.png){kind=icon}
.png){kind=icon}
Return
Predicting turbulent flow over a backward-facing step using grid-adaptive simulation method
.png){kind=icon}
PRE
AI
DOI：10.1016/j.ast.2024.109913.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Flow separation in backward-facing step (BFS) is a common phenomenon in engineering. The large scale of turbulent separation and reattachment makes it a challenge for the accurate prediction of such kinds of flows. Traditional hybrid Reynolds averaged Navier-Stokes (RANS) and large eddy simulation (LES) methods require relatively high grid resolutions, while the grid-adaptive simulation (GAS) method, a recently proposed hybrid RANS-LES method, can obtain high accuracy with reduced grid-resolution requirements. In this study, sensitivity studies of spanwise extent and grid resolution are conducted on the GAS method with the shear-stress transport (SST) k-omega turbulence model in simulating the BFS flow. Then, the delayed detached eddy simulation (DDES), improved DDES (IDDES), and scale-adaptive simulation (SAS), based on the SST model are compared with the GAS method for predicting the BFS flow with relatively low resolution of grid. The skin-friction and wall pressure coefficients, the velocity, and the resolved turbulent stress profiles predicted from different hybrid RANS-LES methods are compared with the experimental results. Results indicate that the GAS method can significantly outperform other hybrid RANS-LES methods when using coarser grid. The dynamic mode decomposition (DMD) analysis is conducted based on the GAS-SST results. Results show that GAS-SST with a low-resolution grid of 0.3 million cells can accurately predict key parameters such as mode shape, mode frequencies, spectral characteristics, and energy distribution of the main unsteady structures. Hence, a large amount of computational cost can be saved by the GAS method in comparison to the previous high-fidelity IDDES-SST simulation with a highresolution grid of 14 million cells.
Keywords:
Hybrid RANS-LES method
Backward-facing step flow
Grid-adaptive simulation
Dynamic mode decomposition
Turbulent simulation
Journal
IF:
5.8
Papers:
1.0W
Citations:
3.0W