Computational study of boundary layer effects on stochastic rotor blade vortex shedding noise

Abstract

This work entailed the use of lattice-Boltzmann simulations to investigate the effects of different sUAS rotor blade surface treatments, volumetric grid resolutions, and surface roughness heights on broadband noise prediction. Aerodynamic results showed trailing edge flow separation and a mispredicted laminar separation bubble for the laminar-to-turbulent transitional wall-function case and the fully-turbulent wall-function case as well as a transition front for the laminar-to-turbulent transitional wall-function case. It was seen that the current version of PowerFLOW (6-2021) cannot predict turbulent structures associated with turbulent wall-bounded flow unless a triggering mechanism (e.g., boundary layer trip) is employed, possibly explaining aerodynamic force and broadband noise underprediction. Tonal noise results showed thickness (i.e., geometry dependent) noise dominance at out-of-plane observers located above the rotor plane for the fundamental blade passage frequency and at all observer locations for the second harmonic of the blade passage frequency. Extrapolation techniques were also successfully used over broadband noise results to separate regions of underpredicted broadband noise caused by inadequate spatial resolution from other underpredicted regions caused by the misprediction of flow physics using the current computational strategy.