Nudging-based data assimilation of the turbulent flow around a square cylinder

We investigate the prediction of turbulent flow around a canonical square cylinder at Re= 22000 solving unsteady Reynolds-averaged Navier-Stokes (URANS) equations. The limitations URANS modelling are overcome through application data assimilation technique involving feedback control term, allowing to drive predictions towards reference data. Using regularly-spaced temporally-resolved that extracted from DNS, we assess abilities methodology in improving predictions. While low-frequency oscillations wake flow, also demonstrate ability recover high-frequency dynamics Kelvin-Helmholtz instabilities. influence spatial resolution is systematically investigated. which required improve structures interest related their wavelength. spacing order one length between points, already observe synchronisation vortex shedding leads significant decrease temporal and spectral errors as computed by proper orthogonal decomposition. Furthermore,improved accuracy terms mean achieved. wavelength vortices, even those shear layer absent baseline simulation.