A STATISTICAL APPROACH TO THE EXPERIMENTAL EVALUATION OF TRANSONIC TURBINE AIRFOILS IN A LINEAR CASCADE

In aircraft engine design (and in other applications), small improvements in turbine efficiency may be significant. Since analytical tools for predicting transonic turbine losses are still being developed, experimental efforts are required to evaluate various designs, calibrate design methods, and validate CFD analysis tools. However, these experimental efforts must be very accurate to measure the performance differences to the levels required by the highly competitive aircraft engine market. Due to the sensitivity of transonic and supersonic flow fields, it is often difficult to obtain the desired level of accuracy. In this paper, a statistical approach is applied to the experimental evaluation of transonic turbine airfoils in the VPI & SU transonic cascade facility in order to quantify the differences between three different transonic turbine airfoils. This study determines whether the measured performance differences between the three different airfoils are statistically significant. This study also assesses the degree of confidence in the transonic cascade testing process at VPI & SU.