Navier-Stokes computation and turbulence modeling for turbine viscous flow and heat transfer

The objective of this research is to improve the predictive capability of turbine viscous flows which is essential for achieving improved design and increased efficiency. A 3-D Navier-Stokes procedure incorporating a Reynolds stress model (RSM), an algebraic Reynolds stress model (ARSM) and a nonlinear k-{dollar}varepsilon{dollar} model, has been developed using a four-stage Runge-Kutta scheme. This procedure has been employed to analyze turbine viscous flows with an emphasis on turbulence modeling for streamline curvature effects, boundary layer transition/heat transfer, endwall secondary flows and turbine rotor and tip-leakage flows. A detailed assessment has been made for the performance of the RSM, ARSM and the nonlinear k-{dollar}varepsilon{dollar} model for the prediction of streamline curvature effects.; The Navier-Stokes procedure is used to analyze turbine boundary layer transition, heat transfer and the evolution of wake profiles. The effects of blade surface pressure gradient, freestream turbulence level and Reynolds number on the boundary layer transition are examined. The k-{dollar}varepsilon{dollar}ARSM model is found to be able to capture the anisotropy of turbulence associated with the by-pass transition. The simulations indicate that the turbine blade heat transfer, under real engine conditions, can be predicted well by the Navier-Stokes method.; The 3-D viscous flow through an annular turbine nozzle passage has been computed with the k-{dollar}varepsilon/rm ARSM{dollar} and the k-{dollar}varepsilon{dollar} model. These two models are found to provide similar predictions for the mean flow parameters, although slight improvement in the prediction of some secondary flow quantities has been obtained by the ARSM model. Finally, the 3-D viscous flowfield in the downstream turbine rotor, including the tip-leakage flow, is computed using the Navier-Stokes procedure. A new grid-generation code has been developed to obtain the embedded H grids inside the rotor tip gap. Detailed comparisons with the data indicate that the current procedure is sufficient to obtain much of the 3-D flowfield information in the entire turbine stage and can be used as a valuable design tool for turbomachinery engineers.