Aerodynamic performance measurements of a film-cooled turbine stage

Abstract

The goal of this research is to measure the aerodynamic performance of a film-cooled turbine stage and to quantify the loss caused by film-cooling. A secondary goal of the research is to provide a detailed breakdown of the losses associated with film-cooling for the turbine stage being tested. The experimental work was carried out at the MIT Blowdown Turbine Facility using a highly loaded turbine stage. The Blowdown Turbine Facility is a short duration test facility capable of testing turbine stages under fully scaled conditions for a test duration of 0.5 seconds. The facility was modified to enable the measurement of the turbine mass flow and shaft torque. These newly developed measurement techniques, along with previously developed total pressure and temperature instruments, have enabled the measure- ment of the stage isentropic efficiency. A highly loaded turbine stage (without film-cooling) was designed, fabricated, and tested using the newly developed measurement techniques. The turbine stage was then modified to incorporate vane, blade and rotor casing coolant manifolds using precision electrical discharge machining. The film-cooling hole geometry was created using a laser drilling process to produce the required 43,000 cooling holes. The film-cooled stage was then tested over a range of operating conditions (pressure ratios and corrected speeds) and over a range of coolant-to-mainstream mass flow and temperature ratios.

(cont.) The loss due to film-cooling is defined as the difference in performance between the film-cooled turbine and an ideal turbine with the same velocity triangles and airfoil Mach number distributions. However, there is no uncooled turbine geometry that will produce the same flow conditions as the film-cooled turbine stage, and consequently, there is no experimental baseline that can be tested to determine the loss due to film- cooling. A meanline velocity triangle model of the turbine stage was developed using published correlations and loss models to estimate the performance of this ideal stage. The model was calibrated against the baseline test results without coolant and it was then used to estimate the loss due to film-cooling. The estimated loss due to film-cooling was 3.0% at the design point, which corresponds to 0.3% per percent of coolant. The estimated repeatability (U95) for the efficiency measurement of the uncooled tur- bine geometry is ± 0.14%. Based on this measurement repeatability, the net effect of a design change can be determined with an uncertainty of just ± 0.1% if four measurements are repeated for each design configuration. The estimated measurement uncertainty for the film-cooled stage efficiency is 0.55% and for back-to-back measurements the uncertainty is 0.45%.