Roughness effect modelling for wall resolved RANS – Comparison of methods for marine hydrodynamics

Journal article, 2022

This paper deals with several aspects of surface roughness modelling in RANS codes applied to full-scale ship simulations. To select a method that is suitable for wall-resolved RANS solvers and gives reliable results at high Reynolds numbers, five different roughness models are compared. A grid uncertainty analysis is performed and the sensitivity to the grid resolution close to the wall (y+) is investigated. The results are compared to extrapolated results of experiments carried out with rough plates with various heights and roughness types. A correlation factor between the Average Hull Roughness and the equivalent sand roughness height is investigated, and a value of five is deemed the most suitable. The work suggests that the Aupoix-Colebrook roughness model gives the best results for full-scale ship simulations, at least with the current code, and that the near-wall grid resolution required for smooth surfaces can be applied also for the rough case.