A review and comparison of error estimators for anisotropic mesh adaptation for flow simulations

Automated mesh adaptation is known to be an efficient way control discretization errors in Computational Fluid Dynamics (CFD) simulations. It offers the added advantage that user only needs have a minimal expertise generating appropriate meshes, which biggest bottleneck current CFD workflows. For anisotropic, or highly stretched, simplex (triangles 2D and tetrahedra 3D) concept of metric field, describes continuous mesh, convenient describe shape size elements. Error estimators can derived optimized on this get optimal used generate corresponding using metric-conforming generator. The work aims review various error fields available for anisotropic adaptation, compare their convergence behavior flow problems. Mathematical formulations are also briefly described here many those methods understand differences, including underlying local global analytic optimization. include both solution-based ones interpolation scalar field Lq norm, adjoint-based particular output functional. All considered implemented NASA’s open-source grid mechanics package, refine, FUN3D-SFE, finite-element solver, primal adjoint solutions. Mesh convergences quantities compared benchmark problems, inviscid laminar flows around ONERA M6 wing, supersonic over low-boom aircraft, RANS turbulent high-lift configuration.