Scientific and technological advances from DCNS_no. 2
Transcription
Scientific and technological advances from DCNS_no. 2
FO RM AT IO NM AN AG ST EM EA LTH EN T AN DA IN NTEN TE GR NA AT IO N IN PER FOR MA AN NCE A DM T ARI SEA NE PRO PLA DU TFO CTI RM VIT DYN YA ND AM ICS C OF OMP IND ETI UST TIVE ENE NES RIA RGY L P OPT RO S CES IMI SES SAT ION TH RO UG HL ON IFE BO RES AR IST DI AN NT CE ELL IGE NC E RESEARCH Scientific and technological advances from DCNS_no. 2 CONTENTS 04_ EDITORIAL 06_ FOREWORD 08_ NEWS 11_ PERFORMANCE AT SEA AND MARINE PLATFORM DYNAMICS Virtual Ship: integrated ship design in the preliminary project phase MAX: generic autonomous model for submarine manoeuvrability studies 17_THROUGH LIFE RESISTANC Modelling of mechanical propulsion transmissions Composite material antenna systems Multifunctional composite materials for military naval applications COVER ILLUSTRATION: computer-generated image representing fibre optics. 25_ENERGY OPTIMISATION Optimised operation of ships: simulations and optimisations 29_ONBOARD INTELLIGENCE Architecture of a guidance system for autonomous vehicles 33_INFORMATION MANAGEMENT New data association processing solutions Tracking manoeuvring targets in 3D Innovation and human factors 43_STEALTH AND ANTENNA INTEGRATION Vibroacoustic radiation by plates in transient states 47_PRODUCTIVITY AND COMPETITIVENESS OF INDUSTRIAL PROCESSES Evaluation of friction stir welding (FSW) on high yield strength steels for shipbuilding The development of TOFD at DCNS Biofilms and corrosion of corrosion-resistant alloys in sea water Digital engineering and future shipyard 60_OUR SCIENTIFIC PUBLICATIONS RESEARCH_2. The DCNS scientific and technological review. Executive editor: Gilles LANGLOIS _Editorial board: Christian AUDOLY, Julien BÉNABÈS, Alexia BONNIFET, Luc BORDIER, Marc BOUSSEAU, Jean-Michel CORRIEU, François CORTIAL, Xavier DAL SANTO, Sylvain FAURE (CEA), Fabien GAUGAIN, Anne-Marie GROLLEAU, Joëlle GUTIERREZ, Emmanuel HERMS (CEA), Guillaume JACQUENOT, Dann LANEUVILLE, Cédric LEBLOND, Jean-Jacques MAISONNEUVE, Thierry MILLOT, Pol MULLER, Adrien NEGRE, Antoine PAGÈS, Fabian PÉCOT, Mathieu PRISER, Ygaal RENOU, Lucie ROULEAU, Céline ROUSSET, David ROUXEL, Florent SAINCLAIR, David-François SAINT-CYR, Jean-François SIGRIST, Camille YVIN _Design and production: _Photo credits: DCNS – all rights reserved. RESEARCH_2 03 EDITORIAL “The acceleration of worldwide competition both in shipbuilding and in the marine renewable energy sector necessitates the development of new design and production methods.” Gilles Langlois, head of DCNS Research DCNS Research is one of the growth drivers of the also regularly produce innovative test protocols for the Group. Its contribution will enable us to take up the aerospace industry. boldest challenges in shipbuilding, energy and the sustainable development of the oceans. To speed up the international development of DCNS, DCNS Research practises innovation in all its forms: DCNS Research brings together internationally-reputed 1 • technological: we invite you to read about some researchers, experts with experience of complex pro- examples i n issue no. 2 of the DCNS RESEARCH grammes and engineers eager to identify the emergent magazine; trends with DCNS customers. 2 • marketing: our experts are working continuously on the market positioning of our innovations; 04 Bringing together areas of excellence as varied as 3 • operating concepts: our teams are working on the corrosion, materials, acoustics and algorithms, not for- exploration of new technological building blocks which, getting hydrodynamics, DCNS Research makes a wide like drones, will revolutionise military operations and range of contributions to the DCNS Group: from expert exploitation of the oceans in the years to come; assessment on submarines in service to the develop- 4 • process: through open innovation and collaborative ment of technology for significantly increasing the ope- research, international partnerships and integration rational capabilities of the ships produced by DCNS, we since 2008 of a true SME, SIREHNA®, into the Group. RESEARCH_2 EDITORIAL Spearhead of DCNS Group R&T thanks to the com- The world is accelerating, let’s accelerate! bination of expertise, research and modelling, DCNS In conclusion, I want to stress that the application of Research contributes to the vision and to the deve- innovations in emerging markets and the development lopment of the Group. It is a vector of international of industrial and research partnerships enable us to growth for the DCNS Group, for example through its observe that the results and work of DCNS Research cooperation projects with research centres in France, lead to competitive advantage, customer satisfaction, in Europe and around the world. but also a real career boost for our employees. DCNS Research is heavily involved in the development I wish you excellent reading of the second issue of of a collaborative framework, in particular with the IRT RESEARCH magazine. Jules Verne, the IRT SystemX, the “l’Usine du futur” (future factory plan) research group, competitiveness clusters such as EMC2 and academic and industrial partners in all sectors of activity. The research work in the shipbuilding and marine renewable energy sector concerns the design and manufacturing processes for metal structures, and the behaviour of these structures in the marine environment (resistance to the sea, shocks, corrosion, etc.). The research & development and innovation of the Group gets organised in France In a few months, the DCNS Research activ ities in Nantes will be relocated to the Technocampus Océan (Nantes/Bouguenais). In addition to its fi nancial investment in this project, the Pays de la Loire regional council is project manager for the architectural project comprising 16,000 m 2 of offices and laboratory space. The platform will combine common, office and workshop spaces for collaborative projects and facilities, as well as private spaces, rented by companies and university researchers. The complex will be a true scientific showcase, a tremendous vector of the region’s capacity for innovation. Information-related activities applicable to embedded systems and to maritime surveillance and security networks will be relocated to the Technopôle de la Mer at Ollioules (south-eastern France). RESEARCH_2 05 FOREWORD The Océanides project Claudie Haigneré, President of Universcience Throughout history, the oceans have always been innovation amplify the flourishing of ideas and speed up sou rces of power a nd prosper it y for hu m a n it y. the emergence of new technologies. That is what the Océanides project is intended to study and illustrate. Sponsored by the minister of Ecology, But there are also considerable human issues. One of Sustainable Development, Transport and Housing, the these issues is in education and training. Our world is Océanides project, a DCNS initiative, is being carried out changing every day, our time is fi lled with uncertainties in partnership with some twenty companies, local autho- and, little by little, some citizens are fi nding themselves rities, higher education institutions, research institutes on the margins of this society of knowledge. We must and professional federations, with active support from also prepare our young people to master, and tomorrow the French Cluster Maritime. invent, this ever-changing technology. How can not only sound knowledge but also the thirst for understanding, Sustainable development of the oceans and protection of the appetite for learning and the enthusiasm for creating marine resources are today still areas full of promise for be transmitted to as many as possible? the future and new opportunities for our industries. It is now necessary to invent, in society as in compa- 06 However, the growth and the industrial recovery that nies, new forms of communication, more interactive, our country needs, in a globalised environment where more participative and more creative, restore pas- the most precious resource is knowledge, necessitate an sion to the professions of industry, give new life to the increased effort in the areas of research and innovation. technical sector. There are scientific and technical issues underlying this I salute the DCNS initiative to publish the latest results effort. Industry, public-sector laboratories, universities of its technological research. Much of this work is being and other higher education institutions must collaborate done by brilliant doctoral students. I hope that they to invent, innovate and develop new concepts, new pro- contribute to mobilising yet more young people for the ducts and new services. Collaborative research and open fi ne professions of the sea. RESEARCH_2 FOREWORD RESEARCH_2 07 NEWS DCNS is developing research programmes with Dalhousie University At the end of November 2013, Patrick Boissier announced the signature of a partnership with Dalhousie University in Halifax for developing bilateral research programmes. The MoU (Memorandum of Understanding) defines the general framework for collaboration programmes for five years. According to Jean-François Sigrist, manager of the research team at DCNS Research, this MoU represents a nonnegligible competitive advantage for DCNS: “We are going to develop new research work with Dalhousie University on the ‘dynamics of structures’ (surface ships, submarines, offshore platforms, etc.) and their behaviour in the marine environment. We are also planning scientific exchanges between France and Canada, and hosting of students (master, doctorate).” These encouraging prospects for DCNS, in the position of challenger in the face of competition from the United States, echo the announcement by Prime Minister Harper and President Hollande in June of a roadmap for bilateral cooperation, covering for example international security and defence. They also contribute to a climate of strengthened economic partnership following the recent signature of a free trade agreement between Canada and Europe. DITCHING TESTS DCNS Research and SIREHNA® are working together on rigorous ditching tests. In recent years, the study of the behaviour of helicopters in the event of ditching at sea has become essential in the context of the certification of the floatation and safety systems fitted to these aircraft. Ditching tests also provide data that can only be obtained experimentally, as full-scale ditching tests are inconceivable. For the manufacturer and/or the supplier of floatation systems, the objective is to validate the dynamic stability of the helicopter for various float and centre of gravity configurations and environmental conditions (sea state and wind), stability which will enable the crew to evacuate the helicopter in the event of ditching. At present, there are two types of sea landing test: dynamic stability tests, the objective of which is to validate the seaworthiness of the helicopter once it has landed in the sea, and ditching tests proper, the objective of which is to characterise the sea landing phase. For these two types of test, the goal is for DCNS Research/SIREHNA® to supply a complete test package: specification and production of the model and the floatation system; balancing; choice of test facility; running and analysing the tests. 2013 RESEARCH PROJECTS Fabien Gaugain “Experimental analysis and numerical simulations of fluid-structure interaction of an elastic hydrofoil in cavitating and subcavitating flow.” Doctoral Thesis, École nationale supérieure d’arts et métiers de Paris, 2010-2013. Marine Robin “Validation of a fluid-structure coupled calculation chain for dimensioning structures under flow.” École nationale supérieure de l’énergie, l’eau et l’environnement, Grenoble INP, 2013. Élise Chevallier “Towards numerical simulation of fluid-structure interaction in large displacements.” EnseirbMatméca intership report, September 2013. David Louboutin “Experimental data acquisitions on DCNS configurations: comparisons with simulation (civa and athena) and modelling (mina) software.” Internship report 2013: “Ultrasound waves propagation in austenitic welds”, Institut français de mécanique appliquée. Alexandre Rochas “Feasibility of a seawater pump impeller made of composite materials – surface ship application”, Ensiacet internship report, August 2013. Lucie Rouleau “Vibro-acoustic modelling of sandwich structures with viscoelastic material layers”. Doctoral thesis, Conservatoire national des arts et métiers de Paris, October 2013. 08 RESEARCH_2 NEWS KEY EVENTS 2013 RESEARCH DAYS Nantes-Indret (France), 25-26 June 2013. On 25 June, in partnership with the Audencia management school and the consultancy firm Bessé, more than 220 persons from regional and national entities involved in the economy, research and education met at the initiative of DCNS Research with 15 experts and specialists to initiate a constructive dialogue on risk management, a key factor in innovation. INDIA-FRANCE TECHNOLOGY SUMMIT New Delhi (India), 23-24 October 2013. Participation in the Technology Summit and presentation of technological advances and innovations, and also partnerships with Indian universities (MoU). From rapid prototyping… to direct production! Additive manufacturing, or 3D printing, combines all the processes enabling a physical object to be produced from a digital object by adding material layer by layer. These numerous processes can be categorised by type of material (metallic, polymer, ceramic, in liquid, powder, sheet or wire form) or by deposition technique (seven major families including directed energy deposition and powder bed fusion for metallic materials). This approach has a number of advantages: unrivalled freedom of geometrical design, manufacturing of parts close to the final dimensions and with short lead times. Current applications involve many fields, including medical, tools, vehicles and fashion, as well as energy and aerospace. The potential for applications at DCNS is very broad: for parts with complex geometry, for coatings or even parts with composition gradients, for repairs in EUROPORT 2013 Rotterdam (Netherlands) 5-8 November 2013. DCNS Research was at the international meeting place for marine science and technology. The programme included improvement of energy consumption and limitation of pollution emissions. METS 2013 Amsterdam (Netherlands), 18-20 November 2013. SIREHNA® exhibited its latest innovations and technological advances in the Super Yacht Pavilion at the Marine Equipment Trade Show. OMAE 2013 Nantes (France), 9-14 June 2013. DCNS Research was present at the OMAE 2013 forum to present its latest technological advances and discuss the development of the oceans with experts, researchers, engineers, technicians and students. new-build or maintenance, process on its own or in combination with others (hot isostatic compaction of powders, for example), etc. Encouraging initial work has already been done by DCNS Research to assess the potential of the process for stainless steels and titanium alloys, and its technical and economic interest, in particular for the manufacture of plate heat exchangers, difficult to produce by conventional methods. RESEARCH_2 OFFSHORE EUROPE 2013 Aberdeen (UK), 3-6 September 2013. For the 40th anniversary of the Offshore Europe show, attended by more than 63,000 persons, DCNS Research participated in talks on the theme “The next 50 years”. 09 10 RESEARCH_2 PERFORMANCE AT SEA AND MARINE PLATFORM DYNAMICS Hull drag computation, model tests in a hydrodynamic basin, simulating the seakeeping characteristics of a structure, optimising propulsion units, design of dynamic stabilisation systems, analysing the launch of underwater weapons: this sector embraces all activities enhancing the effectiveness and reliability of powered and unpowered marine platforms in mission execution. RESEARCH_2 11 PERFORMANCE AT SEA AND MARINE PLATFORM DYNAMICS Virtual Ship: integrated ship design in the preliminary project phase AUTHORS: Benoît Rafine and Julien Bénabès Virtual Ship is an R&D project initiated by DCNS Research. Its goal is the production of a software environment for optimum whole warship architecture design in the pre-sale phase. Virtual Ship is a multiphysical, multidisciplinary and multiscale collaborative digital integration platform. The project is led by the CEMIS in collaboration with SIREHNA®. Project context The design of a naval system is one of the most complex major industrial projects. For example, Le Terrible, the latest French SSBN, weighing more than 12,000 tons, incorporates a million components; its construction requires 15 million hours of work and involves 6,000 contractors under DCNS project management. In comparison, a car weighing 1.9 tons has “only” 3,000 parts, with assembly requiring 23 hours of labour. The construction budget of a warship is generally between 1 and 10 billion euros, for a limited quantity of 1 to 10 vessels with service lifetimes of around fi fty years. A warship must incorporate a large number of subsystems and equipment items manufactured independently of the project. It also incorporates a very large number of different and sometimes confl icting functions, and it has to be able to operate in a hostile and hazardous environment. Many varied scientific disciplines are involved in the technical and operational defi nition of the ship. These disciplines do not handle the same design variables and do not need the same level of detail nor the same computing time. Warship design is consequently a multidisciplinary design activity, in which the challenge for the naval system architect is to have an overview of system performance in order to facilitate the design choices. The preliminary “pre-sale” design phase is fundamental in the overall design process of the ship. During this phase, there is a close relationship between the customer and DCNS in order to structure and clarify the customer’s need and identify the optimum technical-economic configurations for the ship. This phase provides the greatest freedom of design choices, 12 so innovative solutions can be explored. Lastly, knowledge of the overall behaviour of the ship must increase rapidly in this phase: the objective is to minimise the technological risks as quickly as possible so as to control costs and be in a position to submit a binding proposal to the customer. Product control over the project life cycle. Design Review Contract Development Concept Design Feasibility Contract Design Definition Validation Contract reference cost Budget Assessment System knowledge Design Review Detailed Design Development reference cost Production reference cost 100% Ideal Real In this context, DCNS Research initiated the Virtual Ship R&D project. The goal of the project is the production of a multiphysical, multidisciplinary and multiscale collaborative integration software environment for optimum whole warship architecture design in the pre-sale phase. RESEARCH_2 PERFORMANCE AT SEA AND MARINE PLATFORM DYNAMICS Meta-model NATO Architecture Framework Technical Requests Rules & Standards Environment Budget Technology Ship Concept Cost Estimate Technology Plan Models Functional Physical (Digital Mock-Up) Behavioral Probabilistic NDMS ASRU EHCLS* EHCLS* Decoy launcher DECOY'S LAUNCHERS Design Space Mutli-physics Simulation Optimization Strategy Visualisation Multi-objectives Analysis Architecture of the Virtual Ship environment. Objectives and expected results The following main scientific results are expected for the Virtual Ship project: • better management, structured and shared at system level, of the design requirements (customer need, manufacturing constraints, budgetary constraints, etc.); • connection between the functional modelling (requirements and functions) of the ship and its digital mock-up (physical description); • seamless direct relationship between the ship performance simulation results and the ship’s functional architecture; • a more automated approach to the pre-sale phase in order to shorten the response time and gain design study time; • effective exploration (more complete and optimal) of the design space; • better justification and traceability of design choices; • facilitation of knowledge management from the pre-sale phase; • closer collaboration between the different entities involved. Technological obstacles The development of the Virtual Ship software platform necessitates the overcoming of several technological and industrial obstacles, in particular: • the production of a global warship architecture model: the NATO Architecture Framework (NAF) will be adapted for description of the operational (customer need) and functional (ship functions and technical requirements) sub-models. This NAF will be connected to the physical model of the ship (physical tree structure and digital mock-up); • linkage and consistency of all the models contributing to the technical and operational definition of the ship. This involves enabling dialogue between the ship performance models and the architecture model of the ship itself (interoperability between models); • seamless relationship between the whole warship architecture modelling and the multiphysical simulation programmes in order RESEARCH_2 to enable direct coupling between the functional analysis of the ship and its requirements and the numerical performance simulations; • implementation of multiphysical and multidisciplinary optimisation strategies for Pareto-optimal exploration of the design space under time and resource constraints; • incorporation of often contradictory and sometimes subjective expert opinions into collaborative decision-making; • data management in a context of complex system and collaborative working environment; • minimal modification of existing whole warship performance models for their integration and use in the software platform; • introduction of industrial constraints into the Virtual Ship tool; • qualification of all the performance models and consistency of the scope of their validity. Virtual Ship is an ambitious project federating the activities of DCNS Research. It is based on close collaboration with DCNS engineering. The goal is to supply a practical aid tool for management of a complex system that imagination and experience alone cannot grasp totally. _REFERENCES A. BOVIS. Naval Systems: The Virtual Ship. Proceedings of the CSD&M Conference, 2013. C. KERNS, A. BROWN, D. WOODWARD. Application of a DoDAF Total-Ship System Architecture in Building a Design Reference Mission for Assessing Naval Ship Operational Effectiveness. Proceedings of the ASNE Global Deterrence and Defense Symposium, 2011. M. BOLE, C. FORREST. Early Stage Integrated Parametric Ship Design. Proceedings of the ICCAS Conference, 2005. A. PAPANIKOLAOU, S. HARRIES, M. WILKEN, G. ZARAPHONITIS. Integrated Design and Multiobjective Optimization Approach to Ship Design. Proceedings of the ICCAS Conference, 2011. 13 PERFORMANCE AT SEA AND MARINE PLATFORM DYNAMICS MAX(1): generic autonomous model for submarine manoeuvrability studies AUTHOR: Jérémie Raymond The hydrodynamic design of a submarine means that the skin shape has to be frozen very early in the preliminary project phase. Digital methods are employed to test different hydrodynamic configurations for a vessel project. They are used in the preliminary project phase to compare designs and rank their performance. Nevertheless, the hydrodynamic design of a submarine remains an empirical science. It is consequently not possible to rely completely on the digital approach. Once a design has unanimous acceptance, it must be characterised in the “real” world, by testing. This is the stage at which the new model MAX becomes involved. It can obtain this “real world” characterisation very rapidly. In addition, several steering gear variants can be compared. MAX enables the analysis of complex manoeuvres that cannot as yet be modelled by numerical simulations. It can also be used to validate computer codes and is positioned as a development in parallel to and complementing the “digital tank”. MAX: a multifunction underwater drone Design and manufacturing, started in December 2011, were cont r a cted by SM A En g i neer i n g to DCNS R ese a rch / SIREHNA®, which is also responsible for operational implementation of MAX when it is in use. Simple and rapid to put into operation, MAX is autonomous, autopilot-controlled underwater and on the surface, reprogrammable for rapid testing of different confi gurations and adaptable to the hydrodynamic scales representative of the whole range of DCNS submarines. It can be used to: • dimension the steering gear; • perform manoeuvrability tests (DG, DT) for identification of the mathematical model of the vessel; • study the impact of specific appendages (stringers, deck shelters, etc.); 14 • defi ne the control laws; • conduct specific studies such as the effect of sea bed proximity on vessel behaviour. 2012 was spent designing the model, while 2013 saw model assembly and qualification trials at sea. General concept The general concept adopted is that of a generic tool, i.e. easily adaptable to the shapes and the characteristics of the DCNS submarine range. The model has a generic core structure comprising a watertight aluminium body 4 metres long to which specific elements (bridge fin, external shapes, steering gear) of the submarine to be represented can be fitted. The assembled model is between 7 metres and 10 metres long, according to the submarine being modelled, with a diameter of about 80 cm and a mass of about 2 tons. Designed to operate at depths down to 70 metres, the working depth is between 15 metres and 40 metres. RESEARCH_2 PERFORMANCE AT SEA AND MARINE PLATFORM DYNAMICS Onboard systems MAX has six independent actuators for rudder and plane control. Propulsion is handled by a 2 kW motor and the propulsion unit is adaptable to each new project. Control and safety device management functions are performed by two instrumentation and control computers. Test data is acquired by a dedicated acquisition unit. An IMU90 inertial measurement unit and an electromagnetic log give the speed and attitude of the submarine at any time. The weight regulating system is automatic, using a closed ballast. An acoustic communication device is used to determine the approximate position of the model when diving and tracking it during its movements. A Wi-Fi communication system is used for control on the surface and test data retrieval as soon as the model surfaces and for planning the next mission. View of the steering gear. Tests and initial conclusions After a seven-month assembly and workshop acceptance phase (January-July 2013), MAX went to sea for the first time in September. For the occasion, it was con figured as the Scorpene® Chile, a submarine for which a large volume of data is available for comparing the behaviour of the model with that of the real submarine. The tests took place from September to November at the La Ciotat site (between Toulon and Marseille). They validated the major operational qualities of the model (ability to perform complex manoeuvres, ease of use, productivity) and the representativeness of MAX with respect to the real submarine. 2014 prospects With its shelter and dedicated test logistics, MAX can be deployed rapidly to different sea or lake test locations. 2014 should see testing conducted by the SCO400 programme to d i mension the steer i n g gear a nd for tests speci f ica l ly designed to determine the manoeuvrability model. The possibility of using a torpedo version of MAX to study swim out is also being considered. A major development for DCNS, confi rmed by Vincent Geiger, Deputy Director Deterrent Overall Architecture (AED) at SMA: “MAX should divide the hydrodynamic design work cycle time by two and the costs by four.” MAX operating on the surface. (1) In homage to the father of submarine manoeuvrability studies, Mr Max Aucher. Max Aucher (Ecole Polytechnique, class of 1942), director of the Bassin d’essais des carènes (model testing basin) from 1979 to 1982, completed the mathematical model of submarine manoeuvrability, developed a method of extrapolation to full scale of surface ship resistance and self-propulsion model tests, and contributed to propulsion system noise reduction research. Launching. RESEARCH_2 15 16 RESEARCH_2 THROUGH LIFE RESISTANCE Throughout their working life, structures, whether metallic or not, undergo natural or accidental attack: corrosion, water impact, fire, shocks. Day after day, they also face the phenomena of fatigue and ageing. These attacks and other phenomena call for calculations and testing to assess the durability of structures, as well as identify and try out technological solutions to improve it. RESEARCH_2 17 THROUGH LIFE RESISTANCE Modelling of mechanical propulsion transmissions AUTHOR: Romain Fargère Power transmission is a major concern on a ship and affects the whole of the transmission system, from the reduction gearbox to the propulsion propeller. Many factors are involved, including reliability, acoustic discretion and production cost. For an effective response to each of the issues, DCNS Research has been working for several years on the development of dedicated numerical tools specifically designed to take account of coupling between the various components and the physical processes. Used for a long time in research applications, these tools are being rolled out in the engineering departments concerned. They provide new perspectives, both in terms of working methods, favouring numerical/experimental correlations, and in terms of design choices for the various components on future ships. Power transmission Power transmission is a concern with very major issues involving a number of complex components and consequently a number of physical processes: shaft lines, housings, gearing, bearings, coupling components, etc. Overall shaft line and housing vibration processes giving rise to acoustic discretion problems are combined with very localised processes. The latter mainly concern tooth contacts and bearings, and necessitate particular attention in both the design and the manufacturing phases. Development of dedicated tools Propeller (low speed) without links between them, necessitating a large number of manual iterations and providing a relatively simplified description of the physics of the problems. The increasingly demanding discretion or cost constraints and the complexity of current systems make it necessary to develop new tools at the junction of several fields of physics. These tools are being developed by the DCNS Research engineers using an approach that can be summarised in a few main stages: • problem analysis/literature survey; • defi nition of equations; Gearbox teeth in contact: excitation Gearbox casing Excited shaftline For a long time, DCNS has been using high-performance tools based on a high level of empiricism and on the experience of its technicians and engineers. More recently, with the development of numerical tools, the engineering and design departments have had access to tools w ith high performance but 18 Propeller (low speed) Fluid bearings Schematic diagram of the transmission system of a ship. RESEARCH_2 THROUGH LIFE RESISTANCE • digitisation/discretisation; • development of the solving method; • development of the communication interfaces (HMI). An example of transfer: reduction gearing simulation software A hydrodynamic bearing software program has already been transferred from the DCNS Research teams to the engineering departments concerned, and a reduction gearing dynamic behaviour simulation program is being transferred. The latter software is also an example of collaborative work, as it has been developed in part in collaboration with a laboratory that has extensive experience on this topic: LaMCoS at INSA Lyon. The tool in question comprises a human-machine interface (HMI), facilitating data input and post-processing, and a computing core for setting up and solving a nonlinear secondorder differential system w ith parametric excitation of the form: HMI PRE-PROCESSING Geometry, meshing, running conditons... Computing code Thermal loop Static calculation Bearings solutions Time increase Update teeth contact Solution of linéarized equation Bearings convergence? Teeth convergence? POSt-PROCESSING (bearing reactions and temperatures, tooth loading, varios coefficients…) Gearbox behaviour simulation software architecture. [M] Ẍ + [C] Ẋ + [K + Keng (t ; X)] X = Q + Qpal (X ;Ẋ) + Qeng (t ; X) The difficulties raised by certain nonlinear terms (for example contact terms such as bearings [pal] and gears [eng]) make it necessary to develop a specific numerical method [1], based on the substantial feedback from the university partner [2]. Tool qualification, the last stage before use In general, these tools require a long qualification phase, to which DCNS Research contributes. During this phase, each function is tested according to a strict specification and the results compared with proven results. To this end, a series of experiments is carried out on a high-precision, high-power (several hundred kW) test bench in order to compare the results of simulations with the vibration, temperature, load and other readings, supplementing a previous series [3]. From the point of view of the end user, this work remains transparent, and in the end the users have a near-custom tool meeting their specific needs in terms of both modelling precision and pertinence of the input and output data (loading [time and spectral distributions and analyses], pressures, temperatures, etc.), enabling completely conventional application of the tool in phases: • parameter input; • calculation by the tool; • processing and interpretation of the results. Conclusion The work described above represents some of the tools applicable to power transmission transferred to the engineering departments. The development of extensions to these tools is under study (gearboxes with complex architectures, bearings with various geometries, housing behaviour, etc.). This work is evidence of the RESEARCH_2 Single-stage reduction gear test bench. importance of the links between DCNS Research and the engineering departments, and of the role played by the research partners of the Group. _REFERENCES [1] R. FARGÈRE: “Simulation du comportement dynamique des transmissions par engrenages sur paliers hydrodynamiques”. Doctoral thesis, Institut national des sciences appliquées de Lyon, Villeurbanne, France, 203 p., 2012. [2] P. VELEX, M . MATAAR: “A mathematical model for analyzing the influence of shape deviations and mounting errors on gear dynamic behaviour”. Journal of Sound and Vibration, 191(5), p. 629-660., 1996. [3] S . BAUD; “Développement et validations sur banc d’essai de modèles du comportement dynamique de réducteurs à engrenages à axes parallèles”. Doctoral thesis, Institut national des sciences appliquées de Lyon, Villeurbanne, France, 194 p., 1998. 19 THROUGH LIFE RESISTANCE Composite material antenna systems AUTHOR: Patrick Parneix The SAMCOM collaborative project covers two aspects of the problem of insertion of antenna functions into composite material structures: design of antennas made entirely of composite materials, and most compact achievable integration of networked antenna elements into composite material load-bearing structures. Strong technical ambitions Changing technical and operational needs are leading to increasing use of telecommunications facilities and consequently of antennas on the platforms. This rapid increase is generating growing integration difficulties: problems of intrinsic performance of equipment (SWR(1)), masking problems, problems of electromagnetic compatibility between systems, physical installation problems, signature (radar or visual) degradation risks, problems of vulnerability in operational environments. The ambition of the SAMCOM project is to provide solutions to these various problems by means of composite materials and technologies, and more specifically: • by developing compact wideband communication antennas usable at sea and on land; • by developing methods for compact integration of antennas and antenna arrays in composite panels. Towards “custom” multifunctional composite materials For more than thirty years, radiocommunication systems operating in demanding environments have been using composite materials successfully as framework and protection for the metal radiating elements of antennas. Making use of DCNS experience, for example in naval shipbuilding, the goal of the SAMCOM project is better exploitation of the potential of composite materials: local adjustment of radioelectric characteristics (dielectric, conducting, insulating, etc. materials), integration of periodic components or patterns, multifunctional elements; in other words, production of a “custom” composite panel. 20 Principal details of the SAMCOM project. SAMCOM (composite material antenna systems): • collaborative project FUI 9; • co-labelled by the EMC2 and Mer Paca clusters; • DCNS Research project lead partner; • 6 partners: Thales Communications & Security, Institut d’électronique et des télécommunications de Rennes (IETRUMR-6164, Rennes-1 university), Plastima Composites, CERPEM, CEMCAT and DCNS; • T0: 1/12/2010 – Duration: 54 months; • DCNS teams involved: DCNS Research CESMAN (lead), DCNS Research CEMIS; • DCNS ING SMA, DCNS ING SNS (Composite Developments – Communications Department), SER Brest. Each of the functions necessary for a composite antenna and for installation of antennas or antenna arrays in panels imposes specific requirements sometimes in conflict with the other functions. Consequently, one of the essential steps in SAMCOM is to compile the widest possible database of existing materials or materials to be developed within the project. Substantial work has been done by the IETR to develop reliable measurement methods for materials of very different natures and over extended frequency bands. In addition to the concepts of “conventional” materials, customary in composites even though here the performance targets are very demanding, periodic material concepts such as highimpedance surfaces have also been considered. RESEARCH_2 THROUGH LIFE RESISTANCE “All-composite” wideband antennas The use of carbon fibre fabrics as radiating element(s) of antenna structures has been investigated, initially in antennas with simple geometry as illustrated below, then in more complex systems. New communications antenna designs made entirely of composite materials have been developed both for civil applications and for more specifically military applications. Prototypes of each design are in the validation phase. The developed antennas succeed in the challenge of obtaining high microwave performance while achieving the initial objectives of compactness and frequency bandwidth. Operation of the antennas in panels will be validated on demonstrators representative of their actual environment, in a civil context for land application and in a naval and military context (work planned in 2014). Load-bearing panels favouring compact antenna installation The aim of this part of the project is to develop the necessary technologies and defi ne the rules for installation of antennas in composite panels, so as to: • optimise the radiation patterns of the integrated individual antennas; • ensure decoupling enabling physical proximity of the antennas. New solutions have been developed and will themselves be validated on the demonstrators mentioned above. This ultimate phase of the project should enable evaluation of the ability of the system to operate in an array and provide continuous search capability, as well as the operation of the individual antennas in different layout configurations. A highly collaborative project Like all the FUI (French single interministerial fund) projects, SAMCOM is an applied research project aiming to develop products or technologies likely to be put on the market in the short or medium term. At this stage of the project, it is evident that tangible advances have been recorded, some leading to patent applications. Some of these advances are likely to be exploited commercially in the short term, and perhaps even before the end of the project. The quality and the involvement of the partners, the support of the co-funders (State, Pays de la Loire Region), the excellent collaborative spirit favoured by the small number of partners, their complementarity and geographical proximity are no doubt the keys to the effectiveness of the project. DCNS is taking an active part in the project, as lead partner (DCNS Research) and also through the decisive contribution of t h e t e a m s of t h e C E S M A N ( M a t e r i a l s), t h e C E M I S (Electromagnetism), SMA (Engineering), SNS (Composite Materials Workshop, Communication Systems Dept) and SER (Brest Microwave Laboratory). (1) Stationary wave rate. _REFERENCES P. PARNEIX. Systèmes antennaires en matériaux composites. 13es Journées de caractérisation micro-ondes et matériaux. Nantes (France), 24-26 March 2014. L. MANAC’H, X. CASTEL, M. HIMDI. Carbon-Fiber Tissue as Radiating Element: Toward Pure Composite Materials Antennas. “International Conference on Electronic Materials.” IUMRS-ICEM 2012, 23-28 September 2012, Yokohama, Japan. L. MANAC’H, X. CASTEL, M. HIMDI. “Performance of a lozenge monopole antenna made of pure composite laminate.” PIERS Letters. Volume 35, 115-123, 2012. Example of evaluated conducting material. Composite antennas with copper or carbon fibre radiating elements [2]. RESEARCH_2 Examples of the radiation patterns in different frequency bands of an antenna array integrated into a composite mast. 21 THROUGH LIFE RESISTANCE Multifunctional composite materials for military naval applications AUTHORS: Patrick Parneix and Mathieu Priser The intrinsic properties of composite materials make them naturally multifunctional materials. These properties were first exploited to produce lightweight non-magnetic structures, durable in the marine environment. Gradually, the choice of components and the design of panels have been optimised in order to make use of this multifunctionality, either through the use of fibres, resins and core materials, themselves multifunctional, or by exploiting the capacity of composites to incorporate foreign components into their structure. This trend is growing, as increasing efforts are being made, often for reasons of increased durability or performance, to incorporate into the panel itself functions previously externalised (mainly as coatings), or to design coatings that are themselves multifunctional. New material concepts are emerging, such as “smart” materials and metamaterials. Naturally multifunctional materials The very fi rst applications of composite materials, and more specifically in naval shipbuilding, were based on their absence of corrosion and more broadly on their good ageing in the marine environment. This property remains a strong argument for the durability of naval structures made of composites and their moderate cost of ownership. The non-magnetism of these materials was a determining factor for their use as hull materials for minehunters. Lightness, low thermal conductivity, transparency to acoustic waves: there are many examples where the specific characteristics of composite materials have led to structural applications, multifunctional since the choice of the designer was guided as much by the capacity of the composites to fulfi l a structural role as by the particular function provided by this family of materials. Towards “custom” multifunctional composite materials Step by step, the preoccupation with optimisation of structures has led to work on incorporating new functions into composites by choosing the components no longer on the sole criterion of their mechanical performance, but also on the 22 basis of other performance criteria for functions that they would contribute to the end product. Furthermore, use has been made of the ability of composites to incorporate into their very structure foreign components providing a new function while minimising the impact on their mechanical performance. The development of production processes such as vacuum infusion moulding is making a large contribution to this innovation. > New functions by incorporation of foreign components into the structure One of the earliest examples is the incorporation of the electromagnetic shielding function into dielectric panels. Various technologies have been employed, such as the incorporation of fine metal grids or metallised fabrics between the layers of glass fibre composites. The ability to incorporate sensors and to carry information on the strain or the internal health of composite structures was subsequently exploited, for example by means of optical fibre Bragg gratings [1]. More recently, work has also been done on other internal health monitoring methods, including the incorporation of piezoelectric sensors for obtaining information on RESEARCH_2 THROUGH LIFE RESISTANCE Eridan-class minehunter. SSBN/NG Le Triomphant. local damage states of composite material structures [2] [3]. The antenna function is another function that can be incorporated. It will be seen below that recent work advocates antennas made entirely of composite materials. The insertion of periodic patterns in composite panels can form frequencyselective surfaces, which can be used to produce frequency fi lters for radome walls, for example. Organic matrix composite materials show relative versatility regarding their ability to incorporate foreign components. Nevertheless, the “foreign bodies” introduced into the composite structure generally act as defects, the impact of which on the overall mechanical performance of the panel should be minimised. > New functions through choice of components that are themselves multifunctional If the choice is available, preference is given to components that are themselves multifunctional, in order to confer upon the composite part purposes other than purely structural. For electromagnetic shielding, for example, it is advantageous to replace the grids by structural carbon fabrics. Antenna functions can be fulfi lled entirely using composite materials, and decoupling between antennas incorporated into panels can be achieved by using very specific composite materials. The use of certain core materials and appropriate choices of reinforcing fibres or resins or panel designs results in load-bearing structures that are transparent to electromagnetic waves or, conversely, absorbent. It is very clear that the choice of multifunctional components offers greater assurances about the lifetimes of composite material structures, and developments in naval shipbuilding are moving significantly in this direction. The principal difficulty is to identify and qualify the right materials. Future concepts The search for multifunctional materials is particularly active in all areas. In the naval area, three lines of development can be highlighted: RESEARCH_2 Insertion of piezoelectric sensors. Frequency-selective surfaces. • continuation of incorporation within the structure of functions at present more usually fulfi lled by coatings; • development of smart materials, ranging from simple sensitive materials to materials capable of reaction; • emergence of new material concepts. Following the logic leading to the search for increasingly functionalised load-bearing panels, it is natural to consider inserting into the structures function previously handled by coatings. Among the functions mentioned in this communication, the example of radar stealth is significant. The thickness and the sandwich structure of superstructure panels mean that various absorbing structural panel designs can be considered, providing broadband performance difficult to obtain with a simple coating. Incorporation of the function into the core of the structure also provides an additional assurance of durability in comparison with coating solutions: no degradation of the absorbent, no delamination, easier maintenance, etc. The smart materials concept covers a very broad area which can be defined as encompassing materials designed “with one or more properties that can be changed significantly in a controlled manner by external stimuli such as mechanical stress, temperature, humidity, pH, magnetic or electrical fields (currents), etc.” All of these functions are evidently of interest to naval shipbuilding for various applications, and work has been undertaken from the early 1990s. The evolution towards increasingly autonomous smart materials is a major line of development, mainly in the areas of acoustics and electromagnetism: adaptive acoustic materials (piezoelectricity, electrorheology), adaptive microwave materials (stealth, antenna incorporation, etc.), nanotechnologies. Over the last few decades, new material concepts have emerged. These materials draw their characteristics from their sub-wavelength-scale structural elements rather than from the intrinsic properties of their constituents. They form a new category of materials, with properties that cannot be found in natural 23 THROUGH LIFE RESISTANCE materials. These metamaterials fi nd applications in many areas of physics, and more particularly in electromagnetism, even though new concepts are starting to emerge in acoustics. This was the context in which DCNS funded the work for a thesis defended in 2013 entitled “Étude des interactions élastoacoustiques dans des métamatériaux formés d’inclusions résonnantes réparties aléatoirement” (Study of elastoacoustic interactions in metamaterials formed of randomly-distributed resonant inclusions) [5]. This work showed that particle local resonance mechanisms in an elastomer matrix could be exploited to increase the performance of hull acoustic coatings. This may turn out to be of particular interest in the area of low frequencies, by adjusting the geometry of the particles and the constituent materials. Radial stress field related to a dipolar resonance mechanism of a core-shell particle subjected to an incident acoustic wave [6]. In-pool characterisation of the reflection and transmission coefficients of acoustic panels consisting of a random dispersion of spherical particles. The basic principle applied in these materials is to use the substructure (random or periodic) of the material to generate interferences between objects at this sub-scale and produce apparent “exotic” properties at a macroscopic scale, representative of the material. Nevertheless, these materials usually show effectiveness in a relatively narrow band related to the characteristic size of their substructure (pitch of a periodic array, size of the objects of a random structure, etc.). One of the challenges is consequently to be able to demonstrate these properties over a broader frequency range. At present, metamaterials are a very active research topic and one of the keys to overcoming the technological barriers to the invisibility cloak concept, which would provide military naval structures with a level of stealth (radar or acoustic) constituting a technological breakthrough compared with present levels. In conclusion, as in other areas, the composites employed in naval shipbuilding are evolving rapidly. Over time, there has been a transition from multifunctionality limited to exploitation of the natural properties of these materials to a “targeted” multifunctionality, leading to complex panels themselves constituting integrated systems. New concepts are emerging, making use of unsuspected properties, shaking up the conventional perceptions and classifications in the field of materials. Obviously DCNS has to be active in this area, as illustrated by the example of the thesis mentioned above. 24 _REFERENCES [1] M. BUGAULT, P. FERDINAND, S. ROUGEAUD, V. DEWYNTERMARTY, P. PARNEIX, D. LUCAS. Health Monitoring of Composite Plastic Waterworks Lock Gates Using in-Fibre Bragg Grating Sensors. 4th European Conference on Smart Structures and Materials, Harrogate, United Kingdom, July 1998. [2] M. GRESIL, P. PARNEIX, M. LEMISTRE, D. PLACKO, J.-C. WALRICK. Lamb wave propagation in a hybrid Glass/Carbon composite laminate for electromagnetic shielding. 7th International Workshop on Structural Health Monitoring, Stanford, United States, September 2009. [3] M. GRESIL, P. PARNEIX, M. LEMISTRE, J.-C. WALRICK, D. PLACKO. Effet de l’insertion de blindage électromagnétique sur la propagation des ondes de Lamb dans un composite à renforts de fibres de verre. 16e Journée nationale sur les composites, Toulouse, June 2009. [4] P. PARNEIX, M. PRISER. Matériaux composites multifonctionnels pour applications navales militaires. ATMA 2013, Paris. [5] G. LEPERT. “Étude des interactions élasto-acoustiques dans des métamatériaux formés d’inclusions résonnantes réparties aléatoirement”, doctoral thesis, I2M, Bordeaux, December 2013. [6] G. LEPERT, C. ARISTÉGUI, O. PONCELET, T. BRUNET, C. AUDOLY and P. PARNEIX. Study of the acoustic behavior of materials with core-shell inclusions. Journées anglo-françaises d’acoustique physique (AFPAC) conference, Fréjus,January 2013. RESEARCH_2 ENERGY OPTIMISATION At a time when fossil fuel is becoming rarer and more costly, it is vital to consider all solutions to reduce energy consumption. This entails shape optimisation of hulls, energy saving control systems, lightweight structures, or even recovering a vessel’s stabilisation energy. New energy, including marine renewables, in search of higher yields, are also at the forefront of research solutions to recover energy better, and to store and transfer it. RESEARCH_2 25 ENERGY OPTIMISATION Optimised operation of ships: simulations and optimisations AUTHORS: Charles-Édouard Cady and Christophe Gaufreton As part of the EONAV FUI project, a decision support tool has been developed to provide recommendations for optimising the control of a ship. The tool generates savings of around 2% on the fuel consumption of a ship. Three types of optimisation are currently being developed: speed, electrical load-shedding and generator load optimisation. The genericity of the architecture used enables other types of optimisation (e.g. refrigeration, water production) to be accommodated. Introduction Changes in international regulations (target of a 20% reduction of CO2 emissions by 2020) and the lasting increase in the cost of oil are leading all shipbuilding industry leaders to propose innovations to improve the energy efficiency of ships both in the design phase and in operational use. The EONAV optimised ship operation system for reducing energy consumption and emissions is a response to this concern with protection of the environment and with control of energy costs for ship operators. The emphasis placed by DCNS Research on energy optimisation conforms to its strategic development priority for managing tomorrow’s energy issues and for broadening its range of innovating products. Decision support tools (DST) Optimisation The EONAV system has been designed to help a crew optimise the energy management of a ship through various levers: optimum speed, configuration of the generating plant and configuration of the electrical substations for load-shedding. Other optimisations such as water production and trim can be added to these in order to improve performance, but speed management is a major factor, as it accounts for 60% to 80% of the total energy consumption of a ship. DST energy model The core of the application is based on a complex multiphysical energy model of the ship comprising the following components: • hydrodynamic model incorporating the environmental constraints; 26 • mechanical model containing the “propulsion” part of the ship; • electrical model including the generators and the loads; • thermal model including the cold/heat producers and the loads; • emission model. The optimisation algorithms use all or part of the energy model to minimise fuel consumption. The reduction of emissions is a direct consequence of the lower fuel consumption. User modes With the energy model of the ship, there are three DST user mode options: • planning mode: for voyage preparation, recommending optimum speeds for the various routes and suggesting electricity generating plant configurations according to operational activity and system availabilities; • monitoring mode: used during the voyage for displaying the environmental indicators (EEOI, CO2, SOx and NOx) and for following up the recommendations. The tool adjusts and updates the optimisation solutions according to the route actually followed and to be taken; • analysis mode: for analysing and comparing completed voyages using the data recorded by the system. It contributes to the SEEMP imposed by the regulations. This mode can also be used to compare the model with the recorded data and detect any drift in the equipment. RESEARCH_2 ENERGY OPTIMISATION Monitoring mode HMI. Speed Technical innovations Design of a formal computing core The optimisation algorithms require the calculation of gradients, Jacobians and Hessians. These functions can be coded manually, but this takes a long time, is a source of error and is not very maintainable. For these reasons, a formal calculation core has been developed specifically for this task. It uses principles derived from the λ-calculation, with simplifications applied when the expressions are defined in order to keep the computing time close to that of direct coding. Manipulation of embedded simulators A simulator manipulation grid has been produced making it easy to add models and force values (or even replace a model by playback of recordings in real time) without having to establish a nd ma i nta i n con nections bet ween models ma nua l ly. To achieve this, a forward/backward chaining algorithm automatically infers the dependencies between the modules. Substantial performance gains are achieved using a lazy evaluation algorithm. average Time Start End Speed average Time Start End Speed Optimisations performed Speed The speed profile, with respect to the seabed and for a defined route, is optimised taking into account the environmental forces. Each speed profile is simulated and the profiles are compared according to the total fuel consumption that they generate on the trip. To reduce the size of the domain to be explored, speed profiles RESEARCH_2 average Time Start End Profile of speed over bottom on defined track. 27 ENERGY OPTIMISATION that are constant by segment are applied, and refined recursively while keeping the total trip time constant. Diesel generators The electrical load is distributed between the diesel generators so as to minimise total instantaneous consumption. This is done using consumption/load curves which can be manipulated as mathematical functions thanks to the formal computing core. A solver implementing the inner point method is then used to solve the optimisation problem. Load shedding The electrical loads to be connected to the network are optimised taking into account their importance (vital, semi-vital, non-vital) and their propensity to disturb the network. Mathematically, this is a linear discrete problem under linear constraints which is solved using a linear programming solver. The formal computing core enables this problem to be written simply. Functions supplied such that supplied requested Optimization problem Profile of speed over bottom on defined track. Recommendation example: the breakers marked in red correspond to the differences between the current state and the proposed optimisation. Conclusion The tool has been validated using consumption data recorded on the cruise ship Norwegian Epic. The saving was calculated by comparison between the actual and optimised consumption on the same voyage. The results show that the decision support tool improved fuel consumption by 2% on a ship with an already substantially optimised design. The cost optimisation is derived from the application of speed optimisation and electrical energy production optimisation. This project is an initial step in ship energy management using decision support tools. These results enable ship consumption and emission reductions of 5% to be targeted by adding new optimisation strategies: refrigeration, water production, trim, etc. Lastly, the decision support tool using energy indicators is a response to the regulations for ship operators, enabling them to manage the energy efficiency management plans of their ships. 28 Profile of speed over bottom on defined track. RESEARCH_2 ONBOARD INTELLIGENCE Faced by increasingly complex maritime operations related to reduced crews, operators must be supported by partly or completely automated systems, using advanced control rules, which need to be constantly improved and adapted to new situations. Security and energy saving requirements are also to be integrated into control algorithms and system architectures. If we push this logic further, we enter the world of drones or unmanned vehicles (UxVs), autonomous devices capable of carrying out tasks without human intervention. RESEARCH_2 29 ONBOARD INTELLIGENCE Architecture of a guidance system for autonomous vehicles AUTHOR: Denis Gagneux The use of aerial drones (Unmanned Aerial Vehicles) has grown exponentially over the last few years. Long restricted to military uses, they are now becoming available in the civil sector. Surface drones (Unmanned Surface Vehicles [USV]) and underwater drones (Unmanned Underwater Vehicles [UUV]), complementing aerial drones, have also seen their use become more widespread. They save manpower, either for carrying out repetitive tasks or for work in hostile environments. Since the early 2000s, combining its knowledge of marine platforms and automatic pilots, SIREHNA® (DCNS Research) has been contributing to this revolution through its work on guidance systems. The guidance systems are the heart of autonomous vehicles. Developed by SIREHNA®’s engineers for marine drones, these software and hardware architectures are the fruit of a decade of research work and experimentation. USVs and UUVs are autonomous vehicles that can be used in very diverse environments, ranging from secured zones (reserved exclusively for the drone) to zones of dense maritime traffic (ports, commercial shipping routes, etc.). In the near future, the adaptability of the level of autonomy of these drones will enable military and civil users to vary their uses: as remote-controlled vehicles, as vehicles whose movements are monitored by remote operators, and as fullyautonomous vehicles. The architecture of their guidance systems enables all navigation requirements to be met, while ensuring the safety of the vehicle and of its environment. vehicles or objects in its environment. If an abnormal event is detected, the secondary computer takes control of the vehicle in order to make it safe. It is powered by a different, redundant power source from that of the main computer, so that it can be operational in the event of main system electrical failure. It is also linked with a remote supervision station via a dedicated communication link separate from the one used by the main computer, enabling an operator to take action in the event of failure of the main system communication link. The technology used for the communication link depends on the environment of the autonomous vehicle. If it operates on the surface, a microwave link is used, while for underwater operation the link is acoustic. The architecture may incorporate various sensors to avoid collisions with objects in its environment: radar, AIS, laser, video camera or sonar (the only sensor that can be used on underwater vehicles). The technologies of these sensors are complementary, increasing the vehicle’s perception of its environment. The guidance system consists of a main computer (Command & Control) and a secondary computer (Safety) The main computer performs the functions related to navigation, for example by incorporating an autopilot. The secondary computer ensures the safety of the vehicle and of its environment: it checks the consistency of the commands output by the autopilot and detects vehicle failures and potential collisions with other 30 RESEARCH_2 ONBOARD INTELLIGENCE Graphic interface of the supervision station. The operator monitors the movement of the vehicle using the digital map module. The architecture is currently deployed on a surface drone (Remorina) and on a non-tethered underwater model (MAX). RESEARCH_2 31 32 RESEARCH_2 INFORMATION MANAGEMENT The expanded use of electromagnetic, optronic and acoustic sensors in diverse operational situations has led to tremendous increases in data availability. Effective information management has thus become essential and now requires automated processes for better localisation, identification, characterisation and tracking. The underlying goal is to provide operators with reliable, relevant and real-time information that exposes the best possible decision. RESEARCH_2 33 INFORMATION MANAGEMENT New data association processing solutions AUTHOR: Olivier Marceau, Daniel Vanderpooten(1) and Jean-Michel Vanpeperstraete The sensors on board a vessel supply descriptive measurements of the objects around the platform. The data association function plays a key role in obtaining a synthetic representation of the environment of the vessel on the basis of these measurements. Work at DCNS Research aims to propose data association processing solutions which consume kinematic and qualitative information, without constraining assumptions (scan assumption) about the structure of the sensor measurements on reception. Description of the problem The data association function of a multisensor system (radar, infrared imager, etc.) is a critical technical component of search management. While each sensor supplies elementary information (positions, angles, identifications) describing the targets that it has detected, the data association function aims to combine information from sensors which designate the same target. ( ). describes the accumulation of scans: (1) Mathematical modelling of the association problem All the measurements are assumed to be made at fi xed times with A candidate hypothesis models a combination hypothesis for measurements that designate the same target. More precisely, is defi ned as a subset of measurements from such that: (in reality, the measurements are generally asynchronous). At each time , the sensor supplies a set of measurements called a “scan”. where is the number of measurements received in scan and is the -th measurement received in scan . The composition of a measurement depends on the sensor: for example, a 2D radar measures the distance and the azimuth angle of each potential target 34 for any . (2) (3) (2) indicates that each candidate hypothesis cannot contain more than one measurement of each scan. (3) states that a candidate hypothesis contains at least one measurement. RESEARCH_2 INFORMATION MANAGEMENT The quality of a partition is assessed through the quality of its constituent candidate hypotheses . Each candidate hypothesis of partition describes a combination of kinematic measurements which is more relevant the more consistent the measurements are with a kinematic behaviour of the targets to be detected. This kinematic consistency is assessed using a probability Candidates hypotheses . The optimum partition is thus the one that maximises a consistency criterion of the following form ([9]): (7) Tracks Measurements Figure 1. Examples of candidate hypotheses. where designates the set of partitions and (6). The output of the data association processing is a set sisting of candidate hypotheses. con- (4) must form a partition The candidate hypotheses of . In other words, each measurement of belongs to one, and only one, constituent candidate hypothesis of . for any (5) (6) Candidates hypotheses satisfying (4), (5) Problem (7) is a discrete optimisation problem, which has no guaranteed exact solution in polynomial time for all instances where is strictly greater than 2 ([9]). This optimisation problem in fact belongs to the category of NP-complete problems ([4]). Solving the association problem Equations (1) to (7) constitute the standard mathematical model of the data association problem. When the number of dimensions equals 2, the data association problem corresponds to the classic assignment problem, for which there are very effective exact polynomial algorithms (e.g. [1,2]…). Many publications (e.g. [3,5,8,9]…) propose algorithms that are suboptimal but efficient in terms of computing time when the number of dimensions is strictly greater than 2. Despite the large number of solutions available, DCNS Research is developing innovative solutions for associating data, for the following main reasons: • the scan structure constraint; • the use of qualitative data. Tracks Measurements Figure 2. Example of partition. RESEARCH_2 The scan structure constraint The scan concept is central to the modelling of the association problem, and most algorith ms make critical use of the concept. The scan models how the sensor measurements are presented to the data association processing. In the case of a rotating radar sensor, the scan is the set of measurements acquired during one complete antenna rotation. However, today operational problems are arising where the asynchronism of the measurements renders the modelling constraint imposed by the scan structure problematic. To deal with this problem, 35 INFORMATION MANAGEMENT DCNS Research has developed innovative and particularly effective data association methods which are not constrained by the scan concept. The use of qualitative data Most of the publications describe solutions which automatica l ly process the k i nematic measu rements (d ista nce, azimuth), the errors of which are characterised by a statistical model. In contrast, few publications ([6]) describe solutions for using qualitative information that is not characterised in statistical terms. Identification information is an example of qualitative information. However, qualitative data can be an essential additional information source for improving data association processing. Thanks to a partnership with the Lamsade laboratory, DCNS Research has obtained promising initial results on the use of qualitative data ([7]). DCNS Research is continuing its work with Lamsade in order provide technical data association solutions in the medium term capable of using the available qualitative information (1) PSL, Paris university Dauphine, Lamsade, place du Maréchal-de-Lattre-deTassigny, 75775 Paris Cedex 16. _REFERENCES [1] D. P. BERTSEKAS: “The auction algorithm: a distributed relaxation method for the assignment problem”. Annals of Operations Research, vol. 14, p. 105-123, 1988. [2] R. BURKARD, M. DELL’AMICO, S. MARTELLO. Assignment Problems, SIAM 2009. [3] A. CAPPONI: “Polynomial time algorithm for data association problem in multitarget tracking”. AES IEEE, p. 1398-1410, 2004. [4] M. GAREY, D. S. JOHNSON: Computers and Intractability. A Guide to the Theory of NP-Completeness. Ed Freeman. [5] H. GAUVRIT: “Extraction multipistes : approches probabiliste et combinatoire”. Thesis, Rennes university, 1997. [6] H. HUGOT, D. VANDERPOOTEN, J. M. VANPEPERSTRAETE: “A bi-criteria approach for the data association problem”. Annals of Operations Research, vol. 147, no. 1, p. 217-234, 2006. [7] O. MARCEAU, J. M. VANPEPERSTRAETE: “Automatisation des traitements et aides à la décision”. Internship report, 2013. [8] K. R PATTIPATI, S. DEB, Y. BAR-SHALOM,R. B. WASHBURN: “A new relaxation algorithm and passive sensor data association”. IEEE Trans on Automatic Control, vol AC-37, no. 2, p. 198-213, February 1992. [9] A. POORE: “Multidimensional assignment formulation of data association problems arising from multitarget and multisensor tracking”. Computational Optimization and Applications, p. 27-57, 1994. 36 RESEARCH_2 INFORMATION MANAGEMENT Tracking manoeuvring targets in 3D AUTHOR: Dann Laneuville Surveillance systems, the purpose of which is to determine the tactical situation in an extended zone covered by search sensors, and tracking systems focusing on a particular object, are sometimes confronted with particular situations in which certain objects in the scene are highly manoeuvrable. This may be the case of a personal watercraft or an inflatable boat in the context of asymmetric threat countermeasures, or of an ASBM (antiship ballistic missile) in the context of ABMD (antiballistic missile defence), for example. Such manoeuvres, combined with the potential presence of false alarms, are a real problem for the tracking algorithm, which may “lag”, showing an estimation bias (loss of precision), or even break lock, i.e. lose the track of the pursued object, obliging the system to regenerate the track after a possible search phase in the event of tracking. In both cases, the loss of performance can prove fatal if a threatening target has to be engaged. Consequently, the availability of a robust, highperformance manoeuvring target tracking algorithm appears to be an essential component of surveillance and tracking systems. The purpose of this work is to study a new algorithm for tracking a manoeuvring target that can manoeuvre vigorously in two or three dimensions. State of the art The state of the art in manoeuvring target tracking algorithms is represented by the IMM (Interacting Multiple Model) fi lter introduced in the 1990s ([2]). This is a recursive algorithm using several Kalman fi lters in parallel, each dedicated to a particular phase of the trajectory, for example a fi lter for the uniform movement phases (constant speed = no manoeuvre) and a fi lter for the manoeuvre phases, such as uniform turns (in which the speed does not var y and the turn rate is constant). This fi lter has recently been revised to take strict account of the case of fi lters of different dimensions in each mode ([7]), as will be the case here, and improved in its approach with respect to the optimum filter ([5]) with the RESEARCH_2 GMIMM (Gaussian Mixture based IMM), which retains the most probable r hypotheses in each mode, whereas the IMM merges them all into a single hypothesis. Today, the two most widely used models in an IMM fi lter are the NCV (Nearly Constant Velocity) model, which describes the u n i for m movement phases w ith state vector ([1]) X(t) = [x y z vx v y v y]’, and the NCT (Nearly Coordinated Turn) model for the coordinated turns in the horizontal plane ([1]) with state vector X(t) = [x y z vx v y ω]’ where ω is the turn rate in the plane. Until [3], the turns tracked by the IMM were in the horizontal plane. The recent approach developed in [3] also enables manoeuvres to be tracked in a vertical plane, but at present no fi lter satisfactor i ly processes manoeuvres made simultaneously in the two planes (genuinely 3D manoeuvres). That is the goal of our approach. New approach The two models above are described at present in Cartesian coordinates (CC) in the literature. We propose a new mixed representation: Cartesian for the position and spherical for the speed. This gives state vector X(t) = [x y z s ψ θ]’ for the fi rst model, where s is the modulus of the speed, ψ et θ are the two angles defi ning the direction of the speed vector (see fi gure 1 below), and X(t) = [x y z s ψ θ ω1 ω2]’ for the second model, where ω1 is 37 INFORMATION MANAGEMENT Figure 2. Scenario 1. Figure 3. Scenario 2. Figure 4. Scenario 3. 38 RESEARCH_2 INFORMATION MANAGEMENT the turn rate in the horizontal plane and ω2 is the vertical turn rate. the second-order discretisation of which is given by ([6]): dx1 = x 4 sin(x5) cos(x6)dt x1D = x 01 + tx 04 sin(x5) cos(x 06) t x x cos(x ) cos(x ) +— 05 06 2 04 07 2 t – — x x sin(x ) sin(x ) 2 dx2 = x 4 cos(x5) cos(x6)dt dx 3 = x 4 sin(x6)dt 2 dx 4 = σ 1 dW1 08 05 06 x2 D = x 02 + tx04 cos(x 05) cos(x 06) dx5 = x 7 dt t x x sin(x ) cos(x ) –— 05 06 2 04 07 t2 x x cos(x ) sin(x ) –— 2 dx6 = x 8 dt dx 7 = σ 2 dW2 04 and dx 8 = σ 3 dW3 2 04 08 05 06 (3) t2 x x cos(x ) x3D = x 03 + tx 04 sin(x 06) + — 06 2 04 08 x 4D = x 04 x5D = x 05 + tx07 Figure 1. Speed in spherical coordinates. x6D = x06 + tx08 x 7D = x07 x8D = x 08 The corresponding models of change over continuous time are given by the following stochastic differential equations: dx1 = x 4 sin(x5) cos(x6)dt dx1 = x 4 sin(x5) cos(x6)dt dx2 = x 4 cos(x5) cos(x6)dt dx2 = x 4 cos(x5) cos(x6)dt dx 3 = x 4 sin(x6)dt dx 3 = x 4 sin(x6)dt dx 4 = σ1 dW1 dx5 = σ2 dW2 Examples of results dx 4 = σ1 dW1 and dx6 = σ 3 dW3 dx5 = x 7 dt For further details of the discretisation of (1), and in particular the stochastic part, refer to [4] from which this summary is derived. (1) dx6 = x 8 dt dx 7 = σ 2 dW2 The results of two IMM fi lters, each using the two models NCV and NCT, are compared, the fi rst (IMM1) in Cartesian coordinates and the second (IMM2, new approach) in mixed coordinates. A radar, symbolised by a blue triangle on the scenario curves, supplies the 1 s rate of the distance measurements, where σr = 20 m, and circular and elevation angle measurements, where σang = 10 mrad. dx 8 = σ 3 dW3 These models have the general form of a stochastic differential equation: dXt = a(Xt) dt + b(Xt) dWt (2) In the fi rst scenario, the target performs a manoeuvre in the horizontal plane with a turn rate of 0.2 rads -1 at a speed of 250 ms-1, giving a load factor of 5 g. The performance of the two IMM fi lters is illustrated in the left-hand graph. It can be seen that the new approach significantly improves the performance in comparison with what is at present the best filter (IMM1) for manoeuvres in the horizontal plane. In the second scenario, the target performs a manoeuvre in a vertical plane with the same turn rate of 0.2 rads-1 at the same speed of 250 ms-1, giving the same load factor of 5 g. It can again be seen that the new approach significantly improves the performance. RESEARCH_2 39 INFORMATION MANAGEMENT In the third scenario, the target performs a 3D manoeuvre, i.e. in both the horizontal and vertical planes, with a turn rate of 0.2 rads-1 in both planes and at a speed of 250 ms-1, giving a load factor of 7 g. It can again be seen that the new approach behaves as well as previously (manoeuvre in one of the two planes) and significantly improves the performance. Conclusions and prospects We have presented a new approach to modelling the manoeuvres of a target, which not only improves performance on coordinated turn manoeuvres in the horizontal or vertical plane, but also maintains the same level of performance on genuinely 3D manoeuvres. This approach, more natural and more physical than Cartesian coordinates for modelling the manoeuvres of a moving object, consequently appears to be very promising for manoeuvring target tracking applications. Subsequent work consists in incorporating it into a tracking algorithm in order to test it in a multitarget environment with false alarms and detection gaps. _REFERENCES [1] Y. BAR-SHALOM, P. WILLETT and X. TIAN. Tracking and Data Fusion. A Handbook of Algorithms. YBS Publishing, 2011. [2] H. A. P. BLOM, Y. BAR-SHALOM. “The interacting multiple model algorithm for systems with Markovian switching coefficients”. IEEE Transactions on Automatic Control. 33, p. 780-783, August 1988. [3] J. GLASS, W. D. BLAIR, Y. BAR-SHALOM. “IMM Estimators with Unbiased Mixing for Tracking Targets Performing Coordinated Turns”. Proceedings of IEEE Aerospace Conference, Big Sky, United States, March 2013. [4] D. LANEUVILLE. “New Models for 3D Maneuvering Target Tracking”. Proceedings of IEEE Aerospace Conference, Big Sky, United States, March 2014. [5] D. LANEUVILLE, Y. BAR-SHALOM. “Maneuvering Target Tracking: A Gaussian Mixture Based IMM Estimator”. Proceedings of IEEE Aerospace Conference, Big Sky, United States, March 2012. [6] A. TOCINO and J. VIGO-AGUIAR. “New Itô-Taylor expansion”. Journal of Computational and Applied Mathematics, 158, p. 169-185, 2003. [7] T. YUAN, Y. BAR-SHALOM, P. WILLETT, E. MOZESON, S. POLLAK and D. HARDIMAN. “A multiple IMM approach with unbiased mixing for thrusting projectiles”. IEEE Transactions on Aerospace and Electronic Systems, 48(4):3250-3267, October 2012. 40 RESEARCH_2 INFORMATION MANAGEMENT Innovation and human factors AUTHOR: Chantal Maïs Taking human factors into account in the design of complex systems is necessary in order to ensure optimum operation. For warships, high-technology systems that are very complex to operate involving both individual and collective performance, in a harsh environment and in stress situations, it is a major issue: the effectiveness of the crew-ship unit in fulfilling its missions. In this view, the innovations must improve and increase the capabilities of the ships. What does taking the “human factor” into account contribute to innovation? It is necessary to: • identify the “level” of innovation: system upgrade or technological breakthrough? • characterise the type of innovation: technological, organisational, individual (related to the individual person)? But, whatever its nature, any innovation has an impact on the human aspects. This impact must be assessed sufficiently early in the design process to: • choose the appropriate ty pes and methods of human factors involvement; • conduct the upstream HF studies, when the predicted impact is high, in order to anticipate new uses, ensure the acceptability of the innovation and manage the change, in addition to the more conventional criteria of HMI usefulness and usability. DCNS at the cutting edge of virtual reality. Innovation and human factors When thinking of innovation, the main things that come to mind are technological and technical innovations. These innovations enable the development of increasingly automated and even autonomous systems (drones, for example), but also provide access to large quantities of data from multiple “big data” sources. New interaction techniques RESEARCH_2 such as augmented reality, virtual reality and touch screens offer new modes of interaction. These innovations entail modifications, changes and/or breakthroughs in how systems are used. The acceptability of such innovations is an important point to be taken into account, in addition to the utility and usability aspects. These changes entail changes in human organisations (in particular geographically-distributed collaborative or cooperative organisations, or access to low-level data and no longer only 41 INFORMATION MANAGEMENT summary data for the decision-making levels, etc.), and changes in the allocation of tasks between individuals (potentially with reduction of workforce) or in practices. But innovation is not limited to technological innovation. Any change at the level of the individual (changes in tasks, change in profile), at the level of a collective (workforce reduction, decentralised organisation), or at the level of the company (company development) can also be seen as an innovation, the impacts of which on the future users must be analysed and anticipated, incorporating the following data: • the various human aspects – anthropometry, physiology, cognition, conation, sociology, etc. – are not mutually independent. Their interactions in part determine the behaviour (and potentially the performance) of the users. For example, the confidence users have in the system can modify how they interact with the system; • individuals differ from each other in the various aspects. This interindividual variability generates different requirements, needs and behaviours; 42 • individuals change over time in the various aspects. As they gain experience, they acquire new knowledge, modify their cognitive structures, etc. In a user-centred approach, it is necessary to predict the impacts on the users, including on future practices, through a validation procedure (mock-up/evaluation) with the users as far upstream as possible. At DCNS, a company with a high level of technicality, innovation is most often considered in terms of technology. Taking human factors into account not only guides innovation, but is also a source of innovation! RESEARCH_2 STEALTH AND ANTENNA INTEGRATION How can you be somewhere without others being aware of your presence? How can you prevent disturbing the environment through which you are moving? Whether in the field of electromagnetics or acoustics, there are several technological solutions to choose from. These call for special materials, carefully designed forms, positionings and judicious decoupling. Seeking and identifying these solutions requires upstream modelling of the phenomena involved, using tools and methodologies capable of being updated to keep pace with threats and analytical needs. RESEARCH_2 43 STEALTH AND ANTENNA INTEGRATION Vibroacoustic radiation by plates in transient states AUTHORS: Thomas Leissing, Roch Scherrer and Christian Audoly The use of acoustic waves at sea for detection of enemy forces by means of passive sonar has been widespread for decades, and sonar technology is evolving constantly. The upgrading of passive sonars, now capable of analysing transient noise, makes it necessary to optimise acoustic discretion in nonsteady-state conditions. DCNS must consequently be capable of specifying transient acoustic discretion requirements for equipment and their support structures installed on vessels. In this context, DCNS Research is working on the study of vibroacoustic radiation mechanisms of structures in transient states, for example in a doctoral thesis in partnership with the INSA-Lyon acoustic vibration laboratory. The objective of this study is to improve knowledge in order to represent the mechanisms of vibroacoustic radiation of structures in transient states in terms of radiation sources, transfers and factors. Solving method To study the vibrations and the acoustic radiation of an immersed structure excited by a transient source, a simple structure is considered first: an infinite plate. The plate is immersed and subjected to a point pulsed excitation. The movement of the plate is expressed by the Love-Kirchhoff equation [1,2]: where D is the bending stiffness of the plate, and h are respectively the density and thickness of the plate, ω is the angular frequency, ξ is a structural damping term, w(r,t) represents the displacement of the plate as a function of the distance from the excitation source and time, f 0 is the amplitude of the exciting force at location r 0 and p(r,t) is the parietal pressure. Particular attention must be paid to the viscoelastic damping model, which must obey the causality principle. Solving this equation for the dynamics of thin plates in harmonic states gives the movement of the plate in the frequency domain. An inverse Fourier transform is then applied to obtain the pulse 44 response of the plate and the pressure radiated in the fluid in the time domain. A schematic representation of the solving principle is shown in fi gure 1. Application to an immersed infinite plate The method used is validated by comparison with a plate in a vacuum, for which there are analytical solutions [3]. The study of vibration on an immersed plate shows the effect of the fluid, the dispersion of the plate waves and the fluid/solid interface waves. The study of the radiated pressure shows directivity of the radiation and propagation of the waves in the plate before being radiated. An example of simulation of an immersed aluminium plate is shown in fi gure 2. The pressure levels in the fluid at four different times can be seen (left to right and top to bottom: t = 0.23 ms, t = 1.40 ms, t = 2.79 ms and t = 6.12 ms). The planar plate is located along z = 0, so its edge is represented by the black dotted line. The exciting force consists of a pulse at (t = 0, z = 0). The colour map shows the pressure levels in the fluid. The propagation of the waves in the fluid can be observed in the four panels. Four virtual hydrophones are placed in RESEARCH_2 STEALTH AND ANTENNA INTEGRATION Unknown transfer function Source Answer Fourier Transform Inverse Fourier Transform Source Answer Transfer function known Figure 1. Schematic representation of the solving method. Figure 2. Simulation of the pressure field radiated by an immersed infiDinite plate excited by a point force. RESEARCH_2 45 the fluid medium; they are represented by the filled black circles. The signals received at these points are shown on time/amplitude graphs. It can be seen that the temporal characteristics of the signals differ greatly according to the distance between the excitation point and the observation point. The signal received by the nearest receiver (bottom-left) is very short, practically a pulse, whereas the signal observed by the receiver located bottom-right has a smaller amplitude, a longer duration and a more complex frequency content. This simulation reveals, for a simple case, various physical processes that would be difficult to identify and interpret in a frequency representation. Prospects The case of the infinite plate can be used to validate the models and the numerical methods developed, but is not sufficiently representative of the complex mechanical systems on board ships. In order to approach as closely as possible to real structures, the case of fi nite plates, beams and their coupling is also being processed by methods similar to those described here. In addition, experiments are being carried out on immersed plates in an acoustic tank; these measurements will be used to validate the simulation models developed in this study. Fine-tuning and experimental validation of these models and tools will enable the DCNS Research teams to obtain finer characterisation of the vibration and radiation processes of structures subjected to pulsed excitations. Eventually, it will be possible to apply these methods to the design of ships for acoustic discretion in transient states. _REFERENCES [1] M. C. JUNGER, D. FIET: Sound, structures and Their Interaction. Second ed., Cambridge: The MIT Press, p. 235-250, 1986. [2] X. L. BAO, H. FRANKLIN, P. K. RAJU, H. UBERALL: “The splitting of dispersion curves for plates fluid-loaded on both sides”. Journal of the Acoustical Society of America. 102 (2), 1997. [3] R. SCHERRER, L. MAXIT, J.L. GUYADER, C. AUDOLY, M. BERTINIER: Analysis of the Sound Radiated by a Heavy Fluid Loaded Structure Excited by an Impulsive Force. Internoise 2013, Innsbruck, Austria, September 2013. 46 RESEARCH_2 PRODUCTIVITY AND COMPETITIVENESS OF INDUSTRIAL PROCESSES Industrial processes affect all product manufacturing activities and components, broadly encompassing the choice of materials, manufacturing techniques, assembly and appropriate testing methods. Such processes include the extensive use of virtual or augmented reality, i.e. reliance on modelling and visualisation methods to choose the best path forward, well upstream of the prototype phase. Moreover, the changes taking place in the way overall production chains are organised, in incorporating new manufacturing technologies, has prompted research into the “future factory” and the “extended factory” within many industrial sectors. RESEARCH_2 47 PRODUCTIVITY AND COMPETITIVENESS OF INDUSTRIAL PROCESSES Evaluation of friction stir welding (FSW) on high yield strength steels for shipbuilding AUTHORS: Guillaume Rückert, Myriam Chargy, François Cortial and François Jorez FSW process has been evaluated on three shipbuilding steels (DH36, S690QL and 80HLES steels) by fullypenetrated butt welds on 8 mm thick plates. Non destructive tests were carried out to highlight the presence of intrinsic defects known for the welding process (e.g. kissing bond). The validated inspection methodology (volume and surface testing) confirm the integrity of welds and the absence of geometrical defects for examinations and mechanical tests as part of a qualification procedure. Introduction The competitive market of naval vessels requires more efficient designs and manufacturing conditions in order to increase industrial and service performances. In particular the improving of the manufacturing conditions and the weight saving can be achieved by innovative design and processes. The use of performance materials such as high yield strength (HYS) steels in shipbuilding (hull, superstructures, floatboard, propellers…) is a solution which allows a significant weight saving by reducing the structures thicknesses. The manufacturing processes must accordingly adapt to the characteristics of these grades. In this context, the development of innovative welding processes must also be conducted ensuring the quality of assemblies: limitation of weld defects, limitation of distortions and improving the flatness of structures, improvement of hygiene and safety with regard to manufacturing processes, better controllability, improvement of serviceability and repairing. The fusion welding processes used in shipyards could require preheating, generally a filler metal, and lead currently to defects (lack of fusion, porosity, inclusions, cold cracking) or geometric defects on the surface (undercuts…), as well as distortions of structures. These defects undertake works, for repairing and/or finishing. The friction stir welding is an alternative method that uses solid-phase 48 welding without filler metal, which eliminates the disadvantages associated with metal fusion. While there are still many scientific and technological locks, the FSW is a mature technology for homogeneous and heterogeneous aluminium alloys [1] assembly. The main applications in the field of shipbuilding concern assembly extruded shapes to form stiff floors [2]. Studies [3, 4] have demonstrated the importance of this process in low alloy steels for a thickness of about 6 mm to 12 mm, for the grain refi nement in heat affected zone (HAZ), an enhancement of the metallurgical weldability with regard to arc welding processes, and producing no welding fumes, particularly containing hexavalent chromium. However, the results are derived from laboratory works on small specimens and do not demonstrate the industrial feasibility, taking into account the operating conditions in general and the use of tool with limited life (based tungsten-molybdenum) in particular. The recent developments of technologies coming from drilling tools (poly-crystalline boron nitrides – PCBN – doped W-Re), proposed by MegaStir lead to a technology breakthroughs [5]. Studies related to the use of these promising tools are limited to the work of the team behind the tool invention and limited mechanical performance materials [6]. To our knowledge, concerning welding of HYS steels required especially for shipbuilding, no study has been published to date. RESEARCH_2 PRODUCTIVITY AND COMPETITIVENESS OF INDUSTRIAL PROCESSES This study presents an evaluation of this innovative welding process for both civil and military shipbuilding applications. The main objective is to demonstrate that the FSW process technically and economically meets the specific requirements of the shipyards. Indeed, the introduction of this process in the workshop can be done only if its interest is demonstrated in terms of quality and manufacturing costs, controllability, health and safety of personnel dealing with proven and economically viable processes. It is therefore necessary to verify that the process is robust and repeatable, and that it can satisfy the required performances of the assemblies. The present study focused on three steels currently used for French naval application: two high yield strength steels (S690QL and 80HLES, a French grade close to HY100) and a DH36 steel, employed for hull and structures. The presentation provides a characterisation of the welded joints for optimised welding parameters and a complementary analysis to consider a future industrialisation of the FSW process in our shipyards. Experimental procedure Materials Three grades of steel (two high yield strength steels and one construction steel) used for hull and structure applications have been selected for this study according to their respective mechanical properties: • DH36 steel, a structural steel for hull according to Bureau Veritas NR 216 [7]; • S690QL steel, a structural steel according to standard EN 10025-6 [8]; • 80HLES steel, a French equivalent grade of HY-100 steel for hull and structures. The following tables give respectively guaranteed values for mechanical properties and chemical composition for each grade. Table 1. Guaranteed values for mechanical properties for DH36, S690QL and 80HLES steels Grade\Mech. properties YS* (MPa – ksi) DH36 S690QL 80HLES UTS** (MPa – ksi) El.*** % 355 – 51.5 490-620/71-90 21 690 – 100 770-940/112-136 14 700 – 101.5 780-900/113-130 14 * Yield strength. ** Ultimate tensile strength. *** Elongation. Table 2. Chemical composition for DH36, S690QL and 80HLES steels % max (pds) DH36 C Si Mn P S 0.18 0.50 1.60 0.035 0.035 N B – – Cr Cu Mo Nb Ti – – – – – V – Ni Zr – – S690QL 0.18 0.50 1.60 0.020 0.010 0.015 0.005 0.8 0.50 0.70 0.062 0.05 0.10 2.0 0.15 80HLES 0.15 0.25 0.50 0.01 0.01 – – 0.5 0.25 0.40 – – 0.09 4.8 – Welding process The butt welding tests (with full penetration) have been performed RESEARCH_2 with a Gantry machine on 1,500 x 150 x 8 mm plates on the three grades, joined together in the lengthwise (rolling direction). Friction stir-tool in PCBN based material comprises a threaded shoulder (25 mm diameter) prolonged by an 8 mm long threaded conical pin. Non-destructive tests (NDT) and mechanical tests Characterisation tests were carried out in order to compare the performance of the FSW process with arc-welding processes for which the expected performances are known. The introduction of welding processes in manufacturing requires a welding procedure qualification (WPQ), the rules and requirements are governed by codes or standards. For shipbuilding steels, the ISO 15614-1 standard [9], dealing with the qualification of welding procedures for arc welding of steels offers a level of minimum requirements. NDT required in this context are two complementary types of tests, i.e. volume testing and surface testing on 100% of the weld. These tests are always preceded by a visual inspection which, in the case of FSW, can identify unacceptable defects like partial penetration type, or lack of recovery, burrs… For the volume part, digital radiographic testing (X-rays) are preferred to ultrasonic testing because of the welded thickness. Indeed, for 8 mm thick welds, the resolution of the X-rays is much better. For surface measurements on the upper and lower surfaces of the weld, penetrant and magnetic particle testings are performed. The surface of the weld can cause background noise, especially for dye penetrant testing. It is often necessary to grind the surfaces before examinations. Destructive tests are performed to ensure that the weld does not cause significant change in the characteristics of the assembly. Transverse tensile tests (two specimens), transverse bend tests (two specimens per face – upper and lower –, bending angle 180°) impact tests (V-notch at –20 °C) and Vickers hardness (HV5) were carried out according to [9] and associated standards cited by this standard. For impact tests, specimens are machined on the nugget and the thermo-mechanically affected zone, i.e. TMAZ (three specimens per batch). For hardness tests, two rows of indentations were made at a depth of up to 2 mm below the upper and lower surfaces of the welded joint (at least 3 individual indentations in each area). In addition, complementary longitudinal tensile tests have been carried out on the nugget in order to characterise the stirred zone. Results NDT tests Specimens welded during preliminary testing dedicated to settings of welding parameters have been used to validate that radiographic testing is relevant to indicate the volume defects which may appear in FSW such as internal voids. Defects such as kissing bonds, which are not detected by visual examination 49 PRODUCTIVITY AND COMPETITIVENESS OF INDUSTRIAL PROCESSES (figure 1.a) or penetrant testing, are revealed through magnetic particles testing. Figure 1.b illustrates this observation by the presence of a black vertical line in the middle of the weld corresponding to a discontinuity of the magnetic field lines. It could be related to incoherence in the microstructure between the two edges to be welded, confirmed by micrographics on cross-section coupons (figure 1.c). The ACFM (alternating current field measurement) method is an alternative method to magnetic particles testing for kissing bond examination. This typical defect in FSW caused particular developments in innovative techniques for aluminum alloys. The magnetic nature of the steel allows the application of simple and proven methods to ensure the absence of kissing bond. (a) (b) Destructive tests In this study, the selected areas for mechanical tests on each grade of steels are free of NDT indications. Tensile tests on transverse specimens lead systematically to a failure on the base material. For DH36 steel, the two specimens exhibit respectively tensile strength values at 512 and 513 MPa (742 and 744 ksi) at 20 °C. The ultimate tensile strength reached on longitudinal tensile specimens (on nugget) is about 700-712 MPa (101-103 ksi). It confi rms a significant gap in mechanical behaviour between the base material and the nugget, explained by a severe thermo-mechanical cycle constituting a Widmanstätten microstructure, as shown in fi gure 2.a. Elongations values are not too much affected (22% and 24% in the nugget) and slightly higher than the requirements for the base material (i.e. 21%). We can expect such behaviour for welded S690QL and 80HLES steels, each exhibiting a very hard martensitic structure as shown in figure 2 for 80HLES steel and confirmed latter by hardness tests (table 4) for both steels. Bend tests are discriminating enough for kissing bond (when the upper surface is stretched); without defect, no indication upper than 3 mm, with respect to the standard [9], is noticeable. Macroscopic examinations on the three welded steels conclude to a defect free nugget. No internal or geometric defects were noted in observation areas. In retreating side, it could appear, as shown in fi gure 3 for 80HLES steel, a lack of thickness, but well below the acceptance criterion (i.e. 0.1 t from [9]). Burrs may appear locally (corresponding with areas exhibiting a lack of thickness). If the lack is acceptable, we just remove the burr by grinding. Impact tests results are given in table 3 for DH36 steel. Values in TMAZ are slightly lower in retreating side (two individual values around 35 J), probably associated with the conditions of the bond formation to the back of the stir tool. The base material exhibits values upper than 70 J. Values obtained on the friction stirred weld are systematically lower than those for base material. In arc weldi ng, the mean values of absorbed energy vary between 40 J and 100 J, according to the processes 50 (c) Figure 1. Kissing bond defect on DH36 steel; no indication by visual examination of the lower surface (a), indication (black vertical line in the middle of the weld) in magnetic particles testing on the lower surface (b), observation on cross-section micrograph (c). (a) (b) Figure 2. Micrograph on friction stirred nuggets; DH36 (a) and 80HLES (b). Figure 3. Macrograph of the friction stirred weld on 80HLES steel (advancing side on left side). and welding energies. The FSW process leads to results with similar magnitude than arc welding process, consistent with the elongation values previously identified and consistent with requirements of Bureau Veritas (i.e. 34 J at –20 °C) [7]. Table 3. Impact test (V-notch) results (individual and mean values) for DH36 steel at –20 °C Nugget Middle upper side 79 71 64 TMAZ Middle lower side 42 61 57 65 TMAZ/Nugget interface TMAZ/Nugget interface advancing side retreating side 78 50 45 49 51 35 59 33 42 Hardness tests results are shown in table 4 for the three welded steels. These results indicate a hardening in nugget and affected zones (HAZ/TMAZ) related to thermal and mechanical conditions imposed during friction stir welding, even more pronounced when the steel is susceptible to quenching (as shown in figure 2.b for 80HLES steel). However, large increases in hardness for S690QL and 80HLES steels are compatible with the requirements specified in the standard [7]. There is no significant difference in results between advancing and retreating side. The dispersion in results into the nugget, particularly between upper and lower rows, even in a same row (e.g. S690QL steel), can be explained by the coexistence of different mixed areas. RESEARCH_2 PRODUCTIVITY AND COMPETITIVENESS OF INDUSTRIAL PROCESSES Table 4. Hardness test results (HV5) in friction-stirred welds for DH36, S690QL and 80HLES steels Advancing side Grade Location BM Retreating side HAZ/TMAZ Nugget HAZ/TMAZ BM Upper row 178 172 178 187 202 229 214 202 211 220 212 221 192 181 186 Lower row 172 169 169 188 195 202 202 205 199 207 198 188 174 173 176 Upper row 279 277 268 374 374 386 393 416 413 399 386 378 277 290 294 Lower row 279 263 271 391 445 435 395 418 430 430 440 375 268 266 269 Upper row 277 275 276 410 441 433 421 410 431 440 438 433 275 275 280 Lower row 279 277 268 374 386 393 416 413 399 386 383 378 277 290 294 DH36 S690QL 80HLES Permitted maximum value [7] 380 450 450 Conclusions Preliminary investigations on three friction-stir welded steels were performed in order to confirm the applicability of the innovative FSW process for shipbuilding steels. The main results of this study are: – NDT tests (volume and surface testing) are proposed for the examination of the qualification coupons and manufacturing assemblies; in particular, the application of magnetic particles testing which is a common and simple to perform accurate and relevant test to the detection of defects such as kissing bond; – for DH36 steel, a complete character isation has been performed and concludes on a good behaviour of the welded joints, compatible with the minimum charges required in shipbuilding. – the first partial results on S690QL and 80HLES steels are encouraging; welds can be free from defects and hardness rows display results in compliance with the requirements. However, microscopic observations exhibit complex structures related to the thermo-mechanical cycle and the primary structures of steels that can lead, in the worst case, to an excessive fragility of the weld. It requires further investigation in progress to understand the formation mechanisms of the different areas and accordingly optimise the welding parameters. Acknowledgements This study is part of SIPSAN project supported by IRT Jules Verne (French Institute in Research and Technology in Advanced Manufacturing Technologies for Composite, Metallic and Hybrid Structures). Authors wish to associate the industrial and academic partners of this project, respectively DCNS, STX France, Bureau Veritas, GeM Institute (UMR CNRS-ECNNantes University 6183); IMN Institute (UMR CNRS-Nantes University 6502). Authors wish also to thank gratefully MegaStir (Provo, USA) for the achievement of friction stirred welds. RESEARCH_2 _REFERENCES [1] R. RAI, A. DE, H. K. D. H. BHADESHIA ET T. DEBROY. “Review: friction stir welding tools, Science and Technology of Welding and Joining”, Volume 16, no. 4, p. 325-342, February 2011. [2] K. J. COLLIGAN, M. T. SMITHERMAN, S. B. HOYLE. “Low-cost friction stir welding for littoral combat ship applications”. Naval Engineers Journal, March 2009. [3] T. J. LIENERT, W. L. STELLWAG, B. B. GRIMMETT, R. W WARKE. “Friction stir welding studies on mild steel”. Welding Journal, Volume 82, no. 1, p. 1-9, January 2003. [4] T. SHINODA, H. TAKEGAMI et al. Development of FSW Process for Steel Assemble to Shipbuilding and Off shore Structure. Proceedings of the 15th International Off shore and Polar Engineering Conference, Seoul, Korea, 19-24 June 2005. [5] J. DEFALCO, R. STEEL. “Friction stir process now welds steel pipe”. Welding Journal, Volume 88, no. 5, p. 44-48, May 2009. [6] C. C. TUTUM, J. H. HATTEL. “Numerical optimisation of friction stir welding: Review of future challenges”. Science and Technology of Welding and Joining, Volume 16, no. 4, p. 318, 2011. [7] NR 216 DT R06 E. “Rules on Materials and Welding for the Classification of Marine Units”, edited by Bureau Veritas, February 2013. [8] EN 10025-6 standard: Produits laminés à chaud en aciers de construction – Partie 5 : conditions techniques de livraison pour les aciers de construction à résistance améliorée à la corrosion atmosphérique (Hot rolled products of structural steels – Part 5: Technical delivery conditions for structural steels with improved atmospheric corrosion resistance), edited by Afnor, March 2005. [9] ISO 15614-1 standard: Descriptif et qualifi cation d’un mode opératoire de soudage pour les matériaux métalliques – Épreuve de qualifi cation d’un mode opératoire de soudage – Partie 1 : soudage à l’arc et aux gaz des aciers et soudage à l’arc des nickels et alliages de nickel (Specification and qualification of welding procedures for metallic materials – Welding procedure test – Part 1: Arc and gas welding of steels and arc welding of nickel and nickel alloys), edited by Afnor, February 2005. 51 PRODUCTIVITY AND COMPETITIVENESS OF INDUSTRIAL PROCESSES The development of TOFD at DCNS AUTHOR: Patrick Recolin One of the missions of DCNS Research is the development of new non-destructive testing techniques in the laboratory followed by their deployment on the Group’s production and maintenance sites. The growing use of the TOFD ultrasonic technique is an illustration of this. TOFD (Time of Flight Diffraction) is an ultrasonography technique using two transducers (a transmitter and a receiver) placed either side of a weld (figure 1). In combination with an encoding system, this system can rapidly obtain an image of the weld, which is comparable to a cross-section view of the weld (figure 2). This technique originated in the 1980s in the United Kingdom. It has been under study at DCNS since the 1990s and used for the first assessments on onboard nuclear steam supply systems from 1995 (figure 3). Many other developments have taken place since then, and the TOFD technique is now widely employed when necessary in the context of in-service monitoring (figure 4). For hulls, discussions with the UK MoD were initiated in 1995, in which one of the major topics was the replacement of radiography for hull weld inspection. The discussions resulted in various benchmarks for comparing and improving testing procedures and equipment. At the time, DCNS carried out various statistical studies [1] demonstrating the high level of effectiveness of the ultrasonic technique. These studies, backed by all the studies conducted during the same period in other fields (petrochemicals, offshore, etc.), led DCNS to propose the TOFD ultrasonic technique to French defence procurement agency DGA as an alternative to radiography for submarine hull weld inspection. Reduction of radiation sources was of course a major objective, but the productivity gain obtained by increased concurrent activity in the construction phase was the other important factor. Coordinated by DCNS Research, a team combining ultrasonics experts, engineering and the production site compiled a supporting documentation package which was finally accepted by the customer in 2012 (figure 5) [2]. Feedback is today considered 52 positive by the various entities involved: ultrasonic examination provides greater flexibility in the manufacturing process while ensuring detection sensitivity at least equal to that of radiography. A similar approach is now introduced for certain joints in the hulls of surface ships. The difficulty in this case is to ensure examination of the whole volume required on non-flush joints of low thickness. In parallel with this work, DCNS Research is continuing to develop TOFD, for example by incorporating multi-element technology [3]. The beam focusing and deflection capabilities enable new applications to be considered, such as examination of austenitic joints [4] and inspection of very thick joints (up to 200 mm). _REFERENCES [1] P. RECOLIN, S. RIVALIN, Y. LEZIN. Utilisation de TOFD pour le contrôle des joints de coque, COFREND 2001. [2] B. MARIE, P. RECOLIN, B. LEYRONAS, A. LEBIEZ. Remplacement de la gammagraphie par les ultrasons sur des soudures de coques, COFREND 2011. [3] S. RIVALIN, P. RECOLIN. Application de la technologie des capteurs ultrasonores multi-éléments au suivi de soudures de collecteur vapeur, COFREND 2005. [4] P. RECOLIN, S. RIVALIN, B. MARIE. Examen partiel en TOFD d’une soudure austénitique, COFREND 2011. RESEARCH_2 PRODUCTIVITY AND COMPETITIVENESS OF INDUSTRIAL PROCESSES – Control TOFD L45 L60 Figure 1. Typical TOFD configuration. Figure 2. TOFD image of a welded joint. Figure 3. First assessments. Figure 4. Inspection of the containment of the aircraft carrier. Figure 6. Use of multi-elements in TOFD. Figure 5. Inspection of a hull joint. RESEARCH_2 53 PRODUCTIVITY AND COMPETITIVENESS OF INDUSTRIAL PROCESSES Biofilms and corrosion of corrosion-resistant alloys in sea water AUTHORS: Émilie Malard, Hervé Gueuné, Jean-François Ghiglione, Christelle Caplat, Valérie Debout, Zakoua Guede and Chantal Compère Nickel-based alloys (e.g. Inconel 625 – base Ni, 22% Cr, 9% Mo, Nb) and austenitic-ferritic duplex steels have been used extensively for applications in sea water since the 1980s. These alloys are specifically resistant to all forms of uniform corrosion. Nevertheless, they may be sensitive to localised corrosion and in particular to crevice corrosion. Feedback from users, including DCNS, shows that this type of corrosion occurs in all the oceans and seas around the world and at any time of year. It is characterised by an increase in the free corrosion potential and the generation of a corrosion current between the “crevice” zone (seal, flange, pump, heat exchanger, etc.) and the metal surface outside the crevice, as shown in fi gure 1. Figure 1. Crevice corrosion of alloy 625 in natural sea water on a flange and at the metal/seal contact. These increases in potential and current have been eliminated by application of biocides. This provided evidence of the involvement of biofi lms formed in the systems, leading to a focusing of research on a more specific study. The objective of this project, supported by French defence procurement agency DGA through the REI exploratory research and innovation scheme, was to understand the role of the biofi lm in the crevice corrosion processes on Inconel 625. The project involved multidisciplinary characterisation of biofi lms formed on alloy Inconel 625 under controlled conditions leading to active or inactive biofi lms. The diversity of bacterial species in each biof i l m was a na lysed, as wel l as its phy toplanktonic diversity, its chemical composition (lipids, carbohydrates, amino acids, mineral and 54 STRUCTURAL ORGANISATION Bacterial abundance Microscopic Observation Area Coverage PHYTOPLANKTON DIVERSITY GLOBAL CHEMICAL COMPOSITION Lipids, glucids , Amino acids Inorganic elements Hydrogen peroxides ( H2O2 ) Multidisciplinary Characterization BACTERIAL DIVERSITY DNA fingerprinting Taxonomic Identification ENZYMATIC ACTIVITIES Lipase, glucosidase aminopeptidase Figure 2. Test plan. RESEARCH_2 PRODUCTIVITY AND COMPETITIVENESS OF INDUSTRIAL PROCESSES metallic constituents, oxidising molecules) and its spatial organisation, according to the test plan in fi gure 2. Difference observed on samples A: active and NA: Not active on the structure of the bacterial biofilm The fi rst stage of the study consisted in conditioning surfaces using an electrochemical method in order to obtain “active” biofi lms on samples showing a high cathode current and “inactive” biofi lms on samples showing an open-circuit potential of 300 mV/SCE but a low cathode current. Nearly 700 samples were conditioned for analysis. All the biofi lm characterisation analyses illustrated in fi gure 3 were perfor me d on s a mple s a t d i f fer e nt t i me s of ye a r (f r om November 2011 to end-2012). The multidisciplinary characterisation observed differences between “active” and “inactive” biofi lms, including: • higher concentration of microorganisms in the “active” biofi lms, associated with greater coverage of the surface of these active samples in the form of aggregates; • greater diversity of bacterial communities present and active in the “inactive” biofi lms; • fatty acid composition mainly comprising saturated fatty acids in the “inactive” biofi lms and mainly comprising monounsaturated fatty acids in the “active” biofi lms; • higher concentrations of some mineral or metallic constituents in the “active” biofi lms; • higher aminopeptidase enzyme activity in “active” biofi lms than in “inactive” biofi lms. Biofilm stained with DAPI (4’,6-diamidino-2-phenylindole: staining of the DNA) and observed with an Apotome fluorescence microscope. The results obtained for each type of analysis show the same trend. Statistical analysis was used to select specific criteria of “active” and “inactive” biofi lms. The overall results obtained suggest selection of metabolicallyactive bacterial populations, apparently related to the active character of the samples and/or leading to activation of the samples. Preliminary analysis of the 454 pyrosequencing results identified the predominant bacterial populations in the active biofi lms as related to the species Halomonas venusta and Halomonas sp. (class Gammaproteobacteria, order Oceanospirillales, family Halomonadaceae). Figure 3. Illustration of test results. RESEARCH_2 55 PRODUCTIVITY AND COMPETITIVENESS OF INDUSTRIAL PROCESSES Digital engineering and future shipyard AUTHOR: Alain Bovis The acceleration of worldwide competition both in shipbuilding and in the marine renewable energy sector necessitates, as the CORICAN (French council for the orientation of research and innovation in shipbuilding and related activities) has stressed in its roadmap, the development of new design and production methods intended to increase the competitiveness of this industrial sector. DCNS Research is heavily involved in this development in a very broad collaborative framework, in particular with the IRT Jules Verne, competitiveness clusters EMC2 and Systematic, and academic and industrial partners in all sectors of activity. Digital engineering For almost three centuries, DCNS has been designing, on a scientific basis, and building, in outstanding industrial facilities, warships recognised in each epoch as among the most complex systems produced by humans. They are complex in terms of the technologies used and the scientific tools necessary for their design calculations, complex in terms of the number of specifications with which they have to comply, around 150,000 today for a front-line ship, and complex in the number of equipment items and individual components (1,000,000). They are also complex in terms of the variety of trades and the number of contributors involved in their production (the system supplier and integrator accounts for only 20% to 40% of the added value of a ship). Last but by no means least is operational complexity in a difficult environment, the ocean, and in a context of widened and increasingly integrated cooperation (system of systems concept). Naval architecture thus appears as a precursor of what is today called “system engineering”. A warship is an assembly of subsystems and equipment units, but any naval architect knows from experience that it is not limited to the sum of these subsystems and equipment units. The integrated ship, comprising both the “carrier” component, the platform and its propulsion, and the “disposable load”, the combat system, has major properties, referred to as cross-functional performance, which cannot be obtained or analysed from the building blocks: this is the case for the overall hydrodynamic performance, the various signatures (acoustic, electromagnetic, 56 infrared) and, more generally, the overall operational capabilities that confer tactical superiority. The design of a complex ship is an iterative sequential process consisting of consecutive phases during which the defi nition of the ship, its dimensions, its constituents and their layout is refi ned until the detailed working drawings are produced. The fi nal detailed design must satisfy all the requirements of the owner. The design process extends over several years and does not end until after performance verification at sea on the first of class vessel, which is also always the prototype. However, very early in the process, contractual agreement on a price and a delivery date has to be reached between owner and prime contractor. This commitment is generally made on completion of the fi rst phase, referred to as the “feasibility study phase”, during which the main characteristics of the ship are established, the technical sticking points and specific development needs are identified, and the project costs and duration are estimated. Although ideally it might be hoped that at this stage 80% of the elements determining the defi nition of the ship be known, leaving 20% as estimates, in practice barely more than 50% of such elements are known at this stage with sufficient precision, both because of requirement drift and because the need for new technological developments opens up a wide range of risks for the prime contractor. Extensive use of digital engineering can significantly shorten the feasibility study phase, thereby allowing exploration of a much larger volume of the envelope of possible confi gurations. RESEARCH_2 PRODUCTIVITY AND COMPETITIVENESS OF INDUSTRIAL PROCESSES The convergence of simulations of different mathematical types, necessitating different representations of the same ship, is in itself a complex problem. One solution is to reproduce the different simulations in a single surrogate model with a very large number of parameters. The objective is to be able to explore the largest number of possible confi gurations in order to identify the optimum solution taking account of all the requirements and their priorities. This is a multiphysical, multiobjective optimisation approach. It provides the architect, the customer and the main future cocontractors with a platform for collaborative exchanges and speeds up the transition to the “digital mock-up” for detailed defi nition studies. At the level of complexity encountered, joint management with the customer of the countless requirements necessitates the development of what are called “metamodels”, or “architecture models”, of the type developed by the American Department of Defence and adopted by NATO in the field of software systems (DoDAF, NAF). These models are used to set boundaries on the envelope of possible variation of the design parameters. Today, with its academic and industrial partners, DCNS Research is focusing on the development of computing and algorithm resources leading to a design optimisation tool. The design itself, based on the digital defi nition of a confi guration, requires the development of simulation tools: physical simulation of cross-functional performance (hydrodynamic, acoustic, EMC, shock and impact response behaviour, vibration behaviour, etc.), functional simulation of systems (propulsion, weapon systems, etc.), behaviour simulations (human factors), statistical simulations (fatigue, ageing, obsolescence, etc.). DCNS Research is engaged in several programmes to develop such tools, for example the “digital tank”, new vibro-acoustic models and functional simulations of energy and propulsion systems. Future shipyard The construction of a ship is similar to a major public works project in that the object built is unique, or at best only a small number will be produced. And even in the latter case, the spreading out over time of the different units or the special requirements of different customers generally leads to significant changes between units of a given class. However, ship construction, in contrast to major public works projects, is carried out in a dedicated industrial facility, the shipyard. Meta-model NATO Architecture Framework Technical Requests Rules & Standards Environment Budget Technology Ship Concept Cost Estimate Technology Plan Models Functional Physical (Digital Mock-Up) Behavioral Probabilistic NDMS ASRU EHCLS* EHCLS* Decoy launcher DECOY'S LAUNCHERS Design Space Mutli-physics Simulation Optimization Strategy Visualisation Multi-objectives Analysis RESEARCH_2 57 PRODUCTIVITY AND COMPETITIVENESS OF INDUSTRIAL PROCESSES Dedicated, but not specific, as the same facility, substantial and capital-intensive, has to be usable for building ships of sometimes very different types and configurations. Unlike other industrial sectors, where the effect of scale is determining and each production line is specific, the production process in shipbuilding is determined mainly by the infrastructure. Technological developments (use of high-performance alloys and the introduction of composites, for example) and international competition mean that the process needs to be changed. The main objective is to improve the conditions under which the operations are carried out, in order to increase productivity and reduce the rework rate. For example, the transition to construction in pre-outfitted blocks or sections in the 1980s was a major change in the industrial process of shipbuilding. The emergence of new areas of development for the shipbuilding industry, such as marine renewable energy, is also demanding the development of new industrial approaches. tools in real time. A significant proportion of the construction costs is related to manufacturing margins (stiffening, allowances), rework and interruptions due to concurrent activity or to supply problems. While significantly increasing productivity, the new production technologies will also make operations safer. The development of these innovative technologies is one of the major objectives of DCNS Research. The discussions now under way in various sectors on “the factory of the future”, focusing on robotisation, large-scale introduction of information technologies and the energy transition, are also opening up new prospects of change for the shipbuilding industry. DCNS Research, with the IRT Jules Verne, is involved in the various national and European initiatives intended to develop the future production tools (advanced manufacturing). The future shipyard will be economical, with extensive use of renewable energy, optimising its consumption in the context of its energy distribution area. It will be clean, reducing its consumption of materials, thanks to lightening of structures, by reducing the quantities of consumables needed for assemblies and coatings and by recycling its own by-products and waste. It will be safer, by automating and by moving all operations that are difficult to carry out on board to open areas or workshops. The future shipyard will be “digitised”, i.e. it will make full use of all the power of IT resources to prepare the operations, simulating the strains induced by welding or by handling parts in order to better manage them, planning sequences in advance and using haptic virtual reality (which reproduces the forces) to optimise them and train the operators. The future shipyard will be “informed”. Tracing and real-time localisation of each part or equipment item thanks to the technologies of the “Internet of Things” (IoT) enables leaner flows, reduced stocks and accurate and timely delivery, while optimising task start dates. This instantaneous knowledge of the state of production is extended to all co-contractors (extended yard). This information is available to the operators, who have in situ access to the definition and production engineering data using touch-screen tablets. Augmented reality will enable operators to monitor the position and the settings of their 58 _REFERENCE A. BOVIS: The Virtual Ship. 4th Conference on Complex Systems Design and Management, Paris, 4-6 December 2013. RESEARCH_2 RESEARCH_2 59 OUR SCIENTIFIC PUBLICATIONS CONFERENCES – COMMUNICATIONS _ “DYPIC project” Artic Technology Conference 2014, Houston (United States), Kerkeni et al. _ N. Kamkar, F. Bridier, P. Bocher, P. Jedrzejowski “Water droplet erosion mechanisms in rolled Ti-6Al-4V” Int. Conference on Wear of Materials, Portland (United States), 14-18 April 2013 _ “Automatic heading control for DP in ice” MTS DP Conference 2013, Houston (United States), Kerkeni et al. _ “Capability plots of dynamic positioning in ice” _ P. Bocher, D. Mingardi, B. Larregain, F. Bridier, F. Dughiero “Simulation of fast induction surface heating and comparison with experimental full-field surface temperature measurements” ASME OMAE Conference 2013, Nantes (France), Kerkeni et al. Int. Conference on Heating by Electromagnetic Sources, Padua (Italy), 21-24 May 2013 ASME OMAE Conference 2013, Nantes (France), Metrikin et al. _ “Experimental and numerical investigation of dynamic positioning in level ice” _ “Comparison of control laws in open water and ice” _ F. Bridier, J.-C. Stinville, N. Vanderesse, P. Villechaise, P. Bocher “Measurement of microscopic strain localization and crystal rotation within metallurgical grains” Materials Structure & Micromechanics of Fracture, Brno (Czech Republic), 1-3 July 2013 _ F. Bridier, J.-C. Stinville, N. Vanderesse, M. Lagacé, P. Bocher “Measuring and comparing local strain field and crystal rotation at the microscopic scale” Microscopy & Microanalysis, Indianapolis (United States), 4-8 August 2013. ASME OMAE Conference 2013, Nantes (France), Kerkeni et al. _ Contribution J.-C. Poirier, DCNS Research/SIREHNA® “Tank testing of a new concept of floating offshore wind turbine” Proceedings of the ASME 2013 32nd International Conference on Ocean, Offshore and Arctic Engineering, OMAE 2013, Nantes (France), 9-14 June 2013 _ “Comités scientifiques : modélisation et simulation numérique du soudage” 11th AFM colloquium _ M. Andriamisandrata, F. Bridier “Insight on the mechanical behavior of copper bi-crystal using crystal plasticity” Materials Science & Technology Conference, Montreal (Canada), 27-31 October 2013 _ E. Liberge, M. Pomarède, E. Longatte, C. Leblond “Réduction de modèle en interaction fluide structure via une formulation POD multiphasique pour les écoulements en faisceaux de tubes” _ THERMEC 2013 conference Las Vegas (United States), 2-6 December 2013 _ M. Allart, G. Rückert, P. Paillard “Étude métallurgique du soudage par friction malaxage sur un acier à haute limite élastique destiné à la construction navale : le 80 HLES” SF2M annual conference, Lille (France), 29-31 October 2013 11th Mechanical Engineering Congress, Agadir (Morocco), 23-26 April 2013 _ G. Rückert, M. Chargy, F. Cortial, F. Jorez “Evaluation of FSW on high yield strength steels for shipbuilding” _ R. Fargère, P. Velex “Some experimental and simulation results on the dynamic behaviour of spur and helical geared transmissions with hydrodynamic journal bearings” THERMEC conference, Las Vegas (United States), 2-6 December 2013 5th VDI Wissensforum, München (Germany), 7-9 October 2013 _ C. Allery, A. Ammar, A. Dumon, A. Hamdouni, C. Leblond “Proper generalized decomposition for the resolution of incompressible flow” Eccomas – 2nd International Workshop on Reduced Basis, POD and PGD Model, Blois Castle (France), 3-6 November 2013 _ G. Rückert, N. Perry, S. Sire, S. Marya “Enhanced weld penetrations in GTA welding with activating fluxes – Case studies: plain carbon & stainless steels, titanium and aluminum” THERMEC conference, Las Vegas (United States), 2-6 December 2013 _ M. Allart, A. Benoit, P. Paillard, G. Rückert, M. Chargy “Metallurgical Study of Friction Stir Welded High Strength Steels for Shipbuilding” _ S. Lakovlev, J.-F. Sigrist, C. Leblond, H. Santos, C. T. Seaton, K. Williston “Efficient semi-analytical methodology for the pre-design analysis of the shock response of marine structures” _ “DCNS experience about metal/composite assemblies on board of navy ships” ASME 2013, 32nd International Conference on Ocean, Offshore and Arctic Engineering, Nantes (France), 9-14 June 2013 _ The 13th Euro-Japanese Symposium on Composite Materials THERMEC conference, Las Vegas (United States), 2-6 December 2013 Nantes (France), 4-6 November 2013 OUR SCIENTIFIC PUBLICATIONS _ “DCNS experience about composite patch onboard of navy ships” FP7-SST-2008-RTD-1 Sustainable Surface Transport, European CO-PATCH project (Composite Patch Repair for Marine and Civil Engineering Infrastructure Applications), Stakeholder Workshop, London (United Kingdom), 17 April 2013 _ F. Cortial, T. Giraud, P. Recolin, S. Drobysz “Soudage par faisceau d’électrons de l’acier inoxydable austénitique stabilisé au niobium X6CrNiMoNb 17.12.2” _ J.-C. Petiteau, E. Verron, R. Othman, P. Guéguan, H. Le Sourne, J.-F. Sigrist, G. Barras “Dynamic uniaxial extension of elastomers at constant true strain rate” Polymer Testing, vol. 32, p. 394-401, 2013 _ S. Lakovlev, C. Seaton, J.-F. Sigrist “Submerged circular cylindrical shell subjected to two consecutive shock waves: resonance-like phenomena” Revue générale nucléaire, Ed. SFEN, 2013, no. 2, March-April, p. 70-75 Journal of Fluids and Structures, vol. 42, p. 70-87, 2013 _ G. Rückert, M. Chargy, F. Cortial, F. Jorez “Evaluation of FSW on high yield strength steels for shipbuilding” _ Metrikin et al. “Numerical simulation of dynamic positioning in ice” MTS Journal, vol. 47, no. 2, March-April 2013, p. 14-30(17), THERMEC conference, Las Vegas (United States), 2-6 December 2013, Materials Science Forum, vol. 783-786, p. 1776-1781 _ X. Ledoux, F. Buy, A. Perron, E. Suzon, J. Farré, B. Marini, T. Guilbert, P. Wident, G. Texier, V. Vignal, F. Cortial, P. Petit “Kinetics of sigma phase precipitation in niobiumstabilized austenitic stainless steel and effect on the mechanical properties” THERMEC conference, Las Vegas (United States), 2-6 December 2013, Materials Science Forum, vol. 783-786, p. 848-853 _ A. Pagès, J.-J. Maisonneuve, X. Dal Santo, DCNS Research/SIREHNA® “Étude et modélisation des petits navires de pêche pour l’amélioration de leur comportement par mer forte” ATMA 2014 _ D. Laneuville “Polar versus cartesian velocity models for manoeuvering target tracking with IMM” _ J. Raymond (DCNS Research/SIREHNA®), P.-M. Guillouet (DCNS) “Bassin numérique et modèle libre générique : DCNS fait évoluer son processus de conception hydrodynamique des sous-marins” IEEE Aerospace conference 2013 ATMA 2014 _ T. Millot “Titanium in renewable energies DCNS’s OTECH issues” _ C. Drouet, N. Cellier, J. Raymond, D. Martigny OMAE 2013 “Sea state estimation based on ship motions measurements and data fusion” ITA 2013, International Titanium Association, Hamburg (Germany) DCNS Research/SIREHNA® PUBLICATIONS _ N. Kamkar, F. Bridier, P. Bocher, P. Jedrzejowski “Water droplet erosion mechanisms in rolled Ti-6Al-4V” _ S. Kerkeni, X. Dal Santo, L. Vilain DP Asia Conference “Improved cost efficiency of DP operations by enhanced thrust allocation strategy” DCNS Research/SIREHNA® Wear, vol. 301, p. 442-448, 2013 _ B. Larregain, N. Vanderesse, F. Bridier, P. Bocher “Method for accurate surface temperature measurements during fast induction heating” Journal of Materials Engineering and Performance, vol. 22, p. 1907-1913, 2013 _ B. Chassignole (EDF R&D), P. Recolin (DCNS), N. Leymarie (CEA), D. Elbaz (Extende), P. Guy (INSA Lyon), G. Corneloup et C. Gueudre (Aix-Marseille universities) BINDT “3D modelling of ultrasonic testing of austenitic welds” _ R. Fargère, P. Velex “Influence of clearances and thermal effects on the dynamic behaviour of gear-hydrodynamic journal bearing systems” _ L. Rouleau, J.-F. Deü, A. Legay, F. Le Lay “Application of Kramers-Kronig relations to timetemperature superposition for viscoelastic materials” ASME, Journal of Vibration and Acoustics, 135(6), 061014-1, 2013 Mechanics of Materials, vol. 65, p. 66-75, 2013 TO LEARN MORE www.dcnsgroup.com Join us on www.BlooPlanet.com It’s time for a sea change in your social networking. DCNS 40-42, RUE DU DOCTEUR FINLAY F-75732 PARIS CEDEX 15 TEL.: + 33 (0)1 40 59 50 00 Member of French Cluster Maritime Communications Department – November 2014 – Design and production: – Photo credits: DCNS, Getty Images. FIND US AT