libFastMesh: An optimized finite-volume framework for computational aeroacoustics

Published: 13 October 2023| Version 1 | DOI: 10.17632/7w9fy2xtcf.1
Contributors:
, Alon Zameret, Yann Delorme, Yanir Edri, Solal A.T. Amouyal, Mark Wasserman, Zhaohui Ding,

Description

This paper presents an optimized framework for simulating aeroacoustic flow on unstructured meshes. The libFastMesh (LFM) framework is based on a low-dissipation finite–volume discretization, together with high-order explicit time integration, for direct computation of aeroacoustic fields. These methods were previously implemented in caafoam, an OpenFOAM-based solver developed by D'Alessandro et al. [1] to resolve aeroacoustic flows. The new framework is specifically developed from scratch to alleviate two of the main bottlenecks currently limiting the performance of OpenFOAM-based solvers at large scale: memory bandwidth and inter-node communication. Compact data structures and compute kernel fusion are employed to enhance cache utilization and re-use, respectively. Separate treatment of inter-process boundary cells, together with non-blocking MPI communication enable effective overlap between communication and flux computations. The accuracy and robustness of libFastMesh are established via simulations of well-established aeroacoustic flow benchmarks. Comparison to previous simulation results obtained with caafoam, and to DNS data serve to validate the code. Finally, the effectiveness of the aforementioned optimizations is demonstrated through rigorous analysis of the proposed solver performance in single-node and multi-node operation modes. The obtained results show that libFastMesh offers a speed-up of up to 20x with respect to the previously developed OpenFOAM-based solver, caafoam, in large-scale computations. Moreover, LFM is shown to scale extremely for cell-per-core ratios of less than 500. These results are particularly appealing given the highly resource-demanding nature of direct computational aeroacoustics (CAA). Consequently, LFM could be an extremely attractive tool for scientists who wish to conduct large-scale CAA simulations on modern, exascale architectures.

Files

Categories

Computational Physics, High Performance Computing, Aeroacoustics

Licence