A full-field noise conversion model of different Mach numbers for flows past a cylinder at Reynolds number 3900

For the benchmark simulation of flow past a stationary body, dipole source constitutes the principal noise-generating mechanisms. However, in underwater environments, the extremely low signal-to-noise ratio (SNR) renders experimental noise measurements particularly challenging, making it quite difficult to obtain accurate results. To address this limitation, the present study employs the noise conversion exponents Na and Nu, thereby extending existing noise conversion methodologies of different Mach numbers for noise conversion between two different media. These exponents are applied to enable mutual conversion between low-Mach-number aerodynamic noise and hydrodynamic noise at identical Reynolds numbers, allowing high-SNR aerodynamic acoustic experimental data to be leveraged for more reliable hydrodynamic noise prediction. In this work, the open-source solver OpenFOAM is employed to simulate flow past a circular cylinder, and dynamic mode decomposition in conjunction with wavelet analysis is utilized to perform both qualitative and quantitative investigations of hydrodynamic and aerodynamic flow-induced noise. The similarities and differences between the two cases are examined in detail in both the flow field and the acoustic field. The noise conversion exponents Na and Nu are then applied to achieve effective cross-domain noise conversion. Results indicate that, for observers with identical radii but different azimuthal positions, the Nu facilitates accurate conversion from aerodynamic noise to hydrodynamic noise, whereas for observers with identical azimuthal positions but different radii, the Na works. The proposed noise conversion exponents thus provide a valuable framework for predicting underwater noise from wind tunnel acoustic measurements.