Physics-Informed Broadband Noise Source Identification and Prediction of an Ideally Twisted Rotor

What is it about?

This work aims to provide a physics-based approach for the prediction and isolation of broadband noise emanating from different regions of an ideally twisted rotor. A preliminary prediction was conducted using a lattice-Boltzmann method–very-large-eddy simulation (LBMVLES) implemented within the software suite, PowerFLOW. Regions of particular interest to broadband noise were investigated by calculating one-third octave sound pressure levels of the unsteady pressure fluctuations acting on the rotor. These regions were then treated as individual Ffowcs Williams and Hawkings (FW-H) surfaces and simulated for three run conditions at a finer spatial resolution to identify the broadband noise from these separate regions and isolate it from the total acoustic spectra. These predictions were then compared to experimentally acquired data and to semiempirical prediction methods to highlight the acoustic contributions of various broadband noise generation mechanisms as well as to exemplify and explain shortcomings in the semiempirical methodology.

Featured Image

Why is it important?

This work investigates broadband acoustic phenomena not previously emphasized, such as blade wake interaction noise and laminar boundary layer vortex shedding noise (LBL-VS). It also explains a new, physics-informed approach at implementing the semi-empirical method of Brooks, Pope, and Marcolini for predicting self-noise with the inclusion of tip vortex formation effects and a Reynolds number limiter for tuning LBL-VS predictions.