Development of Loss Correlation and Tool Validation at Transonic Condition Based on Cascade Test

Overall efficiency of an axial compressor is largely affected by its front stage when it is operating under transonic flow conditions. For this reason, many manufacturers and researchers are advancing research and development of transonic airfoils in these days. Doosan, in frame of a development of high efficiency gas turbine, developed high efficiency airfoil for a transonic rotor and conducted cascade tests. Therefore, this study deals with a test of two compressor transonic blade cascades at inlet Mach number over 1.1. To improve the efficiency and operating range, two kinds of thickness distribution type based on Enhanced Doosan Airfoil (EDA), which applied unique rule, were applied and assessed. The first airfoil consists of polynomial thickness distribution and the second airfoil consists of new thickness distribution with specially tailored leading edge. In order to ensure accurate geometry of a model, a detailed checkout process upon production of model blades used in the test was performed. This is because, in the case of transonic airfoil, if the inlet leading edge shape differs by more than 0.2% than designed airfoil of leading edge, the result will be completely different. Therefore, not only the tolerance within 0.1% was confirmed but also the shape produced through simulation and 3D CMM scan data. The main parameters for the comparison are an inlet Mach number, an axial velocity density ratio (AVDR) and the kind of thickness distribution. Results of tests and CFD blade to blade analysis using MISES 2.70 are compared. The flow field was visualized using schlieren technique and parameters of the suction side boundary layer were evaluated at several locations based on Pitot probe traverses. The results confirm that a suction peak at the round leading edge disappears in the case of the new thickness type distribution with tailored leading edge. This confirms that the profile shaping without jump in curvature in the leading edge region leads to smooth acceleration without peaks. Nevertheless, results show that the new thickness distribution type is not absolutely good in comparison with the polynomial thickness distribution type with respect to the total pressure loss coefficient. Moreover, bucket range (operating range) is also almost the same. Results of the suction side boundary layer traversing suggest that the transition of the boundary layer takes place beyond the location x/cax > 0.088. The MISES results show that a shock location and the boundary layer parameters are similar to test results. However, values of the loss coefficient show some difference. Therefore, a new correlation in particular transonic flow condition was developed.