GPU-accelerated large-eddy simulations of the NASA fan noise source diagnostic test benchmark

Abstract

Wall-modeled large-eddy simulations of NASA Fan Broadband Source Diagnostic Test (SDT) model are performed using the GPU-accelerated moving-mesh compressible flow solver charLES. Three different outlet guide vane (OGV) configurations (i.e., baseline, low count and low noise) are simulated under the approach condition. The full annulus of the fan and the OGVs, the nacelle and the entire test section are included in the computational domain to avoid numerical artifacts caused by artificial periodicity or domain truncation. The LES results are compared to available experimental data in terms of aerodynamic performance, unsteady flow fields and far-field noise, and show good overall agreement. In particular, the predicted sound power levels and the fan and stage performance match closely with the measurements for the three OGV configurations. The SDT configuration is also used to assess the solver scalability and increased computational throughput with GPU acceleration: for the present O(140) million cell mesh, the simulation results for 10 full rotations are obtained in approximately 6 hours on 40 standard GPUs.