Numerical Study of Turbulence Transition Models
by
Kshitij Neroorkar
UAB
Coauthors: Dr. Gary Cheng

The transition of an attached boundary layer from laminar to turbulent greatly affects the performance of aerospace and turbomachinery devices. In the past, a wide range of methods have been employed for accurately modeling this physical phenomenon. These methods mainly include the Stability theory, Direct Numerical Simulation (DNS), and numerous proposed modifications to existing Reynolds Averaged Navier Stokes (RANS) turbulence models. Out of all the existing methods, it is believed that the RANS-based models are easiest for coupling with Computational Fluid Dynamics (CFD) codes in order to achieve rapid and reasonably accurate estimates. Based on our literature study of the existing RANS-based turbulence transition models, the Walters and Leylek model and the Local Correlation based Transition Model of Menter et al. have the best physical justification and generality. The main goal of this work is to implement these two transition models into RANS-based CFD code (FDNS) and to evaluate them with some benchmark test cases.

PDF

Date received: August 31, 2007