UROP Project

Anisotropic mesh adaptation for turbulent flows

Adaptive meshing has been considered as a significant bottleneck in the advancement of Computational Fluid Dynamics (CFD). An adapted mesh is critical, in particular, for flows that exhibit strong anisotropic features such as singularities and boundary layers where the phenomena also span multiple length scales. Here, having an optimal mesh can lead to considerable improvements in computational cost. In this context, we have previously used an adaptation technique using the Riemannian metric field based on a continuous mesh model. The technique can construct anisotropic meshes to minimize the error at a fixed number of elements. These techniques have to be investigated for various turbulent aerodynamic cases and the results have to be verified and validated against previous studies. Construction of these meshes could lead to significant gains in automatic optimal mesh generation without the intervention of an expert CFD practitioner.

Task

The task for this project will mainly involve four steps: 1. Setup a few benchmark CFD simulations 2. Run the CFD solver and the in-house mesh adaptation tool 3. Post-process the results from CFD solver and previous studies (whose data is already available) 4. Compare the results and document the best practices for various test cases. The main task will be about the documenting best practices for the work.

Requirements

The student will ideally have: 1. Interest in fluid dynamics 2. Basic background in CFD 3. Basic knowledge in applied mathematics