2020 Volume 11 Issue 2 Pages 44-55
This paper presents detailed measurements and post-test simulations of the penny cavity leakage flow and its interaction with the mainstream flow in an annular cascade wind tunnel. The annular cascade wind tunnel consists of a single row of 30 variable stator vanes, derived from a high-pressure compressor stator with inner and outer vane disks, called pennies, which - when assembled in the hub and casing walls - leave cylindrical-shaped ring gaps called penny cavities. The wind tunnel runs at a Mach number of 0.34 at the stator inlet and a Reynolds number of 3.82 x 105 based on axial chord length at 50% span. Two different penny gap sizes on the hub are compared to a reference case without a penny gap. Detailed 2D-traverses were performed with multi-hole pressure and hot-wire-probes covering 2.5 passages in the inflow and outflow of the stator row. Pressure taps were embedded in the airfoil surface and inside the penny cavity. Surface oil flow measurements were conducted with different colors for the vane suction side, pressure side, hub and the penny cavity to detect the secondary flow phenomena. Reynolds- averaged Navier-Stokes (RANS) simulations, using the measured boundary conditions, were compared to experimental data. As a result, a relative increase in the total pressure loss coefficient of 1.9% for the nominal and 6.8% for the double penny gap was measured compared to no-penny cavity. The additional penny losses are limited to the lower 40% span. The post-test simulations are in good agreement with the measurements, showing that the outflow from the penny cavity on the suction side generates vortices, which cause additional losses. The penny vortices are detected in the outlet plane by an increase in turbulence intensity and streamwise vorticity. However, the additional penny losses are overestimated in the simulation by up to 7.3%. A change in the pressure fields with an increasing penny gap size, both around the airfoil and inside the penny cavity, can be seen in the numerical and experimental results. The outflow regions of the penny cavity, estimated by simulations, are confirmed by the results of the surface oil flow measurements. In summary, this paper consolidates previous numerical analyses carried out by the authors [13-16] on penny cavity leakage flow effects with experimental data for different penny gap sizes.