Lattice-boltzmann and navier-stokes simulations of the partially dressed, cavity-closed nose landing gear benchmark case

Hou, Yu, Angland, David, Sengissen, Alois and Scotto, Aline (2019) Lattice-boltzmann and navier-stokes simulations of the partially dressed, cavity-closed nose landing gear benchmark case. In 25th AIAA/CEAS Aeroacoustics Conference, 2019. American Institute of Aeronautics and Astronautics.. (doi:10.2514/6.2019-2555).

Abstract

A Lattice Boltzmann Method (LBM) solver LaBS and a weakly compressible Navier-Stokes (NS) solver implemented in OpenFOAM are both used to simulate the Partially Dressed, Cavity-Closed Nose Landing Gear (PDCC-NLG) from the Benchmark Problems for Airframe Noise Computations (BANC). The main purpose of this study is to test and compare the accuracy and efficiency of the two solvers. Unlike most previous work that has attempted to compare and contrast the two methods, in this work the computational grids used by both methods are refined in an identical octree manner with identical refinement regions and resolutions. Also, Large-Eddy Simulation (LES) is used by both methods to model the subgrid-scale (SGS) turbulence using the same wall function to bridge the viscous and buffer sublayers near the walls. The unsteady pressure sampled on the landing gear surface is used to calculate the far-field noise using the same Ffowcs-Williams and Hawkings (FW-H) solver. The LBM solver was found to trigger early separation on wheels for the coarse resolution. LBM and NS predictions of the unsteady surface pressure spectra have some discrepancies with the coarse grid, but they became more consistent and agreed better with wind tunnel data with the finer resolution. In the far-field, LBM and NS results showed reasonable agreement with each other in the low frequency range. Differences existed in the mid-frequency range. However, the agreement with each other and with experiments improved with increased grid resolution. LBM provided more accurate predictions at high frequencies with a higher cut-off frequency. CPU time statistics in times shows that the computational efficiency was better with LBM solver compared to the NS solver.

This record has no associated files available for download.

Contributors

Author: David Angland

Download statistics