Numerical investigation of marine propeller underwater radiated noise using acoustic analogy part 1: the influence of grid resolution

This study investigates the effects of grid resolution on hydroacoustic performance of the benchmark INSEAN E779A propeller operated in uniform flow, open water and non-cavitating conditions. In the numerical calculations, an incompressible hydrodynamic solver together with the porous FW-H (Ffowcs-Williams Hawkings) equation is used to predict the propeller URN (Underwater Radiated Noise). The first aim within this study is to explore the sensitivity of the grid resolution on the prediction of propeller hydroacoustic performance. Furthermore, amongst the contribution of nonlinear noise sources on overall acoustic pressure, the role of the tip vortex is believed to be dominant under non-cavitating conditions. The inadequate extension of the tip vortex is one of the drawbacks in the RANS (Reynolds-averaged Navier-Stokes) solver for accurate prediction of propeller URN, especially at the receivers located in the propeller's slipstream. Thus, the second aim within this study is to examine the contribution of tip vortex on overall acoustic pressure. In order to visualise the numerical noise and determine the realistic extension of the vortex distributions in the propeller's slipstream, the time derivative of the pressure technique is proposed in this study. The results indicate that insufficient grid resolution in the numerical simulations causes unphysical numerical noise which is attributed to the sliding interfaces, and it leads to contamination of the overall acoustic pressures. Moreover, an increase in grid resolution reduces numerical diffusion in the RANS solver, allowing for an extended tip vortex distribution. However, an increase in tip vortex extension and intensity alongside the downstream of the propeller is not adequate itself to make a reliable prediction of propeller URN by using RANS. Consequently, a more realistic prediction of propeller URN requires the use of advanced models (i.e. LES (Large Eddy Simulation) and DES (Detached Eddy Simulation)) together with the porous FW-H equation, particularly if the receivers located in the downstream are of great interest.