Numerical prediction of periodic transpired channel flow with low-Reynolds-number turbulence models

A fully developed channel flow with periodic transpiration, under the influence of injection and suction, and positive and negative pressure gradients, is a challenging test case for turbulence models. A direct numerical simulation (DNS) of the flows with weak and strong transpiration is performed. Three low-Reynolds-number turbulence models without wall distance, the stress transport model proposed by Shima, the nonlinear K-ε model of Okamoto and Shima and the Launder and Sharma linear K-ε model, are tested by using the DNS data obtained. The prediction of the models is generally good near the nontranspired wall, while the stress model reproduces the turbulence quantities near the transpired wall.