Comparison of low and high-order data prolongation methods for two-dimensional flow

Turbulent flows prescribed by Navier-Stokes equations are complex and highly nonlinear to solve numerically, and hence, large computing resources are required for accurate turbulent simulations. There are several strategies to reduce the computational costs of high-fidelity simulations of turbulent flows, and one of the widely used methods is Adaptive Mesh Refinement (AMR). AMR reduces the computational cost by dynamically refining the mesh resolution in the physics-rich regions to capture chaotic/turbulent structures, such as vortices and eddies while keeping the mesh in the rest of the domain at a relatively coarse level. However, during post-processing, it is beneficial to have solutions on a uniform grid, for example, to perform the Fourier Transform of different fluid fields. Thus, a data prolongation procedure of solutions obtained with an AMR grid onto a uniform grid is required. In this study, we explore the use of low- and high-order Lagrangian interpolation methods for projecting 2D Direct Numerical Simulation (DNS) solutions of compressible interfacial Rayleigh-Taylor Instability onto uniform grids.

The effects and accuracy of using different interpolation methods are compared.