Computation of heat transfer in a linear turbine cascade

The efficiency of a turbine increases in general with an increase of the temperature of the working gas. In modern turbines this gas temperature may well exceed the melting temperature of the metal walls (Harasgama, 1995). Locally high heat transfer can lead to an excessive temperature and high thermal stresses in the walls, causing an early fatigue of the high pressure turbine components. Thus, the design of these components requires accurate evaluation of heat transfer at the walls. The prediction of heat transfer at the endwall and the blade surface requires simulation of the viscous interaction between the boundary layer approaching the blade and that developing on the blade itself. Secondary flows, horseshoe type vortices, and strong turbulence effects generate complex endwall heat transfer distributions with several local maxima occurring at the endwall and the blade surface. Accurate prediction of these peaks is crucial for the design of the turbine cooling system. Recently, a detailed experimental investigation of the endwall heat transfer in a linear cascade was carried out at the NASA Glenn Transonic Turbine Blade Cascade Facility. This investigation includes the effects of Mach and Reynolds number and the influence of the free stream turbulence on the endwall and blade surface heat transfer rate. A large database has been created through this work for the validation of numerical methods for the simulation of turbine flows. The objective of the present work is to use this database to evaluate the influence of turbulence models on the accuracy of heat transfer predictions in complex three-dimensional flows in turbine geometries. The sensitivity of the heat transfer coefficient prediction to the turbulence model used is analyzed using two different models: the Spalart-Allmaras one equation model (Spalart & Allmaras 1992) and Durbin’s four equation v-f model (Durbin 1995). The use of two different flow solvers, the NASA research code CFL3D and the commercial package FLUENT, increases confidence in the results and allows the elimination of effects related to the numerical discretization of the equations.