Cosimo Tarsia Morisco, INRIA Saclay

Abstract:

Edge-based method is known to be one of the most robust and accurate discretization algorithms for general tetrahedral meshes. Actually, second-order edge-based methods have proven to be the most suitable schemes to handle  accurate and robust numerical simulations on anisotropic adapted meshes. Accurate boundary conditions are mandatory to preserve accuracy. Indeed, weak boundary conditions are known to provide more accurate solutions in adjoint-based mesh adaptation, in comparison with a strong boundary condition procedures. Inviscid weak boundary conditions preserving second and third-order accuracy are already known in literature. However, there is no general equivalent formula for viscous simulations. In the first part of this talk, I will present a new weak boundary procedure for P^1 Galerkin discretization, with no tunable parameters and preserving second-roder accuracy. In the perspective to run simulations even more accurate, the design of upwind third-order schemes is the next step. To handle viscous simulations we need a hyperbolic Navier-Stokes (HNS) formulation. In the second part of this talk, we will explain the different steps required to achieve third order accuracy for adaptative simulations, the progress made so far and interesting insights for future developments. 

Bio: 

Currently I am a postdoctoral researcher in the GammaO team at INRIA.
I earned a Ph.D. in Fluid Dynamics at the ENSAM ParisTech in 2020. During my Ph.D., I worked on linear stability analysis and numerical methods for turbulent compressible flows, involving RANS, URANS and DDES modeling. I then moved to INRIA and joined the GammaO team, which aims at certifying the numerical results of the entire simulation pipeline and providing high-fidelity numerical solutions. Since then, I work on numerical schemes for anisotropic meshes, RANS and transition modeling. Later, I spent a brief period abroad as a visiting researcher at the National Institute of Aerospace (NIA). There, I established professional networks with NIA and NASA LaRC researchers. During this period, I worked on high-order edge-based algorithms. My area of expertise is turbulence modeling and the development of numerical schemes for unstructured and mixed-element meshes.