CFD study of effusion cooling

Abstract

The desire to increase the efficiency, i.e., reduce the specific fuel consumption and raise the thrust-to-weight ratio, of gas turbines has led to an increase in pressure and temperature in the combustion chamber and turbine. The operational life of the combustion chamber walls decreases with increased temperature thus an effective method of cooling must be used to protect the wall. Effusion cooling provides a practical solution to this engineering problem. A fundamental understanding of the physical mechanisms involved in effusion flow fields is required to make significant advances in cooling technology. At the same time, designers need a predictive design tool that allows quick turnaround times without the current build and break approach. Computational Fluid Dynamics (CFD) presents the designer with the potential for an effective, fast and relatively accurate method of achieving this. This paper presents a computational study of effusion cooling applications using the Reynolds Averaged NavierStokes (RANS) approach. The need to evaluate the predictive capability of the Reynolds Stress Transport (RST) model when applied to Full Coverage Film-Cooling (FCFC) effusion scenarios is highlighted since two-equation Eddy-Viscosity (EV) models fail to predict turbulent anisotropy and therefore the complex flow mechanisms involved in effusion cooling flow fields. An isothermal and non-isothermal numerical study of effusion cooling flow is conducted. In the isothermal case the RST model is shown to be capable of predicting the injection, penetration, downstream decay and lateral mixing of the effusion jets reasonably well. In the non-isothermal case the laterally averaged cooling effectiveness across the plate is slightly under-predicted but still conforms to the general increasing trend.