Development of a robust adaptive mesh refinement method for Large Eddy Simulations: application to hydrogen deflagrations

A significant reduction of greenhouse gases emissions is needed to keep global warming at an acceptable level in the next decades. In particular, the decarbonization of the energy and transport sectors is necessary and the use of hydrogen is a plausible solution as its consumption, through combustion processes or in fuel cells, is carbon-free. It may also provide flexibility to systems based on intermittent resources. However, hydrogen is a volatile and highly flammable compound. Its storage and use are associated with high risks of explosions which must be adequately addressed. As experiments on hydrogen deflagrations are difficult and expensive to set-up, it is interesting to use numerical tools to assess the hazards related to an industrial hydrogen installation. These numerical models must be able to predict flame propagation in an environment which is either inherently turbulent, or where turbulence is generated by the interactions between the flame and obstacles.
The simulation of turbulent flows requires the resolution of a wide range of physical scales, from the smallest (Kolmogorov) scales to the largest (integral) scales. Adequate simulation techniques, in particular Large Eddy Simulation (LES), enable the numerical resolution of complex systems by modeling all or part of the turbulent scales. Nevertheless, many issues remain to be overcome, especially in the case of large-scale and highly unsteady systems, as is the case for explosions. Indeed, the definition of numerical meshes fine enough to solve all the turbulent structures leads to prohibitive computation times. The turbulence is often generated by the flame, which propagates rapidly in the domain. One solution consists in dynamically refining the mesh using the Adaptive Mesh Refinement (AMR) technique, thus making optimal use of the available resources, and allowing the simulation of large domains. Nevertheless, the use of AMR requires the definition of an efficient activation criterion. While several attempts have been published, the definition of a universal AMR activation criterion based on physical quantities related to turbulence is an open problem. 

The objective of this thesis is to propose an efficient AMR activation criterion to solve turbulence in the LES of large-scale systems. The PhD student in expected to:
i)    Propose one or more AMR activation criteria, based on the literature and original ideas
ii)    Test the retained strategies on academic cases of increasing complexity
iii)    Validate the retained model on an industrial case in the field of hydrogen security. The case will be provided by INERIS, which is a French institute specialized in industrial security.

Keywords : Large Eddy Simulation, Hydrogen, Deflagration, Adaptive Mesh Refinement, Turbulence

• Academic supervisor    Pr Pierre SAGAUT, M2P2 laboratory, Aix-Marseille université
• Doctoral School    École Doctorale « Sciences pour l’ingénieur » ED353 (Aix-Marseille université) - https://ecole-doctorale-353.univ-amu.fr/ 
• IFPEN supervisor    Dr MEHL Cédric, Numerical modeling of energy systems department, cedric.mehl@ifpen.fr
• PhD location    IFP Energies nouvelles, Rueil-Malmaison, France
• Duration and start date    3 years, starting in fourth quarter 2022
• Employer    IFPEN
• Academic requirements    University Master degree involving CFD, physics and/or numerical modelling
• Language requirements    Fluency in French or English, willingness to learn French
• Other requirements    Programming skills (Python, C++)