Modelling and Simulation of Virtual Natural Lighting

Transcription

Modelling and Simulation of Virtual
Natural Lighting Solutions in Buildings
PROEFSCHRIFT
ter verkrijging van de graad van doctor
aan de Technische Universiteit Eindhoven,
op gezag van de rector magnificus, prof.dr.ir. C.J. van Duijn,
voor een commissie aangewezen door het College voor Promoties
in het openbaar te verdedigen
op woensdag 14 mei 2014 om 16.00 uur
door
Rizki Armanto Mangkuto
geboren te Bogor, Indonesië
Dit proefschrift is goedgekeurd door de promotoren en de samenstelling van de
promotiecommissie is als volgt:
voorzitter:
1e promotor:
2e promotor:
copromotor:
leden:
prof.dr.ir. J.J.N. Lichtenberg
prof.dr.ir. J.L.M. Hensen
prof.dr.ir. E.J. van Loenen
dr.ir. M.B.C. Aries
prof.dr. J. Mardaljevic (Loughborough University)
prof.dr. M. Andersen (École Polytechnique Fédérale de Lausanne)
prof.dr. E.H.L. Aarts
dr.ir. Y.A.W. de Kort
Modelling and Simulation of Virtual
Natural Lighting Solutions in Buildings
Rizki A. Mangkuto
The work described in this thesis has been carried out in the Unit Building Physics
and Services at the Department of the Built Environment, Eindhoven University of
Technology. This research was supported by the Sound Lighting research line of the
Intelligent Lighting Institute (ILI) at Eindhoven University of Technology.
Copyright © 2014 by Rizki A. Mangkuto
Eindhoven University of Technology, the Netherlands
All rights reserved. No part of this document may be photocopied, reproduced, stored,
in a retrieval system, or transmitted, in any form or by any means whether, electronic,
mechanical, or otherwise without the prior written permission of the author.
A catalogue record is available from the Eindhoven University of Technology Library
ISBN: 978-90-386-3604-7
NUR: 955
Bouwstenen 194
Cover design provided by A. Davie (Modul8) and adapted by P. Verspaget.
Printed by Gildeprint Drukkerijen, Enschede, the Netherlands.
Modelling and Simulation of Virtual Natural Lighting Solutions in Buildings /
by Rizki A. Mangkuto – Eindhoven University of Technology
– proefschrift –
Subject headings: Virtual Natural Lighting Solutions / virtual window prototype /
computational modelling / lighting simulation / visual comfort / building performance
“… But it is possible that you hate a thing that is good for you,
and that you love a thing that is bad for you.
Allah knows, while you know not.”
Al-Qur’an, 2: 216
To the memory of my dear mother
and to my dear wife,
for their endless and invaluable love,
just like the sun shining over the world
…
“Jiwaku tetap mengabdi pada Ibunda.
Dan aku pun tetap Timur, Adinda.”
Nur St. Iskandar (1922)
Table of Contents
Acknowledgements
xi
Summary
xv
Samenvatting
xvii
Ikhtisar
xix
Nomenclature
xxi
Chapter 1 – Introduction
1
1.1. Natural Lighting Demand
1
1.2. Shortcomings of Natural Lighting
3
1.3. Proposed Solution
4
1.4. Aim and Objectives
8
1.5. Research Methodology
8
1.6. Thesis Outline
10
Chapter 2 – General Concept of Virtual Natural Lighting Solutions
2.1. Definition of VNLS
13
2.2. Classification of VNLS
13
2.2.1. Prototypes with a simplified view
14
2.2.2. Prototypes with a complex view
16
2.3. Expectation of VNLS
19
2.3.1. Light quality
19
2.3.2. View quality
22
2.3.3. Target range
24
2.3.4. Comparison of prototypes
25
2.4. Concluding Remarks
29
Chapter 3 – Measurement and Simulation of a First Generation
Virtual Natural Lighting Solutions Prototype
3.1. Introduction
31
31
3.1.1. Modelling concept in Radiance
31
vii
13
Table of Contents
3.2. Case Description
34
3.3. Measurement Protocol
37
3.4. Simulation Protocol
38
3.5. Results and Discussion
42
3.5.1. Measurement of prototype
42
3.5.2. Simulation of prototype
44
3.5.3. Simulation of real windows
46
49
Chapter 4 – Discomfort Glare Evaluation and Simulation of a First Generation
4.1. Introduction
51
4.2. Method
56
4.2.1. Model description
56
4.2.2. Glare rating correlation
57
58
4.3.1. Rendering and glare source detection
58
4.3.2. Unadjusted rating
60
4.3.3. Polynomial regression
60
4.3.4. Adjusted rating
62
4.3.5. Percentage of disturbed subjects
66
69
Chapter 5 – Design, Measurement, and Simulation of a Second Generation
71
5.1. Introduction
71
5.2. Design Method
72
5.2.1. Test environment
72
5.2.2. Light sources
73
5.2.3. Control circuit
75
5.2.4. Display and structure
77
5.2.5. Programming and setting
79
viii
51
Table of Contents
80
82
82
5.4.2. Validation
83
5.5. Analysis of Various Configurations
84
5.6. Analysis of Various Operating Scenarios
85
5.6.1. Settings and data collection
85
5.6.2. Daily profiles and annual modes
86
91
5.7.1. Measurement of actual test room
91
5.7.2. Simulation of actual test room
95
5.7.3. Comparison of various configurations
97
5.7.4. Comparison of various operating scenarios
99
104
Chapter 6 – Modelling and Simulation of a Virtual Natural Lighting Solutions
with a Simplified View and Directional Light
6.1. Introduction
107
6.2. Methods
108
6.2.1. Modelling
108
6.2.2. Settings
113
6.2.3. Assessment
114
119
6.3.1. Sensitivity analysis
121
6.3.2. Comparison with real windows
123
128
Chapter 7 – Modelling and Simulation of a Virtual Natural Lighting Solutions
with Complex Views and Directional Light
131
7.1. Introduction
131
7.2. Methods
134
7.2.1. Modelling
134
ix
107
Table of Contents
7.2.2. Settings
136
7.2.3. Assessment
136
139
7.3.1. Transmissive approach
139
7.3.2. Comparison of transmissive and emissive approaches
145
148
Chapter 8 – Conclusions and Recommendations
8.1. Conclusions
151
8.2. Recommendations
154
References
157
Appendices
169
Curriculum Vitae
187
Publication List
188
x
151
Acknowledgements
All praise is due to Allah, the Light upon light, the Lord of the heaven and earth and
everything in between; for without Him, no single particle of light will ever exist in the
universe, and nor will this small thesis.
It was a little bit more than four years ago, when I considered submitting an online
application for a doctoral research position on the topic of modelling and simulation of
something I had never heard about: Virtual Natural Lighting Solutions. Knowing that such a
position in the topic of lighting can be relatively rare, and that no external scholarships were
required from my side to run the project, I had no doubt to proceed with applying. It was then
followed with a short, phone interview at almost midnight (western Indonesia time). The
outcome was destiny, a move to the City of Light in the ‘Land of Windmills’.
I would therefore like to express my gratitude to my first promotor, Prof. Dr. Ir. Jan
Hensen, who actually spoke on the other side of the line in that crucial phone interview, for
providing the invaluable opportunity to conduct the doctoral research, and for giving the trust
on me to execute the work in his leading, internationally diverse and recognised group of
Building Performance (simulation). Many of his (former) doctoral candidates, including
myself, appreciate the monthly progress meetings. Not only because the meetings give plenty
of opportunity to practice presentation skill and to learn from each other, but also because
they ensure us to stay on track in the progress; since as one says, it is very easy to get lost in
the jungle of research.
I would also express my gratitude to my second promotor, Prof. Dr. Ir. Evert van Loenen,
for his ideas, suggestions, and most of all his detailed comments on the experiments,
simulations, and any writings that were related with this doctoral research. Among others, he
also provided opportunity and support to conduct the experimental setup and measurements
of prototypes in the (old and new) ExperienceLab of Philips Research at the High Tech
Campus Eindhoven. The work there has proven to be an integral part of the project and
significantly contributed to the build-up of this thesis.
My gratitude and thanks also go to my copromotor, Dr. Ir. Myriam Aries, for her advice,
suggestions, and recommendations during the course of this long-term project. Many of her
inputs were so crucial in determining the direction of the project, which I appreciate very
much. Outside the office, I also had some nice chance to get in touch with her family and I
really enjoyed that.
Furthermore, I would like to express my appreciation to Prof. Dr. John Mardaljevic
(Loughborough University, the United Kingdom), Prof. Dr. Marilyne Andersen (École
Polytechnique Fédérale de Lausanne, Switzerland), Prof. Dr. Emile Aarts (Eindhoven University of Technology), and Dr. Ir. Yvonne de Kort (Eindhoven University of Technology)
xi
Acknowledgements
for willing to sit in the Doctorate Committee and for giving their invaluable comments on this
thesis.
This project was supported by the Sound Lighting research line of the Intelligent Lighting
Institute (ILI) at Eindhoven University of Technology. The works with prototypes were
mostly conducted with the help of Bernt Meerbeek, MSc, PDEng and other colleagues in
Philips Research at the High Tech Campus Eindhoven.
In the Unit of Building Physics and Services, my thanks are given to the past and present
colleagues in the group of Building Performance, in alphabetical order of the first name: Ana
Paula Melo, Azzedine Yahiaoui, Bruno Lee, Chul-sung Lee, Daniel Cóstola, Giovanni
Pernigotto, Hamid Montazeri, John Bynum, Katarína Košútová, Marcel Loomans, Massimo
Chiappini, Meng Liu, Mike van der Heijden, Mohammad Mirsadeghi, Mohammed Hamdy,
Petr Zelensky, Pieter-Jan Hoes, Qimiao Xie, Rajesh Kotireddy, Rebeca Barbosa, Roel
Loonen, Song Pan, Vojta Zavrel, and Wiebe Zoon. Among them, I particularly acknowledge
the suggestions from and brief discussion with Hamid (also thanks for the coffee-break chats,
proof-reading the thesis, and giving tutorial in creating high quality figures), Daniel, Roel,
and Wiebe, at certain points at some time within my stay in the group. My thanks are also
given to the colleagues in the group of Building Lighting: Mariëlle Aarts, who together with
Myriam had kindly provided the samenvatting of this thesis; Carlos Ochoa Morales, who was
partially involved in giving directions during the first two years of my project; and Parisa
Khademagha, who just started her doctoral research in the group, also under the umbrella of
ILI.
My next thanks are given to the colleagues, particularly the ‘senior members’ of the
group of Building Material, with whom I shared the open-plan working space in almost the
first two years of my projects: Qingliang Yu, Przemek Spiesz, Miruna Florea, Alberto Lazaro
Garcia, George Quercia Bianchi, Azee Taher, and Štěpán Lorenčik. I thank Przemek in
particular for his advice on my trip to Kraków and Oświęcim; and Qingliang for the rather
unique research collaboration on the application of modelling and simulation in Radiance for
photocatalytic material. My appreciations also go to the past and present secretaries of the
unit: Renée van Geene, Yeliz Varol, Janet Smolders, and Ginny Vissers; and to the staff in
HR service and International Office, particularly Carol van Iperen and Kara de Rooy. My
special gratitude is expressed to the head of BPS Laboratory, Jan Diepens, for providing
Radiance in Oracle VM VirtualBox and for providing the measurement devices; and to Harry
Smulders for setting up the connection to the UNIX server.
During my project, I had the honour to supervise three graduation projects of master
students: Chris van Dronkelaar, Ruben Pelzers, and Shen Wang; two pre-graduation projects
of master students: Richard Claessen and Annelous Bossers; and an honour project of two
bachelor students: Bas Kil and Martijn Gootzen. I thank all of them for the cooperation. In
particular, I express my gratitude to Wang, whose passion in lighting design and technology
had significantly contributed to two chapters of this thesis.
xii
Acknowledgements
Living in a country faraway from home can be dull without a nice and welcoming social
environment. I hereby would like to appreciate all of the Indonesian families and friends in
Eindhoven and surrounding, for their support to my family and myself in many ways. To all
of them, I would like to thank; terima kasih banyak dan sampai berjumpa lagi di lain
kesempatan.
In the Laboratory of Building Physics and Acoustics at Engineering Physics Research
Group, Institut Teknologi Bandung, I would like to acknowledge my respected gurus that
have shaped my academic experience, among others: Prof. Dr. Ir. R.M. Soegijanto, Dr. Ir.
F.X. Nugroho Soelami, Dr. Ir. Joko Sarwono, and Dr. Iwan Prasetiyo. I am very grateful for
have been learning from and working with them in the past, and will always look forward to
doing so again in the future.
As the Prophet said, heaven is beneath the sole of one’s mother. I am greatly indebted to
my mother, Saptarina (Allahu yarham), no mountains of gold can ever pay her unconditional
love. It is my biggest regret not to have her here to witness the completion of this thesis; and I
pray to Allah for keeping her in the best place in His heaven. To my father, Harry Mangkuto,
I am also indebted for his love and support. To my sister, Puti Kemalasari, and my brother-inlaw, Hario Rahadanto, I thank them for keeping on contact while separated in time and space.
To my parents-in-law, Asmiati dan Arman Nur, I give my greatest respect and gratitude. To
my aunt and uncle, Solita dan Santo Koesoebjono, I respectfully thank them for their support
during my stay in the Netherlands, and for warmly welcoming us whenever we visited them
in Wassenaar. To my grandmother, Sophie Sarwono, in Bogor and the rest of my big family
in Padang and Jakarta and other various places in Indonesia, I thank and appreciate their
support, prayers, and wishes during my period of stay here.
To my two boys, Mufid Raqillasyah dan Mirzavan Raskha, it is a great pleasure and
indescribable experience to witness their birth (both in Eindhoven) and watch them
constantly growing day by day. They are indeed the living time indicators: Mufid was born
only a year after I started my project, whereas Raskha was born when I was struggling with
writing and refining all of the messy chapters. I give my prayers for their health, safety, and
well-being; and hope that their period of stay in Eindhoven becomes an important and
somehow memorable part of their life.
Finally, to my dear and beloved wife, Asmelia, who always gives her infinite and endless
love, care, support, advice, and patience, despite so little I could only give her in return: no
words of thanks could describe my gratefulness. This thesis is dedicated to her, who has
dedicated herself to the sake of her family. Bunda, tesis ini Ayah persembahkan untuk
Bunda!
Eindhoven, March 2014
Rizki A. Mangkuto
xiii
Acknowledgements
xiv
Summary
Modelling and Simulation of Virtual Natural Lighting Solutions in
Buildings
In situations where daylight is not or insufficiently available, the concept of Virtual
Natural Lighting Solutions (VNLS), which are systems that can artificially provide natural
lighting as well as a realistic outside view, can be promising. The benefit of installing VNLS
in a building is the possibility to use more floor area that currently has very limited or no
access to daylight, with the additional possibility to control the light and view quality.
The aim of this research is predicting the impact of various VNLS applications on
lighting performance and visual comfort in buildings. Design methods including rapid
prototyping and multiple design concepts using computational modelling and building
performance simulation are proposed to obtain a better design of VNLS. The development
process is based on existing VNLS prototypes, which are further improved by applying the
relevant simulation tool.
Throughout the process, the following steps are carried out: (1) literature review of
current development of VNLS prototypes; (2) measurement of VNLS prototypes, which are
presented in case studies of a prototype with diffuse light (the so-called ‘first generation’
prototype) and a prototype with directional light (the so-called ‘second generation’
prototype); (3) simulation of VNLS prototypes and models, which are performed to predict
the lighting performance, measurements of the prototypes are incorporated to validate the
VNLS model; (4) computational modelling of future VNLS, which involves arrays of small
light sources with various tilt angles to deliver the light in various directions; (5) sensitivity
analysis, to understand the influence of the relevant input parameters to the relevant
performance indicators of the VNLS model; (6) performance comparison with real windows
scenes in simulation, to understand the benefit of installing VNLS in a given space; and (7)
theoretical calculation, to estimate the annual space availability and electrical energy
consumption of a VNLS prototype under various operating scenarios.
The most important results are summarised as follows:
• Based on the conducted measurement of a first generation VNLS prototype displaying
simplified sky scenes in a test room with no façades, it is known that the provided settings
do not satisfy the minimum lighting criteria. Based on Radiance lighting simulation, the
investigated prototype performs better in terms of light distribution uniformity than a
corresponding, hypothetical real window under overcast or partly cloudy scenes. Under
the clear sky scene, the difference between the real and virtual windows is less, due to the
influence of direct sunlight.
• A method is proposed to correlate the commonly applied glare metrics, which can be
predicted using simulations, to the glare rating that was used in an experiment conducted
xv
Summary
elsewhere, to assess discomfort glare from a first generation VNLS prototype with various
complex views. The results suggests the simulated, normalised values of DGI, UGR, and
CGI using Evalglare are all overestimated relative to the values converted from the
experiment data, while the simulated values of DGP are in a better agreement with the
converted DGP values.
• A second generation VNLS prototype has been designed and built. This prototype was
assessed to validate the computational model that can be extended for further development
of non-existing VNLS. Arrays of LED tiles were installed to provide diffuse light and a
simplified view, whereas arrays of linear LED fixtures were installed to provide direct
light that yielded some visible patches on the side walls of the test room. Simulation and
measurement values of horizontal illuminance at certain distances were evaluated and
showed a good agreement. The space availability can be optimised by placing a prototype
on each short wall facing each other, or by placing two prototypes on a long wall. All of
the investigated operating scenarios yield a relatively similar impact on the average annual
space availability and electrical energy consumption of the prototypes.
• More complex VNLS configurations composed of small light emitting sources have been
developed in computational model. The model has a simplified view, delivering the light
from the ‘ground’ to the ceiling and from the ‘sky’ to the floor. Sensitivity analysis shows
that total luminous flux of the ‘sky’ largely influences the space availability of the test
room, whereas the source beam angle largely influences other output variables, including
discomfort glare. Most of the modelled VNLS with a beam angle of 76° perform the
closest to the corresponding real windows. The performance can be optimised by
increasing the beam angle to 114°, yielding a space availability of around two times larger
than the corresponding real windows.
• A model of VNLS configurations with complex views has been created, where the light
was provided by arrays of white-coloured directional light emitting sources, while the
view was provided by pasting a two-dimensional image on a transparent glass in front of
the light sources. Comparisons are shown between 10 image scenes. The use of the
transmissive approach in this point offers more flexibility to apply complex views on the
display, but also requires more light to satisfy the illuminance criteria. On the other hand,
the use of emissive approach in the previous point may introduce more light, but the view
complexity is limited by the number of pixels.
xvi
Samenvatting
Modellering en Simulatie van Virtuele Natuurlijke
Verlichtingsoplossingen in Gebouwen
In situaties waarbij geen of onvoldoende daglicht aanwezig is, is het concept van Virtuele
Natuurlijke Verlichtingsoplossingen (VNLS) veelbelovend. VNLS zijn systemen die op
kunstmatige wijze zowel natuurlijk licht als een realistisch uitzicht bieden. Het voordeel van
het toepassen van VNLS in een gebouw is dat een groter vloeroppervlak met daglichtkwaliteit gebruikt kan worden dat eerder een beperkte of geen toegang tot daglicht had. Het
biedt tevens de mogelijkheid om het licht te regelen als ook de kwaliteit van het uitzicht.
Het doel van dit onderzoek is het voorspellen van de invloed van VNLS toepassingen op
de lichttechnische prestatie en het visuele comfort in gebouwen. Ontwerpmethodes
waaronder ‘snelle prototypering’ en meerdere ontwerpconcepten, waarbij gebruik gemaakt
wordt van computermodellering en -simulatie van gebouwprestatie, zijn toegepast om het
ontwerp van VNLS te verbeteren. Het ontwikkeltraject is gestart op basis van bestaande
VNLS prototypes en is middels toepassing van relevante simulatiemiddelen ontwikkeld.
Gedurende het proces zijn de volgende stappen genomen: (1) literatuuronderzoek naar de
huidige ontwikkeling van VNLS prototypes; (2) metingen aan bestaande VNLS prototypes,
weergegeven in case studies van een prototype met diffuus licht (eerste generatie prototype)
en een prototype met gericht licht (tweede generatie prototype); (3) simulatie van nietbestaande VNLS prototypes en modellen, uitgevoerd om de lichtprestatie te voorspellen
naast metingen van de bestaande prototypes welke een onderdeel vormen van de validatie
van het VNLS model; (4) computermodellering van de VNLS van de toekomst bestaande uit
de simulatie van een verzameling van kleine lichtbronnen met diverse richtingshoeken om
licht vanuit verschillende richtingen te creëren; (5) gevoeligheidsanalyse om de invloed van
de inputparameters op de relevante prestatie-indicatoren van het VNLS model te
kwantificeren; (6) prestatievergelijking van VNLS met een eveneens gesimuleerd echt raam
om inzicht te krijgen in het voordeel van het installeren van VNLS in een ruimte; en (7)
theoretische berekeningen om de jaarlijkse ruimte-beschikbaarheid en het elektrische
energieverbruik van een VNLS prototype onder verschillende gebruiksscenario’s in te
schatten.
Een opsomming van de belangrijkste resultaten is de volgende:
• Uit metingen aan een eerste generatie VNLS prototype met vereenvoudigde hemelscènes
in een testruimte zonder gevels is gebleken dat de hierbij beschikbare instellingen niet aan
de minimale lichtcriteria voldoen. Lichtsimulaties in Radiance laten zien dat het
onderzochte prototype beter presteert wat betreft lichtverdeling dan een vergelijkbaar,
gesimuleerd echt raam onder een bewolkte of gedeeltelijk bewolkte hemelscènes. Met een
heldere hemelscène is het verschil tussen een echt en een virtueel raam kleiner. Dit wordt
veroorzaakt door de invloed van het directe zonlicht.
xvii
Samenvatting
• Een methode is voorgesteld om vier algemeen toegepaste verblindingindexen, die middels
simulaties kunnen worden voorspeld, te correleren met de resultaten van het helderheidoordeel in een elders uitgevoerd experiment met proefpersonen. Op die manier kan het
visueel discomfort van een eerste generatie VNLS prototype met verschillende complexe
mogelijkheden tot uitzicht worden beoordeeld. De resultaten suggereren dat de gesimuleerde, genormaliseerde indexwaarden DGI, UGR en CGI, berekend met behulp van
Evalglare allemaal zijn overschat in vergelijking tot waarden verkregen uit de
experimentele data. De gesimuleerde waarden van DGP komen beter overeen met de
geconverteerde DGP waardes uit het experiment.
• Een tweede generatie VNLS prototype is ontworpen en gebouwd. Dit prototype werd
gebruikt voor ter validatie van het computermodel en kan worden uitgebreid voor verdere
ontwikkeling van niet-bestaande VNLS. Een raster van LED-tegels werd geïnstalleerd om
een diffuse lichtbron en een vereenvoudigd uitzicht. Lineaire LED-armaturen zijn
geïnstalleerd om te voorzien in direct licht, dat zichtbaar wordt als enkele heldere vlakken
op de zijwanden van de testruimte. Simulatieresultaten en meetwaarden van de horizontale
verlichtingssterkte op vergelijkbare meetpunten werden geëvalueerd en toonden een goede
overeenkomst. De ruimte-beschikbaarheid in een standaard kantoorruimte kan worden
geoptimaliseerd door een prototype op elke korte wand tegenover elkaar of door twee
prototypes op een lange wand naast elkaar te plaatsen. Alle onderzochte operationele
scenario’s laten een relatief vergelijkbaar effect zien op de gemiddelde jaarlijkse ruimtebeschikbaarheid en het elektrische energieverbruik van de prototypes.
• Meer complexe VNLS configuraties, bestaande uit kleine licht-uitstralende bronnen, zijn
ontwikkeld in een simulatiemodel. Het model heeft een vereenvoudigd uitzicht en voorziet
in het gereflecteerde licht van de ‘grond’ op het plafond en van de ‘hemel’ op de vloer. Uit
een gevoeligheidsanalyse bleek dat de totale lichtstroom van de ‘hemel’ voornamelijk de
ruimte-beschikbaarheid van de testruimte beïnvloedt, terwijl de stralingshoek van de
lichtbron voornamelijk van invloed is op de andere outputvariabelen, waaronder
verblinding. Het merendeel van de gemodelleerde VNLS met een stralingshoek van 76°
presteerden nagenoeg identiek aan echte ramen. De prestaties kunnen worden geoptimaliseerd door de stralingshoek te vergroten naar 114°, waardoor een ruimte-beschikbaarheid
ontstaat die ongeveer twee keer groter is dan die van corresponderende echte ramen.
• Een model van VNLS configuraties met een complex uitzicht is gesimuleerd, waarbij het
licht afkomstig was van een raster van gerichte, witgekleurde lichtbronnen. Het uitzicht
werd gerealiseerd door een tweedimensionale afbeelding aan te brengen op een
transparante glasplaat welke geplaatst is voor de lichtbronnen. Vergelijkingen tussen 10
afbeeldingsscènes zijn getoond door het gebruik van de transmissieve aanpak. Deze
aanpak biedt weliswaar meer flexibiliteit om een complex uitzicht op het scherm te
projecteren maar behoeft meer licht om aan de vereiste verlichtingssterkte te voldoen.
Anderzijds kan het gebruik van een emissieve aanpak, zoals gedemonstreerd in het vorige
punt, het licht efficiënter in de ruimte brengen, maar de complexiteit van het uitzicht
wordt beperkt door het aantal pixels. xviii
Ikhtisar
Pemodelan dan Simulasi Solusi Pencahayaan Alami Virtual dalam
Bangunan
Dalam situasi di mana pencahayaan alami tidak tersedia secara mencukupi, konsep Solusi
Pencahayaan Alami Virtual (VNLS), yaitu sistem yang dapat menyediakan cahaya alami dan
pemandangan ke luar secara realistis, adalah sangat menjanjikan. Manfaat dari pemasangan
VNLS dalam bangunan yaitu adanya kemungkinan untuk menggunakan lebih banyak luas
lantai yang sebelumnya hanya memiliki sedikit akses kepada cahaya alami. Hal ini
menciptakan kemungkinan untuk mengendalikan kualitas dari cahaya dan pemandangan
yang tersedia.
Penelitian ini bertujuan untuk memprediksi pengaruh dari berbagai aplikasi VNLS pada
kinerja pencahayaan dan kenyamanan visual dalam bangunan. Dalam penelitian ini, diajukan
metode perancangan menggunakan kreasi purwarupa secara cepat serta konsep perancangan
berdasarkan pemodelan komputasional dan simulasi kinerja bangunan, untuk mendapatkan
desain VNLS yang lebih baik. Proses pengembangan tersebut didasarkan pada purwarupa
VNLS yang telah tersedia, kemudian disempurnakan lebih lanjut menggunakan perangkat
simulasi yang relevan.
Dalam penelitian ini dilakukan langkah-langkah sebagai berikut: (1) tinjauan literatur
yang terkait dengan perkembangan purwarupa VNLS terkini; (2) pengukuran purwarupa
VNLS, berupa studi kasus dari suatu purwarupa bercahaya difus (disebut juga purwarupa
‘generasi pertama’) serta purwarupa bercahaya terarah (disebut juga purwarupa ‘generasi
kedua’); (3) simulasi purwarupa dan model VNLS, menggunakan perangkat simulasi
komputasi untuk memprediksi kinerja pencahayaan dalam bangunan; yang mana pengukuran
dari purwarupa ‘generasi kedua’ digunakan untuk memvalidasi model VNLS; (4) pemodelan
komputasi dari VNLS, melibatkan rangkaian sumber cahaya berukuran kecil dengan berbagai
sudut kemiringan, guna menghantarkan cahaya ke berbagai arah; (5) analisis sensitivitas,
untuk mengetahui pengaruh dari parameter-parameter masukan terhadap indikator kinerja
dari model VNLS yang terkait; (6) perbandingan kinerja dengan jendela sejati untuk
mengetahui seberapa besar manfaat dari pemasangan VNLS dalam suatu ruangan; dan (7)
perhitungan teoretis, untuk memperkirakan konsumsi energi listrik tahunan dari suatu
purwarupa VNLS dengan berbagai skenario pengoperasian.
Hasil yang terpenting dari penelitian ini dapat dirangkum sebagai berikut:
• Berdasarkan pengukuran dari suatu purwarupa ‘generasi pertama’ yang menampilkan
pemandangan langit yang disederhanakan, didapatkan bahwa kriteria pencahayaan
minimum yang dikehendaki tidaklah terpenuhi. Berdasarkan simulasi pencahayaan dengan
Radiance, purwarupa tersebut menghasilkan kemerataan yang lebih baik daripada jendela
sejati, di bawah kondisi langit mendung dan berawan sebagian. Di bawah kondisi langit
xix
Ikhtisar
cerah, perbedaan antara jendela sejati dan virtual menjadi lebih rendah karena pengaruh
dari cahaya matahari langsung.
• Sebuah metode diajukan untuk menghubungkan empat metrik kesilauan yang umum
diterapkan, yang dapat diprediksi menggunakan simulasi, dengan nilai kesilauan subjektif
dari suatu purwarupa VNLS generasi pertama yang didapatkan dalam suatu eksperimen
menggunakan subjek manusia. Hasil simulasi dengan Evalglare menunjukkan bahwa
nilai-nilai DGI, UGR, dan CGI yang dinormalisasi adalah lebih tinggi daripada nilai-nilai
yang dikonversi dari data eksperimen, sedangkan nilai-nilai DGP berdasarkan simulasi
lebih sesuai dengan nilai-nilai DGP yang dikonversi.
• Sebuah purwarupa VNLS generasi kedua telah dirancang dan dibangun. Purwarupa ini
dievaluasi untuk memvalidasi model komputasi yang dapat digunakan lebih lanjut untuk
memodelkan VNLS yang belum ada. Suatu rangkaian ubin LED secara khusus digunakan
untuk menghasilkan cahaya difus dan pemandangan, sedangkan rangkaian armatur LED
linear digunakan untuk memberikan cahaya terarah yang menghasilkan berkas-berkas
yang dapat terlihat pada dinding samping dari ruang uji. Simulasi dan pengukuran
iluminansi horisontal pada lokasi yang bersesuaian menunjukkan hasil yang serupa satu
sama lain. Ketersediaan ruang dalam sebuah ruang kantor standar dapat dioptimalkan
dengan cara menempatkan satu purwarupa pada setiap dinding pendek, atau dengan
menempatkan dua purwarupa pada sebuah dinding panjang. Seluruh skenario pengoperasian menghasilkan pengaruh yang relatif sama pada rata-rata ketersediaan ruang dan
konsumsi energi listrik tahunan dari purwarupa.
• Konfigurasi VNLS yang lebih kompleks menggunakan sumber-sumber cahaya berukuran
kecil telah dirancang dalam model komputasi. Model tersebut menampilkan pemandangan
yang disederhanakan, menghantarkan cahaya dari ‘tanah’ ke langit-langit dan dari ‘langit’
ke lantai. Analisis sensitivitas menunjukkan bahwa fluks cahaya total dari ‘langit’ sangat
mempengaruhi ketersediaan ruang, sedangkan sudut pancaran dari sumber sangat
mempengaruhi variabel-variabel keluaran yang lain, termasuk tingkat kesilauan. Sebagian
besar model VNLS dengan sudut pancaran 76° menghasilkan kinerja yang paling serupa
dengan jendela sejati. Dengan meningkatkan sudut pancaran menjadi 114°, ketersediaan
ruang dapat dioptimalkan menjadi sekitar dua kali lebih besar dibandingkan dengan
ketersediaan ruang yang dihasilkan oleh jendela sejati.
• Sebuah model dari konfigurasi VNLS dengan pemandangan kompleks telah dirancang, di
mana cahaya dihasilkan dari susunan sumber-sumber cahaya terarah berwarna putih,
sedangkan pemandangan dihasilkan dengan cara menempelkan gambar dua dimensi pada
permukaan kaca transparan di depan sumber-sumber cahaya. Perbandingan antara 10 jenis
gambar ditunjukkan dengan menggunakan metode transmisif yang menawarkan
fleksibilitas dalam menampilkan pemandangan yang kompleks, namun juga membutuhkan
lebih banyak cahaya untuk memenuhi kriteria pencahayaan dalam ruang. Pada sisi lain,
penggunaan metode emisif pada poin sebelumnya dapat menghasilkan lebih banyak
cahaya, namun kompleksitas dari pemandangan yang ditampilkan menjadi terbatas oleh
banyaknya piksel yang digunakan.
xx
Nomenclature
Roman symbols
%A
%A R
%A V
%G
%Gav
Ai
BA
C1
C2
CGI
CGIn
CGIn conv
CGIn sim
d
DF
DGI
DGIn
DGIn conv
DGIn sim
DGP
DGP conv
DGP sim
Eav
Ecrit
Ed
Eground
Ei
Emea
Emin
Esim
Etotal
Ev
ERC
IA
space availability [%]
space availability under real windows scene [-]
space availability under VNLS scene [-]
ground contribution on the ceiling [%]
average ground contribution on the ceiling [%]
projected light source surface [m2]
beam angle of the ‘sky’ element [°]
Einhorn’s weighting coefficient, i.e. 8 [-]
Einhorn’s weighting coefficient, i.e. 2 [-]
CIE glare index [-]
normalised CIE glare index [-]
converted normalised CIE glare index [-]
simulated normalised CIE glare index [-]
distance between individual windows [m]
daylight factor [%]
daylight glare index [-]
normalised daylight glare index [-]
converted normalised daylight glare index [-]
simulated normalised daylight glare index [-]
daylight glare probability [-]
converted daylight glare probability [-]
simulated daylight glare probability [-]
average illuminance [lx]
criterion illuminance [lx]
direct vertical illuminance [lx]
illuminance contribution from the ‘ground’ element on the ceiling [lx]
diffuse vertical illuminance [lx]
measured illuminance [lx]
minimum illuminance [lx]
simulated illuminance [lx]
total illuminance [lx]
vertical illuminance on the observer’s eye [lx]
externally reflected component [%]
interval of tilt angle of the ‘sky’ element [°]
xxi
Nomenclature
IR,G,B
IR
IG
IB
IRC
Lav
Lb
Le
Li
Lmax
Lmin
Lo
Lr
total spectral irradiance [W/m2]
red spectral irradiance values [W/m2]
green spectral irradiance values [W/m2]
blue spectral irradiance values [W/m2]
internally reflected component [%]
average surface luminance [cd/m2]
background luminance [cd/m2]
emitted radiance [W/(sr·m2)]
incoming radiance [W/(sr·m2)]
maximum surface luminance [cd/m2]
minimum surface luminance [cd/m2]
outgoing radiance [W/(sr·m2)]
reflected radiance [W/(sr·m2)]
total spectral radiance [W/(sr·m2)]
LR
red spectral radiance values [W/(sr·m2)]
LG
green spectral radiance values [W/(sr·m2)]
LB
blue spectral radiance values [W/(sr·m2)]
surface or glare source luminance [cd/m2]
Ls
N
total number of points [-]
n(E ≥ 500 lx) number of points with illuminance ≥ 500 lx [%]
n(E ≥ Ecrit)
number of points with illuminance exceeding the criterion illuminance [%]
P
position index [-]
PDGav
average probability of discomfort glare [-]
PDGav R
average probability of discomfort glare under real windows scene [-]
PDGav V
average probability of discomfort glare under VNLS scene [-]
SC
sky component [%]
U0
uniformity [-]
U0 R
uniformity under real windows scene [-]
U0 V
uniformity under VNLS scene [-]
UGR
unified glare rating [-]
UGRn
normalised unified glare rating [-]
UGRn conv
converted normalised unified glare rating [-]
UGRn sim
simulated normalised unified glare rating [-]
Wreal
real-time electrical power consumption [W]
LR,G,B
xxii
Nomenclature
Greek symbols
β
β’
θi
θo
ρ
ρR
ρG
ρB
τ
τR
τG
τB
Φ
Φi
Φv
ψi
ψo
Ωi
ωpos
ωs
regression coefficient [-]
standard regression coefficient [-]
incoming angle [rad]
outgoing angle [rad]
weighted average spectral reflectance [-]
spectral reflectance in red [-]
spectral reflectance in green [-]
spectral reflectance in blue [-]
weighted average spectral reflectance [-]
spectral transmittance in red [-]
spectral transmittance in green [-]
spectral transmittance in blue [-]
total luminous flux of the ‘sky’ element [lm]
total radiative flux of light source [W]
total luminous flux of light source [lm]
incoming angle [rad]
outgoing angle [rad]
solid angle of the incoming radiance [sr]
modified solid angle of the glare source [sr]
solid angle of the glare source [sr]
Abbreviations
CCT CEN
CIE CRI CQS DALI
DIR
DML
DMM
DMR
DMX
DNL
DNR
correlated colour temperature
Comité Européen de Normalisation
Commission Internationale de l'Eclairage
colour rendering index
colour quality scale
digital addressable lighting interface
directionality
distant mixed land distant man-made distant mixed river digital multiplex distant near land distant near river
xxiii
Nomenclature
DoE
DPC
EN
GSD
HD
HDR
INF
ISO
LED MRI
NML
NMM
NMR
NNL
NNR
ORG
PAR
RGB
RMS
RW
SPD
TL VNLS
Department of Energy
depth perception cues
European standard
green, sky, and distant objects
high-definition
high dynamic range
information
international standard (for measuring film speed)
light emitting diode
magnetic resonance imaging
near mixed land
near man-made
near mixed river
near natural land
near natural river
organisation
parabolic aluminium reflector
red, green, blue
root mean square error
real windows
spectral power distribution
tubular fluorescent lamp
virtual natural lighting solutions xxiv
Chapter 1
Introduction
This chapter discusses the use of natural light in buildings, its benefits and shortcomings,
particularly the limitation in time and space. This chapter introduces the concept of Virtual
Natural Lighting Solutions (VNLS), and the use of computational modelling and simulation
to steer the development of the solutions. The aim and objectives of the research, as well as
the thesis outline, are further described.
1.1. Natural Lighting Demand
Human beings have a strong preference for natural light. Many researchers have shown
that natural light is highly preferred over electrical lighting in the built environment for its
positive effects on user satisfaction and health (e.g. Farley & Veitch, 2001; Boyce, 2003;
Chang & Chen, 2005; Galasiu & Veitch, 2006; Aries et al., 2010). Natural light provides
various stimulations throughout the day, and it is believed that access to natural light can
reduce stress and increase productivity (e.g. Boyce et al., 2003; Heschong, 2003a; Heschong,
2003b). A relationship between stress, depression, and little exposure to natural light has been
discovered, on which thorough overviews are provided by, e.g. Edwards & Torcellini (2002);
Boyce et al. (2003); Boubekri (2008); and Beute & de Kort (2014). A thorough literature
review of studies on the effects of daylight exposure on human health since 1989 until 2013
is presented by Aries et al. (2013), in which a statistically significant and well-documented
evidence for the relationship between daylight and its potential effect on health was found to
be limited. Nonetheless, some first practical implementations for building design can already
be shown (Aries et al., 2013).
In buildings, the admission of natural light also provides a view with information about
the outside situation, such as time of the day and weather condition. In general, it is found
that weather is influential on people’s health and mood (e.g. Eagles, 1994; Keller et al., 2005;
Denissen et al., 2008). Several studies have reported on beneficial and restorative effects of
views onto a natural scene (e.g. Ulrich, 1984; Ulrich et al., 1991; Tennessen & Cimprich,
1995) whereas views onto human-built environments yield effects which are similar to
having no window at all (Kaplan, 1993). Kim & Wineman (2005) showed in an internal
report that empirically, views and windows have psychological and economic values. In the
first part of their study, they showed that availability of view from a building was positively
related to assigned property values. In the second part, they recorded seating selection
occupancy rates in a cafeteria and a library, and found that people were more likely to choose
a seat near windows and views.
In general, a window is an opening in the wall that allows the admittance or flows of air,
light, and sound, which mostly influence the indoor environment (Tregenza & Loe, 1998). In
its very basic function, a window provides the possibility for having light and air from
1
Chapter 1
outside into the inside space. Nevertheless, Collins (1975) found that windows provided
many more functions for people than just sources of light and air. In her study with 88
window-related cases conducted in a variety of settings, she found that windows provided a
view to the outside, knowledge of the weather and time of day, feeling of connection to the
outside environment, relief from feelings of claustrophobia, monotony or boredom.
Many researchers show that the view is an important aspect provided by a real window,
and even cannot be separated from the natural light itself (e.g. Tuaycharoen & Tregenza,
2007). The findings of Markus (1967) and Keighley (1973a, 1973b) showed that views
should have three specific layers: a layer of sky, a layer of city or landscape, and a layer of
ground. Each layer has its own specific function.
Depending on the position of the observer inside the building, as well as the location of
the building itself, Keighley (1973b) pointed out that for a typical office building in an urban
area, the view can be classified into three types. Figure 1.1 illustrates these three types: the
first one represents a cityscape scene with a natural horizon, such as normally seen from the
uppermost floors of a tall building; the second represents a panorama of mid-ground
buildings, giving an elevated skyline seen from approximately ground floor position; and the
third is entirely occupied by the façade of a nearby building. The results showed that the type
of view, together with other factors, can influence the satisfaction of the observer.
(a)
(b)
(c)
Figure 1.1. Three types of view investigated by, and taken from Keighley (1973b). View (a) is a
cityscape scene with a natural horizon, (b) is a panorama of mid-ground buildings, and (c) is
entirely obstructed by the façade of a nearby building.
2
Introduction
1.2. Shortcomings of Natural Lighting
Despite all of its advantages, the quality and quantity of natural light is highly variable.
Its availability is limited by space and time. For instance, there is not enough or no daylight at
all during nighttimes, buildings can be too deep to supply sufficient daylight throughout the
space, and some rooms are simply not provided with windows, skylights, or any form of
daylight transporting systems. The latter situation can be found, among others, in operating
rooms in hospitals and in control rooms in industrial plants; due to hygienic or safety
considerations. Or, as suggested in Figure 1.1c, people may be working nearby a window
whose view is obstructed by the façade of a neighbouring/adjacent building, and find there is
not much natural light penetrating into their work stations, for instance because the sky is
obstructed in such a way that there is no functional daylighting possible.
In general, useful natural light in a sidelit space for typical office and educational
activities can be roughly estimated by the so-called ‘window-head-height’ rule of thumb
(Reinhart & Weismann, 2012). This rule relates how far ‘adequate, useful and balanced
daylight enters the spaces for most of the year’, as a function of the distance from any point
on the floor to the top of the window (Reinhart, 2005). A simulation-based validation study
of this rule of thumb for unobstructed facades yielded that the depth of the daylit area usually
lies between 1 and 2 times the size of the window-head-height, in a typical sidelit office
space with Venetian blinds. For spaces that are not equipped with movable shading devices,
such as atria or circulation areas, the ratio can increase up to 2.5 times the size of the
window-head-height.
Admission of natural light into work places is recommended in most national building
standards, even though the legislation differs from country to country (Boubekri, 2004;
Boubekri, 2008). In commercial buildings such as offices, the layout design is however very
much dependent on the national context, as thoroughly discussed by van Meel (2000). For
example, according to Saxon (1994), employees in the United States tend to sit within 14 ~
16 m from a window, British employees are used to sitting within 8 ~ 10 m, and German
employees within 4 ~ 6 m. According to Duffy et al. (1993), British employees tend to work
in open-plan offices while their counterparts in North European countries are used to work in
cellular offices. A clear example of this tendency is illustrated in Figure 1.2, adapted from
van Meel (2000), which shows the work place layout of the main headquarters of a
commercial bank in Amsterdam and London.
As shown, most employees in Amsterdam (representing most of the North European
countries) sit next to a window, in a building with corridors and long, narrow floor plan. Such
layout can ensure sufficient access to natural light for everyone in the building; however, the
design will also require more land area. In London (representing the United Kingdom and the
United States), most employees are working in relatively small cubicles in large open areas.
This concept can efficiently reduce the use of total land area, but on the other hand, it will
provide very little access of natural light to the employees who work in the cubicles.
3
Chapter 1
(a)
(b)
Figure 1.2. Work place layout of the main headquarters of a same commercial bank in (a)
Amsterdam and (b) London, taken and adapted from van Meel (2000)
Another important fact is that significant fractions of the working population in the world
do their work during nighttime. In the European Union, approximately 15% of women and
29% of men of the working population (< 45 years old) do night shift work (Härmä &
Ilmarinen, 1999). For the elder working population (≥ 45 years old), the figures were 12% for
women and 24% for men (Härmä & Ilmarinen, 1999). Night shift workers experience various
discomfort issues, such as sleep problems, fatigue, and poor performance, and even increased
long-term risk of some types of cancer due to a lack of synchronisation between the shift
work schedule and the worker’s light-dark cycle (Stevens, 2009; Blask, 2009).
Moreover, many studies have reported that the value of increased productivity due to an
improved indoor climate can be much greater than the costs of energy it consumes (e.g.
Woods, 1989; Skåret, 1992; Kosonen & Tan, 2004; CABA, 2008; EC, 2013). The quality of
the working environment in offices, schools, and factories could therefore outweigh any
savings in energy and should become one of the major drivers in research on buildings (EC,
2013). All of these considerations lead to a demand for having an artificial solution that can
bring natural light with all of its qualities to the inside space.
1.3. Proposed Solution
In situations where daylight is not or insufficiently available, Virtual Natural Lighting
Solutions (VNLS) concept can be promising. VNLS are systems that can artificially provide
4
Introduction
natural lighting as well as a realistic outside view, with properties comparable to those of real
windows and skylights. The benefit of installing VNLS in a building is the possibility to use
more floor area that currently has very limited or no access to daylight, with the additional
possibility to control the light and view quality.
The real, ideal product that gives both light and view in a very high quality does not yet
exist at the moment. Nonetheless, it is known that in its intense appearance without a
sufficient view, bright light can have a positive effect on human well-being (e.g. Glass et al.,
1985; Badia et al., 1991; Avery et al., 1992; Eastman et al., 1998; Lingjærde et al., 1998;
Avery et al., 2001; Mottram et al., 2011; Smolders et al., 2012; Smolders, 2013; Smolders et
al. 2013; Beute & de Kort, 2014). The inverse is also true, that artificial views which emit no
light themselves can also have a positive effect on humans (e.g. Heerwagen & Orians, 1986;
Heerwagen, 1990; Ulrich et al., 1993). The concept of VNLS is to combine both light and
view together, to provide even more positive effects on the users.
Investigations of the psychological effects of existing VNLS prototypes are still an
ongoing process. For example, de Vries et al. (2009) conducted experiments to study the
work performance of test subjects in a standard office room with two units of ‘emulated
windows’, obstructed with a diffuse screen. Prototypes of the same type were used in the
experiments of Smolders et al. (2012) to investigate the effects of brightness, focusing on the
effect of eye illuminance on subjective measures, task performance, and heart rate variability.
Experiments on glare sensation from another prototype with a simplified view were
conducted by Rodriguez & Pattini (2014), observing its effects on glare-sensitive and glareinsensitive subjects when performing a computer task. A number of short- (one day) and
long-term (3 and 10 years) acceptance studies of a VNLS prototype with a simplified sky
scene and sunlight were conducted by Enrech Xena (1999), as also shown in (Fontoynont,
2011a, 2011b), which results showed a high acceptability in windowless space for long-term
use, under some certain settings.
Some user perception studies on view and light (quality) aspects of VNLS prototypes
have been reported. For example, Tuaycharoen & Tregenza (2005) studied subjective
discomfort glare from screen projected images, and concluded that a good view (also
described as a view with high interest), which mainly consists of the natural scenes, tends to
reduce discomfort glare perception. IJsselsteijn et al. (2008) focused on depth perception cues
from screen projected images, and concluded that motion parallax, occlusion, and blur had a
significant effect on the viewer’s see-through experience, with motion parallax yielding the
greatest effect size. Shin et al. (2012) investigated subjective discomfort glare from a backlit,
transparent printed image, and concluded that the tolerance of discomfort glare sensation for
the distant views including skyline was greater than the near views. In all of those
experiments, the prototypes/displays were assumed to be a representation of what the subjects
normally see through a real window.
Nevertheless, the existing prototypes are considered not suitable for meeting the whole
expectation, since they are only able to meet part of the natural light and view expectation
(Mangkuto et al., 2011). For instance, the light produced is much less than that coming from
5
Chapter 1
a real window, there is no directional light component, the view displayed is limited; and so
on. These limitations indicate that the ideal VNLS concept should go beyond its
predecessors.
In this thesis, the light and view qualities are considered two key aspects in assessing
existing and future solutions. In particular, the presence of a directional light component and
the complexity of the view are taken as the general descriptors for the qualities. Therefore,
VNLS prototypes (that exist) and models (that do not yet exist) can be generally classified
into four types, as illustrated in Figure 1.3, which are:
1.
2.
3.
4.
Type that provides a simplified view and (mainly) diffuse light
Type that provides a complex view and (mainly) diffuse light
Type that provides a simplified view and (mainly) directional light
Type that provides a complex view and (mainly) directional light
Note that the first two types are existing prototypes, which can alternatively be considered
as the ‘first generation’ prototypes, which are available for research and/or commercial
purpose. The last two types are future generations, which are proposed and built in this thesis
using computational modelling and simulation, and for which the physical models are not yet
available.
Figure 1.3. Classification of VNLS based on light and view qualities
6
Introduction
While the relationship between the first generation VNLS prototype and user perception
has been investigated elsewhere (e.g. Tuaycharoen & Tregenza, 2005; IJsselsteijn et al.,
2008; Shin et al., 2012), there is very little discussion about the impact of VNLS application
on building performance. This thesis focuses on the latter, in which the influence of such
solutions on the lighting performance and visual comfort inside the relevant spaces is
evaluated. Since the ideal VNLS are future, not-yet-existing systems with lots of possible
input variables, computational modelling and building performance simulation are applied to
predict the performance of the solutions and to accelerate the development process.
In his thesis, Crawley (2008) described computational modelling as an approach that
allows evaluation of alternative designs or technologies without having to create the artifacts.
It is generally cheaper to create a model and to test alternative designs configurations than to
build a real prototype and revise it later based on trial and error. Furthermore, in view of
application in buildings, Crawley described building performance simulation as a powerful
tool that emulates the dynamic interaction of natural, physical phenomena such as heat, light,
and sound within the building to predict its energy and various environmental performances,
available today for use by policy setters and decision makers. In this thesis, design methods
including rapid prototyping and providing multiple design concepts are proposed as means to
obtain better design solutions.
Within the building design context, a number of design stages can be distinguished, as
suggested e.g. by Stoelinga (2005). These stages mainly consist of decision, programme of
requirements, preliminary design, final or detailed design, and the contract document. The
objectives and requirements are defined in the programme of requirements or project brief.
The main systems are selected and a number of concepts are developed in the preliminary or
conceptual design. Next, the development and integration of design elements to operate
design solutions takes place in the final design stage, which is finally closed with the contract
document in which the production drawings, specification, and construction resource
documentation are finalised. In particular, there is a demand to use building performance
simulation for design support of the generation and selection of alternative design concepts
during early phases in the design process, where decisions often have to be made with limited
resources and based on limited knowledge (Stoelinga, 2005; Hopfe, 2009).
In the context of this thesis, computational modelling and building performance
simulation are applied to predict the performance of both existing and non-existing solution,
in terms of lighting performance and visual comfort in buildings. The validated lighting
simulation and rendering software Radiance (Ward, 1994; Ward & Shakespeare, 1998) is
employed as the main tool to (re-)create the model of existing prototypes and non-existing
solutions, as well as to predict the relevant performance indicators.
7
Chapter 1
1.4. Aim and Objectives
The aim of the research in this thesis is predicting the impact of various VNLS
applications on lighting performance and visual comfort in buildings, by means of
computational modelling and building performance simulation.
The main objectives of the research are investigating and enabling innovative application
of modelling approaches for VNLS, which involve:
• Determining the relevant properties and performance indicators for VNLS.
• Finding the appropriate modelling approach for VNLS.
• Evaluating the lighting performance and visual comfort of various VNLS model.
• Finding the potential of applying VNLS under various configurations by predicting their
performance, and under various operating scenarios by estimating total annual electrical
energy consumption.
1.5. Research Methodology
In this thesis, design methods including rapid prototyping and providing multiple design
concepts using computational modelling and building performance simulation are proposed
for obtaining a better design of VNLS. The research deals with two main parts, i.e. existing
VNLS prototypes and non-existing VNLS models. The existing VNLS prototypes are
classified into the ‘first’ and ‘second’ generations, which correspond to those with low and
high directionality. Measurement and simulation of the first generation are performed under
various display settings to validate the lighting performance calculation, whereas simulation
of real window is performed to understand the difference between the prototype and its
corresponding real window. Measurement and simulation of the second generation are also
performed under various display settings to validate the lighting performance calculation, and
to analyse various configurations of the prototypes. In addition, theoretical calculation is
conducted based on the measurement data, to estimate the annual lighting performance and
electrical energy consumption under various operating scenarios.
In turn, results from the existing VNLS prototypes are used as inputs to develop nonexisting VNLS models, which are classified based on view complexity. Simulation is
performed for both models with simplified and complex views, taking various input variables
into account to investigate the influence of each variable on lighting performance, whereas
simulation of real windows is also performed to analyse the difference between models and
their corresponding real windows. For the model with a complex view, additional simulation
is conducted using two modelling approaches, i.e. ‘transmissive’ and ‘emissive’ approaches,
to understand the impact of employing both approaches on the lighting performance of the
VNLS models. Based on these results, conclusions and recommendations for further research
are drawn.
The research framework diagram is illustrated in Figure 1.4.
8
Introduction
Figure 1.4. Research framework diagram
Throughout the process, the following steps are carried out:
• Literature review of current development of VNLS prototypes (existing virtual windows
and skylights).
• Measurement of VNLS prototypes, which are presented in case studies of a prototype with
diffuse light (the first generation prototype) and a prototype with directional light (the socalled second generation prototype).
• Simulation of VNLS prototypes, which are performed using computational building
performance simulation tools to predict the lighting performance. The measurement and
simulation results of the prototypes are incorporated as calibration of the future VNLS
model.
• Computational modelling of future VNLS, which involves arrays of small light sources
with various tilt angles to deliver the light in various directions into the space.
• Sensitivity analysis, to understand the influence of the relevant input parameters on the
relevant performance indicators of the VNLS model.
• Performance comparison with real windows scenes, to understand the benefit of installing
VNLS in a given space, relative to the similar scene with real windows, by comparing the
performance indicators of interest.
9
Chapter 1
• Theoretical calculation, to estimate the annual space availability and electrical energy
consumption of a VNLS prototype under various operating scenarios.
1.6. Thesis Outline
The thesis focuses on the impact of various types of VNLS on lighting performance and
visual comfort in buildings. In addition, to estimate the total electrical energy consumption on
an annual basis, a theoretical study based on measured power consumption data of an existing
VNLS prototype is applied. Figures 1.3 and 1.4 can be referred back to describe the
framework of this thesis. Chapter 2 discusses the left-hand side of the chart in Figure 1.3; by
giving a literature review of existing first generation prototypes. The general concept of
VNLS is discussed, explaining the definition and ambitions of having VNLS installed in a
building.
Chapter 3 presents an example of using Radiance to reproduce the scenes and to evaluate
the lighting performance of a first generation prototype with a simplified view and diffuse
light (lower-left quadrant in Figure 1.3). The performance of the measured prototype was
compared to a hypothetical real window under the same settings in simulation.
Chapter 4 provides an evaluation of discomfort glare from a first generation VNLS
prototype with complex views and diffuse light (upper-left quadrant in Figure 1.3), correlated
to the results of an experiment conducted elsewhere (Shin et al., 2012) on subjective glare
perception from the same prototype. Radiance and Evalglare were applied to recreate the
scenes and evaluate the glare metrics. The results from both the subjective evaluation and
simulation were compared to each other.
Chapter 5 discusses the lighting performance of a second generation VNLS prototype
(between the lower-left and lower-right quadrant in Figure 1.3), in which more directional
light is installed. The lighting performance obtained from measurements in a reference space
was compared to simulation using Radiance, as a basis to calibrate the model that can be
extended for future VNLS. Various possibilities of placing the prototype inside the room
were investigated in Radiance to determine the effect on space availability and visual
comfort. Various operating scenarios were introduced and calculated to determine the effect
on the average space availability and total annual electrical energy consumption that are
produced and consumed by the prototype.
Chapter 6 introduces a new VNLS model providing a simplified view and directional
light (lower-right quadrant in Figure 1.3), and discusses the calculated lighting performance
of the model. The model was created and simulated using Radiance to understand the effect
of varying input variables of the VNLS model on lighting performance of a reference space,
and to compare the lighting performance of the simulated VNLS, relative to that of real
windows under the standard CIE overcast sky.
Chapter 7 discusses the lighting performance of a VNLS model with complex views and
directional light (upper-right quadrant in Figure 1.3). The model was created and simulated
10
Introduction
using Radiance to understand the effect of varying input variables of the more comprehensive
VNLS model on the lighting performance of a reference space. Comparisons of various
image views, as well as two modelling approaches, i.e. the transmissive and emissive
approaches, are shown.
Chapter 8 contains the main conclusions of the thesis and recommendations for further
research.
11
Chapter 1
12
Chapter 2
General Concept of Virtual Natural Lighting Solutions
This chapter discusses the general concept of Virtual Natural Lighting Solutions (VNLS),
including the definition, expectation, classification, as well as state of the art of development.
The VNLS prototypes and models can be generally classified in terms of light directionality
and view complexity. Comparisons of light and view qualities of the existing prototypes are
shown, based on their observed features.
2.1. Definition of VNLS
In most buildings, a real natural lighting solution can be thought of as any opening in the
façade or ceiling of a building, which can bring natural light to the space inside. Two
common examples of elementary natural light openings are the (vertical) window and the
(horizontal) skylight. A window is a transparent opening in the building façade, door or wall,
which allows the passage of light and, if not closed or sealed, air and sound. Windows are
usually glazed or fitted with some other translucent or transparent material like glass. A
skylight is an opening located in the roof, covered with translucent or transparent material
like glass or plastic, which is designed to admit daylight. An example of a complex natural
light opening is a light reflecting tubular device (light pipe).
In the cases where a real natural lighting solution is absent or ineffective, for instance due
to space and time limitation, the concept of Virtual Natural Lighting Solutions (VNLS) can
be promising to overcome the problem of lack of daylight. VNLS are defined here as
‘systems that can artificially provide natural lighting as well as a realistic outside view, with
properties comparable to those of real windows and skylights’.
2.2. Classification of VNLS
A number of efforts have been made to imitate one or more elements of natural light
inside buildings, in the form of artificial solutions. Originally, the efforts were more focused
on bringing the ‘view’ of an outside condition into the room. Attempts to create a realistic
artificial view have been under development for centuries. For example, in art history, trompe
l'oeil is known as an art technique involving realistic imagery to create the optical illusion
that the depicted objects appear in three dimensions, while actually being a two-dimensional
painting. This technique can be traced back to the ancient Greek era around the year 400 BC,
and was developed further mostly by Italian artists between the 15th and 17th century. Despite
being very inspiring, this example is not discussed further in detail, since it is not an actual
light source, nor a device that can transmit light from the outside environment. Nevertheless,
13
Chapter 2
the concept of displaying artificial sceneries of nature is still used in the later form of VNLS
prototypes. Some researchers have shown that artificial views, which do not emit light
themselves, can actually have a positive effect on human health (e.g. Heerwagen & Orians,
1986; Heerwagen, 1990; Ulrich et al., 1993).
Interestingly, the inverse is also true. In its intense appearance without a sufficient view,
artificial bright light can create a positive effect on human well-being and healing (e.g. Glass
et al., 1985; Badia et al., 1991; Avery et al., 1992; Eastman et al., 1998; Lingjærde et al.,
1998; Avery et al., 2001). Specific lighting products have been manufactured to generate
large amounts of light with a particular spectral power distribution for this application. In
general, the idea behind this type of VNLS prototype is to recreate the situation with natural
light and its qualities inside a space, and to harvest the benefit it may offer.
Directionality of the light is another important property that typically distinguishes a real
window or skylight from an artificial version. In fact, directional light is something that rarely
appears in existing VNLS prototypes; most of them only generate light in a nearly diffuse
direction. Therefore, a non-diffuse, or directional, light is considered a key feature that should
appear in an ideal VNLS prototype.
Based on these considerations, any VNLS prototypes (that exist) and models (that do not
yet exist) can be classified based on their light and view qualities, as previously illustrated in
Figure 1.3.
2.2.1. Prototypes with a simplified view
One of the simplest versions of VNLS prototype is the ‘light box’, which is generally
constructed of a series of artificial light sources behind a translucent diffuse surface. In view
of health application, it is known that human bodies use natural (sun-) light to regulate a
variety of functions that affect mood and energy level, cure skin disorders, and make vitamin
D (e.g. Cajochen, 2007; Vandewalle, 2009). Without enough (sun-) light, humans often feel
down, lack energy, and sometimes even suffer physical disorders. To help reduce these
symptoms, specific light boxes have been designed to provide illuminances up to 10000 lx at
a distance of approximately 50 cm, where the individual sits for a specified duration. It has
been shown that the so-called bright light therapy can have a positive effect on human wellbeing and healing (e.g. Glass et al., 1985; Badia et al., 1991; Avery et al., 1992; Eastman et
al., 1998; Lingjærde et al., 1998; Avery et al., 2001; Mottram et al., 2011; van Hoof et al.,
2012). A similar solution uses sets of blue light emitting diodes (LEDs) in a light box,
designed with an enhanced blue spectrum component, based on independent clinical research
showing that blue light from the summer sky can regulate mood and can trigger human
bodies to become active and energetic (e.g. Webb, 2006; Glickman et al., 2006; Viola et al.,
2008; Iskra-Golec et al., 2012). Another new application to create the effect of natural light
uses gradually increasing levels of brightness, to wake up people in the morning in a natural
way.
14
General Concept of VNLS
A number of studies have been performed using such prototype as a method to study
various effects of light and view on subjects. For instance, in their experiments, de Vries et al.
(2009) installed two units of ‘emulated windows’, each measuring 1.20 m × 1.20 m with 12
rows of tubular fluorescent lamps, covered with a diffuse screen (see Figure 2.1). The
experiments were conducted to evaluate the work performance of the subjects, which results
showed that the performance of the test subjects increased when the view was removed and
when daylight was replaced by an artificial light source. It should be noticed however that the
study was only a pilot with a relatively small number of samples (N = 10).
Figure 2.1. Interior view of the room with obstructed windows in the experiments of de Vries et
al. (2009)
Prototypes of the same type were used in the experiments of Smolders et al. (2012),
focusing on the effect of eye illuminance on subjective measures, task performance, and heart
rate variability. The results showed that a higher eye illuminance could improve not only
subjective feelings of alertness and vitality, but also objectively measured performance. The
performance measures suggested that white light could improve performance and yield faster
responses and higher accuracy on simple cognitive tasks. In addition, the exposure to a higher
illuminance can also increase physiological arousal.
Experiments on glare sensation from another prototype with a simplified view were
conducted by Rodriguez & Pattini (2014), observing its effects on glare-sensitive and glareinsensitive subjects when performing a computer task. The results showed that luminance and
size of the window had the same, statistically significant effect on glare sensation for both
groups. However, when occasionally looking directly at the glare source, glare-sensitive
people had a higher relative risk of being disturbed. In all of those mentioned studies, the
prototype was installed to provide the intended light qualities such as vertical illuminance and
view luminance. A prototype that provided not only light with a simplified sky scene but also sunlight has
been developed by Philips (van Loenen et al., 2007). The prototype was a 1.20 m × 1.20 m
luminaire with 12 rows of tubular colour fluorescent lamps. Each lamp could be tuned to
mimic the colour gradients of, for example, the sunrise, noon, or sunset. A high intensity
discharge (HID) spot light was added and could be controlled to mimic direct sunlight. The
view variation of this prototype was higher compared to those mentioned in the previous
15
Chapter 2
paragraph, since there was a possibility to control the colour gradient and to create the
impression of having a spot of sunlight inside the space.
Another prototype with a simplified sky scene and sunlight has been also developed by
ENTPE-EDF Lyon (Enrech Xena, 1999; Fontoynont, 2011a, 2011b). Short- (one day) and
long-term (3 and 10 years) acceptance studies were performed under various colour modes
(Figure 2.2). The results showed a high acceptability in windowless space for long-term use,
individual control was indispensable, and non-natural light spectra were sometimes preferred
in the end of the day.
Figure 2.2. Example of prototype with a simplified view and diffuse light, and a possibility of
adding sunlight, taken from Fontoynont (2011a, 2011b)
To briefly summarise, the type of solutions with a simplified view and mainly diffuse
light can be classified as shown in Table 2.1 as follows.
Table 2.1. Classification of solutions with a simplified view and mainly diffuse light
Source
Features
References
Large brightness, static view
de Vries et al. (2009); Smolders et al.
(2012); Rodriguez & Pattini (2014)
Varying brightness, static view
Enrech Xena (1999); van Loenen et al.
(2007); Fontoynont (2011a, 2011b)
Fluorescent/LED
2.2.2. Prototypes with a complex view
While light from a window is beneficial for building occupants, view is another important
feature of a window (e.g. Collins, 1975; Collins, 1976; Kaplan & Kaplan, 1989; Kaplan,
1993; Farley & Veitch, 2001; Boyce, 2003; Galasiu & Veitch, 2006; Aries et al. 2010). A
number of commercial efforts have been developed to provide a view from a VNLS
prototype, using static, semi-transparent photographs in front of a group of light sources,
mostly fluorescent or LED lamps, an example of which is shown in Figure 2.3. Application
of this prototype can be found in windowless healthcare environments such as critical care
units and magnetic resonance imaging (MRI), particularly to reduce the anxiety of the
patient.
16
Figure 2.3. Example of prototype with a complex view and diffuse light using backlit, transparent
photos showing static image: round skylight for healthcare environment, taken from TESS (2012)
Research regarding subjective discomfort glare from such prototypes has been performed,
for example by Shin et al. (2012) and Kim et al. (2012), using backlit, transparent printed
photographs on top of a light box constructed of incandescent lamp arrays. Experiments on
subjective discomfort glare were also performed by Tuaycharoen & Tregenza (2007), using a
number of screen projected images displaying natural and man-made sceneries. A similar
technique of using projected image on a screen was applied by IJsselsteijn et al. (2008), in
their investigation on subjective depth perception cues. In all of those experiments, the
prototypes/displays were assumed to be the representation of what the subjects normally see
through a real window.
Next to the backlit and projection image technique, other researchers and manufacturers
have utilised electronic large, high-definition (HD) monitor displays for the purpose of
simulating window-views in a more flexible manner. A number of commercial products of
virtual window have been developed (Figure 2.4), consisting of a number of LCD screens,
displaying a recorded, realistic moving image that could be chosen by the users, e.g. Rational
Craft (2013) and Windauga (2013).
(a)
(b)
Figure 2.4. Examples of solutions with a complex view and diffuse light using HD monitor display
showing dynamic image: (a) Winscape ®, taken from Rational Craft (2013) and (b) Windauga ®,
taken from Windauga (2013)
17
Chapter 2
A specific type of VNLS prototype with motion parallax has been developed by Gaver et
al. (1995), for the purpose of remote communication. The motion parallax can be simulated if
the location of the viewer’s head in relation to the display is known. Local head locations
were detected by a tracking camera and were used to control a moving camera in the remote
office. The effect was that the image on the local monitor changed as if it were a window. In
particular, the prototype might offer an expanded field of view and reduced visual
discontinuities. The prototype was however still relatively large, slow, and inaccurate for an
extended application. The number of people who can correctly experience the motion
parallax was also limited to one.
Another research on a VNLS prototype with a complex view was performed by
Radikovic et al. (2005). They presented a system using a head-coupled display and image
rendering to simulate a photorealistic view of nature with motion parallax. A pan-tilt-zoom
camera tracked the observer as long as the face was visible to the camera. Below the camera
was a large display showing the window view that should be seen from the observer’s
position. Evaluation data obtained from test subjects suggested the prototype was a better
window substitute than a static image, and had significantly more positive effects on the
observers’ arousal, positive affects, and interest. The test subjects judged the system
prototype as an acceptable replacement for a real window, and gave it higher ratings for
realism and preference than a static image.
Research on HD monitor displays without motion parallax was conducted, for example
by Friedman et al. (2008) and Kahn et al. (2008). The monitors were installed on the walls of
seven inside offices of faculty and staff at a university, and displayed, as the default image,
real-time views of the immediate outside scene (Figure 2.5). Data were collected over a 16week period to explore the user experience with these large display windows. The results
showed that users deeply appreciated many aspects of the experience. One of the benefits was
the reported increase in users’ connection to the wider social community, connection to the
natural world, psychological well-being, and cognitive functioning.
Figure 2.5. Examples of solution with a complex view and diffuse light, in the forms of HD
monitor display showing real-time, dynamic image, taken from Friedman et al. (2008)
18
To summarise, the classification of solutions with complex view and mainly diffuse light
is shown in Table 2.2 as follows.
Table 2.2. Classification of solutions with a complex view and mainly diffuse light
Source
Features
References
Incandescent/
fluorescent/LED
Transparent printed photograph,
static view
Shin et al. (2012); Kim et al. (2012);
TESS (2012)
Projection
Recorded, static and dynamic view
Tuaycharoen & Tregenza (2007);
IJsselsteijn et al. (2008)
Recorded, dynamic view,
no parallax
HD monitor display
Recorded, dynamic view,
with parallax
Rational Craft (2013); Windauga
(2013)
Gaver et al. (1995); Radikovic et al.
(2005)
Real-time, dynamic view,
no parallax
Friedman et al. (2008); Kahn et al.
(2008)
2.3. Expectation of VNLS
A virtual natural lighting solution should give effects comparable to or even better
effects than the real natural lighting solutions. The latter has a number of essential properties,
which can be generally classified into two categories, i.e. light and view qualities. The light
quality mainly indicates the photometric output of the solution, and normally can be
described in numerical values. The view quality is mostly related to the process of seeing and
perceiving the viewed objects, and may not always be reported in numbers. Nonetheless, they
do have a significant contribution in developing a virtual natural lighting solution.
2.3.1. Light quality
The properties of light quality are derived from characteristics of the real natural light
opening, in its role as a light source. Since natural light is highly variable, the values range of
the properties is also variable, depending on the time and weather condition. Those properties
and their general range are listed as follows.
1. Surface luminance. Luminance of a surface is the luminous flux emitted in a given
direction divided by the projected area of the surface element in that direction. Surface
luminance of a natural light opening (viewed from inside) mostly depends on the sky
condition and the transmittance of the glazing material. For example, under the CIE
overcast sky, the window surface luminance can range from zero until around 3000 cd/m2.
Under the CIE clear sky, the values can be four or five times higher, or even more if direct
sunlight is present. However, based on the findings of Shin et al. (2012), it is preferable to
19
Chapter 2
have an average absolute surface luminance of not more than 3200 cd/m2, that is the value
at which on average people perceive the glare from simulated windows as ‘acceptable’,
i.e. scored as 2.5 out of 4.5 on their discomfort glare scale. On the other hand, it should be
noted that it is usually not the absolute luminance that causes the discomfort, but rather the
difference of luminance values between two adjacent surfaces, i.e. the contrast.
2. Colour temperature. The (correlated) colour temperature is the temperature of the
Planckian radiator whose perceived colour most closely resembles that of a given stimulus
at the same brightness, under a specified viewing condition. Natural light has a wide range
of CCT (Chain et al., 1999, 2001), from warm colours (red to yellowish white, during
sunrise/sunset: 2700 ~ 3000 K) until cool colours (bluish white, during sunny day/around
noon: 5000 ~ 6500 K; and blue, during overcast day or very blue sky condition: 6500 ~
20000 K), which varies over time in a day. Note that for electric lighting, the colour
temperature is predominantly constant, as opposed to natural lighting.
3. Colour quality. Natural light generally shows a ‘complete’ spectral power distribution
(SPD), while artificial light sources have an ‘incomplete’ SPD, e.g., incandescent lamps
usually lack on blue light component, fluorescent lamps have mercury spikes at some
wavelengths which outnumber the quantity of other wavelengths. Bouma (1948) described
natural (day-) light as the ideal source of illumination for good colour rendering, because
‘it displays a great variety of colours, makes it easy to distinguish slight shades of colour,
and the colours of objects around us obviously look natural’.
For artificial light sources, the CIE Colour Rendering Index (CRI) has been widely used
for many years to measure their ability to reproduce the colours of various objects, in
comparison with an ideal (blackbody radiator) or natural light source. It ranges from 0 to
100. However, the CRI is 40 years old and various problems with the CRI when used for
light-emitting diode (LED) sources have been identified, as reported in many publications
(e.g. Rea & Freyssinier-Nova, 2008; Zukauskas et al., 2009; Davis & Ohno, 2009; Davis
& Ohno, 2009).
A new alternative to describe colour rendering is the Colour Quality Scale (CQS)
developed by The National Institute of Standards and Technology (NIST) (Ohno & Davis,
2010). It addresses the problems of the CRI for solid state lighting sources, also has 0 ~
100 scale, and yet maintains good consistency of scores with the CRI for traditional
sources. Nevertheless, more concepts have been proposed earlier and later, and yet there is
no consensus found on this issue.
4. Directionality. The directionality of light is a balance between the directional and diffuse
components within the luminous environment. In the design stage, these properties are
normally described with the sky component (SC [%]), externally reflected component
(ERC [%]), and internally reflected component (IRC [%]); which are illustrated in Figure
2.6. This approach is known as the split-flux method, and is applied to predict the daylight
factor (DF [%]) at any given point inside a building, according to Equation 2.1.
20
DF = SC + ERC + IRC (2.1)
(a)
(b)
(c)
Figure 2.6. Section view of a building with a calculation point (U) and the presence of (a) sky
component, (b) externally reflected component, and (c) internally reflected component
21
Chapter 2
Under the CIE clear or intermediate sky, light comes from a certain direction and may
create shadow. Under the CIE overcast sky, which is the assumption in daylight factor
calculation, the sky provides diffuse light coming in uniform direction (circle-shaped
luminous intensity distribution), but the shape and surface of the glazing may still direct
the incoming light.
Related to the built environment, Inanici & Navvab (2006) defined directionality as a ratio
between directional and diffuse luminance components in a given space. It gives an
indication about the spatial distribution of light flow onto an element or into a space. They
suggested an image subtraction method for obtaining the diffuse and directional
components of light, by analysing two images of a given luminous environment scene; one
image includes the directional and diffuse components of the luminous environment, while
the other excludes the diffuse component. The ratio of the directional and diffuse
components is then obtained from the average luminance values of both images. The
target range is between 1.4 : 1 (typical CIE overcast sky) and 2.5 : 1 (typical CIE clear
sky).
However, that indicator depends largely on the environment setting (e.g. room dimension,
indoor surface reflectance) where the solution is applied. For describing the characteristic
of the existing prototypes, parameters with more qualitative levels are proposed, classified
as follows: A (the best choice; generates non-uniform, directional light and possibility to
control or vary the directionality), B (generates non-uniform, directional light but no
possibility to control or vary the directionality), C (generates a uniform pattern of diffuse
and directional light), and D (the worst choice; generates a uniform pattern of only diffuse
light).
2.3.2. View quality
The view quality is defined as quality of the outside view (image) experienced by the
viewer in the room. Studies on what kind of component should be present in the viewed
image have been done previously by many researchers (e.g. Ulrich, 1984; Ulrich et al., 1991;
Tennessen & Cimprich, 1995; Chang & Chen, 2005; de Kort et al., 2006; Aries et al., 2010;
Beute & de Kort, 2013). In their experimental studies, Tuaycharoen & Tregenza (2007) stated
that “view is simply one of the ways we interpret or perceive the light that flows in through
the window”, and therefore cannot be separated from the natural (day-) light itself. They
concluded that a good view (also referred as a view with high interest), which mainly consists
of natural scenes, tends to reduce glare perception.
Related to daylight and view, Hellinga & de Bruijn-Hordijk (2009) proposed certain
quality levels for themes that influence visual comfort. They proposed parameters with
qualitative levels, which are classified as: A (the absolute best choice for that parameter), B
(good), C (sufficient), and D (insufficient); as shown in Table 2.3. The values for the
22
different quality levels are based on values found in the literature (e.g. Kaplan & Kaplan,
1989; Tregenza & Loe, 1998).
Table 2.3. Quality levels for parameters of view that influence visual comfort, adapted from
Hellinga & de Bruijn-Hordijk (2009)
Parameter
A
B
C
D
- Green, sky, and
distant objects
The view
contains all 3
elements
The view
contains 2 of
the 3 elements
The view
contains 1 of
the 3 elements
The view
contains none
of the 3
elements
- Information
The view
gives
maximum
information
about outside
environment:
weather, season, time of
day, activities
The view
gives
information
about
weather,
season, time
of day, and
activities
The view
gives
information
about
weather,
season, and
time of day
The view
gives no
information
- Organisation
The view is
highly
complex and
coherent
Medium
complexity
and coherence
Low
complexity
and coherence
The view is
simple and
incoherent
The parameters proposed by Hellinga & de Bruijn-Hordijk (2009) were defined
specifically for windows. In their analyses of windows, ‘green’ has a specific meaning and
importance as natural elements, irrespective of being nearby or distant. In this thesis, VNLS
covers not only windows, but also skylights. For example, ‘green’ in a virtual skylight
display can be the leaves and branches of a tree (see Figure 2.3), whereas there are more
examples for those in a virtual window display. In terms of information, activity through a
skylight is rarely human, but can for example be flying birds.
In addition to the view quality, IJsselsteijn et al. (2008) performed experiments on the
efficacy of three ‘depth perception cues’, i.e. motion parallax, occlusion, and blur, using
projected photorealistic scenes to create a window-like ‘see-through experience’, as
illustrated in Figure 2.7. The three cues can be briefly explained as follows:
1. Motion parallax is the phenomenon that occurs as one moves his/her body (or only head)
from side to side, more distant objects will traverse smaller angles across the retina than do
objects closer by (Markus, 1967; IJsselsteijn et al., 2008). Technically, it is an apparent
displacement in the position of an object viewed along two different lines of sight, and can
be measured by the angle or semi-angle of inclination between those two lines.
23
Chapter 2
2. Occlusion or window framing is the addition of a frame, which is expected to provide
additional depth information regarding the position (depth layer) of the frame related to the
outside view, via the occlusion or interposition cue, particularly in the case where motion
parallax is present. For more details, see Cutting & Vishton (1995).
3. Blur is the addition of something that is hazy and indistinct to the boundaries of the frame,
which would give a signal to the visual system that the frame is located at a different depth
layer than the view being displayed (IJsselsteijn et al. 2008).
Figure 2.7. Schematic representation of the experimental 2 × 2 × 2 design by, and taken from
IJsselsteijn et al. (2008), varying in blurring of the frame, the presence of an occluding cross-shaped
frame, and the presence of movement parallax
The results of IJsselsteijn et al. (2008) indicated all of the three cues have a significant
main effect on the viewer’s see-through experience, with motion parallax yielding the
greatest effect size (F(1,19) = 24.86, p < 0.001). Following the classification of three
aforementioned properties, the depth perception cues therefore can also be evaluated in four
levels as follows:
• Level A: the view contains all of the three depth perception cues
• Level B: the view contains two depth perception cues
• Level C: the view contains one depth perception cues
• Level D: the view contains none of the three depth perception cues
2.3.3. Target range
Based on the discussed expectation in Sections 2.3.1 and 2.3.2, the target range of the
VNLS properties can be summarised in terms of light and view qualities. In this section, the
light quality is described by the surface luminance (Ls), colour temperature (CCT), colour
quality scale (CQS), and directionality (DIR), whereas the view quality is described by
24
presence of green, sky, and distant objects (GSD), information (INF), organisation or
complexity and coherence (ORG), and depth perception cues (DPC). Table 2.4 summarises
the possible and target range of VNLS properties as follows.
Table 2.4. Target range of VNLS properties
Symbol
Unit
Possible
range
Target
range
Surface luminance
Ls
cd/m2
0~∞
1000 ~ 3200
Colour temperature
CCT
K
2700 ~ 17000
2700 ~ 6500
Colour quality scale
CQS
-
0 ~ 100
71 ~ 93
Directionality
DIR
-
A, B, C, D
A or B
Presence of green, sky, and
distant objects
GSD
-
A, B, C, D
A or B
Information
INF
-
A, B, C, D
A or B
Organisation
ORG
-
A, B, C, D
A or B
Depth perception cues
DPC
-
A, B, C, D
A or B
Properties
Light quality
View quality
2.3.4. Comparison of prototypes
In order to compare the light and view qualities, the existing prototypes that have been
discussed in Section 2.2 are classified based on their specific features. The classification is
intended to be generic, and does not specifically point to the products taken as examples. All
values, particularly the ones related with light qualities, are roughly estimated based on given
specification from which the properties in question are derived. For instance, the luminous
flux of the light source Φv [lm] may be given in the product specification, or can be estimated
by multiplying the light source’s electrical power with its typical luminous efficacy. For
diffuse (or very low directionality) light source, the light is transmitted approximately evenly
over the entire hemisphere, i.e. the solid angle Ωi is equal to 2π sr. The luminous intensity
therefore can be predicted, and hence also the surface luminance Lv [cd/m2], expressed as
follows:
Lv =
v
 i Ai
where Ai is the projected light source surface [m2].
25
(2.2)
Chapter 2
Regarding the view quality, it is obvious that the prototypes with a simplified view obtain
the lowest score (level D), since most of them display very little or even no view at all. For
the prototypes with a complex view, the view quality is estimated by observation. The
summary of the comparisons is given in Table 2.5.
Table 2.5. Comparisons of the properties of existing prototypes with a simplified view + mainly
diffuse light and a complex view + mainly diffuse light
Properties
Features
Simplified view,
diffuse light
Fluorescent/LED,
large brightness,
static view
Fluorescent/LED,
varying brightness,
static view
Complex view,
diffuse light
Incandescent/
fluorescent/LED,
transparent
photograph, static
view
Ls
[cd/m2]
≤ 10000
≤ 10000
Light quality
CCT
CQS
[K]
[-]
2700
~6500
2700
~6500
View quality
INF
ORG
[-]
[-]
DIR
[-]
GSD
[-]
DPC
[-]
64~93
D
D
D
D
B
64~93
C
C
C
C
B
≤ 10000
2700
~6500
64~93
D
A~B
C
A~B
C
Projection, recorded,
static view
≤ 500
6500
71~93
D
A~B
C
A~B
A
HD monitor,
recorded, dynamic
view, no parallax
≤ 1000
5500
71~93
~10500
D
A~B
B
A~B
C
HD monitor,
recorded, dynamic
view, with parallax
≤ 1000
5500
71~93
~10500
D
A~B
B
A~B
B
HD monitor, realtime, dynamic view,
no parallax
≤ 1000
5500
71~93
~10500
D
A~B
A
A~B
C
A~B
A~B
A~B
A~B
A~B
Ideal VNLS
1000
~3200
2700
~6500
71~93
Referring to the description in Sections 2.3.1 and 2.3.2, the following explanation can be
given:
26
1. Surface luminance: A number of artificial light sources can be combined to create a
prototype generating up to 10000 cd/m2 of luminance, for instance in the case of the light
box display of Shin et al. (2012). A single HD monitor display normally gives less than
10% of that amount. An image projection normally gives a very low luminance (less than
500 cd/m2), considering the reduction of light after reflection from the (diffuse) screen.
2. Colour temperature: Most prototypes with transparent printed photographs generate light
with colour temperatures between 2700 K and 6500 K, depending on the type of the light
source. The high intensity lamps for healing purpose can give a very bright light, which is
required to cure diseases such as winter blues. Colour temperature of this type of lamp is
usually very high; in some cases can be up to 17000 K, to enhance the blue light
properties. Most HD monitor displays generate light with high colour temperatures,
somewhere between 5500 K and 10500 K. Reducing the colour temperature gives the
entire screen an increasingly reddish cast, while increasing it makes the colour cast
increasingly blue. For ordinary personal computer use, a colour temperature of 6500 K is
standard. For television broadcasting, the value is defined based on national standards;
most countries use a colour temperature standard of 6500 K according to the US
broadcasting standard (NTSC), whereas 9300 K is used under the Japanese standard
(NTSC-J); see more details in EIZO (2013).
3. Colour quality: In their experiment on a number of samples, Ohno & Davis (2010)
showed that most recent phosphor LED type products had a CQS of 71 ~ 93. Their tested
fluorescent type products had a CQS of 64 ~ 80, whereas an incandescent type had 98. An
HD monitor display is assumed to have a similar colour quality as LED type products.
4. Directionality: Most prototypes do not feature any directional light at all. The prototype
of Philips (van Loenen et al., 2007) included an HID lamp for simulating the sun position.
This lamp created the impression of a sun ‘spot’ on the blurred display and in the room,
although smaller in size and weaker in luminous intensity than real sun patches.
5. Presence of green, sky, and distance objects: Prototypes with a simplified view mostly
display none of the three layers. The prototype of Philips (van Loenen et al., 2007)
displayed a blurred sky scene by varying the intensity of each coloured light source
behind the diffuser, hence essentially showed one layer. Most of the prototypes with a
complex view display at least two layers, i.e. sky and distance objects or sky and green
objects.
6. Information: Obviously there is very little to see from the prototypes with a simplified
view, except vague information about the weather, depicted on the sky display. Most of
the prototypes with a complex view display sceneries from which the weather, season,
and time of day can be deduced. In the case of a static view, the information is
unchanged, whereas in the case of recorded, dynamic view, the information may be
repeated after some time. Table 2.3 describes human activities as important information
to display; since human activities are constantly changing, a dynamic view is required.
27
Chapter 2
The maximum information about the outside environment is obtained when a real-time
view is presented, ensuring no pattern repetition.
7. Organisation: In line with the information property, there is very little or no complexity
on most of the prototypes with a simplified view. Prototypes with a complex view mostly
display sceneries of medium or high complexity (see Figures 2.4 and 2.5).
8. Depth perception cues: All of the prototypes with a simplified view do not feature any
motion parallax, but mostly have a blurred display. The occlusion effect can be present
when the prototype is put behind window frames. Most of the prototypes with a complex
view do not come with motion parallax either, but do have occlusions or window framing.
In the prototype with a transparent photograph and a static view, the frame boundaries are
very often not enough to give an impression that there is a different depth layer between
the frame and the view. In the prototype with projected image, the three depth perception
cues can be created (IJsselsteijn et al. 2008). In the prototype with a recorded dynamic
image, the occlusion effect is present, but the blur effect is missing, as shown for instance
in Figure 2.4.
This overview shows that an ideal VNLS does not yet exist at the moment. The existing
prototypes have their own limitations, and each prototype addresses only a subset of all
aspects required for an ideal VNLS. For instance, directionality of the incoming light is an
issue to be addressed, since most displays will provide only diffuse light, and it is very
challenging to imitate the constantly moving direction of natural light. In some prototypes,
the light direction may be varied by aiming the lamps behind the display into different angles,
but this also needs a sophisticated control system.
Dynamics of the view and motion parallax are even harder to imitate. A video display
may create a moving image, but the given motion parallax effect is mostly limited to one
position in front of the display, depending on the number of sensors installed. High detailed
images may not be required for skylights, but it is required for the vertical windows. A very
complex view can be provided by an HD monitor display, either recorded or real-time, but
the light quality criteria are yet to be satisfied. One of the most interesting challenges in this
direction is probably to create an HD monitor display with real-time image at a sufficient
depth to the frame, sufficient light output, and also with the appearance of motion parallax.
Looking at the discussed aspects, direction of further development should be steered
toward improving the light directionality and view dynamics. In order to approach the ideal
condition, a number of evaluation stages must be performed, including theoretical analysis,
initial design, numerical testing of the design, prototype construction, physical testing,
subjective laboratory testing, field trials, and so on. In the early design stage, computational
modelling and simulation is a powerful tool to predict the system performance, with regards
to the relevant physical phenomena. By using computational modelling and simulation, one is
able to rapidly test multiple design concepts for better solutions, in an efficient way in terms
of time and cost.
28
A number of efforts have been made to recreate the elements of natural light inside
buildings, in the form of artificial solutions. Such solutions, the so-called VNLS, can be
generally classified based on their light and view qualities into four types, which are those
providing: (1) simplified view and diffuse light, (2) complex view and diffuse light, (3)
simplified view and directional light, and (4) complex view and directional light. The first
two types already exist in reality as prototypes, while the last two are still under development
in conceptual models. Building performance simulation tools have their role to predict the
performance. Most prototypes with a simplified view appear in the shape of ‘light box’,
which is generally constructed of a series of artificial light sources behind a diffuse surface.
Despite the insufficient view on their display, research in health-related contexts has shown
that installation of this particular prototype, under a large brightness setting, can be useful for
the purpose of healing or therapy. Adding a complex view on VNLS prototypes is also
beneficial, for example to reduce the anxiety of patients in windowless healthcare
environments. Prototypes with a complex view are also available for entertainment purpose
as well as for application in office environments, to provide the feeling of being connected to
the outside world.
Based on the given overview of relevant properties, further research and development
should be directed toward improving the light directionality and view dynamics of the
prototype. At the current situation, all of the existing prototypes only address a part of the
ideal VNLS properties. Particularly in the early design stage, computational modelling and
building performance simulation can help steering the process of VNLS design development,
with regards to the relevant physical phenomena. In this context, the use of computational
modelling and simulation can help to rapidly test multiple design concepts in an efficient way
in terms of time and cost, even though the results still need to be validated by real user
experiments.
Finally, it is learned that a number of subject-based experiments have been performed
using various simple forms of VNLS prototype, to gain knowledge on how people perceive it,
and/or to investigate which aspects of natural light people appraise in reality. Nonetheless,
very little is known about how the prototypes physically influence the indoor lighting
condition where the prototypes are installed. This suggests another research direction, which
is to evaluate the objective performance of the prototypes, and to propose better design
solutions to improve it.
29
Chapter 2
30
Chapter 3
Measurement and Simulation of a First Generation Virtual
Natural Lighting Solutions Prototype
This chapter discusses an example of application of Radiance as a simulation tool to
reproduce the scenes and to evaluate the lighting performance of a first generation VNLS
prototype with a simplified view and diffuse light displaying various sky scenes, located in a
test room with no façades. The performance of the measured prototype was compared to a
hypothetical real window under the same settings in simulation.
3.1. Introduction
The overview of various examples of prototypes in Chapter 2 shows that an ideal VNLS
does not yet exist at the moment. In order to approach the ideal condition, a number of
evaluation stages must be performed, such as theoretical analysis, initial design, numerical
testing of the design, prototype construction, physical testing, subjective laboratory testing,
field trials, and so on. In the early design stage, computational modelling and simulation is a
powerful tool to predict the system performance, with regards to the relevant physical
phenomena. Computational modelling and simulation are subsets of design methods that can
be used to rapidly analyse multiple design concepts, and to predict their performance.
In the context of this thesis, it is intended to know how a certain VNLS prototype will
influence the indoor lighting condition and visual comfort under certain scenarios. Moreover,
it is also intended to further improve the performance towards the ideal condition, by
introducing better design options. Therefore, there is a need to create a representative model
of the prototypes, and to predict their performance by mean of simulations. For a lightingand view-related system such as VNLS, a sophisticated lighting modelling and simulation
technique is therefore needed.
3.1.1. Modelling concept in Radiance
In general, lighting simulation technique can be divided into two main types: the first one
is photorealistic rendering, which produces images that are mostly used for artistic
impressions or perception studies, and the second one is physically-based visualisation, which
produces a physically accurate representation and predicts reality under given conditions
(Ward & Shakespeare, 1998; Moeck & Selkowitz, 1996; Ochoa et al., 2012). Each technique
has algorithms and their supporting calculation methods, which also come with their own
specific applications and limitations (Ochoa et al., 2012). The most generally used algorithms
are:
31
Chapter 3
1. Direct calculations used for artificial lighting; these are specific physical formulas and
simplifications, often delineated in national standards to cover most usual illumination
situations. The algorithms are simple and often used as rules-of-thumbs, but can lack
accuracy in the real situation.
2. View-dependent algorithms; these are classified based on direction from which tracing
rays are computed; i.e. from the light source (forward raytracing), from the observer’s eyes
(backward raytracing), or from light source and observer (bidirectional raytracing). They
are used for lighting calculations and renderings, and require a specific observer position.
3. Scene-dependent algorithms; these are mainly radiosity calculations, adapted from heat
transfer techniques. They are mainly used for calculations but not for rendering due to
complex formulas.
As reported by Maamari et al. (2005) and Maamari et al. (2006), the CIE technical
committee 3.33 has defined a set of simple test cases, based on analytical or experimental
references, with the objective of assessing lighting computer programmes. It was shown that
the use of the CIE test cases allows verifying the accuracy level of the tested programmes,
with respect to physical laws in lighting. They observed that for Radiance (Ward &
Shakespeare, 1998), good accuracy was observed in general, except for the indirect lighting
test with reflectance values of 0.8 and above. However, they suggested that a single ideal
lighting programme does not exist, but some are adapted to given tasks and constraints. For
the purpose of daylighting modelling and simulation in particular, Radiance has been
validated elsewhere (e.g. Mardaljevic, 1995, 1997; Reinhart & Herkel, 2000; Reinhart &
Walkenhorst, 2001; Reinhart & Andersen, 2006).
In general, Radiance calculates the outgoing radiance (Lo) as the sum of total reflected
radiance (Lr) and emitted radiance (Le) from a surface (dA) to another surface, defined by the
radiance equation as follows:
Lo (ψo ,θo) = Le (ψo ,θo) + Lr (ψo ,θo ,ψi ,θi)
(3.1)
where Lo (ψo, θo) [W/(sr·m2)] is the outgoing radiance in a direction given by the angles ψo and
2
θo expressed in [rad]; which is composed of Le (ψo, θo) [W/(sr·m )], the emitted radiance, and
2
Lr (ψo, θo, ψi, θi) [W/(sr·m )], the reflected radiance as function of the incoming angles ψi and
θi and outgoing angles ψo and θo, all given in [rad]. In turn, the reflection is further defined as
the function of the incoming radiation (Li):
Lr (ψo, θo, ψi, θi) =   fr ( o , o , i ,  i )Li ( i ,  i ) cos  i sin  i d i d i
 i i
(3.2)
where fr (ψo, θo, ψi, θi) [sr-1] is the reflection as a function of outgoing angles ψo and θo and the
incoming angles ψi and θi [rad], and Li [W/(sr·m2)] is the incoming radiance from a specific
projection, as a function of the incoming angles of the incoming ray representing that
32
Measurement and Simulation of a First Generation VNLS Prototype
projection. The incoming radiance is integrated over the incoming angles or solid angle Ωi
[sr] of the incoming radiance, as illustrated in Figure 1.
Figure 3.1. Schematic illustration of radiance equation and the corresponding variables in
Equations (3.1) and (3.2), adapted from Pelzers et al. (2014)
Emission of a given light source can be modelled with a material type or obtained from
photometric measurements. While experimental data for visible light is normally obtained
with a moving-cell photometer, the emission data may also be approached with the ‘light’
material type. Normally, it is applied as general light source material which emits radiation
(Li [W/(sr·m2)]) on which the material type is applied, defined by the three radiance values:
2
LR, LG, and LB, for the red, blue and green component respectively [W/(sr·m )]. The radiation
is then related to the total radiative flux of light source (Φi [W]) and the projected light source
surface (Ai [m2]) by the following equation:
Li =
i
 i Ai
(3.3)
In principle, Radiance solves the radiance equation for the red, green, and blue (RGB)
values separately to obtain the radiance Li [W/(sr·m2)], or the irradiance IR,G,B [W/m2], if
integrated over the solid angle Ωi [sr]. When a picture is rendered, the spectral irradiance
values in red, green, and blue (IR, IG, IB, respectively) are summed and weighted to obtain the
single value of IR,G,B, according to Ward & Shakespeare (1998):
I R,G,B = 0.265 IR + 0.670 IG + 0.0648 IB
(3.4)
Assuming a conversion factor of 179 lm/W (Ward & Shakespeare, 1998), the irradiance
values can be easily converted to illuminance (E [lx]) as follows:
E = 179 I R,G,B = 179 (0.265 IR + 0.670 IG + 0.0648 IB)
33
(3.5)
Chapter 3
Recent development within the Radiance community has led to application of spectral
rendering with Radiance for other purposes (Geisler-Moroder & Dur, 2010). In fact,
Ruppertsberg & Bloj (2006) showed that spectral rendering could actually be more accurate
than the original Radiance RGB model. Nevertheless, Equations 3.4 and 3.5 remain as the
governing equations, which in most cases are applicable to determine illuminance values on
individual sampling points.
Another important feature in Radiance is the ambient calculation, which consists of all
diffuse and indirect specular reflection, diffuse transmission, and emitted light from the
secondary light sources. The main ambient parameters are ambient bounces, ambient
divisions, ambient super samples, ambient accuracy, and ambient resolution. A more detailed
explanation on these parameters can be found in Ward & Shakespeare (1998). In the context of this thesis, Radiance is employed to address indoor lighting and visual
comfort aspects of VNLS. Very little is known about how these solutions influence the
indoor condition in various scenarios; for instance how the image variation affects the
lighting performance on the workplane. Another important question is how a VNLS
prototype actually correlates to a real window or skylight; could a VNLS prototype perform
as good as, or even better than the real one? A comparison to the real window is then required
on that aspect; such a comparison will be useful for designing a better solution in the future.
Therefore, the study in this chapter aims to address lighting measurements and simulation
of a ‘first generation’ VNLS prototype. The objective is to evaluate the lighting performance
of the prototype under various settings, to analyse if the results can be accurately reproduced
in simulation, and to compare the performance with the corresponding simulated real
windows.
3.2. Case Description
An example of the so-called first generation prototype is the one developed by Philips
(van Loenen et al., 2007), which is briefly discussed in Section 2.2.1. Due to the possibility to
vary the view display and to add directional light using a spot lamp for sun simulation, this
prototype was selected as the first case study to demonstrate how Radiance can be employed
to reproduce the scenes and obtain the lighting performance of the space, validated by actual
measurements.
The prototype was installed in a kitchen laboratory setting, located in the ExperienceLab
of Philips Research in Eindhoven, the Netherlands. The prototype was constructed of 12
colour tubular fluorescent (TL5) lamps of 54 W each, put in an array of 12 rows, and covered
with a diffuse panel of 1.20 m × 1.20 m. A halogen, parabolic aluminised reflector (PAR)
spot lamp of 70 W was installed in the upper right corner to simulate the sunlight.
The construction was put vertically in an adjacent control room behind a transparent,
clear glass window which was a part of the kitchen room interior. During the experiment, the
general lighting in the kitchen was switched off all the time. The room had no façades and
34
real windows, ensuring no daylight admission. No motion parallax was associated with this
prototype.
The TL5 lamp array was covered by a white, diffuse panel, installed 0.35 m behind the
window glass plane. The dimension of the diffuse panel was 1.20 m × 1.20 m, while the
window opening was 0.65 m × 0.65 m. The 12 TL5 lamps were divided into four groups;
each group consisted of three lamps emitting red, green, and blue light, respectively. Every
lamp had its own ballast so that it could be dimmed independently, using the Digital
Addressable Lighting Interface (DALI) system. The overcast, clear, and partly cloudy sky
scenes were realised by adjusting the intensity of each lamp and were subjectively evaluated
to imitate the real sky scenes. Table 3.1 shows the type of colour emitted by each lamp, the
electrical power rating, and the intensity level settings for the three scenes.
Table 3.1. Intensity level settings for the three sky scenes of the prototype
Lamp’s
row (from
top)
1
2
3
4
5
6
7
8
9
10
11
12
n/a
Type
Red
Green
Blue
Red
Green
Blue
Red
Green
Blue
Red
Green
Blue
PAR
Power
rating
[W]
54
54
54
54
54
54
54
54
54
54
54
54
70
Overcast
Intensity level
[%]
30
30
30
30
30
30
20
45
30
20
20
20
0
Clear
Intensity level
[%]
15
3
100
80
15
100
0
100
100
80
0
3
100
Partly cloudy
Intensity level
[%]
15
15
100
80
80
100
0
100
100
80
80
0
100
Three scenes were defined for the lamp setting to simulate the colour gradients of typical
sky conditions, i.e. ‘overcast’, ‘clear’, and ‘partly cloudy’. The interior, equiangular fisheye
views of the kitchen room under the three scenes of the prototype are shown in Figure 3.2. In
general, the overcast scene (the PAR spot lamp was turned off) produced a white-grey
appearance on the window, while the other two scenes (the PAR spot lamp was turned on)
produced a combination of light blue and red appearance with a bright spot in the upper right
corner, suggesting the sun’s position, which was not moved during the experiment.
35
Chapter 3
(a)
(b)
(c)
Figure 3.2. Interior view of the kitchen room with the prototype under the (a) overcast, (b) clear,
and (c) partly cloudy sky scenes
The actual lighting performance was measured and obtained by collecting the following
data at certain lighting conditions:
• Horizontal illuminance on the workplane; data were collected for 55 horizontal points on
the workplane height, i.e. the countertop in the kitchen (0.95 m from the floor).
• Vertical illuminance on the observer’s eye plane; data were collected for two vertical
points on the typical observer height (1.20 m from the floor).
• Minimum, maximum, and average luminance perceived by the observer; data were
collected for two vertical points on the typical observer height (1.20 m from the floor).
• Reflectance of interior surface materials; data were collected for the relevant interior
surface, such as floor, walls, ceiling, and furniture.
Furthermore, the horizontal illuminance data were post-processed to obtain the average
illuminance values (Eav [lx]), the uniformity (U0), and the space availability (%A [%]). The
latter is defined as the percentage of the measuring points satisfying a minimum illuminance
value of 500 lx, which is the required indoor illuminance criterion for kitchens (CEN, 2002).
These three indicators can be expressed in equations as follows:
N
 Ei
Eav =
U0 =
%A =
i 1
N
E min
E av
N E  500 lx
× 100%
N
36
(3.6)
(3.7)
(3.8)
where Ei [lx] is the horizontal illuminance on each measuring point, Emin [lx] is the minimum
horizontal illuminance, N E  500 lx is the number of measuring points satisfying the criterion
of minimum illuminance value of 500 lx, and N is the total number of measuring points.
To evaluate the visual comfort in this case, the Daylight Glare Probability (DGP)
(Wienold & Christoffersen, 2006) was used as an indicator, which can be expressed as
follows:
2

n Ls,i s,i

DGP = 5.87 × 10 Ev + 9.18 × 10 log 2  1   1.87 2

Pi
i  1E v

–5
–5





(3.9)
where Ev is the total vertical eye illuminance [lx], ωs is the solid angle of the glare source [sr],
2
Ls is the glare source luminance [cd/m ], and P is the position index, i.e. a weight factor based
on position in a viewing hemisphere.
During the measurement, the following instruments were used:
• SpectraDuo PR-680 photometer; for measuring luminance and illuminance values, as well
as spectral power distribution.
• Canon EOS50D digital single-lens reflex camera + Sigma 4.5mm fisheye lens + Photolux
3.1 software; for taking multiple (20 in this case) photographs in equiangular 180° view
with various exposure values, which in turn were post-processed to obtain the luminance
pictures. The luminance values were calibrated with the SpectraDuo photometer.
• Konica Minolta CM-2600D spectrophotometer; for measuring reflectance values of the
interior surface materials.
Horizontal illuminance data were collected on 55 points at a height of 0.95 m (countertop
level) as displayed in Figure 3.3. Vertical illuminance and luminance perceived by the
observer were measured by taking 20 photographs (ISO 400, f/5.6, shutter time varied from 4
s to 1/8000 s) each at positions 1 and 2, at a height of 1.20 m, with the view direction
specified by the arrows in Figure 3.3.
To determine the glare index value at both observer’s positions, the obtained photographs
were exported to Radiance, combined into High Dynamic Range (HDR) images using the
Hdrgen programme, and analysed using Evalglare (Wienold & Christoffersen, 2006).
37
Chapter 3
Figure 3.3. Floor plan of the kitchen with the measuring points for horizontal illuminance
Since the test room was not connected to the building’s façade, the condition under a real
window could not be observed. Therefore, the real window scene was modelled and
simulated in Radiance. In addition, the actual conditions under all scenes of the prototype
were also modelled and simulated, to give an insight in the difference between simulation and
actual measurement. Comparisons were made between the values of horizontal illuminance at
the central line, where points P1 and 1 were located (i.e. the blue-coloured points on Figure
3.3). The difference between the average illuminance, uniformity, and space availability was
also evaluated.
The front, top, and perspective views of the modelled prototype are displayed in Figure
3.4. The 12 TL5 lamps were modelled as 12 rows of cylinders, with a length of 1.20 m and a
diameter of 0.016 m, constructed with a ‘light’ material. Assuming a total luminous flux of
4250 lm for each lamp (Philips, 2013a), a conversion factor of 179 lm/W between
photometric and radiometric units (Ward & Shakespeare, 1998), and a solid angle of the
incoming radiation of π sr (Ward & Shakespeare, 1998), Equation 3.3 was applied to obtain
the total radiance value of each lamp, i.e. 394 W/(sr·m2) at the maximum setting.
Equation 3.4 was applied to obtain the red, green, and blue radiance components for the
‘light’ material. For the red-coloured lamps, the green and blue radiance components were
assumed to be zero; for the green-coloured lamps, the red and blue were assumed to be zero;
and for the blue-coloured lamps, the red and green were assumed to be zero. Hence, at the
maximum setting, the red-coloured lamps were set to have a red component of 1487
W/(sr·m2), the green-coloured lamps have a green component of 588 W/(sr·m2), and the bluecoloured lamps have a blue component of 6059 W/(sr·m2). For other settings, the values were
adjusted proportionally.
38
The PAR lamp was modelled as a thin cylinder with a diameter of 0.12 m, aimed at an
angle of 45°, and constructed with a ‘light’ material. Assuming a total luminous flux of 1415
lm (Philips, 2013b), and by applying Equation 3.3, a total radiance value of 223 W/(sr·m2) is
obtained. The red, green, and blue components were assumed to be equal.
Table 3.2 displays the assigned values for the light sources in the prototype.
(a)
(b)
(c)
Figure 3.4. (a) Front, (b) top, and (c) perspective views of the modelled prototype. The TL5
lamps and the diffuse panel are coloured in red, the PAR lamp is coloured in green.
Table 3.2. Red, green, and blue irradiance components of ‘light’ material defined in Radiance for
the 12 TL5 lamps in the prototype
Lamp’s
row
(from top)
Red
Green
Blue
Red
Green
Blue
Red
Green
Blue
1
446
0
0
223
0
0
223
0
0
2
0
176
0
0
18
0
0
88
0
3
0
0
1818
0
0
6059
0
0
6059
4
446
0
0
1189
0
0
1189
0
0
5
0
176
0
0
88
0
0
470
0
6
0
0
1818
0
0
6059
0
0
6059
7
297
0
0
0
0
0
0
0
0
8
0
265
0
0
588
0
0
588
0
9
0
0
1818
0
0
6059
0
0
6059
10
297
0
0
1189
0
0
1189
0
0
11
0
118
0
0
0
0
0
470
0
12
0
0
1212
0
0
182
0
0
0
PAR
-
-
-
223
223 223 223 223 223 Overcast
Clear
39
Partly cloudy
Chapter 3
The detailed values assigned for the window construction properties are specified in
Table 3.3, together with the room’s interior surfaces reflectance as obtained from the
measurement. The properties of the diffuse panel were estimated based on the ‘trans’ model
of the translucent panel in Reinhart & Andersen (2006), by fine-tuning the diffuse
transmissivity to 0.35.
Table 3.3. Material definitions in Radiance for the window construction and room’s interior
Material
Red
Green
Blue
Diffuse panel
Window glass
Window frame
Ceiling
Walls
Floor
Door
Countertop
0.21
0.88
1.00
1.00
0.90
0.56
0.56
1.00
0.21
0.88
0.78
1.00
0.90
0.55
0.48
1.00
0.21
0.88
0.60
0.95
0.90
0.48
0.56
0.97
Specularity
0.08
0
0
0
0
0
Roughness
0
0
0
0
0
0
Diffuse
Transmit.
transmiss. specularity
0.35
0
-
Simulations were run for the three sky scenes, i.e. overcast, clear, and partly cloudy.
Calculation was performed for the 55 measuring points on the workplane. One-to-one
comparison between measurement and simulation was done for all values of horizontal
illuminance at the line where the points P1 and 1 were located. This line, at which there were
seven measuring points, was located directly in the central projection of the window.
In addition, the prototype scenes were compared to real window scenes. The latter were
modelled in Radiance by replacing the entire construction of artificial light sources, with the
corresponding sky models, i.e. overcast, clear, and partly cloudy. These sky models were
generated in Radiance using the Gensky programme by using the options –c, –s, and +i,
respectively. Site location was set for Eindhoven, the Netherlands (51.45°N, 5.47°E), with
south-facing window orientation, on 21 June at 12.20 hrs local time.
The zenith radiance [W/(sr·m2)] of each sky was defined so that the illuminance values at
the nearest point to the window (P1) were the same under the corresponding real and virtual
window scenes. The relevant zenith radiance was respectively 7.5, 5.5, and 11 W/(sr·m2) for
the overcast, clear, and partly cloudy skies. Illuminance values on the rest of the points at the
central column were determined for the comparison. DGP values at positions 1 and 2 (see
Figure 3.3) were also analysed using Evalglare.
Furthermore, simulation parameters in Radiance were set as shown in Table 3.4.
40
Table 3.4. Radiance simulation parameters
Parameter
-ab
-aa
-ar
-ad
-as
Description
Ambient bounces
Ambient accuracy
Ambient resolution
Ambient divisions
Ambient super-samples
Value
4
0.08
128
1024
256
In order to assess whether the simulation results are fit for the purpose of recreating the
measured scene, several criteria can be applied. There is no definitive agreement on an
acceptable degree of accuracy (Ochoa et al., 2012). For example, in their report on testing
accuracy of various lighting simulation programmes, Maamari et al. (2006) suggested a
criterion of two times the global error, based on the estimated error in the measurements and
in the scenario description, e.g. sensor cosine and colour corrections, sensor calibration,
lumen output fluctuation, luminaire position and flux output distribution, room dimensions,
and surface reflectance. These were approximately ±21% from the true value, see also Slater
& Graves (2002) and CIE TC-3-33 (2005). According to Fisher (1992), an acceptable criteria
range would be 10% for average illuminance calculations and 20% for measured point
values. The criterion of 20% for use in real cases has been validated by Reinhart & Andersen
(2006), as appeared in studies replicating built realities.
In view of subjective lighting perception, the European Standard EN 12464-1 (CEN,
2002) mentions that “a factor of approximately 1.5 represents the smallest significant
difference in subjective effect of illuminance”, as given in the recommended scale of
illuminance [lx] for various conditions in work places. This is approximately in line with the
findings of Slater et al. (1993) in their subjective study, where illuminance ratios between two
work stations of at least 0.7 (or 1.4 if the ratio is inversed) were ‘generally acceptable’. They
mentioned that even though there was a trend of decreasing acceptability at lower
illuminance ratios, there were indications that lower illuminance ratios may also be
acceptable under some conditions.
Taking this recommendation into account, the criterion for which difference between
simulation (Esim [lx]) and measurement (Emea [lx]) values does not lead to a significant
difference in subjective effect is:
0.67 <
E sim
< 1.50
E mea
(3.10)
In other words, the ratio of simulation and measurement values at any measuring point
should not be less than 2 : 3 (or approximately 0.67) and not more than 3 : 2 (or 1.50), so that
the values do not lead to a significant difference in their subjective effect. This criterion is
applied in the following sections to evaluate the simulation results.
41
Chapter 3
Section 3.5.1 presents the measurement results of the prototype. Section 3.5.2 presents
simulation results of the prototype, whereas Section 3.5.3 presents those of the
corresponding, hypothetical real window in the same positions, under the same sky scenes.
3.5.1. Measurement of prototype
Measurement results of the average illuminance values (Eav [lx]), the uniformity (U0),
and the space availability (%A [%]) under the three sky scenes of the prototype are
summarised in Table 3.5.
Table 3.5. Measurement results of the average illuminance, uniformity, and space availability
under the overcast, clear, and partly cloudy sky scenes
Eav [lx]
U0 [-]
%A [%]
Overcast
52
0.28
0
Clear
70
0.28
0
Partly cloudy
102
0.27
0
The measurement results show that at the nearest point to the window, the horizontal
illuminance value is found to be 400 lx under the partly cloudy scene, compared to 180 lx
under the overcast one. Despite this large variation, the uniformity in the three scenes are
relatively similar (0.27 ~ 0.28), which means the influence of the HID spot lamp on
uniformity is limited, mainly increasing the total light output.
Moreover, none of the points receives a horizontal illuminance larger than 500 lx, under
all sky scenes. As a result, the space availability (taking 500 lx as the minimum criterion) in
all cases is zero. It should be noted that the general lighting in the room was completely
switched off, to ensure that only the prototype contributed to the light inside the room.
Vertical illuminance on the observer’s eye plane (Ev [lx]), together with minimum (Lmin
2
2
2
[cd/m ]), maximum (Lmax [cd/m ]), and average luminance (Lav [cd/m ]) perceived by the
observer at positions 1 and 2 (referring to Figure 3.3) are displayed in Table 3.6. These values
were extracted from the post-processing software Photolux 3.1. In addition, the DGP values
obtained from Evalglare are also given.
In line with the measurement results of horizontal illuminance, the lowest measured
vertical illuminance is also found under the overcast sky scene, while the highest is found
under the partly cloudy one. This is also true for the minimum, maximum, and average
luminance, as well as DGP perceived by the observer. While the vertical illuminance at
position 1 under the partly cloudy scene is around 1.5 times the value under the overcast
42
scene, the maximum luminance under the former is 4 times higher than that under the latter
(6000 to 1550 cd/m2). The maximum luminance is actually found on the location of the ‘sun
spot’, whereas the vertical illuminance at position 1 is determined by the total window
surface area.
Table 3.6. Vertical illuminance on the observer’s eye, minimum, maximum, average luminance,
and DGP perceived by the observer at positions 1 and 2 under the three sky scenes of the
prototype
Position – scene
Ev [lx]
Lmin
[cd/m2]
Lmax
[cd/m2]
Lav
[cd/m2]
DGP
[-]
1 – overcast
2 – overcast
1 – clear
2 – clear
1 – partly cloudy
2 – partly cloudy
403
208
427
246
600
348
0.23
0.18
0.28
0.22
0.40
0.30
1550
1540
5500
3400
6000
4200
73
39
80
47
111
66
0.24
0.21
0.32
0.28
0.34
0.30
According to a discomfort glare classification of Jakubiec & Reinhart (2012), DGP values
of < 0.30 are considered ‘imperceptible’, a DGP range of 0.30 ~ 0.35 corresponds to a
‘perceptible’ category, DGP values between 0.35 and 0.45 represent disturbing glare, and
DGP values over 0.45 represent intolerable glare. Hence, only the observers at position 1
under the partly cloudy and the clear sky scenes are expected to experience perceptible
discomfort glare from the prototype.
Figure 3.5 displays the luminance false colour pictures of the prototype as seen from
position 1; note there are different scales used in the three pictures. The window surface
under the overcast scene obviously appears more uniform, whereas a bright spot of the PAR
lamp in the upper right corner of the window is revealed under the other two scenes.
Combined high dynamic range (HDR) images of the same views are displayed in Figure 3.2,
in which the directional light from the PAR lamp leaves its pattern on the countertop (Figures
3.2b and 3.2c).
43
Chapter 3
(a)
(b)
(c)
Figure 3.5. Luminance false colour pictures of the prototype observed at position 1, under (a)
overcast, (b) clear, and (c) partly cloudy sky scene
From the pictures in Figure 3.5, one can conclude that, as the mean view luminance of the
prototype is more than 1800 cd/m2, the display is capable of creating discomfort glare (Shin
et al., 2012; Kim et al., 2012). This level is present in the clear and partly cloudy sky scenes.
The contrast between the surrounding wall and the window is very often over 1 : 20 or 1 : 40,
which is another sign of potential discomfort glare.
3.5.2. Simulation of prototype
Table 3.7 summarises the simulation results of the horizontal illuminance point at the
central line on the workplane, together with the overall average illuminance values (Eav [lx]),
uniformity (U0), and space availability (%A [%]) under the three sky scenes of the prototype.
44
For comparison, the measurement results, and the ratio between simulation and measurement
values are also shown.
The lighting simulation and measurement results of the prototype generally show similar
trends with a maximum relative difference of 26%, found on the farthest point from the
window, under the overcast sky scene. The maximum relative difference for the average
illuminance is 18%, also found under the overcast sky scene. However, the ratio of the
simulated value to the measured one at all points is always in the range of 0.67 ~ 1.50, which
represents the smallest significant difference in subjective effect of illuminance (CEN, 2002).
Looking at the criterion, the models are therefore considered sufficient for the purpose of
reproducing the scenes without giving a significant subjective difference, even though more
care should be taken when interpreting the modelling results of scenes with relatively low
lighting levels, as shown here in the overcast sky scene.
Table 3.7. Simulation (sim.) and measurement (meas.) results of horizontal illuminance point at
the central column, together with the average illuminance values (Eav [lx]), uniformity (U0), and
space availability (%A [%]) under the three sky scenes of the prototype
Overcast
Distance to
Sim.
window
Clear
Meas. Ratio Sim. Meas. Ratio Sim. Meas. Ratio
[lx]
[-]
[lx] [lx]
[-]
[lx]
[lx]
[-]
[m]
[lx]
0.4
245
204
1.15
432
354
1.22
533
491
1.09
0.9
155
155
1.00
253
236
1.07
346
355
0.97
1.4
79
88
0.89
139
129
1.07
187
190
0.99
1.9
52
61
0.85
104
102
1.01
132
149
0.89
2.4
38
44
0.86
65
56
1.16
81
80
1.01
2.9
28
36
0.77
48
47
1.02
65
65
1.00
3.4
25
34
0.74
43
44
0.98
61
65
0.94
Eav [lx]
42
52
0.82
79
71
1.11
99
102
0.96
U0 [-]
0.19
0.28
0.27
0.71
0.20
0.27
0.74
%A [%]
0
0
0
n/a
2
0
n/a
0.68 0.20
n/a
0
45
Partly cloudy
Chapter 3
3.5.3. Simulation of real windows
Figure 3.6 displays the graphs showing the relationship between horizontal illuminance
and the distance to the window under the three sky scenes, based on the measurement and
simulation of the prototype (VW) and simulation of real windows (RW).
(a)
(b)
(c)
Figure 3.6. Graphs showing the relationship between horizontal illuminance and distance to
window under the (a) overcast, (b) clear, and (c) partly cloudy sky scene
Compared to the corresponding real window scenes, it is seen that all of the modelled sky
scenes, the prototype gives a more uniform illuminance distribution throughout the space.
Figure 3.6 shows how the light from the real window rapidly drops at the distance of more
than 1 m from the window, while the decreases are less dramatic under the virtual window
scenes. Under the clear sky scene, the difference between the real and virtual windows is less,
due to the influence of direct sunlight, which reduces the illuminance decrease throughout the
space with the real window.
46
(a)
(b)
(c)
(d)
(e)
(f)
Figure 3.7. False colour maps of the workplane illuminance [lx] under the (a) overcast, (c) clear,
and (e) partly cloudy sky scenes of the measured prototype; and under the (b) overcast, (d) clear,
and (f) partly cloudy sky scenes of the simulated real window
47
Chapter 3
Figure 3.7 displays false colour maps of workplane illuminance values under the three
sky scenes, from both the measured prototype and the simulated real window. Comparison of
the corresponding illuminance contour maps reveals that the prototype yields a wider
illuminance distribution throughout the workplane. This is mainly due to the fact that the
light sources of the prototype are placed at a certain distance from the window glass; whereas
under the real windows scenes, the sun and sky are at infinity, therefore the light distribution
rapidly drops throughout the space. Under the clear sky scene, the real window gives a wider
distribution at the left-hand side of the workplane, as shown in Figure 3.7d, since the direct
sunlight comes from the upper-right corner of the window. Under the overcast and partly
cloudy scenes, the diffuse panel plays a role not only in creating the blur and cloudy display
of the window, but also in spreading the generated light onto the back of the room.
Table 3.8 displays the maximum luminance and DGP perceived at positions 1 and 2, from
the simulated real window under all sky scenes, as compared to those in the measured
prototype. It is interesting to see that even though both scenes have the same illuminance
value at the nearest point (P1 in Figure 3.3), the luminance values perceived by the observers
greatly differ. Under nearly all sky scenes and observer’s positions, the maximum luminances
of the real window are lower, and so are the DGP values, than those of the virtual one.
Table 3.8. Maximum luminance and DGP perceived by the observer at positions 1 and 2, under
the three sky scenes in the simulated real window and the measured prototype
Real window
VNLS prototype
Position – scene
Lmax
[cd/m2]
DGP
[-]
Lmax
[cd/m2]
DGP
[-]
1 – overcast
2 – overcast
1 – clear
2 – clear
1 – partly cloudy
2 – partly cloudy
601
556
3596
3523
2587
2111
0.23
0.21
0.28
0.27
0.25
0.24
1550
1540
5500
3400
6000
4200
0.24
0.21
0.32
0.28
0.34
0.30
Despite the relatively high luminance a real window can produce, particularly in
appearance of the sun, the discomfort glare perception is still relatively low compared to the
situation with the prototype. This is also due to the fact that the light sources of the prototype
are placed at a certain distance from the window glass, instead of at infinity. Under the real
window scenes, the light is scattered in a more diffuse way, therefore the discomfort glare
under the real window scenes is less than that under the prototype scenes. It is also noticed
that the placement of the PAR spot lamp as a virtual sun at the upper corner of the prototype
cannot always represent the real sun’s position at the site location of the test room,
particularly for low solar elevation angles; it should be placed at a sufficient distance behind
the window glass in order to do so.
48
In general, the measurement and simulation results give an idea of how a VNLS
prototype with various sky scenes compares to a real window under a similar sky scene, in
terms of physical lighting phenomena. The VNLS prototype analysed here had a limited
complexity level of the view. Additional features such as motion parallax and sound
transmission could also improve the degree of similarity between the virtual and real
windows, even though it may not be directly related to the lighting performance on the
workplane.
It can also be argued that while the investigated prototype lacked some features that are
usually associated with a real window, this prototype was designed and constructed to create
a subjective, rather than accurately measured, impression or feeling of being connected to the
outside world, without necessarily reproducing all of the details. For instance, the addition of
curtains or Venetian blinds on the window frame makes it less visible, which in some cases
can remove the impression that the window is artificial (van Loenen et al., 2007). Compared
to other prototypes with a simplified view discussed in Section 2.2.1, this particular prototype
scores better in terms of visual appearance, due to the possibility to vary the sky view, colour
gradients, and directional light. The future work on this subject will be to investigate how
building occupants actually appraise such artificial solutions in reality. Therefore, thorough
user’s performance and perception studies are required.
Computational modelling and building performance simulation can help steer the process
of VNLS design development. An example of the influence of simulation in VNLS
development is shown in this chapter, where Radiance was applied to reproduce the scenes
and to evaluate the lighting performance of a first generation VNLS prototype displaying a
view of overcast, clear, and partly cloudy skies. Based on the performed measurements, it is
observed that for the selected three sky scenes, the overcast scene produced the lowest
average horizontal illuminance, while the partly cloudy one produced the highest. Using the
designed setting, none of the measuring points received a horizontal illuminance of 500 lx or
larger, suggesting the need of a higher intensity setting for each scene, to ensure sufficient
amount of light for typical working activities.
The key point of this chapter is to show that simulations can be used to compare an actual
VNLS prototype with a hypothetical real window under the same settings, which was not
possible physically, since the test room was not located at the building’s façade. Based on the
lighting simulation in Radiance, the investigated prototype performed better in terms of light
distribution uniformity than a corresponding, hypothetical real window under the overcast
and partly cloudy scenes. Under the clear sky scene, the difference between the real and
virtual windows is less, due to the influence of direct sunlight.
Further works should be focused on improving the sun mimicking under the clear sky
scene. Moreover, the greatest next challenge possibly is to understand how people will
49
Chapter 3
actually appraise VNLS in reality. Therefore, thorough user’s performance and perception
studies are required in the future.
50
Chapter 4
Discomfort Glare Evaluation and Simulation of a First
Generation Virtual Natural Lighting Solutions Prototype
This chapter provides an evaluation of discomfort glare from a first generation VNLS
prototype with complex views and diffuse light, correlated to the results of experiment
conducted by Shin et al. (2012) on subjective glare perception from the same prototype.
Correlation was made between four normalised glare metrics and the reported glare
perception ratings. Radiance and Evalglare were employed to model and recreate the scenes,
and to evaluate the glare metrics.
4.1. Introduction
Many researchers have shown the significant role of windows in buildings. Windows are
important in controlling the amount of natural light admitted from the exterior environment
into the buildings. A proper use of natural light would potentially save considerable amount
of energy from artificial lighting use (e.g. Assem & Al-Mumin, 2010; Hammad & AbuHijleh, 2010; Yun et al., 2010). Moreover, it has been shown that building occupants feel
windows are important due to their preference for having natural light over electric light (e.g.
Markus, 1967; Ulrich, 1984; Farley & Veitch, 2001; Hartig et al., 2003; Chang & Chen,
2005; Aries et al., 2010). Several studies have reported beneficial and restorative effects of
views on a natural scene (e.g. Kaplan, 1993; Kaplan, 1995; Tennessen & Cimprich, 1995;
Berman et al., 2008), whereas views on human-built environments yield effects, which are
similar to having no window at all (Kaplan, 1993). Kim & Wineman (2005) showed
empirically that views and windows have psychological and economic values. In general
terms, the correct application of a daylighting strategy in buildings can increase visual
comfort and energy efficiency (Galasiu & Veitch, 2006).
Nonetheless, natural light is known to have a certain limitation, mostly on its availability
in time and space. As mentioned earlier in Chapter 1, there are situations in which natural
light is absent; for instance during nighttimes, in the inner part of buildings where access to
the façade is limited, and in working spaces where having a real daylight opening is not
possible due to hygienic or safety reasons.
Another well-known disadvantage of natural light is discomfort glare, which very often
reduces the effective use of natural light inside buildings. This leads to discomfort glare being
one of the most important topics to address in daylighting research. Many researchers have
conducted experiments to study the human response to discomfort glare from windows.
However, due to the limited possibility of controlling natural light, some researchers opted to
apply prototypes of ‘simulated’ or ‘virtual’ windows to create a controlled daylit scene in
their experiments, which mostly include discomfort glare (e.g. Tuaycharoen & Tregenza,
51
Chapter 4
2005; Tuaycharoen & Tregenza, 2007; Kim et al., 2008; Kim & Kim, 2011; Yun et al., 2011;
Kim et al., 2012). Such prototypes have a complex view display but generate a mainly diffuse
light output, hence are still classified as the first generation VNLS prototypes.
Note that the aforementioned studies were not conducted to measure the difference in
perception of views from real and virtual windows. Instead, those studies focused on the
variation of the view itself, with regard to subjective discomfort glare perception. As
mentioned by Shin et al. (2012), despite the fact that some luminous characteristics of virtual
windows can be different from those of real windows, the luminance ratios within the view
area can be set equal to the real scenes, to reproduce same luminance conditions to the
subjects. Another advantage of using virtual windows is the possibility to repeat exactly the
same experiment condition for other subjects, which will be difficult if using real windows.
In the experiment of Shin et al. (2012), the prototype had a dimension of 1.2 m × 1.2 m ×
0.25 m, of which 0.9 m × 0.9 m was viewable. Incandescent lamps were installed behind the
prototype in rows of 14 × 14 at an interval of 80 mm, which in total could generate a
luminance level of approximately 15000 cd/m2. In the experiment, mean view luminance
values of 1000, 1800, 3200, 5600, and 10000 cd/m2 were used, which had equal increment
factors of approximately 1.8.
The view image was digitally photographed and printed on a transparent film, and was
pasted in front of the prototype surface. A total of 10 view images were used in the
experiment, which represented some distant or near views, natural or man-made, as well as
land or river landscapes. The view images were named as follows: ‘Distant Mixed Land’,
‘Near Mixed Land’, ‘Distant Natural River’, ‘Near Natural River’, ‘Distant Man-made’,
‘Near Man-made’, ‘Distant Natural Land’, ‘Near Natural Land’, ‘Distant Mixed River’, and
‘Near Mixed River’; as displayed in Figure 4.1.
Figure 4.1. View images on the prototype, from (Shin et al., 2012): (a) ‘Distant Mixed Land’, (b)
‘Near Mixed Land’, (c) ‘Distant Natural River’, (d) ‘Near Natural River’, (e) ‘Distant Man-made’,
(f) ‘Near Man-made’, (g) ‘Distant Natural Land’, (h) ‘Near Natural Land’, (i) ‘Distant Mixed
River’, and (j) ‘Near Mixed River’.
52
Discomfort Glare Evaluation and Simulation of a First Generation VNLS Prototype
The experiment was conducted in a room with dimensions of 6.25 m × 4.56 m × 2.5 m.
The prototype was located 0.6 m above the floor, and the subjects were seated facing the
prototype at a distance of 1.5 m, as illustrated in Figure 4.2. The subjects were then asked to
look at the centre of the view for 5 seconds and to rate the discomfort glare that they sensed
on a rating questionnaire, presented in both English and Korean (native) languages. The
discomfort glare rating was based on Hopkinson (1972), in which the scale of the discomfort
glare was divided into eight stages, using semantic scales proposed by Flynn et al. (1979) as
shown in Table 4.1. A total of 10 views with five luminance mean values were shown to each
subject. A total of 48 subjects (24 men and 24 women, ages 23 ± 2.4 yr) participated in the
experiment.
(a)
(b)
Figure 4.2. (a) Floor plan of the test room (taken from Kim et al. (2012)), (b) section plan of
setup of the subject and the prototype (taken from Shin et al. (2012))
Table 4.1. Discomfort glare rating used in the experiments of Shin et al. (2012)
Glare perception
Just perceptible
Noticeable
Just acceptable
Acceptable
Just uncomfortable
Uncomfortable
Just intolerable
Intolerable
Rating
1
1.5
2
2.5
3
3.5
4
4.5
In Shin et al.’s study, it was not discussed that people generally accept higher luminance
values (more glare) from daylight compared to electric lighting. It is uncertain if this also
53
Chapter 4
applies to VNLS, since they are supposed to be a category in between daylight and electric
lighting, and therefore the question of the study on this chapter.
Meanwhile, in (day-) lighting simulation, the use of existing glare metrics such as the
Daylight Glare Probability (DGP) (Wienold & Christoffersen, 2006), the Daylight Glare
Index (DGI) (Hopkinson, 1972), the CIE Glare Index (CGI) (Einhorn, 1979), or the Unified
Glare Rating (UGR) (CIE TC-3-13, 1995) is practical, since they are based on objective,
physical phenomena leading to glare itself, which is mainly indicated with the size, location,
and luminance of the sources and background in the field of view.
In historical order of the development, the four glare metrics are expressed as follows:
n
DGI = 10 × log 
i  1Lb
0.8
L1s.,6i  pos
,i
(4.1)
 (0.07s0,.i5 Ls,i )
where ωs is the solid angle of the glare source [sr], ωpos is the solid angle of the glare source
modified for its position in the field of view [sr], Ls is the glare source luminance [cd/m2], and
Lb is the background luminance, i.e. the average luminance of areas not indentified as glare
sources [cd/m2].
2
2
(1  E d / 500) n Ls,i s,i
CGI = C1 × log C2

2
Ed  E i
i  1 Pi
(4.2)
where Ed is the direct vertical illuminance [lx], Ei is the diffuse vertical illuminance [lx], P is
the position index, i.e. a weight factor based on position in a viewing hemisphere, and C1 and
C2 are the weighting coefficients defined as 8 and 2 by Einhorn (1979).
0.25 n Ls,i s,i
UGR = 8 × log

Lb i  1 Pi2
2
2
2

n Ls,i s,i

DGP = 5.87 × 10 Ev + 9.18 × 10 log 2  1   1.87 2

Pi
i  1E v

–5
–5
(4.3)





(4.4)
where Ev is the total vertical eye illuminance [lx].
To evaluate the aforementioned glare metrics, Jakubiec & Reinhart (2012) performed
simulations of three indoor spaces under 144 clear sky conditions. In order to directly
compare the results from the various glare metrics, they proposed a normalisation procedure
54
for DGI, UGR, and CGI results, so that the ranges were between 0 (no likelihood of
discomfort) and 1 (100% probability of discomfort). The metrics were multiplied by a certain
multiplier, chosen to correlate the intolerable value ranges with those of DGP (Jakubiec &
Reinhart, 2012). The normalisation equations read as follows:
DGIn = 0.01452 × DGI
(4.5)
UGRn = 0.01607 × UGR
(4.6)
CGIn = 0.01607 × CGI
(4.7)
where DGIn, UGRn, CGIn are the normalised DGI, UGR, and CGI values, respectively. Next
to that, they also proposed the resulting value ranges in which glare was considered to be
‘imperceptible’, ‘perceptible’, ‘disturbing’, and ‘intolerable’ for the corresponding metrics.
These ranges are given in Table 4.2.
Table 4.2. Value ranges of DGP, DGI, UGR, and CGI, after Jakubiec & Reinhart (2012)
Glare perception
Imperceptible
Perceptible
Disturbing
Intolerable
DGP
DGI
UGR
CGI
< 0.30
0.30~0.35
0.35~0.45
> 0.45
< 18
18~24
24~31
> 31
< 13
13~22
22~28
> 28
< 13
13~22
22~28
> 28
In their article, Shin et al. (2012) did not report a direct relationship between their glare
ratings and any of the four glare metrics. In another article, Kim et al. (2012) briefly
discussed correlations between their glare study findings and DGI, but a one-to-one
relationship between the glare ratings and DGI itself was not determined. In this thesis, the
glare metrics are incorporated to indicate visual discomfort from VNLS, performed in
simulation. Therefore, it is intended to have a comparison between the simulation-based glare
metrics and the reported subjective glare ratings, so that the results can be used to indicate
which glare metrics are the most suitable for VNLS application, and what the effect of
various VNLS display views on discomfort glare perception will be.
The objective of this study is to correlate simulated glare metrics from a VNLS prototype
to the experimental results of Shin et al. (2012). As a method, Radiance and Evalglare were
applied to model the test room and to evaluate the relevant glare metrics, i.e. DGP, DGI,
UGR, and CIE from the prototype. Polynomial regression was applied to correlate the glare
perception ratings of Shin et al. and the normalised glare metrics of Jakubiec and Reinhart.
The polynomial equation was used to convert the ratings of Shin et al. to the four glare
metrics. Finally, the values obtained from simulation were compared with the converted
values, for each image view.
55
Chapter 4
4.2. Method
The VNLS prototype was modelled in Radiance, by using the standard ‘light’ material to
form a relatively large light emitting area, with a uniform surface luminance. To add the view
image on top of the light emitting surface in Radiance, the technique of two-dimensional
image mapping was applied. It is basically a technique of pasting a picture on top of a plain
surface, for example to display a painting or photograph inside a scene (Jacobs, 2012a). By
assigning a ‘light’ material for the plain surface, a luminous display is created.
The test room (see Figure 4.2) was modelled in Radiance, assuming typical values for the
interior reflectance, as given in Table 4.3.
Table 4.3. Material reflectance definitions in Radiance for the room’s interior
Material
Window frame
Ceiling
Walls
Floor
Door
Red
Green
Blue
Specularity
Roughness
0.600
0.800
0.700
0.200
0.600
0.430
0.800
0.700
0.200
0.430
0.210
0.800
0.700
0.200
0.210
0
0
0
0
0
0
0
0
0
0
The relevant two-dimensional image was mapped on a flat, infinitely thin surface
constructed with standard ‘light’ material of a very small thickness. The luminance of this
thin surface can be obtained with applying the following equation, which converts red, green,
and blue radiance components of a surface to its corresponding luminance value (Ward &
Shakespeare, 1998).
Ls = 179 LR,G,B = 179 (0.265 LR + 0.670 LG + 0.0648 LB)
(4.8)
where Ls is the surface luminance [cd/m2], LR,G,B is the surface radiance [W/(sr·m2)], and LR,
2
LG, LB are the spectral radiance values [W/(sr·m )] in red, green, and blue components,
respectively.
In this case, for a certain image and luminance setting, a same value was assigned for the
three spectral radiance values (i.e. LR = LG = LB) to produce only white light behind the pasted
image. Different values were assigned to different images and luminance settings, so that the
mean window luminance as seen by the observer became equal to 1000, 1800, 3200, 5600, or
10000 cd/m2. Table 4.4 gives the assigned spectral radiance values for each image and
luminance setting.
56
Table 4.4. Spectral radiance values assigned for the painting mat at each image and luminance
setting
Image
Distant Man-Made
Near Man-Made
Distant Natural Land
Near Natural Land
Distant Mixed Land
Near Mixed Land
Distant Natural River
Near Natural River
Distant Mixed River
Near Mixed River
1000
9.80
13.30
10.95
18.02
9.16
14.70
11.89
13.97
13.30
11.64
Mean window luminance [cd/m2]
1800
3200
5600
10000
17.64
31.36
54.89
98.01
23.94
42.57
74.49
133.01
19.72
35.05
61.34
109.54
32.44
57.67
100.92
180.21
16.49
29.31
51.29
91.58
26.46
47.05
82.33
147.02
21.40
38.04
66.56
118.86
25.14
44.69
78.21
139.66
23.94
42.57
74.49
133.01
20.95
37.24
65.18
116.39
Simulations were run individually in Radiance and Evalglare for the 10 view images and
five mean luminance values, to calculate DGP, DGI, UGR, and CGI at the observer’s
position. The ambient parameters used in Radiance for all variations are set as shown in
Table 4.5.
Table 4.5. Radiance ambient parameters used in the simulation of all variations
Parameter
-ab
-aa
-ar
-ad
-as
Description
Ambient bounces
Ambient accuracy
Ambient resolution
Ambient divisions
Value
4
0.08
128
1024
256
4.2.2. Glare rating correlation
To correlate the glare rating of Shin et al. and the four normalised glare metrics, values
with approximately similar glare perception were selected. As a different researcher can use a
different terminology for the glare perception, some assumptions were made. Those were as
follows:
• The lowest end of each scale is assumed to be zero for all metrics, and is perceived as
‘imperceptible’, i.e. not perceptible at all.
• ‘Just perceptible’ value (1 in Shin et al.’s rating) is assumed equal to the minimum value
defined in the original DGP experiment (Wienold & Christoffersen, 2006), which is 0.20.
This is approximately 2/3 of the noticeable (perceptible) value.
57
Chapter 4
• ‘Noticeable’ in Shin et al.’s rating is assumed equal in perception to ‘perceptible’ in other
metrics.
• ‘Uncomfortable’ in Shin et al.’s rating is assumed equal in perception to ‘disturbing’ in
other metrics.
• ‘Just intolerable’ value in all metrics is assumed equal to the mid-point between
‘uncomfortable’ and ‘intolerable’.
Using the aforementioned assumptions, the corresponding values can be listed for all
metrics, as shown in Table 4.6.
Table 4.6. Discomfort glare rating used by Shin et al. and the corresponding values of DGP,
DGIn, UGRn, and CGIn as correlated to the glare perception
Glare perception
Shin
DGP
DGIn
UGRn
CGIn
Imperceptible
Just perceptible
Noticeable (perceptible)
Uncomfortable (disturbing)
Just intolerable
Intolerable
0
1
1.5
3.5
4
4.5
0
0.20
0.30
0.35
0.40
0.45
0
0.17
0.26
0.35
0.40
0.45
0
0.14
0.21
0.35
0.40
0.45
0
0.14
0.21
0.35
0.40
0.45
Applying the technique of curve fitting with polynomial regression, one can derive an
equation correlating the glare metric (DGP, DGIn, UGRn, or CGIn) as a function of Shin et
al.’s rating. Knowing this equation, the reported mean glare rating of Shin et al. can be
converted to estimate the glare metric, symbolised with DGPconv, DGIn conv, UGRn conv, or
CGIn conv. These values were then compared with the simulated values from Evalglare,
symbolised with DGPsim, DGIn sim, UGRn sim, or CGIn sim, by putting them in a scatter plot for
all 50 variations. The converted and simulated values should be the same in the most ideal
case. By examining mean square error between those values, one can observe which of the
four glare metrics is the most reliable for describing the glare perception in Shin et al.’s
experiment.
4.3.1. Rendering and glare source detection
Two examples of a 180° equiangular view of the rendered model of the test room are
shown in Figures 4.3a and 4.3b, displaying the prototype with the ‘Distant Man-Made’
(DMM) and ‘Near Man-made’ (NMM) image scenes at mean luminance value of 3200
cd/m2, observed from the subject’s position. The false colour luminance pictures of both
scenes are displayed in Figures 4.3c and 4.3d.
58
(a)
(b)
(c)
(d)
(e)
(f)
Figure 4.3. Impressions of the test room as observed from the subject’s position under the
‘Distant Man-Made’ (DMM) scene and ‘Near Man-made’ (NMM) image scenes. (a) and (b) display
the rendered images, (c) and (d) display the false colour luminance pictures, (e) and (f) display the
glare source detection pictures of the DMM and NMM scenes respectively. The mean luminance
value of the window is 3200 cd/m2 under both scenes.
59
Chapter 4
Evalglare displays detected glare sources in different colours where the rest of the image
is set to grey. The colour is randomly chosen by the programme, and does not indicate a
specific range of luminance values. The glare source detection pictures of the same scenes are
displayed in Figures 4.3e and 4.3f.
From these two examples, it is seen that different view images produce a different glare
source, which in turn will result in a different glare perception. In the case of distant view
such as the DMM image scene, the ‘sky’ element is mostly the main glare source. The sky
element also delivers some light to the ceiling, as seen in Figure 4.3c, even though that part of
the ceiling is not considered as a glare source (Figure 4.3e). In the case of near view such as
the NMM image scene, a more uniform luminance distribution is observed on the window
surface. The white part of the building in the image becomes the glare source.
4.3.2. Unadjusted rating
The results of DGP, DGIn, UGRn, and CGIn under the DMM and NMM image scenes are
plotted in Figure 4.4. The mean glare rating of Shin et al. (value range of 1 ~ 4.5) is plotted
without any adjustment in a linear scale together with the normalised glare metric (value
range of 0 ~ 1). The resulting curve shows a significantly higher gradient, compared to the
four metrics. Note that the mean values have a relatively large standard deviation, showed by
the error bars. Similar graphs and trends are also found under the other image scenes. Figure 4.4. Results of DGP, DGIn, UGRn, CGIn, and uncorrected Shin et al.’s rating under the
DMM and NMM image scenes. Error bars indicate standard deviation from the mean values of
Shin et al.’s rating.
4.3.4. Polynomial regression
The values of glare rating used by Shin et al. and the glare metrics as displayed in Table
4.4 are plotted in Figure 4.5. By applying curve fitting with polynomial regression, the
equation relating the metric and the rating can be derived, as also displayed in the charts. For
60
every glare metric, a third-order polynomial equation is obtained, with a coefficient of
determination R2 > 0.99.
Figure 4.5. Graphs showing the selected glare rating of Shin et al. and the corresponding values of
DGP, DGIn, UGRn, and CGIn
Based on the obtained polynomial equations, the reported mean glare rating of Shin et al.
were converted to yield the glare metrics. These converted values and the simulated values
from Evalglare for all 50 variations are plotted in Figure 4.6. By applying curve fitting with
linear regression, the equation relating the simulated and converted metric can be derived, as
also displayed in the charts.
61
Chapter 4
Figure 4.6. Graphs showing the simulated and converted values of DGP, DGIn, UGRn, and CGIn
for all 50 variations
4.3.4. Adjusted rating
In the most ideal case, both simulated (y) and converted (x) values are equal, hence the
root mean square error (RMSE) will be zero, and the linear equation will have a gradient of 1
and an offset of 0, i.e. y = x. It is found that DGP has the smallest RMSE (0.04) among the
four metrics, indicating the smallest difference between simulated and converted values. The
CGIn linear equation has a gradient of approximately 1, but the RMSE is large (0.29).
Moreover, the linear equations of DGIn, UGRn, and CGIn have relatively large offset (0.22 ~
0.39), while DGP has a relatively small offset (–0.13). This suggests that the simulated values
of DGIn, UGRn, and CGIn are all overestimated, compared to converted values from the
62
experiment data of Shin et al. In turn, the simulated values of DGP show a good agreement
with the converted ones, even though for the higher range (around 0.40), the simulated values
are overestimated.
The high accuracy of DGP compared to other glare metrics was also reported elsewhere,
e.g. Kleindienst & Andersen (2009) and Jakubiec & Reinhart (2012), mainly due to the fact
that DGP uses vertical illuminance on the observer’s eye as the main input, while this type of
illuminance has a strong, positive relationship with the glare metric (Kleindienst & Andersen,
2009). However, all of the earlier findings on DGP were based on real daylight scenes, as
obviously implied in the metric’s name. The finding in this chapter gives a new insight on the
applicability of DGP for VNLS prototype, together with its correlation to the subjective glare
perception.
From the relationship between DGP and Shin et al.’s rating in Figure 4.5, the third-order
polynomial regression equation reads as follows:
y = 0.013x3 – 0.107x2 + 0.3193x – 0.038
(4.9)
where y is DGP and x is Shin et al.’s rating. Taking the derivative function of Equation 4.9
yields the following equation:
y dy
≈
= 0.039x2 – 0.214x + 0.3193
x dx
(4.10)
where Δy and Δx are the deviations of y and x, respectively. Since the actual standard
deviations from Shin et al.’s rating are known from the article (see Figure 4.4), the estimated
deviations of the converted values can be obtained.
To give a better illustration, the simulated and converted values of DGP under the DMM
and NMM image scenes are plotted in Figure 4.7.
63
Chapter 4
Figure 4.7. Graphs showing the simulated and predicted values of DGP under the DMM and
NMM scenes. Error bars indicate the estimated deviation of the converted values.
The resulting curve shows similar values of the simulated and converted DGP,
particularly at the low and medium luminance. It is noticed that the largest differences
(approximately 0.1) are found at the 10000 cd/m2 luminance, possibly due to the lower
tolerance that Evalglare gives for scenes with a relatively high vertical illuminance on the
subject’s eye (Wienold & Christoffersen, 2006). All predicted values have standard
deviations of less than 0.05, shown by the error bars. Under the other image scenes, the
differences between the simulated and predicted values are also found to be less than 0.1. The
graphs are displayed in Figure 4.8.
Figure 4.8. Graphs showing the simulated and predicted values of DGP under all image scenes,
excluding DMM and NMM scenes. Error bars indicate the estimated deviation of the predicted
values.
64
Figure 4.8. (continued)
65
Chapter 4
4.3.5. Percentage of disturbed subjects
In their article, next to reporting the mean values of the glare rating as a function of
mean window view luminance, Shin et al. also reported the distribution of their subjects who
voted for a certain glare rating, which corresponded to a certain glare perception. In this
study, it has been demonstrated that the mean values of the rating can be converted into
normalised glare metrics (i.e. DGP, which is the most accurate one) and compared them with
the values from simulation. However, it is unknown whether these values can be correlated
with the actual percentage of subjects who felt disturbed from glare in Shin et al.’s
experiment.
To investigate this, the percentage of subjects in Shin et al.’s experiment who voted for
each corresponding glare perception, under all image scenes, are shown in Table 4.7. From
there, the total percentage of the subjects who voted for ‘just uncomfortable’ or worse (i.e.
higher discomfort glare rating) was calculated for each mean window view luminance.
Similarly, total percentage of the subjects who voted for ‘uncomfortable’ or worse, and those
who voted for ‘just intolerable’ or worse, were also calculated. These values represent the
actual percentage of disturbed subjects in the experiment, and are drawn in graphs in Figure
4.9.
Table 4.7. Percentage of subjects in Shin et al.’s experiment who voted for each corresponding
glare perception, under all image scenes
Rating
Glare perception
1
1.5
2
2.5
3
3.5
4
4.5
Just perceptible
Noticeable
Just acceptable
Acceptable
Just uncomfortable
Uncomfortable
Just intolerable
Intolerable
Total
Window luminance [cd/m2]
1800
3200
5600
24.3
11.7
0.7
35.3
18.2
4.2
30.1
32.9
14.5
7.4
25.4
28.3
4.4
9.8
29.5
0.2
1.8
14.5
0.2
0.2
8.2
0.0
0.0
0.0
100
100
100
1000
71.3
16.5
10.3
1.8
0.2
0.0
0.0
0.0
100
66
10000
0.1
0.0
0.7
4.4
20.1
25.3
32.7
17.7
100
Figure 4.9. Graphs showing the percentage of subjects in Shin et al.’s experiment who perceived
glare equal to or worse than ‘just uncomfortable’, ‘uncomfortable’, or ‘just intolerable’ under all
image scenes
Based on Table 4.7, if the ‘just uncomfortable’ category is taken as the lower threshold of
the disturbed feeling from discomfort glare, there are approximately 0% subjects who felt
disturbed at 1000 cd/m2, 4% at 1800 cd/m2, 12% at 3200 cd/m2, 53% at 5600 cd/m2, and 96%
at 10000 cd/m2. If the threshold is raised to ‘uncomfortable’, the figures will be 0%, 0%, 2%,
23%, and 76% at the same luminance values. If ‘just intolerable’ is the threshold, the figures
will be 0%, 0%, 0%, 8%, and 51%.
Moreover, when the data are analysed separately for each image scene, a roughly similar
trend is observed; two examples are illustrated in Figure 4.10 for the DMM and NMM image
scenes.
Figure 4.10. Graphs showing the percentage of subjects in Shin et al.’s experiment who perceived
glare equal to or worse than ‘just uncomfortable’, ‘uncomfortable’, or ‘just intolerable’ under the
DMM and NMM image scenes
67
Chapter 4
Whichever glare perception rating is chosen as the lower threshold, the curves in Figures
4.9 and 4.10 do not correspond well to the curves of DGP in Figures 4.7 and 4.8. This may be
explained by drawing the distribution of subjects in the experiment who voted for each glare
perception, under all image scenes, grouped based on the mean window view luminance
values as shown in Figure 4.11. From the charts, it is observed that there is a ‘floor effect’
phenomenon on the lowest window value (1000 cd/m2), where more than 70% of the subjects
voted for ‘just perceptible’. Such a dominant vote is not found elsewhere. Note that the
minimum rating given was 1; if there was a rating of 0 (imperceptible), many would probably
have chosen it.
The subjects agreed less at the higher luminance settings; at 1800 cd/m2 most subjects
voted for ‘noticeable’ (i.e. 1.5 on the rating), but there were a large difference between the
percentage of subjects who voted ‘just acceptable’ (rating of 2) and ‘acceptable’ (rating of
2.5). The votes at 3200 cd/m2 and 5600 cd/m2 were more normally distributed, but were still
relatively widespread. At the highest luminance of 10000 cd/m2, the subjects were still
divided between perceiving the glare as ‘uncomfortable’ and ‘just intolerable’. A possible
explanation can be that the meaning of the words are very close to each other and can be
understood differently between the subjects. Another explanation would be that the scale was
agreed on, but different people simply perceive glare differently.
Figure 4.11. Graphs showing the distribution of subjects in Shin et al.’s experiment who voted for
each glare perception under all image scenes, grouped based on the mean window view luminance
Due to the interpersonal variance in subjective glare perception ratings, it is relatively
hard to predict the actual percentage of the subjects who felt disturbed. A comparison with
the experiment of Wienold & Christoffersen (2006) in developing the DGP suggests the use
of fewer rating scales. In that experiment, the subjects were asked to rate the glare on a fourpoint scale, i.e. ‘imperceptible’, ‘noticeable’, ‘disturbing’, and ‘intolerable’. They were also
asked to rate the lighting condition as ‘comfortable’ or ‘uncomfortable’, if they had to
68
perform daily work at the test location. The subjects would then have fewer criteria on
assessing the glare, which presumably led to a sharper distinction between disturbed and
undisturbed feelings from the glare. Furthermore, there were in total 76 subjects (48 men and
28 women, ages ranged from 20 to 59 yr), slightly more than the number of subjects in Shin
et al.’s experiment.
While it has been shown that DGP yields the most accurate results compared to other
glare metrics, as also found in the study of Jakubiec & Reinhart (2012), it is noticed that DGP
also has some limitations. Next to the low availability of user acceptance study at lower
luminance ranges (Wienold & Christoffersen, 2006), it is also mentioned that DGP is less
accurate for luminance contrast-based glare (Kleindienst & Andersen, 2009). A typical
example of this kind of glare can be found in scenes with a generally low ambient
illuminance, and with a high contrast point such as bright window glass with a dark frame,
which is similar to the situation in the Shin et al.’s experiment. To address this problem,
Kleindienst & Andersen (2009) proposed the so-called model-based method to approximate
the glare. In a recent study, Suk et al. (2013) proposed the use of absolute and relative glare
factors to develop a more practical analysis method. In this method, there is a different
analysis for glare from extremely high luminance light sources, and that from glare sources
people can still adapt to. Further development on this specific topic of discomfort glare
metrics is still expected or ongoing.
Based on the discussed observation, it is concluded that the mathematical models
correlating the glare metrics and the glare rating of Shin et al. can accurately predict the
reported mean values in that particular experiment. The reported mean values can be
converted to DGP, which values are comparable to those obtained from simulation. The
simulation values however are not representations of the actual number of subjects who felt
disturbed in the experiments, even though the simulation values correspond very well to the
mean values, with certain standard deviations, as voted by the subjects.
Furthermore, this simulation study confirms the general finding from the experiment; that
at mean view luminance of 3200 cd/m2, glare from the prototype display is perceived as
noticeable or perceptible, i.e. corresponds to a DGP value of 0.30. This suggests an upper
limit, where any exceeding mean values will only create visual discomfort and inefficient use
of electrical energy.
A method is proposed to correlate the commonly used glare metrics, i.e. DGP, DGI,
UGR, and CGI, which can be obtained from simulation, to subjective glare ratings from a
first generation VNLS prototype with complex views (Shin et al., 2012). Based on the scatter
plots comparing the simulated and converted values, it was found that the linear equations of
DGIn, UGRn, and CGIn have a relatively large root mean square error (RMSE), while DGP
has a relatively small one. This suggests that the simulated values of DGP are in a better
69
Chapter 4
agreement with the converted ones. Hence, the DGP yields the most accurate results in this
particular case. Even though the accuracy of DGP has been widely reported, all of the earlier
findings were based on real daylight scenes. The finding in this chapter demonstrates the
applicability of DGP for VNLS prototype, and how it correlates to the subjective glare
perception.
Neither the simulated values nor the predicted values of the glare metrics can be
correlated with the actual percentage of subjects who felt disturbed from glare in the
experiment. This is mainly due to different glare perception, particularly at the higher
luminance values. There is also an observed ‘floor effect’ at the lowest mean luminance
value. Meanwhile, it has been demonstrated that the mean values of the Shin et al.’s rating
can be converted into normalised glare metrics and then compared with the values from
simulation.
70
Chapter 5
Design, Measurement, and Simulation of a Second
Generation Virtual Natural Lighting Solutions Prototype
This chapter discusses the design and evaluation of a second generation VNLS prototype.
The design features a more sophisticated direct light component and more energy-efficient
light sources. The prototype was realised, and was evaluated by measuring workplane
illuminance in a test room. Simulation using Radiance was performed and validated with the
measurement results. Various possibilities of placing the prototypes inside the room were
investigated in Radiance to determine the effect on space availability and visual comfort.
Various operating scenarios were introduced and calculated to determine the effect on the
average space availability and total annual electrical energy consumption that are produced
and consumed by the prototypes.
5.1. Introduction
In Chapter 3 of this thesis, an existing first generation VNLS prototype as described in
van Loenen et al. (2007) has been investigated and modelled, to give an idea on how such a
prototype compares to a (simulated) real window. This prototype featured a blurred, diffuse
view of sky scenes, i.e. overcast, clear, and partly cloudy, with an additional PAR spot lamp
to create the impression of direct sunlight. The partly cloudy sky scene used in this study
produced the highest average horizontal illuminance, whereas the overcast sky scene
produced the lowest. None of the measuring points received a horizontal illuminance of 500
lx or larger, suggesting the need for a larger area or a higher intensity setting for each scene,
particularly the overcast one, to ensure a sufficient amount of light for typical working
activities.
While the findings from that particular prototype give an insight in the actual lighting
performance, it gives direction to design improvements needed to overcome its current
limitations, particularly its relatively small window-to-wall ratio, the missing of ground
elements and horizon in the view, and the use of fluorescent light sources instead of more
energy-efficient ones. The spot lamp also had limited ability to create a realistic impression
of sun patches in the space. A new generation prototype would then be required to observe
whether the performance can be improved by adding the missing features, and to validate a
computational model that can be extended for further development of future VNLS.
Another issue that has not been addressed so far is the influence of operating scenarios on
the lighting performance and energy demand of a certain VNLS prototype. Next to the
average space availability, it is intended to predict the amount of electrical energy the entire
prototype system will consume on an annual basis, provided the prototype is operated on a
routine schedule on each working day.
71
Chapter 5
Therefore, the study in this chapter aims to design and build a second generation VNLS
prototype with an improved performance compared to the first one. Three objectives are
defined: the first one is to validate the illuminance distribution results obtained from
Radiance simulation with the ones obtained from measurement, by evaluating the interior
lighting condition inside a test room. The second objective is to determine the effect of
various configurations of the prototypes inside the test room on the space availability,
uniformity, and visual comfort. Finally, the third objective is to estimate the influence of
various operating scenarios on the average space availability and total annual electrical
energy consumption that would respectively be produced and consumed when applying such
prototypes. 5.2. Design Steps
5.2.1. Test environment
The test environment was built in the new ExperienceLab of Philips Research at the High
Tech Campus in Eindhoven, the Netherlands. The dimension of the test room was 6.81 m ×
3.63 m × 2.70 m (L×W×H), slightly longer than the standard reference office room (van Dijk
& Platzer, 2003). Two original real window openings were placed on a short wall (south) to
enable daylight admission. The VNLS prototype was installed on the opposite wall (north).
In order to avoid daylight entering from outside during the measurements, in the entire
experiments, the two real windows were blocked with two white covers of the same colour
and reflectance as the surrounding wall finishing. Figure 5.1 illustrates the floor plan and
section view of the test room.
N
Prototype
(a)
Figure 5.1. (a) Floor plan and (b) cross section view of the standard test environment
72
Design, Measurement, and Simulation of a Second Generation VNLS Prototype
(b)
There were two openings for the prototype; each had a dimension of 0.90 m × 1.20 m
(W×H) excluding the window frames, while the height of the window bottom was 0.93 m
from the floor. The distance between the frames of the two openings was 0.14 m. Figure 5.2
shows the prototype openings on the wall’s front section.
Figure 5.2. Front view of the prototype openings
5.2.2. Light sources
The light source of this prototype was an array of light emitting diode (LED) tiles. The
decision to use LEDs was taken mainly due to its relatively long life time, high efficacy, high
flexibility, and possibility to individually control and to display multiple colours. Moreover,
LED technology has a great potential for saving energy consumption in buildings (Jenkins &
Newborough, 2007; Pandharipande & Caicedo, 2011). In this study, a total of eight Philips
Origami BPG762 luminaires were incorporated to provide the light as well as to construct the
view. An individual Origami BPG762 is actually a combination of four Origami BPG732
tiles. Each luminaire houses 108 LUXEON RGB power LEDs, and can display a virtually
limitless range of dynamic changing colours, given in red, green, and blue (RGB)
73
Chapter 5
components. Because it is designed to be a uniform edge-free lighting tile, it can be combined
to form an array and perform as a ‘pixel’ unit.
An individual Origami BPG762 has a maximum power consumption of 128 W and is
driven by a 24 V DC source. Each tile can be independently controlled by a Digital Addressable Lighting Interface (DALI) or Digital Multiplex (DMX) protocol with a proper
programming. The dimension of an individual Origami BPG762 tile is shown in Figure 5.3.
Figure 5.3. Dimensions of Origami BPG762 tile
Inside each individual Origami BPG732, 27 LUXEON power LEDs are distributed. These
LEDs are covered by multiple layers of diffusers with different properties to ensure a uniform
light output distribution. Red, green and blue channels can be controlled independently.
To fit the size of the prototype window openings, a total of eight (2 rows × 4 columns)
Origami BPG762 tiles were applied. Since each individual Origami BPG762 is actually a
combination of four Origami BPG732 tiles, the entire prototype consisted of 32 pixels in
total. Each Origami BPG762 tile could be controlled by 12 channels. The conceptual
structure of the display lighting panel together with window opening is illustrated in Figure
5.4.
Figure 5.4. Rear view illustration of the 2 × 4 Origami BPG762 array
74
Beside the main display elements, additional direct light sources were also required to
realise another feature of a (real) window, which is providing direct sunlight into the indoor
space. For this purpose, the use of the LED tiles alone was not sufficient. To select the direct
light sources, some requirements were taken into account, i.e.:
• The shape should be linear
• The emitted light should be directional
• The array should be independently dimmable
• The colour temperature should be controllable
• The beam angle of the light source should be wide (100° ~ 110°)
The Philips iW Cove MX Powercore (wide beam version) module was selected for the
purpose. It is a high-performance, white light LED fixture with adjustable colour
temperatures. It has independent channels of warm, neutral, and cool white LEDs to produce
colour temperatures in a range from 2700 K to 6500 K. Moreover, an individual iW Cove is
only 0.30 m long (see Figure 5.5) and can be easily connected in a line with other lamps of
the same type to reach the total length needed.
Figure 5.5. Dimensions of an individual iW Cove
5.2.3. Control circuit
In this study, the DMX512 protocol was used to realise the control function. DMX512 is
a standard for digital communication networks that is commonly used to control lighting
dimmers. Under this protocol, a digital dimming lighting network can be set up by using
DMX512 controllers and control software. Each colour channel can be dimmed from the full
RGB value of 255 to 0. Compared to the DALI protocol, the advantage of using DMX512 is
that each can control 512 channels at the same time, whereas each DALI controller can only
control 63 channels. DMX also has a high data rate, as needed for dynamic views. Therefore
the DMX512 is a better choice for the control circuit in this case.
75
Chapter 5
The DMX512 modules are already integrated with the Origami and iW Cove arrays. The
Origami accepts the so-called regular DMX ESTA signals, whereas iW Cove uses DMX
Kinetics signals. To control both types of lamps, a DMX Splitter was applied to split the
signal from the computer, convert it to the respective DMX formats, and transfer the signal to
the luminaires. By connecting them with a DMX controller, the basic network was formed.
Commands could be sent by the computer to control the light emitting level of each ‘pixel’ or
display unit.
The Origami array was powered by 24 V DC supplies, provided by rectifiers. The iW
Cove array, connected in series, was powered by a Philips Data Enabler Pro, providing
integrated DMX signals. The conceptual control circuit structure illustration is shown in
Figure 5.6, in which gray lines symbolise the signal circuit and black lines symbolise the
power circuit. COMPUTER
iW Cove
ORIGAMI
ORIGAMI
iW Cove
iW Cove
Rectifier
iW Cove
ORIGAMI
DMX
CONTROLLER
ORIGAMI
iW Cove
Rectifier
iW Cove
iW Cove
ORIGAMI
ORIGAMI
iW Cove
DMX
SPLITTER
Rectifier
ORIGAMI
ORIGAMI
DATA
PRO
ENABLER
Rectifier
SIGNAL:
POWER:
Figure 5.6. Control and power supply circuits
76
5.2.4. Display and structure
The prototype was designed to provide a blurred suggestion of a view, rather than a high
resolution view. A single diffusive panel was mounted in front of the Origami array,
replacing the original separate outer diffusers, to eliminate the visibility of the tile
boundaries. To achieve that, the diffusive panel should also have a high transmittance, and be
placed in a certain distance from the Origami array. A 5 mm thick PLEXIGLASS Satinice
colourless diffuser (type 0F00 SC) was selected for this purpose. It had a one-sided matte
surface and was made of frosted surface material, which means that its thickness does not
significantly influence the amount of light loss.
The layers of Origami array – diffuser – window glass formed the basic structure of the
prototype display, as shown in Figure 5.7.
Figure 5.7. Basic three-layer structure of the prototype display
In this prototype, all of the Origami tiles were fixed to a MDF back plate holding the 2.40
m × 1.20 m array. During system operation, heat is generated by the LEDs, as discussed for
example by Ahn et al. (2014). Therefore, openings were made in the back plate (see Figure
5.8), allowing cooling by natural convection.
Figure 5.8. Front view of the MDF frame
77
Chapter 5
The prototype appearance from outside and inside the test room (at ‘off’ and ‘on’
conditions) are displayed in Figure 5.9.
(a)
(b)
(c)
Figure 5.9. The prototype appearance from (a) outside and (b) inside the test room at ‘off’
condition, (c) inside the test room at ‘on’ condition
The current prototype could display a diffuse, 32 pixel view. Therefore, any given view
image should be converted or filtered into 32 pixels, and should comply with some general
rules as follows:
• The aspect ratio is 2 : 1 (width : height), as is the cased for the prototype source.
• The horizon is at eye height, which is 1.20 m for sitting viewers. Using the defined space
configuration and given the display is four pixels high, the ‘ground’ element should be
displayed only at the lowest row.
• The ‘sky’ element is composed of blue sky and white clouds. To increase the white light
output, the clouds should preferably take more than half of the sky area.
The general steps to simulate a view on the VNLS prototype are described in Table 5.1.
78
Table 5.1. General steps to simulate a view on the prototype
No
1
Step
Select a view image with an aspect ratio of 2 : 1,
adjust the height of the horizon into
approximately 0.25 of the total view height.
2
Use image processing software to apply a mosaic
filter to turn the image into 32 pixels. Use colour
picker to obtain the RGB values of each pixel, as
input for the DMX.
3
Apply filter to simulate the diffuser and glass
plate in front of the view image.
4
Apply the direct light source.
5
Add the window frame.
View example
5.2.5. Programming and setting
To define the colour display settings, each pixel was given a code name and was assigned
with a DMX value. For the Origami array, each pixel was given a code name of A, B, C, or
D, followed by a number. Code name ‘A’ referred to the lowest row, while ‘D’ was the
highest row. The number was named 1 to 8, indicating the position from left to right, seen
from inside the test room. Each pixel had three channels corresponding to the red, green, and
blue (R, G, and B) components.
79
Chapter 5
For the iW Cove array, the codes started with ‘P’ followed by number from 1 to 8. Each
pixel had three channels corresponding to the colour temperature of warm (2700 K), neutral
(4000 K), and cool (6500 K).
Referring to the view image shown in Table 5.1, the DMX values addressed for each
pixel of the display at the maximum setting (i.e. the highest luminance) are given in Table
5.2. The values range from 0 (entirely off) to 255 (100% on), where RGB values of [255,
255, 255] correspond to a full-white colour display.
Table 5.2. DMX values addressed for each pixel on the display at the maximum setting
Position
Red
Green
Blue
Position
Red
Green
Blue
Position
Red
Green
Blue
Position
Red
Green
Blue
D1
20
255
230
C1
230
255
255
B1
255
255
255
A1
0
200
100
D2
220
255
255
C2
220
255
255
B2
255
255
255
A2
0
216
10
D3
255
230
255
C3
230
255
255
B3
225
255
255
A3
0
221
10
D4
26
233
255
C4
220
255
255
B4
205
255
255
A4
0
215
10
D5
25
255
255
C5
220
255
255
B5
200
255
255
A5
0
227
10
D6
220
255
255
C6
230
255
255
B6
180
255
255
A6
0
224
10
D7
220
255
255
C7
200
255
255
B7
230
255
255
A7
0
225
10
D8
25
248
227
C8
220
255
255
B8
200
255
255
A8
0
229
10
For further analysis, any daily profile scenario can be defined by proportionally scaling
the DMX values of the maximum setting. Three daily profile scenarios were defined,
depending on the climate type, to represent ‘summer’, ‘spring’, and ‘winter’ conditions,
which are discussed later in detail in Section 5.6.
To evaluate the actual lighting performance of the prototype, a number of data were
collected at three settings, i.e. 100% (see Table 5.2), 62.5% (representing a medium setting),
and 25% (representing a low setting) of the maximum. The DMX values were proportionally
scaled, while the display showed the view image in Table 5.1. The collected data were as
follows:
• Horizontal illuminance on the workplane; data were collected for 91 points (see Figure
5.10) on the workplane height, i.e. 0.75 m from the floor.
80
• Vertical illuminance on the observer’s eye plane; data were collected for four positions
(namely 1 until 4, see Figure 5.10) on the typical observer height when seated (1.20 m
from the floor).
• Minimum, maximum, and average luminance perceived by the observer at the same four
vertical points on the typical observer height (1.20 m from the floor).
• Reflectance of interior surface materials; data were collected for the relevant interior
surface, such as floor, walls, ceiling, and furniture.
3
1
4
Prototype
2
Figure 5.10. Horizontal and vertical illuminance measurement points in the test room
Furthermore, the horizontal illuminance data were post-processed to obtain the average
illuminance values (Eav [lx]), uniformity (U0), and space availability (%A [%]), referring to
Equations 3.6 until 3.8 in Chapter 3.
Vertical illuminance and luminance perceived by the observer were measured by taking
20 photographs (ISO 400, f/5.6, shutter time varied from 4 s to 1/8000 s) each at positions 1
until 4, at a height of 1.20 m, with the view direction specified by the arrows in Figure 5.10.
To determine glare index, i.e. the Daylight Glare Probability (DGP) (Wienold &
Christoffersen, 2006), at the observer’s position, the obtained photographs were exported to
Radiance, combined into HDR images using the Hdrgen programme, and then analysed using
Evalglare (Wienold & Christoffersen, 2006).
During the measurement, the following instruments were used:
• SpectraDuo PR-680 photometer; for measuring luminance values as well as spectral power
distribution.
• Lutron LX-1118 light meter; for measuring illuminance values on the workplane.
81
Chapter 5
• Canon EOS50D digital single-lens reflex camera + Sigma 5.5mm fisheye lens + Photolux
3.1 software; for taking multiple (20 in this case) photographs in equiangular 180° view
with various exposure values, which in turn were post-processed to obtain the luminance
pictures. The luminance values were calibrated with the SpectraDuo photometer.
• Konica Minolta CM-2600D spectrophotometer; for measuring reflectance values of the
interior surface materials.
The first objective of this chapter is to validate the illuminance distribution results
obtained from simulation using Radiance with the ones obtained from measurement, since
the computational model is to be extended for development of future (not-yet existing)
VNLS. Therefore, the actual conditions under the three lighting scenes were also modelled
and simulated, to give an insight in the difference between simulation and actual
measurement. Comparison was made between the simulated and measured values of
horizontal illuminance at the middle row in Figure 5.10, as well as between the simulated and
measured values of the space availability and uniformity.
The front, top, and perspective views of the modelled prototype are displayed in Figure
5.11. The 32 Origami tiles were modelled as boxes of 0.30 m × 0.30 m × 0.05 m (L×W×D),
constructed with a ‘light’ material. The eight iW Cove lamps were modelled as a continuous
row of eight cylinders, with a length of 0.30 m each and a diameter of 0.016 m, also
constructed with a ‘light’ material. Each lamp had various red, green, and blue radiance
components [W/m2/sr], depending on the row position and the sky scene. The assigned values
for each component under the maximum setting in Radiance are defined in Table 5.3, which
is fine-tuned proportionally to the actual DMX values (Table 5.2). For other settings, the
assigned values are proportionally scaled.
(a)
(b)
(c)
Figure 5.11. (a) Front, (b) top, and (c) perspective views of the modelled prototype
82
Table 5.3. Red, green, and blue irradiance components of ‘light’ material defined in Radiance for
each pixel on the display at the maximum setting
Position
D1
D2
D3
D4
D5
D6
D7
D8
4.70
60.0
55.1
51.8
60.0
60.0
60.0
55.1
60.0
6.12
46.6
60.0
5.88
54.8
60.0
51.8
60.0
60.0
51.8
60.0
60.0
5.88
58.3
53.4
C1
C2
C3
C4
C5
C6
C7
C8
55.1
60.0
60.0
51.8
60.0
60.0
55.1
60.0
60.0
51.8
60.0
60.0
51.8
60.0
60.0
55.1
60.0
60.0
47.1
60.0
60.0
51.8
60.0
60.0
B1
B2
B3
B4
B5
B6
B7
B8
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
48.2
60.0
60.0
47.1
60.0
60.0
42.3
60.0
60.0
55.1
60.0
60.0
47.1
60.0
60.0
A1
A2
A3
A4
A5
A6
A7
A8
Red
Green
0
47.1
0
50.8
0
52.0
0
50.8
0
53.4
0
52.7
0
52.9
0
53.9
Blue
23.5
2.35
2.35
2.35
2.35
2.35
2.35
2.35
Red
Green
Blue
Position
Red
Green
Blue
Position
Red
Green
Blue
Position
The detailed values assigned for the window construction and the room’s interior surfaces
reflectance as obtained from the measurement are specified in Table 5.4.
Table 5.4. Material definitions in Radiance for the window construction and room’s interior
Material
Red
Green
Blue
Diffuse panel
Window glass
Window frame
Ceiling
Walls
Floor
Door
0.25
0.90
0.79
0.91
0.79
0.13
0.79
0.25
0.90
0.79
0.91
0.79
0.08
0.79
0.25
0.90
0.79
0.91
0.79
0.03
0.79
Specularity
0
0
0
0
0
0
Roughness
0
0
0
0
0
0
Diffuse
Transmit.
transmiss. specularity
0.55
0
-
5.4.2. Validation
The first objective of this chapter is to validate the simulation of workplane illuminance
with the measurement results. Simulations were run for the three settings, i.e. 100%, 62.5%,
and 25%; by addressing the input defined in Table 5.3. Calculation was performed for the 91
measuring points on the workplane, referring to Figure 5.10. One-to-one comparison between
measurement and simulation was done for all values of horizontal illuminance at the middle
83
Chapter 5
row where point 4 was located. This row, at which there were 13 measuring points, was
located directly in the central projection of the windows. Furthermore, simulation parameters
in Radiance were set as previously shown in Table 3.4 in Chapter 3. As previously discussed in Section 3.4 in Chapter 3, the European Standard EN 12464-1
(CEN, 2002) mentions that “a factor of approximately 1.5 represents the smallest significant
difference in subjective effect of illuminance”, as given in their recommended scale of
illuminance. This criterion is applied to evaluate the simulation results, in which the ratio of
simulation and measurement values at any measuring point should not be less than 2 : 3 (or
approximately 0.67) and not more than 3 : 2 (or 1.50).
5.5. Analysis of Various Configurations
The second objective of this chapter is to determine the effect of various configurations of
the prototype inside the test room, on the space availability, uniformity, and visual comfort.
In order to maximise the light distribution on the workplane, two prototypes were modelled
inside the test room, and were placed either on the short walls or on the long walls. Seven
configurations were introduced, named Configuration 1 until 7, which floor plans are
illustrated in Figure 5.12. Note that in most configurations, the prototype was split into two
equal parts; each consisted of 4 × 4 tiles. Each opening had a dimension of 0.90 m × 1.20 m
(W×H), and the height of the window bottom in all configurations was 0.93 m from the floor,
that is the same as in the tested configuration (Figure 5.2).
(a)
(b)
(c)
(d)
Figure 5.12. Floor plan of the test room with the prototypes in configurations (a) 1, (b) 2, (c) 3,
(d) 4, (e) 5, (f) 6, and (g) 7
84
(e)
(f)
(g)
Using the properties in Tables 5.2 and 5.3, simulations in Radiance and Evalglare were
run to obtain the space availability, uniformity, and DGP at the defined four observer’s
positions. In Section 5.3, the test room was assumed to be used for typical office work that
requires workplane illuminance of 500 lx, but in practice it can also be used for other
purposes, for instance waiting rooms in healthcare facilities, that require a lower workplane
illuminance. Therefore, for this analysis, the space availability was evaluated not only for the
criterion of 500 lx, but also for 300 and 200 lx.
5.6. Analysis of Various Operating Scenarios
5.6.1. Settings and data collection
The third objective of this chapter is to get an overview of the space availability and
electrical energy consumption of a VNLS prototype under various operating scenarios in a
given year. To estimate the space availability, the workplane illuminance measurement data
of the prototype were revisited. With the same reason as in Section 5.5, to maximise the light
distribution on the workplane, it was assumed that there were two prototypes inside the test
room, placed facing each other on the short walls; i.e. according to Configuration 2 in Figure
5.12. Linear regression was performed for each point to estimate the illuminance values
under any given DMX setting.
To estimate the energy consumption, the real-time power consumption of the same
prototype was measured, as a function of the corresponding DMX values, i.e. 0%, 10%, 20%,
40%, 50%, 62.5%, 80%, 90%, and 100%. The measurement was conducted under the defined
85
Chapter 5
displayed setting, i.e. the view image was according to the values in Table 5.2 and was
proportionally scaled. During the measurement, the ELV EM800 energy monitor was used for
measuring the real-time power and current of the entire system.
5.6.2. Daily profiles and annual modes
Assuming an operating schedule from 09.00 to 18.00 hrs local time on every working
day, it is intended to predict the amount of average space availability and total electrical
energy the two prototypes will produce and consume in a given year. To obtain this, certain
schedules based on daily profiles and annual modes should be defined beforehand.
Given the maximum DMX value settings in Table 5.2, three daily profiles were defined to
represent ‘spring’, ‘summer’, and ‘winter’ conditions. In this study, the daily profiles were
obtained by calculating vertical illuminance on a point facing south in an exterior field, on 21
March (spring), 21 June (summer), and 21 December (winter); at every hour from 09.00 to
18.00 hrs local time, using Gensky programme of Radiance. All values were normalised to
the maximum value among those three days, so that they could be presented in a scale of 0 ~
1.
In addition, five locations were selected to represent various Köppen-Geiger climate types
(Peel et al., 2007). The representative cities for each selected climate type were Singapore
(tropical rainforest), Cairo (arid, hot desert), Amsterdam (temperate, oceanic), Sevilla
(temperate, Mediterranean), and Chicago (cold, continental). For each climate type, a sky
model for Gensky input was selected by considering the typical weather data provided by the
Department of Energy of the United States (US Department of Energy, 2011). Table 5.5
gives the description of those five locations.
Table 5.5. Five climate types selected for total electrical energy consumption estimation
Climate
type
General description
Typical
city
Latitude,
longitude
Sky model
Af
Tropical, rainforest
Singapore
1.37°N, 103.97°E
Standard CIE overcast
(–c)
Bwh
Arid, hot desert
Cairo
30.12°N, 31.38°E
Partly cloudy with sun
(+i)
Cfb
Temperate, warm summer/
Amsterdam
oceanic
52.28°N, 4.77°E
Partly cloudy without
sun (–i)
Csa
Temperate, hot summer/
Mediterranean
Sevilla
38.42°N, 5.90°E
(+i)
Dfb
Cold, warm summer/
continental
Chicago
41.97°N, 87.92°W
(+i)
86
(a)
(b)
(c)
(d)
(e)
Figure 5.13. Daily profile settings as defined for (a) Singapore, (b) Cairo, (c) Amsterdam, (d)
Sevilla, and (e) Chicago
87
Chapter 5
(a)
(b)
(c)
(d)
(e)
Figure 5.14. Monthly average solar radiation of (a) Singapore, (b) Cairo, (c) Amsterdam, (d)
Sevilla, and (e) Chicago, based on the weather data (US Department of Energy, 2011)
88
As a direct consequence of using various climate types, the daily profiles for the selected
locations are different from each other, see Figure 5.13. It should be also noted that the
profiles were based on vertical illuminance at each hour on the three selected days, which
makes it possible that the values on ‘spring’ and ‘winter’ are larger than those on ‘summer’,
because the sun’s altitude is actually lower on spring and winter days, hence more direct
sunlight.
For estimating the total electrical energy consumption in a year, two annual operating
modes were defined. These modes were based on the order of assigning the daily pattern,
either following the real seasons, or the opposite. In detail, they are as follows:
• Mimicking mode: the assigned daily pattern of the prototypes follows the real seasons, i.e.
the prototype displays the ‘summer’ setting during summer months, and displays the
‘winter’ setting during winter months.
• Compensating mode: the assigned daily pattern of the prototypes follows the opposite of
the real seasons, i.e. the prototype displays the ‘winter’ setting during summer months, and
displays the ‘summer’ setting during winter months.
In both modes, ‘spring’ is considered representing also the autumn months, therefore they
are interchangeable. To simplify the calculation, it is assumed that every day in a given
month has exactly the same daily profile. For instance, if June is assumed to be a summer
month, then the prototypes will display the ‘summer’ setting for the entire month of June
under the mimicking mode, or the ‘winter’ setting for the entire month of June under the
compensating mode.
Note that in the calculation, both prototypes are assumed to display exactly the same
setting at all time, i.e. there is no difference between light output and view from both of them.
The terms mimicking and compensating only refer to the difference between the prototypes
display and the real seasons, not to the difference between the two prototypes display.
The choice of which months to be considered as ‘spring’, ‘summer’, and ‘winter’ was
made by considering the variation in monthly average global horizontal solar radiation of the
particular location (US Department of Energy, 2011), as illustrated in Figure 5.14. For
example, the months with the highest solar radiation in Cairo (i.e. May, June, and July) can
be regarded as summer months; however, one can also argue that April, August, and
September should be included in the ‘summer’, given the relatively high solar radiation in
those months as compared to the other climates. To test the sensitivity of the results for such
choices, three annual scenarios were defined for each location. Under the mimicking mode,
Scenario 1 (Sc1) has the shortest summer and the longest winter months, Scenario 2 (Sc2) has
roughly equal length of summer and winter, and Scenario 3 (Sc3) has the longest summer and
the shortest winter months.
Table 5.6 gives the selected scenarios for the five climate types, where ‘Sp’, ‘Su’, and
‘Wi’ stand for spring, summer, and winter, respectively.
89
Chapter 5
Table 5.6. Annual profile scenarios for each selected climate type under mimicking and
compensating modes
Location
Singapore
Cairo
Mode
Mimicking
Compensating
Mimicking
Compensating
Month Sc1 Sc2 Sc3 Sc1 Sc2 Sc3 Sc1 Sc2 Sc3 Sc1 Sc2 Sc3
Jan
Sp Sp Sp Sp Sp Sp Wi Wi Wi Su Su Su
Feb
Su Su Su Wi Wi Wi Sp Wi Sp Sp Su Sp
Mar
Su Su Su Wi Wi Wi Sp Sp Sp Sp Sp Sp
Apr
Su Su Su Wi Wi Wi Sp Sp Su Sp Sp Wi
May
Sp Su Su Sp Wi Wi Su Su Su Wi Wi Wi
Jun
Sp Sp Sp Sp Sp Sp Su Su Su Wi Wi Wi
Jul
Sp Sp Su Sp Sp Wi Su Su Su Wi Wi Wi
Aug
Sp Sp Sp Sp Sp Sp Sp Su Su Sp Wi Wi
Sep
Sp Sp Sp Sp Sp Sp Sp Sp Su Sp Sp Wi
Oct
Sp Sp Sp Sp Sp Sp Sp Sp Sp Sp Sp Sp
Nov
Wi Wi Wi Su Su Su Wi Wi Wi Su Su Su
Dec
Location
Amsterdam
Sevilla
Mode
Mimicking
Compensating
Mimicking
Compensating
Month Sc1 Sc2 Sc3 Sc1 Sc2 Sc3 Sc1 Sc2 Sc3 Sc1 Sc2 Sc3
Jan
Feb
Wi Sp Sp Su Sp Sp Sp Wi Sp Sp Su Sp
Mar
Apr
May
Su Su Su Wi Wi Wi Su Su Su Wi Wi Wi
Jun
Jul
Aug
Sp Sp Su Sp Sp Wi Sp Su Su Sp Wi Wi
Sep
Oct
Nov
Dec
90
Table 5.6. (continued) Location
Chicago
Mode
Mimicking
Compensating
Month Sc1 Sc2 Sc3 Sc1 Sc2 Sc3
Jan
Wi Wi Wi Su Su Su
Feb
Sp Wi Sp Sp Su Sp
Mar
Sp Sp Sp Sp Sp Sp
Apr
Sp Sp Sp Sp Sp Sp
May
Su Su Su Wi Wi Wi
Su Su Su Wi Wi Wi
Jun
Jul
Su Su Su Wi Wi Wi
Aug
Sp Su Su Sp Wi Wi
Sep
Sp Sp Sp Sp Sp Sp
Oct
Sp Sp Sp Sp Sp Sp
Nov
Wi Wi Wi Su Su Su
Dec
Wi Wi Wi Su Su Su
Results of the measurement and simulation of the actual test room and prototype are
given in Sections 5.7.1 and 5.7.2. Results of analyses of various configurations and operating
scenarios are respectively given in Sections 5.7.3 and 5.7.4. 5.7.1. Measurement of actual test room
Measurement results of the average workplane illuminance, uniformity, and space
availability (%A [%]) at 25%, 62.5%, and 100% of the maximum setting in the test room are
summarised in Table 5.7. It is found that under the maximum setting of the defined
configuration, the average workplane illuminance was 239 lx and 11% of the workplane met
the target illuminance of 500 lx. This suggests an addition of local or task lighting on the
workspace is still required to satisfy the criterion for typical working activity (CEN, 2002).
Under the lower settings, the space availability becomes zero; i.e. no spots on the workplane
satisfy the illuminance criterion. The uniformity is about the same under all settings, within
the range of 0.30 ~ 0.37.
Table 5.7. Measurement results of the average illuminance (Eav), uniformity (U0), and space
availability (%A) at 25%, 62.5%, and 100% of the maximum setting in the test room
Setting
Eav [lx]
U0 [-]
25%
49
0.37
0
62.5%
136
0.33
0
100%
239
0.30
11
91
%A [%]
Chapter 5
Figure 5.15 displays false colour maps of horizontal illuminance values on the workplane
under the three settings.
(a)
(b)
(c)
Figure 5.15. False colour maps of the measured workplane illuminance [lx] under (a) 25%, (b)
62.5%, and (c) 100% of the maximum setting
The measurement results show that at the nearest point to the prototype, the workplane
illuminance value is 800 lx under the maximum setting, with space availability (taking 500 lx
as the minimum criterion) of around 11%. Under 62.5% and 25% settings, none of the points
receives illuminance larger than 500 lx, which is required for typical office work. However,
when the test room is used for other activities which require lower illuminance criteria, the
space availability will be higher. For instance, taking 300 lx as the criterion, approximately
30% space availability can be achieved under the maximum setting, as observed from Figure
5.15c. The light distribution is symmetrical along the prototype’s central axis.
Vertical illuminance on the observer’s eye plane (Ev [lx]), together with minimum (Lmin
2
2
2
[cd/m ]), maximum (Lmax [cd/m ]), and average luminance (Lav [cd/m ]) perceived by the
observer at all positions (referring to Figure 5.10) are displayed in Table 5.8. In addition, the
DGP values obtained from Evalglare are also given.
92
Table 5.8. Vertical illuminance on the observer’s eye, minimum, maximum, average luminance,
and DGP perceived by the observer at all positions at 25%, 62.5%, and 100% of the maximum
setting
Ev [lx]
Position
1 – at 25%
2 – at 25%
3 – at 25%
4 – at 25%
1 – at 62.5%
2 – at 62.5%
3 – at 62.5%
4 – at 62.5%
1 – at 100%
2 – at 100%
3 – at 100%
4 – at 100%
124
37
55
57
316
94
144
148
521
177
233
239
Lmin
Lmax
Lav
2
2
[cd/m ] [cd/m ] [cd/m2]
1.09
57000
60
0.69
3060
18
0.44
480
13
0.50
540
13
2.44
143000
155
1.88
17500
49
1.13
1740
34
1.48
1310
34
2.94
320000
255
1.57
36000
89
1.80
2270
55
2.58
2130
55
DGP
[-]
0.13*
0.01*
0.03*
0.02*
0.21
0.10*
0.14*
0.15*
0.26
0.14*
0.21
0.21
*values are lower than 0.20, i.e. the minimum defined for DGP
(a)
(b)
(c)
(d)
Figure 5.16. HDR view of the test room with the prototype under the maximum (100%) setting,
observed at position (a) 1, (b) 2, (c) 3, and (d) 4
93
Chapter 5
Figure 5.16 displays combined high dynamic range (HDR) images of the test room under
the maximum (100%) setting, observed from position 1 until 4. Figure 5.17 displays the
luminance false colour pictures of Figure 5.16, generated by Photolux 3.1 software.
(a)
(b)
(c)
(d)
Figure 5.17. Luminance false colour pictures of the prototype under the maximum (100%) setting,
observed at position (a) 1, (b) 2, (c) 3, and (d) 4
From Figures 5.16a and 5.16b, one can observe visible patches of light on the two side
walls. These patches are in fact the intended result from installing the direct lamps arrays.
The effect can be varied or extended if necessary, for example by controlling the light output
of each lamp, or adding Venetian blinds to create structured patches on the walls.
Based on the measurement data in Table 5.8 and view images in Figures 5.16 and 5.17, it
is revealed that position 1 experienced the worst glare perception, due to the large source area
observed. The luminance of the observed window area was in the range of 1000 ~ 5000
cd/m2 (see Figure 5.17a). The maximum average luminance was 255 cd/m2; this is
approximately two times larger than the maximum average luminance measured from the
first generation prototype in Chapter 3 (see Table 3.6).
Based on the criteria of Jakubiec & Reinhart (2012), a DGP range of 0.30 ~ 0.35
corresponds to a ‘perceptible’ category, while DGP values of < 0.30 are considered
94
‘imperceptible’; therefore the largest DGP in this case (i.e. 0.26) was still classified as
imperceptible. At the other positions and settings, the perceived glare was below the
minimum defined value of DGP (i.e. 0.20).
5.7.2. Simulation of actual test room
Table 5.9 summarises the simulation results of the horizontal illuminance point at the
middle row on the workplane, together with the average illuminance values (Eav [lx]),
uniformity (U0), and space availability (%A [%]) under the three settings. For comparison, the
measurement results and the relative differences are also shown.
Table 5.9. Simulation (sim.) and measurement (meas.) results of horizontal illuminance point at
the middle row, together with the average illuminance values (Eav [lx]), uniformity (U0), and space
availability (%A [%]) under the three settings
25%
62.5%
100%
Distance to Sim. Meas. Ratio Sim. Meas. Ratio Sim. Meas. Ratio
window [m] [lx]
[lx]
[-]
[lx]
[lx]
[-]
[lx]
[lx]
[-]
0.5
171
165
1.03
430
459
0.94
807
858
0.94
1.0
147
137
1.07
365
382
0.95
684
677
1.01
1.5
106
97
1.10
273
272
1.01
511
480
1.07
2.0
78
71
1.10
198
203
0.98
364
349
1.04
2.5
58
52
1.12
147
155
0.95
270
254
1.06
3.0
43
41
1.07
108
118
0.91
208
204
1.02
3.4
36
36
1.02
93
94
0.98
172
163
1.06
3.9
28
29
0.97
69
79
0.88
127
135
0.94
5.4
22
25
0.88
56
66
0.84
105
111
0.94
4.9
19
22
0.86
46
57
0.81
89
98
0.90
5.3
17
20
0.85
41
51
0.81
74
88
0.84
5.8
15
20
0.74
38
49
0.77
68
81
0.84
6.3
14
19
0.72
36
47
0.76
65
78
0.83
Eav [lx]
50
49
1.01
124
136
0.91
236
239
0.99
U0 [-]
0.26
0.37
0.69
0.26
0.33
0.79
0.27
0.30
0.88
%A [%]
0
0
n/a
0
0
n/a
12
11
1.10
95
Chapter 5
The lighting simulation and measurement results of the prototype generally show a good
agreement, with a maximum relative difference of 28% at the farthest point under the 25%
setting, possibly dominated by measurement accuracy limits, since the absolute difference is
only 5 lx. However, ratio of the simulated value to the measured one (or measured to
simulated, whichever is greater) at all points is actually always less than 1.5, which represents
the smallest significant difference in subjective effect of illuminance (CEN, 2002). Therefore,
the computational model is considered sufficient for the purpose of reproducing the scenes
without a significant subjective difference, and can be further extended for non-existing
solutions.
Figure 5.18 displays the graphs showing the relationship between horizontal illuminance
and the distance to the windows under the three defined settings, based on the measurement
and simulation.
(a)
(b)
(c)
Figure 5.18. Graphs showing the relationship between horizontal illuminance and distance to
windows under the (a) 25%, (b) 62.5%, and (c) 100% of the maximum setting
96
5.7.3. Comparison of various configurations
The relationship between space availability, uniformity, and maximum DGP in the test
room under the simulated configurations (Figure 5.12) is illustrated in Figure 5.19. It is
observed that when 200 lx is taken as the criterion for workplane illuminance, nearly all
configurations yield a space availability of 100% or very close to it, except Configuration 1 in
which all of the four openings are placed on a short wall. Consequently, the far side of the
room is left without sufficient light. When 300 lx is taken as the criterion, only
Configurations 2 (two openings on each short wall facing each other) and 5 (four openings on
a long wall) yield space availabilities of more than 90%. When 500 lx is taken as the
criterion, all configurations yield space availabilities of less than 50%, the highest being
Configuration 3 (two openings on each long wall facing each other, 0.14 m distance between
openings on the same wall).
Figure 5.19. Graphs showing the relationship between space availability, uniformity, and
maximum DGP in the test room, under the simulated configurations scenes
The highest uniformity is achieved under Configuration 2 (0.59), and second to that are
Configurations 5 and 7 (0.55). The maximum DGP values under all configurations range are
very similar, within the range of 0.25 ~ 0.30, mostly found at the observer’s positions that are
the closest to the openings, or those that are able to view the entire four openings. The highest
value is found at position 1 under Configuration 1, since the position is located near the wall
97
Chapter 5
where both of the prototypes are placed. Under Configuration 2, the observer at the same
position would experience the second highest DGP value, since each prototype is placed on
the short walls, one of which is at a distance of 1.0 m to the observer’s position.
Figure 5.20 displays false colour maps of horizontal illuminance values on the workplane,
under Configurations 2, 5, 3, and 7. The false colour maps of those under Configurations 1, 4,
and 6 are displayed in Appendix E.
(a)
(b)
(c)
(d)
Figure 5.20. False colour maps of the simulated horizontal illuminance [lx] under Configurations
(a) 2, (b) 5, (c) 3, and (d) 7
The simulation results show that at the nearest points to the openings, the workplane
illuminance values are approximately 800 lx under Configuration 2 (Figure 5.20a), whereas
the values are between 600 ~ 650 lx under Configuration 5 (Figure 5.20b). The distance
between two adjacent openings is 0.14 m in Configuration 2, and 0.80 m in Configuration 5.
In the latter, the individual prototype (4 × 8 tiles) is split into two parts; each consisted of 4 ×
4 tiles. Therefore, the maximum illuminance in Configuration 5 is less than that in
Configuration 2, since the light is spread more evenly and is not concentrated as much as in
Configuration 2. As shown in Figure 5.19, both configurations have a space availability of
100% for 200 lx criterion, and approximately 90% for 300 lx criterion. For 500 lx criterion,
the satisfying workplane area in Configuration 5 is slightly larger than that in Configuration
2.
98
Configuration 3 (Figure 5.20c) yields the largest space availability for 500 lx criterion,
but not the largest for 300 lx criterion. Under this configuration, the prototypes are placed in
the centre of the long walls; therefore the light is more concentrated in the middle part of the
room, and drops towards the edges. Configuration 7 (Figure 5.20d) is close in terms of
performance to Configuration 5, only the light is less concentrated at the nearest points to the
openings, and there is less workplane area with illuminance of higher than 300 lx.
Under Configuration 5, the highest DGP value (0.26) is observed at position 2 (see Figure
5.12), which directly faces four openings; whereas under Configuration 2, the highest DGP
value (0.29) is observed at position 1, as mentioned earlier. This finding leads to a suggestion
of placing workstations and viewing directions that will give the least discomfort glare
perception. For instance, under Configuration 2, the viewing direction at position 2 is
recommended; whereas position 3 or 4 is recommended under Configuration 5.
5.7.4. Comparison of various operating scenarios
5.7.4.1. Space availability
Relationship between the normalised DMX settings and the space availability produced
by two prototypes (four openings) in Configuration 2 is illustrated in Figure 5.21. It is
observed that the space availability is highly sensitive to the choice of illuminance criterion.
Note that the graphs are stepwise due to the discrete number of points used to calculate the
space availability. The higher the illuminance criterion, the fewer points receive illuminance
above that level, and the higher the minimum setting required to reach non-zero space
availability.
Figure 5.21. Space availability and normalised DMX setting relationship, using the criteria of
200, 300, and 500 lx
Using the daily profile in Figure 5.13, one can estimate the space availability at each
hour. In turn, the daily and annual average values can also be estimated for each defined
climate type. Table 5.10 gives the results for each climate type and daily profile setting.
99
Chapter 5
Table 5.10. Estimated daily average space availability [%] in Configuration 2 for each climate
type and daily profile, i.e. ‘spring’ (Sp), ‘summer’ (Su), and ‘winter’ (Wi)
Location
Singapore
Space availability – 500 lx
[%]
Sp
Su
Wi
17
13
13
[%]
Sp
Su
Wi
61
51
49
[%]
Sp
Su
Wi
88
86
83
Cairo
11
0
15
40
3
52
80
17
73
Amsterdam
15
15
0
53
52
2
79
95
8
Sevilla
18
0
13
63
15
47
88
39
70
Chicago
18
6
3
61
31
17
82
82
45
Table 5.11 gives the average annual space availability for each climate type, annual
mode, and scenario. The results suggest that variation within the same climate type is very
small. The difference between mimicking and compensating modes is insignificant, since the
annual space availability is averaged out of the values in a year. As discussed earlier, the
values are indeed sensitive to the choice of required illuminance. The values in Singapore are
found to be the highest, while those in Cairo are the lowest. Note that the profiles are based
on exterior vertical illuminance on the defined day and sky condition. Consequently, the daily
values for ‘spring’ can be larger than those for ‘summer’ in non-tropical locations, since the
sun is in a lower altitude at the spring time, resulting in a larger vertical illuminance.
Table 5.11. Estimated average annual space availability [%] in Configuration 2 for each climate
type, annual mode, and scenario
Space availability – 500 lx [%]
Location
Singapore
Cairo
Amsterdam
Sevilla
Chicago
Location
Singapore
Cairo
Amsterdam
Sevilla
Chicago
Mimicking
Compensating
Sc1 Sc2 Sc3 Sc1
Sc2 Sc3
15
15
15
15
15
15
9
9
7
9
9
10
10
11
11
11
11
10
12
10
10
12
10
12
11
9
10
11
9
10
Mimicking
Sc1 Sc2 Sc3
87
87
86
61
56
47
60
66
67
70
65
66
73
70
73
Compensating
Sc1
Sc2 Sc3
87
86
86
62
57
61
65
64
59
71
66
69
72
69
69
100
Mimicking
Sc1 Sc2
Sc3
57
56
55
33
31
25
36
40
40
46
41
42
42
36
39
Compensating
Sc1
Sc2
Sc3
57
56
54
34
32
37
39
39
35
47
42
45
41
36
38
5.7.4.2. Electrical energy consumption
The relationship between the normalised DMX settings and the real-time power of the
actual, individual prototype is given in Figure 5.22. Linear regression curves are drawn based
on the relationship. One can observe a highly linear relation (R2 > 0.99) between the real-time
power and the normalised DMX settings. Moreover, the linear regression equation can be
applied to predict the real-time power of the system, which can be expressed as follows:
Wreal = 528.12x + 117.46
(5.1)
where Wreal [W] is the real-time power consumed by the individual prototype, and x is the
normalised DMX setting that ranges from 0 to 1. Equation 5.1 can be applied in particular for
estimating the total electrical energy consumption of the system in a day, given a certain daily
profile as illustrated in Figure 5.13.
Figure 5.22. Relationship between total real-time power and normalised DMX setting of the
actual prototype
Given the obtained Equation 5.1 and assuming that a working day consists of nine bins of
DMX value, i.e. 09.00 ~ 10.00 hrs, 10.00 ~ 11.00 hrs, and so forth until 17.00 ~ 18.00 hrs,
then the total daily electrical energy consumption for two prototypes can be estimated. The
daily profile settings in Figure 5.13 are reflected in the estimated total daily electrical energy
consumption. In the tropical climate region such as Singapore, where the difference between
seasons is minimal, the estimated total daily energy consumption is relatively similar within
the three designated days. Since the daily profiles were determined based on the simulated
vertical illuminance on an exterior point, the values for ‘spring’ are larger than those for
‘summer’ in the non-tropical locations, due to the lower sun altitude at the spring time. The
values for ‘winter’ are generally the smallest because the day is shorter, except for the cases
of Cairo and Sevilla, where the total daily energy consumption in ‘winter’ is larger than that
in ‘summer’.
101
Chapter 5
Assuming that every month has either 20 or 25 working days (i.e. 4 or 5 working
weeks), then the monthly electrical energy consumption can be estimated for each location,
annual mode, and scenario. Summing up the monthly electrical energy consumption gives the
total annual value [kWh] for 260 working days in a year. Table 5.12 shows the results for two
prototypes (Configurations 1 to 7), indicating that in a given climate type, the total annual
electrical energy consumption is relatively similar for the two annual modes and three
scenarios. Applying the compensating mode instead of the mimicking mode essentially
means switching the ‘summer’ and ‘winter’ months, which does not significantly influence
the annual sum. Differences between scenarios 1, 2, and 3 are also small, since most of these
scenarios were created by adding or removing only one or two ‘summer’ (and consequently,
‘winter’) months, based on the actual profile of monthly average solar radiation in the
relevant location.
Table 5.12. Estimated total daily and annual electrical energy consumption [kWh] of two
prototypes for each climate type, annual mode, and scenario
Total daily energy
consumption [kWh]
Location Singapore
Cairo
Amsterdam
Sevilla
Chicago
Spring Summer Winter 9.6
8.2
9.1
9.7
9.4
9.1
5.2
9.3
6.5
8.1
8.9
8.5
4.1
8.4
6.0
Total annual energy consumption [kWh]
Mimicking
Sc1
Sc2
Sc3
2380 2367 2354
1945 1891 1767
1952 2051 2054
2218 2127 2153
2131 2037 2104
Compensating
Sc1
Sc2
Sc3
2434 2417 2400
1961 1907 1977
2029 2026 1926
2228 2136 2201
2121 2027 2054
If real windows are also present in the same room, then under the mimicking mode, one
can expect to experience similar weather conditions shown by the real and virtual windows,
but the interior ambient lighting level may be amplified too much, hence a concern on the
energy use. Under the compensating mode, the ambient lighting level is more stable, but one
may experience a contradicting view and perception between the real and virtual windows.
Nevertheless, the total annual electrical energy consumption under both annual modes is
found very similar within all selected locations, as shown by the results. The value for all
climate types is approximately 2000 kWh, the highest being 2434 kWh in Singapore and the
lowest being 1767 kWh in Cairo.
It should be noticed that the settings of daily profile for all selected climate types are
normalised to the same maximum DMX setting. This means at the maximum setting, for
example, the prototype set for Cairo would give the same luminance as the one set for
Amsterdam. Meanwhile, based on the simulation and weather data, the peak vertical
illuminance and horizontal solar radiation values in both locations would differ by a ratio of
around 1.6. However, it is unknown whether the difference in vertical illuminance and/or
solar radiation has a one-to-one relationship with the expected maximum luminance of the
102
prototype. Given the nature of human perception and preference in visual comfort, most
probably there is a certain upper limit for a ‘tolerable’ window luminance (Kim et al., 2012;
Shin et al., 2012), where any values exceeding that limit would only create discomfort and
energy spill. With this consideration, next to the technical limitation of the applied
luminaires, setting the same maximum value in the daily profile for all selected climate types
is considered appropriate.
A direct consequence of setting the same maximum value for all climate types is that the
one with the least variation between seasons, hence the most consistent daily profile values,
requires the most energy. Figure 5.23 illustrates the average of total annual electrical energy
consumption in the selected climate types, together with their upper and lower limits, which
are taken from the maximum and minimum values when the same daily profile is applied the
entire year. To present the results more generally, all values are normalised to the maximum
value that can be achieved in the defined settings, i.e. the total electrical energy consumed by
the prototypes when they constantly display the maximum setting (x = 1) in Table 5.2 at each
working hour and on each working day in a year. Figure 5.23. Estimated normalised total annual electrical energy consumption with the upper and
lower limits in the five selected climate types
It is observed from Figure 5.23, that within the selected climate types, the annual profile
of Amsterdam yields the most variation in electrical energy consumption, whereas that of
Singapore yields the least. The highest value is achieved when displaying ‘spring’ in Sevilla,
even though the highest on average is found for Singapore. The lowest is for ‘winter’ in
Amsterdam. Displaying ‘summer’ in Cairo for the entire year gives the lowest annual
electrical energy consumption, while displaying ‘winter’ in the same location yields a
relatively high value, due to the sun’s appearance and position at that time. For Sevilla and
Chicago, the highest value is achieved when displaying ‘spring’ for the entire year. Note that
the values in Figure 5.23 are normalised, therefore it is still applicable for other light sources
with a different efficacy.
103
Chapter 5
Even though the presented results have given a general idea on the influence of operating
scenarios on the performance of VNLS, it is noticed that the findings are bounded to the
setup of the particular prototype. There are in fact many more factors that may contribute to
the resulting space availability, for instance, the windows configuration and placement, the
source beam angle, and the display resolution. Nevertheless, the obvious advantage of
applying VNLS in a space is the possibility to put the displays on any wall or ceiling surface,
independent of where the actual façade is located. In this manner, a high uniformity within
the workplane can reasonably be achieved, whereas the same thing is practically impossible
in a sidelit room having only real windows on one façade.
A second generation VNLS prototype has been designed and built by installing an array
of 32 LED tiles and a line of LED linear fixtures with adjustable colour temperatures to
provide direct light into the test room. This particular prototype has an important role in
validating the computational model that can be extended for further development of not-yetexisting VNLS. It is found that under the tested maximum (100%) setting with a view, the
average workplane illuminance of the test room was 239 lx and 11% of its total workplane
area met the target illuminance of 500 lx. Patches of direct light on the side walls could be
created as an intended result from the installation of the LED line.
Simulation and measurement values of horizontal illuminance at certain distances from
the prototype were compared under three settings. The maximum relative difference is 28%,
found at the farthest point under the tested minimum (25%) setting. The ratio of the simulated
value to the measured one at all points is always between 0.67 and 1.50, ensuring no
significant difference in subjective effect of illuminance. Therefore, the computational model
is considered sufficient for the purpose of reproducing the scene without creating a
significant subjective difference.
Based on the comparison of seven configurations of two prototypes with equal total
opening size in the test room, it is found that nearly all configurations, except Configuration 1
(four openings on a short wall), yield a space availability of 100%, taking 200 lx as the
criterion. When 300 lx is taken as the criterion, Configurations 2 (two openings on each short
wall facing each other) and 5 (four openings on a long wall) yield space availabilities of more
than 90%. When 500 lx is taken as the criterion, the configurations yield space availabilities
between 25% and 50%. The highest uniformity is achieved under Configurations 2 (0.59) and
5 and 7 (0.55). The maximum DGP values under all configurations range between 0.25 and
0.30.
Based on the comparison of various operating scenarios with regards to annual space
availability and total electrical energy consumption, calculated for a room with two prototypes in Configuration 2, it is found that:
• Variation of average annual space availability within a given climate type is found to be
very small; the highest being under the annual profile of Singapore, whereas the lowest
104
being under that of Cairo. The values are however sensitive to the chosen criterion of
workplane illuminance.
• Based on the designated daily profiles and annual modes, the normalised, total annual
electrical energy consumption in all climate types is on average within the range of 0.63 ~
0.79, relative to the total electrical energy consumed by the prototype when it constantly
displays the maximum setting. There is no significant difference found between applying
the mimicking and compensating modes. Within the selected climate types, the annual
profile of Amsterdam yields the most variation in electrical energy consumption, whereas
Singapore yields the least.
This prototype could be further improved in terms of display resolution and luminous
efficacy. By applying the latest developed LED technology, it is possible in the future to
display more detailed image views with higher light output and/or lower energy use. Another
feature not yet tested in this prototype is view dynamics. This specific feature can be applied
by further programming, so that the displayed view is constantly changing.
105
Chapter 5
106
Chapter 6
Solutions with a Simplified View and Directional Light
This chapter discusses the modelling and simulation of future (not-yet-existing) VNLS with a
simplified view and directional light, using Radiance to predict the lighting performance. The
model features an array of small light sources with various direct light components,
resembling a simplified view of the sky and ground. Four input variables, i.e. distance
between windows, interval of tilt angle, beam angle, and total luminous flux, were varied to
observe their effect on the lighting performance of a reference office space. The VNLS
models were also compared to corresponding real windows under the standard CIE overcast
sky.
6.1. Introduction
The concept of VNLS is relatively new and the real, ideal product does not exist yet. The
currently available prototypes can not meet all the expectations yet; they are only able to
meet part of the natural light expectation (Mangkuto et al., 2011, 2012). The previous
chapters of this thesis have demonstrated the role of using modelling and simulation of
existing prototypes; Chapter 3 shows comparisons of a first generation prototype (van Loenen
et al., 2007) to a hypothetical real window of the same sky scenes, Chapter 4 shows
comparisons of subjective discomfort glare perception from a first generation prototype of
Shin et al. (2012) to glare metrics obtained by simulation, while Chapter 5 shows the
possibility to recreate the scenes of a second generation prototype in simulation, of which the
model was validated with the measurement results.
In this chapter, development of not-yet-existing next generation VNLS is discussed, by
introducing a new VNLS model, investigating various possible input variables, and
evaluating their lighting performance. The VNLS model in this case is an improvement of the
‘second generation’ prototype as discussed in Chapter 4. The model was created in Radiance,
in the form of arrays of small directional light sources, providing a ‘simplified view’,
resembling the sky, the horizon, and the ground. The light directionality was optimised by
varying the tilt angle of the light sources in different rows.
In particular, the objectives of this study are twofold. The first objective is to understand
the effect of changing input variables of the VNLS, which in this case are: the window’s
configuration, tilt angles of the sources, beam angle of the sources, and total luminous flux of
the sources; on the lighting performance of a reference office space. The second objective is
to compare the lighting performance of the simulated VNLS in a reference office space,
relative to that of real windows under the standard CIE overcast sky.
107
Chapter 6
The lighting performance is described in terms of the ability to meet the space availability
demand, the illuminance uniformity on the workplane, the illuminance contribution from the
ground elements on the ceiling, and the ability to produce minimal glare at predefined
observer positions in the given space. The space availability is defined here as the percentage
of workplane (at a height of 0.75 m from the floor) meeting a certain minimum illuminance
criteria. VNLS ideally provide space availability comparable to or better than real windows
with the same configuration.
6.2. Methods
6.2.1. Modelling
While all detailed characteristics of the view from a window are considered very
important for developing the requirement of VNLS, the focus in this study is only on
modelling the characteristics of direct and diffuse light from the sky and reflected light from
the exterior ground. One of the reasons that people distinguish the difference between a real
and an existing VNLS prototype with a view, either static or dynamic, may be because the
directionality of the light coming out of the surfaces is different. In general, most of the
existing prototypes behave like a diffuse light source, without the possibility of seeing the
impression of direct and reflected light components on the interior surfaces.
Therefore, in this chapter, a model of VNLS in the form of an array of small light
emitting areas is proposed, displaying a simplified view of green ground, (horizon) and blue
sky, as illustrated in Figure 6.1. The third layer (distant objects such as built landscape) was
not yet represented. The bottom array acted as the ‘ground’ which was tilted upward to mimic
reflected light, directed to the ceiling. The rest of the light sources acted as the ‘sky’, and
were tilted downward to direct the light to the workplane area.
Real windows
VNLS
Figure 6.1. Schematic overview of the VNLS as used in the simulation. The light sources are
constructed in arrays, such that the light from the ‘ground’ is delivered to the ceiling and light
from the ‘sky’ is delivered to the floor
108
Modelling and Simulation of VNLS with a Simplified View and Directional Light
The light emitting areas were modelled to fit two individual vertical windows, each with
the size of 0.80 m × 1.17 m (W×H), which corresponds to a window-to-wall ratio of 20% of a
short wall with the size of 3.6 m × 2.7 m. Each light emitting area in each individual window
had a size of 0.05 m × 0.05 m and resembled a blue sky. In the lowest row, there were four
light emitting areas (0.20 m × 0.20 m each) resembling a green ground surface.
In order to model the directionality of the entering light, the ‘sky’ sources were tilted
downward with a certain interval of tilt angle. Three variations of this interval were
introduced, i.e. 1.0°, 1.5°, and 2.0°. The sources in the row directly above the ‘ground’ were
never tilted (i.e. 0°), while the sources in the second row above the ‘ground’ were tilted
downward by 1.0°, or 1.5°, or 2.0°. The sources in the third row above the ‘ground’ were
tilted downward by 2.0°, or 3.0°, or 4.0°, and so on. As a result of using the defined window
height, the sources in the top row were tilted downward by 20°, 30°, and 40°. The ‘ground’
sources were tilted at a 40° angle pointing upward. Figure 6.2 displays the views of an
individual VNLS with an interval of tilt angle of 2.0°.
Figure 6.2. Front and side views of the individual VNLS with an interval of tilt angle of 2.0°
The sources had a certain beam angle, i.e. the angle between the two directions at which
the luminous intensity is half that of the maximum luminous intensity. To see the effects of
varying beam angle, three values of beam angle for the ‘sky’ were introduced, i.e. 38°
(relatively narrow spread), 76° (medium), and 114° (wide).
The luminous intensity distribution of each light source was written in an IES format
file, based on the character of downlights with a certain beam angle. The distributions in
every row had similar patterns. The luminous intensity values for the sources with a 114°
109
Chapter 6
beam angle were set in such a way that the combination of these sources gives an average
surface luminance (L [cd/m2]) of either 1000 cd/m2 (low luminance setting), 1800 cd/m2
(medium luminance setting), or 3200 cd/m2 (high luminance setting). These are the first three
values used in the experiments with an ‘emulated window’ by Shin et al. (2012).
The luminous intensity values for the ‘sky’ sources with 38° and 76° beam angles were
adjusted accordingly, so that the total luminous flux coming from the ‘sky’ sources altogether
remained the same. The technique for calculating the total luminous flux from the source was
based on the zonal cavity method described by Lindsey (1997). Given the luminous intensity
values at various angles of a luminaire, and assuming that the luminous intensity distribution
is direct (no values for angles more than 90°) and symmetrical around the luminaire’s axis,
the area surrounding the luminaire can be divided into nine zonal cavities, which are the
volumes of conic solid angles with a width of 10°, starting from 0° ~ 10° up to 80° ~ 90°.
The total luminous flux produced by the luminaire (Φ [lm]) can be determined as follows:
9
Φ =  (I N  2 (cos  minN  cos  maxN ))
N 1
(6.1)
The minimum and maximum angles in each zonal cavity (θminN, θmaxN [°]) determine the
zonal constant which is multiplied by the average luminous intensity (IN [cd]) to yield the
luminous flux of that particular zonal cavity. The total luminous flux is then the sum of
luminous flux of all zonal cavities.
For this case, the calculated total luminous fluxes of all ‘sky’ sources (two windows)
were approximately 6200 lm, 11100 lm, and 19900 lm, respectively for the low, medium, and
high luminance settings. Figure 6.3 displays the nine evaluated luminous intensity
distributions of the ‘sky’ sources. Note that the luminous intensity values of the polar
diagrams changed with varying beam angle.
Each ‘ground’ source had a maximum luminous intensity of 110 cd at the low luminance
setting, 199 cd at the medium one, and 354 cd at the high one; all had a similar pattern of
luminous intensity distribution. The beam angle of the ‘ground’ source remained constant at
76° for all variations. A tilt angle of 40° upward was chosen so that the ‘ground’ sources do
not stand completely vertical, which could possibly create too much glare; and that they were
not tilted too much, which could reduce the visibility of the ‘ground’ itself.
110
(a)
(b)
(c)
(d)
(e)
(f)
Figure 6.3. Polar diagram of luminous intensity (in candela) of the light sources resembling the
sky, with beam angles of 38°, 76°, 114° and total luminous flux of 6200, 11100, and 19900 lm
111
Chapter 6
(g)
(h)
(i)
The variation in all input variables is summarised in Table 6.1. In total, there were 54
different combinations possible based on the input variables.
Table 6.1. Input variables and their values
Variable
Symbol
Unit
Values
Distance between windows
d
m
0, 0.75
Interval of tilt angle
IA
deg
1.0, 1.5, 2.0
Beam angle
BA
deg
38, 76, 114
Φ
lm
6200, 11100, 19900
Total luminous flux
112
6.2.2. Settings
The space discussed in this study was a reference office space with dimensions of 5.4 m ×
3.6 m × 2.7 m (L×W×H). There were two vertical window configurations chosen from the
earlier studies of Diepens et al. (2000) and Lawrence Berkeley National Laboratory (LBL)
(2010), see Figure 6.4. Each VNLS was modelled with a simplified view image on its
surface, which is explained in the section 6.2.1. No real windows were present together with
the VNLS in the modelled spaces.
(a)
(b)
Figure 6.4. Elevation views of the VNLS window configurations on the wall
In the given space, VNLS were put on one side of the wall (W 3.6 m × H 2.7 m). Frames
of 5 cm wide were defined at the perimeters of the windows. Reflectance values of the
room’s interior were: ceiling: 85%, walls: 50%, floor: 20%, door: 50%, window and door
frames: 50%; all of which based on the IEA Task 27 reference office (van Dijk & Platzer,
2003).
Three different observer positions, namely A, B, and C, were defined at the eye height of
1.2 m above the floor. Position A was near the window and viewing parallel to the window
plane, B was in the middle of the room and viewing parallel to the window plane opposite to
the viewing position A, while C was near the rear wall and directly facing the window plane,
as shown in Figure 6.5.
113
Chapter 6
Figure 6.4. (a) Plan view and (b) section view of the simulated space
For all simulations, the simulation parameters in Radiance simulations were set as shown
in Table 6.2.
Table 6.2. Radiance simulation parameters
Parameter
Description
Value
-ab
Ambient bounces
4
-aa
Ambient accuracy
0.08
-ar
Ambient resolution
128
-ad
Ambient divisions
1024
-as
256
-ds
Direct sub-sampling
0.2
6.2.3. Assessment
6.2.3.1. Performance indicators
The assessment for this study is based on the performance indicators of interest, which
are:
- Space availability (%A): The percentage of workplane area (at a height of 0.75 m, with a
size equal to the total floor area) with illuminance ≥ 500 lx (typical criterion for office
work). Calculation was performed at 1944 (= 54 × 36) points which were evenly distributed
on the workplane. The %A is the percentage of the number of points with illuminance ≥ 500
lx (n(E ≥ 500 lx)), compared to the total number of points (N).
%A =
n(E  500 lx)
× 100%
N
114
(6.2)
- Uniformity (U0): The ratio between the minimum illuminance (Emin [lx]) to the average (Eav
[lx]); based on the defined calculation points.
U0 =
E min
E av
(6.3)
- Ground contribution (%G): The percentage ratio of illuminance contribution from the
‘ground’ element sources (Eground [lx]) to the total illuminance (Etotal [lx]) received at a
certain point on the ceiling, with the surface normal facing downward (z- axis). Calculation
was performed for N = 10 points on the ceiling, located as displayed in Figure 6.6. The
average value is reported as %Gav.
%G =
E ground
× 100%
E total
(6.4)
N
 %Gi
%Gav =
i 1
N
(6.5)
Figure 6.5. (a) Plan view and (b) section view of calculation points for ground contribution
- Probability of discomfort glare: The normalised values of all potentially relevant glare
indices, i.e. Daylight Glare Probability (DGP), Daylight Glare Index (DGI), Unified Glare
Rating (UGR), and CIE Glare Index (CGI) were calculated with the Evalglare programme
(Wienold & Christoffersen, 2006). Those four indices were taken into account since to the
best of the author’s knowledge, very little is known about which glare indices are most
suited for the case of not-yet-existing VNLS models. Another often-used index, the Visual
Comfort Probability (VCP) was not considered, since it was specially developed for
115
Chapter 6
typically sized, ceiling-mounted, artificial lighting installations with uniform luminances
(Harrold, 2003). Because the calculated glare indices have different ranges of values, it is
intended to normalise the values for the purpose of comparison. The normalisation factors
were suggested by Jakubiec and Reinhart to determine the ‘probability of discomfort glare’,
by multiplying the DGI value with 0.01452, and multiplying UGR and CGI values with
0.01607 (Jakubiec & Reinhart, 2012). No normalisation is required for DGP, since it is
already defined within the range of 0 ~ 1. Thus, the relationships can be expressed as
follows:
DGIn = 0.01452 × DGI
(6.6)
UGRn = 0.01607 × UGR CGIn = 0.01607 × CGI
(6.7)
(6.8)
where DGIn, UGRn, CGIn are the normalised DGI, UGR, and CGI values, respectively.
Next, the four normalised glare indices are averaged, and then the value is reported as the
average probability of discomfort glare (PDGav).
PDGav = (DGP + DGIn + UGRn + CGIn) / 4
(6.9)
6.2.3.2. Sensitivity analysis
To evaluate the effect of the current input variables on the defined performance
indicators, a sensitivity analysis using multiple linear regressions was performed. This
regression model assumes a linear relationship between the output variable yi and the p-vector
of input variables xi. This relationship is modelled through an error variable εi, which is an
unobserved random variable that adds noise to the linear relationship between the output and
input variables. The mathematical model takes the general form as follows:
yi =  1x i 1   2 x i 2   3 x i 3   4 x i 4   i ,
i = 1, 2, …, n
(6.10)
where βi is a p-dimensional regression coefficient. In this case: p = 4, n = 2 × 3 × 3 × 3 = 54,
x1 is the distance between windows (d, in metres), x2 is the interval of tilt angle (IA, in
degrees), x3 is the beam angle (BA, in degrees), and x4 is the total luminous flux (Φ, in
lumens); while y is evaluated for %A, U0, %Gav, and PDGav, individually. Since the variables
have different units, it is intended to standardise the values for each of the output and input
variables, that is:
116
y y
x  xn
yi ' = i
; x ni ' = ni
(6.11)


xn
y
where y i ' and x ni ' are the standardised output and input variables, y i and x ni are the actual
output and input variables, y and x n are the arithmetic mean of the output and input
variables, and  y and  xn are the standard deviation of the output and input variables. The
standardised values were then put in the regression model, which can be expressed in matrix
form as follows:
 y 1' 
 
 ...  =
y n '
  1 '
x 11' x 12 ' x 13 ' x 14 '  

   2 '
...
...
...
...

   ' +
x n1' x n 2 ' x n 3 ' x n 4 '  3 
  4 '
  1'
 
 ...  , n = 54
 n '
(6.12)
The built equations were then solved using a MATLAB toolbox in order to determine  1' ,  2 ' ,  3 ' , and  4 ' ; which are the standard regression coefficients that determine the
sensitivity of the output as function of the input. The standard regression coefficients range
between 1 (strong positive correlation) and –1 (strong negative correlation). The interaction
between input variables was not investigated.
6.2.3.3. Comparison with real windows
As a means of comparison, the VNLS in all configurations were replaced with real
windows (double clear glass 6 mm, transmittance 88.5%) under a CIE overcast sky condition.
The comparison with real windows is considered necessary, since the general concept of this
VNLS type is to increase the possibility of seeing the impression of direct light components
from the sky and reflected light components from the ground, on the interior surfaces. This
impression often appears in a space with real windows, but typically not in a space with a
conventional electrical lighting installation. Since the typical general lighting installation is
ceiling-mounted, a fair comparison with wall-mounted VNLS will be difficult to achieve.
Moreover, the display of a simplified view of blue sky and green ground is also an important
feature of the proposed VNLS model, which should also be compared with a relatively
simple view of overcast sky and plain ground outside the real windows.
The sky condition of the real windows scenes was defined in the Gensky programme in
Radiance, by inserting the zenith radiance value [W/(sr·m2)]. This value was chosen so that
the interior surface of the window will give approximately the same average luminance as the
corresponding VNLS. It should be noted that VNLS with the same total luminous flux can
have a different average surface luminance, particularly when the beam angles are different.
Therefore, each VNLS scene was compared only to the real window scene with
117
Chapter 6
approximately the same average surface luminance. The assessments for determining the
performance indicators were then performed for the real window scenes.
Since the three observer positions in the room were located in such a way that they were
facing different directions, one can presume that position C, which was directly facing the
window, may experience the most severe glare amongst the three positions. Therefore, the
glare assessment may be reduced to focus only on position C, as it will be sufficient to reflect
the worst situation. To demonstrate this, the glare assessment for positions A, B, and C were
performed on the following sample of variations:
•
•
•
•
VNLS: configuration 1a (d = 0), IA = 2.0°, BA = 76°, Φ = 11100 lm, Lav = 3200 cd/m2
VNLS: configuration 2a (d = 0.75 m), IA = 2.0°, BA = 76°, Φ = 11100 lm, Lav = 3200 cd/m2
Real window: configuration 1a (d = 0), Lav = 3200 cd/m2
Real window: configuration 2a (d = 0.75 m), Lav = 3200 cd/m2
To evaluate the performance of all VNLS variations, four performance criteria were
applied on the relative comparison between performance indicators of the VNLS and the real
windows with the same average surface luminance. These were based on the expected benefit
of having VNLS, i.e. gaining more well-lit and uniform space; while maintaining the ground
contribution on the ceiling and the probability of discomfort glare comparable to those in real
windows scenes. The criteria are defined in terms of a ratio, which was evaluated up to one
significant digit after the decimal point. The criteria are as follows:
• The VNLS should create equal or larger space availability, compared to the real windows.
• The VNLS should create equal or better illuminance uniformity, compared to the real
windows.
• The VNLS should create average ground contribution on the ceiling that is within ±0.1
(10%) of that in the real windows scene.
• The VNLS should create equal or smaller average probability of discomfort glare as
observed in position C, compared to the real windows.
The criteria are expressed in mathematical forms as follows.
%A V
≥ 1.0
%A R
(6.13)
U0 V
≥ 1.0
U0 R
(6.14)
0.9 ≤
%Gav V
≤ 1.1 %Gav R
PDGav R
≥ 1.0
PDGav V
118
(6.15)
(6.16)
where the subscripts of V and R correspond to the VNLS and real windows scene with the
same average surface luminance.
Moreover, it is preferable to have an average surface luminance which does not exceed
3200 cd/m2. This was the value given in the experiments of Shin et al. (2012), where the
subjects on average perceived glare from the simulated windows as ‘acceptable’, i.e. scored
as 2.5 out of 4.5 on their discomfort glare rating scale.
The results of glare assessment for positions A, B, and C performed on the four
aforementioned variations are shown in Table 6.3. The average probabilities of discomfort
glare are shown together with their standard deviations.
Table 6.3. Results of glare assessment for positions A, B, and C performed on the four variations
in both VNLS and real windows (RW) scenes
Type
RW
VNLS
RW
VNLS
Conf. IA [°] BA[°] Φ [lm] Pos.
1a
1a
2a
2a
Lav = 3200 cd/m2
2.0
76
11100
Lav = 3200 cd/m2
2.0
76
11100
DGP
DGIn UGRn CGIn
A
0.24
0.21
0.35
B
0.21
0.19
C
0.26
A
PDGav
SD
0.39
0.30
0.08
0.31
0.33
0.26
0.07
0.31
0.43
0.45
0.36
0.09
0.24
0.21
0.36
0.39
0.30
0.09
B
0.21
0.20
0.32
0.35
0.27
0.08
C
0.27
0.33
0.46
0.48
0.38
0.10
A
0.22
0.26
0.36
0.39
0.31
0.08
B
0.21
0.22
0.32
0.34
0.27
0.07
C
0.26
0.33
0.43
0.45
0.37
0.09
A
0.21
0.17
0.32
0.35
0.26
0.09
B
0.21
0.21
0.31
0.34
0.27
0.07
C
0.27
0.34
0.45
0.47
0.38
0.09
Since the standard deviations in VNLS scenes are found to be very similar and never
differing more than 0.01 from their real windows counterpart, the average probability of
discomfort glare can be taken as an indicator for both the VNLS and the real windows scene.
The results also show that the probability of discomfort glare at position C is always found to
be the largest; hence it is considered sufficient to take into account only this position in the
complete glare assessment for the entire variations.
Table 6.4 summarises the space availability (%A), uniformity (U0), ground contribution
(%Gav) and average probability of discomfort glare (PDGav) for all window variations/
119
Chapter 6
configurations with total luminous flux of 11100 lm. Note that configurations with the same
distance between windows, beam angle, and total luminous flux are compared to the same
reference real window, of which the performance indicators are shown directly above them in
the table. For instance, configurations (1a, IA = 2.0°, BA = 38°, Φ = 11100 lm), (1a, IA =
1.5°, BA = 38°, Φ = 11100 lm), and (1a, IA = 1.0°, BA = 38°, Φ = 11100 lm), are all
compared to real windows with an average surface luminance of 10000 cd/m2.
Table 6.4. Summary of space availability, uniformity, average ground contribution, and
probability of discomfort glare for all variations and position C in both VNLS and real windows
(RW) scenes with Φ = 11100 lm
Conf.
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
2a
2a
2a 2a 2a
2a 2a 2a 2a
2a 2a 2a IA [°]
BA [°]
Φ [lm]
RW – 10000 cd/m2
2.0
38
11100
1.5
38
11100
1.0
38
11100
RW – 3200 cd/m2
2.0
76
11100
1.5
76
11100
1.0
76
11100
RW – 1800 cd/m2
2.0
114
11100
1.5
114
11100
1.0
114
11100
RW – 10000 cd/m2
2.0
38
11100
1.5
38
11100
1.0
38
11100
RW – 3200 cd/m2
2.0
76
11100
1.5
76
11100
1.0
76
11100
RW – 1800 cd/m2
2.0
114
11100
1.5
114
11100
1.0
114
11100
120
%A
[%]
70
32
33
34
27
31
32
32
14
28
29
30
63
35
35
32
30
34
34
33
15
29
31
31
U0
[-]
0.19
0.21
0.23
0.26
0.16
0.28
0.30
0.32
0.18
0.37
0.37
0.37
0.17
0.23
0.25
0.28
0.15
0.32
0.33
0.35
0.16
0.36
0.36
0.38
%Gav
[%]
51
61
59
55
51
50
47
44
50
49
47
45
48
59
57
55
50
49
47
44
48
48
46
45
PDGav
[-]
0.42
0.43
0.45
0.46
0.36
0.38
0.39
0.39
0.34
0.35
0.35
0.35
0.43
0.44
0.46
0.47
0.37
0.38
0.39
0.39
0.35
0.35
0.35
0.35
6.3.1. Sensitivity analysis
Figure 6.7 displays the standard regression coefficient (β’) of all input variables (i.e.,
distance between windows (d), interval of tilt angle (IA), beam angle (BA), and total
luminous flux (Φ)), evaluated for the four performance indicators, i.e. %A, U0, %Gav, and
PDGav. The coefficients of determination values R2 are respectively 0.97, 0.95, 0.80, and
0.96.
Figure 6.6. Standard regression coefficient of all input variables (i.e. distance between windows,
interval of tilt angle, beam angle, and total luminous flux), evaluated for the four performance
indicators (i.e. %A, U0, %Gav, and PDGav)
As can be seen in the graph, luminous flux and beam angle are the most influential input
variables. The graphs showing the relationship between arithmetic mean of the output and the
most influential input variable(s) with a 95% confidence level are displayed in Figure 6.8.
Table 6.5 gives the summary of arithmetic mean, minimum, maximum, standard deviation,
and 95% confidence level of the output.
Table 6.5. Summary of mean, minimum, maximum, standard deviation, and 95% confidence level
of the output and the most influential input variable(s)
Input
Output
Mean
Min.
Max.
SD
Confd.
95%
%A [%]
10
32
70
1
28
56
16
35
84
3
2
9
2
1
4
Φ = 6200 lm
Φ = 11100 lm
Φ = 19900 lm
121
Chapter 6
Table 6.5. (continued)
Input
BA = 38°
BA = 76°
BA = 114°
BA = 38°
BA = 76°
BA = 114°
BA = 38°
BA = 76°
BA = 114°
Output
U0 [-]
%Gav [%]
PDGav [-]
Mean
Min.
Max.
SD
Confd.
95%
0.24
0.32
0.37
57
47
46
0.45
0.39
0.35
0.21
0.28
0.36
54
44
44
0.40
0.36
0.32
0.28
0.35
0.38
61
50
49
0.51
0.42
0.38
0.02
0.02
0.00
2
2
2
0.03
0.02
0.02
0.01
0.01
0.00
1
1
1
0.01
0.01
0.01
(a)
(b)
(c)
(d)
Figure 6.7. Graphs showing the relationship between arithmetic mean of the output and the most
influential input variable(s), with a 95% confidence level
122
The space availability is highly, positively influenced (β’ = 0.98) by the total luminous
flux of VNLS. Figure 6.8a shows that on average, a total luminous flux of 6200 lm, 11100
lm, and 19900 lm will create space availability of around 10%, 32%, and 70%, respectively.
Note that the total luminous flux values are set on a logarithmic scale, with an increment
factor of around 1.8. The mean space availability values increase with a larger factor; that is
3.2 when increased from 6200 lm to 11100 lm, and 2.2 when increased from 11000 lm to
19900 lm.
6.3.1.2. Uniformity
The uniformity is highly, positively influenced (β’ = 0.94) by the beam angle of VNLS.
On average, a beam angle of 38°, 76°, and 114° will create uniformity of around 0.24, 0.32,
and 0.37, respectively (see Figure 6.8b). The relationship between these input and output
variables is almost perfectly linear.
6.3.1.3. Ground contribution on the ceiling
The average ground contribution on the ceiling is highly, negatively influenced (β’ =
–0.82) by the beam angle of the VNLS. On average, a beam angle of 38°, 76°, and 114° will
create an average ground contribution of around 57%, 47%, and 46%, respectively. The mean
output values are decreased by around 10% (absolute difference), when the input is increased
from 38° to 76°; but they are only decreased by around 0.4% when the input is increased
from 76° to 114°, see Figure 6.8c.
6.3.1.4. Probability of discomfort glare
The average probability of discomfort glare is highly, negatively influenced (β’ = –0.85)
by the beam angle of the VNLS. Figure 6.8d shows that on average, a beam angle of 38°, 76°,
and 114° will create an average probability of discomfort glare of around 0.45, 0.39, and
0.35, respectively, as observed at position C. The relationship between these input and output
variables is almost perfectly linear.
6.3.2. Comparison with real windows
Table 6.6 summarises the ratio of space availability, uniformity, and average ground
contribution of each VNLS configuration to those of real windows with the same average
surface luminance; and the ratio of the average probability of discomfort glare at position C
in the real windows scene to that in a VNLS scene with the same average surface luminance;
with total luminous flux of 11100 lm. The bold-typed values are those satisfying the criteria,
given that the average surface luminance should not exceed 3200 cd/m2.
123
Chapter 6
Table 6.6. Ratio of %A, U0, %Gav,of each VNLS configuration to those of real windows with the
same average surface luminance; and ratio of PDGav at position C in real windows scene to that in
VNLS scene with the same average surface luminance, with Φ = 11100 lm
Conf. IA [°] BA [°] Φ [lm]
%A V
%A R
U0 V
U0 R
%Gav V
%Gav R
PDGav R
PDGav V
1a
2.0
38
11100
0.5
1.1
1.2
1.0
1a
1.5
38
11100 0.5
1.2
1.1
0.9
1a
1.0
38
11100 0.5
1.4
1.1
0.9
1a
2.0
76
11100 1.1
1.7
1.0
0.9
1a
1.5
76
11100 1.2
1.8
0.9
0.9
1a
1.0
76
11100 1.2
2.0
0.9
0.9
1a
2.0
114
11100 2.0
2.1
1.0
1.0
1a
1.5
114
11100 2.0
2.1
0.9
1.0
1a
1.0
114
11100 2.1
2.1
0.9
1.0
2a
2.0
38
11100
0.6
1.3
1.2
1.0
2a 1.5
38
11100 0.6
1.4
1.2
0.9
2a 1.0
38
11100 0.5
1.6
1.1
0.9
2a 2.0
76
11100 1.1
2.1
1.0
1.0
2a 1.5
76
11100 1.1
2.1
0.9
0.9
2a 1.0
76
11100 1.1
2.3
0.9
0.9
2a 2.0
114
11100 2.0
2.2
1.0
1.0
2a 1.5
114
11100 2.1
2.3
1.0
1.0
2a 1.0
114
11100 2.1
2.3
0.9
1.0
The results show that most of the VNLS with a beam angle of 38° (narrow spread) fail to
create larger space availability relative to the real windows. In terms of space availability,
most of the VNLS with a beam angle of 76° (medium spread) yield a ratio of around 1.0,
relative to the corresponding real window, which mean they perform the closest to real
windows.
Meanwhile, all of the VNLS with a beam angle of 114° (wide spread) satisfy all criteria
and yield a ratio of space availability of around 2.0, relative to the corresponding real
window. This means the VNLS with a beam angle of 114° can outperform real windows, by
giving a larger space availability. A luminous intensity from a VNLS with a 114° beam angle
is more evenly distributed throughout the space; hence more space can satisfy the illuminance
criterion on the workplane. The appearance of the CIE overcast sky model for the real
windows, which is typically characterised by an almost diffuse luminous intensity
distribution pattern over the workplane, can be best approached by using a wide spread beam
angle for the VNLS model. Note that all of the modelled VNLS, regardless their beam angles,
124
yield larger uniformity than the corresponding real windows, due to the variation of tilt
angles of the ‘sky’ sources, which in turn deliver more light to the rear side of the room.
6.3.2.1. Space availability and uniformity
Within the variations satisfying all criteria, the gain of space availability is between 1.1 ~
2.3, and the gain of uniformity is between 1.4 ~ 2.5, compared to real windows. For example,
for an office of 19.4 m2 floor area as in this case, the real windows with an average surface
luminance of 1800 cd/m2 produce a daylit area of approximately 2.9 m2. If VNLS with 114°
beam angle and the same average surface luminance are used instead of the real windows,
they can produce a daylit area of approximately 5.7 m2 ~ 6.0 m2, which is a 100% increase.
The uniformity is also increased from 0.16 in the real windows scene to 0.36 in VNLS scene.
Figure 6.9 displays two sets examples of images with isolux contour lines on the
workplane of the following configurations which satisfy all criteria:
•
•
•
•
Real window: configuration 1a (d = 0), Lav = 1800 cd/m2
VNLS: configuration 1a (d = 0), IA = 2.0°, BA = 114°, Φ = 11100 lm, Lav = 1800 cd/m2
Real window: configuration 2a (d = 0.75 m), Lav = 1800 cd/m2
VNLS: configuration 2a (d = 0.75 m), IA = 2.0°, BA = 114°, Φ = 11100 lm, Lav = 1800
cd/m2
(a)
(b)
(c)
(d)
Figure 6.8. Isolux contour lines on the workplane of configurations (a) real window 1a (d = 0),
Lav = 1800 cd/m2; (b) VNLS 1a (d = 0), IA = 2.0°, BA = 114°, Φ = 11100 lm; (c) real window 2a
(d = 0.75 m), Lav = 1800 cd/m2; (d) VNLS 2a (d = 0.75 m), IA = 2.0°, BA = 114°, Φ = 11100 lm
125
Chapter 6
From these shown examples, it can be seen that the VNLS have an isolux pattern similar
to the corresponding real windows. The contour lines for 500 lx values are however located at
different distances from the window. The area covered by the 500 lx contour line, which is
the space availability, in the VNLS scene is approximately double the size of that in the real
windows scene. The uniformity under the VNLS is also larger compared to the real window,
by approximately the same factor of 2.
Within the variations satisfying all criteria, the ratio of average ground contribution on the
ceiling in the VNLS and real windows scene is within the range of 0.9 ~ 1.1. However, a
relatively large difference is found between the pattern of ground contribution propagation in
the VNLS and real windows scene. Figure 6.10 displays graphs showing the ground
contribution propagation for a selected number of VNLS variations; all with Φ = 11100 lm,
together with the reference real window case.
(a)
(b)
(c)
(d)
Figure 6.9. Graphs showing propagation of ground contribution on the ceiling for a selected
number of variations of VNLS (solid lines), displayed together with the reference real windows
(dotted line)
126
From Figure 6.10, it can be seen that the ground contribution on the ceiling propagates
differently under real windows and under VNLS with different beam angles. In the real
windows scene, the ground contribution values are always within 30% to 70% (configuration
1a) or within 40% to 70% (configuration 2a). The propagation also shows more fluctuations,
since the received light on the ceiling mostly come from reflections from the ground outside
or from other interior surfaces. In the VNLS scene, the values can reach up to 90% near the
window, but then rapidly decline. The received light on the ceiling mostly come from the
‘ground’ light sources, therefore the propagation is smoother. The row of light sources that
represents the ground in the VNLS sources have their role here, where the beam angle is set
at 76° (medium spread, constant in all variations) and tilted with a 40° angle pointing upward.
In general, given a constant beam angle and tilt angle of the ‘ground’ source, the lower
the beam angle of the ‘sky’ source, the higher the average ‘ground’ contribution on the
ceiling; since the ‘sky’ will contribute less to the ceiling. The results show that there is an
inverse correlation between the beam angles chosen for the light sources that represent the
‘sky’ and the average ‘ground’ contribution on the ceiling. However, a low beam angle of the
‘sky’ source creates a rapid decline in propagation, making it less similar to the situation with
real windows (with an overcast sky condition).
Within the variations satisfying all criteria, the ratio of average probability of discomfort
glare (real windows to VNLS) is found to be 1.0. In the other variations, this ratio is 0.9, thus
in no cases are these ratios found to be larger than 1.0. It means that relative to the
corresponding real windows, the VNLS generally create similar or slightly worse average
probability of discomfort glare. Figure 6.11 displays the impression of the selected
configurations whose isolux contour lines are displayed in Figure 6.9.
From these shown examples, it can be seen that the VNLS with 114° beam angle have
some similarities and differences compared to the corresponding real windows. While the
average window surface luminance and the average probability of discomfort glare are
approximately similar, a few differences are still recognisable. For instance, the luminance
from the ‘ground’ element sources of VNLS are significantly larger than that from the real
ground element, if viewed from position C. This high luminance is needed for the VNLS to
be able to resemble the impression of ground reflection on the ceiling. Despite the high
luminance of the ‘ground’ sources in VNLS, the average probability of discomfort glare
viewed from the observer positions is still comparable to that in real window scenes. The
wide spread beam angle of the ‘sky’ sources reduces the green-coloured impression on the
ceiling, but on the other hand also distributes the light to a wider area of the space, hence
creating a more uniformly lit space.
127
Chapter 6
(a)
(b)
(c)
(d) Figure 6.10. Impression of configurations (a) real windows 1a (d = 0), Lav = 1800 cd/m2; (b)
VNLS 1a (d = 0), IA = 2.0°, BA = 114°, Φ = 11100 lm; (c) real windows 2a (d = 0.75 m), Lav =
1800 cd/m2; (d) VNLS 2a (d = 0.75 m), IA = 2.0°, BA = 114°, Φ = 11100 lm
One of the key research objectives in this thesis is to investigate if VNLS could be
designed with a similar lighting and view performance to real windows. In this simulation
study, VNLS configurations composed of light emitting sources with a size of 0.05 m × 0.05
m have been developed. It shows the possibility to model the direction of light from the
‘ground’ to the ceiling and from the ‘sky’ to the floor. It is concluded that:
• The total luminous flux of VNLS greatly influences the space availability (standard
regression coefficient β’ = 0.98).
• The beam angle of VNLS greatly influences the uniformity (β’ = 0.94), average ground
contribution on the ceiling (β’ = –0.82), and average probability of discomfort glare (β’ =
–0.85).
128
• Most of the VNLS with a beam angle of 76° (medium spread) yield a ratio of space
availability of around 1.0, relative to the corresponding real window, which mean they
perform the closest to real windows.
• All of the VNLS with a beam angle of 114° (wide spread) satisfy all criteria and yield a
ratio of space availability of around 2.0, relative to the corresponding real window, which
mean they outperform real windows.
Compared to real windows (under the CIE overcast sky), the gain of space availability is
between 1.1 ~ 2.3, and the gain of uniformity is between 1.4 ~ 2.6. For example, real
windows with an average surface luminance of 1800 cd/m2 produce a daylit area of
approximately 2.9 m2 in an office of 19.4 m2 floor area. The VNLS with 114° beam angle and
the same average surface luminance can produce a daylit area of 5.7 m2 ~ 6.0 m2 in the same
space without real windows. The uniformity is also increased from 0.16 to 0.36. To some
extent, it shows the benefit of VNLS compared to the real windows.
As mentioned in the introduction, the work presented in this chapter is a report on
progress in the VNLS development. The results of this study are based on simulation of
VNLS with a rather simple image and the light sources in the same row are all set with a
same horizontal angle. The rendered images are very different from actual window views.
Therefore, further studies involving more detailed images on the window view, as well as
more features of real daylight that influence visual comfort as previously mentioned in Table
2.3 in Chapter 2, are required to improve the similarity to real windows. More sophisticated
configurations and source parameters will be studied to further improve the visual comfort
characteristics. Nevertheless, the results presented in this chapter show clear examples on
how building performance simulation contributes in the research and development of nonexisting solutions, by demonstrating that the numerical model of VNLS can perform better in
some ways than that of real windows.
129
Chapter 6
130
Chapter 7
Solutions with Complex Views and Directional Light
This chapter discusses the modelling and simulation of future (not-yet-existing) VNLS with
complex views and directional light, using Radiance to predict the lighting performance. The
model features an array of small light sources with various direct light components, a
transparent glass surface in front of the light sources array and a two-dimensional image
pasted on this surface. Two input variables were varied to observe their effect on the lighting
performance of a reference office. Ten view images were introduced and compared to each
other.
7.1. Introduction
In Chapter 6 of this thesis, a model of VNLS with a small light source array constructing
a simplified view that resembles the blue sky and green ground has been described. The bluecoloured ‘sky’ elements were tilted downward to deliver light to the workplane, while the
green-coloured ‘ground’ elements were tilted upward to deliver light to the ceiling
(Mangkuto et al., 2014), referring to the ideal CIE overcast sky where the split-flux method
applies (Tregenza, 1989). The results show that the total luminous flux greatly influences the
space availability, while the beam angle is highly influential on the uniformity, average
ground contribution on the ceiling, and average probability of discomfort glare. Most of the
VNLS satisfying all criteria (in terms of ratio, compared to the real windows scene) are those
having a beam angle of 114° (wide spread).
While the findings may give an illustration on how VNLS will perform in a space, it is
noticed that VNLS ideally should generate directional (non-diffuse) light as well as a
relatively complex view, which is not yet the case for the VNLS model in the previous
chapter. In this particular chapter, more complex views are incorporated into the model, while
maintaining the directional light component.
Many researchers have conducted investigation on components that should be present in
viewed images (e.g. Ulrich, 1984; Ulrich et al., 1991; Tennessen & Cimprich, 1995; Chang &
Chen, 2005; de Kort et al., 2006; Aries et al., 2010; Beute & de Kort, 2013). Tuaycharoen &
Tregenza (2007) suggested that view cannot be separated from the natural (day-) light itself.
Beute & de Kort (2013) found consistent preferences for natural over urban scenes, sunny
over overcast scenes, and bright over dim scenes.
Hellinga & de Bruijn-Hordijk (2009) proposed that for the view elements of windows, the
highest quality level will be achieved if the view contains the following:
1. Green, sky, and distant objects
131
Chapter 7
2. Maximum information about outside environment, such as weather, season, time of day,
and (human) activities
3. Highly complex and coherent image scene
The use of complex view on virtual windows in a laboratory environment to investigate
the psychological effects has been explored by some researchers. For example, IJsselsteijn et
al. (2008), who focused on depth perception cues from screen projected images, used five
image scenes, showing the presence of: (1) trees, ground, and three people standing, (2) trees
and ground without people, (3) creek, (4) desert, and (5) city skyline on a river at night. All
scenes other than the last one display a daytime sky.
In their experiments on discomfort glare from projected images, Tuaycharoen & Tregenza
(2005) used 10 pairs of image scenes displaying either ‘natural’ (showing the presence of
mountains, river, and/or trees) or ‘urban’ (showing the presence of buildings, i.e. houses,
skyscrapers, castle, or school) views. A daytime sky (with or without clouds) is visible in
every image scene. They concluded that a good view (also described as a view with high
interest), which mainly consists of the natural scenes, tends to reduce discomfort glare
perception.
Shin et al. (2012), who focused on subjective discomfort glare evaluation from a backlit,
transparent printed image, used five pairs of image scene displaying either ‘distant’ (i.e. the
viewed objects are relatively faraway from the window) or ‘near’ (i.e. the viewed objects are
relatively close to the window). The scene displayed a ‘mixed land’ (skyscrapers and trees),
‘man-made’ (skyscrapers), ‘mixed river’ (city skyline on a river), ‘natural land’ (trees and
green ground), or ‘natural river’ (mountains or plants on a river). All pictures in the scenes
were taken during daytime, but the sky is only (partly) visible on the ‘distant’ scenes, and not
at all on the ‘near’ scenes. They concluded that the tolerance of discomfort glare sensation for
the distant views including skyline was greater than the near views.
Considering the variation and clear distinction of the objects’ distance, the 10 image
scenes of Shin et al. were incorporated to model the VNLS with complex views in this
chapter. Some adaptations were made, including stretching, mirroring, and cropping the
upper and lower part of the original image to get the same height of horizon (i.e. the border
between the ground and the sky), and to get the same image size in every scene. The adapted
image scenes are displayed in Figure 7.1.
132
Modelling and Simulation of VNLS with Complex Views and Directional Light
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
(j)
Figure 7.1. Image scenes adapted from Shin et al. (2012): (a) ‘Distant Mixed Land’ (DML), (b)
‘Near Mixed Land’ (NML), (c) ‘Distant Man-made’ (DMM), (d) ‘Near Man-made’ (NMM), (e)
‘Distant Mixed River’ (DMR), (f) ‘Near Mixed River’ (NMR), (g) ‘Distant Natural Land’ (DNL),
(h) ‘Near Natural Land’ (NNL), (i) ‘Distant Natural River’ (DNR), (j) ‘Near Natural River’ (NNR)
The work described in this chapter aims to demonstrate the role of building performance
simulation in the research and development of VNLS, by predicting the performance of
numerical model of VNLS with complex views and directional light on the display. In
particular, the objectives of this study are twofold. The first objective is to understand the
effect of changing input variables of the VNLS with complex views, which in this case are:
beam angle and total luminous flux of the ‘non-ground’ elements, as well as the variation on
the image scene itself; on the lighting performance of a reference office space. The second
objective is to compare two techniques of modelling the view: using the ‘emissive’ approach,
i.e. where the light sources are coloured and constructing the view itself, and using the
‘transmissive’ approach, i.e. where the light sources are all white and the view is made by
pasting an image on a transparent surface in front of the sources.
Similar to the case of VNLS with a simplified view, the lighting performance is hereby
described in terms of the ability to meet the space availability demand, the illuminance
uniformity on the workplane, the illuminance contribution from the ground elements on the
ceiling, and the ability to produce minimal glare at predefined observer’s positions in the
space.
133
Chapter 7
7.2. Methods
7.2.1. Modelling
While all detailed characteristics of view from a window are considered very important
for developing the requirement of VNLS, this study focuses on modelling the characteristics
of direct light from the sky and reflected light from the exterior ground. In general, most of
the existing VNLS prototypes behave like a diffuse light source, reducing the possibility of
seeing the impression of direct and reflected light components on the interior surface.
Therefore, a more complex model of VNLS is proposed, in the form of arrays of small
directional light emitting areas that are tilted, as specified in Chapter 6 of this thesis. To
realise a complex view in the Radiance simulation tool, a two-dimensional image scene was
imported and pasted/mapped on a very thin, vertically flat, transparent glass (τ = 0.90)
material. The glass itself did not emit light, and it was put in front of the light source arrays.
Ten image scenes adapted from Shin et al. (2012) were used, as displayed in Figure 7.1.
All of the light emitting areas were white, since the actual colour display was given by the
mapped image. The bottom array acted as the ‘ground’ which was tilted upward to deliver the
light to the ceiling. The rest of the sources acted as the ‘sky’ which was tilted downward to
deliver the light to the workplane.
The light emitting areas were modelled to fit two individual vertical openings, each with
the size of 0.80 m × 1.17 m (W×H), corresponding to a window-to-wall ratio of 20%. Each
light emitting area in each individual window had a size of 0.05 m × 0.05 m and had the role
of lighting the ‘non-ground’ part of the image. At the lowest row, there were four light
emitting areas (0.20 m × 0.20 m each) to light the ‘ground’ part of the image.
To model the directionality of light entering through a window, the sources at the row
directly above the ‘ground’ were at all times not tilted (i.e. 0°), while the sources at the
second row above the ‘ground’ were tilted downward by 2.0°. The sources at the third row
above the ‘ground’ were tilted downward by 4.0° and so forth. The ‘ground’ sources were
always tilted with a 40° angle pointing upward.
Figure 7.2 displays the front and side views of the VNLS model.
134
Figure 7.2. Front and side views of the VNLS model
The sources had a certain beam angle, i.e. the angle between the two directions opposed
to each other over the beam axis for which the luminous intensity is half that of the maximum
luminous intensity. To observe the effects of varying beam angle, three values of beam angle
for the ‘non-ground’ sources were used, i.e. 38° (relatively narrow spread), 76° (medium),
and 114° (wide).
The luminous intensity distribution of each light source was written in an IES format
file, based on the character of downlights with a certain beam angle. For the sources with a
114° beam angle, the distributions were set so that the combination of these sources, without
the addition of the transparent glass, gave an average surface luminance (L [cd/m2]) of 1000
cd/m2 (low luminance setting), 1800 cd/m2 (medium), or 3200 cd/m2 (high). These were the
first three values used in the experiments of Shin et al. (2012). The intensity values for the
‘non-ground’ sources with 38° and 76° beam angles were adjusted accordingly, so that the
total luminous flux coming from the ‘non-ground’ sources altogether remained the same. The
total luminous flux from the source was then calculated based on the zonal cavity method
described by Lindsey (1997).
In this case, the calculated total luminous fluxes of all ‘non-ground’ sources, without the
addition of the transparent glass, are approximately 6200 lm, 11100 lm, and 19900 lm. Figure
6.3 is referred to display the luminous intensity distributions of the ‘non-ground’ sources. The
settings for each ‘ground’ source in this case were also identical with the case for VNLS with
a simplified view, having a fixed beam angle of 76 degrees, and maximum intensity of 110
cd, 199 cd, and 354 cd for the three conditions respectively.
The variation in all input variables, including the image scene is summarised in Table
7.1. It should be noticed that the interval of tilt angle and the distance between windows are
not considered as input variables that vary, since the finding in the case of VNLS with a
simplified view suggests that both of them are less influential than the total luminous flux and
beam angle (Mangkuto et al., 2014). In total, there are 90 possible combinations based on the
input variables, taking the image scene variations into account.
135
Chapter 7
Table 7.1. Input variables and their values variation
Variable
Symbol
Unit
BA
deg
38, 76, 114
Total luminous flux
Φ
lm
6200, 11100, 19900
Image scene
-
-
Beam angle
Values
DML, NML, DMR, NMR, DMM, NMM,
DNL, NNL, DNR, NNR
7.2.2. Settings
The space discussed in this study was the same reference office space as discussed in
Chapter 6. Since the findings in Chapter 6 show no significant influence from the distance
between windows to the lighting performance, there was only one window configuration,
chosen for the study in this chapter, illustrated in Figure 6.4a. Reflectance values of the
room’s interior were: ceiling: 85%, walls: 50%, floor: 20%, door: 50%, window and door
frames: 50%; which are based on the IEA Task 27 reference office (van Dijk & Platzer,
2003).
Three observer positions were defined at the eye height of 1.2 m above the floor, as
previously illustrated in Figure 6.5. According to the finding in the case of VNLS with a
simplified view, position C that directly faced the window plane receives the most severe
discomfort glare. Therefore, the glare analysis was performed only for the observer at
position C. For all simulations, the parameters in Radiance were set as previously shown in
Table 6.2.
7.2.3. Assessment
7.2.3.1. Performance indicators
As for the case of VNLS with a simplified view, the assessment for this study is also based
on the relevant performance indicators as discussed in Chapter 6, which are the space
availability, uniformity, average ground contribution on the ceiling, and average probability
of discomfort glare.
7.2.3.2. Sensitivity analysis
Sensitivity analysis using multiple linear regressions was performed to evaluate the
influencing effect of the current input variables on the defined performance indicators. This
regression model assumes a linear relationship between the output variable yi and the p-vector
of input variables xi. The mathematical model reads as follows:
136
y 1   1x i 1   2 x i 2   i ,
i = 1, 2, …, 9
(7.1)
where βi is a p-dimensional regression coefficient. In this case: p = 2, n = 3 × 3 = 9, x1 is the
beam angle (BA, in degrees), and x2 is the total luminous flux (Φ, in lumens); while y is
evaluated individually for %A, U0, %Gav, and PDGav. The calculation is performed
individually for the 10 image scenes, i.e. no mix between different image scenes.
The values were standardised, and then were put in the regression model expressed in the
matrix form as follows:
 y 1'
 
 ...  =
y n '
x 11' x 12 ' 

   1'
...
...

   2 ' +
x n 1' x n 2 '  
  1'
 
 ...  , n = 9
 n '
(7.2)
The equations were then solved using a MATLAB toolbox to determine  1' and  2 ' ,
which are the standard regression coefficients that determine the sensitivity of the output, i.e.
BA and Φ respectively, as function of the input. The values range from 1 (strong, positive
influence) to –1 (strong, negative influence).
7.2.3.3. Comparison of transmissive and emissive approaches
As mentioned in Section 7.1, the second objective of this study is to compare the
approach of modelling the view on VNLS, by observing the effect on the lighting
performance. In Chapter 6 of this thesis, the so-called ‘emissive’ approach was employed to
model the VNLS with a simplified view. In that case, the light sources were either blue or
green-coloured in order to construct the view of the blue sky and green ground. While in this
chapter, the so-called ‘transmissive’ approach is introduced, where the light sources are all
white and do not build the view themselves. The view is realised by pasting a twodimensional image on a transparent, glass surface in front of the sources. It is then intended to
know the impact of using these two approaches on the lighting performance of the given
space.
To make the comparison, two characteristically different image scenes were introduced,
i.e.: (1) a blue sky and green ground (referred as BG), and (2) a green ‘sky’ (entirely
obstructed) and blue ‘ground’ (referred as GB), as in a simplified view of green trees on a
river, seen from a relatively near distance. The image scenes are illustrated in Figure 7.3.
While the first scene is more common, the second scene can be considered located on the
other side of the spectrum. In the GB scene, the general composition of blue and green
colours is inverted; the sky is entirely covered and the ‘ground’ appears brighter.
137
Chapter 7
(a)
(b)
Figure 7.3. Image scenes of (a) ‘blue sky and green ground’ (BG) and (b) ‘green obstructed sky
and blue ground’ (GB)
The image scenes were drawn using the Colour Picker from JALOXA website (Jacobs,
2012b), where colour properties are given in red, green, and blue values between 0 ~ 1 as
used in Radiance. If these values are seen as the spectral reflectance of a given material, then
the weighted average reflectance (ρ) can be obtained using the formula described by Jacobs
(2012a):
ρ = 0.265 ρR + 0.670 ρG + 0.065 ρB
(7.3)
where ρR, ρG, ρB are respectively the spectral reflectance in red, green, and blue. By analogy, a
similar expression can be applied for relating the weighted average transmittance and spectral
transmittance of a given material, which is the case for the transmissive approach. The
equation then reads as follows:
τ = 0.265 τ R + 0.670 τ G + 0.065 τ B
(7.4)
The properties of each element of both image scenes are described in Table 7.2. As
shown there, using the transmissive approach, the shade of green in the GB scene is equal to
that in the BG scene, as displayed in Figure 7.3. To display a darker colour using the emissive
approach, one needs to reduce the luminous intensity of the source, hence also reduce the
total luminous flux. In this comparison, it was chosen to maintain the same total luminous
flux of the light sources in both approaches. Therefore, the colour displays in the emissive
approach are practically different with that in the transmissive approach.
Table 7.2. Colour properties of each element of the BG and GB image scenes
Scene
Element
Red
Green
Blue
Average
BG
‘Sky’
0.500
0.800
1.000
0.734
BG
‘Ground’
0.300
0.500
0.200
0.428
GB
‘Sky’
0.300
0.500
0.200
0.428
GB
‘Ground’
0.500
0.800
1.000
0.734
138
In the emissive approach, the light emitting areas were created from standard IES files
describing the luminous intensities at all relevant angles, then converted into standard
Radiance object files using the programme ies2rad. By default, the light emitting areas are
defined as ‘glow’ materials with colour properties of (1, 1, 1), i.e. white. To change the
colour, one may use the colour properties as suggested by JALOXA, but the values should be
normalised with their weighted average value, before inserting in Radiance. This is necessary
to ensure the final weighted average value stays equal to 1, thus maintaining correct values
for the resulting luminances (Jacobs, 2012a). For example, to display a light-blue colour of
(0.5, 0.8, 1), one has to calculate the weighted average transmittance value using Equation
7.4, which gives 0.734. The normalised values are then (0.5/0.734, 0.8/0.734, 1/0.734), or
(0.682, 1.091, 1.363). These values should be filled in the Radiance object file of the light
source, replacing the default values of (1, 1, 1).
The input variables were beam angle of 38°, 76°, and 114°, interval of tilt angle of 2.0°,
and total luminous flux of the ‘non-ground’ elements of 11100 and 19900 lm. In the emissive
approach, the models were built by assigning the relevant colours to the light sources; while
in the transmissive approach, the relevant image in Figure 7.7 was pasted on the glass surface
in front of the white light sources. Note that for the latter approach, the total luminous flux of
the ‘non-ground’ elements belongs to the light sources only, without considering the
transparent glass surface.
For all variations, the simulations were performed to evaluate the four performance
indicators mentioned in Section 6.3.1.
The results are divided into two main parts, i.e. the sensitivity analysis for the ten image
scenes (Figure 7.1) modelled using the transmissive approach (Section 7.2.1), and analysis of
the BG and GB scenes (Figure 7.3) modelled using the transmissive and emissive approach
(Section 7.2.3.3).
7.3.1. Transmissive approach
The standard regression coefficients of all input variables, i.e. beam angle (BA) and total
luminous flux (Φ), are shown in Figure 7.4. They were evaluated for the four performance
indicators, i.e. %A, U0, %Gav, and PDGav, under the 10 image scenes. The coefficients of
determination values R2 are also shown for each performance indicator and each image scene.
139
Chapter 7
(a)
(b)
(c)
Figure 7.4. Standard regression coefficient of all input variables (i.e. BA and Φ), evaluated for the
four performance indicators, i.e. (a) %A, (b)U0, (c) %Gav, and (d) PDGav under the 10 image scenes
140
(d)
As illustrated in Figure 7.4, total luminous flux is the most influential input variable to the
space availability, while beam angle is the most influential input variable to the other output
variables, independent of the image scene. It is however noticed that under the scenes of
‘Near Natural Land’ and ‘Near Natural River’, the standard regression coefficients (β’) for
total luminous flux related to space availability are respectively 0.56 and 0.59, while the
value under other scenes is always larger than 0.80, which is discussed later on. Figure 7.5
displays the graphs showing the relationship between arithmetic mean of the output and the
most influential input variable(s) with a 95% confidence level, under the 10 image scenes.
(a)
(b)
Figure 7.5. Graphs showing the relationship between arithmetic mean of the output (i.e. %A, U0,
%Gav, and PDGav) and the most influential input variable(s), with a 95% confidence level
141
Chapter 7
(c)
(d)
The space availability is highly, positively influenced (β’ = 0.56 ~ 0.96, depends on the
image scene) by the total luminous flux of the ‘non-ground’ elements. The largest values are
achieved under the ‘Near Mixed River’ and ‘Distant Natural River’ scene, where a total
luminous flux of 19900 lm will create a space availability of around 27% and 24%,
respectively. The mean space availability values increase with a factor of 4 (under NMR
scene) and 5 (under DNR scene) when increased from 6200 lm to 11100 lm, then 2.2 (NMR)
and 2.4 (DNR) when increased from 11000 lm to 19900 lm. Both image scenes provide either
a large white-coloured area (i.e. the skyscraper buildings’ façade in NMR) or a bright, bluishcoloured area (i.e. the sky in DNR). Hence, more light is transmitted through the image plan,
generating a larger workplane area with illuminance exceeding 500 lx.
On the other side of the scale, the ‘Near Natural Land’ and ‘Near Natural River’ scenes
generate the smallest space availability, which is only 1% and 3% when the total luminous
flux is 19900 lm, and zero when the total luminous flux is lower. Both scenes have the sky
entirely covered with either green trees or red plants on a hill. Those elements block most of
the light transmission, resulting in a low value of space availability. Under these two scenes,
increasing total luminous flux does not necessarily increase the space availability, which in
turn makes the standard regression coefficient not as large as under the other eight scenes.
142
7.3.1.2. Uniformity
The uniformity is highly, positively influenced (β’ = 0.75 ~ 1.00) by the beam angle of
the ‘non-ground’ elements. Under a given scene and with a given beam angle, the uniformity
practically stays the same when the total luminous flux is changed. The minimum and
average illuminance values on the workplane are directly proportional to the total light output
of the source, hence the constant value of uniformity. On average, the ‘Near Natural Land’
and ‘Near Natural River’ scenes generate the largest uniformity, which are 0.44 and 0.48 for
beam angle of 114°. While both scenes have the sky entirely covered with relatively uniform,
dim objects (i.e. green trees or red plants), the resulting minimum and average illuminance
values on the workplane are also relatively close to each other.
Under the other eight scenes, the uniformity values range from 0.12 to 0.19 for beam
angle of 38°, 0.22 to 0.28 (76°), and 0.32 to 0.38 (114°). The relationship between the beam
angle and uniformity is almost perfectly linear under all image scenes.
The average ground contribution on the ceiling is highly, negatively influenced (β’ =
–0.91 ~ –0.97) by the beam angle, quite independent of the image scene. As found for the
uniformity, under a given scene and with a given beam angle, the average ground
contribution on the ceiling stays constant when the total luminous flux is changed. A similar
trend is also observed in terms of the image scenes creating the largest %G values. The ‘Near
Natural Land’ and ‘Near Natural River’ scenes generate the largest average ground
contribution on the ceiling, which are respectively 73% and 78% for beam angle of 38°, 65%
and 70% (76°), and 64% and 67% (114°). In both scenes, the view is almost entirely filled
with dim objects (i.e. green trees or red plants), but the ‘ground’ part (i.e. light green grass or
blue river) actually appears brighter than the ‘non-ground’ part. This results in a larger
contribution of illuminance values on the ceiling, compared to the contribution of the ‘nonground’ part.
Meanwhile, in the other scenes, the ‘ground’ part is generally darker than the ‘nonground’ part, creating a lower value of average ground contribution. The scenes of ‘Near
Mixed Land’ and ‘Distant Natural River’ generate the smallest average ground contribution
on the ceiling, which are respectively 30% and 34% for beam angle of 38°, 21% and 22%
(76°), and 17% and 19% (114°).
The average probability of discomfort glare is highly, negatively influenced (β’ = –0.78 ~
–0.89) by the beam angle. Figure 7.5d shows that at a given beam angle, the values under
various image scenes are nearly the same, except again under the ‘Near Natural Land’ and
‘Near Natural River’ scenes, which give lower values. For example, the two scenes give
143
Chapter 7
average probability of discomfort glare of 0.33 and 0.32 (beam angle of 38°), while the
figures are between 0.35 and 0.40 under the other scenes with the same beam angle. As
explained earlier, both scenes have the sky entirely covered with relatively dim objects (i.e.
green trees or red plants), which effectively reduce the transmitted light, and in turn also
reduce the space availability and discomfort glare perception. The relationship between the
beam angle and probability of discomfort glare is almost perfectly linear under all image
scenes. Figure 7.6 displays the rendered impression, observed from position C, of some
selected image scenes, i.e. ‘Near Mixed River’, ‘Near Natural Land’, and ‘Near Natural
River’, with beam angle of 38°, 76°, 114°, and total luminous flux of 19900 lm.
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
Figure 7.6. Impression of some image scenes, all with Φ = 19900 lm: (a) NMR, BA = 38°; (b)
NMR, BA = 76°; (c) NMR, BA = 114°; (d) NNL, BA = 38°; (e) NNL, BA = 76°; (f) NNL, BA =
114°; (g) NNR, BA = 38° (h) NNR, BA = 76°; (i) NNR, BA = 114°
144
While the results of the sensitivity analysis seem to be obvious, the quantitative influence
of each input parameter on each output variables under various image scenes can be obtained
only by simulation. An interesting example is the probability of discomfort glare;
theoretically, as well as empirically, one can assume that the more light output coming from
the display, the more discomfort glare it will create. The reverse can be assumed for beam
angle; the largest the beam angle, the more discomfort glare. These are indeed also found in
the simulation. However, it turns out from the sensitivity analysis that the beam angle is
actually the most influential factor on discomfort glare, and not the total luminous flux. One
potential reason is that the probability of discomfort glare is mostly dominated by the amount
of contrast, and increasing beam angle will result in reducing contrast. Based on the
performed simulation, one can decide better on which input variables to focus on, with regard
to a certain output variables; rather than merely guessing based on visual observation.
7.3.2. Comparison of transmissive and emissive approaches
Table 7.3 summarises the space availability, uniformity, average ground contribution, and
average probability of discomfort glare of the two image scenes, i.e. ‘blue sky and green
ground’ (BG) and ‘green obstructed sky and blue ground’ (GB); and the modelling
approaches, i.e. transmissive and emissive; all with total luminous flux of 11100 and 19900
lm.
Table 7.3. Summary of %A, U0, %Gav, and PDGav of each VNLS configuration with the BG and
GB image scenes, using emissive and transmissive approaches
Φ
[lm]
11100
%A
[%]
17
U0
[-]
0.18
%Gav
BG
BA
[°]
38
[%]
49
PDGav
[-]
0.37
BG
76
11100
13
0.27
38
0.28
BG
114
11100
9
0.37
35
0.25
GB
38
11100
0
0.26
64
0.33
GB
76
11100
0
0.35
55
0.26
GB
114
11100
0
0.42
54
0.23
BG
38
11100
32
0.21
61
0.43
BG
76
11100
31
0.28
50
0.38
BG
114
11100
28
0.37
49
0.35
GB
38
11100
32
0.21
61
0.43
GB
76
11100
31
0.28
50
0.38
GB
114
11100
28
0.37
49
0.35
Approach
Scene
Transmissive
Emissive
145
Chapter 7
Table 7.3. (continued)
%Gav
[%]
32
U0
[-]
0.18
[%]
49
PDGav
[-]
0.41
19900
31
0.26
38
0.30
114
19900
28
0.37
35
0.27
GB
38
19900
7
0.25
64
0.36
GB
76
19900
4
0.35
55
0.28
GB
114
19900
2
0.44
54
0.25
BG
38
19900
72
0.21
60
0.46
BG
76
19900
70
0.28
50
0.41
BG
114
19900
59
0.37
49
0.37
GB
38
19900
72
0.21
60
0.46
GB
76
19900
70
0.28
50
0.41
GB
114
19900
59
0.37
49
0.37
Φ
[lm]
19900
%A
BG
BA
[°]
38
BG
76
BG
Approach
Scene
Transmissive
Emissive
Table 7.3 shows that with the emissive approach, given the same beam angle and total
luminous flux, the performance indicators do not change when a different image scene is
displayed. To display various colours of a ‘glow’ material, one can edit the red, green, and
blue radiance components of the sources, but they have to be normalised using the procedure
given by Jacobs (2012a), so that two light sources with the same light output will produce the
same illuminance values on the same point, even though the colours are different in display.
The consequence of using this technique of producing colours from light is that darker
colours can only be realised by reducing the light intensity. In practice, one may not get the
intended dark colours; for example a completely black colour will be difficult to display,
instead it may just become a dark grey.
Using the transmissive approach, the light output is reduced by the transparent glass on
which the image scene is pasted. It is clearly shown that the GB scene reduces a significant
amount of light, compared to the BG scene. As a result, the space availability under the GB
scene is much smaller than that under the BG scene. Nevertheless, the ground contribution
under the GB scene is relatively larger, due to the fact that the ‘ground’ part appears brighter
than the rest of the display. The probability of discomfort glare under the GB scene is also
smaller than that under the BG scene, obviously due to the smaller amount of light
transmitted from the window display.
To understand the difference between the two approaches, Table 7.3 can be further
simplified by showing the ratio of each performance indicator obtained using the transmissive
approach, compared to that obtained using the emissive approach, for a given beam angle and
total luminous flux. These ratio values are displayed in Table 7.4.
146
Table 7.4. Ratio of %A, U0, %Gav, and PDGav of each VNLS configuration with the BG and GB
image scenes using the transmissive approach (subscript t), compared to those using the emissive
approach (subscript e)
Scene
BA
[°]
Φ
[lm]
%A t
%A e
U0 t
U0 e
%Gav t
%Gav e
PDGav t
PDGav e
BG
38
11100
0.5
0.9
0.8
0.9
BG
76
11100
0.4
1.0
0.8
0.7
BG
114
11100
0.3
1.0
0.7
0.7
GB
38
11100
0.0
1.2
1.1
0.8
GB
76
11100
0.0
1.2
1.1
0.7
GB
114
11100
0.0
1.1
1.1
0.7
BG
38
19900
0.4
0.9
0.8
0.9
BG
76
19900
0.4
0.9
0.8
0.7
BG
114
19900
0.5
1.0
0.7
0.7
GB
38
19900
0.1
1.2
1.1
0.8
GB
76
19900
0.1
1.2
1.1
0.7
GB
114
19900
0.0
1.2
1.1
0.7
It is seen from Table 7.4, that under the BG scene, the space availability using the
transmissive approach is around 30% ~ 50% of the corresponding values using the emissive
approach. Under the GB scene, the space availability using the transmissive approach is very
near or equal to zero. It should be noticed however that the space availability is calculated
based on 500 lx as minimum criterion; hence a value of zero does not necessarily imply that
the entire workplane has zero illuminance, but the maximum illuminance is certainly smaller
than 500 lx. Moreover, even though the glass transmittance for all cases is fixed at τ = 0.90,
the actual light transmitted by the display is apparently much less than this proportion, due to
the additional reduction which largely depends on the view elements of the image scene.
The uniformity and ground contribution under the GB scene is slightly larger when using
the transmissive approach, compared to the emissive one; while the opposite is true under the
BG scene. Under both scenes, the average probability of discomfort glare using the
transmissive approach is around 70% ~ 90% of the values obtained using the emissive
approach.
To give a clearer illustration, Figure 7.7 displays some rendered impression of the scenes
with total luminous flux of 19900 lm and BA = 114°, under both the BG and GB image
scenes, using the emissive and transmissive approaches. In the scenes using the emissive
approach, the side walls apparently show a strong impression of the colour of the ‘nonground’ elements, i.e. blue under the BG scene and green under the GB scene. The floor areas
near the window also appear less bright and less bluish/greenish in the scenes using the
147
Chapter 7
transmissive approach, due to the light reduction by the display. While the emissive approach
generally creates larger space availability, it also generates more contrast between the display
and its immediate surrounding, which leads to a larger probability of discomfort glare.
Scene
Emissive approach
Transmissive approach
BG
GB
Figure 7.7. Impression of the image scenes, all with Φ = 19900 lm and BA = 114°: BG, emissive
approach; BG, transmissive approach; GB, emissive approach; and GB, transmissive approach
A model of VNLS configurations with complex views has been created, where the light
was provided by arrays of white-coloured directional light emitting sources with specific tilt
angles, while the view was provided by mapping a two-dimensional image on a transparent
glass in front of the light sources. It is concluded that:
• Under every image scene, the total luminous flux of the ‘non-ground’ element has a large
influence on the space availability (standard regression coefficient β’ = 0.56 ~ 0.96).
• Under every image scene, the beam angle of the ‘non-ground’ element has a large
influence on the uniformity (β’ = 0.75 ~ 1.00), average ground contribution on the ceiling
(β’ = –0.91 ~ –0.97), and average probability of discomfort glare (β’ = –0.78 ~ –0.89).
148
• The largest space availabilities are achieved under the ‘Near Mixed River’ and ‘Distant
Natural River’ scene, while the smallest are under the ‘Near Natural Land’ and ‘Near
Natural River’ scenes. In turn, the ‘Near Natural Land’ and ‘Near Natural River’ scenes
generate the largest uniformity and ground contribution on the ceiling, as well as the
smallest probability of discomfort glare.
Comparison of the emissive and transmissive approaches shows that the transmissive
approach generally results in smaller values of space availability, relative to the emissive one.
The actual light transmitted is smaller than the transmittance value of the glass; it largely
depends on the colour of the view elements of the image scene. In turn, the average
probability of discomfort glare using the transmissive approach is also smaller than that using
the emissive approach.
These findings can give a first indication on what kind of performance future VNLS
would bring. The use of emissive approach may introduce more light inside the space, but the
view complexity is limited by the number of pixels or individual elements. If this limitation
can be addressed in the future, then this approach will be the reasonable choice. On the other
hand, the transmissive approach offers more flexibility to apply complex views on the
display, but also requires more light, and thus consumes more energy to satisfy the lighting
criteria.
Lastly, the use of computational modelling and simulation has the advantage of saving
time in evaluating various input alternatives, which may include more complex situations
where interactions can become hard to visually observe and theoretically forecast. As
mentioned earlier, there is a long process in developing future solutions such as VNLS. The
use of modelling and simulation is important in influencing the design decision; therefore it
deserves to be discussed on its own.
149
Chapter 7
150
Chapter 8
Conclusions and Recommendations
This chapter summarises the general conclusions of every chapter of this thesis, highlights the
main contribution, and gives recommendations for further research.
8.1. Conclusions
Despite the benefit it gives to human well-being, natural light entrance in buildings in
practice has a limited availability, particularly in space and time. Examples of this situation
can be found, among others, in cubicle workspaces in open-plan offices, operating rooms in
hospitals, and control rooms in industrial plants. To solve the problem in building spaces
without sufficient access to natural light, a number of efforts have been made to recreate the
elements of natural light, in the form of artificial solutions. Such solutions, the so-called
VNLS, can be generally classified based on their light and view qualities into four types,
which are those providing: (1) simplified view and mainly diffuse light, (2) complex view
and mainly diffuse light, (3) simplified view and mainly directional light, and (4) complex
view and mainly directional light.
Subject-based experiments have been performed elsewhere using various form of VNLS
prototypes, to gain knowledge on how people perceive it, and/or to investigate which aspects
of natural light people appraise in reality. Previous research in the health-related area has
suggested that artificial light and view from VNLS prototypes can have positive effects on
subjects. Nonetheless, there is very little exploration on how the prototypes physically
influence the indoor lighting condition of the space where they are installed, as this will also
be related to the total performance of the building where they are located. This suggests a
research direction on how to evaluate the objective performance of the prototypes, and how
to propose better design solutions to improve it. To answer this challenge, this thesis aims to
predict the impact of various VNLS applications on lighting performance and visual comfort
in buildings, by means of computational modelling and simulation.
Based on a comparison of existing prototypes in Chapter 2, it is found that an ideal VNLS
prototype does not yet exist at the moment. Most of the existing prototypes generate only
diffuse light, and each prototype addresses only a subset of the required properties of an ideal
VNLS. There are several properties that need to be improved, for instance, the information
content on the view and the depth perception cues such as blur and motion parallax. Direction
of further development should be therefore steered toward improving the light directionality
and view dynamics, including dynamic elements, e.g. rustling leaves, running water, flying
birds, and moving clouds.
151
Chapter 8
For the purpose of lighting modelling and simulation, Radiance has been validated
elsewhere for daylight and electric lighting scenes, but not yet for VNLS scenes in particular.
An example of the influence of simulation in VNLS development is shown in Chapter 3,
where Radiance was applied to reproduce the scenes and to evaluate the lighting performance
of a first generation VNLS prototype displaying a simplified view of overcast, clear, and
partly cloudy skies, located in a test room with no façades. The key point of this chapter is to
show that simulations can be used to compare an actual VNLS prototype with a hypothetical
real window under the same sky settings, which was physically not possible. Based on the
lighting simulation in Radiance, the investigated prototype performs better in terms of light
distribution uniformity than a corresponding, hypothetical real window under the overcast
and partly cloudy scenes. Under the clear sky scene, the difference between the real and
virtual windows is less, due to the influence of direct sunlight. The findings demonstrate how
a VNLS prototype compares to its real counterpart, in terms of illuminance distribution.
In terms of glare perception, it is generally unknown which glare metrics or ratings are
most suited for the case of VNLS prototypes. In Chapter 4, a method is proposed to correlate
the commonly applied glare metrics, i.e. DGP, DGI, UGR, and CGI, which can be predicted
using simulations in Radiance and Evalglare, to the glare rating that was used in the
experiment of Shin et al. (2012), in order to assess discomfort glare from a first generation
VNLS prototype with various complex views. It was found the simulated values of
normalised DGI, UGR, and CGI are all overestimated relative to the values converted from
the experiment data of Shin et al., while the simulated values of DGP are in a better
agreement with the converted values of DGP. Even though the accuracy of DGP has been
widely reported, all of the earlier findings were based on real daylight scenes. The findings in
this chapter demonstrate the applicability of DGP for the investigated VNLS prototype, and
how it correlates to the subjective glare perception.
A second generation VNLS prototype has been designed and built by installing an array
of LED tiles providing diffuse light and a view, and a line of LED linear fixtures with
adjustable colour temperatures to provide direct light into the test room. This particular
prototype has an important role in validating the computational model that can be extended
for further development of future (not-yet-existing) VNLS. As intended, patches of direct
light could be created and were visible on the side walls. Simulation and measurement values
of horizontal illuminance at certain distances were evaluated and showed a good agreement.
Based on simulation of seven configurations of the second generation prototype with
equal total opening size in the test room, it was found that nearly all configurations yield a
space availability of 100%, taking a workplane illuminance of 200 lx as the criterion. When
300 lx is taken as the criterion, Configurations 2 (two openings on each short wall facing
each other) and 5 (four openings on a long wall) yield space availabilities of more than 90%.
When 500 lx is taken as the criterion, the configurations yield space availabilities between
25% and 50%. The maximum DGP values under all configurations range between 0.25 and
0.30.
152
Various operating schedules were defined and calculated for the second generation
prototype, from 09.00 to 18.00 hrs local time, with one-hour time step on every working day.
Based on the designated daily profiles (i.e. ‘spring’, ‘summer’, and ‘winter’), annual modes
(i.e. mimicking and compensating the real seasons), and climate type scenarios (i.e.
Singapore, Cairo, Sevilla, Amsterdam, and Chicago), the variation of average annual space
availability within a given climate type was found to be very small. The values are however
sensitive to the chosen criterion of workplane illuminance. The normalised, average annual
electrical energy consumption in all climate types is within the range of 0.63 ~ 0.79, relative
to the total electrical energy consumed by the prototype when it constantly displays the
maximum setting at each working hour and on each working day in a year. To further improve the performance of the solution, more complex VNLS configurations
composed of small, light emitting sources have been developed and investigated in a
computational model in Chapter 6. The model comprises of more than 600 small light
emitting sources placed in rows, each of which has a different tilt angle. In this way, the
model can display a simplified view of blue sky and green ground, while also delivering the
light from the ‘ground’ to the ceiling and from the ‘sky’ to the floor. Sensitivity analysis
shows that total luminous flux of the ‘sky’ significantly influences the space availability of
the test room, whereas beam angle of the source largely influences the uniformity, ground
contribution on the ceiling, and probability of discomfort glare.
To increase the view quality, a model of VNLS configurations with complex views has
been created in Chapter 7, by pasting a two-dimensional image on a transparent glass in front
of the light sources. The light was provided by the same configuration of light sources as
used in Chapter 6, but all sources are white-coloured. Comparisons were shown between 10
image scenes. The use of the so-called transmissive approach offers more flexibility to apply
complex display views, but also requires more light, and thus consumes more energy to
satisfy the lighting criteria. On the other hand, the use of the emissive approach in Chapter 6
may introduce more light, but the view complexity is limited by the number of pixels. The
more complex the view, the more individual pixels are required.
To conclude, the main contribution of this thesis is demonstrating the application of
computational modelling and building performance simulation in providing multiple design
concepts to improve the objective performance of VNLS. The defined objectives have been
addressed, and can be summarised as follows:
• Determining the relevant properties and performance indicators for VNLS.
Chapter 2 is dedicated to give an overview of various properties of existing prototypes.
Light directionality and view dynamics are two properties that are the most complicated to
feature. To indicate the benefit of VNLS in term of gaining more ‘daylit’ space, the space
availability is determined as the main light quality performance indicator, and is used in
evaluating all types of VNLS that are discussed in Chapters 3 until 7.
• Finding the appropriate modelling approach in order to model VNLS.
Chapter 3 is opened by giving a brief introduction on lighting modelling and simulation,
and Radiance is selected as the most appropriate tool. Chapter 4 demonstrates how
153
Chapter 8
Evalglare can be employed to predict glare metrics, in comparison to subjective glare
ratings based on an experiment with subjects. Chapter 7 shows a comparison between two
modelling approaches, i.e. an emissive and a transmissive approach for modelling a more
complex VNLS. The transmissive approach has the advantage of creating more possibility
to display complex views, whereas the emissive approach has the advantage of delivering
more light into the space.
• Evaluating the lighting performance and visual comfort of various VNLS models.
Chapters 3 and 4 discuss lighting performance and visual comfort evaluations of the first
generation VNLS prototypes, whereas Chapter 5 discusses those of the second generation
prototype that also serves for validation purpose. Chapters 6 and 7 describe the
development and assessment of the next generation VNLS in computational models. Most
of the modelled, future VNLS with a beam angle of 76° (medium spread) perform the
closest to the corresponding real windows. The performance can be optimised by
increasing the beam angle to 114° (wide spread), yielding a space availability of around
two times larger than the corresponding real windows. The probability of discomfort glare
from most of the models is relatively close to that from the corresponding real windows.
• Finding the potential of applying VNLS under various configurations by predicting their
performance, and under various operating scenarios by estimating total annual electrical
energy consumption.
Chapter 5 particularly addresses these objectives. Space availability under various
configurations of the location of the prototype is evaluated using simulation in a
rectangular test room, whereas the total annual electrical energy consumption is estimated
based on the measured power of the prototype that has been built. The space availability
can be optimised by placing the prototypes facing each other on both short walls, or
placing all prototypes on a long wall. All of the investigated operating scenarios yield a
relatively similar impact on the annual space availability and electrical energy
consumption. However, the values are sensitive to the chosen criterion of workplane
illuminance.
8.2. Recommendations
This thesis focuses on better understanding of how VNLS influence the indoor lighting
performance and visual comfort, and how to design better solutions to improve the light and
view qualities. For that purpose, most of the results in this thesis are based on physical
measurement and/or computational modelling and simulation of VNLS prototypes and
models. Room for improvement is mostly open in the topic of assessing subjective user’s
perception on the particular prototypes and models. Even though there is already first
supporting evidence on the positive effects of some existing (i.e. first generation) VNLS
prototypes in Chapter 2, as the solutions become more complex and sophisticated, further
evaluation and analysis on how people actually appraise VNLS in reality is still necessary,
before proceeding to the design implementation.
154
Recommendations for future research can be summarised as follows:
• In general, the first generation prototype with a simplified view is limited not only in its
view complexity and information content, but also in the possibility to switch between
views. To improve the degree of similarity to the real window, additional features such as
motion parallax and sound transmission can be included. While these features are not
directly influential to the indoor lighting performance, they can be helpful in giving the
impression of having a real window inside the space.
• With regard to the subjective glare experiment of Shin et al. (2012), neither the simulated
values nor the converted values of the glare metrics can be correlated with the actual
percentage of subjects who felt disturbed. This finding suggests two things to improve; on
the subjective experiment, giving fewer semantic scales for glare perception may be more
efficient as it will reduce the chance of confusion and disagreement on the meaning itself.
On the discomfort glare model development, a more appropriate method to approximate
glare in a high contrast environment is required, for instance as proposed by Kleindienst &
Andersen (2009) and Suk et al. (2013).
• It is noticed that the second generation prototype in Chapter 5 was limited in its display
view resolution and luminous efficacy. By applying the latest lighting technology, it is
possible to create a more detailed view on the display with higher light output, while
consuming less energy. To introduce view dynamics, digital programming can be applied
to automate the display variation at every given time step. The application of such
dynamic lighting solutions is also in line with the roadmap of the European Commission
(EC, 2013), which has put healthy and comfortable indoor environment (including air
quality, ventilation, lighting, and acoustic) as one of its cross-platform target areas for
2020. Under this target, future research and innovation topics should be aimed at, among
others, efficient and comfortable indoor lighting; for example by developing flexible
lighting based on LEDs.
• The findings in Chapter 5 give a rough idea on the impact of varying operating schedules
of the second generation prototype, but the results are based on the defined settings that
are rather simplified, with one-hour time step in the daily profiles. Further research should
be directed towards introducing more realistic display scenarios, for instance by
introducing constantly changing weather conditions with a small time step, to closely
follow the variation of real natural light.
• The next generation VNLS has been developed in computational models, using an array of
small light emitting areas. As the size of the individual light source in the model is only
0.05 m × 0.05 m, further models can incorporate more dimensions and shapes to optimise
the output variables, as well as to allow possibility of displaying more detailed views.
• The use of an emissive approach to model VNLS with a complex view in Chapter 6 has
the advantage of introducing more light inside the space, but the view complexity is
limited by the number of pixels or individual elements. In line with the previous point, if
future research can address this challenge, then the emissive approach will become the
155
Chapter 8
most reasonable option in creating efficient solutions. On the other hand, the transmissive
approach offers more flexibility to apply complex views, but requires more light output to
satisfy the lighting criteria. Further research on this approach can be directed towards
finding the most suitable transmissive material for applying the two-dimensional image,
while not losing too much light in the process. Alternatively, further research can be
directed towards combining the two approaches, for example using the emissive one for
the ‘ground’ and the transmissive one for the ‘sky’, or the emissive one for the lower layer
and the transmissive one for the details.
• Finally, to predict the impact of VNLS on total building energy consumption, a detailed
study on thermal properties of a VNLS prototype is required. Such information will be
useful particularly as an input for building energy simulation tools, in determining the total
amount of heating and cooling energy demand of the relevant space.
156
References
Ahn, B.-L., Jang, C.-Y., Leigh, S.-B., Yoo, S., & Jeong, H. (2014). Effect of LED lighting on
the cooling and heating loads in office buildings. Applied Energy, 113, 1484-1489.
Aries, M. B. C., Veitch, J. A., & Newsham, G. (2010). Windows, view, and office
characteristics predict physical and psychological discomfort. Journal of Environmental
Psychology, 30, 533-541.
Aries, M. B. C., Aarts, M. P. J., & van Hoof, J. (2013). Daylight and health: A review of the
evidence and consequences for the built environment. Lighting Research and Technology,
1-22, doi: 10.1177/1477153513509258.
Assem, E. O. & Al-Mumin, A. A. (2010). Code compliance of fully glazed tall office
buildings in hot climate. Energy and Buildings, 42, 1100-1105.
Avery, D. H., Bolte, M. A., & Millet, M. (1992). Bright dawn simulation compared with
bright morning light in the treatment of winter depression. Acta Psychiatrica
Scandinavica, 85, 430-434.
Avery, D. H., Eder, D. N., Bolte, M. A., Hellekson, C. J., Dunner, D. L., Vitiello, M. V. et al.
(2001). Dawn simulation and bright light in the treatment of SAD: a controlled study.
Biological Psychiatry, 50, 205-216.
Badia, P., Myers, B., Boecker, M., Culpepper, J., & Harsh, J. R. (1991). Bright light effects
on body temperature, alertness, EEG and behavior. Physiology and Behavior, 50, 583588.
Berman, M. G., Jonides, J., & Kaplan, S. (2008). The cognitive benefits of interacting with
nature. Psychological Science, 19, 1207-1212.
Blask, D. E. (2009). Melatonin, sleep disturbance and cancer risk. Sleep Medicine Reviews,
13, 257-264.
Beute, F. & de Kort, Y. A. W. (2013). Let the sun shine! Measuring explicit and implicit
preference for environments differing in naturalness, weather type and lightness. Journal
of Environmental Psychology, 36, 162-178.
Beute, F. & de Kort, Y. A. W. (2014). Salutogenic effects of the environment: Review of
health protective effects of nature and daylight. Applied Psychology: Health And WellBeing, 6, 67–95.
Boubekri, M. (2004). An overview of the current state of daylight legislation. Journal of the
Human-Environment System, 7, 57-63.
Boubekri, M. (2008). Daylighting, Architecture and Health: Building Design Strategies.
Oxford: Architectural Press.
157
References
Bouma, P. J. (1948). Physical Aspects of Colour: An Introduction to the Scientific Study of
Colour Stimuli and Colour Sensations. Eindhoven: Philips Gloeilampenfabrieken (Philips
Industries) Technical and Scientific Literature Dept.
Boyce, P. R. (2003). Human Factors in Lighting. (2nd ed.) London: Taylor & Francis.
Boyce, P. R., Hunter, C., & Howlett, O. (2003). The Benefits of Daylight Through Windows
Capturing the Daylight Dividend Program.
CABA (2008). Bright Green Buildings: Convergence of Green and Intelligent Buildings.
Retrieved 27-1-2013, from http://www.caba.org/brightgreen
Cajochen, C. (2007). Alerting effects of light. Sleep Medicine Reviews, 11, 453-464.
CEN (2002). EN 12464-1: Light and Lighting - Lighting of work places, Part 1: Indoor work
places. Brussels: Comité Européen de Normalisation.
Chain, C., Dumortier, D., & Fontoynont, M. (1999). A comprehensive model of luminance,
colour temperature and spectral distribution of skylight – comparison with experimental
data. Solar Energy, 65, 285-295.
Chain, C., Dumortier, D., & Fontoynont, M. (2001). Consideration of daylight's colour.
Energy and Buildings, 33, 193-198.
Chang, C.-Y. & Chen, P.-K. (2005). Human response to window views and indoor plants in
the workplace. HortScience, 40, 1354-1359.
CIE TC-3-33 (2005). Test cases to assess the accuracy of lighting computer programs
Vienna: Commission Internationale de l'Eclairage.
Collins, B. L. (1975). Windows and people: a literature survey. Psychological reaction to
environments with and without windows (Rep. No. NBS-BSS-70). National Bureau of
Standards, Washington, D.C. (USA); U.S. Dept. of Commerce, National Bureau of
Standards, Institute for Applied Technology, Washington, DC.
Collins, B. L. (1976). Review of the psychological reaction to windows. Lighting Research
and Technology, 8, 80-88.
Crawley, D. B. (2008). Building Performance Simulation: A Tool for Policymaking. PhD
thesis, University of Strathclyde, Glasgow.
Cutting, J. E. & Vishton, P. M. (1995). Perceiving layout: The integration, relative
dominance, and contextual use of different information about depth. In W. Epstein & S.
Rogers (Eds.), Handbook of Perception and Cognition: Vol. 5. Perception of Space and
Motion (pp. 69-117). New York: Academic Press.
Davis, W. & Ohno, Y. (2009). Approaches to color rendering measurement. Journal of
Modern Optics, 56, 1412-1419.
158
References
de Kort, Y. A. W., Meijnders, A. L., Sponselee, A.-M., & IJsselsteijn, W. A. (2006). What’s
wrong with virtual trees? Restoring from stress in a mediated environment. Journal of
Environmental Psychology, 26, 309-320.
de Vries, H. J. A., Aarts, M. P. J., Knoop, M., & Cornelissen, H. (2009). Beneficial nonvisual effects of daylight, Research into the influential parameters. In Proceedings of Lux
Europa 2009 - the 11th European Lighting Conference. Istanbul: Aydinlatma Türk Milli
Komitesi.
Denissen, J. J. A., Butalid, L., Penke, L., & van Aken, M. A. G. (2008). The effects of
weather on daily mood: A multilevel approach. Emotion, 8, 662-667.
Diepens, J. F. L., Bakker, F., and Zonneveldt, L. (2000). Daylight Design Variations Book.
Retrieved 12-3-2012, from http://sts.bwk.tue.nl/daylight/varbook/index.htm
Duffy, F., Laing, A., & Crisp, V. (1993). The responsible workplace: The redesign of work
and offices. Oxford: Butterworth Architecture.
Eagles, J. M. (1994). The relationship between mood and daily hours of sunlight in rapid
cycling bipolar illness. Biological Psychiatry, 36, 422-424.
Eastman, C. I., Young, M. A., Fogg, L. F., Liu, L., & Meaden, P. M. (1998). Bright light
treatment of winter depression – A placebo-controlled trial. Archive of General
Psychiatry, 55, 883-889.
EC (2013). Energy-Efficient Buildings: Multi-annual roadmap for the contractual PPP under
Horizon 2020. Luxembourg: European Commission.
Edwards, L. & Torcellini, P. (2002). A Literature Review of the Effects of Natural Light on
Building Occupants (Rep. No. NREL/TP-550-30769). Golden: National Renewable
Energy Laboratory.
Einhorn, H. D. (1979). Discomfort glare: a formula to bridge differences. Lighting Research
EIZO (2013). Altering color dramatically with a single setting: Examining color temperature
on an LCD monitor. Retrieved 06-01-2014, from
http://www.eizo.com/global/library/basics/color_temperature_on_an_LCD_monitor/
Enrech Xena, C. (1999). Simulation de la lumiere naturelle par des sources artificielles:
Enjeux architecturaux. PhD thesis, Université de Nantes, Nantes.
Farley, K. M. J. & Veitch, J. A. (2001). A Room with a View: A Review of the Effects of
Window on Work and Well-being (Rep. No. RR-136). IRC Research.
Fisher, A. (1992). Tolerances in lighting design. In Proceedings of CIE Seminar on Computer
Programs for Light and Lighting. Vienna: Commission Internationale de l'Eclairage.
Flynn, J. E., Hendrick, C., Spencer, T., & Martyniuk, O. (1979). A guide to methodology
procedures for measuring subjective impressions in lighting. Journal of the Illuminating
Engineering Society, 8, 95-110.
159
References
Fontoynont, M. (2011a). Benefits of quality lighting in hospitals. In Konference og Workshop
- Lys som Forbreder Sundhed & Trivsel i Hospitalssektoren. Copenhagen.
Fontoynont, M. (2011b). Solid State Lighting and changed situation for lighting conditions.
In The Future of Energy Efficient Lighting. Katrineholm.
Friedman, B., Freier, N. G., Kahn, J. P., Lin, P., & Sodeman, R. (2008). Office window of the
future? – Field-based analyses of a new use of a large display. International Journal of
Human-Computer Studies, 66, 452-465.
Galasiu, A. D. & Veitch, J. A. (2006). Occupant preferences and satisfaction with the
luminous environment and control systems in daylit offices: a literature review. Energy
and Buildings, 38, 728-736.
Gaver, W. W., Smets, G., & Overbeeke, K. (1995). A virtual window on media space. In
Proceedings of CHI '95 - SIGCHI Conference on Human Factors in Computing Systems
(pp. 257-264). New York: ACM Press/Addison-Wesley Publishing.
Geisler-Moroder, D. & Dur, A. (2010). Estimating Melatonin Suppression and
Photosynthesis Activity in Real-World Scenes from Computer Generated Images Society
for Imaging Science and Technology.
Glass, P., Avery, G. B., Subramanian, K. N. S., Keys, M. P., Sostek, A. M., & Friendly, D. S.
(1985). Effect of Bright Light in the Hospital Nursery on the Incidence of Retinopathy of
Prematurity. New England Journal of Medicine, 313, 401-404.
Glickman, G., Byrne, B., Pineda, C., Hauck, W. W., & Brainard, G. C. (2006). Light therapy
for seasonal affective disorder with blue narrow-band light-emitting diodes (LEDs).
Biological Psychiatry, 59, 502-507.
Hammad, F. & Abu-Hijleh, B. (2010). The energy savings potential of using dynamic
external louvers in an office building. Energy and Buildings, 42, 1888-1895.
Härmä, M. I. & Ilmarinen, J. E. (1999). Towards the 24-hour society – new approaches for
aging shift workers? Scandinavian Journal of Work, Environment & Health, 25, 610-615.
Harrold, R. (2003). IESNA Lighting Ready Reference: A Compendium of Materials from the
IESNA. New York: Illuminating Engineering Society of North America.
Hartig, T., Evans, G. W., Jamner, L. D., Davis, D. S., & Gärling, T. (2003). Tracking
restoration in natural and urban field settings. Journal of Environmental Psychology, 23,
109-123.
Heerwagen, J. H. (1990). The psychological aspects of windows and window design. In
Annual Conference of the Environmental Design Research Association (pp. 269-280).
Heerwagen, J. H. & Orians, G. (1986). Adaptation to windowlessness: A study on the use of
visual decor in windowed and windowless offices. Environment and Behavior, 18, 623639.
160
References
Hellinga, H. Y. & de Bruijn-Hordijk, G. J. (2009). A new method for the analysis of daylight
access and view out. In Proceedings of Lux Europa 2009 - the 11th European Lighting
Conference. Istanbul: Aydinlatma Türk Milli Komitesi.
Heschong, L. (2003a). Window and Classrooms: A Study of Student Performance and the
Indoor Environment (Rep. No. P500-03-082-A-7). California, USA: Heschong Mahone
Group.
Heschong, L. (2003b). Window and Offices: A Study of Office Worker Performance and the
Indoor Environment (Rep. No. P500-03-082-A-9). California, USA: Heschong Mahone
Group.
Hopfe, C. J. (2009). Uncertainty and sensitivity analysis in building performance simulation
for decision support and design optimization. PhD thesis, Eindhoven University of
Technology, Eindhoven.
Hopkinson, R. G. (1972). Glare from daylighting in buildings. Applied Ergonomics, 3, 206215.
IJsselsteijn, W. A., Oosting, W., Vogels, I., de Kort, Y. A. W., & van Loenen, E. J. (2008). A
room with a cue: The efficacy of movement parallax, occlusion, and blur in creating a
virtual window. PRESENCE, 17, 269-282.
Inanici, M. N. & Navvab, M. (2006). The Virtual Lighting Laboratory: Per-Pixel Luminance
Data Anaysis. LEUKOS, 3, 89-104.
Iskra-Golec, I. M., Wazna, A., & Smith, L. (2012). Effects of blue-enriched light on the daily
course of mood, sleepiness and light perception: A field experiment. Lighting Research
Jacobs, A. (2012a). RADIANCE Cookbook. Retrieved 12-3-2012, from
http://www.jaloxa.eu/resources/radiance/documentation/docs/radiance_cookbook.pdf
Jacobs, A. (2012b). Colour Picker for Radiance. Retrieved 2-4-2013, from
http://www.jaloxa.eu/resources/radiance/colour_picker/index.shtml
Jakubiec, J. A. & Reinhart, C. F. (2012). The 'adaptive zone' – A concept for assessing
discomfort glare throughout daylit spaces. Lighting Research and Technology, 44, 149170.
Jenkins, D. & Newborough, M. (2007). An approach for estimating the carbon emissions
associated with office lighting with a daylight contribution. Applied Energy, 84, 608-622.
Kahn, J. P., Friedman, B., Gill, B., Hagman, J., Severson, R. L., Freier, N. G. et al. (2008). A
plasma display window? -- The shifting baseline problem in a technologically mediated
natural world. Journal of Environmental Psychology, 28, 192-199.
Kaplan, R. (1993). The role of nature in the context of the workplace. Landscape and Urban
Planning, 26, 193-201.
161
References
Kaplan, S. (1995). The restorative benefits of nature: Toward an integrative framework.
Journal of Environmental Psychology, 15, 169-182.
Kaplan, R. & Kaplan, S. (1989). The Experience of Nature: A Psychological Perspective.
Cambridge: Cambridge University Press.
Keighley, E. C. (1973a). Visual requirements and reduced fenestration in office buildings - A
study of window shape. Building Science, 8, 311-320.
Keighley, E. C. (1973b). Visual requirements and reduced fenestration in offices - A study of
multiple apertures and window area. Building Science, 8, 321-331.
Keller, M. C., Fredrickson, B. L., Côté, S., Johnson, K., Mikels, J., Conway, A. et al. (2005).
A warm heart and a clear head: The contingent effects of weather on mood and cognition.
Psychological Science, 16, 724-731.
Kim, J. & Wineman, K. (2005). Are Windows and View Really Better? A Quantitative
Analysis of the Economic and Psychological Value of Views. New York: Lighting
Research Center, Rensselaer Polytechnic Institute.
Kim, J. T., Shin, J. Y., & Yun, G. Y. (2012). Prediction of discomfort glares from windows:
Influence of the subjective evaluation of window views. Indoor and Built Environment,
21, 92-97.
Kim, W., Ahn, H. T., & Kim, J. T. (2008). A first approach to discomfort glare in the
presence of non-uniform luminance. Building and Environment, 43, 1953-1960.
Kim, W. & Kim, J. T. (2011). The scope of the glare light source of the window with nonuniform luminance distribution. Indoor and Built Environment, 20, 54-64.
Kleindienst, S. A. & Andersen, M. (2009). The adaptation of daylight glare probability to
dynamic metrics in a computational setting. In Proceedings of Lux Europa 2009 - the 11th
European Lighting Conference. Istanbul: Aydinlatma Türk Milli Komitesi.
Kosonen, R. & Tan, F. (2004). Assessment of productivity loss in air-conditioned buildings
using PMV index. Energy and Buildings, 36, 987-993.
Lawrence Berkeley National Laboratory (LBL) (2010). Virtual Lighting Simulator. Retrieved
12-3-2012, from http://gaia.lbl.gov/vls/
Lindsey, J. L. (1997). Applied Illumination Engineering (2nd ed.). Lilburn: The Fairmont
Press, Inc.
Lingjærde, O., Føreland, O., & Dankertsen, J. (1998). Dawn simulation vs. lightbox treatment
in winter depression: a comparative study. Acta Psychiatrica Scandinavica, 98, 73-80.
Maamari, F., Fontoynont, M., & Adra, N. (2006). Application of the CIE test cases to assess
the accuracy of lighting computer programs. Energy and Buildings, 38, 869-877.
162
References
Maamari, F., Fontoynont, M., Tsangrassoulis, A., Marty, C., Kopylov, E., & Sytnik, G.
(2005). Reliable datasets for lighting programs validation--benchmark results. Solar
Energy, 79, 213-215.
Mangkuto, R. A., Aries, M. B. C., van Loenen, E. J., & Hensen, J. L. M. (2011). Properties
and performance indicators of Virtual Natural Lighting Solutions. In Proceedings of
CISBAT 2011 (pp. 379-384). Lausanne: LESO-PB, École Polytechnique Fédérale de
Lausanne.
Mangkuto, R. A., Aries, M. B. C., van Loenen, E. J., & Hensen, J. L. M. (2012). Lighting
performance of Virtual Natural Lighting Solutions with a simplified image in a reference
office space. In Proceedings of Experiencing Light 2012 (pp. 1-5). Eindhoven: Eindhoven
University of Technology.
Mangkuto, R. A., Aries, M. B. C., van Loenen, E. J., & Hensen, J. L. M. (2014). Simulation
of virtual natural lighting solutions with a simplified view. Lighting Research and
Technology, 46, 198-218.
Mardaljevic, J. (1995). Validation of a lighting simulation program under real sky conditions.
Lighting Research and Technology, 27, 181-188.
Mardaljevic, J. (1997). Validation of a lighting simulation program: A study using measured
sky brightness distributions. In Proceedings of Lux Europa 1997 - the 8th European
Lighting Conference (pp. 555-569). Amsterdam: Nederlandse Stichting voor
Verlichtingskunde.
Markus, T. A. (1967). The function of windows – A reappraisal. Building Science, 2, 97-121.
Moeck, M. & Selkowitz, S. E. (1996). A computer-based daylight systems design tool.
Automation in Construction, 5, 193-209.
Mottram, V., Middleton, B., Williams, P., & Arendt, J. (2011). The impact of bright artificial
white and 'blue-enriched' light on sleep and circadian phase during the polar winter.
Journal of Sleep Research, 20, 154-161.
Ochoa, C. E., Aries, M. B. C., & Hensen, J. L. M. (2012). State of the art in lighting
simulation for building science: a literature review. Journal of Building Performance
Simulation, 5, 209-233.
Ohno, Y. & Davis, W. (2010). Rationale of color quality scale. National Institute of
Standards and Technology (NIST).
Pandharipande, A. & Caicedo, D. (2011). Daylight integrated illumination control of LED
systems based on enhanced presence sensing. Energy and Buildings, 43, 944-950.
Peel, M. C., Finlayson, B. L., & McMahon, T. A. (2007). Updated world map of the KöppenGeiger climate classification. Hydrology and Earth System Sciences, 4, 439-473.
163
References
Pelzers, R. S., Yu, Q. L., & Mangkuto, R. A. (2014). Radiation modeling of a photo-reactor
using a backward ray-tracing method: an insight into indoor photocatalytic oxidation.
Environmental Science and Pollution Research, 1-12, doi: 10.1007/s11356-014-2552-1.
Philips (2013a). Master TL5 HO 54W/850 UNP - Product catalog. Retrieved 13-1-2014,
from http://www.ecat.lighting.philips.com/l/lamps/fluorescent-lamps/tl5/master-tl5-highoutput/927929085018_eu/
Philips (2013b). Long Life IR PAR38 - Product catalog. Retrieved 15-1-2014, from
http://www.ecat.lighting.philips.com/l/lamps/halogen-lamps/mv-halogen-withreflector/long-life-ir-par38/924757119802_na/
Radikovic, A. S., Leggett, J. J., Keyser, J., & Ulrich, R. S. (2005). Artificial window view of
nature. In Proceedings of CHI '05 (pp. 1993-1996). New York: ACM.
Rational Craft (2013). Winscape. Retrieved 18-11-2013, from
http://rationalcraft.com/Winscape.html
Rea, M. S. & Freyssinier-Nova, J. P. (2008). Color rendering: A tale of two metrics. Color
Research & Application, 33, 192-202.
Reinhart, C. F. & Herkel, S. (2000). The simulation of annual daylight illuminance
distributions – a state-of-the-art comparison of six RADIANCE-based methods. Energy
and Buildings, 32, 167-187.
Reinhart, C. F. & Walkenhorst, O. (2001). Validation of dynamic RADIANCE-based
daylight simulations for a test office with external blinds. Energy and Buildings, 33, 683697.
Reinhart, C. F. & Andersen, M. (2006). Development and validation of a Radiance model for
a translucent panel. Energy and Buildings, 38, 890-904.
Reinhart, C. F. & Weismann, D. A. (2012). The daylit area – Correlating architectural student
assessments with current and emerging daylight availability metrics. Building and
Environment, 50, 155-164.
Rodriguez, R. G. & Pattini, A. (2014). Tolerance of discomfort glare from a large area source
for work on a visual display. Lighting Research and Technology, 46, 157-170.
Ruppertsberg, A. & Bloj, M. (2006). Rendering complex scene for psychophysics using
RADIANCE: How accurate can you get? Journal of the Optical Society of America A, 23,
759-768.
Saxon, R. (1994). The atrium comes of age. Harlow: Longman.
Shin, J. Y., Yun, G. Y., & Kim, J. T. (2012). View types and luminance effects on discomfort
glare assessment from windows. Energy and Buildings, 46, 139-145.
Skåret, J. E. (1992). Indoor environment and economics (Project No. N 6405), Oslo: NBIByggforsk.
164
References
Slater, A. I. & Graves, H. (2002). Benchmarking Lighting Design Software (Rep. No. TM
28/00). London: CIBSE.
Slater, A. I., Perry, M. J., & Carter, D. J. (1993). Illuminance differences between desks:
Limits of acceptability. Lighting Research and Technology, 25, 91-103.
Smolders, K. C. H. J. (2013). Daytime light exposure: Effects and preferences. PhD thesis,
Eindhoven University of Technology, Eindhoven.
Smolders, K. C. H. J., de Kort, Y. A. W., & Cluitmans, P. J. M. (2012). A higher illuminance
induces alertness even during office hours: Findings on subjective measures, task
performance and heart rate measures. Physiology and Behavior, 107, 7-16.
Smolders, K. C. H. J., de Kort, Y. A. W., & van den Berg, S. (2013). Diurnal light exposure
and feelings of vitality: Results of a field study during regular weekdays. Journal of
Stevens, R. G. (2009). Working against our endogenous circadian clock: Breast cancer and
electric lighting in the modern world. Journal of Mutation Research/Genetic Toxicology
and Environmental Mutagenesis, 680, 106-108.
Stoelinga, P. (2005). Current practice of simulation in the building design process: a matter
of communication. Delft: IBPSA NVL.
Suk, J. Y., Schiler, M., & Kensek, K. (2013). Development of new daylight glare analysis
methodology using absolute glare factor and relative glare factor. Energy and Buildings,
64, 113-122.
Tennessen, C. M. & Cimprich, B. (1995). Views to nature: Effects on attention. Journal of
TESS (2012). Therapeutic Environmental Solutions - Creating Healing Environments.
Retrieved 18-11-2013, from http://www.tesspatientexperience.com/
Tregenza, P. R. (1989). Modification of the split-flux formulae for mean daylight factor and
internal reflected component with large external obstructions. Lighting Research and
Tregenza, P. R. & Loe, D. (1998). The Design of Lighting. London: E & FN Spon.
Tuaycharoen, N. & Tregenza, P. R. (2005). Discomfort glare from interesting images.
Tuaycharoen, N. & Tregenza, P. R. (2007). View and discomfort glare from windows.
Ulrich, R. S. (1984). View through a window may influence recovery from surgery. Science,
224, 420-421.
Ulrich, R. S., Lundén, O., & Eltinge, J. L. (1993). Effects of exposure to nature and abstract
pictures on patients recovering from heart surgery. Psychophysiology, 30, 7.
165
References
Ulrich, R. S., Simons, R. F., Losito, B. D., Fiorito, E., Miles, M. A., & Zelson, M. (1991).
Stress recovery during exposure to natural and urban environments. Journal of
US Department of Energy (2011). Weather Data - All Regions: Southwest Pacific WMO
Region 5. Retrieved 7-3-2013, from
http://apps1.eere.energy.gov/buildings/energyplus/cfm/weather_data2.cfm/region=5_sout
hwest_pacific_wmo_region_5
van Dijk, D. & Platzer, W. J. (2003). Reference office for thermal, solar and lighting
calculations (Rep. No. swift-wp3-tno-dvd-030416).
van Hoof, J., Westerlaken, A. C., Aarts, M. P. J., Wouters, E. J. M., Schoutens, A. M. C.,
Sinoo, M. M. et al. (2012). Light therapy: Methodological issues from an engineering
perspective. Technology and Health Care, 20, 11-23.
van Loenen, E. J., Oosting, W., Vogels, I., Lashina, T., van Hasen, R., Staats, C. et al. (2007).
Virtual Windows. In E. Aarts & E. Diederiks (Eds.), Ambient Lifestyle: From Concept to
Experience (pp. 148-161). Amsterdam: BIS Publishers.
van Meel, J. (2000). The European Office: Office Design and National Context. Rotterdam:
010 Publishers.
Vandewalle, G., Maquet, P., & Dijk, D. J. (2009). Light as a modulator of cognitive brain
function. Trends in Cognitive Sciences, 13, 429-438.
Viola, A. U., James, L. M., Schlangen, L. J. M., & Dijk, D. J. (2008). Blue-enriched white
light in the workplace improves self-reported alertness, performance and sleep quality.
Scandinavian Journal of Work, Environment & Health, 34, 297-306.
Ward, G. (1994). The RADIANCE lighting simulation and rendering system. In SIGGRAPH
(pp. 459-472). Orlando: ACM SIGGRAPH.
Ward, G. & Shakespeare, R. A. (1998). Rendering with Radiance: The Art and Science of
Lighting Visualization. San Francisco: Morgan Kaufman Publishers.
Webb, A. R. (2006). Considerations for lighting in the built environment: Non-visual effects
of light. Energy and Buildings, 38, 721-727.
Wienold, J. & Christoffersen, J. (2006). Evaluation methods and development of a new glare
prediction model for daylight environments with the use of CCD cameras. Energy and
Buildings, 38, 743-757.
Windauga (2013). Windauga® - Products and services. Retrieved 18-11-2013, from
http://www.windauga.com/english/products.htm
Woods, J. E. (1989). Cost avoidance and productivity in owning and operating buildings. Occupational Medicine, 4, 753-770.
166
References
Yun, G. Y., Hwang, T., & Kim, J. T. (2010). Performance prediction by modelling of a lightpipe system used under the climate conditions of Korea. Indoor and Built Environment,
19, 137-144.
Yun, G. Y., Shin, J. Y., & Kim, J. T. (2011). Influence of window views on the subjective
evaluation of discomfort glare. Indoor and Built Environment, 20, 65-74.
Zukauskas, A., Vaicekauskas, R., Ivanauskas, F., Vaitkevicius, H., Vitta, P., & Shur, M. S.
(2009). Statistical approach to color quality of solid-state lamps. IEEE Journal of Selected
Topics in Quantum Electronics, 15, 1189-1198.
167
References
168
Appendix A
This appendix summarises the complete measurement and simulation results of horizontal
illuminance values on the workplane in the kitchen test room as discussed in Chapter 3.
Figure 3.3 is referred for the orientation; the lower left corner on the floor plan is defined as
the origin. Position x and y are both measured in metres, the workplane height is 0.92 m.
Table A.1 shows the results under the overcast clear, and partly cloudy sky scenes of the
VNLS prototype as well as the simulated real window.
Table A.1. Measurement and simulation results of horizontal illuminance values on the workplane
in the kitchen test room with VNLS prototype and real window (RW)
Overcast
x
[m]
y
[m]
0.2
0.7
1.2
1.7
2.2
2.7
3.2
3.7
0.2
0.7
1.2
1.7
2.2
2.7
3.2
3.7
0.2
0.7
1.2
1.7
2.2
2.7
3.2
3.7
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
VNLS VNLS
Emea
Esim
[lx]
[lx]
34
22
34
24
35
25
34
25
32
24
31
23
28
19
24
12
30
22
35
22
38
26
36
28
37
25
33
21
29
18
25
15
28
26
38
33
44
38
44
38
48
31
37
26
31
21
25
16
Clear
RW
Esim
[lx]
10
10
8
10
10
8
9
8
10
11
12
12
12
12
11
13
11
13
15
15
15
13
12
11
VNLS
Emea
[lx]
43
44
45
44
41
36
31
26
38
42
45
47
47
41
32
27
40
47
51
56
55
48
33
28
169
VNLS
Esim
[lx]
45
37
47
43
47
44
37
31
40
54
50
48
52
49
40
30
49
59
68
65
59
52
42
28
Partly cloudy
RW
Esim
[lx]
57
64
68
66
66
61
58
50
61
67
76
71
72
63
60
57
66
80
88
87
82
74
62
58
VNLS VNLS
Emea
Esim
[lx]
[lx]
60
54
65
59
66
60
65
61
63
60
55
54
50
48
43
30
52
50
62
61
72
62
65
65
68
67
57
63
51
46
41
37
51
62
68
75
85
80
80
81
79
80
65
66
54
53
41
38
RW
Esim
[lx]
33
35
35
36
32
38
33
37
36
39
43
43
39
38
37
41
42
50
51
55
49
45
39
37
Appendices
Table A.1. (continued)
Overcast
x
[m]
y
[m]
0.2
0.7
1.2
1.7
2.2
2.7
3.2
3.7
0.2
0.7
1.2
1.7
2.2
2.7
3.2
3.7
0.2
0.7
1.2
1.7
2.2
2.7
3.2
3.7
0.7
1.2
1.7
2.2
2.7
3.2
3.7
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.5
3.5
3.5
3.5
3.5
3.5
3.5
VNLS VNLS
Emea
Esim
[lx]
[lx]
39
31
47
43
57
51
61
52
76
46
45
33
32
18
23
15
42
36
61
50
103
76
88
79
76
68
52
40
32
23
20
13
23
19
63
61
135
119
155
155
142
104
62
43
30
20
17
9
48
43
115
127
204
234
118
103
40
26
21
12
15
8
Clear
RW
Esim
[lx]
21
27
41
45
36
24
18
11
13
40
81
105
64
31
16
14
39
125
230
93
23
11
10
21
27
41
45
36
24
18
11
VNLS
Emea
[lx]
54
59
74
102
67
57
27
26
61
85
109
129
129
73
38
23
39
94
156
236
176
76
39
23
86
215
354
151
52
36
20
170
VNLS
Esim
[lx]
60
81
102
104
82
66
40
29
72
99
136
139
125
76
42
18
37
94
209
253
181
76
37
15
87
260
432
200
42
23
16
Partly cloudy
RW
Esim
[lx]
106
121
132
137
106
84
69
52
74
136
207
193
137
88
62
52
153
367
376
165
89
60
46
106
121
132
137
106
84
69
52
VNLS VNLS
Emea
Esim
[lx]
[lx]
70
80
86
104
130
109
149
132
107
94
87
71
56
49
40
34
81
85
110
119
174
166
190
187
191
146
116
91
58
52
35
22
53
46
119
125
241
256
355
346
256
220
109
90
53
44
31
21
120
115
287
316
491
533
237
242
72
69
39
29
27
19
RW
Esim
[lx]
72
86
108
116
86
61
47
44
51
126
237
227
139
74
47
40
153
407
500
165
69
44
36
72
86
108
116
86
61
47
44
Appendix B
This appendix gives the high dynamic range (HDR) and luminance false colour pictures of
the VNLS prototype in the kitchen test room under the three sky scenes in Chapter 3, as seen
from position 2, shown in Figures B.1 and B.2. Figure 3.3 is referred for the orientation. Note
there are different scales used in the three pictures in Figure B.2.
(a)
(b)
(c)
Figure B.1. High dynamic range pictures of the prototype under the (a) overcast, (b) clear, and
(c) partly cloudy sky scenes; as seen from position 2
171
Appendices
(a)
(b)
(c)
Figure B.2. Luminance false colour pictures of the prototype observed at position 2, under (a)
overcast, (b) clear, and (c) partly cloudy sky scene
172
Appendix C
This appendix summarises the mean subjective glare ratings, converted and normalised
simulation values of glare metrics from the first generation VNLS prototype with complex
views (Shin et al., 2012) as discussed in Chapter 4. Table C.1 shows the results under the five
mean view luminance and 10 image scenes.
Table C.1. Mean subjective glare ratings, converted and normalised simulated values of glare
metrics from the VNLS prototype in the experiment of Shin et al. (2012)
Image
DMM
NMM
DNL
NNL
DML
NML
Lav
[cd/m2]
1000
1800
3200
5600
10000
1000
1800
3200
5600
10000
1000
1800
3200
5600
10000
1000
1800
3200
5600
10000
1000
1800
3200
5600
10000
1000
1800
3200
5600
10000
Shin’s DGP DGP DGIn DGIn CGIn
rating conv sim conv sim conv
[-]
[-]
[-]
[-]
[-]
[-]
1.1
1.3
1.6
2.5
3.4
1.3
1.6
2.1
2.8
3.5
1.2
1.6
2.3
2.9
3.5
1.3
1.8
2.3
3.1
3.6
1.2
1.4
1.7
2.7
3.5
1.2
1.5
2.4
3.0
3.7
0.24
0.26
0.29
0.33
0.36
0.26
0.29
0.32
0.34
0.36
0.25
0.29
0.32
0.34
0.36
0.26
0.30
0.32
0.35
0.37
0.25
0.27
0.29
0.33
0.36
0.25
0.28
0.33
0.34
0.37
0.26
0.28
0.31
0.36
0.46
0.26
0.28
0.31
0.37
0.46
0.26
0.28
0.31
0.36
0.46
0.26
0.28
0.31
0.36
0.46
0.25
0.27
0.31
0.36
0.46
0.26
0.28
0.31
0.36
0.45
0.20
0.22
0.25
0.31
0.35
0.22
0.25
0.29
0.32
0.36
0.21
0.25
0.30
0.33
0.36
0.22
0.27
0.30
0.33
0.36
0.21
0.23
0.26
0.32
0.36
0.21
0.24
0.30
0.33
0.37
173
0.34
0.36
0.38
0.40
0.43
0.35
0.37
0.39
0.41
0.43
0.35
0.37
0.39
0.41
0.44
0.35
0.37
0.39
0.41
0.43
0.33
0.36
0.38
0.40
0.42
0.33
0.36
0.38
0.40
0.42
0.16
0.18
0.21
0.28
0.35
0.18
0.21
0.25
0.30
0.35
0.17
0.21
0.27
0.31
0.35
0.18
0.23
0.27
0.32
0.36
0.17
0.19
0.22
0.30
0.35
0.17
0.20
0.27
0.32
0.37
CGIn
sim
[-]
UGR
conv
[-]
UGRn
sim
[-]
0.45
0.50
0.55
0.60
0.66
0.46
0.51
0.56
0.61
0.67
0.46
0.50
0.56
0.61
0.67
0.46
0.50
0.55
0.61
0.67
0.45
0.49
0.54
0.60
0.66
0.45
0.50
0.55
0.60
0.66
0.16
0.18
0.21
0.28
0.35
0.18
0.21
0.25
0.30
0.35
0.17
0.21
0.27
0.31
0.35
0.18
0.23
0.27
0.32
0.36
0.17
0.19
0.22
0.30
0.35
0.17
0.20
0.27
0.32
0.37
0.48
0.51
0.54
0.57
0.61
0.50
0.53
0.56
0.59
0.63
0.51
0.54
0.57
0.60
0.63
0.49
0.53
0.56
0.59
0.62
0.47
0.51
0.54
0.57
0.60
0.48
0.51
0.55
0.58
0.61
Appendices
Table C.1. (continued)
Image
DNR
NNR
DMR
NMR
Lav
[cd/m2]
1000
1800
3200
5600
10000
1000
1800
3200
5600
10000
1000
1800
3200
5600
10000
1000
1800
3200
5600
10000
Shin’s DGP DGP DGIn DGIn CGIn CGIn
rating conv sim conv sim conv sim
[-]
[-]
[-]
[-]
[-]
[-]
[-]
1.1
0.24 0.25
0.20 0.33 0.16
0.44
1.5
0.28 0.27
0.24 0.36 0.20
0.49
2.0
0.31 0.31
0.28 0.38 0.24
0.54
2.8
0.34 0.36
0.32 0.40 0.30
0.60
3.8
0.38 0.45
0.38 0.42 0.38
0.66
1.2
0.25 0.26
0.21 0.34 0.17
0.46
2.0
0.31 0.28
0.28 0.36 0.24
0.50
2.5
0.33 0.31
0.31 0.38 0.28
0.55
3.3
0.35 0.37
0.34 0.40 0.34
0.61
3.9
0.39 0.46
0.38 0.43 0.39
0.67
1.2
0.25 0.25
0.21 0.34 0.17
0.44
2.0
0.31 0.27
0.28 0.36 0.24
0.49
2.2
0.32 0.30
0.29 0.38 0.26
0.54
3.0
0.34 0.35
0.33 0.40 0.32
0.59
3.7
0.37 0.45
0.37 0.42 0.37
0.65
1.2
0.25 0.26
0.21 0.35 0.17
0.45
1.4
0.27 0.28
0.23 0.37 0.19
0.50
1.8
0.30 0.31
0.27 0.40 0.23
0.55
2.7
0.33 0.36
0.32 0.42 0.30
0.61
3.6
0.37 0.45
0.36 0.44 0.36
0.66
174
UGR
conv
[-]
UGRn
sim
[-]
0.16
0.20
0.24
0.30
0.38
0.17
0.24
0.28
0.34
0.39
0.17
0.24
0.26
0.32
0.37
0.17
0.19
0.23
0.30
0.36
0.47
0.50
0.53
0.56
0.60
0.48
0.51
0.54
0.57
0.61
0.48
0.51
0.54
0.57
0.61
0.52
0.55
0.59
0.62
0.65
Appendix D
This appendix summarises the complete measurement and simulation results of horizontal
illuminance values on the workplane in the test room as discussed in Chapter 5. Figure 5.10 is
referred for the orientation; the upper right corner on the floor plan is defined as the origin.
Position x and y are both measured in metres, the workplane height is 0.75 m. Table D.1
shows the results under the 25%, 62.5%, and 100% of the maximum setting of the VNLS
prototype.
Table D.1. Measurement and simulation results of horizontal illuminance values on the workplane
in the test room with the prototype under 25%, 62.5% and 100% maximum setting
x
[m]
y
[m]
2.3
2.8
3.3
3.8
4.3
4.8
5.2
5.7
6.2
6.7
7.1
7.6
8.1
2.3
2.8
3.3
3.8
4.3
4.8
5.2
5.7
6.2
6.7
7.1
7.6
8.1
0.375
0.375
0.375
0.375
0.375
0.375
0.375
0.375
0.375
0.375
0.375
0.375
0.375
0.875
0.875
0.875
0.875
0.875
0.875
0.875
0.875
0.875
0.875
0.875
0.875
0.875
25%
Emea
[lx]
66
80
69
56
45
36
33
27
24
21
19
19
18
127
106
82
63
48
39
34
29
25
22
19
19
18
62.5%
Emea
Esim
[lx]
[lx]
193
172
223
212
197
186
161
149
129
118
104
91
86
77
72
61
61
53
58
40
47
36
47
34
47
32
351
331
302
295
230
227
182
174
142
129
111
105
90
84
76
66
64
53
56
44
50
40
48
36
48
34
Esim
[lx]
70
82
72
59
46
37
32
25
20
17
15
14
13
133
117
90
69
51
40
33
27
22
16
16
14
14
175
100%
Emea
[lx]
343
382
343
284
227
187
154
127
107
95
85
76
73
632
524
404
317
241
196
160
132
111
98
87
81
78
Esim
[lx]
336
379
347
280
216
169
140
120
91
80
73
64
63
617
548
432
319
241
190
160
123
105
82
67
69
67
Appendices
Table D.1. (continued)
x
[m]
y
[m]
2.3
2.8
3.3
3.8
4.3
4.8
5.2
5.7
6.2
6.7
7.1
7.6
8.1
2.3
2.8
3.3
3.8
4.3
4.8
5.2
5.7
6.2
6.7
7.1
7.6
8.1
2.3
2.8
3.3
3.8
4.3
4.8
5.2
5.7
6.2
6.7
7.1
7.6
8.1
1.375
1.375
1.375
1.375
1.375
1.375
1.375
1.375
1.375
1.375
1.375
1.375
1.375
1.875
1.875
1.875
1.875
1.875
1.875
1.875
1.875
1.875
1.875
1.875
1.875
1.875
2.375
2.375
2.375
2.375
2.375
2.375
2.375
2.375
2.375
2.375
2.375
2.375
2.375
25%
Emea
[lx]
171
130
93
69
51
40
35
29
25
22
20
20
19
165
137
97
71
52
41
36
29
25
22
20
20
19
175
132
92
68
51
40
35
29
25
22
20
20
19
62.5%
Emea
Esim
[lx]
[lx]
470
444
362
354
261
258
196
188
152
139
116
108
93
84
78
70
65
54
57
43
50
40
49
37
48
35
459
441
382
364
272
268
203
197
155
145
119
112
94
91
79
69
67
57
57
46
51
41
49
35
47
35
484
431
366
349
258
257
196
179
150
135
118
103
93
89
79
67
66
53
57
47
51
41
47
36
47
35
Esim
[lx]
180
145
104
77
57
43
35
27
22
19
16
14
14
174
146
108
80
58
45
36
28
22
18
15
15
14
170
140
102
75
58
44
34
27
23
17
16
14
14
176
100%
Emea
[lx]
866
637
460
340
249
203
163
135
111
100
87
81
78
858
677
480
350
254
204
163
135
112
98
88
82
78
903
650
458
337
248
203
159
134
110
98
86
82
79
Esim
[lx]
834
677
496
366
276
203
168
127
104
84
78
71
68
842
701
516
373
267
205
170
136
109
86
76
70
67
820
649
478
349
261
200
172
132
90
87
82
70
68
Appendices
Table D.1. (continued)
x
[m]
y
[m]
2.3
2.8
3.3
3.8
4.3
4.8
5.2
5.7
6.2
6.7
7.1
7.6
8.1
2.3
2.8
3.3
3.8
4.3
4.8
5.2
5.7
6.2
6.7
7.1
7.6
8.1
2.875
2.875
2.875
2.875
2.875
2.875
2.875
2.875
2.875
2.875
2.875
2.875
2.875
3.375
3.375
3.375
3.375
3.375
3.375
3.375
3.375
3.375
3.375
3.375
3.375
3.375
25%
Emea
[lx]
132
105
81
62
48
39
35
29
25
22
20
20
19
66
77
68
56
46
37
34
28
24
21
19
19
18
62.5%
Emea
Esim
[lx]
[lx]
364
334
304
284
229
223
181
173
143
134
113
102
91
86
77
65
66
55
57
45
50
40
49
34
46
36
191
182
214
204
194
179
162
148
133
116
107
95
88
79
75
62
63
50
55
42
49
37
45
33
47
33
Esim
[lx]
131
116
88
69
54
41
34
26
22
17
17
14
14
73
82
74
60
47
37
32
25
20
16
15
14
13
177
100%
Emea
[lx]
670
518
398
304
234
291
153
129
109
95
85
81
77
337
378
329
273
219
178
146
124
105
91
83
76
73
Esim
[lx]
628
552
411
330
247
191
153
125
102
86
76
64
67
344
384
347
282
223
179
145
119
96
79
71
66
65
Appendix E
This appendix gives the false colour maps of horizontal illuminance values on the workplane,
under Configurations 1, 4, and 6 of the prototype in the test room as discussed in Chapter 5,
shown in Figure E.1.
(a)
(b)
(c)
Figure E.1. False colour maps of the simulated horizontal illuminance [lx] under Configurations
(a) 1, (b) 4, and (c) 6
178
Appendix F
This appendix summarises the monthly average space availability for 300 lx criterion and
total electrical energy consumption for each climate type, annual mode, and scenario of the
display of the second generation prototype in Configuration 2; as described in Chapter 5. The
monthly values of average space availability are averaged for the entire year, whereas those
of total electrical energy consumption are summed up. Table F.1 gives the complete values.
The light gray-, white-, and dark gray-coloured cells respectively correspond to the ‘spring’,
‘summer’, and ‘winter’ months.
Table F.1. Estimated average monthly space availability (%A) for 300 lx criterion and total
monthly electrical energy consumption (Wmonth) in Configuration 2 for each climate type, annual
mode, and scenario
Location:
Singapore
Month Days
Mimicking
%A [%]
Wmonth [kWh]
Sc1 Sc2 Sc3 Sc1
Sc2
Sc3
Compensating
%A [%]
Wmonth [kWh] Sc1 Sc2 Sc3 Sc1 Sc2
Sc3
Jan
25
61
61
61
241
241
241
61
61
61
241
241
241
Feb
20
51
51
51
182
182
182
49
49
49
179
179
179
Mar
20
51
51
51
182
182
182
49
49
49
179
179
179
Apr
20
51
51
51
182
182
182
49
49
49
179
179
179
May
25
61
51
51
241
228
228
61
49
49
241
223
223
Jun
20
61
61
61
192
192
192
61
61
61
192
192
192
Jul
25
61
61
51
241
241
228
61
61
49
241
241
223
Aug
20
61
61
61
192
192
192
61
61
61
192
192
192
Sep
20
61
61
61
192
192
192
61
61
61
192
192
192
Oct
25
61
61
61
235
235
235
61
61
61
235
235
235
Nov
20
49
49
49
179
179
179
51
51
51
182
182
182
Dec
20
49
49
49
121
121
121
51
51
51
182
182
182
260
57
56
55
2380
2367
2354
57
56
55
2434 2417 2400
179
Appendices
Table F.1. (continued)
Location:
Cairo
Month Days
Mimicking
%A [%]
Wmonth [kWh]
Sc1 Sc2 Sc3 Sc1
Sc2
Sc3
Jan
25
52
52
52
212
212
212
3
3
3
131
131
131
Feb
20
40
52
40
164
169
164
40
3
40
164
105
164
Mar
20
40
40
40
164
164
164
40
40
40
164
164
164
Apr
20
40
40
3
164
164
105
40
40
52
164
164
169
May
25
3
3
3
131
131
131
52
52
52
212
212
212
Jun
20
3
3
3
105
105
105
52
52
52
169
169
169
Jul
25
3
3
3
131
131
131
52
52
52
212
212
212
Aug
20
40
3
3
164
105
105
40
52
52
164
169
169
Sep
20
40
40
3
164
164
105
40
40
52
164
164
169
Oct
25
40
40
40
205
205
205
40
40
40
205
205
205
Nov
20
52
52
52
169
169
169
3
3
3
105
105
105
Dec
20
52
52
52
169
169
169
3
3
3
105
105
105
260
33
31
25
1945
1891
1767
34
32
37
1961 1907 1977
Location:
Amsterdam
Month Days
Mimicking
%A [%]
Wmonth [kWh]
Sc1 Sc2 Sc3 Sc2
Sc3
Sc1
Compensating
%A [%]
Sc1
Jan
25
1
1
1
103
103
103
52
52
52
232
232
232
Feb
20
1
53
53
82
182
182
52
53
53
185
182
182
Mar
20
53
53
53
182
182
182
53
53
53
182
182
182
Apr
20
53
53
53
182
182
182
53
53
53
182
182
182
May
25
52
52
52
232
232
232
1
1
1
103
103
103
Jun
20
52
52
52
185
185
185
1
1
1
82
82
82
Jul
25
52
52
52
232
232
232
1
1
1
103
103
103
Aug
20
53
53
52
182
182
185
53
53
1
182
182
82
Sep
20
53
53
53
182
182
182
53
53
53
182
182
182
Oct
25
53
53
53
227
227
227
53
53
53
227
227
227
Nov
20
1
1
1
82
82
82
52
52
52
185
185
185
Dec
20
1
1
1
82
82
82
52
52
52
185
185
185
260
36
40
40
1952
2051
2055
39
39
35
2029 2026 1926
180
Compensating
%A [%]
Sc3
Appendices
Table F.1. (continued)
Location:
Sevilla
Month Days
Mimicking
%A [%]
Wmonth [kWh]
Sc1 Sc2 Sc3 Sc1
Sc2
Sc3
Compensating
%A [%]
Sc3
Jan
25
47
47
47
210
210
210
15
15
15
162
162
162
Feb
20
63
47
63
195
168
195
63
15
63
195
130
195
Mar
20
63
63
63
195
195
195
63
63
63
195
195
195
Apr
20
63
63
63
195
195
195
63
63
63
195
195
195
May
25
15
15
15
162
162
162
47
47
47
210
210
210
Jun
20
15
15
15
130
130
130
47
47
47
168
168
168
Jul
25
15
15
15
162
162
162
47
47
47
210
210
210
Aug
20
63
15
15
195
130
130
63
47
47
195
168
168
Sep
20
63
63
63
195
195
195
63
63
63
195
195
195
Oct
25
63
63
63
243
243
243
63
63
63
243
243
243
Nov
20
47
47
47
168
168
168
15
15
15
130
130
130
Dec
20
47
47
47
168
168
168
15
15
15
130
130
130
260
46
41
42
2218
2127
2153
47
42
45
2228 2136 2201
Location:
Chicago
Month Days
Mimicking
%A [%]
Wmonth [kWh]
Sc1 Sc2 Sc3 Sc1
Sc2
Sc3
Compensating
%A [%]
Wmonth [kWh] Sc1 Sc2 Sc3 Sc1 Sc2 Sc3
Jan
25
17
17
17
151
151
151
31
31
31
201
201
201
Feb
20
61
17
61
188
121
188
61
31
61
188
161
188
Mar
20
61
61
61
188
188
188
61
61
61
188
188
188
Apr
20
61
61
61
188
188
188
61
61
61
188
188
188
May
25
31
31
31
201
201
201
17
17
17
151
151
151
Jun
20
31
31
31
161
161
161
17
17
17
121
121
121
Jul
25
31
31
31
201
201
201
17
17
17
151
151
151
Aug
20
61
31
31
188
161
161
61
17
17
188
121
121
Sep
20
61
61
61
188
188
188
61
61
61
188
188
188
Oct
25
61
61
61
235
235
235
61
61
61
235
235
235
Nov
20
17
17
17
121
121
121
31
31
31
161
161
161
Dec
20
17
17
17
121
121
121
31
31
31
161
161
161
260
42
36
39
2131 2037
2104
42
36
38
2121 2027 2054
181
Appendix G
This appendix summarises the space availability (%A), uniformity (U0), ground contribution
configurations with total luminous flux (Φ) of 6200 and 19900 lm; as described in Chapter 6.
Note that configurations with the same distance between windows, beam angle, and total
luminous flux are compared to the same reference real window, of which the performance
indicators are shown directly above them in Table G.1.
Table G.1. Summary of space availability, uniformity, average ground contribution, and
probability of discomfort glare for all variations and position C in both VNLS and real windows
(RW) scenes with total luminous flux of 6200 and 19900 lm
Conf.
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
2a
2a
2a 2a 2a
2a 2a 2a 2a
2a 2a 2a IA [°]
BA [°]
Φ [lm]
RW – 5600 cd/m2
2.0
38
6200
1.5
38
6200
1.0
38
6200
RW – 1800 cd/m2
2.0
76
6200
1.5
76
6200
1.0
76
6200
RW – 1000 cd/m2
2.0
114
6200
1.5
114
6200
1.0
114
6200
RW – 5600 cd/m2
2.0
38
6200
1.5
38
6200
1.0
38
6200
RW – 1800 cd/m2
2.0
76
6200
1.5
76
6200
1.0
76
6200
RW – 1000 cd/m2
2.0
114
6200
1.5
114
6200
1.0
114
6200
182
%A
[%]
48
16
13
8
14
12
11
9
6
9
9
9
43
15
9
1
15
12
9
5
3
8
8
8
U0
[-]
0.19
0.21
0.23
0.26
0.18
0.28
0.29
0.32
0.18
0.37
0.37
0.37
0.16
0.23
0.24
0.28
0.16
0.32
0.33
0.35
0.15
0.36
0.37
0.37
%Gav
[%]
50
60
58
55
50
50
47
44
50
49
47
44
49
59
57
54
49
49
47
44
48
48
46
45
PDGav
[-]
0.39
0.40
0.42
0.43
0.34
0.36
0.36
0.36
0.31
0.32
0.32
0.32
0.39
0.41
0.43
0.44
0.35
0.36
0.36
0.36
0.32
0.32
0.32
0.32
Appendices
Table G.1. (continued)
Conf.
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
2a
2a
2a 2a 2a
2a 2a 2a 2a
2a 2a 2a IA [°]
BA [°]
Φ [lm]
RW – 18000 cd/m2
2.0
38
19900
1.5
38
19900
1.0
38
19900
RW – 5600 cd/m2
2.0
76
19900
1.5
76
19900
1.0
76
19900
RW – 3200 cd/m2
2.0
114
19900
1.5
114
19900
1.0
114
19900
RW – 18000 cd/m2
2.0
38
19900
1.5
38
19900
1.0
38
19900
RW – 5600 cd/m2
2.0
76
19900
1.5
76
19900
1.0
76
19900
RW – 3200 cd/m2
2.0
114
19900
1.5
114
19900
1.0
114
19900
183
%A
[%]
100
72
79
76
48
70
74
80
27
60
62
63
89
79
84
81
43
66
71
74
30
56
58
60
U0
[-]
0.18
0.21
0.23
0.26
0.19
0.28
0.30
0.32
0.16
0.37
0.37
0.38
0.15
0.23
0.24
0.28
0.16
0.32
0.33
0.35
0.15
0.36
0.36
0.37
%Gav
[%]
51
60
58
55
50
50
47
44
51
49
47
44
49
59
57
54
49
49
47
44
50
48
46
45
PDGav
[-]
0.46
0.46
0.49
0.50
0.39
0.41
0.42
0.42
0.36
0.37
0.38
0.37
0.46
0.48
0.50
0.51
0.39
0.42
0.42
0.42
0.37
0.38
0.38
0.38
Appendix H
This appendix summarises the space availability (%A), uniformity (U0), ground contribution
configurations with total luminous flux (Φ) of 6200, 11100, and 19900 lm, and with 10 image
scenes; as described in Chapter 7. Table H.1 gives the complete values.
Table H.1. Summary of space availability, uniformity, average ground contribution, and
probability of discomfort glare for all variations and position C in VNLS scenes with total luminous
flux of 6200, 11100, and 19900 lm, and with 10 image scenes Scene
BA [°]
Φ [lm]
%A
[%]
U0
[-]
%Gav
[%]
PDGav
[-]
DML
38
38
38
76
76
76
114
114
114
38
38
38
76
76
76
114
114
114
38
38
38
76
76
76
114
114
114
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6
15
27
0
9
23
0
4
18
2
10
21
0
3
14
0
0
8
4
11
22
0
6
18
0
3
14
0.14
0.14
0.14
0.23
0.23
0.23
0.34
0.34
0.34
0.12
0.12
0.12
0.22
0.22
0.22
0.32
0.32
0.32
0.15
0.15
0.15
0.25
0.25
0.25
0.35
0.35
0.34
41
40
40
30
30
30
26
26
26
30
29
30
21
21
21
17
17
17
44
44
44
33
33
33
30
30
30
0.35
0.37
0.40
0.31
0.33
0.36
0.27
0.30
0.32
0.33
0.35
0.38
0.30
0.32
0.35
0.26
0.28
0.31
0.35
0.38
0.41
0.30
0.33
0.35
0.27
0.29
0.31
NML
DMM
184
Appendices
Table H.1. (continued) Scene
BA [°]
Φ [lm]
%A
[%]
U0
[-]
%Gav
[%]
PDGav
[-]
NMM
38
38
38
76
76
76
114
114
114
38
38
38
76
76
76
114
114
114
38
38
38
76
76
76
114
114
114
38
38
38
76
76
76
114
114
114
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
1
9
20
0
5
16
0
2
12
3
12
25
1
8
22
0
6
18
6
17
31
2
11
27
0
7
23
4
14
27
0
7
23
0
2
18
0.19
0.19
0.19
0.28
0.28
0.28
0.38
0.37
0.37
0.16
0.16
0.16
0.25
0.25
0.25
0.34
0.34
0.35
0.15
0.15
0.15
0.25
0.24
0.25
0.35
0.35
0.34
0.19
0.18
0.19
0.28
0.28
0.29
0.39
0.39
0.38
55
55
55
44
44
44
42
42
42
43
43
43
31
31
31
29
29
28
44
44
44
33
33
33
29
29
29
60
60
60
49
49
49
45
45
45
0.35
0.38
0.40
0.30
0.32
0.35
0.26
0.28
0.31
0.37
0.40
0.43
0.31
0.34
0.36
0.27
0.29
0.32
0.37
0.40
0.42
0.32
0.35
0.38
0.28
0.31
0.33
0.33
0.36
0.38
0.30
0.33
0.35
0.27
0.29
0.32
DMR
NMR
DNL
185
Appendices
Table H.1. (continued) Scene
BA [°]
Φ [lm]
%A
[%]
U0
[-]
%Gav
[%]
PDGav
[-]
NNL
38
38
38
76
76
76
114
114
114
38
38
38
76
76
76
114
114
114
38
38
38
76
76
76
114
114
114
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
6200
11100
19900
0
0
3
0
0
1
0
0
0
5
14
28
1
10
24
0
7
20
0
0
6
1
0
3
0
0
0
0.29
0.29
0.29
0.38
0.38
0.38
0.44
0.44
0.44
0.14
0.14
0.13
0.23
0.23
0.22
0.33
0.33
0.33
0.32
0.32
0.32
0.23
0.41
0.42
0.48
0.47
0.48
73
73
73
65
65
65
64
64
64
34
34
34
22
22
22
19
19
19
78
78
78
70
70
69
67
67
67
0.30
0.33
0.35
0.24
0.27
0.29
0.22
0.24
0.26
0.37
0.40
0.43
0.32
0.34
0.37
0.28
0.30
0.33
0.30
0.32
0.34
0.24
0.26
0.29
0.22
0.24
0.27
DNR
NNR
186
Curriculum Vitae
Rizki A. Mangkuto was born on 17 April 1984, in Bogor, West Java, Indonesia. In
August 2002, he started his bachelor study in the Department of Engineering Physics at
Institut Teknologi Bandung in Bandung, West Java, Indonesia. Since his third year in the
university, he had been involved in assisting courses and practicums. In November 2006, he
obtained the degree of Sarjana Teknik (equivalent to Bachelor of Science) with honour (cum
laude). His bachelor final project was entitled ‘Effect of Luminaire Spacing on the Lighting
Condition at Several Roadways in Bandung City’. Soon after that and until early 2010, he
was involved in various research, education, and consultancy activities in the Laboratory of
Building Physics and Acoustics, under Engineering Physics Research Group of the same
university, within the subject of building and/or environmental acoustics and lighting.
In August 2007, he was granted a’voucher’ (full) scholarship to pursue a master degree
within two years in the Graduate Programme of Engineering Physics at the same university,
with concentration in building physics. In July 2009, he obtained the degree of Magister
Teknik (equivalent to Master of Science) with honour. His master thesis was entitled ‘Study
of Daylight Effect on Building Occupants Based on Electroencephalograph Signal’.
In the first semester of 2010/2011 academic year, he was a teaching assistant in the
subject of Elementary Physics for first-year students of Diploma-3 (pre-bachelor) programme
of Metrology and Instrumentation, held within the Undergraduate Programme of Engineering
Physics. In April 2010, he was appointed as ‘academic assistant’ in Engineering Physics
Research Group at Institut Teknologi Bandung. In the end of April 2010, he moved to the
Netherlands to start his doctoral research in the Unit of Building Physics and Services at
Eindhoven University of Technology. His project was supported by the Sound Lighting
research line of the Intelligent Lighting Institute at Eindhoven University of Technology. The
doctoral research was on ‘Modelling and Simulation of Virtual Natural Lighting Solutions in
Buildings’, which resulted in this thesis.
187
List of Publications
Refereed academic journals
Pelzers, R. S., Yu, Q. L., & Mangkuto, R. A. (2014). Radiation modeling of a photo-reactor
using a backward ray-tracing method: An insight into indoor photocatalytic
oxidation. Environmental Science and Pollution Research, 1-14, doi: 10.1007/s11356014-2552-1.
van Dronkelaar, C., Cóstola, D., Mangkuto, R. A., & Hensen, J. L. M. (2014). Heating and
cooling energy demand in underground buildings: Potential for saving in various climates
and functions. Energy and Buildings, 71, 129-136.
Mangkuto, R. A., Aries, M. B. C., van Loenen, E. J., & Hensen, J. L. M. (2014). Simulation
of virtual natural lighting solutions with a simplified view. Lighting Research and
Conference proceedings
Pelzers, R. S., Yu, Q. L., Mangkuto, R. A., & Brouwers, H. J. H. (2013). Employing
RADIANCE to refine indoor photocatalytic oxidation modeling. In Proceedings of JEP
2013 – the 3rd European Symposium on Photocatalysis, 25-27 September 2013, Portoroz,
Slovenia (pp. P3-30). Portoroz: European Photocatalysis Federation (EPF) & University
of Nova Gorica.
Mangkuto, R. A., Claessen, R. N. H., Aries, M. B. C., van Loenen, E. J., & Hensen, J. L. M.
(2013). Space availability of buildings with virtual natural lighting solutions. In
Proceedings of Lux Europa 2013 – the 12th European Lighting Conference, 17-19
September 2013, Kraków, Poland (pp. 275-280). Kraków: Polski Komitet Oswietleniowy.
Mangkuto, R. A., Aries, M. B. C., van Loenen, E. J., & Hensen, J. L. M. (2013).
Development of virtual natural lighting solutions with a simplified view using lighting
simulation. In Proceedings of Building Simulation 2013 – the 13th International
Conference of the International Building Performance Simulation Association, 26-28
August 2013, Chambery, France, (pp. 3383-3390). Chambéry: IBPSA & Institute
Nationale de l'Energie Solair (INES).
Mangkuto, R. A., Aries, M. B. C., van Loenen, E. J., & Hensen, J. L. M. (2012). Lighting
performance of virtual natural lighting solutions with a simplified image in a reference
office space. In Proceedings of Experiencing Light 2012, 12-13 November 2012, (pp 1-4),
Eindhoven, the Netherlands. Eindhoven: Eindhoven University of Technology.
Mangkuto, R. A., Ochoa Morales, C. E., Aries, M. B. C., van Loenen, E. J., & Hensen,
J. L. M. (2011). Review of modelling approaches for developing virtual natural lighting
solutions. In Proceedings of Building Simulation 2011 – the 12th International
188
Conference of the International Building Performance Simulation Association, Sydney,
Australia, 14-16 November 2011, (pp. 2643-2650). Sydney: IBPSA Australasia &
AIRAH.
Mangkuto, R. A., Aries, M. B. C., van Loenen, E. J., & Hensen, J. L. M. (2011). Properties
and performance indicators of virtual natural lighting solutions. In Proceedings of CISBAT
2011, Lausanne, Switzerland, 14-16 September 2011, (pp. 379-384). Lausanne: Solar
Energy and Building Physics Laboratory (LESO-PB), École Polytechnique Fédérale de
Lausanne (EPFL).
Harahap, Y., Mangkuto, R. A., & Soelami, F. X. N. (2009). Lighting design for axis of
Gedung Sate and Monument of West Java People's Struggle. In Proceedings of ITB
International Conference on Regional Development, Environment and Infrastructures,
Bandung, Indonesia, 18-19 June 2009. Bandung: Institut Teknologi Bandung.
Mangkuto, R. A., Soelami, F. X. N., & Suprijanto, S. (2009). Study of effect of daylight on
building User's performance based on electroencephalograph signal. In Proceedings of the
10th SENVAR / 1st CONVEESH, Manado, Indonesia, 26-27 October 2009. Manado:
Universitas Sam Ratulangi.
Mangkuto, R. A., Paripurna, A., & Soelami, F. X. N. (2009). Evaluation on glare from
vehicle lamps and effectiveness of road components as glare barriers. In Proceedings of
the 3rd SEATUC Symposium, Johor Bahru, Malaysia, 25-26 February 2009. Johor Bahru:
Universiti Teknologi Malaysia.
Mangkuto, R. A., Soelami, F. X. N., & Soegijanto, R. M. (2008). Evaluation on lighting
condition and visual legibility of road surfaces and traffic signs in Bandung City. In
Proceeding of the 9th SENVAR / 2nd ISESEE, Shah Alam, Malaysia, 1-3 December 2008.
Shah Alam: Universiti Teknologi Mara.
Professional journals
van Dronkelaar, C., Cóstola, D., Mangkuto, R. A., & Hensen, J. L. M. (2013). Ondergronds
als alternatief voor bovengronds bouwen. TVVL Magazine, 42(10), 44-47.
Mangkuto, R. A., Ochoa Morales, C. E., Aries, M. B. C., van Loenen, E. J., & Hensen,
J. L. M. (2012). Simulaties voor R&D: ‘Virtual Natural Lighting’ systemen. TVVL
Magazine, 41(5), 38-40.
189
190
Bouwstenen is een publikatiereeks
van de Faculteit Bouwkunde,
Technische Universiteit Eindhoven.
Zij presenteert resultaten van
onderzoek en andere aktiviteiten op
het vakgebied der Bouwkunde,
uitgevoerd in het kader van deze
Faculteit.
Bouwstenen zijn telefonisch te
bestellen op nummer
040 - 2472383
Kernredaktie
MTOZ
Reeds verschenen in de serie
Bouwstenen
nr 1
Elan: A Computer Model for Building
Energy Design: Theory and Validation
Martin H. de Wit
H.H. Driessen
R.M.M. van der Velden
nr 9
Strukturering en Verwerking van
Tijdgegevens voor de Uitvoering
van Bouwwerken
ir. W.F. Schaefer
P.A. Erkelens
nr 2
Kwaliteit, Keuzevrijheid en Kosten:
Evaluatie van Experiment Klarendal,
Arnhem
J. Smeets
C. le Nobel
M. Broos
J. Frenken
A. v.d. Sanden
nr 10
Stedebouw en de Vorming van
een Speciale Wetenschap
K. Doevendans
nr 3
Crooswijk:
Van ‘Bijzonder’ naar ‘Gewoon’
Vincent Smit
Kees Noort
nr 12
Staal in de Woningbouw,
Korrosie-Bescherming van
de Begane Grondvloer
Edwin J.F. Delsing
nr 4
Staal in de Woningbouw
Edwin J.F. Delsing
nr 13
Een Thermisch Model voor de
Berekening van Staalplaatbetonvloeren
onder Brandomstandigheden
A.F. Hamerlinck
nr 5
Mathematical Theory of Stressed
Skin Action in Profiled Sheeting with
Various Edge Conditions
Andre W.A.M.J. van den Bogaard
nr 6
Hoe Berekenbaar en Betrouwbaar is
de Coëfficiënt k in x-ksigma en x-ks?
K.B. Lub
A.J. Bosch
nr 7
Het Typologisch Gereedschap:
Een Verkennende Studie Omtrent
Typologie en Omtrent de Aanpak
van Typologisch Onderzoek
J.H. Luiten
nr 8
Informatievoorziening en Beheerprocessen
A. Nauta
Jos Smeets (red.)
Helga Fassbinder (projectleider)
Adrie Proveniers
J. v.d. Moosdijk
nr 11
Informatica en Ondersteuning
van Ruimtelijke Besluitvorming
G.G. van der Meulen
nr 14
De Wijkgedachte in Nederland:
Gemeenschapsstreven in een
Stedebouwkundige Context
K. Doevendans
R. Stolzenburg
nr 15
Diaphragm Effect of Trapezoidally
Profiled Steel Sheets:
Experimental Research into the
Influence of Force Application
Andre W.A.M.J. van den Bogaard
nr 16
Versterken met Spuit-Ferrocement:
Het Mechanische Gedrag van met
Spuit-Ferrocement Versterkte
Gewapend Betonbalken
K.B. Lubir
M.C.G. van Wanroy
nr 17
De Tractaten van
Jean Nicolas Louis Durand
G. van Zeyl
nr 18
Wonen onder een Plat Dak:
Drie Opstellen over Enkele
Vooronderstellingen van de
Stedebouw
K. Doevendans
nr 19
Supporting Decision Making Processes:
A Graphical and Interactive Analysis of
Multivariate Data
W. Adams
nr 20
Self-Help Building Productivity:
A Method for Improving House Building
by Low-Income Groups Applied to Kenya
1990-2000
P. A. Erkelens
nr 21
De Verdeling van Woningen:
Een Kwestie van Onderhandelen
Vincent Smit
nr 22
Flexibiliteit en Kosten in het Ontwerpproces:
Een Besluitvormingondersteunend Model
M. Prins
nr 23
Spontane Nederzettingen Begeleid:
Voorwaarden en Criteria in Sri Lanka
Po Hin Thung
nr 24
Fundamentals of the Design of
Bamboo Structures
Oscar Arce-Villalobos
nr 25
Concepten van de Bouwkunde
M.F.Th. Bax (red.)
H.M.G.J. Trum (red.)
nr 26
Meaning of the Site
Xiaodong Li
nr 27
Het Woonmilieu op Begrip Gebracht:
Een Speurtocht naar de Betekenis van het
Begrip 'Woonmilieu'
Jaap Ketelaar
nr 28
Urban Environment in Developing Countries
editors: Peter A. Erkelens
George G. van der Meulen (red.)
nr 29
Stategische Plannen voor de Stad:
Onderzoek en Planning in Drie Steden
prof.dr. H. Fassbinder (red.)
H. Rikhof (red.)
nr 30
Stedebouwkunde en Stadsbestuur
Piet Beekman
nr 31
De Architectuur van Djenné:
Een Onderzoek naar de Historische Stad
P.C.M. Maas
nr 32
Conjoint Experiments and Retail Planning
Harmen Oppewal
nr 33
Strukturformen Indonesischer Bautechnik:
Entwicklung Methodischer Grundlagen
für eine ‘Konstruktive Pattern Language’
in Indonesien
Heinz Frick arch. SIA
nr 34
Styles of Architectural Designing:
Empirical Research on Working Styles
and Personality Dispositions
Anton P.M. van Bakel
nr 35
Conjoint Choice Models for Urban
Tourism Planning and Marketing
Benedict Dellaert
nr 36
Stedelijke Planvorming als Co-Produktie
Helga Fassbinder (red.)
nr 37
Design Research in the Netherlands
editors: R.M. Oxman
M.F.Th. Bax
H.H. Achten
nr 38
Communication in the Building Industry
Bauke de Vries
nr 39
Optimaal Dimensioneren van
Gelaste Plaatliggers
J.B.W. Stark
F. van Pelt
L.F.M. van Gorp
B.W.E.M. van Hove
nr 40
Huisvesting en Overwinning van Armoede
P.H. Thung
P. Beekman (red.)
nr 41
Urban Habitat:
The Environment of Tomorrow
George G. van der Meulen
Peter A. Erkelens
nr 42
A Typology of Joints
John C.M. Olie
nr 43
Modeling Constraints-Based Choices
for Leisure Mobility Planning
Marcus P. Stemerding
nr 44
Activity-Based Travel Demand Modeling
Dick Ettema
nr 45
Wind-Induced Pressure Fluctuations
on Building Facades
Chris Geurts
nr 46
Generic Representations
Henri Achten
nr 47
Johann Santini Aichel:
Architectuur en Ambiguiteit
Dirk De Meyer
nr 48
Concrete Behaviour in Multiaxial
Compression
Erik van Geel
nr 49
Modelling Site Selection
Frank Witlox
nr 50
Ecolemma Model
Ferdinand Beetstra
nr 51
Conjoint Approaches to Developing
Activity-Based Models
Donggen Wang
nr 52
On the Effectiveness of Ventilation
Ad Roos
nr 53
Conjoint Modeling Approaches for
Residential Group preferences
Eric Molin
nr 54
Modelling Architectural Design
Information by Features
Jos van Leeuwen
nr 55
A Spatial Decision Support System for
the Planning of Retail and Service Facilities
Theo Arentze
nr 56
Integrated Lighting System Assistant
Ellie de Groot
nr 57
Ontwerpend Leren, Leren Ontwerpen
J.T. Boekholt
nr 58
Temporal Aspects of Theme Park Choice
Behavior
Astrid Kemperman
nr 59
Ontwerp van een Geïndustrialiseerde
Funderingswijze
Faas Moonen
nr 60
Merlin: A Decision Support System
for Outdoor Leisure Planning
Manon van Middelkoop
nr 72
Moisture Transfer Properties of
Coated Gypsum
Emile Goossens
nr 61
The Aura of Modernity
Jos Bosman
nr 73
Plybamboo Wall-Panels for Housing
Guillermo E. González-Beltrán
nr 62
Urban Form and Activity-Travel Patterns
Daniëlle Snellen
nr 74
The Future Site-Proceedings
Ger Maas
Frans van Gassel
nr 63
Design Research in the Netherlands 2000
Henri Achten
nr 64
Computer Aided Dimensional Control in
Building Construction
Rui Wu
nr 65
Beyond Sustainable Building
editors: Peter A. Erkelens
Sander de Jonge
August A.M. van Vliet
co-editor: Ruth J.G. Verhagen
nr 66
Das Globalrecyclingfähige Haus
Hans Löfflad
nr 67
Cool Schools for Hot Suburbs
René J. Dierkx
nr 68
A Bamboo Building Design Decision
Support Tool
Fitri Mardjono
nr 69
Driving Rain on Building Envelopes
Fabien van Mook
nr 70
Heating Monumental Churches
Henk Schellen
nr 71
Van Woningverhuurder naar
Aanbieder van Woongenot
Patrick Dogge
nr 75
Radon transport in
Autoclaved Aerated Concrete
Michel van der Pal
nr 76
The Reliability and Validity of Interactive
Virtual Reality Computer Experiments
Amy Tan
nr 77
Measuring Housing Preferences Using
Virtual Reality and Belief Networks
Maciej A. Orzechowski
nr 78
Computational Representations of Words
and Associations in Architectural Design
Nicole Segers
nr 79
Measuring and Predicting Adaptation in
Multidimensional Activity-Travel Patterns
Chang-Hyeon Joh
nr 80
Strategic Briefing
Fayez Al Hassan
nr 81
Well Being in Hospitals
Simona Di Cicco
nr 82
Solares Bauen:
Implementierungs- und UmsetzungsAspekte in der Hochschulausbildung
in Österreich
Gerhard Schuster
nr 83
Supporting Strategic Design of
Workplace Environments with
Case-Based Reasoning
Shauna Mallory-Hill
nr 84
ACCEL: A Tool for Supporting Concept
Generation in the Early Design Phase
Maxim Ivashkov
nr 85
Brick-Mortar Interaction in Masonry
under Compression
Ad Vermeltfoort
nr 86
Zelfredzaam Wonen
Guus van Vliet
nr 87
Een Ensemble met Grootstedelijke Allure
Jos Bosman
Hans Schippers
nr 88
On the Computation of Well-Structured
Graphic Representations in Architectural
Design
Henri Achten
nr 89
De Evolutie van een West-Afrikaanse
Vernaculaire Architectuur
Wolf Schijns
nr 90
ROMBO Tactiek
Christoph Maria Ravesloot
nr 91
External Coupling between Building
Energy Simulation and Computational
Fluid Dynamics
Ery Djunaedy
nr 92
editors: Henri Achten
Kees Dorst
Pieter Jan Stappers
Bauke de Vries
nr 93
Ein Modell zur Baulichen Transformation
Jalil H. Saber Zaimian
nr 94
Human Lighting Demands:
Healthy Lighting in an Office Environment
Myriam Aries
nr 95
A Spatial Decision Support System for
the Provision and Monitoring of Urban
Greenspace
Claudia Pelizaro
nr 96
Leren Creëren
Adri Proveniers
nr 97
Simlandscape
Rob de Waard
nr 98
Design Team Communication
Ad den Otter
nr 99
Humaan-Ecologisch
Georiënteerde Woningbouw
Juri Czabanowski
nr 100
Hambase
Martin de Wit
nr 101
Sound Transmission through Pipe
Systems and into Building Structures
Susanne Bron-van der Jagt
nr 102
Het Bouwkundig Contrapunt
Jan Francis Boelen
nr 103
A Framework for a Multi-Agent
Planning Support System
Dick Saarloos
nr 104
Bracing Steel Frames with Calcium
Silicate Element Walls
Bright Mweene Ng’andu
nr 105
Naar een Nieuwe Houtskeletbouw
F.N.G. De Medts
nr 108
Geborgenheid
T.E.L. van Pinxteren
nr 109
Modelling Strategic Behaviour in
Anticipation of Congestion
Qi Han
nr 110
Reflecties op het Woondomein
Fred Sanders
nr 111
On Assessment of Wind Comfort
by Sand Erosion
Gábor Dezsö
nr 112
Bench Heating in Monumental Churches
Dionne Limpens-Neilen
nr 113
RE. Architecture
Ana Pereira Roders
nr 114
Toward Applicable Green Architecture
Usama El Fiky
nr 120
A Multi-Agent Planning Support
System for Assessing Externalities
of Urban Form Scenarios
Rachel Katoshevski-Cavari
nr 121
Den Schulbau Neu Denken,
Fühlen und Wollen
Urs Christian Maurer-Dietrich
nr 122
Peter Eisenman Theories and
Practices
Bernhard Kormoss
nr 123
User Simulation of Space Utilisation
Vincent Tabak
nr 125
In Search of a Complex System Model
Oswald Devisch
nr 126
Lighting at Work:
Environmental Study of Direct Effects
of Lighting Level and Spectrum on
Psycho-Physiological Variables
Grazyna Górnicka
nr 115
Knowledge Representation under
Inherent Uncertainty in a Multi-Agent
System for Land Use Planning
Liying Ma
nr 127
Flanking Sound Transmission through
Lightweight Framed Double Leaf Walls
Stefan Schoenwald
nr 116
Integrated Heat Air and Moisture
Modeling and Simulation
Jos van Schijndel
nr 128
˙
Bounded Rationality and Spatio-Temporal
Pedestrian Shopping Behavior
Wei Zhu
nr 117
Concrete Behaviour in Multiaxial
Compression
J.P.W. Bongers
nr 129
Travel Information:
Impact on Activity Travel Pattern
Zhongwei Sun
nr 118
The Image of the Urban Landscape
Ana Moya Pellitero
nr 130
Co-Simulation for Performance
Prediction of Innovative Integrated
Mechanical Energy Systems in Buildings
�
Marija Trcka
nr 119
The Self-Organizing City in Vietnam
Stephanie Geertman
nr 131
Allemaal Winnen
M.J. Bakker
nr 132
Architectural Cue Model in Evacuation
Simulation for Underground Space Design
Chengyu Sun
nr 143
Modelling Life Trajectories and Transport
Mode Choice Using Bayesian Belief Networks
Marloes Verhoeven
nr 133
Uncertainty and Sensitivity Analysis in
Building Performance Simulation for
Decision Support and Design Optimization
Christina Hopfe
nr 144
Assessing Construction Project
Performance in Ghana
William Gyadu-Asiedu
nr 134
Facilitating Distributed Collaboration
in the AEC/FM Sector Using Semantic
Web Technologies
Jacob Beetz
nr 135
Circumferentially Adhesive Bonded Glass
Panes for Bracing Steel Frame in Façades
Edwin Huveners
nr 136
Influence of Temperature on Concrete
Beams Strengthened in Flexure
with CFRP
Ernst-Lucas Klamer
nr 137
Sturen op Klantwaarde
Jos Smeets
nr 139
Lateral Behavior of Steel Frames
with Discretely Connected Precast Concrete
Infill Panels
Paul Teewen
nr 140
Integral Design Method in the Context
of Sustainable Building Design
Perica Savanovic´
nr 141
Household Activity-Travel Behavior:
Implementation of Within-Household
Interactions
Renni Anggraini
nr 142
Henri Achten
nr 145
Empowering Seniors through
Domotic Homes
Masi Mohammadi
nr 146
An Integral Design Concept for
Ecological Self-Compacting Concrete
Martin Hunger
nr 147
Governing Multi-Actor Decision Processes
in Dutch Industrial Area Redevelopment
Erik Blokhuis
nr 148
A Multifunctional Design Approach
for Sustainable Concrete
Götz Hüsken
nr 149
Quality Monitoring in Infrastructural
Design-Build Projects
Ruben Favié
nr 150
Assessment Matrix for Conservation of
Valuable Timber Structures
Michael Abels
nr 151
Co-simulation of Building Energy Simulation
and Computational Fluid Dynamics for
Whole-Building Heat, Air and Moisture
Engineering
Mohammad Mirsadeghi
nr 152
External Coupling of Building Energy
Simulation and Building Element Heat,
Air and Moisture Simulation
Daniel Cóstola
nr 153
Adaptive Decision Making In
Multi-Stakeholder Retail Planning
Ingrid Janssen
nr 154
Landscape Generator
Kymo Slager
nr 155
Constraint Specification in Architecture
Remco Niemeijer
nr 156
A Need-Based Approach to
Dynamic Activity Generation
Linda Nijland
nr 157
Modeling Office Firm Dynamics in an
Agent-Based Micro Simulation Framework
Gustavo Garcia Manzato
nr 158
Lightweight Floor System for
Vibration Comfort
Sander Zegers
nr 159
Aanpasbaarheid van de Draagstructuur
Roel Gijsbers
nr 160
'Village in the City' in Guangzhou, China
Yanliu Lin
nr 161
Climate Risk Assessment in Museums
Marco Martens
nr 162
Social Activity-Travel Patterns
Pauline van den Berg
nr 163
Sound Concentration Caused by
Curved Surfaces
Martijn Vercammen
nr 164
Design of Environmentally Friendly
Calcium Sulfate-Based Building Materials:
Towards an Improved Indoor Air Quality
Qingliang Yu
nr 165
Beyond Uniform Thermal Comfort
on the Effects of Non-Uniformity and
Individual Physiology
Lisje Schellen
nr 166
Sustainable Residential Districts
Gaby Abdalla
nr 167
Towards a Performance Assessment
Methodology using Computational
Simulation for Air Distribution System
Designs in Operating Rooms
Mônica do Amaral Melhado
nr 168
Strategic Decision Modeling in
Brownfield Redevelopment
Brano Glumac
nr 169
Pamela: A Parking Analysis Model
for Predicting Effects in Local Areas
Peter van der Waerden
nr 170
A Vision Driven Wayfinding Simulation-System
Based on the Architectural Features Perceived
in the Office Environment
Qunli Chen
nr 171
Measuring Mental Representations
Underlying Activity-Travel Choices
Oliver Horeni
nr 172
Modelling the Effects of Social Networks
on Activity and Travel Behaviour
Nicole Ronald
nr 173
Uncertainty Propagation and Sensitivity
Analysis Techniques in Building Performance
Simulation to Support Conceptual Building
and System Design
Christian Struck
nr 174
Numerical Modeling of Micro-Scale
Wind-Induced Pollutant Dispersion
in the Built Environment
Pierre Gousseau
nr 175
Modeling Recreation Choices
over the Family Lifecycle
Anna Beatriz Grigolon
nr 185
A Distributed Dynamic Simulation
Mechanism for Buildings Automation
and Control Systems
Azzedine Yahiaoui
nr 176
Experimental and Numerical Analysis of
Mixing Ventilation at Laminar, Transitional
and Turbulent Slot Reynolds Numbers
Twan van Hooff
nr 186
Modeling Cognitive Learning of Urban
Networks in Daily Activity-Travel Behavior
¸
Sehnaz
Cenani Durmazoglu
�
nr 177
Collaborative Design Support:
Workshops to Stimulate Interaction and
Knowledge Exchange Between Practitioners
Emile M.C.J. Quanjel
nr 187
Functionality and Adaptability of Design
Solutions for Public Apartment Buildings
in Ghana
Stephen Agyefi-Mensah
nr 178
Future-Proof Platforms for Aging-in-Place
Michiel Brink
nr 188
A Construction Waste Generation Model
for Developing Countries
Lilliana Abarca-Guerrero
nr 179
Motivate:
A Context-Aware Mobile Application for
Physical Activity Promotion
Yuzhong Lin
nr 189
Synchronizing Networks:
The Modeling of Supernetworks for
Activity-Travel Behavior
Feixiong Liao
nr 180
Experience the City:
Analysis of Space-Time Behaviour and
Spatial Learning
Anastasia Moiseeva
nr 190
Time and Money Allocation Decisions
in Out-of-Home Leisure Activity Choices
Gamze Zeynep Dane
nr 181
Unbonded Post-Tensioned Shear Walls of
Calcium Silicate Element Masonry
Lex van der Meer
nr 191
How to Measure Added Value of CRE and
Building Design
Rianne Appel-Meulenbroek
nr 182
Construction and Demolition Waste
Recycling into Innovative Building Materials
for Sustainable Construction in Tanzania
Mwita M. Sabai
nr 192
Secondary Materials in Cement-Based
Products:
Treatment, Modeling and Environmental
Interaction
Miruna Florea
nr 183
Durability of Concrete
with Emphasis on Chloride Migration
Przemys�aw Spiesz
nr 184
Computational Modeling of Urban
Wind Flow and Natural Ventilation Potential
of Buildings
Rubina Ramponi
nr 193
Concepts for the Robustness Improvement
of Self-Compacting Concrete:
Effects of Admixtures and Mixture
Components on the Rheology and Early
Hydration at Varying Temperatures
Wolfram Schmidt
Propositions
Propositions associated with the thesis
Modelling and Simulation of Virtual Natural Lighting Solutions in
Buildings
1. Applying the ideal VNLS, which combines artificial light and view together, will have
more positive effects on human well-being, compared to only applying artificial bright
light without a view, or applying an artificial view without emitting light.
This thesis, Chapter 1.
2. Directionality of the light is an important property that typically distinguishes a real
window or skylight from an artificial version.
3. Developing future solutions such as VNLS is a long process. The use of modelling and
simulation is important in influencing the design decision; therefore it deserves to be
discussed on its own.
4. In climbing towards the goal of making robots appear human, our affinity for them
increases until a certain point where we realise that the human-like robot may at the first
instance look real, but is in fact artificial, so that we experience an eerie sensation. Such a
sensation occurs since we at that point perceive the robot as a proximal threat – it would
hardly occur if the ideal object was, for instance, a window.
Mori, M. (1970). The uncanny valley. Energy, 7, 33-35 [in Japanese].
5. It must not be believed that because almost all problems can be solved with computers,
there is no need to examine the properties of the solutions. It is always essential to choose
reasonable approximations to solve the problem; it is a nonsense to develop or run a big
routine to compute results that can be obtained simply through analytical methods.
Filippi, F., Habault, D., Lefebvre, J-P., Bergassoli, A. (1999). Acoustics: Basic
Physics, Theory and Methods, London: Academic Press.
Propositions
6. Architects and the building industry have started moving in the direction of sustainable
practice. The importance of having a critical attitude is even greater now than it was.
Unfortunately there are many who use the label of ‘sustainable’ without the substance.
Few dare to say to them that ‘the emperor has no clothes’.
Szokolay, S. V. (2008). Introduction to Architectural Science: The Basis of
Sustainable Design (2nd ed.). Oxford: Architectural Press.
7. In the long journey of a scientific adventure, of which a doctoral project is a typical
example, it is very easy to get lost in the jungle of research.
8. Lecture notes can only give methods, but we must define the boundary condition
ourselves. We should go to the streets, out to the villages, take notes on every symptom,
and understand the real problems. What is the meaning of art, if separated from the
anguish of environment. What is the meaning of thinking, if separated from the problems
of life.
Rendra, W. S. (1977). Sajak Sebatang Lisong (Poem of a Cigar) [in Indonesian].
9. Today individuals can detach themselves from the village unit in order to move to a
nearby city, another island, or even to an urban setting halfway around the world. Those
who have lived away from their village, in the rantau, for much of their lives, do return to
the village because they still consider it as a comfortable place to retire in their old age.
Tanner, N. M. (1982). The nuclear family in Minangkabau matriliny: A mirror of
disputes. Bijdragen tot de Taal-, Land- en Volkenkunde, 138, 129-151.
10. The most succesful person is the one who manages to give the most benefits to his family,
society, and environment. As the Prophet said, “When a son of Adam dies, his actions are
cut off except for three: a continuing charity, knowledge which gives benefit, and a
virtuous child who prays for him”.
Prophet Muhammad S.A.W. Shahih Muslim, hadits narrated by Imam Muslim, from
Abu Hurairah [in Arabic].

Modelling and Simulation of Virtual Natural Lighting

Transcription

Similar documents

Second Annual Indiana Winery Tour

BevMo! - Bill Brown Sales

Alfredo De Vido Architects Kate`s Paperie

resources - Old

sendai mediatheque - toyo ito

Residential Lighting (Good Neighbor Guide)

Flowing Design - Creative Lighting Design and Engineering

COOLUX specifications

A Powerpoint Presentation by Astronomer Robert Gillette

Let there be... LED