Turbulent mean flow prediction in impinging jets using data assimilation methods

Abstract

In this paper, a data-driven turbulence model is devised based on data assimilation (DA) for predicting impinging jet characteristics for various Reynolds numbers and nozzle-to-plate distances. The shear stress transport (SST) model with Tam–Thies correction is applied. The SST model with the X term makes accurate predictions in the region of wall jet near the overlying stationary fluid but fails to predict the velocity distribution near the wall. The DA-optimized SST model with the X term is used for predicting the impinging jet to minimize the deviation between the model prediction and experimental data. Only the model constants corresponding to the region near the wall are optimized through DA. The model constants at H/D = 2, 3, and 6 are fitted using logarithmic curves with respect to the nozzle-to-plate distance to obtain a universal formulation for predicting the impinging jet under various flow conditions. The model using the fitted model constants, referred to as the SST-H/D model, accurately predicts the mean flow for different nozzle-to-plate distances, nozzle types, and Reynolds numbers.