A Wall-Modeled Large-Eddy-Simulation Tool for Fan Broadband Noise Prediction

This SBIR project proposes to develop a wall-modeled large-eddy-simulation (WMLES) tool for fan broadband noise prediction. Since fan broadband noise results from the interaction of turbulence with solid surfaces, it is important to resolve turbulent eddies up to a certain scale in order to achieve an accurate fan broadband noise prediction. Unfortunately the computational cost of the wall-resolved LES approach is found to scale with the Reynolds number as Re2.72, similar to Re2.91 of the direct numerical simulation (DNS) approach. To obtain a faster tool for fan broadband noise prediction, this SBIR effort will pursue the WMLES approach, in which the inner portion of the boundary layer will be modeled rather than resolved. The approach can reduce the computational cost to Re1.14. As a feasibility study, the Phase I outcome will demonstrate the feasibility of the proposed WMLES approach for accurate simulation of NASA 22-in fan noise source diagnostic test (SDT) case. Therefore, it is meaningful to further refine the methodology and develop a computational software tool for commercialization in Phase II.