CFD simulations of cyclist aerodynamics: Impact of computational parameters

Computational Fluid Dynamics (CFD) allows researchers and practitioners to analyze cyclist aerodynamics and identify areas for improvement. Despite the numerous CFD simulations of cyclist aerodynamics in the scientific literature, the extent to which user choices in the large number of computational parameters affect the simulation results remains largely unexplored. This paper aims to establish a set of best practice guidelines for CFD simulations of an isolated cyclist in time trial position through a systematic and comprehensive sensitivity analysis. It includes the computational grid in terms of surface, near-wall, and far-field volume grids and the turbulence modeling. The study reveals a high sensitivity of the computed drag area to the surface grid resolution and y+ value, while the impact of the growth rate and the grid resolution in the wake is relatively smaller. The results emphasize the importance of complete reporting of grid characteristics and the need for grid-sensitivity analyses, and provide prioritization of the key parameters for such analyses. Satisfactory agreement with wind tunnel measurements is achieved using Scale-adaptive simulations (SAS) and steady RANS with the Transition SST (T-SST) or SST k-omega turbulence model for closure. This work intends to contribute to accurate and reliable CFD simulations of cycling aerodynamics.