Evaluation of different FW-H surfaces and modal decomposition techniques for the acoustic analysis of UAV propellers through detached eddy simulations

Abstract

Noise has been a much-studied problem since the beginning of aviation because it is one of the main factors affecting its social acceptance. In recent years, the expansion of Unmanned Aerial Vehicles (UAVs) has led to increased research on propellers working at low Reynolds numbers, which are typically found in this type of aircraft. This paper begins with an Improved Delayed Detached Eddy Simulation (IDDES) of a commercial nine-inch UAV propeller. The Ffowcs-Williams and Hawkings (FW-H) approach is then used to compute the acoustic propagation from the simulation results. Both aerodynamic performance and acoustic signature results in hover flight are validated experimentally. Then, the influence of different FW-H permeable surfaces is methodologically evaluated, finding that magnified SPL values at low frequencies are obtained when using cylinders due to vortical structures passing through the bottom cap, and that the use of spheres appears to be the most consistent approach. Once a validated, time-resolved flow field has been obtained, different data-driven modal decomposition techniques, such as Proper Orthogonal Decomposition (POD) and Dynamic Mode Decomposition (DMD), are applied to the 3D pressure field. This enables a better understanding of the propeller acoustic modes and the assessment of the suitability of each technique for this problem, especially when Reduced Order Models (ROMs) are sought.