Summary Data from the Sixth AIAA Computational Fluid Dynamics Drag Prediction Workshop: Code Verification

Results from the Sixth AIAA CFD Drag Prediction Workshop (DPW-VI), Case 1 Code Verification are presented. This test case is for the turbulent flow over a 2D NACA 0012 airfoil using Reynolds-Averaged Navier-Stokes (RANS) turbulence models. A numerical benchmark solution is available for the standard Spalart-Allmaras (SA) turbulence model that can be used for code verification purposes, i.e., to verify that the numerical algorithms employed are consistent and that there are no programming mistakes in the software. For the Case 1 code verification study, there were 31 data submissions from 16 teams: 23 with the SA model (using various versions), 4 with the k-omega SST model (two variants), and one each with k-kl, k-epsilon, an explicit algebraic Reynolds stress model, and the lattice Boltzmann method (LBM) with very large eddy simulation (VLES). Various grid types were employed including structured, unstructured, Cartesian, and adapted grids. The benchmark numerical solution was deemed to be the correct solution for the 21 submissions with the standard SA model, the SA-noft2 variant (without the ft2 term), and the SA-neg variant (designed to avoid nonphysical transient states in discrete settings). While many of these 21 submissions did demonstrate first-order convergence on the finer meshes, others showed either nonconvergent solutions in terms of the aerodynamic forces and moments or converged to the wrong answer. Results for this case highlight the continuing need for rigorous code verification to be conducted as a prerequisite for design, model validation, and analysis studies