Understanding the aerodynamic interaction between and fluid and a body in realtive motion is important in many engineering contexts. This project will explore the efficacy and accuracy of a computational approach to this problem when anisotropic mesh adaptation is adopted. The quality of the mesh in proximity of the wall is critical to ensure a proper resolution of the velocity and temperature gradients within the boundary layer. Aiming for a reasonable resolution of the boundary layer profile based on the RANS approach over unstructured meshes made by tertahedra and iteratively optimised via anisotropic mesh optimsation could enable a significant speed up in the mesh generation process and consequently will allow for faster predictions of aerodynamic performance. Key objectives:

Streamline (via bash or python scripting) the mesh adaptation process using INRIA AMG library (https://pyamg.saclay.inria.fr) and its interaction with the CFD solver SU2 (https://su2code.github.io); Perform a series of adaptive simulations over airfoils and possibly wings; Assess the robustness and accuracy of the approach by means of a comparative study between the adpative results and reference data in terms of pressure, shear stress and heat fluxes.