Date: 7 June 2018 Location: EERA new offices Rue de Namur Brussels Belgium

Size: px

Start display at page:

Download "Date: 7 June 2018 Location: EERA new offices Rue de Namur Brussels Belgium"

Joel Hunt
5 years ago
Views:

1 Joint EoCoE and AMPEA-EERA workshop Accelerating the energy transition: challenges in materials design enabled by recent advancements in high performance computing Date: 7 June 2018 Location: EERA new offices Rue de Namur Brussels Belgium Registration and Welcome Coffee 09:00 Introduction of the workshop Frédéric Chandezon(CEA, AMPEA) and Massimo Celino (ENEA, EoCoE) 09: Presentation of EERA Adel El Gammal (EERA) Session 1: 09:20 S1 Accelerating the energy transition (Part 1): materials meet supercomputers Chair : F. Chandezon 9: EoCoE: the Energy oriented Centre of Excellence Edouard Audit (MdlS, EoCoE) 9: AMPEA: the EERA Joint Programme and Energy Materials and Processes Frédéric Chandezon (CEA, AMPEA) 10: EoCoE: the Materials for Energy pillar Massimo Celino (ENEA, EoCoE) Thermoelectric materials for energy harvesting 10: new modelling tools with predictive power Ole Martin Løvvik (SINTEF) 10:30 End of session 1 and coffee break 11:00 S2 Session 2 Accelerating the energy transition (Part 2): the role of Chair : M. Celino supercomputers 11: : EoCoE performance benchmarking methodology for renewwable energy applications Multiscale modelling of materials, processes and devices in AMPEA 11: EoCoE support activity on Metalwalls Paul Gibbon (JFC, EoCoE) David Lacroix (Lorraine University, AMPEA) Abel Marin-Laflèche (MdlS, EoCoE) Massive MD simulations of an ionic liquid in 12: confinement Stefano Mossa (CEA Grenoble) Challenges on battery modeling: which tools and 12: methodologies to strengthen the links between Marion Chandesris (CEA, EoCoE) atomistic and device simulations 12:40 End of session 2 and lunch break 14:00 S3 Session 3 Accelerating the energy transition (Part 3): the role of Chair : M. Fabrizio materials science EMIRI: the Energy Materials Industrials research 14: Initiative Simon Perraud (EMIRI) 14: European Energy Materials Development platform Ihssen Holger (HGF, EERA) Computational characterization of passivating contacts 14: Urs Aeberhard (JFC, EoCoE) for silicon solar cells 15: Computational Materials Design Tomi Suhonen (VTT, AMPEA) 15: Catalytic Activity of Defective Nanomaterials towards Oxygen Reduction Reaction: An ab initio study Gracia-Espino Eduardo (Umeå University, AMPEA)

2 15:40 End of session 3 and coffee break Session 4 16:00 S4 Accelerating the energy transition (Part 4): strategies to tackle complexity Mission Innovation. Clean Energy Materials Innovation Challenge 6: to accelerate the exploration, discovery and 16: use of new high performance, low-cost clean energy materials 16: Round table and general discussion: New challenges in materials for energy transition: how high performance computing can help to efficient design? 17:00 End of session 4 and end of the event Chair : D. Lacroix, M. Ramalho Monica Fabrizio (CNR, AMPEA) F. Chandezon, E. Audit, H. Ihssen, O. M. Løvvik, S. Perraud

3 Round table notes and workshop synthesis The AMPEA/EoCoE workshop was held in the framework of Accelerating energy transition through the design of new innovative material enable by high performance computing. First, presentations dealing with general description of AMPEA and EoCoE environments, objectives and main challenges in terms of computing, numerical modeling and exascale perspective gave an overview of projects already started and future realization. At the same time several technical and scientific based presentations were given in order to shed light on specific computing and simulation issues applied to major challenges in innovative material modeling which might enable and foster energy transition. Among them, there were: new catalysts, improved solar cell design, ionic liquid for battery, new thermoelectric materials All, these topics demonstrate the crucial importance of numerical methods and tools to tackle the complexity underlying the research of new and efficient materials for energy transition. The workshop ends on two points; first an overview of Mission Innovation Challenge 6 (MI-C6) on Clean Energy Materials was given by Monica Fabrizio (CNR, Italy). The aim of this presentation was to give a global picture of world wide policies that aim at make a new revolution in material sciences, in the continuation of the material genome project initiated in US, with a goal of self-supporting research institution able to design, elaborate, characterize and tailor in a single place new materials identified through machine learning and artificial intelligence. Then, a general discussion allowed all participants to discuss on the general context and opportunities to reinforce AMPEA and EoCoE collaborations in the future around common interests. Discussions and presentations all along the day raised several issues, which are listed hereafter, they are not ordered in terms of importance. Material simulation is now often based on small-scale approaches, i.e. considering atoms and electron interactions. These so-called DFT (Density Functional Theory), Quantum Monte Carlo, Molecular Dynamics, etc. are very promising and give a very accurate understanding of phenomena that rules material properties (e.g. organization of ionic liquid in pores, amorphization thin layers on solar cells, impact of defects and substitutions in the elaboration of new catalysts, etc.). Yet, all these methods need important (considerable) computing resources that are not always available or remains expensive. o Even if those methods are mostly matures, there is still work to do go on larger systems, interest of exascale? o In a cross-fertilizing approach, what will be the meaningful experimental/characterization techniques suitable to put more reality in models? Databases and material screening. Among big challenges in new material discovery, numerical screening of existing material databases is very promising and is a pillar of Mission Innovation Challenge 6, through Inverse Design. In this approach well-chosen descriptors can help to identify the right material with the expected properties. On this point question are also risen. o What kind of databases to use (calculation based ones or experimental ones? o How to improve access to database to the existing modeling tools?

4 o Impact of descriptor in database analysis, how improve new material seeking? Machine learning and artificial intelligence (AI). If AI can appear as a complicated issue to manage in term of new materials exploration, especially in a whole integrated process (simulation, elaboration, characterization & feedback) suggested by MI-C6, machine learning is a key issue to accelerate computational design of material on integrated platform. o What would be energy material sector where such approaches could be interesting to develop (enough feedback, data to develop appropriate algorithms)? o Issue of trial/error approach. In order to succeed in the development of a machine learning process, not only experimental/numerical techniques that succeed are needed. Errors and misleading attempts are also useful to develop a critic reasoning. New knowledge must come back to the scientific community and stakeholders by reporting problems, coding new types of analyses and applications, and proposing new but reliable materials for computation. Multiphysical and multiscale modeling. Both issues are well known, yet it remains still vey difficult to efficiently combine different simulation tools that operates on very different length and/or time scales (e.g. steel micro organization and risk of failure in operation condition or after years of corrosion etc.). o Integrated platform approach might be a way to interconnect tools, libraries, solvers, etc. as it is currently done in some macro-scale simulation tools. Does it make sense to extend such approach to micro/nano modeling techniques? Exascale. Exascale is at the core of several projects and Centers of Competences within Europe. Presently, the technology (hardware) seems not to have reach enough maturity to induce a new way of design materials, nevertheless it is undisputable that in some years such technology will be available. The question is to know on which part of material design we shall put effort to be ready (code compatibility). Similarly, in which material science domains do this technology can lead to real breakthroughs? Clear objectives are to be defined to succeed in this evolution.

Registered partickipants Aeberhard Urs Forschungszentrum Jülich Germany u.aeberhard@fz-juelich.de Anciaux Sedrakian Ani IFPEN France ani.anciaux-sedrakian@ifpen.fr Audit Edouard MdlS France edouard.

5 Registered partickipants Aeberhard Urs Forschungszentrum Jülich Germany Anciaux Sedrakian Ani IFPEN France Audit Edouard MdlS France Bellosta von Colbe Jose HZG Germany Besmehn Astrid Forschungszentrum Jülich Germany Celino Massimo ENEA Italy Chandesris Marion CEA France Chandezon Frédéric CEA France Fabrizio Monica CNR ICMATE Italy Fontaine Marie-Laure SINTEF Industry Norway Gracia Eduardo Umeå University Sweden Gibbon Paul Forschungszentrum Jülich Germany Ihssen Holger HGF Germany Lacroix David Université de Lorraine France Lemoine Fabrice Université de Lorraine France Løvvik Ole Martin SINTEF Industry Norway Marin-Laflèche Abel CNRS France Meulenberg Wilhelm A. Forschungszentrum Jülich Germany Mossa Stefano CEA Grenoble France Navas Marta CIEMAT SPAIN Perraud Simon CEA France Pohjanne Pekka VTT Finland Ramalho Maria Forschungszentrum Jülich Germany Roeb Martin DLR Germany Suhonen Tomi VTT Finland

AMPEA: The Joint Programme on Advanced Materials and Processes for Energy Applications

AMPEA: The Joint Programme on Advanced Materials and Processes for Energy Applications JP coordinator: Frédéric Chandezon (CEA) JP deputy coordinator : Monica Fabrizio (CNR) Présentation : Engin Molva