Context :
The need to decarbonise transport is leading designers to seek ways of reducing the weight and improving the efficiency of current transportation modes (whether by air, sea or land). Additive manufacturing approaches will be one of the solutions for achieving these objectives. They give more freedom to the designers, enabling them to significantly improve the performance of mechanical parts in terms of stiffness/mass trade-offs. In particular, this process makes it possible to produce structures with a thin surface and an architectured volume. Selecting the properties of this architecture in the right way enables the mass and stiffness of the structure to be controlled locally.However, the procedures and numerical tools used to define the optimum architectures, then dimension and characterise the durability of such structures are extremely costly (specifically in the case where there is no scale separation between the scale of the meso-structure and that of the part).
Job description :
You will be part of GeM's MECNUM thematic research unit (https://gem.ec-nantes.fr/utr-mecnum-2/). Based at Ecole Centrale de Nantes, you will also work in close collaboration with researchers from the « Faculty des Sciences et Techniques » of Nantes University.In this context, you will develop a set of computational tools that will make it possible to dimension meso-architectured structures. These tools will be validated through laboratory-scale tests.
Two major topics will be studied:
I Robust and accurate simulation of parts with very complex geometries (due to the meso-structure), parts in which thin and massive zones coexis;
II Effectively capture phenomena coming from the different scales of the structure.To meet these challenges, innovative numeric approaches will be implemented. In order to avoid meshing problems, we will use a fictitious domain approach [1,2]. The presence of of slender zones leads us to use high-order approaches [3].
The meso and macro scales being coupled, we will consider domain decomposition approaches that allow these two scales to be resolved concurrently [4]. We will then study the coupling between these methods and fictitious domain approaches.Finally, model structures will be built in collaboration with the RAPMAN TRU at GeM (https://gem.ec-nantes.fr/utr-rapman-2/) to validate the numerical approaches implemented.
Profil recherché :
Compétences :
Développement d’approches éléments finis, Programmation (C++ / python)
Maîtrise orale et écrite de la langue anglaise
Qualités requises :
Rigueur, Sens de l’initiative, Capacité d’organisation, AutonomieRequired qualification:
You hold a computational mechanics or applied mathematics phd degree, ideally with experience of non-standard finite element development.Contract description :
- 12 months contract
- Full time, based in Nantes
- Gross salary : 2655€
- Remote work possible under the terms of our protocol
- Starting between January and May 2024
Send your application to Grégory Legrain (gregory.legrain@ec-nantes.fr)
[1] Grégory Legrain, Nicolas Chevaugeon et Kristell Dréau. “High Order X-FEM and Levelsets for Complex Microstructures : Uncoupling Geometry and Approximation”. In : Computer Methods in Applied Mechanics and Engineering 241-244 (oct. 2012), p. 172-189. doi : 10.1016/j.cma.2012.06.001.
[2] Jamshid Parvizian, Alexander Düster et Ernst Rank. “Finite Cell Method”. In : Computational Mechanics 41.1 (2007), p. 121-133. doi : 10.1007/s00466-007-0173-y.
[3] G. Legrain et N. Moës. “Adaptive Anisotropic Integration Scheme for High-Order Fictitious Domain Methods : Application to Thin Structures”. In : International Journal for Numerical Methods in Engineering (2018). doi : 10.1002/nme.5769.
[4] Mickaël Duval et al. “Non-Intrusive Coupling : Recent Advances and Scalable Nonlinear Domain Decomposition”. In : Archives of Computational Methods in Engineering 23.1 (mar. 2016), p. 17-38. doi : 10/gf25qv.
