A semi-implicit discrepancy model of Reynolds stress in a higher-order tensor basis framework for Reynolds-averaged Navier–Stokes simulations

With the rapid development of artificial intelligence, machine learning algorithms are becoming more widely applied in modification turbulence models. In this paper, with aim improving prediction accuracy Reynolds-averaged Navier–Stokes (RANS) model, a semi-implicit treatment Reynolds stress anisotropy discrepancy model is developed using higher-order tensor basis. A deep neural network constructed and trained based on model. The parameters embedded computational fluid dynamics solver to modify original RANS Modification computations performed for two cases: one interpolation extrapolation different numbers. For these cases, ability modified capture anisotropic features has been improved. Moreover, when compared mean velocity large eddy simulations (LES), root square error significantly lower than Meanwhile, can better simulate flow field separation fluctuation shear layer reattachment point profile addition, also improves pressure coefficient friction underlying wall surface. previously directly computation massive periodic hill flows. It shown that results simulated by LES approach consistent both trend magnitude approach.