Robust Bayesian Calibration of a k−ε Model for Compressible Jet-in-Crossflow Simulations

What is it about?

Jet-in-crossflow simulations, when conducted using a Reynolds-Averaged Navier-Stokes model, are rarely predictive. We show how this may be mainly due to the use of inappropriate parameters in the k-epsilon turbulence model. We provide more appropriate parameters which we calculate using experimental data, as well as estimate theoretically. Our parameters lead to simulations that are far more predictive than those obtained using the standard k-epsilon model.

Featured Image