First of all, this paper presents the overall efficiency correlation to the balance hole diameter of the 3-stages air turbine test results. All 3 wheels have the same balance hole geometry, with total 5 holes in the wheel, and the variation of diameter is 0 mm, 16 mm, 25 mm, 30 mm, 40 mm, 50 mm, and 60 mm, respectively. The test results show that the lowest efficiency condition is the one without balance hole, and increases with the hole diameter until 25–30 mm, but decreases after it. Then, a single stage CFD model, including the first stage stator, moving blade, stator hub leakage, and wheel chamber, is built to simulate the stage efficiency trend under steady flow condition. The relative efficiency correlation of the simplified single stage CFD model matches very well with the test results. The CFD results show that, the maximum efficiency happens when there is minor gas suction from the main gas path, about 0.1%–0.2% compared with the main mass flow, but when the gas suction is above 0.2% of the main mass flow, the efficiency decreases, as less flow is passing through the moving blade to convert energy.
Finally, the radial seal clearance’s influences in the wheel chamber between the diaphragm and the wheel is checked by CFD using the same geometry pattern. Five different radial seal clearances model, including 0 mm (ideal condition), 0.6 mm, 1.2 mm, 1.8 mm and 6 mm (without radial seal) are built, under two different balance hole diameter (14 mm and 25 mm), and three different hub seal clearances (0 mm, 0.6 mm and 1.2 mm) conditions. The CFD results show that the throttling effect is very small when the radial clearance is above or equal to 0.6 mm, with almost constant efficiency variation with the clearance up to 6 mm, under specified stator hub clearance and balance hole diameter. Very small deficiency (0.2%) is detected under small balance hole diameter (14 mm) and large stator hub clearance (1.2 mm) condition, as the hub leakage mass flow is much beyond the balance hole flow capability.