Leveraging unstructured grids for direct numerical simulations of wall turbulence
arXiv:2605.01015
Abstract
Towards computational cost saving for direct numerical simulations (DNSs) of wall turbulence, we formulate an unstructured grid-generation framework, termed -grid, where the wall-normal () and spanwise () grid sizes are proportional to the local Kolmogorov scale . The framework consists of an inner layer, with a thickness viscous units, with viscous-scaled grid sizes similar to a conventional DNS grid: over a smooth wall, and over uneven surfaces, where is the smallest surface wavelength. Above the inner layer, . We test -grid with finite volume and spectral element solvers, and conduct DNSs of turbulent channel flows and boundary layers over smooth wall and various streamwise-aligned riblets, up to friction Reynolds number . We assess the accuracy of -grid against the conventional Cartesian grids, through comparison with the reference DNS and experimental data. Results from -grid and the Cartesian grids differ by less than , in terms of turbulence statistics up to second-order, and the energy spectra. For turbulent channel flows with , the number of grid points with -grid () scales over a smooth wall, and over riblets, whereas the number of grid points with a Cartesian grid and hyperbolic-tangent -grid () scales . By , over a smooth wall, and over typical drag-reducing riblets, with viscous-scaled spacing .