Metamodelling techniques for propeller array far-field noise

The present paper proposes the development of surrogate models for the description of acoustic emission of multi-propeller configurations, like those adopted by the vast majority of the Urban-Air-Mobility concepts nowadays under investigation. This approach represents an interesting and efficient solution for the aeroacoustic characterization of electrically-powered propeller array, for which reliable and cost-efficient aeroacoustic models are still lacking, and the application of costly high-fidelity tools is mandatory. The numerical investigation focuses on surrogate models that take into account the effects of geometric parameters, such as propeller hubs relative positions and de-phasing, on the propeller array noise emissions. A well-known Artificial Neural Network metamodelling technique is applied on a numerical database obtained through a boundary integral formulation for the solution of incompressible potential flows around lifting/thrusting bodies, followed by the application of the Farassat 1A boundary integral formulation for the noise field evaluation. Due to the proposed non-standard application of the metamodelling technique applied, particular attention is paid to identifying the best strategy to exploit a database with a minimal number of samples with respect to those managed in the standard application of machine learning technique.