SHOWERHEAD FILM COOLING PERFORMANCE OF A TRANSONIC TURBINE VANE AT HIGH FREESTREAM TURBULENCE (Tu=16%): 3-D CFD AND COMPARISON WITH EXPERIMENT

The main objective of the study reported here is to use 3-D CFD to calculate and explain adiabatic film cooling effectiveness and Nusselt number distributions downstream of the showerhead film cooling rows of a turbine vane at high freestream turbulence and realistic exit Reynolds number/Mach number condition. The paper discusses a three-simulations technique to calculate vane surface recovery temperature, adiabatic wall temperature, and surface Nusselt number to completely characterize film cooling performance in a high speed flow. The RANS based nu(2)-f turbulence model, originally suggested by Durbin [1] is used in all numerical predictions. The vane midspan numerical calculations are compared with the experimental results obtained with the showerhead film cooled vane instrumented with single-sided platinum thin film gauges at the midspan and arranged in a. two-dimensional, linear cascade in a heated, transonic, blow-down wind tunnel. Exit Mach number of M(ex) = 0.76-corresponding to exit Reynolds numbers based on vane chord of 1.1 x 10(6)-was tested with an inlet free stream turbulence Intensity (Tu) of 16% and an integral length scale normalized by vane pitch (Lambda(x)/P) of 0.23. A showerhead cooling scheme with five rows of cooling holes was tested at blowing ratios of BR = 1) and 1.5, and a density ratio of DR= 1.3. CFD predictions performed with experiment-matched boundary conditions show an overall good trend agreement with experimental adiabatic film cooling effectiveness and Nusselt number distributions downstream of the showerhead film cooling rows of the vane. For the experimental data, the primary effects of coolant injection are to augment Nusselt number and reduce adiabatic wall temperature downstream of the injection on the vane surface as compared to no film injection case (BR = 0) at M(ex) = 0.76. Similar to experimental results, the adiabatic film cooling effectiveness prediction on the Suction surface at BR = 1.5 is found to be influenced by favorable pressure gradient due to Mach number through changes in local adiabatic wall and recovery temperature. The Nusselt number prediction on the suction Surface shows a peak and a valley downstream of the film cooling rows in a favorable pressure gradient region for both tested blowing ratio conditions. This trend is also observed in the experimental results.