysis was also performed in order to assess spatial variability of the Hurst exponent for
interevent time over the two study areas in a square grid, superimposed on the study area with
20 km spacing. At each node the data on earthquakes located within the 30–km radius are taken
into account and the estimation is performed for a minimum number of earthquakes inside the
circle equal to or larger than 80. H values range between 0.5 – 0.8 showing a spatial pattern
with high values along Psathopyrgos (Ps.F in Fig.2) and Aigion faults (Aig.F in Fig.2) (H~0.8)
where microseismicity is also highly clustered spatially (black dots) but also revealing areas
where earthquake occurrences are likely uncorrelated (H~0.5, greenish parts).
Figure 3. Spatial distribution of the Hurst exponent computed from the interevent times between events.
Concluding Remarks
Corinth gulf seismicity is clustered in all three considered process parameterizations. A significant long memory content was revealed in the seismicity process, especially regarding the
IET between consecutive events for recent microseismicity. It seems that seismicity along Psathopyrgos (Ps.F) and Aigion (Aig.F) faults, where H is about 0.8, tended to maintain constant
activity rate, either larger or smaller compared to the eastern part of the gulf and the transitions
between different levels of activity rate did not involve many events.
Acknowledgements
This work is co–financed by the European Union (European Social Fund – ESF) and Greek
national funds through the Operational Program “Education and Lifelong Learning” of the
National Strategic Reference Framework (NSRF) – Research Funding Program: Heracleitus
II. Investing in knowledge society through the European Social Fund. The work was also partially supported within statutory activities No 3841/E-41/S/2015 of the Ministry of Science and
Higher Education of Poland.
References
1. Cisternas A, Polat O, Rivera L (2004) The Marmara Sea region: Seismic behaviour in time and the likelihood
of another large earthquake near Istanbul (Turkey). J Seismology 8: 427–437.
2. Hurst H., 1952, The Nile (Constable, London).
3. Lomnitz C (1994) Fundamentals of earthquake prediction. John Willey & Sons, Inc., New York 326.
270
ISBN: 978-960-98739-8-7
4. Reasenberg, P. (1985) Second-order moment of central California seismicity 1969– 1982. J. Geophys. Res.,
90, 5479– 5495.
5. Węglarczyk, S., Lasocki, S. (2009) Studies of short and long memory in mining-induced seismic processes.
Acta Geophysica, 57, 696–715.
6. Xu Y, Burton PW (2001) Rescaled range analysis of the frequency of occurrence of moderate–strong
earthquakes in the Mediterranean area. In: Novak, M.M. (Ed.), Emergent Nature: Patterns, Growth and Scaling
in the Sciences. World Scientific, New Jersey, 305–314.
ISBN: 978-960-98739-8-7
271
The joint modeling of earthquake time series using bivariate
Poisson Hidden Markov Models
Katerina Orfanogiannaki1 and Dimitris Karlis2
1
2
Institute of Geodynamics, National Observatory of Athens, Greece, *[email protected]
Department of Statistics, Athens University of Economics and Business, Greece
Keywords: Hidden Markov Models, Bivariate Poisson Model, Copula, earthquake
counts.
Abstract
Discrete valued Hidden Markov Models (HMMs) are used to model time series of event counts
in several scientific fields. The model has two parts: the observed sequence of event counts and
an unobserved (hidden) sequence of states that consist a Markov chain. Each state is characterized by a specific distribution and the progress of the hidden process from state to state is
controlled by a transition probability matrix. While models for univariate discrete valued time
series are well known we extend the HMMs to the bivariate case by assuming appropriate bivariate discrete distributions for each state. We examine properties of the model and propose
inference. Maximum likelihood estimators of the models’ parameters are derived using an EM
algorithm.
On 26 December 2004 and 28 March 2005 occurred two of the largest earthquakes of the last
40 years between the Indo-Australian and the southeastern Eurasian plates with moment magnitudes Mw = 9.1 and Mw = 8.6, respectively. We model jointly the time series containing the
number of events occurring in the rupture zones of each mainshock in 23-day periods in order
to examine possible interactions between them.
Introduction
HMMs are well known models with many applications in different scientific fields including
seismology (see MacDonald and Zucchini, 1997, Ebel et al, 2007, and Orfanogiannaki et al,
2010). They allow for overdispersion (variance larger than the mean), autocorrelation and zero
inflation in the data. Characteristics that are often present in many real data. The seismic activity is not constant but it fluctuates between different time periods. These variations depend
on the state of the area the certain period of time. Naturally the state/phase in which an area is
at a given period depends on the one in the previous period introducing some kind of dependence between successive time intervals. In addition, the seismic activity in a particular region
may also be affected by the seismic activity of a neighbouring region. By extending univariate
HMMs to the bivariate case we aim to incorporate in the model interactions between different
regions.
HMMs are extended to the bivariate case by assuming different bivariate distributions associated with each state of the model starting from the well known bivariate Poisson model and
extending to Copula based distributions to allow for full flexibility. In our application the Frank
Copula has been selected since it allows for both positive and negative correlation. Of course,
any other copula can be used. We use the proposed models to jointly model earthquake frequencies in 23-day periods in 2 adjacent areas in the Sumatra rupture zone. The models‘ parameters
are estimated using an EM algorithm.
272
ISBN: 978-960-98739-8-7
The Proposed Models
We will formulate 2 different models. The Bivariate Poisson Hidden Markov model (BPHMM)
and a Hidden Markov model based on the Frank copula (HMMC). What differentiates the
models is the selection of the distribution family. In the BPHMMs each state corresponds to a
standard bivariate Poisson distribution while in the HMMCs each state corresponds to a bivariate distribution with Poisson marginals defined via Frank copula.
Bivariate Poisson Model. In the BPHMM the distribution associated with each state is a bivariate Poisson distribution. Assume that Xi, are independent Poisson distributions with parameters
λi, respectively where i = 0, 1, 2. The random variables Y1, Y2 defined as: Y1 = X1 + X0 and Y2 =
X2 + X0 follow the bivariate Poisson distribution with parameters (λ1, λ2, λ0). In BPHMM each
state is associated with a different bivariate Poisson distribution. Consider, that λj=( λ1j, λ2j, λ0j)
is the vector of parameters of the bivariate Poisson distribution that corresponds to state j and yi
= (y1i, y2i)′ for i = 1,...,n is the vector of observed data that corresponds to the i-th observation.
Cov(Y1i, Y2i) = λ0j is the covariance between the random variables Y1 and Y2 and since it is always
equal to a non negative number, only positive correlation is allowed. When the covariance is
equal to zero independence is implied.
Hidden Markov Models with Copula. Copulas are bivariate (multivariate) distributions with
uniform marginals. For a more formal definition see Nelsen, (2006). In the discrete case in
order to derive the joint probability mass function (pmf) we need to take differences i.e.: for
the bivariate case with marginals F(x) and G(y) the joint pmf is given by the formula: f(y1i,
y2i) =C(F(y1i),G(y2i))-C(F(y1i-1),G(y2i))-C(F(y1i),G(y2i-1))+C(F(y1i-1),G(y2i-1)) where F(∙) and G(∙)
are the marginal cdfs. This can be generalized to larger dimensions, but as dimensions increase excessive summation is needed. This creates a challenge on selecting copulas that can be
sufficient for the calculations. So, far we have worked with a bivariate Frank copula given by
equation 1.
1  (e −τu − 1)( e −τv − 1) 
C (u , v) = − log 1 +
 , τ
(e −τ − 1)


(1)
where τ is the copula parameter representing the dependence implied. Frank copula allows for
both negative and positive correlation. In addition the parameter of the Frank copula is unbounded and can take any real value. Finally, dependence in the Frank copula is symmetric in both
tails. Of course any other copula can be used. In our case we have selected Poisson marginals.
Parameter estimation
Due to the underlying structure of HMMs that allow for a missing data representation of the
model an EM algorithm (Dempster et al, 1997) is adopted for Maximum Likelihood estimation
of the parameters of interest both in the univariate and in the bivariate case. In the univariate
case (for details see MacDonald and Zucchini, 1997) and in the case of bivariate Poisson the parameters can be estimated by closed form equations. However, this is not the case for the copula
based models. In the later no close form equations of the model’s parameters are available and
numerical maximization techniques are adopted. A detailed description of the EM algorithm for
the bivariate case can be found in Orfanogiannaki et al (2013).
ISBN: 978-960-98739-8-7
273
Application
Data. Two large earthquakes occurred between the Indo-Australian and the southeastern Eurasian plates on 26 December 2004 and 28 March 2005 with moment magnitudes Mw = 9.1 and
Mw = 8.6, respectively. Both earthquakes were shallow and were followed by many aftershocks. The spatial distribution of aftershocks gives an approximation of the fault zone of each
earthquake. In the case of the 2004 earthquake the rapture started from the South and propagated further North. While in the case of the 2005 earthquake the rupture followed the opposite
direction. The fact that the two rupture zones do not overlap indicate that both earthquakes are
mainshocks with their one aftershocks each. The data sources are the USGS and ISC earthquake
data files for the region defined by the rectangle E with coordinates -1.00N - 15.00N and 91.00E
- 100.00E. This region was divided into two sub-regions, North (N) and South (S), corresponding to the rupture zones of the two mainshocks, respectively (Lay et al, 2005). Data completeness analysis based on the magnitude-frequency relationship showed that the data in all regions
are complete for mb≥4.2 for the time interval from 01.01.2000 to 25.12.2004. Following the
methodology introduced in Orfanogiannaki et al, (2013) the extended HMMs allow us to jointly
model the number of earthquakes in 23-day periods that occurred in the 2 sub-regions and estimate the correlation between them. The 23-day period was selected by Orfanogiannaki et al.
(2014), in order to divide the time interval between the two mainshocks in such a way that the
two mainshocks do not fall in the middle of any time interval but only at the edges.
Independent PHMMs
Bivariate Poisson
Frank copula Biv Pois
m
Loglik
BIC
Loglik
BIC
Loglik
BIC
1
-472.246
953.231
-468.253
951.695
-469.929
952.967
2
-396.276
818.770
-384.429
803.814
-395.098
825.151
3
-387.344
827.122
-367.496
800.534
-384.181
833.904
4
-376.632
840.653
-353.469
811.806
-373.650
852.167
Table 1. Comparison of the fitted models on the basis of BIC. Key: m is the number of states
Results. We applied PHMMs to the 2 univariate time series of earthquake counts for different number of states. We also modeled jointly the two time series considering both BPHMMs
and HMMC. The comparison between the different models was based on the values of the
Bayesian information criterion (BIC) defined as: BIC (m) = −2l (k ) + ln(n)d f , where l(k) is the
maximized log-likelihood for the model with m states, df is the number of free parameters of
that model and n is the size of the sample. The values of the maximized log-likelihood and of
BIC for the different models with different number of states are summarized in Table 1. The
model that best describes the data is the HMMC with 2 states. The parameter estimates for the
HMMC model with 2 states are shown in Table 2. Each column of the table corresponds to a
different state. The first 2 rows correspond to the seismicity rates for the 2 sub-regions while
the last row corresponds to the covariance of the bivariate Poisson distriubution. The estimated transition probability matrix determined for the HMMC model with 2 states is:
 0.000 0.401 0.599 


Γˆ =  0.082 0.415 0.503  . State 1 is a very active state for both sub-regions. Under the
 0.064 0.383 0.553 


274
ISBN: 978-960-98739-8-7
assumption of state 2 the North sub-region has about 4 times seismicity than the South subregion whiles under the assumption of state 3 the South sub-region becomes more active. When
the two regions are in state 2 they remain in that state with probability 0.415 or they transit to
state 3 with probability 0.503. So, from state 2 in which the North sub-region is more active it
is more probable to transit to state 3 in which the South sub-region becomes more active. The
higher value of the covariance parameter corresponds to state 3 in which both sub-regions have
relatively low seismicity levels.
North
South
Covariance
1st state
21.409
5.324
0.009
2nd state
6.019
1.515
0.011
3rd state
1.070
1.795
0.776
Table 2. Comparison of the fitted models on the basis of BIC. Key: m is the number of states
Conclusions
We have applied BPHMMs and HMMC to model together bivariate time series of earthquake
counts that correspond to 23-day periods. The data correspond to the rupture zones of the two
mainshocks that occurred in the region of Sumatra on 26 December 2004 and 28 March 2005.
We applied both models with different number of components and based on the BIC we have
selected the BPHMM with 3 states. The states correspond to different levels of seismicity and
show no monotonic behavior. The joint modeling of the two time series improves the results
obtained, providing evidence of interaction between the two sub-regions.
Acknowledgements
This research has been co-financed by the European Union (European Social Fund ESF) and
the National Strategic Reference Framework (NSRF) - Research Funding Program: Heracleitus
II. Investing in knowledge society through the European Social Fund.
References
1. Dempster, A.P., Laird, N.M. and Rubin, D.B, 1997, Maximum likelihood from incomplete data via the EM
algorithm (with discussion). Journal of the Royal Statistical Society B, 39, 1-38.
2. Ebel, J.E., Chambers, D.W., Kafka, A.L. and Baglivo, J. A., 2007, Non-Poissonian Earthquake Clustering and
the Hidden Markov Model as Bases for Earthquake Forecasting in California. Geophys. Res. Lett. 78, 57-65.
3. Lay, T., Kanamori, H., Ammon, C.J., Nettles, M., Ward, S.N., Aster, R.C., Beck, S.L., Bilek,S.L., Brudzinski,
M.R., Butler, R., DeShon, H.R., Ekstrom, G., Satake, K., Sipkin, S., 2005. The great Sumatra-Andaman
earthquake of 26 December 2004. Science, 308, 1127-1133.
4. MacDonald, I.L. and Zucchini, W., 1997, Hidden Markov models and other models for discrete-valued time
series (London, U.K.: Chapman and Hall).
5. Nelsen, R., 2006, An Introduction to copulas (Springer).
6. Orfanogiannaki, K., Karlis, D. and Papadopoulos, G.A., 2010, Identifying seismicity patterns using Poisson
Hidden Markov Models. Pure and Applied Geophysics 167, 919-931.
7. Orfanogiannaki, K., and Karlis, D., 2013. Hidden Markov models in modeling time series of earthquakes.
Proceedings of the 26th Panhellenic Statistics Conference, Piraeus, Greece, 8-11 May, pp. 312-321.
8. Orfanogiannaki, K., Karlis, D. and Papadopoulos, G.A., 2014, Identification of temporal patterns in the
seismicity of Sumatra using Poisson Hidden Markov Models. Research in Geophysics 4:4969.
ISBN: 978-960-98739-8-7
275
An overview of landslide detection and monitoring using geodetic
satellite observations
AC Peidou*, G Fotopoulos
Queen’s University, Department of Geological Sciences and Geological Engineering,
Kingston, Ontario, Canada
* Corresponding author: [email protected]
Keywords: landslides, submarine landslides geodetic satellite missions, remote
sensing.
Abstract
The purpose of this study is to provide an overview of geodetic satellite observations and their
application to both the detection and monitoring of landslides. From a geodetic perspective,
the detection of a landslide includes the determination of vertical and horizontal displacements
caused by the deformation event, while monitoring employs systematic observations of the rate
of movements over time. Landslides occur at various spatial and temporal scales. Geodetic satellite observations considered useful for geohazard monitoring can be categorized as follows,
a) Global Navigation Satellite Systems (GNSS) b) optical remote sensing images c) radar-based
remote sensing images d) altimetric observations, and e) low-Earth-orbit (LEO) gravity observations. The use of GNSS requires in-situ observations for the monitoring of landslides of all
scales and can provide sub-cm level accuracy. Digital Elevation Models (DEMs) generated by
optical and radar-based remote sensing satellites have been used for detecting and monitoring catchment and regional scale landslides via change detection algorithms on processed and
rectified optical and radar based images. The potential for altimetric and gravity-based satellite
observations for detecting large-scale solid mass transfers in submarine environments is an
emerging area of research.
Introduction
Landslides are triggered by physical, geological and geomorphological causes, including volcanic eruptions, weak geological layers, earthquakes, glacial loading, flooding, tectonic uplifts,
erosion and structural discontinuities (Meunier, 2008). Failure mechanisms vary based on the
geomorphology of the region, the trigger and the deformation features, which entail the spatial
and temporal scales and the volume of the mass transferred or displaced (Huang, 2015). The
contribution of geodetic observations in the detection and monitoring of landslides involves
understanding the physical, geological and geomorphological features, including mass properties and horizontal and vertical displacements. In this study, the spatial scale of landslides is
categorized as follows: point (1 m), local (100 m), catchment (10 km) and regional (1000 km);
(Glade and Cozier, 2005b). A landslide zone consists of several minor landslides of local and
catchment scale that occur in a specific region. Figure 1 displays the operable spatial and temporal scales of five categories of geodetic satellite observations. The rate of mass movement
indicating the temporal scale of the geohazard is classified in seconds, days, years and decades
(Glade and Cozier, 2005b). Terrestrial surveying techniques, such as geodetic total stations
and Light Detection and Ranging (LiDAR) systems offer precise measurements of subaerial
landslides but are typically limited in terms of spatial coverage to the local scale. In the case
of submarine landslides, in-situ techniques measuring solid mass changes are recorded either
by seismic or acoustic profiles derived from sonar surveys. In the following, a brief overview
of the most promising geodetic satellite observations for detection and monitoring of landslide
276
ISBN: 978-960-98739-8-7
events is provided. Of note is the inability of a single technique to provide sufficient information in all cases. Rather solutions involving a combination of various platform data (both terrestrial and airborne as well as satellite) and multiple sources should be considered.
Figure 1. Operable spatial and temporal scales of geodetic satellite observations identified as useful for detecting and monitoring landslide events
Geodetic Satellite Observations
GNSS monitoring adapts well to the detection and periodic or continuous monitoring of geophysical processes (Moss, 2000). This process requires in-situ observations and the realization
of a network, optimized in terms of station geometry such that the deformation features of interest can be encompassed. The two most common GNSS surveying techniques for this application
are static differential GPS (DGPS) and real-time kinematic (RTK). Both techniques use carrierphase observations and can provide sub-cm level accuracy in real-time (or post-processed).
The accurate positioning of the elevation changes of slopes depend also on the observational
rate (Hansen and Riggs, 2008). An optimal observational rate for landslide movements spans
1~25 Hz, while optimal baseline lengths between GNSS stations should not exceed ~10 km
(Teke et al., 2008). The number and the length of baselines depend on the geomorphology of
the region. The continuous or periodical mode of observations depends on the temporal scale
of the landslide.
Images obtained through optical remote sensing, span from 1m to ~250 m spatial resolution
(e.g., 1m provided by IKONOS and 250m provided by MODIS). Temporal resolution of most
of the optical remote sensing data is ~5 days on average and ranges between daily up to 35 days.
The three types of optical satellite images are:
Panchromatic imagery: provides high spatial resolution data with limited spectral information
resulting in ambiguities regarding the source of surface changes.
Multispectral imagery: allows for the discrimination of surface changes due to external sources,
such as vegetation, soil moisture and water flow.
Hyperspectral imagery: can be used for a better classification of the changes detected, improving the confidence level of the mass changes.
Factors such as morphologic heterogeneity, cloud cover and other sources of changes such as
soil moisture, often limit the capability of optical imagery to adequately detect movements independently, requiring integration with air-photos or ground-based surveys (Metternicht et al.,
2005). Spectral information of the solid mass change and geomorphological features of interest
can be used as initial screenings for mass movement. Change detection involves the comparison
ISBN: 978-960-98739-8-7
277
of co-registered successive optical images (assuming all corrections have been applied) with
the end result identifying stable areas and areas with ground surface change.
Interferometric synthetic aperture radar (InSAR) is considered the most effective radar-based
technique for providing 3D information of topography and deformation patterns of a region.
An InSAR image is created by co-registering two SAR images and calculating the difference
between their corresponding phase values on a pixel-by-pixel basis. The causes of the phase
change can be due to a) satellite orbit differences, b) geomorphology and vegetation, c) solid
mass deformation, d) atmospheric delay, and e) systematic and environmental noise (Lu et al.,
2007). The most limiting factor of InSAR data for this application is the inherent incidence
angle formed between the sensor and the target surface, which in this case is at a slope. Ideally,
in rough terrain, the incidence angle should be set to 45o (Eineder, 2003). Practically, the range
of incidence angles of current SAR missions varies between 15º and 60º. The spatial resolution of SAR missions ranges between 1m up to 100 m, for TerraSAR-X and RADARSAT 1/2,
respectively. This would allow for assessing regions of deformation at the local scale, however
depending on the terrain (incidence angle and slope), integration of InSAR observations with
GNSS observations is common (Komac et al., 2014).
Submarine landslides usually occur in inclined areas of the seafloor due to environmental stresses such as earthquakes, weak geologic materials, rapid accumulation of sedimentary deposits,
and groundwater seepage (Hampton et al., 1996). The volume and the spatial extent of masses
transferred or displaced typically exceeds the catchment scale (Hampton et al., 1996). Submarine landslides have an impact on the ocean floor altering bathymetry and can trigger tsunamis
(Tappin et al., 2014). Most of the reported submarine landslides occurred in prehistoric time
and were discovered by investigating sediments or rock mass displacements using acousticreflection profiles (Hampton et al., 1996). The potential of using satellite altimetry and satellite
gravity observations for detecting submarine landslides is intriguing. Altimetric measurements
provide sea surface height within a typical spatial resolution of ~30 km on a weekly basis
(Prigent et al., 2007). Sea surface heights and other observations such as the deflections of the
vertical are used for the determination of ocean floor topography (Smith and Sandwell, 1997).
By modelling the impact of the induced hazard, the initial submarine mass failure can be also
modelled (Fine et al., 2005). Key challenges with altimetric data for this application include the
variable correlation between geology (e.g. sediment thickness) and ocean topography as well as
noisy observations near coastal areas.
The Gravity Recovery and Climate Experiment (GRACE) satellite mission measured the temporal variations of the gravity field from 2002 to 2012 and offered unprecedented and invaluable information about mass changes caused by hydrological and solid mass transfers. The low
spatial resolution (~400 km) limits the operable scale for geohazards, while the flight altitude
(~500 km) further constrains the magnitude of the gravity change. Using simulations, the impact
of a specified solid earth mass transfer on the gravity field can be modelled. Initial case studies
indicate that a simulated landslide zone with a spatial extent of 250 km x 40 km (catchment
scale) would have to occur in order to be resolvable by GRACE observations given inherent
noise levels at the several μGals (Wahr et al., 2006). Although, catchment and regional scale
landslides have been recorded in the past, the only such landslide occurring within the GRACE
mission lifetime is the Tohoku landslide (Tappin et al., 2014). Future low-Earth-orbiting dedicated satellite gravity missions may possess more favourable characteristics (altitude and spatial resoltuion) for this type of application.
278
ISBN: 978-960-98739-8-7
Concluding Remarks
The utility of geodetic satellite missions for detecting and monitoring landslide events depends
on several factors, including the a) spatial resolution of the sensor, b) temporal resolution of the
sensor and c) observable parameters. Overall, terrestrially-based techniques that employ in-situ
measurements such as GNSS are optimal, although ground-based surveying is required. Strictly
‘remote’ observations from satellite-based platforms such as optical and radar observations also
benefit from in-situ measurements. While, it is yet to be proven, in theory altimetric and gravity
observations can serve as an additional source of information for such applications.
References
1. Eineder, M., 2003, Problems and solutions for InSAR digital elevation model generation of mountainous
terrain. In Proc. Fringe 2003 Workshop, Frascati, Italy.
2. Glade, T. and Crozier, M. J., 2005, A review of scale dependency in landslide hazard and risk analysis.
Landslide hazard and risk. Wiley, Chichester, 75-138.
3. Hansen, M. C. and Riggs, R. A., 2008, Accuracy, Precision, and Observation Rates of Global Positioning
System Telemetry Collars. The Journal of Wildlife Management, 72: 518–526. doi: 10.2193/2006-493.
4. Huang, R., 2015, Understanding the Mechanism of Large-Scale Landslides. In Engineering Geology for
Society and Territory-Volume 2 (pp. 13-32). Springer International Publishing.
5. Komac, M., Holley, R., Mahapatra, P., van der Marel, H. and Bavec, M., 2014, Coupling of GPS/GNSS and
radar interferometric data for a 3D surface displacement monitoring of landslides. Landslides, 12(2), 241257.
6. Lu, Z., Kwoun, O. and Rykhus, R., 2007, Interferometric synthetic aperture radar (InSAR): its past, present
and future. Photogrammetric engineering and remote sensing, 73(3), 217.
7. Metternicht, G., Hurni, L. and Gogu, R., 2005, Remote sensing of landslides: An analysis of the potential
contribution to geo-spatial systems for hazard assessment in mountainous environments. Remote sensing of
Environment, 98(2), 284-303.
8. Meunier, P., Hovius, N. and Haines, J. A., 2008, Topographic site effects and the location of earthquake
induced landslides. Earth and Planetary Science Letters, 275(3), 221-232.
9. Prigent, C., Papa, F., Aires, F., Rossow, W. B. and Matthews, E., 2007, Global inundation dynamics inferred
from multiple satellite observations, 1993–2000. Journal of Geophysical Research: Atmospheres (1984–2012),
112(D12).
10. Smith, W. H. and Sandwell, D. T., 1997, Global sea floor topography from satellite altimetry and ship depth
soundings. Science, 277(5334), 1956-1962.
11. Tappin, D. R., Grilli, S. T., Harris, J. C., Geller, R. J., Masterlark, T., Kirby, J. T. and Mai, P. M., 2014, Did a
submarine landslide contribute to the 2011 Tohoku tsunami?. Marine Geology, 357, 344-361.
12. Teke, K., Yalcinkaya, M. and Konak, H. (2008). Optimization of GPS networks for landslide areas. Fresenious
environmental bulletin, 17(6), 664.
13. Wahr, J., Swenson, S. and Velicogna, I., 2006, Accuracy of GRACE mass estimates. Geophysical Research
Letters, 33(6).
ISBN: 978-960-98739-8-7
279
Problems of Fracture Mechanics and Statistical Physics
Peter Sammonds and Clare Matthews
Institute for Risk and Disaster Reduction, University College London, Gower Street,
London, WC1E 6BT, UK, [email protected]
Keywords:Fracture Mechanics, Earthquakes, Volcano Seismicity
Abstract
A greater understanding of the physical processes occurring within a volcano is a key aspect
in the success of eruption forecasting. By considering the role of fracture growth, interaction and coalescence in the formation of dykes and conduits as well as the source mechanism
for observed seismicity we can create a more general, more applicable model for precursory
seismicity. The frequency of volcano-tectonic earthquakes, created by fracturing of volcanic
rock, often shows a short-term increase prior to eruption. Using fracture mechanics, the model
presented here aims to determine the conditions necessary for the acceleration in fracture events
which produces the observed pre-eruptive seismicity. By focusing on the cause of seismic
events rather than simply the acceleration patterns observed, the model also highlights the distinction between an accelerating seismic sequence ending with an eruption and a short-term
increase which returns to background levels with no activity occurring, an event also observed
in the field and an important capability if false alarms are to be avoided. However, there are
problems with the application of statistical physics approaches in fracture mechanics in the
Earth Sciences. Principally, while in tension, although the thermodynamics criterion for fracture is a necessary and sufficient condition, in shear it is only a necessary condition. But that
the local stresses at the tip of the fracture exceed the rock tensile strength, is always a sufficient
condition. The implication is that approaches in statistical physics need to consider fracture
mechanics at the micro-scale for application to large-scale processes.
Volcano Seismicity
Seismology is a key tool in eruption forecasting. Volcanic eruptions are very often preceded
and accompanied by an increase in local seismic activity, driven by fracturing within the edifice which must occur, at least in the case of long repose systems, in order for a pathway for
the magma to be formed. Models which calculate the time to failure, or eruption by analysing
the acceleration of event rate with time can be fit to observed data of seismicity prior to extrusion of magma (eg. Pinatubo 1991, Soufriere Hills 1995, Mount St Helens 1982). Multi-scale
fracture mechanics results have been used to explain internal processes producing the observed
patterns.
The literature contains numerous studies into both the temporal and spatial distribution of tectonic and volcanic seismicity. Significant shifts in parameters relating to seismic b-value, fractal
dimension, percolation theory and non-extensive statistical physics have been proposed as indication of a change in some underlying process controlling the development of activity within
a volcanic system.
The variation in spatial distribution of tectonic earthquakes was studied, by estimating the fractal dimension of locations of earthquake epicentres in Japan over a period of approximately 40
years. The spatial and temporal fractal dimension for seismicity during periods before (1983 1988) and after (1993 – 1996) a major intrusive episode at Mt Etna have been calculated. They
280
ISBN: 978-960-98739-8-7
found a rise in temporal dimension, but a fall in spatial dimension between the two periods,
implying that seismic activity became less clustered and more randomly or uniformly spaced in
time, but that the areas of activity became more concentrated. Similarly, changes in the Tsallis
parameters, have been calculated for Santorini. The change in these parameters are seen as an
indication of altering stress conditions within the volcano.
In out model, effects of the surrounding stress field on a volcanic system are described by a simple, one-dimensional configuration representing a potential pathway for magma from an underground storage to the surface. The initial, intact structure is made up of a number of individual
elements, each of which must fail before a fully connected conduit is created. Each element has
a particular fracture strength which must be exceeded for failure to occur. The strengths are allocated using a random walk process on a Gaussian distribution. Elements can exist in one of
two states: failed or intact. In the absence of any healing process, once an element has failed, it
remains failed. The failure of elements corresponds to seismic events.
Initially, the model is arranged to run to failure. Whenever the system reaches equilibrium, the
remote stress is incrementally increased until further fracturing is induced. Despite the random
nature of the initial conditions, when run under uniform rules for stress evolution, a similar sequence of increasing event rate is always seen (Fig 1a). The model was run a number of times,
in each case being allowed to run to failure. The stacked data of event rate leading up to collapse
of the system was compared, after normalising, to the increase of seismic activity leading up the
dome extrusion on Pinatubo on 7 July 1991 (Fig. 1b).
0.14
350
0.12
normalised 4hr event rate
400
event rate
300
250
200
150
100
50
0
80
70
60
(a)
50
40
30
20
10
0
▲
―
model data
■
―
data from Pinatubo 1991
0.10
0.08
0.06
0.04
0.02
0.00
time before failure
1
3
5
7
9
1
14 13 12 11 10 10
(b)
93
85
77
69
61
53
45
37
29
21
13
5
time before failure/onset of magmatic activity (hrs)
Fig. 1 Models with 6000 elements, run under gradually increasing remote stress. (a) hourly event rate for single
run of model. (b) stacked, 4 hourly results from 10 different initial distributions, compared with 4 hour event rate
leading up to the onset of dome extrusion at Pinatubo 1991.
The implications are discussed in terms of approaches in statistical physics.
Acknowledgements
The work of PS was supported by the THALES Program of the Ministry of Education of Greece
and the European Union in the framework of the project “SEISMO FEAR HELLARC’’, (MIS
380208).
ISBN: 978-960-98739-8-7
281
282
ISBN: 978-960-98739-8-7
Volume 1 / Topic C: Arts & Humanities
Topic C
Arts & Humanities
ISBN: 978-960-98739-8-7
283
Paradigm Shift in Innovative Design: Use of Form and Form of Use.
Dr Athanasios Kouzelis, Professor
TEI of Athens, [email protected]
Keywords: Paradigm shift, modern design history, innovation, tacit knowledge,
praxeology.
Abstract
The objective of this essay denies historical formalization that it scarcely can be distinguished
from the descriptive proliferation. It bases on Michael Polanyi’s philosophy of tacit knowledge’s importance, which points out that practical wisdom is more truly embodied in action
than expressed in the rules of action. Hence, a contradiction between tradition and transcendence seems to be fundamental to design praxis, and therefore a new perspective has to be laid
on it, via researching the experienced by experience. If design historians incorporated in their
works Thomas Kuhn’s approach about the revolutionary terms that have changed the progress
of the science of physics in general, so, any prominent form of design activity could be possible
to be brought to a practical, dialectic, plane, with meanings, more closely relating to experiences of giving form and structure to useful things, than to external causes of morphological
signification as a matter of historical justification.
A utilization of similarities and differences between all prominent cases of modern designers and their work in action force us to reformulate the questions that were aroused and the
answers that were finally given by the historical subjects of design concerned with how their
concept ability and materialization possibilities became related to the unique situation : to shift
from traditional design praxis by transcending the ways of thinking and acting in order to come
upon a form’s originality, both as product of use and prominent object of human progress and
culture in general.
Introduction
Our inquiries concerning the emergence of innovations are as much past as the history of our
material production itself. Since the era human beings draw a practical usability from the forms
of physical and material objects, it became also apparent that their act was a pioneering innovative step towards to our modern values of welfare. Innovations, conceived as praxeological
facts, constitute an upset in any form of prevailing usability, because they are bearing a new
aspect of it, which entails a higher value in compare to its former estate. Therefore every innovation engages our creativity in order to achieve placement among existing and pursued terms
of what we use to call objectified material wealth.
Encountering the importance of a paradigm for the historical prominence of a designer’s work,
the history of design gain a status of actualization within the frame of practice. By paradigm
design history is focused on the solving ability that a designer or design practitioner (namely
engineer, physician, architect etc) had shown by acquiring theory, methods and standards
together in an inextricable knowledge novelty. Therefore, when paradigms have changed in
design, there were significant shifts in the criteria that determine the historical legitimacy both
of problems and proposed solutions.
As Thomas Kuhn has remarked we may assume that the aspiratory process is much more significant as historical than a finished product is as a designer’s work. This process advances “by the
proponents of a new paradigm, that they can solve the problems that have led the old one to a
crisis”. The decision between alternate ways of practicing science, art and technology is called
284
ISBN: 978-960-98739-8-7
for, and the circumstances that decision must be based less on past achievement than on future
promise. The designer who has embraced a new paradigm has faith that the new paradigm will
succeed with the many large problems that confront it, knowing only that the older paradigm
has failed with a few. That was what Peter Behrens conceived entering his design work in AEG,
destructing all figurative notions on electric apparatus for the sake of electric efficiency fulfillment. As a new outstanding candidate he questioned the problems that electric industry had
previously not solved pointing the way that they could be met in other terms.
Paradigm components
The components of a paradigm are according to Thomas Kuhn’s theory ‘the symbolic generalizations’, ‘the metaphysical notions’, ‘the shared values’ and ‘the exemplars’. Projecting this
epistemology in design history we may detect a hidden, but prominent historiography that deploys the historical subject’s roll in design development in human civilization’s history. These
readily shapeable components are specifying the discipline of design history as a totality of
human thinking and acting in practice, as well as a consistent science without glorifying objectivity and even rationality.
The essential characteristics of symbolic generalizations are beliefs, values and notions shared
by the designers of a given historic phase. These issues are connected with practical and aesthetical hypotheses as Horatio Greenough’ s ‘Form follows function’, as well as Raymond
Loewy’s formula ‘most advanced, yet acceptable’(MAYA). On the other, they denote the model
of thinking as a basis for the explicit rules of the designing work. It is a result of the undergone
education and professional initiation, a process that sets every designer in a uniquely pursuit of
a set of shared concerns about the object one is acting upon. This common awareness is closely
related to what Thomas Kuhn has defined as a ‘disciplinary matrix’, disciplinary because it
refers to the common possession of the designers of a particular notion or formula. Shared commitments which cannot challenge a novelty or improvisation are offers to abandon themselves.
Until a designer can be freed from previous symbolic generalizations, all design work is a historic tautology to be eliminated.
Having next isolated the design techniques and methodologies as ‘paradigm exemplars’ the
historian may appropriate the kernel of all design development: the concrete problem-solution
act that makes a designer prominent and its work important for historical assessment. This point
shows that aesthetical values or morphological notions are to be applied in that ‘risk business’
which must be undertaken by a designer in concern to technical and practical means. The paradigm exemplar has always been responding to a design technique anomaly or a methodological
discrepancy.
The problem-solution act is laid at the bottom of a designer’s conceptual competence, which
is a complex consisting by visual, cognitive logical and monitor abilities. To master a design
concept means to be able to do things, because conceptual competence is internally related to
action competence. It is the way of dealing with any potential design problem that makes a
work of design as distinct from all other in design history.
Paradigm shift in modern design praxis
David Pye focused a paradigm shift in modern design praxis by the meaning of use or usability of every produced thing, esteeming that any object gets its own value of use solely by the
outcome of our interaction with it as a preoccupational operation. It takes place rather as a matter of verifying the existence of use, as well as a contradiction to the absence of the object one
may act upon. According to his point of view, David Pye has noted that the ‘raison d’etre’ of
any use is not the shape of the object one acts upon, or its physical attributes, but the means that
these properties become exploited within the human activities. So, because of this reason the
ISBN: 978-960-98739-8-7
285
form of a usable product can not follow at any time its function, as emphatically predicted the
functionalistic design theory, but in practice what it refers is essentially the upset purpose of its
use (as shown in practice a cargolift and a lever carry out the same function, but they possess a
different form due to their specific pursue).
The rationalisation of the human activity categories which were undertaken by the functionalist
designers during the era of industrial expansion at the fields of everyday life (1935-1955) followed up a historic necessity to determine in accordance to a ‘minimum existenz’ the shapes of
objects of use at a highest possible geometric precision. Every research and empirical deduction
was focused on the level of an action’s practice and a series of experimentations with people
acting under economised and prescribed spatial circumstances were paraded for the sake of the
functionalist designers. Within the strict frame of the time course these experimentations led in
turn to morphological and volumetric considerations that finally caused a minimizing of needed
organic power and an energy saving at many cases of human activity.
After the end of the functionalist era all with usability associated researches followed theoretic
models that simplified the reality of every day life. Many researchers presumed technical and
physical terms which cause a diverse occurrence for the human activity. They depicted its qualitative attributes as a matter of programming and monitoring of objects of use in order to attain
an estate of effectiveness on them. In this scope the contribution of the technology of minimal
energy consumption after the decade of 1980 became a revelation for the organisation of usability. According to the aforementioned technology the form of use, seen as a physical object,
influenced any means of regulating the human activity, so that the needed energy consumption
could be diminished in accordance to the demanded effective outcome. In many cases a kinetic
aesthetic approach of the usability emerged as a regulative factor so that a more saved energy in
a lesser mass of material consumption should replace any functionalistic product of use.
From another point of view the use of forms in modern design practice has been also connected
to the emergence of creative notions, as a result of free hand selection and construction of
geometric and volumetric compositions which can depict solutions of usability with particular
aesthetic value. The use of shapes as design practice does not constitute a simple trial to beautify or better an invented product of use, as for instance suggested Raymond Loewy to its colleagues, because it is an outcome of the capitalistic need to shine and decorate the commercial
appearance of any commodity’s form, but rather as David Pie has noted it composes a mean to
conceive a product’s usability even as an aesthetic value.
Innovative design paradigm components
According to Per Mollerup’s design notion an innovation which concerns a new product of use
may derives from compositive transformations and considerations, expressing not only practical but even aesthetic attributes as a harmonic or symmetric whole may be. In such compositive
qualitative relationships one can detect new and unique sensations of form which may be juxtaposed to own aesthetic knowledge and not necessarily to own practical preference. Thus, the
form of every new product of use, seen as an innovation due to its significant aesthetic value,
can be esteemed in practice under the criterion of its morphological utility.
Therefore a reference of a significant product of use to a morphological symbolism is not a
consequence of its satisfactory or utilitarian attributes, but it rather derives from a unique composition of its shape as a bearer of aesthetic quality. That’s why the morphological symbolism
have been perceived as a culminant act of a form’s exploitation, because only then the form has
reached its fullmaturity, as a fact of adversity, which means that the form can not develop itself
more, retaining solely the given aesthetic value.
286
ISBN: 978-960-98739-8-7
Selection and use of pleasant lines, surfaces and volumes, as well as colour accents and any
other figurative contribution to elegancy have favoured innovations under the scope of developing higher aesthetic valueby shaping a product of use. It is also remarkable a distinction of
these elements of elegancy in comparison to respective attractive because of their geometric
and symbolic characteristics which are decoded from peculiarity on one hand and hedonism on
the other hand, perceived asconstants of aesthetic value.
The endeavours against symbolism freed the perspective of making innovations by aesthetical considerations as the prominent modern Danish designer Piet Hein once tested trying
to search significance and remarkableness in the forms of products of use. By exploiting as
means of transformation algebraic equations’ diagrams, he adapted their curves into the edges
of regular rectangular parallelograms, so that peculiar shapes could emerge as hyper-ellipses
of the parallelogram. His mature work upon these transformations became remarkable due
to the sense of modern manner of plasticity which determined all of his contributions to the
product design during the decade of the 1960’s (kitchen-wares, furniture and table-wares).
The embedded potentiality of the aforementioned design methodology incited the modern
Swedish designer Olavi Linden, heading the Fiskar’s company design office, to improve the
firm’s produced scissors’ level of efficiency and signification. His innovative concept followed
the figurative conception of applying the shape of a cogwheel in the joint point of the reverse
cutters, as well as sharpen their edges, in order to attain a more sophisticated and elegant form,
which finally served to develop longseries of successful and effective hand-tools for cutting
many kinds of hard or semi-hard materials.
Following a similar methodological course his colleague Ulf Hanses created also an innovative device by revolving a cone top, cut off by an oblique plane, around the perpendicular axis
of the circular plane of the section, in order to attain an easier and less movable form of a tea
pot. By his transformation of the regular cone he facilitated the more convenient flowing out of
any pot’s liquid content, without necessitating its user to lift it over the table for this purpose.
The use of invented forms has been also connected to demands about transformation of a
shape’s characteristics, so that they create what we perceive as styling in the modern design
practice. Styling embraces a range of figurative considerations which deduct a form of a product of use as a post-product of its usability in adjustment to a satisfactory restitution to the user,
as an aesthetic and cultural value. The criteria of this restitution determine also the innovative
content of a form, by initiating a comparison to a sufficient issue of it, for example in the plane
of commerce, as Raymond Loewy suggested once.
By interpreting, according to the German design philosopher Gert Selle’s theoretic approach,
the utility form in the design practice, as a matter of an aesthetic reflection to necessitated or
conventional material needs, it becomes remarkable to note that any innovative use of form is
connected with figurative and geometric solutions. These solutions enable a worth giving to the
produced objects of use in such a way that they constitute a practical capacity by acting upon
them, with or without the help of a machine driven power. Within this frame the interpretation
of usability follows the dialectic relationship between a use of form and its responding form of
use, as a redirection paradigm shift on the search of higher and more advanced aesthetic and
practical values, especially for the sake of our material and cultural welfare.
References
1.
2.
3.
4.
5.
6.
Burdek Bernhardt, 1994, Design, Gesichte Theorie und praxis der produktgestaltung (Basel, Birkhäuser).
Buur J , Windum J, 2002, MMS Design (København, DDC).
Johannessen K. , 1978, Kunst og Kunstforstaelse (Univ. forlaget., Bergen ).
Kouzelis A., 2014, The object as form and value, (Athens, Eumorfia).
Kuhn Th. S., 1962, The Structure of Scientific Revolutions (Chicago, University of Chicago Press).
Loewy Raymond, 1951, Never leave well enough alone (Baltimore, J. Hopkins Univ. Press).
ISBN: 978-960-98739-8-7
287
7.
8.
9.
10.
11.
12.
13.
14.
288
Lundequist J. , 1986, Norm och modell, (KTH Stockholm).
Lundequist J. , 1995, Design och produktutveckling (Studentlitteratur, Lund).
Monö Rune, 1997, Design for product understanding (Liber, Stockholm).
Polanyi M. , 1964, Personal Knowledge (Harper and Row, N.York).
Pye D. , 1978, Nature and Aesthetics of Design (Van Nostrand, London).
Selle Gert, 1973 ‚ Ideologie und Utopie des Design (Du Mont, Frankfurt).
Selle Gert, 1997, Siebensachen. Ein buch über die Dinge (Campus, Frankfurt).
Zetterlund Chr., 2002, Design i informationsåldern (Raster, Uppsala).
ISBN: 978-960-98739-8-7
Technology as a Creative Factor
Prof. Dr Tomas Kačerauskas
Department of Philosophy and Communication, Faculty of Creative Industries,
Vilnius Gediminas Technical University, Lithuania, [email protected]
Key words: technology, creative society, creativity indices, patent industry, Innovation
Index, High Tech Index.
Abstract
Two of possible creativity indices are High Tech and Innovation indices. The High Tech Index
is defined by R. Florida as the percentage of high technologies among regional technologies.
Innovation Index is the number of patents compared with the number of inhabitants in a region.
In both cases we face the problems as follows. A question arises as to how to measure high
technologies as a percentage of all technologies. The innovativeness of high tech is a specific
variety of creativity that has little to do with artistic creativity. Orientation to high technologies does not necessarily provide a competitive advantage or ensure a high standard of living.
High technology, whose creation requires creativity and special skills and knowledge, often
becomes a substitute for creativity and an instrument for the homogenisation of society. Patenting performs a social function by transferring ownership of patents and freeing inventors from
promoting their patents. The number of unimplemented patents creates no additional value or
impact on social development, i.e. does not indicate the quality of the inventions. The great
increase in the number of patents reflects the tendencies of a consumer society rather than of a
creative society.
Introduction
In presenting creative economy, J. Howkins (2010) reviews different branches of creative industries and their place in the economy. R. Florida (2012) raises the question of how to characterize the creative class, the main agent of the creative economy, in terms of indices. The
Bohemian, Innovation, High Tech, Gay, Talent and Melting Pot Indices which he defines could
be treated as qualitative indices of the creative economy. At least two of them are related with
the technologies developed in a society, i. e. Innovation and High Tech indices. Furthermore,
he defines two integrative creativity indices which we will call the Major Index and Minor Index according to their greater or lesser scope (see Table 1). It means that these two indices are
inseparable from are indices in the creative society.
No.
1.
2.
3.
4.
5.
6.
Index
Bohemian
Melting pot
Gay
High tech
Innovation
Talent
Integrative Indices
Minor Integrative
Index (diversity)
–
–
–
Major Integrative Index
Table 1. R. Florida’s indices of creative economy
ISBN: 978-960-98739-8-7
289
The paper deals with High Tech and Innovation Indices proposed by R.Florida which could also
be considered qualitative indices of the creative economy.
The scholars appeal to the different indices of the technologies including Global Innovation
Index (Al-Sudairi; Bakry 2014), hard and soft factors of innovativeness (Fu 2008), Index of
the patents (Al-Sudairi ; Bakry 2014), Creativity Support Index (Carroll 2009 Cherry; Latulipe
2014), Creativity Support Environment Index (Cherry; Latulipe 2014), Digital Media Index
(Weinstein 2014) etc.
High Tech Index
The High Tech Index is defined by R. Florida (2002; 2012) as the percentage of high technologies among regional technologies. This index is comparative in two senses: high technologies
are compared with low or traditional technologies, and furthermore, this proportion is compared
with the national average. According to R. Florida, high technologies bear witness to the creativity of society, provide competitive advantage, and finally, ensure a high standard of living.
We face here several problems. First, a question arises as to how to measure high technologies
as a percentage of all technologies. Even if we agree that high technologies are information
technologies or biotechnologies, yet they permeate all other technologies: one of high technologies’ criteria is their social receptivity and versatility. Being versatile, they combine with other
technologies (for example, with art technologies). Even a traditional area as agriculture (the
technology of cultivation) demands ever more technologies for information, accounting and
management. This symbiosis does not allow one to separate “high” and “low” technologies.
Second, the innovativeness of high tech is a specific variety of creativity that has little to do
with artistic creativity. True, creative economy covers both of them. However, more plausibly,
they are incommensurable or conflicting sides of creativity which bear witness to the social contradiction of creativity. Third, orientation to high technologies does not necessarily provide a
competitive advantage or ensure a high standard of living. Although the state of California with
high tech developed in Silicon Valley is one of the most dynamic in the USA, the state of Texas
with its orientation towards traditional business and production does not lag behind thanks to
its low tax rate. Besides, some areas of creative industry (for example, the crafts) deliberately
distance themselves from high technologies in production (although not necessarily in management) by appealing to a certain social antipathy to them and a nostalgia for a time before them.
High tech culture with its electronic music presupposes an anti-culture towards classical music.
Finally, we should decide our priorities: a creative society, a wealthy society or a happy society.
A creative society is not necessarily a wealthy society (with a developed economy), nor is it
necessarily happy, although creativity is one of happiness’s conditions.
Social priorities are mixed up when the need for high technology is based on the goal of a
wealthy society. The paradox is that high tech, especially IT, serves the strategies of mass consumption without the need for any strategists. In other words, high technology, whose creation
requires creativity and special skills and knowledge, often becomes a substitute for creativity
and an instrument for the homogenisation of society. That is why this index may indicate not
a creative society, but rather a consumer society, that is, a society of appetites that are manipulated by ratings. Finally, high tech is not necessarily ecological, quite the contrary.
Innovation Index
Another creativity index proposed by R. Florida (2002; 2012) is the Innovation Index. It is the
number of patents compared with the number of inhabitants in a region. The number of patents
reflects a particular (technological) creativity that is called innovativeness. It is a very practical
empirical index because it can be easily determined. Besides, there are rather clear criteria and
rules for patenting. The main principle here is as follows: inventions are patented and discov290
ISBN: 978-960-98739-8-7
eries are not. Nevertheless, here there are certain problems as well, especially if we treat this
as an indicator of creative economy. In general, the origin of patents in the 19th century in the
USA should be linked not so much with the creativity of society or some part of it as with entrepreneurship. Private patent offices were established using available information and seeking
to profit from both inventors and producers, i.e. from consumers of inventions1.
The patent institution cares not at all for the welfare of the inventors. All it promises are certain
rights, namely, the right of an inventor to the incomes from his (her) invention. Naturally, this
right must be paid for. The fee is so large that individual inventors usually are unable to pay for
the patent – only companies and universities are. After covering the patent costs, the companies
take over the rights to the inventions. In other words, patenting performs a social function by
transferring ownership of patents and freeing inventors from promoting their patents. The patent institution is a good example of how new additional value has been created with the help of
management technologies: without this institution there would be no demand for patents. The
patent institution itself is the fruit of its establisher’s invention (creativity).
Having all of this in mind, I come back to the question of whether and how much do the absolute or relative number of patents express the creativity of society. Are societies with a small
number of patents not creative? Does the fact that this index is low in Central and Eastern
Europe including Lithuania mean that societies here are not creative? That is not the case. It
means only that the patent industry has been less developed here, i.e. these societies are less
enterprising from this point of view. Also, it means a high cost threshold for inventors and the
financial weakness of companies and universities interested in having patents. Finally, it means
insufficient support and attention in general from the government towards this sector. And yet
in that the patent industry expresses exclusively commercial interests, government interference
can even make things worse: dumping can destroy branches of industry and nationalization of
inventors’ property rights can ruin them as agents of creative society. A counter argument could
go as follows: a low number of patents indicates that a society is not enterprising because it is
not sufficiently creative. Although entrepreneurship is a form of creativity, the Patent Index
could be a creativity index only indirectly, as the patent institution is a legacy of the industrial
society.
Meanwhile, the number of unimplemented patents creates no additional value or impact on social development, i.e. does not indicate the quality of the inventions. The significance of a tree
made from noodles or a dragon made from matches is not equal to the significance of a robot
that overturns industry. In seeking high indices, institutions patent everything without selection.
Having a certain policy of patenting or high barriers to patenting does not stop the avalanche of
inventions to be patented. The exponential increase in the number of patents which corresponds
to the increase in scientific texts in a certain period allows us to speak of an extension of D. S.
Price’s (1963) law to patents. The great increase in the number of patents reflects the tendencies
of a consumer society rather than of a creative society. Inventions are patented for the sake of
statistics, and afterwards they are thrown on the scrapheap.
Finally, we face one more problem, speaking about patents. Namely, we have different patenting
policies in different regions of the world. For example, comparing patent rights in Europe and
the USA shows that not only the standards of patenting but also the attitudes towards patenting
are different. For instance, the human genome has been successfully patented in the USA; this
would be absolutely impossible in Europe where the genome has been treated as an area to be
discovered but not to be invented. This illustrates not so much the more liberal attitudes concerning patenting in the USA as an attempt to extend the region of inventions at the expense of
the region of discoveries whereby parts of the latter have been annexed and privatised.
1
More about it in J. Howkins 2010.
ISBN: 978-960-98739-8-7
291
The question arises here as follows: who is the ruler and who are the vassals? Since they are
connected with the strategies of consuming, the ruler is often anonymous, i.e. the “they” who
pressure us to behave in the usual way. According to Z. Bauman, the real possessor of power in
consumer society is the market for goods (Bauman 2007), i.e. an anonymous entity.
Aside from this, the phenomenon of patenting has pointed up the vanishing border between the
regions of invention and discovery. This could be treated as an attribute of postmodern culture
in which different cultural territories overlap with each other. On the one hand, this overlap
can become a source of new creative ideas. On the other hand, regions without clearly defined
borders not only lead to conflicts between them but also to a large surplus of ideas. This works
against creativity which needs the support of a tradition. Invention and discovery presuppose
not only different approaches but also different attitudes towards knowing the world. Finally,
this nourishes the different sciences with new methods and new terms. If the different regions of
science, world view and culture are jumbled, then they become a media without traditions that
offers itself as the honeycomb for the honey of creative ideas. This artificial hive lacks distinctions between combs, honey, hives and apiaries, which is to say, lacks distinctions between scientists, cultural and creative workers. Different cultural, scientific and social environments bear
witness to different relationships between invention and discovery. Some authors (Cetindamar
and Gunsel 2012) criticize the Innovation Index as not adequate in certain regions.
The fact that this border is more liquid in the USA bears witness to not so much its greater creativity compared with European society as to their different cultural attitudes. So the innovation
index reflects neither the region’s ability to innovate nor its creativity. That is why this index
should not be used as the main index of creativity, but rather perhaps as an auxiliary together
with other indices.
References
1. Al-Sudairi, M.; Bakry, S. H., 2014, Knowledge issues in the global innovation index: Assessment of the state
of Saudi Arabia versus countries with distinct development. Innovation-management policy & practice, 16
(2), 176–183.
2. Bauman, Z., 2007, Consuming Life (Cambridge: Polity Press).
3. Carroll, E. A.; Latulipe, C.; Fung, R. et al., 2009, Creativity Factor Evaluation: Towards a Standardized Survey
Metric for Creativity Support. Proceedings of the 2009 ACM SIGCHI conference on creativity and cognition,
127–136.
4. Cetindamar, D.; Gunsel, A., 2012, Measuring the creativity of a city: a proposal and an application. European
Planning Studies, 20 (8), 1301–1318.
5. Cherry, E.; Latulipe, C., 2014, Quantifying the Creativity Support of Digital Tools through the Creativity Support
Index. ACM Transactions on computer-human interaction, 21 (4), Article 21 (1–25).
6. Florida, R., 2002, The Rise of Creative Class. And how it’s transforming work, leisure, communitiy and
everyday life. New York: Basic Books.
7. Florida, R., 2012, The rise of the creative class revisited. New York: Basic Books.
8. Fu, X., 2008, Foreign direct investment, absorptive capacity and regional innovation capabilities: Evidence
from China. Oxford Development Studies, 36(1), 89–110.
9. Howkins, J., 2007, The Creative Economy (London: Penguin).
10. Price, D. S., 1963, Little Science, Big Science. New York: Columbia University Press.
11. Weinstein, B. L.; Clower, T. L., 2000, Filmed entertainment and local economic development: Texas as a case
study. Economic Development Quarterly, 14 (4), 384–394.
292
ISBN: 978-960-98739-8-7
Science and Technology in Archaeological Museums:
The Re-Exhibition of the Archaeological Museum of Tegea
Gregory Grigorakakis1*, Panagiotis Ioannidis2 and Andreas Tsatsaris3
Ephorate of Antiquities in Arcadia, Greece,
Indepented IT Solutions Architect
3
Technological Educational Institute of Athens, [email protected]
1
2
Keywords: Museums’ Technology, Multi-touch Interactive Tables, IT Applications,
Archaeology, Spatial Technology, Tegea.
Abstract
The re-exhibition of the Archaeological Museum of Tegea, presents for the first time to the
public not the mere fact, but the course of events, the causal relationships leading to the consolidation of the institution that constitutes the hallmark of the Greek way of social organization,
namely the polis (city-state). In this framework the motivation of the visitor to deepen in this
complex phenomenon through the myth, dialects, social, economic, political and military history as well as the recent historical – archaeological approaches lay at the center of the whole
effort.
Simultaneously, besides the detailed archaeological documentation of the above and in order for
us to provide the optimum results, we implemented an integrated multidisciplinary approach.
Geography and Cartography now played an important role. Spatial analysis of the archaeological - historical data was introduced, so as their documented interrelation with geographic space,
geological environment and topography to be accomplished. Thus the goal of presenting in a
comprehensive way the actual space certain events took place, as well as their projection to the
present was fully achieved.
Consequently, the aim of this project was the combination of the archaeological, historical and
geospatial knowledge with the advanced digital technology of the interactive surfaces, and
graphic art, which can represent effectively and comprehensively the anthropogenic processes
or even natural phenomena and situations to the public.
Introduction. The Museum from the 20th to the 21st century
In the early 20th century the Archaeological Society at Athens, with enhanced financial resources, mainly from the lottery for antiquities, and in close dependence on the State, is engaged in
activities which are held primarily in the realm of state responsibility. By order of the SecretaryGeneral of the Society, Konstantin Romeos undertakes the arduous task of completing the construction of the building, which then had to be equipped with the necessary infrastructure and
finally, organizes of the exhibition program.
Almost 30 years later, in the 30s, we learn from reports that the condition of the Archaeological
Museum of Tegea was not good. Marcelo Mitsos was the archaeologist in charge for both the
repair works and the reorganization of the exhibition. The third gallery of the museum seems to
have been built in 1935 or 1936.
The condition of the Museum of Tegea after the war was not improved compared to that of
the 1930s. From 1959, however, an effort to improve the situation would begin. Repair works
started in 1967 and finished in 1968 under the supervision of Angelo Delivorias, then a young
archaeologist in the Greek Archaeological Service.
ISBN: 978-960-98739-8-7
293
The integration of the program for the upgrade of the building of the Museum into the Operational Program “Culture” of the 3rd Community Support Framework (3rd CSF) started in
2005. In 2011 the project for the re-exhibition of the Museum was integrated into the NSRF,
specifically into the regional operational program “Western Greece - The Peloponnese - Ionian
Islands.”
The new museological approach to the re-exhibition is based on the central axis: “Tegea from
Myth to History: genesis and evolution of an Arcadian polis and its territory”, which is analyzed
in: (a) “Museum Memories” and (b) “Genesis, Formation and Evolution of a Greek City-State.”
In the project of the re-exhibition includes three subprojects are embodied. Procurement and
development of digital applications in the Museum of Tegea is the one discussed in this paper.
Applications Approach
Designing the interactive applications constituted the real challenge of the whole project of the
museum re-exhibition. Two interactive application groups were devised, compatible with the
general outlines of the approved museological study: 1. Timeline (e-book) and 2. Virtual history. The amount of information collected and studied was immense. It was not just a bibliographic updating. It comprised research of primary sources, namely archival material, dispersed
in many scientific institutions throughout Athens and abroad. This material had to be located,
digitized and studied. However, collection of sources is never a real problem. It is rather a matter of time. Systematization of such a vast material constituted our main concern, given the
following prerequisites: A. The constraints imposed by the technology used, which literally
dictated the way the material was organized, B. The budget at our disposal, as well as C. the
timelines of the whole project, D. Accuracy of the information provided, E. Fidelity towards
the non-specialist visitor.
Therefore database tables were compiled. Certain thematic axes constituted the primary level of
information. “Thematic Entities” and “Sub-entities” followed in a secondary level:
Table 1. Values entry table
In a tertiary level, the so-called “Thematic Categories” and “Sub-categories” were developed
and analyzed. This tertiary level, constituted a detailed outline of the scenarios of the two groups
of interactive applications, which were to be developed according to a certain structure. This
three level structure allowed for both the control of the scientific accuracy of information as
well as the quantification – measurability of each step of the project, which is essential as far
as the project management, is concerned. Thus, things were organized as presented in the following chart:
294
ISBN: 978-960-98739-8-7
Figure 1. Workflow chart
In this context thematic categories (i.e. Space-Time and Social Evolution, History, Myth, Language, etc.) and subcategories (Geology, Geography, Man and its space, Arcadocypriot Dialect,
Historical Framework of Tegea, Arcadian Genealogy, etc.) are not independent structures. On
the contrary, they are interconnected in a variety of combinations.
Multi touch Interactive Table Applications
The purpose of the applications that are shown in the multi touch interactive table is to present
a more comprehensive insight into the history of ancient Tegea, through information data and
maps, and to present to visitor’s studies and theories on the creation and expansion of urban
centres in antiquity. Due to the above, as well as to achieve scientifically valid performance
information for the visitor and a more interdisciplinary approach, applications emerged as a
well-directed and user-friendly visual representation of bases of archaeological and geographic
data created exclusively for Tegea’s region.
Combined with the experiential data required in each experience interaction with digital media,
applications have the form of interactive objects. They are independent and consists of simulations of physical objects, for example pictures, text pages or notes, which are tied into an attractive GUI that makes comprehensible the connection between data, within the same thematic
category, but may be at different depths in each interactive object, or even more than one such
objects.
ISBN: 978-960-98739-8-7
295
The interactive applications follow an object-oriented and not hierarchical model. The independent components for each of the subsets of modules present as much as possible independent information while a simultaneously connection to the others is allowed. The user can
navigate to all existing information, allowing other users simultaneously to share the same
information with him.
Each independent task while running contains obvious ways to connect with other modules so
that visitors know and may continue to be updated for other issues. The interface is created in
such a way that the visitor can follow his own selection path while simultaneously can see all
the information. Different targeting groups (tourists, school groups, local residents, scientists
and scholars with special interests) are taken into account.
Extensive texts are avoided. Generally, text streamed are boring, tiring the visitor and is not
compatible with the interactive media visibility of information. Where relevant textual is needed the information is given briefly or by diagram. In this way, the texts are concise, the information immediately comprehensive, and the visitor is led straight to the essence of things. There
is logically a cause - effect relationship and immediacy. For better management, control and
performance of the contents in the application, and the simplicity of use of the data from the
Museum, all information are in simple excel spreadsheets converted to XML (Extensible Markup Language) sheets, which are used worldwide for the organized transfer of data between
disparate systems.
In the “Timeline” (first group of applications) the visitor can at a glance see the general developments focusing simultaneously on Tegea and to the evolution in time. In this application,
through archival material, the history of the Museum and the archaeological research in Tegeatis is presented, interrelated with events concerning political and military history, the evolution
of national and international institutions concerning the protection and conservation of cultural
heritage etc. Thus, a general framework, formed out of events - turning points of modern Greek
history, both organizes and interprets the specific archival material, through which the “history
of Archaeological Research” in Tegeatis unfolds parallel to the general socio-economic and political developments in the country. The way archival material and other information is indexed,
categorized and organized as well as the specific interface created, allow the visitor himself/
herself to decide what, when, how and if!
“Virtual history” is the second application group, no doubt more complex than the first one.
The applications use individual independent work (components) based in the objectives of the
museum. Using “Users Robotic Interface” interactive objects, visitors create interactively their
own script scenarios. Each script manages a number of independent operations (components)
with the smallest possible plot to make it more understandable to the visitor and to permit interconnection with other independent work in a simple interactive way. The design of Users
Robotic Interface (interactive objects) defines interactivity applications with the acts of the
visitor.
From the area of the history of the museum and scientific research in Tegeatis, now we pass to
the Polis itself: Its birth, establishment and evolution. The narration covers a span from Prehistory up to the end of Antiquity, with a special reference and analysis of the Early Iron Age
political and socio-economic processes that led to the formation of the Polis. The significance
of the so-called peripheral or rural sanctuaries is underlined emphatically. They are the spots of
contact between man and the divine, they are places were dispersed human communities meet
each other, with the result that both tight bonds of unity are gradually formed and identity of
interests is confirmed. As such, the peripheral sanctuaries constitute the essential field of aristocratic antagonism, ranging from showy dedications, consumption etc. up to armed conflict.
Many of them have been founded in marginal areas; this is either in the wilderness (forests,
mountains or marshes) or at the edge of the cultivated land, which makes their association with
296
ISBN: 978-960-98739-8-7
the concept of “border” or of “territory” comprehensible. Monumental architecture and sculpture develops in these sanctuaries. One of them, the most prominent one, constitutes the core
that unifies the dispersed communities into a single unit, the polis which continues to evolve up
to the end of Late Antiquity.
Concluding Remarks
As much as demanding and multi-parameter though it was, the project of the re-exhibition of
the archaeological museum of Tegea was accomplished as scheduled. Such qualities as “multiparameter” or “demanding” do not merely refer to its budget (1.200.000€) or to the number of
the sub-projects it consisted of (in concise: Documentation, exhibits showcases construction
and creation of digital applications). It must be understood in conjunction with the core of the
project its-self that was not just a public work. We had to deal with an archaeological museum,
a place where modern man is getting in touch with a civilization and culture of the past, a civilization and culture which has long since ceased to exist, but, as a way of living and attending
the perceptible reality, influenced (and of course continues to influence) and determined what
we know as western civilization, regardless of the different points of view that it was analysed
and comprehended. We are talking about civilization and culture, in Greek “politismos”. It
originates from the institution, the social formation and way of organizing human life that was
borne and developed in Greece, to be accurate in southern Greece. This was the “Polis”.
Designing the interactive applications constituted the real challenge of the whole project of
the “Re-exhibition of the Archaeological Museum of Tegea”. Neither was there a precedent
in a Greek museum (not to say worldwide) so far, nor the approved by the Ministry of Culture
studies, which we were supposed to be relied on, dealt sufficiently with so complex an issue.
Also the financing of the “interactive applications subproject” was demanding enough. In terms
of cost-benefit analysis, we were obliged to improvise, to think differently. Furthermore, both
the hardware and software we had decided to use were highly sophisticated. In any case we
couldn’t do otherwise. There was nothing we could elaborate on or even presented in another,
maybe more sophisticated, way.
Conclusively, this project dealt with the idea of the simplicity on the approach of the combination of complex concepts, which are diffused between the sciences of history, archaeology,
geospatial technology and informatics. The result was the creation of technological products
perfectly reachable to the general public. However, the abovementioned could not be achieved
without the value of an interdisciplinary approach for a demanding and composite work, which
will be subject to the ongoing public judgment.
Acknowledgements
The authors express their acknowledgements to the organizers of this conference.
References
1. Bowen, J. et. al., 2008, Digital Technologies and the Museum Experience: Handheld Guides and Other Media,
AltaMira Press, Plymouth, United Kingdom.
2. Γρηγορακάκης, Γ., 2014, Το Έργο της Επανέκθεσης του Αρχαιολογικού Μουσείου Τεγέας: Οι Ψηφιακές
Εφαρμογές και το Αρχειακό Υλικό. Στο: Η Ιστορική και Αρχαιολογική Έρευνα στην Πελοπόννησο, όπως
προκύπτει από τα αρχεία των Γ.Α.Κ. Νομών Πελοποννήσου και αρχεία άλλων φορέων, 4-5/10/2014.
Αρχαιολογικό Ινστιτούτο Πελοποννησιακών Σπουδών: Η Αρχαιολογία στην Πελοπόννησο 2, (Τρίπολη), σσ.
71-87.
3. Γρηγορακάκης, Γ., Ιωαννιδης, Π., Τσάτσαρης, Α., 2014, Ένα καινοτόμο τεχνολογικό ταξίδι στη γη του μύθου
και της ιστορίας. Στο: Αρχαιολογικό Μουσείο Τεγέας. Ένα τεχνολογικό ταξίδι στη γη του μύθου και της
ιστορίας. Επιστημονική Επιμέλεια Γ. Γρηγορακάκης, Π. Ιωαννίδης, Α. Τσάτσαρης, Τρίπολη, σσ. 23-137.
4. Darley, A., 2002, Visual Digital Culture: Surface Play and Spectacle in New Media Genres. Routledge,
London
ISBN: 978-960-98739-8-7
297
5. Longley, P, Goodchild, M, Maguire, D, Rhind, D, 2005, Geographical Information Systems and Science, John
Willey and Sons Ltd, England.
6. Rick, J. , Rogers, Y., 2008, From DigiQuilt to DigiTile: Adapting educational technology to a multi-touch
table. Working paper, 3rd IEEE International Workshop on Horizontal Interactive Human Computer Systems,
pp 73-80, DOI: 10.1109/TABLETOP.2008.4660186
7. Servet; B., Ramadan, I., 2005, Digital Map Literacy Bounded With Culture Under Geographical Information
Systems Perspective. Journal of Visual Literacy, 25 (2), 167-176
8. Thrower, Norman J.W. Maps & Civilization: Cartography in Culture and Society. Chicago, 1996.
9. Τσάτσαρης, Α., Γρηγορακάκης, Γ., Ιωαννιδης, Π., Σκουλουφιανάκης, Δ., Καρεγλή, Σ., 2013, Η Ψηφιακή
Χαρτογραφία του Αρχαιολογικού Μουσείου Τεγέας. Ένα χαρτογραφικό ταξίδι στη γη του μύθου και της
ιστορίας. Πρακτικά του 13ου Εθνικού Συνέδριου Χαρτογραφίας της Χαρτογραφικής Επιστημονικής Εταιρείας
Ελλάδος (ΧΕΕΕ), Πάτρα, (υπό έκδοση).
298
ISBN: 978-960-98739-8-7
Oil paintings on paper support: Determination of condition criteria
via non destructive testing and microanalysis
A. Alexopoulou1, B. Singer1,2, P. Banou3, S. Zervos4, A. Kaminari1
A. Moutsatsou5, A. Terlixi5, E. Tziamourani1, A. Karabotsos1,
M. Doulgeridis5
Department for Conservation of Antiquities and Works of Art, TEI of Athens, Greece,
[email protected]
2
Analysis for Conservation, UK
3
Department for Conservation, General State Archives, Athens, Greece
4
Department for Library Sciences and Information Systems, TEI of Athens, Greece,
5
Department for Conservation, National Gallery – Alexandros Soutzos Museum (NGASM),
Athens, Greece
1
Keywords: paper support, analytical techniques, multispectral imaging, artificial
ageing
Abstract
This paper presents the physicochemical methodology especially nondestructive testing and
microanalysis applied for the documentation of the condition of oil paintings and oil sketches
on paper support. The research is focused on the effect of the oil binder on the paper and the
changes caused during the progress of deterioration in order to provide criteria for the better
evaluation of the condition. Six original artworks representative samples of the Greek artists
of the 18th-20th century, belonging to the National Gallery - Museum and Alexandros Soutzos
Museum were investigated. The research was supported by experimental laboratory study of
artificially aged mock ups made of linseed oil and different paper supports. The research methodology involved non-destructive imaging (UVF, multispectral), FORS, colourimetry, microscopic techniques (optical VLM/FLM, SEM) and micro analytical techniques (HS-SPME-GCMS, SEM–EDX, and FTIR) for the examination and the identification of the materials and the
degradation products of the original artworks and the mock-ups before and after ageing. The
present article discusses the evaluation of the obtained results and the assessment of the techniques employed as far as the applicability and the efficacy in the study of the particular subject.
The appraisal of the methodology is expected to contribute to the development of an experimental tool for the evaluation of the condition of this particular type of works. Introduction
Oil binders have been widely used in various works on paper supports, such as oil sketches, oil
studies, drawings and paintings, as well as, images and texts in books printed with traditional
oil based inks. Certain problems recorded in these works appear to be related to the presence
of the oil binder; discoloration of the paper varying in intensity, reduction of the mechanical
strength and embrittlement of the support on areas where oil diffusion or absorption has occurred. Even though the subject raises conservation and preservation issues, there are only
sporadic references in literature and limited publication upon research on this matter. The majority of the publications refer to the painting materials, the types of support and the preparation
required/applied, the various techniques used, the terminology and their distinction respectively
(Gritsay et al, 2004, Bower, 1992,). There is a limited number of articles referring to the condi-
ISBN: 978-960-98739-8-7
299
tion and the problems presented in this type of works, the conservation treatments and the risks
involved. They are mostly focused on oil stain removal, colour change reversion, repair and lining techniques, methodology to support the fragile paper in rigid secondary supports for display
and storage (McAusland,1989, Banou et al, 2015).
Methodology
Six original works of art, representative samples of the Greek artists of the 18th-20th century,
belonging to the National Gallery and Alexandros Soutzos Museum were investigated. The
research was supported by experimental laboratory study of artificially aged mock ups made of
linseed oil and different paper supports (pure cellulose Munktel, typical watercolour Montval
and a Kraft type paper). The mock ups were submitted to close environment artificially aging according to ASTM D 6819-02 2002 standard (90 °C, 78%RH 1,4,7,14,21 and 28 ageing
days). The methodology of research involved technical examination using non-destructive imaging techniques (Fisher and Kakouli, 2006) like UVR, UVF and Vis/IR multispectral imaging
(Fig.1), FORS and colourimetry, as well as microscopic techniques (optical VLM/FLM, USB
Microscopy in combination with SEM–EDX) and the advanced VOC/MS analysis (HS-SPMEGC-MS) applied for the first time to the study of the effect of drying oils in paper except in our
own preliminary studies, for the examination and the identification of the materials and the
degradation products of the original artworks and the in depth study of extensive multi – parameter series of mock -ups before and after ageing.
Fig.1 N.Gysis, “Sewing studio” Oil painting, detail (NGASM). The comparative study of the same area in Visible, UV Reflection, UV Fluorescence, False Colour Infrared (FCIR) and Hyperspectral imaging (MuSIS HS) at
600 and 1000nm reveal information (the sketch, the pigments, the morphology) from different depth.
Results
Optical microscopy and SEM offered valuable information about the oil – paper system Observations on morphology of the three different types of paper show that papers without an oil application present qualitative changes that are far less in extent than those of the oiled papers indicating a
rather more stable behaviour upon ageing (Fig.2). Generally, the morphological changes and damage of the fibres recorded are limited. Oiled mock-ups of the three types of paper, before ageing,
present a similar condition: a layer of oil that almost completely covers the paper surface. The
oil layer presents a surface profil which responds to the relief of the fibre net or might be attributed to the different adsorption degree of the oil locally, which is connected to the porosity
of each paper and relevant surface tension properties.
Upon ageing the phenomena evolve in the same way for all three types of paper. There is a progressive but uniform sinking of the oil film, parts of the surface fibres are progressively more
distinguished, holes and recesses appear, revealing the fibres and the fibre net. By the final stage
of ageing the bulk of the oil film is still present at a lower level and the presence of holes and
recesses is so extended that the oil layer seems to be fragmented.
300
ISBN: 978-960-98739-8-7
Fig.2. Electronmicrographs of kraft paper without oil (a), (b) and with oil (c), (d). no ageing (a), (c) and after 28
days ageing (b),(d). Magnification x200
In all three paper types, oil application results in the significant reduction of brightness which reaches comparable values of the L* coordinate at the 28th day (Fig. 3). Analogous decrease of yellow
colour upon ageing is also observed. For the white colour papers the increase of the red colour, and
the fluctuation pattern within that, could be indicative of the chemical changes. Also differences in
colour through each mock up are strongly associated with the oil concentration on the paper locally.
The creamy oiled montval paper mock ups get an orange-brown hue from the first day of ageing
and turn gradually darker to a warm brown colour upon the progress of ageing. The colour of
the mock ups turn to an intense dark brown colour at the final stages of ageing. It is likely that
linseed oil application is principally responsible for the colour changes during ageing, though
oxidation of the cellulose in the paper may also contribute to the colour change, especially if ,
as we suspect, the oil is enhancing the rate of oxidation of the paper, as indicated in our VOC
emission studies.
Fig.3 The L* coordinate values for the three types of paper mock ups at all ageing periods.
Left: paper mock ups without oil application, right: paper mock ups with oil application
In transmitted light (Fig.4), the three types of paper (cotton, montval and kraft) allow a limited
amount of light to pass through, differing in quantity among the paper types, with kraft paper
being significantly less transparent than the others. The oiled mock ups of the three paper types
appear to allow gradually less light pass upon ageing, without getting absolutely opaque even
at the 28th day of ageing.
The presence of drying oil in paper greatly accelerates the emission of volatile cellulose degradation products both for cotton based and wood based papers. In the cotton paper it has greatly
accelerated the emission of furfural, 2-ethyl furan, 5- methyl, 5- ethyl furfural and 5-pentylfuranone during ageing, while in the wood based papers has greatly accelerated the emission of
furfural, 5-methy furfural, 5-ethyl furfural and 5-pentylfuranone.
ISBN: 978-960-98739-8-7
301
Fig.4 Transparency of non oiled (left) and oiled mock-ups (right)
Conclusions
Deep understanding of the paper – oil system optical alterations due to the deterioration were
established through extensive studying of mock-ups. The catalytic effect that oil medium has
on the acceleration of the entire degradation was examined and recorded in various ways. The
research has shown that the presence of dried oil films accelerates the oxidative degradation of
artists’ paper including both cotton, and therefore rag paper and also wood based artists’ papers.
We can conclude that since drying oil clearly accelerates the destructive oxidation of the paper
in the mock ups that it is having the same effect in the works of art in the collection studied.
The associated loss of mechanical strength and deterioration of optical properties such as colour
will also be accelerated in these works. A methodology based on non-destructive analysis was
also established that enables the conservation scientists to have a better insight in the deterioration of the originals. These findings will provide sound data on which informed decisions can
be made regarding the conservation treatments for these works and that therefore this research
will have an impact on the lifetime of these works making them available for public viewing
for an extended period.
Aknowledgements
Sincere thanks to Prof. Dr. C.Tzia, Head of the laboratory of food analysis in the Faculty of
Chemical Engineers of the National Technical University of Athens and her research team members Dr.D. Tsimogiannis and Dr. Hara Chranioti, for carrying out the GC-MS analysis.
This research has been co-financed by the European Union (European Social Fund - ESF)
ARCHIMEDES III - Investing in knowledge society through the European Social Fund.
References
1. Banou, P., Alexopoulou, A. & Singer B., 2015, The treatment of oil paintings on paper supports. Considerations
on the treatment applications used from the past until the present, J. of Paper Conservation, IADA, 16, (1),
29-36.
2. Bower, P., 2002, A brush with nature: An historical and technical analysis of papers and boards used as supports
for landscape oil sketching. in Works of art on paper. Books, Documents and Photographs. Techniques and
Conservation, Baltimore Congress, 2-6 September, ed. Daniels, V., Donnithorne, A. & Smith, P., pp.16-20.
3. Fischer, Chr., Kakoulli, I., 2006. Multispectral and hyperspectral imaging technologies in conservation:
current research and potential applications, Reviews in Conservation, Number 7, 3-16.
4. Gritsay, N., Larionov, A., Durante. S.S., Vegelin, E. & Woodall, J. (2004), Peter Paul Rubens: Oil
sketches and related works from the State Hermitage Museum and Courtlauld Institue Gallery,
Prestel Publishing Ltd.
5. McAusland 1989, Problems with works of art in oil on paper and their support: A paper conservator’s
viewpoint. General Meeting of the Association of British Picture Restorers, Report. (UK).
302
ISBN: 978-960-98739-8-7
Swelling inhibition of clay-bearing building materials
N.A. Stefanis1*, M. Stefanidou2, A. Kalagri3, I. Karatasios3, V. Kilikoglou3,
I. Papagianni2 and P. Theoulakis1
Technological Educational Institute of Athens, Department of Conservation of Antiquities
and Works of Art, *[email protected]
2
Aristotle University of Thessaloniki, School of Civil Engineering
3
N.C.S.R. “Demokritos”, Department of Materials Science
1
Keywords: Clay-bearing building materials, thixotropic properties, swelling inhibition,
performance evaluation.
Abstract
Aim of the present paper is the study of deterioration characteristics of building materials with
clay impurities and the search for suitable materials and consolidation treatments to restore the
internal cohesion of the building materials and also to decelerate the decay processes. Particular emphasis is placed on examining treatment materials from the group of alkoxysilanes with
different composition so as to to offer, with escalating intensity, modification and control of
the surface tension, i.e. alteration of the hydrophilic properties of the substrates and absorption
of moisture and aqueous solutions. The effect of surfactants (e.g. fluorinated surfactants) is
independently examined on the phenomenon of swelling of the crystal lattice of certain clays
(swelling inhibitors) such as montmorillonite, with sorption of water molecules into the bilayer
positions. For the evaluation of tested treatment agents in the laboratory and in the field, an
original methodology is developed which results in the production of a model (algorithm) that
is based on data / performance criteria of treatments of the system building material – treatment
agent. Representative case studies from significant archaeological sites of Greece such as building materials from excavations of Toumba Thessaloniki, the prehistoric settlement at Dispilio
Kastoria, respectively and the Dafni Monastery in Athens.
Introduction
Climate changes that occur in the natural environment during the last decades are expected to
have a serious impact, direct and indirect to the preservation state of archaeological and architectural monuments (Casar, 2005). The rise of the sea level, the most frequent and intense flooding, the disorder of the hydrogeological characteristics of the coastal and continental areas as
well as the protracted periods of high temperatures, are expected to have an intense influence on
the weathering rate of the architectural heritage and sculpture which are made of natural stones,
especially those containing clay minerals (Baker et al., 2007). In most of the cases, this kind of
stones has been used in foundations or other parts of the buildings where there are no humidity
fluctuations or direct exposure to rain. Alumino-silicates participate in the formation of a great
category of sedimentary rocks such as marly limestones and sandstones. Due to the easiness
they present in quarrying, and use, these stones were widely used at a great number of archaeological and historic monuments. Some representative cases are the foundation of Erechtheion of
the Acropolis of Athens, the sanctuary of Poseidon in Poros, as well as the megalithic columns
of Heraion in Samos. Apart from stones containing alumino-silicate minerals, clays have been
extensively used since the Neolithic period for the construction of adobe structures. In Greece,
these structures can be found at Dipsilio, Kastoria, in Byzantine structures in Thessaloniki, as
well as in houses built after 1940. According to the type and the crystalline structure of the
ISBN: 978-960-98739-8-7
303
minerals participating in the mass of natural stones, some of these minerals have the tendency
to swell or shrink depending on the environmental conditions.
The presence of clay minerals in natural stones and adobes that are used in monuments produces a series of problems which relate to the internal stresses which are developed during the
absorption of moisture and the consequential swelling and shrinkage, as well as due to the fatigue the building materials which are subjected to repeated wetting and drying cycles (Jiménez
González et al., 2002). At a simplified model, the phenomenon evolves gradually by trapping
water molecules (layered hydrates) in between the crystalline structures of the clay-mineral
platelets, which present different rates of absorption of moisture and consequently of swelling
and shrinkage. The above-mentioned phenomenon can be interpreted through two basic mechanisms: due to the initial swelling because of the hydration of the counterbalancing cations and
due to the long-range, purely electrostatic osmotic swelling (Jiménez González et al., 2004,
Tim et al., 2004). The direct result of the mechanical distress caused by the swelling of clay
minerals found in natural stones and adobes, in combination with the chemical and mechanical
decay that can cause the presence of water in monuments, is that stones are superficially weathered and lose their internal cohesion. Long exposure of the stones to this mechanism leads to
the total decay and crumbling, threatening the stability of the whole monument. In these cases,
strengthening of the building and its decorative materials with a consolidation agent is the only
solution for their preservation, since the restoration of the internal cohesion, the increase of the
mechanical strength and the reduction of the weathering rate, can be achieved. Thus, the topic
of the present proposal is the study of the relationship between the lithological characteristics
of the building materials, of the swelling mechanism of the clay minerals as a function of the
particular characteristics of the direct environment of the monument (distance from the sea,
presence of soluble salts, percentage and fluctuation of relative humidity, rainfall) and of the
efficiency of different types of consolidation materials from the category of silanes (Félix et al.,
1994, Félix et al., 1995, Wheeler, 2005, Karatasios et al., 2009)
The originality of the present paper lies on the fact that it studies and evaluates the relation
between:
• The lithological characteristics of the building materials (mineralogical composition, type
and concentration of swelling minerals, microstructure),
• The swelling mechanism and alteration of clay minerals in correlation to the particular characteristics of the direct environment of the monument, and
• The effect of the abovementioned parameters to the efficiency and durability that different
types of silanes provide after consolidation treatments.
The evaluation of the system stone-environment-treatment is based on the development of a
multiparemeter model – algorithm that will accept and process objectively measurable properties (Theoulakis et al., 2008). This combination is in fact, the originality of the project, rendering it particularly important for the protection and conservation of archaeological and historic
buildings that are constructed of clay bearing building materials. Especially today, the issue is
significant, since climate change is affecting the microclimatic conditions of the monuments
and consequently the weathering rates of the materials and the demands for a consolidation
treatment.
304
ISBN: 978-960-98739-8-7
Acknowledgements
national resources under the framework of the “Archimedes III: Funding of Research Groups in
TEI of Athens” project of the “Education & Lifelong Learning” Operational Programme.
References
1. Baker, P.H, Sanders, C., Graham, H., Galbraith, R., McLean, C., 2007, Investigation of wetting and drying
behaviour of replica historic wall constructions. Engineering Historic Futures - Stakeholders Dissemination
and Scientific Research Report. May Cassar and Chris Hawkings (Eds). UCL, pp. 73-95.
2. Cassar, M, 2005, Climate change and the historic environment, Research report: Centre for Sustainable
Heritage, University College London, London, UK.
3. Félix C., 1995, Choix de gres tenders du Plateau Suisse pour les travaux de conservation (Choice of soft
sandstones from the Swiss plateau for conservation work). In Conservation et restauration des biens culturels,
Actes duCongres LCP, Montreux, Septembre 1995, R. Pancella Ed., EPFL, 45-71.
4. Félix, C.; Furlan, V., 1994, Variations dimensionnelles des gres et calcaires liees a leur consolidation avec un
silicate d’ethyle (Dimensional changes of sandstones and limestones related to their consolidation with an
ethyl silicate). In 3rd international Symposium on the conservation of Monuments in the Mediterranean Basin.
Edited by V. Fassina, F. Zezza. Venice, 22-25-June 1994.
5. Jiménez González, I., Higgins, M., Scherer, G.W., 2002, Hygric swelling of Portland Brownstone. In Materials
Issues in Art & Archaeology VI, MRS Symposium Proc., eds P.B. Vandiver, M. Goodway and J.L. Mass
(Material Res. Soc.), Warrendale, PA, (7), 21-27.
6. Jiménez González, I., Scherer, G.W., 2004, Effect of swelling inhibitors on the swelling and stress relaxation
of clay bearing stones. In Environmental Geology, 46: 364-377
7. Karatasios, I. Theoulakis, P, Kalagri, A,. Sapalidis, A., Kilikoglou, V., 2009, Evaluation of consolidation
treatments of marly limestones used in archaeological monuments, Construction and Building Materials (23),
2803-2812.
8. Theoulakis, P., Karatasios, I., Stefanis N.A., 2008, Performance criteria and evaluation parameters for the
consolidation of stone. Int. Symposium “Stone consolidation in cultural heritage. Research and practice”.
Lisbon, 6-7 May, pp. 279-288.
9. Tim J. Tambach, P., Bolhuis, G., Smit, B., 2004, A Molecular Mechanism of Hysteresis in Clay Swelling,
Angew. Chem. Int. Ed., (43), 2650 –2652.
10. Wheeler G. 2005. Alkoxysilanes and the consolidation of stone, Getty Conservation Institute, Los Angeles.
ISBN: 978-960-98739-8-7
305
Testing composites for the in situ protection of wooden shipwrecks
Alexios-Nikolaos Stefanis1* and Anastasia Pournou1
Technological Educational Institute of Athens, Dept. Conservation of Antiquities & Works
of Art, Ag. Spyridonos str., GR 122 10 Aegaleo, Greece
* [email protected]
1
Keywords: Composite materials, marine environment, mechanical properties
Abstract
This paper presents preliminary findings on the performance of composite synthetic materials
in the pelagic marine environment. The study undertaken, was part of a R&D project which
aimed to investigate, develop and implement innovative methods for protecting in situ wooden
shipwrecks found in high biodeterioration dynamic environments such as the Eastern Mediterranean. Samples made of various composite materials (e.g. FRP) were placed at the underwater
environment of the Rhodes’ harbour. The exposure trial aimed to find an appropriate material
for constructing an encasement system for a byzantine wooden shipwreck found at the harbour. Samples were retrieved after six and twelve months exposure period and their mechanical
properties were tested and compared with respective controls. Results obtained demonstrate
composites suitability for the underwater environment and led to the selection of the most appropriate construction material.
Introduction
Composites are widely used in marine applications and thus there is considerable experience
for the use of glass-reinforced resins in boats and ships (Davies and Rajapakse, 2014). However, the number of studies on the applications of such materials in the underwater environment
is limited (Gellert and Turley, 1999, Kootsookos and Mouritza, 2004). Under the framework
of a THALES project that dealt with the in situ preservation of wooden shipwrecks (Mermaid
2015), a new approach for protecting wooden wrecks found in high-energy areas was investigated. This involved the development of an encasement system that would create anaerobic
conditions unaffected by the hydrodynamic status of underwater sites. The materials tested
were two types of vinyl polymers, two types of glass/ester composite and one type of a glass
fibre reinforced honeycomb sandwich panel (Gu, 2009). All specimens were tested according to
ASTM D638-02a for their flexural strength, except the honeycomb sandwich, which was tested
for its compressive strength.
Results obtained indicate a strong impact of the underwater environment on the mechanical
properties of the materials (fig. 1). The glass/ester composites (GE1 and GE2) showed the
greatest reduction of their tensile strength after twelve months of exposure, reaching 93% of the
respective strength of the control. Accordingly, vinyl polymers (V1 and V2) exhibited reduction
of their mechanical properties as well, but V1 presented the lowest alteration by
-1.20% in
comparison to the reference material. Regarding the honeycomb sandwich, it reduced its original compressive strength by 61% after twelve months of exposure.
306
ISBN: 978-960-98739-8-7
a
b
Figure 1. Samples exposed in Rhodes’ harbour (a) and upon retrieval after12 months’ exposure (b)
V1
GE1
GE2
V2
Reduction of Tensile strength (Mpa)
6m exposure
12 m exposure
0.07%
-1.20%
-72.60%
-77.33%
-89.56%
-93.14%
-28.78%
-7.75%
Table 1. Tensile strength reduction (%) of tested materials after 6 and 12 months exposure
The material V1 showed the lowest decrease in mechanical strength (1.20%) and consequently
the greater stability. The tensile strength of the reference sample was 14.33 MPa, whereas the
sample which was exposed in the marine environment for twelve months showed a strength of
14.15 MPa.
The GE1 control, presented a tensile strength of 140.65 MPa, whereas the sample remained in
the marine environment for twelve months 31.89 MPa with an impressive drop of 77.33% in
tensile strength.
The GE2 control exhibited a tensile strength of 272.66 MPa, which was the highest value
reached of all materials tested, whereas the sample exposed in the underwater environment for
twelve months was 18.70 MPa with an impressive drop of 93.14%.
The V2 control exhibited a tensile strength of 1.81 MPa, which was the lowest value of all materials examined, whereas the sample exposed for twelve months was 1.67 MPa with a decrease
of 7.75%.
Finally, the honeycomb sandwich control sample showed a compressive strength value of 8.16
MPa, whereas the sample exposed for twelve months 3.15 MPa, with a decrease of 61.40%.
Conclusions
Results obtained revealed that the most resistant materials in the marine environment were vinyl
polymers rather than glass/ester composites. However, the construction of an encasement with
vinyl polymers needs more research to be undertaken as these materials need a metallic frame
support to adjusted in a construction and also they have limit puncture strength. Moreover, both
glass/ester composites show very poor performance. The tensile strength values demonstrated
were very low however the anisotropy of these materials needs to be further investigated, as
it is beloved that a cross laminated constructions could perform better. Finally the honeycomb
ISBN: 978-960-98739-8-7
307
sandwich, show very poor buoyancy properties for underwater applications and its not recommended for an underwater construction.
Acknowledgements
This research has been co-financed by the European Union (European Social Fund – ESF) and
Reinforcement of the interdisciplinary and/or inter-institutional research and innovation with
the possibility of attracting high standard researchers from abroad through the implementation
of basic and applied excellence research.
References
1. ASTM D638-02a Standard test method for tensile properties of plastics
2. Davies, P., Rajapakse, D.S., 2014, Durability of composites in a marine environment, Springer
3. Gellert, E.P., Turley, D.M., 1999, Seawater immersion ageing of glass-fibre reinforced polymer laminates for
marine applications, (30), 1259-1265
4. Gu, H., 2009, Behaviours of glass/fibre unsaturated polyester composites under seawater environment,
Materials and Testing (30), 1337-1340
5. Kootsookos, A., Mouritz, A.P., 2004, Sea water durability of glass- and carbon-polymer composites,
Composites Science and Technology, (64), 1503-1511
6. MERMAID, Date of access: 27 September 2015. http://mermaid.thales.teiath.gr/
308
ISBN: 978-960-98739-8-7
Polypeptide models of keratin and collagen as a tool for the
destruction of works of art
E. Fotou1*, E. Panou-Pomonis1, E. Ioakeimoglou2, E. Malea2,
G. Panagiaris2 and M. Sakarellos-Daitsiotis1
Section of Organic Chemistry and Biochemistry, Department of Chemistry, University
of Ioannina, Greece, *[email protected]
2
Department of Conservation of Antiquities and Works of Art, TEI of Athens, Greece
1
Keywords: polypeptide models, ELISA, decay, natural and artificial aging
Abstract
The effects of global climate change through air pollution, extreme values and fluctuations in
temperature and humidity, solar radiation and various microorganisms present in museums, favor almost the complete protein decomposition of organic materials of animal and plant origin.
Therefore, for maintainers of art and relics is vital to find new maintenance and storage conditions designed to preserve the cultural heritage of every nation.
An important area of application of collagen and keratin, is their use as proteins in organic materials such as bone, horn, vellum woolen fabric, paper, wood, cotton, etc. Decay of the above
materials is mainly based on oxidation and hydrolysis procedures, resulting to degradation.
Introduction
Our study is focused on the development of a reliable and reproducible immunoassay for the
evaluation of collagen and keratin-based decay of organic materials constituting natural and
cultural heritage.
For this purpose, were designed, synthesized and studied by the immunological technique
ELISA the following keratin and collagen models:
• the polypeptide model [Pro-Ser(OBzl)-Gly]n, based on the characteristic repetitve motif
(Pro-X-Gly)n encountered to collagen
• the polypeptide model [Pro-Cys(Acm)-Gly]n, based on the characteristic repetitve motif
(Pro-Cys-X)n encountered to C-terminal part of type I keratin and
• the polypeptide model Ac-YRSGGGFGYRSGGGFGYRS-βAla-NH2, based on the characteristic repetitve motif GGGFGYRS encountered to N-terminal part of type II keratin.
The above keratin and collagen models were used for immunization of laboratory animals to
develop specific antibodies against the polypeptide models. The produced anti-collagen and
anti-keratin, as well as commercial anti-collagen Type I and antibodies occurring at N-terminal,
C-terminal and central portion of keratin, were used for the immunochemical detection of decay
of:
• naturally aged samples of wool and parchment at three different areas (one urban - Criminology Museum, a seaside environment - Historical and Folklore Museum of Crete at Heraklion and a mountain area – Library Neophytou Douka at Zagori)
• artificially aged samples of bone at specific laboratory conditions (relative humidity. tempurature, NO2 and SO2).
It is expected, that as the absorption against antibodies of synthetic peptides increases, as the
destuction of naturally and artificially aged samples augments, due to the occurrence of hydrolyzed parts of the original proteins. On the contrary, the absorption against commercial collagen
and keratin antibodies increases, as the deterioration of collagen and keratin-based materials
ISBN: 978-960-98739-8-7
309
decreases, indicating low protein (collagen, keratin) degradation. Therefore the combined application of antibodies against the synthetic peptides and their protein of origin could be a valuable tool for the evaluation of the destruction of organic materials of cultural heritage.
The first two sequential polypeptides were synthesized on 2-chlorotrityl-chloride resin (1.49
mmol Cl-/gr) and after cleavage from the resin (with the mixture CH3COOH/TFE/DCM 2/2/6
v/v), the Boc-Pro-X-Gly-OH [where X=Cys(Acm) and Ser(OBzl)] was transformed in an active pentachlorophenyl ester Boc-Pro-X-Gly-OPCp in solution. The obtained product was polymerized, after cleavage of the Boc- group, with DMF in triethylamine.
The polypeptide Ac-YRSGGGFGYRSGGGFGYRS-βAla-NH2 was synthesized on a Rink
Amide AM resin (0.37mmol/g) using the Fmoc/tBu methodology. The coupling reactions were
performed using a molar ratio of amino acid/HBTU/HOBt/DIEA/resin 3/3/3/6/1. All residues
were introduced as Fmoc-protected amino acids. Fmoc- groups were removed using 20% piperidine in DMF. After the introduction of the last amino acid, the N-terminal part of the peptide
was acetylated. Cleavage of the peptide from the resin, as well as all the protective groups, was
carried out with the mixture TFA/Tis/H2O 95/2.5/2.5 v/v. The purity of the peptide was checked
by analytical HPLC and the correct molecular mass was confirmed by ESI-MS.
The reported keratin and collagen models were used for the immunization of New Zealand
white rabbits in order to develop specific antibodies against the polypeptide models. The collected sera of the rabbits were tested for the presence of antibodies against the polypeptides by
ELISA. It was found that the obtained sera recognize the polypeptides used for the immunizations.
Keratin extraction from naturally aged samples of 4mg of wool, was carried out in solution with
5N NaOH at 65oC, followed by pH adjustment to 9.5with CH3COOH.
Collagen extraction from naturally aged samples of 1.4mg of parchment, was carried out in
solution with 5N HCl at 50oC, followed by pH adjustment to 9.5 with NaOH.
Collagen extraction from artificially aged samples of 1.5mg of bone, was carried out in solution
with 5N HCl at 50oC, followed by pH adjustment to 9.5 with NaOH.
The immunological detection of the deterioration of keratin and collagen from natural and artificial aged samples carried out using a non competitive ELISA at 450 nm (Figs 1-3).
(a)
(b)
Fig.1: Recognition of naturally aged samples of wool (a) into show case and out show case against anti-AcYRSGGGFGYRSGGGFGYRS-βAla-NH2 and (b) into show case and out show case against anti-[Pro-Cys(Acm)Gly]n antibodies.
310
ISBN: 978-960-98739-8-7
(a)
(b)
Fig.2: Recognition of naturally aged samples of parchment (a) into show case and out show case against anticollagen and (b) into show case and out show case against anti-[Pro-Ser(OBzl)-Gly]n antibodies.
(a)
(b)
Fig.3: Recognition of artificially aged samples of bone (a) against anti-collagen and (b) against anti-[ProSer(OBzl)-Gly]n antibodies.
Conclusions
Our study indicates that the combined application of artificial antibodies
• against peptide model, which correspond to fragments of the original proteins (collagen and
keratin) and
• (ii) against collagen and keratin,
can give information about the extent of deterioration and consequently the urgent preservation
of our cultural heritage.
According to this work, given the opportunity to develop new standard reference curves to
estimate the damage protein organic materials, with a view to setting and achieving necessary
environmental conditions for preventive maintenance and storage of these materials in museums, libraries etc.
Acknowledgements
This study is supported by THALIS Program NSRF 2007-2013: “Investigation of the Environmental Factors effects on Organic Materials constituting the Natural and Cultural Heritage
(INVENVORG)”.
References
1. Zevgiti, S., Sakarellos, C., Sakarellos-Daitsiotis, M., Ioakeimoglou, E.and Panou-Pomonis, E., 2007, Collagen
models as a probe in the decay of works of art:synthesis, conformation and immunological studies, J. of
Peptide Science, 13, 121-127.
2. Fotou, E., Zevgiti, S., Panou-Pomonis, E., Panagiaris, G., and Sakarellos-Daitsiotis, M., 2012. Keratin and
collagen models as a tool in the deterioration of works of art. Proceedings of the 32nd European Peptide
Symposium, Athens, Greece, September, pp 660-661.
3. Fotou, E., Moussis, V., Panou-Pomonis, E., Panagiaris, G., Sakarellos-Daitsiotis, M., 2014. Keratin and
collagen polypeptide models as a means for assessing the deterioration of organic materials of cultural heritage.
Proceedings of the 33rd European Peptide Symposium, Sofia, Bulgaria, pp 79-80.
ISBN: 978-960-98739-8-7
311
Assessing Environmental Effects on Organic Materials in Cultural
Heritage: Chemical Deterioration of Artificially Aged Bone
Stamatis Boyatzis*1, Eleni Ioakimoglou1, Yorgos Facorellis1, ClioVossou4,
Maria Sakarellou, Eugenia Panou-Pomoni, Fotou Evmorfia, Efrosini
Karantoni1, Angelos G. Kalampounias,3 George A. Voyiatzis,4
Soghomon Boghosian3,4, Jane Richter5, Athanasios Karampotsos1, Eleni
Tziamourani1 and George Panagiaris1
Technological Educational Institute of Athens (TEI-A), Department of Conservation
of Antiquities and Works of Art, [email protected]
2
Department of Chemistry, University of Ioannina
3
Department of Chemical Engineering, University of Patras
4
FORTH/ICE-HT, Patras
4
Technological Educational Institute of Piraeus (TEI-P), Department of Electronics
5
The Royal Danish Academy of Fine Arts (DRAFA), School of Conservation
1
Keywords: environment, pollutants, cultural heritage, bone, ageing, analysis,
chromatography, SEM, vibrational, spectroscopy, immunoassay.
Abstract
Under the auspices of INVENVORG (Thales Research Funding Program – NRSF), and within
a holistic approach for assessing environmental effects on organic materials in cultural heritage (CH) artefacts, the effect of artificial ageing on elemental and molecular damage and their
effects on the structural integrity of bone was investigated. Metapodial roe deer bone samples
were artificially aged under humidity and atmospheres of sulfur and nitrogen oxides in room
temperature. Elemental micro-analysis of bone material through SEM-EDX and molecular investigations through FTIR and Raman spectroscopy, high performance liquid chromatography
(HPLC) and Enzyme Linked Immunosorbent Assay (ELIZA) were realized. Results show damage within the inorganic and the organic matrix; incorporation of sulfur and nitrogen groups,
minor reduction of specific aminoacids and changes in collagen integrity were evidenced. A
synthesis of the various results is presented.
Introduction
Bone artefacts have enormous importance for our natural and cultural heritage, for understanding the past of life on earth and past human behavior. Bones also consist challenging objects
for the conservator. The proper environment is essential for sustaining preservation of bone
artefacts in museum exhibitions, archival repositories, etc. A wealth of information is available
on the effect of burial conditions on bone artefacts (Child, 1995; Collins et al 1995; Stiner et
al 1995; Smith et al 2007, Weiner 2010 and references therein). On the contrary, no comprehensive assessment of museum environmental conditions and their correlation to the various
aspects of damage on bone has been undertaken.
An initiative focusing on a holistic approach for assessing environmental effects on organic
material within cultural heritage (CH) artefacts has been undertaken in the form of INVENVORG (Investigation Of The Environmental Factors Effects On Organic Materials Constituting
The Natural And Cultural Heritage) under the auspices of Thales Research Funding Program:
Investing in knowledge society through the European Social Fund (Operational Program «Edu312
ISBN: 978-960-98739-8-7
cation and Lifelong Learning» of the National Strategic Reference Framework-NSRF). The
results in this presentation are part of this Programme. After a detailed experimental design
(Dellaportas et al 2013), the congruous application of a wide range of techniques may offer
valuable evidence as to the specific damaging level (mechanical, structural, molecular) through
a systematic examination of factors like heat, humidity and gaseous pollutants (such as nitrogen
and sulfur oxides) through artificial and natural ageing studies. Comprehensive damage assessment on other materials such as parchment has been previously reported [Della Gatta et al 2007,
Odlyha et al 2007] and have been invaluable in providing information and guidelines for study
methodologies, etc.
From the structural and molecular point of view, bone is composed of organic matrix (mainly
collagen, with lower amounts of non-collagenous proteins, lipids and proteoglycans) and inorganic matrix i.e. hydroxyapatite and carbonates (Weiner 2010). Water is also essential (approximately 20%) part of bone, mainly as structural component (crystalline water), bound water, free
water, and water (adsorbed in gaps of both inorganic and organic matrix elements (Timmins
P. A. and Wall 1977, Gonzalez 2013).
Bone artefacts exposed to environmental conditions may suffer damages through exposure
to light, moisture, heat, and atmospheric pollutants (such as sulfur and nitrogen oxides). The
combined effect of these factors causes a complex deterioration pattern in the organic (mainly
protein) and inorganic matrix. The chemical composition that reflects the molecular integrity of
the complex bone matrix is affected. As a result of this, the color and surface morphology, and
internal microstructure deteriorate. On the mechanical properties level, finally, collagen damage undermines the integrity of the material, making bone brittle leading to complete collapse.
The state of preservation of the collagen fibers and that of carbonate hydroxyapatite are key
targets for analysis at the molecular and elemental levels. This specific presentation focuses on
artificially aged bone samples.
Bone samples and artificial ageing
Metapodial bone samples were taken from roe deer – Capreolus and were suitably processed to
remove fat in room temperature without affecting the structure of the items. A universal laboratory (or artificial) ageing scheme has been designed, involving combination of factors like
humidity (45 and 70 %RH) and atmospheres of certain concentrations of sulfur oxides (SOx,
100 and 300 ppm) and nitrogen oxides (NOx, 300 and 100 ppm) for 14 and 28 day periods.
Temperature was kept at 25°C.
SEM-EDX analysis
In SEM-EDX analysis, randomly selected micro-areas on bone surface were examined and
their elemental analysis was performed. Sulfur is clearly detected after exposure of the sample
in SOx (with the presence of NOx) atmospheres. This demonstrates the effective incorporation
of sulfur atoms inside the bone matrix in sulfur-polluted environments.
FTIR results
Previous investigations in collagenous materials, such as parchment aged in SOx and NOx atmospheres, have shown significant changes in FTIR spectra [Odlyha et al 2007, Della Gatta G., et al
2007). Similar observations were done in powder samples taken from the organic matrix of SOx
and NOx-aged bones: moderate collagen removal (decrease in the ratio (A1668 / A1024), suppression of
amide II peak in relation to amide I, and newly-formed nitrate bands within the matrix were evident.
ISBN: 978-960-98739-8-7
313
Also, in the inorganic matrix, small decrease in carbonates type II (located at the trivalent PO43- sites,
detected as broad band at 1420 cm-1) was detected. An extended array of spectra detected in artificially
aged samples in comparison with reference samples along with statistical evaluation of results are
presented. Typical spectra are shown in Figure 1.
1.0
0.9
amide I
0.8
1668
0.7
0.6
A
phosphates
collagen bands
carbonates
1.12
amide II
1549
1
0.5
1024
Type 1 Type 2
1456 1419
0.4
2
0.3
1385
0.2
872
1241
0.1
0.0
Figure 1: FTIR spectral region containing amide I,
II, carbonates and phosphate groups of artificially
aged bone sample.
Curve 1: Reference sample;
Curve 2: Aged sample.
Spectra in absorbance mode.
-0.13
1890.6
1800
1600
1400
Wavenumber cm-1
1200
1000
861.0
Raman Spectroscopy Results
FT-Raman spectra (1064 nm laserline of a Nd:YAG laser) showed peaks at ~1000 and 1045
cm-1, indicative of the incorporation of sulfate and nitrate groups into the matrix, respectively.
Changes to Carbonate/Phosphate, Phosphate/Amide I and Carbonate/Amide I ratios, which are
the strongest predictors of mechanical properties of the samples, are moderate. A small reduction of the inorganic(BAP,960 cm-1)/organic(Collagen, 2940 cm-1) ratio is indicative of reduced
skeletal fragility and decrease of the corresponding Young modulus.
Bone 1 EXP 1 Reference (vs) Aged
0 �0 0 7
0 �0 0 6
1 0 0 3
�S O 24 ��
Relative Intensity
0 �0 0 5
1 0 4 5
�NO �3�
Figure 2: Representative right angle Stokes-side Raman spectra of reference and aged samples.
Green curve: Unaged (reference) sample; red curve:
aged sample (Experiment 1 in Table I).
0 �0 0 4
0 �0 0 3
0 �0 0 2
0 �0 0 1
Aged
Reference
0 �0 0 0
5 0 0
1 0 0 0
1 5 0 0
2 0 0 0
2 5 0 0
3 0 0 0
3 5 0 0
Raman shift, cm-1
High Performance Liquid Chromatography in analysis of collagen and degradation
products
Amino acid composition provides a ‘fingerprint’ of the molecular integrity of the polypeptide
chains of collagen. Any changes in amino acid composition may reveal the presence of phenomena of oxidative or hydrolytic degradation of collagen in bones. Previous studies on parchment damage have showed that SOx and NOx gases interact with collagen molecules, which
contained some amino acids particularly sensitive to the action of these pollutants (Della Gatta
G., et al 2007). Certain aminoacids such as methionine, phenylalanine and tyrosine are known
to react with these gases, but sulfur oxides are interfering suppressing the nitration of aromatic
aminoacids.
Figure 3: HPLC chromatogram of bone aged under a SOx
(100ppm) and NOx (300ppm)
atmosphere. Sample was HClhydrolysed and derivatized with
FMOC (5mM).
314
ISBN: 978-960-98739-8-7
The amino acid composition and content in bone collagen of artificially aged samples in SOx
and NOx atmospheres was evaluated by calibrated amino acid analysis (typical chromatogram
shown in Figure 3). No significant changes in amino acid composition of collagen were observed between the reference and aged bones, with the exception of small decrease of methionine, tyrosine and phenylalanine contents. Finally, free amino acids were detected in aged
samples, mainly glycine, alanine, proline and hydroxyproline. Further investigations are under
way.
ELISA assay
The Enzyme Linked Immunosorbent Assay (ELIZA) investigation is focused on the development
of a reliable and reproducible immunoassay for the evaluation of collagen and keratin-based decay
of organic materials in cultural heritage. For this purpose, polypeptide models of collagen, [ProSer(OBzl)-Gly]n and keratin, [Pro-Cys(Acm)-Gly]n, and Ac-YRSGGGFGYRSGGGFGYRSβAla-NH2, were synthesized and used in immunization experiments. The obtained antibodies
were utilized in ELISA assays (absorption at 450nm) (Fotou et al 2012, Fotou et al 2014). According to our results, high recognition of the aged samples by the anti-polypeptide antibodies
indicates significant deterioration, while high recognition of the anti-collagen and anti-keratin
antibodies indicates lower deterioration.
Concluding Remarks
Artificially aged bone samples in atmospheric pollutants environment were analyzed with elemental and molecular analysis techniques. SEM-EDX microanalysis revealed sulfur incorportated in the case of aged samples. FTR analysis showed a small decrease in collagen content,
incorporation of nitrates within the bone matrix, and no detectable crystallinity changes in the
aged samples. The loss of molecular integrity was reflected through aminoacid analysis of the
organic matrix with HPLC, which showed small decrease of certain aminoacids in the bone
collagen, as well as through an ELIZA assay using anti-polypeptide antibodies.
Acknowledgements
The authors wish to thank Dr. L. Albu (ICCP – Romania) and Prof. D. Anglos and Dr. Z. Papliaka (IESL-FORTH) for providing access and assistance with their ATR-FTIR facilities.
References
1. Child, A.M., 1995, Microbial Taphonomy of Archaeological Bone. Studies in Conservation, 40, pp.19-30.
2. Collins M.J., Riley M.S., Child A.M. and Turner-Walker G., 1995, A Basic Mathematical Simulation of the
Chemical Degradation of Ancient Collagen. Journal of Archaeological Science, Vol. 22, N.2, March, p.175184
3. Della Gatta G., Badea E., Mašić A. and Ceccarelli R., Structural and thermal stability of collagen within
parchment: a mesoscopic and molecular approach [SOx and NOx ageing]. In: Improved damage assessment
of parchment (IDAP): Assessment, Data Collection and Sharing of Knowledge, 2007, Research Report No.18,
89-99
4. Dellaportas, P., Papageorgiou, E. & Panagiaris, G, 2014, Museum factors affecting the ageing process of
organic materials: review on experimental designs and the INVENVORG project as a pilot study. Heritage
Science, 2(1), 2.
5. Fotou, E., Moussis, V., Panou-Pomonis, E., Panagiaris, G., Sakarellos-Daitsiotis, M., 2014, “Keratin and
collagen polypeptide models as a means for assessing the deterioration of organic materials of cultural
heritage”, 33rd European Peptide Symposium, eds Naydenova, E., Pajpanova, T., Sofia, Bulgaria.
6. Fotou, E., Zevgiti, S., Panou-Pomonis, E., Panagiaris, G., Sakarellos-Daitsiotis, M., 2012, “Keratin and
collagen models as a tool in the deterioration of works of art”, 32nd European Peptide Symposium, eds
Kokotos, G., Constantinou-Kokotou, V., Matsoukas, J., Athens Greece 660-661.
7. Gonzalez, Lee Grant, Wess, Timothy James, 2013, The effects of hydration on the collagen and gelatine
phases within parchment artefacts, Heritage Science, 1:14
ISBN: 978-960-98739-8-7
315
8. Odlyha, M., Theodorakopoulos, C., De Groot, J., Bozec, L., Horton, M., 2007, ‘Thermoanalytical (macro
to nano-scale) techniques and non-invasive spectroscopic analysis for damage assessment of parchment’.
In: Improved Damage Assessement of Parchment (IDAP): Assessment, Data Collection and Sharing of
Knowledge. Research 2007, Report No.18, pp.73-87.
9. Smith C.I., C.M. Nielsen-Marsh, M.M.E. Jans and M.J. Collins, 2007, Bone diagenesis in the European
Holocene I: patterns and mechanisms, Journal of Archaeological Science, Volume 34, Issue 9, p.1485-1493.
10. Stiner M.C., Kuhn S.L., Weiner S. and Bar-Yosef O., 1995, Differential burning, recrystallization and
fragmentation of archaeological bone, Journal of Archaeological Science, Vol. 22, N.2, March, p.223-238.
11. Timmins P. A. and Wall J. C., Bone Water, Calcified Tissue Research, 1977, 23(1), 1-5
12. Weiner S., Microarchaeology, Cambridge University Press, Cambridge, 2010, 99-123, 282-297
316
ISBN: 978-960-98739-8-7
Development of a standard laser technology for cleaning evidence
of proteinaceous Cultural and Natural Heritage
Athanassia Papanikolaou1, Kristallia Melessanaki2, Alexandra
Alexandropoulou3, Konstantinos Giannaras3, Athanasios Karambotsos3,
Effrosyni Karantoni3, Ekaterini Malea3, Stavroula Rapti3, Nikolaos-Alexios
Stefanis3, Yorgos Facorellis3, Stamatis Boyatzis3, Paraskevi Pouli2 and
George Panagiaris3*
Department of Physics, University of Crete, P.O. Box 2208, 71003, Heraklion, Greece,
Institute of Electronic Structure and Laser (IESL), Foundation for Research and
Technology-Hellas (FORTH), P.O. Box 1385, 71110, Heraklion, Crete, Greece,
3
Technological Educational Institute of Athens, Department of Conservation
of Antiquities and Works of Art, Ag. Spyridonos str, 12243, Athens, Greece
* [email protected]
1
2
Keywords: laser, wool, parchment, iron oxides, fossilised bones, feather.
Abstract
Natural and cultural heritage collections demonstrate a unique and complex example of conservation problems. This study regards the development of a cleaning methodology for proteinaceous specimens of Cultural and Natural Heritage value using laser radiation. Laser cleaning
have been proved to have unique advantages and properties which contribute in a significant
way to cleaning procedures that require delicate and selective removal of unwanted material.
A series of laser cleaning tests were undertaken on real and artificial samples aiming to investigate the ultimate laser parameters for the cleaning of bone, parchment, wool and feathers. The
results of this study will be presented with emphasis to the development of customised laser
cleaning methodologies per case.
Introduction
Natural history collections comprise of objects with various materials, composition and conservation problems. Insoluble crusts that enclosure bone fossils, dust or soiling media trapped in
taxidermy bird feathers and staining of parchments and wool objects by iron oxidation products
constitute common conservation challenges demanding diverse cleaning approaches. Lasers
due to their unique properties play an important role and through the past 20 years they have
been evolved into unique analytical and conservation tools (Fotakis et al, 2006). Based on the
distinctive attributes of laser light, the developed analytical and cleaning methodologies show
unique advantages such as selectivity and precision while they ensure contactless and detailed
cleaning interventions without damage to the original surfaces. Towards this aim optimisation
of laser parameters to the materials involved, their morphology and the specific cleaning requirements have been investigated and will be presented herein.
Experimental part
Laser systems employed for the cleaning tests were a Q-Switched (QS) Nd:YAG nanosecond
laser and a fibre-coupled short free running (SFR) microsecond laser. Both systems emit in the
infrared region (IR) at 1064 nm. In addition the QS laser provides the opportunity of using the
ISBN: 978-960-98739-8-7
317
second and third harmonic wavelengths which correspond to visible (VIS) radiation of 532 nm
and (UV) radiation of 355 nm respectively.
Evaluation and analytical methodologies
The evaluation of the cleaning methodology is accomplished by microscopic methods (optical
microscopy (OM) and scanning electron microscopy coupled with elemental analysis (SEM/
EDAX)), as well as analytical techniques such as x-ray diffraction analysis (XRD), Fourier
transform infrared spectroscopy (FTIR) and shrinkage temperature measurement of collagen
by Micro Hot Table technique (MHT). These techniques are used in order to ensure that no
physical or chemical alteration is caused to the materials and for assessment of the applied laser
cleaning procedures. The comparative evaluation of the laser cleaning operations designated
the optimum cleaning parameters for each proteinaceous material.
More particularly, the initial determination of the cleaning methodology was first accomplished
by microscopic techniques such as optical microscopy and then by SEM/EDAX which provided
information about the chemical composition and any neo-formed phases of the surface as well
as to assess the cleaning method efficacy. Furthermore, the use of x-ray diffractometer (XRD)
enabled the confirmation that materials of similar or identical chemical composition had not
been altered by the laser beam. FTIR was employed to detect possible shifts in the absorption
bands of the samples after the laser treatment as well as to confirm that the corrosion products
had been removed. Finally, in the case of parchment samples, the MHT measured the collagen
thermal denaturation at fibre, fibrillar and molecular levels after laser cleaning tests.
Fossilised bones hidden inside insoluble crusts. Hippopotamus bone-fossil fragments embedded into hard insoluble calcareous matrix with quartz impurities have been used to test and
evaluate different laser cleaning methodologies. The specimens came from the entrance of a
cave in Kato Zakros in Rethumnon, Crete and are dated from the Upper Pleistocene period.
Irradiation tests with Q-Switched Nd:YAG nanosecond laser emitting at the first (1064nm),
second (532nm) and third (355nm) harmonic wavelengths could not provide sufficient results
(Landucci, 2000). However, fiber coupled SFR Nd:YAG microsecond laser (at 1064 nanometers) can efficiently remove the bulk of the calcareous matrix at high fluence values and wet
conditions without damaging the substrate or causing any discoloration effect.
Figure 1. Detail of an area on the fossilized bone before and after laser cleaning and their SEM-EDAX analysis
318
ISBN: 978-960-98739-8-7
Of particular interest were unusual formations (nanospheres) which appeared under certain,
over-cleaned, conditions. Their size ranged between 10 and100 nm and they were associated to
laser-induced thermal effects. Their elemental composition reflects the chemical composition
of the surrounding area. In any case, care must be taken in order to ensure that the chosen laser
cleaning optimal parameters will not result to such side-effects.
Figure 2. The observed nanospheres under SEM and their EDAX analysis in comparison to a nearby area.
Taxidermy bird feathers with dust. Taxidermy birds in exhibition often captivate dust and
soiling media deposits. Their cleaning using conventional cleaning methods, is a difficult process, resulting in damage to their fragile and delicate structure. Studies have already shown the
adequacy of laser radiation as a cleaning tool on dark colored feathers leaving in abeyance
the conduction of tests on feathers with other coloration (Dignard, 2005). In these terms, light
colored samples of contour and flight feathers originating from duck and flamingo were artificially covered with dust from the top show case of University of Athens Zoological Museum
and used for preliminary laser irradiation studies.
The system employed for these tests was a QS Nd:YAG laser emitting at 1064, 532 and 355
nanometers. The infrared radiation removed the dust and the soiling media from the feather but
left a yellow hue on its surface while visible radiation could not clear away the depositions. On
the other hand, UV radiation provided a reliable cleaning tool and removed the deposits from
the feather without destroying the brittle burbles and hooklets or alternating their color.
Figure 3. Duck feather after laser treatment with UV beam
Parchment stained with rust. Parchments of animal skin were used in the past as a material
for writing on and therefore consists the basic material of the majority of ancient manuscripts
preserved in our days. Objects of cultural heritage made of parchment can be susceptible to
staining by iron corrosion products such as rust. Artificially created parchment samples from
ISBN: 978-960-98739-8-7
319
goat skin were stained with iron oxidation products following two different methodologies: a)
by applying dust iron oxidation products and b) by deposition of aqueous solution of iron oxidation products on the sample’s surface.
Tests were performed with QS Nd:YAG laser employing the IR and VIS radiation (Puchinger,
2005). Our tests indicated that IR radiation could remove the corrosion products effectively
without any surface damage (as confirmed by OM) or physicochemical alterations (as shown
by elemental analysis EDAX and shrinkage temperature Ts).
Figure 4 Parchment samples stained by corrosion products (by dust- on the left, by solution - on the right) after
IR laser cleaning (F=1.9 J/cm2, λ=1064 nm, 10 pulses and F=1.4 J/cm2, λ=1064 nm, 10 pulses respectively)
Woolen fabric stained with rust. Ethnographic composite objects consisted of textile, other
materials and iron elements are stained with iron oxidation products when they are found in a
humid environment. The removal of such stains are a puzzle for conservators (Margariti, 2003),
since all the conventional conservation treatments involve immersion in aqueous solutions,
which are not appropriate to hydroscopic composite materials. Until now literature regarding
the removal of iron corrosion products from composite artifacts is rare and therefore the need
for finding a non destructive method for cleaning different substrates is demanding. Samples
of merino wool as a part of a composite ethnographic object were stained with iron oxidation
products. Tests were conducted with QS Nd:YAG laser with IR, UV and VIS beam. Infrared
and ultraviolet radiation could not remove the corrosion products while it destroyed (burned)
the reference samples. The best results were obtained with the beam of 532 nm.
Concluding Remarks
The above results indicate the operating parameters for each individual cleaning case, while
highlight the versatility of laser cleaning methodologies to the nature and demands of the involved materials. Further studies and optimization are necessary, while it is also important the
implementation of monitoring techniques, so as to ensure a safe intervention.
Acknowledgements
This study is part of the research project entitled “Development of a standard laser technology for cleaning evidence of proteinaceous Cultural and Natural Heritage”, which is operating
under the Action “Archimedes III” of the co-operational program “Education and Life Long
Learning”, co-funded by the Greek Ministry of Education and the EU. We also to acknowledge
Dr Ev. Kambouroglou, Head of Geology & Paleontology Dept., Ephorate of Paleoanthropolgy
& Speleology (Hellenic Ministry of Culture & Sports) for providing the research samples.
320
ISBN: 978-960-98739-8-7
References
1. Dignard, C., Lai, W.F., Binnie, N., Young, G., Abraham, M., Scheerer, S., 2005, Cleaning of soiled white
feathers using the Nd:YAG laser and traditional methods. In: Springer Proceedings in Physics, edited by Κ.
Dickmann, C. Fotakis, J.F. Asmus, Vol. 100, (Springer Berlin Heidelberg), p.227-235.
2. Fotakis C., D. Anglos, V. Zafiropulos, S. Georgiou, V. Tornari, 2006, Lasers in the Preservation of Cultural
Heritage; Principles and applications, eds. R.G.W. Brown, E.R. Pike (New York, USA: Taylor and Francis)
3. Landucci, F., Pini, R., Siano, S., Salimbeni, R., Pecchioni, E., 2000, Laser cleaning of fossil vertebrates: a
preliminary report, Journal of Cultural Heritage, 1 (supplement 1), S263– S267
4. Margariti C., 2003, The use of chelating agents in textile conservation: an investigation of the efficiency and
effects on the use of three chelating agents for the removal of copper and iron staining from cotton textiles.
Dust, Sweat and Tears: Recent Advances in Cleaning Techniques, UKIC Textile Section, AGM Spring Forum,
London, 7 April 2003, 28-38
5. Puchinger L., Pentzien S., Koter R., Kautek W., 2005, Chemistry of parchment-laser interaction. In: Springer
Proceedings in Physics, edited by Κ. Dickmann, C. Fotakis, J.F. Asmus, Vol. 100, (Springer Berlin Heidelberg),
pp 52-58.
ISBN: 978-960-98739-8-7
321
Removing iron corrosion products from museum artefacts:
Investigating the effectiveness of innovative green chelators
Stavroula Rapti1*, Shayne Rivers2 and Anastasia Pournou1
Dept of Conservation of Antiquities & Works of Art, TEI of Athens, Ag. Spiridonos &
Palikaridi str. Egaleo, Athens, Greece
2
West Dean, Chichester, West Sussex, PO18 0QZ, England
* Corresponding author: [email protected]
1
Keywords: iron oxides, chelators, museum artifacts, siderophores
Abstract
This work presents the results obtained from the investigation of an innovative cleaning method
for removing iron corrosion products from museum wooden artefacts.
The method examined, involved the use of siderophores, which are iron carriers produced in
fungi and bacteria for their development. Therefore, a green chelator (deferoxamine), was assessed in comparison with two conventional chelating agents (EDTA, DTPA) used in conservation, in order to prove its effectiveness in removing coloured iron oxides.
The obtained results showed for the first time, this compound found in nature in living organisms, is able to bind successfully to iron oxides found in museum objects. Siderophores appeared to be a promising alternative to common chelating agents with further potential that is
awaited to be investigated.
Introduction
Museum objects are often stained with iron corrosion products originating from various types
of iron fastenings used in their manufacturing. In elevated humidity levels, iron is oxidised to
ferrous ions and in presence of oxygen and water, the later form hydroxide. This is a white or
colourless material, though is readily oxidized to the coloured oxide hydroxide, (oxyhydroxide), that causes staining on any substrate found in close proximity to it (Rivers & Ummey
2003; Schalk 2006; Bacon 2007; Gill 2007).
Corrosion products can also promote further deterioration of hydrogen containing cellulosic
or proteinaceous substrates, as they can catalyse various oxidation processes. In combination
with deterioration of the surrounding materials, corrosion of metallic fastenings can result in
mechanical damage, which threatens the structural integrity of museum objects.
The removal of iron corrosion product is a fundamental step in artefact’s conservation; however, conventional cleaning treatments usually involve chelating agents which are problematic,
as they can be beneficial for one substrate but destructive for another. Furthermore for museum
objects made of wood waterborne treatments with chelators via immersion procedures are not
applicable.
This work is been undertaken under the framework of an ARCHIMEDES project which investigated innovative cleaning methods for the safe and effective removal of iron corrosion
products from museum composite artefacts. More specifically, it investigated, among others
conventional chelating agents, ligands used in Clinical Medicine, such as siderophores.
Siderophores are compounds found in nature in living organisms and bind to metallic trace
elements necessary for organisms’ development (Raymond et al. 1984; Neilands 1995; Schalk
2006; Mossialos & Amoutzias 2009).
322
ISBN: 978-960-98739-8-7
This work presents the investigation of the iron complexing capacity of one siderophore in
comparison with two conventional chelating agents.
Materials & methods
Deferoxamine1, EDTA & DTPA were the three chelators applied on wooden mock-ups stained
with iron corrosion products via cotton swab and CMC gel 4,5% w/v. Detailed parameters used
are shown in Table 1.
For evaluating the efficacy of solutions applied, colour were measured before and after cleaning
with a colorimeter Lovibond (model SP60 - RT Series).
Chelators
EDTA
EDTA
DTPA
DTPA
Deferoxamine
Experimental methodology
Concentrations pH
Reducing agent
2,5% w/v
6,2
SDT/ 5%w/v
2,5% w/v
6,2
2,5% w/v
6,8
SDT/ 5%w/v
2,5% w/v
6,8
-5
8,5
5χ10 M
Application method
cotton swab
CMC gel 4,5% w/v
Table 1. Complexion parameters and chelators used
Iron corrosion products appeared to be effectively removed from wood mock-ups with Deferoxamine. Colorimetry showed that Deferoxamine was more effective than both chelators applied (DTPA, EDTA) (fig. 1).
Moreover it appeared that both conventional chelators were more effective when applied in
combination with SDT. However, residues of Na & S ions due to SDT use, could promote
future deterioration of substrates. For that reason, the use natural chelators’ application is considered more appropriate
Furthermore, based on the results obtained it is showed that the application of the chelators’
solutions with cotton swab was more effective than CMC gels.
1
due to the high cost of pure deferoxamine, a drug manufactured by NOVARTIS, " Desferal ", was used
instead that has the same chemical constituents. Desferal is an injectable drugs used in clinical medicine for
patients suffering from hemochromatosis.
ISBN: 978-960-98739-8-7
323
Figure 1: Application of chelators on wooden mock-ups. a, b, c, d, e: before application of chelators, f, g, h, i,
j: during application with cotton swab and gels and k, l, m, n, o: after cleaning (k: EDTA-SDT cotton swab, l:
EDTA-SDT-CMC gel, m: DTPA-SDT-CMC gel, n: DTPA-CMC gel, o: deferoxamine gel)
Conclusions
The most effective chelator applied on wooden mock-ups showed to be deferoxamine followed
by EDTA with SDT. For the two conventional chelators examined, liquid application was found
to be more effective than gel application. However, deferoxamine was found effective with both
types of applications.
Finally it appears that deferoxamine doesn’t have any negative impact on the wood surface via
macroscopic and microscopic observation; however, chemical analysis on substrate before and
after application is needed in order to confirm this result.
References
1.
Bacon L., 2007, X-radiography of ethnographic objects at the Horniman Museum. In: S. O’ Connor and M. Brooks eds.
X-Radiography of textiles, dress and related objects (Oxford: Butterworth-Heinemann).
2. Gill K., 2007, Evaluating X-Radiography as a tool for examining upholstered furniture. In: S. O’ Connor and M. Brooks
eds. X-Radiography of textiles, dress and related objects (Oxford: Butterworth-Heinemann).
3. Mossialos D. & Amoutzias G.D., 2009, Role of siderophores in cystic fibrosis pathogenesis: Foes or friends? International
journal of Medical Microbiology, (299), 87-98.
4. Neilands J.B., 1995, Siderophores: structure and function of microbial iron transport compounds. Journal of Biological
Chemistry, (270), 26723–26726.
5. Raymond K.N., Miiller G. & Matzanke B.F., 1984, Complexation of iron by siderophores. A review of their
solution and structural chemistry and biological function. [online] Available at: http://www.isolab.uni-luebeck.
de/paper/Top_Curr_Chem_1984_123_49-102.pdf [accessed 12 December 2011]
6. Rivers S. & Ummey N., 2003, Conservation of furniture (Oxford: Butterworths-Heinemann).
7. Schalk I. J., 2006, New Insights on Iron Acquisition Mechanisms in Pathogenic Pseudomonas. In: Juan-Luis
Ramos & Roger C. Levesque eds. Pseudomonas. Molecular Biology of Emerging Issues. Volume 4, Ch. 1
(Netherlands: Springer US), pp.1-34.
324
ISBN: 978-960-98739-8-7
On the role and significance of cultural heritage for a positive
national identity
Some educational remarks on museum education
Wafaa El Saddik
Former General Director of the Egyptian Museum Cairo
President of Children Alliance of Tradition and Social Engagement - C.A.T.S.*
The populations of Egypt and other developing countries are currently involved in a struggle to
overcome the problems of everyday life -money, food, jobs, health and increasingly security.
Hardly any time or energy remains for them to concentrate on their cultural heritage, and that
includes visits to their local museums and archaeological sites. The danger looms large, then the
nations could lose awareness of their own cultural heritage.
It is our task to re-educate the people as accurately as possible on their own cultural history. We
hope in Egypt to found a sort of museo-pedagogical center or museum education department
in every city and town, in order to aid the population in enhancing its national identity, and
broadening its horizons.
Museum education has two components: The matter of the museum and the task of education.
Therefore, museum education is not only an affair of scientists and museum people. Also the
perspective of the educators has to be considered.
In modern time, museums have become a relatively young expression of human life. As such
they comprise two aspects. They are a collection of a certain area to document something of
cultural importance. The second aspect is that museums are collections for the public; this is one
of the results of enlightenment: The public, “the people”, should possess such a collection.
These two elements: collection of an area of cultural value as well as being offered to the public,
are elements of the task of a museum.
One might say the collection, conservation and exhibition of artefacts is the museum’s primary
task. But in fact all museums are first of all learning places. Museums offer different types of
learning opportunities. Visitors who go through an exhibition may experience some individual
learning, but he needs also to learn more and this through the formal services museums offer
such as guided tours, gallery talks, lectures and learning by doing activities. Such programs
help museums more attractive and interactive. That why, museums are becoming centers of
learning. Many educational institutions use museums as an additional teaching resource. Other
initiatives try to take the museum out into community. The aim of such activities is to make museums more accessible to various parts of the population. Cultural collection and public are to
come together, and even more than that: they have to be mediated to each other. The mediation
is necessary because the public is not an expert of the collection like the owner has been. The
task of a museum is the mediation of a cultural collection and the public. This is the educational
task.
The definition of museum education as the task to mediate between collection and the public
may sound very abstract. One might elaborate this through a famous word by Goethe (the German Poet):
“What you inherit from your father, acquire it to possess it”! Superficially taken, this word does
not make sense. For, why shall I acquire something which I have inherited already? It already
belongs to me, so why to try to possess it again? For sure, Goethe knew this, too. Therefore we
have to look more carefully for the meaning of his word. The three main statements of Goethe’s
ISBN: 978-960-98739-8-7
325
word are: to inherit, to acquire and to possess. When we relate the heritage to our topic, i.e. to
museums, then the heritage comprises the goods and traditions which stem from ancient times.
These traditions and goods of culture and civilization are supposed to be present in a museum.
Museum document the heritage, so far, the heritage is present. But Goethe demands from us
that we acquire the heritage. Obviously, for Goethe, the pure presence is not enough. The presence of heritage is not sufficient; it needs the active participation, the activity of the individual.
On the other hand, a museum has to make the heritage available for the public; the average of
citizen needs a good presentation of the exhibits through a skilful selection of exhibits by an
expert and a calm presentation which invites to meditation.
The active participation in the heritage by citizen must be supported by the museum. However,
acquiring the heritage must not be mixed up with the information. This is a fatal misunderstanding of our schools. They feed the children with information, with superficial data about – for
example – history. The result is that those children later as adults will be fed up of history, they
will be bored and will even hate it. Therefore, acquiring the heritage must be more than information; it requires encounter with it.
Museums have to arrange the encounter; they have to give the chance for it and encourage it.
However, they cannot make the encounter happen without the active participation, which is
the key to real encounter. This brings us back to the individual who should actively address his
heritage. In which way has he to be active? There must be, first of all, a personal question, a
curiosity. We can call it a motivation or an interest.
Is an interest sufficient? We know that a superficial interest may quickly disappear; therefore,
it must be serious. The more serious and deeper the motivation is, the more powerful and sustainable it will be in order to bring individual and heritage together. The more we are able and
ready to relate our heritage to our present or even personal situation, the deeper and the more
existential our questions are, the more we are ready to encounter a historic figure or event etc.
An existential acquirement is not only based on motivation, but also on the involvement of the
senses. Also “information” needs eyes and ears, however, in a very superficial way, just for the
transfer of abstract knowledge. A real involvement of our eyes, would mean, for example, to
watch a painting carefully and drew it. Who has drawn it has a different kind of “knowledge”
than the one who has read only the description. The more senses are actively involved the more
existential is the encounter with a subject matter. Drawing is only one simple possibility of
involving the senses. It can be also painting, forming in clay, in paper, other materials, describing it in words and by this reflecting on it; it can be a role or a dramatic performance. Thus, in
manifold forms our senses can be activated in order to take the subject matter into our hands
and not only in our Brain, i.e. learning by doing.
So far, we can reformulate Goethe’s word in a new way:
What is presented to you as your heritage, actively acquire and encounter it with all your senses
in order to possess it. One important aspect is related to this: Without personal relationship to
an object on sensual encounter there is no real love for it.
Because of the importance of museum education, the UNESCO designed a program, which
offers an ideal setting for site-related museums to play a pivotal role as mediator of peace and
intercultural understanding, provider of quality educational contents for local and international
communities, as well as agent of development. The UNESCO project was designed as a pilot
project to accompany the process of revitalizing the World Heritage-related site museums and
interpretation centers dedicated to archaeological sites. The program is aimed primarily for the
development of professional skills of staff from museums, not only curators and conservators,
but also archaeologists and managers of the World Heritage sites.
The participants are to re-scope the mission of the World Heritage-related site museums in the
contemporary world, by means of collective reflection and practical exercise for the enhanced
326
ISBN: 978-960-98739-8-7
and holistic interpretation of collections, collection management, development of participative
approach for local communities, and museum visitor studies. The aim of museum education’s
activities is to make museums more accessible to various parts of the population and more responsive to the visitors’ needs and interests.
Learning around museum objects can be an effective way to provide education to people who
do not have easy access to the written word or do not speak the dominant language. The current relationship between Intangible Cultural Heritage, museums and educational organizations that together play a role in safeguarding Intangible Cultural Heritage, is one that is
still evolving, while museums and other public institutions have keenly promoted UNESCO’s
Convention, 2003 for the Safeguarding of the Intangible Cultural Heritage.
Our goal is to help the Children and also adults in looking at original exhibits and monuments
to have better understanding of their history and their environment. In taking into consideration
the interests and needs of our youth; our goal of making the ancient cultural heritage tangible
for them must provide programs with a clear overall view of the materials used and the roles
people played in the past. This would mean that we must make ancient cultures “come alive”.
We must awaken new understanding and excitement in the young people of our countries, in
order to enhance their cultural identity and increase their awareness of the world they live in.
Awareness of the role and significance of cultural heritage makes a positive future for children
and their family a reality.
Children’s workshop at the Egyptian Museum Cairo
* C.A.T.S. is a nonprofit organization established in 1999 in Cologne Germany. Its goal is to develop programs in
various places in Egypt to acquaint children and their parents with Egypt’s cultural heritage - first and foremost,
through direct, first-hand contact with the remains of Egypt’s ancient, modern and contemporary cultures. By
learning to look at and understand what surrounds them, children will come to appreciate their history, tradition
and environment.
ISBN: 978-960-98739-8-7
327
Preliminary investigation of L-cysteine and mature tobacco as
corrosion inhibitors for marine composite artefacts containing
copper or iron alloys.
V, Argyropoulos1*, S.C. Boyatzis1, M. Giannoulaki1, A. Malea1, A.
Pournou1, S. Rapti1, A. Zacharopoulou2, and Elodie Guilminot3
& Works of Art, Ag. Spyridonos, Egaleo 12210, Greece, [email protected], 30-2105385459
2
Department of Materials Science and Engineering, School of Chemical Engineering,
National Technical University of Athens, 9 Heroon Polytechniou, Zografou Athens, Greece
3
GPLA Arc’Antique, 26 rue de la Haute Forêt, 44300 Nantes, France
1
Keywords: corrosion inhibitors, L-cysteine, mature tobacco, PEG400, composite
artefacts
Abstract
The effectiveness of cysteine and mature tobacco as non-toxic corrosion inhibitors for copper
and iron alloy metals respectively in PEG400 solutions was investigated by electrochemical
studies for the treatment of marine composite artefacts. Potentiodynamic polarization measurements, immersing and monitoring test samples in solutions for one month, as well as testing
on real artefacts to study the effects of the corrosion inhibitors in PEG400 solutions was carried out. Potentiodynamic polarization measurements showed that the presence of cysteine
in PEG400 solutions acts as an inhibitor affecting mainly the cathodic process. However, the
mature tobacco was found not to act as an inhibitor in PEG400 solutions since it increased the
corrosion current when compared to PEG400 solutions alone. The monitoring of treatment solutions with test samples or real artefacts found that the cysteine caused a greying of the patina
copper alloy surfaces but minimized the changes in colour for the associated organic materials
in contrast to PEG400 solutions without inhibitor.
Introduction
In conservation, marine composite artefacts are the most difficult to stabilize and treat especially when an object is made up of different types of organic materials (wood, textile, and
leather) and in some cases more than one type of metal (iron, copper). Corrosion inhibitors are
needed to prevent the corrosion of the remaining metal(s) or the dissolution of metal oxides at
near-neutral pHs both during pretreatment storage and during treatment in Polyethylene Glycol
(PEG) solutions. While past research has focused on the inhibition properties of HostacorIT,
phosphates, benzotriazole, and even PEG itself on wet composite artefacts containing iron or
copper alloys [Argyropoulos et al., 1999; Guilminot et al. 2000; Guilminot et al. 2002], the
effectiveness of these corrosion inhibitors in real situations varies depending on many factors.
The duration of storage or treatment solutions, the presence of certain ions (chloride, metal etc)
at the corrosion/metal interface, as well as the types of corrosion products can result in these
corrosion inhibitors as being ineffective.
The Metals Conservation Lab at the TEI of Athens co-ordinated EU FP6 EC Research project,
PROMET between 2004 to 2008 to test new green corrosion inhibitors. It has continued its
efforts with the Archimedes III project to test mature tobacco and L-cysteine for slowing down
corrosion of iron and copper alloys respectively during stabilization treatments of marine com-
328
ISBN: 978-960-98739-8-7
posite artefacts. Tobacco and its extracts as well as L- cysteine (an amino acid) have been tested
and found by industry to be effective as a corrosion inhibitor for various metals in different
media [Loto et al., 2011; Sahaya et al., 2014]. Furthermore, L-cysteine has been found by industry to inhibit enzymatic browning [Pongsakul et al., 2006] or for its antibacterial properties
[Caldeiraa et al., 2013] of various organic materials. This paper presents the research carried
out on the application of mature tobacco and L-cysteine as a corrosion inhibitor for iron and
copper alloys respectively in water and PEG400 solutions.
Scientific Approach
Electrochemical measurements were carried out for mature tobacco and L-cysteine in deionized
water and 20% (v/v) PEG400 solutions for corrosion-free wrought iron and brass respectively.
For the potentiodynamic experiments carried out at Arc’Antique, France, a Faraday cell was
used with a rotating disk electrode as the working electrode (i.e., wrought iron or brass alloy),
with and without 1% (w/v) mature tobacco and 1%(w/v) cysteine respectively in 20%(v/v)
PEG400 solutions in deionized water. Experiments were controlled at a constant temperature
32 ± 1oC, in naturally aerated solutions. Cathodic and anodic polarization curves were repeated
three times for each inhibitor to check the results. Also, treatments were carried out for copper
and iron corroded samples with and without leather, textile, and wood in 1% (w/v) tobacco
and L-cysteine in 20% (v/v) PEG400 solutions in deionized water for one month; these were
monitored by measuring pH, Ecorr of the metal, as well as documenting the colour changes to
the metal and organic samples before and after treatment using colour chromameter. L-cysteine
solutions in deionized water 1% (w/v), or in 5% (v/v) PEG400 were accordingly tested on
real marine composite artefacts containing either copper alloys. XRD, SEM-EDX, and FTIR
analyses were carried out on the real composite artefacts before and after treatment with Lcysteine.
The results for anodic polarization showed that the presence of cysteine decreases the corrosion
rate (i.e. icorr decreases) and forms a passive film for the brass sample (see Figure 1), whereas
the presence of mature tobacco increased the corrosion rate (i.e. icorr increased) for the wrought
iron sample in 20%(w/v) PEG400 solutions. Also, the open circuit potentials or Ecorr prior to
each experiment with cysteine shifts to more negative values than without. For mature tobacco,
Ecorr values shift to more positive values than without. Although the shape of the curves for
the anodic and cathodic polarization curves were similar when the experiments were repeated
for 1%(w/v) cysteine in 20%(v/v) PEG400 solutions, they were not reproducible whereas in the
case of mature tobacco they were reproducible. Our results indicate that cysteine is acting in
the PEG400 solutions as a cathodic inhibitor in near-neutral pHs. Past studies have confirmed
the cathodic behavior of cysteine in such neutral pHs in aqueous solutions, where the zwitter
ion structure of cysteine is dominant in the range between 1.91 and 8.16, and this ion structure
is highly attracted to cathodic sites on the metal surface [Ismail, 2007]. The lack of reproducibility of the curves for cysteine may be due to the varying amounts of dissolved oxygen in the
solutions for each experiment. The transfer of oxygen from the bulk solution to the copper/
solution interface plays a critical role at the rate of oxygen reduction. The adsorbed layer of
cysteine or its complex behaves as a cathodic inhibitor to Cu corrosion by retarding the transfer
of O2 to the cathodic sites of the Cu surface [Ismail, 2007].
The testing of the corrosion inhibitors for copper and iron corroded samples with and without
leather, textile, and wood in 1% L-cysteine and 1% (w/v) tobacco in 20% (v/v) PEG400 solutions in deionized water for one month showed interesting but varying results. For conservation
treatments, apart from inhibiting the corrosion of the metal, it is important that the inhibitors
ISBN: 978-960-98739-8-7
329
prevent the dissolution of metal oxides so as to preserve the patina or original surface. PEG400
solutions alone are known to dissolve copper patinas or iron oxides [Guilminot et al., 2000].
With composite artefacts, such dissolution of metal oxides during PEG treatments also discolours or contaminates the associated organic materials. With the wood and textile test samples associated to either iron or copper it was difficult to see visible differences with or without
inhibitors. However, our results found that both corrosion inhibitors minimized darkening of
the leather samples in PEG400 solutions associated to iron or copper than without inhibitors.
Furthermore, cysteine did cause a greying of the copper coupons, due to the formation of copper sulphide on the surface, which can be disfiguring to the patina appearance. Finally, the Ecorr
measurements taken for one month (usually on a daily basis) to monitor the metals found that
for copper in cysteine, the Ecorr shifted to more negative values than those without inhibitor,
which was similar to measurements before the potentiodynamic experiments. However, for the
mature tobacco the iron samples also shifted to more negative values compared to those without
inhibitor, differing from the measurements before potentiodyanmic experiments. This behavior
may indicate that with time, mature tobacco begins to passivate the corroded iron.
Finally, 1% (w/v) L-cysteine was tested on real marine copper alloys, and one containing textile either brushed on the surface with deionized water or for four days in 5%(w/v) in PEG400
solution. After drying, the objects were placed in a chamber with relative humidity above 70%
to test for the appearance of active corrosion. The results found no appearance of active corrosion after using the cysteine. However, in the case of the composite object made of brass with
remnants of textile with cysteine and PEG400, a yellow precipitate formed on the areas of the
object containing textile pseudomorphs. Although, the precipitate could be easily washed off,
it did seem to dissolve the concreted surface.
Figure 1 Polished Brass in 20% w/v PEG400 in deionized water with and without Cysteine (1%w/v)
330
ISBN: 978-960-98739-8-7
Conclusions
The results were found to be variable but interesting. L-cysteine was found to inhibit corrosion of copper alloys both in deionized water or PEG400 solutions with or without corrosion
products, but not mature tobacco for corrosion-free iron. On real marine composite artefacts,
L-cysteine was found to be effective as a corrosion inhibitor for copper alloys, and was found
not to affect the appearance of the organic samples. In some cases it reduced the darkening of
the leather when compared to samples in PEG400 solutions without the inhibitor. The varying
effects of L-cysteine on different types of corrosion products at neutral pH have been previously investigated and largely depends on their crystallinity and the types of metal ions that can
catalyze the oxidation of L-cysteine to cystine [Cornell, 1990]. The resulting metal ion-cystine
complex either remains adsorbed onto the metal surface (corrosion inhibition) or results in
desorption into solution of the organic oxidation products (corrosion stimulation). As a result,
the application of amino acids as corrosion inhibitors for marine composite artefacts can be
complex, but offers new potential benefits in green conservation of metal-organic artefacts. The
results for mature tobacco were disappointing, and was abandoned for further application of
treating marine composite artefacts in PEG400 solutions.
Acknowledgements
The on-going research testing new green inhibitors has been co-financed by the European Union (European Social Fund – ESF) and Greek national funds through the Operational Program
“Education and Lifelong Learning” of the National Strategic Reference Framework (NSRF)
- Research Funding Program: ARCHIMEDES III. Investing in knowledge society through the
European Social Fund.
References
1. Argyropoulos, V., Rameau, J.J., Dalard, F. , Degrigny, C., Testing Hostacor IT as a corrosion inhibitor for iron
in polyethylene glycol solutions, Studies in Conservation, 49–57, 1999
2. Caldeiraa, E., Piskinb, E., Granadeiroc, L., Silva c, F., Gouveiaa, I.C., Biofunctionalization of cellulosic
fibres with l-cysteine: Assessment of antibacterial properties and mechanism of action against Staphylococcus
aureus and Klebsiella pneumoniae, Journal of Biotechnology 168, 426-435, 2013.
3. Cornell, R.M., Giovanoli, I.R., Schneider, W., Effect of Cysteine and Manganese on the crystallization of
noncrystalline iron(III)hydroxide at pH 8, Clays and Clay Materials, 38 (1), 21-28, 1990.
4. Guilminot E., Dalard F., Degrigny C., Mechanism of iron corrosion in water-polyethylene glycol (PEG400)
mixtures, Corrosion Science, 44, 2199-2208, 2002
5. E. Guilminot, F. Dalard, and C. Degrigny, Effect of Phosphate Concentration on the Corrosion of Iron in
Water—Polyethylene Glycol Mixtures, Corrosion, 61 (5),
6. Guilminot, E., Rameau, J-J., Dalard, F., Degrigny, C., and Hiron, X. Benzotriazole as inhibitor for copper
with and without corrosion products in aqueous polyethylene glycol, Journal of Applied Electrochemistry, 30
(1), 21-28, 2000.
7. Ismail, K.M., Evaluation of cysteine as environmentally friendly corrosion inhibitor for copper in neutral and
acidic chloride solutions, Electrochimica Acta, 52, 7811-7819, 2007.
8. Loto, C.A., Loto. R.T., and A.P.I. Popoola A.P.I., Electrode Potential Monitoring of Effect of Plants Extracts
Addition on the Electrochemical Corrosion Behaviour of Mild Steel Reinforcement in Concrete, Int. J.
Electrochem. Sci., 6, 3452 – 3465, 2011.
9. Pongsakul, N., Leelasart, B., and Rakariyatham, N., Effect of L-cysteine, Potassium Metabisulfite, Ascorbic
Acid and Citric Acid on Inhibition of Enzymatic Browning, Longan Chiang Mai J. Sci.; 33,(1), 137-141,
2006. www.science.cmu.ac.th/journal-science/josci.html
10. Sahaya Raja, A., Venkatesan, R., Sonisheeba, R., Thomas Paul raj, J., Sivakumar, S., Angel, P., Sathiyabama,
J., Corrosion Inhibition by Cysteine - An Over View, International Journal of Advanced Research in Chemical
Science, 1, 1, 101-109, March 2014.
ISBN: 978-960-98739-8-7
331
Experimental design of natural and accelerated bone and wood
ageing (Project INVENVORG)
Y. Facorellis1*, A. Pournou1, J. Richter2, E. Karantoni1, C. Vossou3, N.
Androutsopoulos1 and G. Panagiaris1
Department of Antiquities and Works of Art Conservation, Faculty of Fine Arts and Design,
Technological Educational Institute of Athens, Aghiou Spyridonos, 12210 Egaleo - Athens,
Greece, *[email protected]
2
Danish Royal Academy of Fine Arts (KADK), School of Conservation, Object Conservation
Laboratory, Philip de Langes Allé 10, 1435 København K, Copenhagen, Denmark
3
Department of Electronics, Technological Educational Institute of Piraeus, 250 Thivon &
P. Ralli Str., 12244 Egaleo - Athens, Greece
1
Keywords: Experimental design; Museum environment; Bone; Wood; Accelerated
ageing
Abstract
This paper presents the experimental design for natural and accelerated ageing of bone and
wood samples found in museum conditions that was conceived as part of the INVENVORG
(Thales Research Funding Program – NRSF) investigating the effects of the environmental factors on natural organic materials.
Introduction
When materials are exposed to the atmosphere in a museum environment, they are subjected
to damage not only from exposure to light, moisture, and heat, but also from interactions with
gaseous pollutants (such as SO2 and NOx). Damage occurs due to dry and/or wet deposition of
the pollutants onto the surface of a material and subsequent formation of an electrolytic solution
in water present on the surface. The materials’ damage may be physicochemical, potentially affecting materials’ durability, or may be purely aesthetic, affecting only their appearance.
In the case of bone, the influence of the atmospheric pollutants has not been studied yet. Βone is
a „hard” proteinaceous material, while wood is a lignocellulosic one. In both cases inadequate
environmental conditions can cause physical, chemical and biological deterioration, depending
on the environment in which they are located.
Although, there have been many studies conducted regarding the aging and thermal effects on
bone, there has been no focused research approach to documenting, through its results, the degradation mechanisms of bones in museum environment (Cronyn, 1990; Weintraub and Wolf,
1995; Hedges, 2002).
In the case of wood, the main abiotic factors in a museum environment affecting wooden artifacts preservation are heat, radiation, relative humidity, and particulate and gaseous pollution
(Rivers and Umney, 2005; George et al., 2005; Yidiz et al., 2006). Literature on the effect of
gaseous pollution on wood and its synergistic action with other environmental factors is scare.
Gaseous pollutants, such as O3, NO2 and SO2, have been studied mostly for surface treated
wood (Grontoft & Raychaudhuri, 2004) or for wood coatings and it has showed that pollutants
surface deposition velocities are highly dependent to relative humidity (RH).
332
ISBN: 978-960-98739-8-7
Experimental Design - Results and discussion
The investigation of the degradation mechanisms of the materials constituting the cultural and
natural heritage, despite the progress made, has a long way still to go. The most common
way followed for the determination of these mechanisms is the application of physicochemical
methods after accelerated ageing. The disadvantage of this practice is the inability to control the
effect of the combination of degradation factors in real time and in real conditions as it happens
in burial environment, in museum environment or in open-air conditions. This problem is getting even worse in the case of organic materials that are characterized by a variety in structure
and composition, and that are extensively sensitive in many environmental factors, such as bone
and wood. The need for understanding the decay process resulted to the standardization of diagnostic techniques and of the accelerated ageing protocols (ASTM, 2010, 2011, 2013). Although
this standardization, it is no coincidence that in recent years it has been noticed a remarkable
research activity towards the development of alternative decay determination methods in organic materials based on statistical elaboration of analytical data obtained experimentally. The
application of statistics in the cultural heritage area is very limited (Dellaportas et al., 2014).
According to the literature cited above, the main abiotic factors in a museum environment affecting bone and wood material preservation are i) temperature, ii) relative humidity (RH) and
iii) air pollution (NO2 and SO2).
The desirable levels, in terms of human comfort, of these factors in a museum are in the intervals of 15-20oC for the temperature and 45-55% for the relative humidity (Ashrae, 2011).
As far as the pollutants, their concentrations are usually those met in the area of the museums.
According to the standards in Greece the level of alarm for the concentrations of these two
pollutants are 200 μg/m3 (0.10 ppm) and 350 μg/m3 (0.18 ppm) for NO2 and SO2, respectively
(ΦΕΚ 125/5.6.02). The levels of the aforementioned factors that do not influence the artifacts
remain to be concluded.
In general, during artificial or accelerated ageing extreme environmental conditions are used,
in order to simulate the natural ageing process in shorter periods of time (Porck, 2000). The
purpose of this experiment is to produce artificially aged samples using specific environmental conditions (temperature, relative humidity and air pollutants) that simulate naturally aged
samples, so then to assess the impact of specific environmental conditions on the physical and
structural integrity of the bone material.
The experimental design has taken into account the ageing of both the organic part of the bone
and wood, as well the inorganic part of the bone. Thus, for the preparation of the accelerated
aged samples, the following conditions (factors and levels) were used (Table 1).
1.
2.
3.
4.
Factors
Relative Humidity (%)
Gaseous pollutants concentration (ppmDay)
Duration (Days)
Levels
45, 70
(SO2)
100, 300*
(NO2)
100, 300*
14, 28
Table 1. Conditions used for the preparation of aged samples. *Following the PROPAINT Dose concentrations
(Dahlin, 2010) for these pollutants and magnifying the alarm levels for Greece by 2000.
Also, the order of pollutants is evaluated, as a fifth factor, in order to decide on whether it is
important or not. Consequently, there are five factors with two levels each (low - high) that were
investigated.
ISBN: 978-960-98739-8-7
333
The planning of this study includes two types of exposure of these samples. The first group of
samples was left exposed for 2 years to closely monitored environments for natural ageing,
inside and outside the showcases of the following museums: 1) the Neofytos Doukas Library at
Ano Pedina Village, Zagori, Epirus, 2) the Criminology Museum in Athens and 3) the Historic
and Folklore Museum in Heraklion Crete.
For the second group the fractional factorial experimental design (Box et al., 2005) was used
and sixteen accelerated ageing cycles were performed, as shown in Table 2, in order to emulate
ageing in a controlled manner. The accelerated ageing chamber used is manufactured and located in the Department of Chemical Engineering of the University of Patras.
Experiment
1
14
12
15
11
6
8
9
7
13
5
16
3
10
2
4
Bone samples
Wood samples
Duration (D)
Relative Humidity,
% (RH)
Concentration SOx
(ppm)
Concentration NOx
(ppm)
Contaminant
Starting Date
Ending Date
Temperature (oC)
3
4
14
3
4
14
3
4
14
3
4
14
3
4
28
3
4
28
3
4
28
3
4
28
3
4
14
3
4
14
3
4
14
3
4
14
3
4
28
3
4
28
3
4
28
3
4
28
45
45
45
45
45
45
45
45
70
70
70
70
70
70
70
70
300
300
100
100
100
100
300
300
100
300
300
100
300
300
100
100
100
300
300
100
100
300
300
100
100
100
300
300
100
300
300
100
N/S
1
15
25
S/N
16
30
25
N/S
31
45
25
S/N
46
60
25
N/S
61
90
25
S/N
91
120
25
N/S
121
150
25
S/N
151
180
25
N/S
181
195
25
S/N
196
210
25
N/S
211
225
25
S/N
226
240
25
N/S
241
270
25
S/N
271
300
25
N/S
301
330
25
S/N
331
360
25
Table 2. Experimental design for bone and wood accelerated ageing.
Property
Physicochemical alterations
of collagen
Physicochemical alterations
of hydroxyapatite
Evaluation of the microstructure
Alteration of mechanical
properties
Method
Estimated trials number
according to the financial
availability
HPLC
TEI-A
40-50 samples
FTIR
TEI-A
90 samples
ELISA
UoI
90 samples
SEM/EDAX
TEI-A
40-50 samples
XRD-XRF
TEI-A
40-50 samples
CT - Scanning
NTUA
Three-point bending
NTUA
Janka Hardness test
NTUA
UoP
Degree of pollutants penetration
Chemical mapping RAMAN
Examination of materials
morphology
Visual observation
SEM
Optical microscopy
Examination of optical
properties (surface gloss
and color)
FE modeling
Institution of implementation
Three angle gloss meter
CIE L*a*b* colour
Computational mechanical analysis
TEI-A
TEI-A
TEI-A
TEI-A
TEI-A
NTUA
90 samples
90 samples
90 samples
90 samples
90 samples
90 samples
90 samples
90 samples
Table 3. Properties, physicochemical methods and estimated number of trials, which are envisaged by the experimental design to be performed on the bone and wood samples.
As bone samples metapodials of modern deers (roe deers - Capreolus) from Denmark were
used. They were chosen because the bones of this animal have about the same length and thickness, and so similar sized samples can be generated.
334
ISBN: 978-960-98739-8-7
As wood samples pieces of a light color softwood (Pinus spp) and a dark color hardwood were
used (Quercus spp). In this way it was possible to study the effect of environmental factors on
different species of wood, and thus to link to their different anatomical, physical and chemical
properties.
According to the experimental design, the properties showed in Table 3 will be determined. The
physicochemical and mechanical techniques will be applied by highly specialized personnel in
appropriately equipped laboratories Their results will be correlated with accelerated ageing and
natural ageing processes in museum environments using factor and regression analysis, anova,
t-test or corresponding non-parametric tests (Dellaportas et al., 2014).
Conclusions
In order to be able to evaluate the results of any reliable research project, this should be clear
and repeatable. This distinguishes science from pseudoscience.
The experimental design should clearly define the research objectives in relation to the state of
the art, the expected results and the methodology according to the available human capital and
scientific resources.
The research design of INVENVORG project for the natural and accelerated bone and wood
ageing took into consideration all the aforementioned conditions and it is characterized by the
following innovations:
natural and accelerated ageing is combined and compared for the first time internationally.
the statisticians participate decidedly from the start to the design of the research and
the research implementation is carried out by a real interdisciplinary team (material scientists,
conservators, environmental engineers and museologists) from the academic community.
Acknowledgements
This research has been co-financed by the European Union (European Social Fund - ESF) and
Reinforcement of the interdisciplinary and/or inter-institutional research and innovation with
the possibility of attracting high standard researchers from abroad through the implementation
of basic and applied excellence research.
References
1. ASTM E313 - 10: Standard practice for calculating yellowness and whiteness indices from instrumentally
measured color coordinates. ASTM, Book Standards 2010, 6:6. doi:10.1520/E0313-10.
2. ASTM D2244 - 11: Standard practice for calculation of color tolerances and color differences from
instrumentally measured color coordinates. ASTM, Book Standards 2011, 6:12. doi:10.1520/D2244-11.
3. ASTM D198 - 13: Standard testmethods of static tests of lumber in structural sizes. ASTM, Book Standards
2013, 4:27. doi:10.1520/D0198.
4. Box, G. E., Hunter, J. S., Hunter,W. G., 2005, Statistics for Experimenters: Design, Innovation, and Discovery,
2nd Edition. Wiley. ISBN 0-471-71813-0.
5. Cronyn, J. M., 1990, The elements of archaeological conservation. London: Routledge.
6. Dahlin, E., ed., 2010, PROPAINT- Improved Protection of Paintings during Exhibition, Storage and Transit.
George B, Suttie E, Merlin A, Deglise X., 2005, Photodegradation and photostabilisation of wood the state of
the art. Polym Degrad Stab, 88(2), pp. 268-274.
7. Grontoft and Raychaudhuri, 2004, Compilation of tables of surface deposition velocities for O3, NO2 and SO2
to a range of indoor surfaces, Atmospheric Environment, 38, pp. 533–544.
8. Hedges, R.E.M., 2002, Bone diagenesis: an overview of process, Archaeometry Vol. 44, No3, pp. 319-328.
9. Final Activity Report. Kjeller. Norwegian Institute for Air Research, NILU OR 42/2010.
10. Dellaportas, P., Papageorgiou, E. and Panagiaris, G., 2014, Museum factors affecting the ageing process of
organic materials: review on experimental designs and the INVENVORG project as a pilot study, Heritage
Science, 2:2. doi:10.1186/2050-7445-2-2.
ISBN: 978-960-98739-8-7
335
11. Rivers, S. and Umney, N., 2005, Conservation of furniture. Elsevier, Butterworth-Heinemann, Oxford.
12. Weintraub, S. and Wolf S. J., 1995, Macro and Microenvironments. In: Storage of Natural History Collections:
A Preventive Conservation Approach, Rose C.L., Hawks C.A., Genoways H.H. (eds.), SPNHC, pp. 123-134.
13. Yidiz, S, Gezer, E.D. and Yildiz, U.C., 2006, Mechanical and chemical behaviour of spruce wood modified by
heat. Building and Environment, 41, pp. 1762-1766.
336
ISBN: 978-960-98739-8-7
Bioart: Borders and definitions. Visitors’ survey at Bioart exhibition
S.O. Papaioannou1, E. Karantoni1*, Z. Tsourti3, E. Kokkinidou4, St.
Drakos1, N. Panourgia2 ,V. Papakiriakou, G. Panagiaris1, Th. Avaritsiotis1
1
2
3
4
& Works of Art, Aigaleo, GR, * [email protected]
Department of Anthropology, Columbia University, New York, USA
Ministry of Agricultural Development and Food, Athens, GR.
ArtLaboratoryBerlin, Berlin, DE
Keywords: Bioart, ethics, visitors’ survey, art museums
Abstract
BioArt stands at the intersection of various fields of science, art and society and brings with
it a range of social, religious, legal and ethical issues, since biological materials although tangible, are also “carriers” of strong intangible values. The research project “BioArt: Borders and
definitions”, aims at the establishment of a commonly accepted deontological framework for
the collection, display and management of Bioart works. A significant part of the project has
been accomplished through field research that utilised questionnaires which were addressed to
museum/galleries visitors. In this abstract, the methodology and the results of this research task
are presented.
From the results of the research it appears that currently there is no consensus about what Bioart
is, and there is also a great variation in public’s perception about what this new form of art is or
represent. Additionally, a variaty of reactions were detected, both negative and positive, when
it comes to ethics, personal values and feelings that Bioart can bring to surface. It seems that all
the above stem from the lack of a commonly accepted definition for Bioart which is a term quite
diversified and includes a variety of media, materials, applications and artworks.
Introduction
Biological materials constitute ‘objects’ of value for collection and research in medicine, archaeology and anthropology, as well as media of artistic expression for centuries (Gessert 2010).
In recent years, given the increased interest among artists and scientists in the use of biological
materials as artistic media, the display of biological materials, such as works of BioArt, has
expanded to include a vast category of museums and galleries (Kallergi 2008).
It is commonly accepted that in this time period of constant change and uncertainty, art and artists are influenced by the rapid changes occurring in science and technology and turn to other
forms of expression, in an effort to meet the new artistic challenges and participate in current
modalities of artistic production (Nelkin and Anker 2002, Wilson 2002).
BioArt stands at the intersection of various fields of science, art and society and brings with it a
range of social, religious, legal and ethical issues, since biological materials although tangible,
are also “carriers” of strong intangible values (Catts and Zurr, 2008).
It is true that the public perception for Bioart exhibitions has not been thoroughly investigated
yet (Kerbe and Schmidt, 2013). Moreover, the most important field that has to be investigated
is related with the ethics and policies in collecting and displaying Bioart.
It is necessary to create a commonly respected code of ethics from which good practices in matters such as acquisition, documentation, lending, dispossession, conservation, storage, exhibition, education and research in Bioart collections, will be developed.
ISBN: 978-960-98739-8-7
337
Although a fruitful dialogue on ethical and deontological issues, regarding the conservation and
management of human remains from archaeological, scientific and ecclesiastic collections has
begun, during the last two decades (Panagiaris 2001, Malea et al. 2004, DCMS 2005, Sakki et
al. 2007, Mertzani et al. 2008, Panagiaris 2009, Jenkins 2011), there still exists a gap when it
comes to the use of biological materials (especially those of human origin) as means of artistic
expression, precisely because the boundaries that circumscribe artistic creation, censorship,
science ethics, and human rights are unclear and vague. The project “BioArt: Borders and definitions. Research project for the development of a widely accepted deontological framework
of its production and management” aims to fulfill this gap and establish good practices for the
management of BioArt, taking into consideration artists’ and scientists’ input, public’s opinion,
as well as the current museological debate on ethics and standards of display and preservation
of biological materials. In this abstract, the methodology and the results of visitors’ survey are
presented.
Methodology
In order to pursue the objectives of the field research, a standardized questionnaire was selected
as the most appropriate means of collecting the necessary data, as it offers the highest degree
of confidentiality and because, when compared to other forms of data collection, it appears to
minimize any potential statistical or methodological biases resulting from the idiosyncrasies
and skill variability of the interviewer (Taylor-Powell, 1998). Moreover, the resulting data can
be analyzed using a variety of statistical procedures (Nicholson and Pearce, 2001).
The visitors’ survey was conducted at the Art Laboratory Berlin during two significant Bioart
exhibitions: [macro] biologies I: the biosphere (8.3.2014 - 4.5.2014) and [macro]biologies
II: organisms (31.5.2014 - 20.7.2014).
The questionnaires contained questions that aimed to elicit responses that could be helpful
in both qualitative and quantitative analysis. Responses were collected with the use of selfcompleted questionnaire, addressed to museum’s and gallery’s visitors, consisting of multiplechoice, close-ended and open ended questions in order to derive a spherical picture of visitors’
opinions on the issues under investigation (Gammon, 2001). The multiple choice questions
were in most of the occasions supplemented with an open-ended section, so as to gather as
much information as possible, and thus to be able to examine the BioArt issues in greater depth
(Mason, 2002).
The questionnaire consists of 16 questions and sub-questions, and is divided into the following sections: section A. Visitors Profile, section B. Beliefs, Attitudes, Opinions, section C.
Personal Data.
At the end of the questionnaire the participants were asked to write any additional comments
they wished. The questionnaire needed approximately 10 minutes to be completed.
As it is mentioned above the survey was contacted only to one art gallery in Berlin. The participants were 65 people in total. The sample size is not large enough to allow advanced statistical
analysis. Although, the results may vary with the increase of the sample, the data is particularly
interesting because the responses themselves present a valuable first image of audiences’ perceptions for this new art form.
The elaboration of the results revealed that:
48 percent of the visitors visit more than four (4) times per year an art museum and/or art gallery. Additionally, they visit art museums mainly because of a general interest or curiosity/
inspiration.
338
ISBN: 978-960-98739-8-7
The majority of the participants (74.5 percent) have a high educational level (bachelor, master
or PhD), although approximately 70 percent of them are not knowledgeable about contemporary art and especially, Bioart (66.15 percent). Nevertheless, most of the participants are interesting in knowing/learning what Bioart is.
Of great and particular interest are the results stemming from the answers provided by visitors
regarding issues that concern ethics and values in displaying work of arts where biological
materials are used. The largest percentage of visitors agrees to the use of organic materials in
artworks (57.4 percent), or is indifferent to them (34.4 percent). Indeed, a lot of them feel excitement (38.3 percent) in the sight of a Bioart work.
Furthermore, a high percentage, (approximately 80 percent), considers Bioart compatible with
their standards and values. It is obvious that at this point we come across a contradiction because even though the majority of the sample is not knowledgeable about Bioart and its applications they express a positive attitude towards ethical matters concerning the use of biological
materials in art and their personal ethical principles and values. Furthermore, they agree with
the use of biological materials without knowing about the methods that are used, artists’ motives, the origin of these materials and more important many of all them haven’t visit any Bioart
exhibition. And even if they had, most of them don’t remember the name of a Bio-artist or the
title of a Bioart work.
Concluding Remarks
In conclusion, Bioart has led to a new media art movement where artists, scientists, curators and
cultural institutions enthusiastically take part. Although the results of the visitors’ survey make
it apparent that there is a lack of knowledge among museum and gallery visitors about this new
form of art we locate an interest and a willingness of knowing it by the public.
The term Bioart is quite complicated, diversified and is constantly changing because it develops in parallel with the developments in modern biology, biotechnology and new media. Bioart
term includes a variety of artworks, practices, materials and methods so it is difficult to identify
a commonly accepted definition. The consequence is the incomplete knowledge of this new art
field, especially by the public, so we come across either with public’s indifference or negative
reactions to such collections and exhibits. The creation of a commonly accepted framework for
Bioart, its methods, materials and management will probably assist and promote acquaintance
with Bioart and perhaps make clearer the opinion of visitors about it.
Acknowledgements
This project is operating under the Action “Archimedes III” of the co-operational program
Education and Life Long Learning, which is co-funded by the Greek Ministry of Education and
the European Union.
References
1. Catts , O. and Zurr, I. 2008. The Ethics of Experiential Engagement with the Manipulation of Life. In: Tactical
Biopolitics Art, Activism, and Technoscience, edited by B. Da Costa and K. Philip, (Cambridge, Massachusetts
London, England: The MIT Press), pp.125-142.
2. DCMS (Department for Culture, Media and Sport). 2005. Guidance for the Care of Human Remains in
Museums. London: Department for Culture, Media and Sport.
3. Gammon, B. 2001. Effective questionnaires for all. A step by step recipe for successful questionnaire. (London:
Science museum).
4. Gessert, G. 2010. Green Light. Toward an Art of Evolution. USA: MIT Press. Date of access: 26/09/2015.
5. http://webarchive.nationalarchives.gov.uk/+/http://www.culture.gov.uk/images/publications/
GuidanceHumanRemains11Oct.pdf
6. Kallergi, A. 2008. Bioart on Display – challenges and opportunities of exhibiting bioart. The Netherlands: Leiden
University. Date of access: 26/09/2015. http://www.kallergia.com/bioart/docs/kallergi_bioartOnDisplay.pdf
ISBN: 978-960-98739-8-7
339
7. Kerbe, W. and M. Schmidt. 2013. Splicing boundaries: The experiences of bioart exhibition visitors. Leonardo
(46). Date of access: 26/09/2015. http://www.mitpressjournals.org/doi/pdfplus/10.1162/LEON_a_00701
8. Malea, K., M. Mertzani, and G. Panagiaris 2004. Dealing with sacred and criminal Human Remains. Proceedings
of the Canadian Association for Conservation of Cultural Property 30th Conference and Workshop, Quebec
26-30 May 2004, p.18.
9. Mason, J. 2002. Qualitative research. (London: SAGE Publications Ltd).
10. Mertzani, M., E. Malea, N. Maniatis and G. Panagiaris, 2008. Towards a Binding Code of Ethics for the
Conservation & Display of Human Remains. Proceedings of the ICOM-CC 15thTriennial Conference, New
Delhi, 22-26 September 2008, India: Allied Publishers Pvt.Ltd., p.364-370.
11. Nelkin, D. and S. Anker, 2002. The influence of genetics on contemporary art. Science & Society. (3): 967971. Date of access: 26/09/2015. http://www.suzanneanker.com/wp-content/uploads/2002-Anker-SuzanneNelkin-Dorothy-The-influence-of-Genetics-in-Contemporary-Art.pdf
12. Nicholson, R.E., and D.G. Pearce, 2001. Why do people attend events: a comparative analysis of visitor
motivations at four south island events? Journal of Travel Research, 39: 449-460.
13. Panagiaris, G., 2009. Methodological and Ethical Issues concerning the conservation and enhancement
of human remains. ANASKAMMA Journal, 3, 135-148. (in Greek) Date of access: 26/09/2015. http://
anaskamma.files.wordpress.com/2009/10/panagiaris.pdf
14. Panagiaris, G., 2001. The Influence of Conservation Treatments on Physical Anthropology Research. BAR
International Series 934, 95-99.
15. Sakki, Z., K. Moraitis, G. Panagiaris, C. Maravelias, C. Spiliopoulou. 2007. Souvenirs of Crime: Deciphering
History of Forensic Medicine through a Museum Collection of Human Remains. Proceedings of the 3rd
Balkan Congress on the History of Medicine, Thessaloniki, November 29 - December 2, p.121.
16. Taylor-Powell, W. 1998. Questionnaire design: Asking questions with a purpose. (Wisconsin: University of
Wisconsin).
17. Wilson, St. 2002. Information Arts. Intersections of Art, Science, and Technology. Cambridge. Massachusetts,
(London, England: The MIT Press).
340
ISBN: 978-960-98739-8-7
ISBN: 978-960-98739-8-7
341
Proceedings of the Conference
SCience in TEchnology
2015
SCinTE
www.scinte.gr
November 5-7, 2015
Athens, Greece
Proceedings Volumes
Volume 1
Applied Mechanics, Civil and Energy
Engineering
Earth and Environmental Sciences
Arts & Humanities
Volume 2
Materials & Micro-nano Technology
Electrical Engineering, Computer Science
& Telecommunications
Learning/Teaching Methodologies
and Assessment
Volume 3
Physical and Mathematical Sciences
Food Science and Technology
Life and Health Sciences, Bioengineering
EFI PANAGIOTIDI
Volume 4
Management Science, Finance,
Economics & Tourism
Conference Chairs: Dimos Triantis, Ioannis Kandarakis
Οrganized by
Supported by
Technological
(TEI) of Athens
Research Funding
Program
ARCHIMEDES III