Modern High Pressure Turbine (HPT) blades operate at high speed conditions. The Over-Tip-Leakage (OTL) flow, which plays a major role in the overall loss generation for HPT, can be high-subsonic or even transonic. In practice from the consideration of problem simplification and cost reduction, the OTL flow has been studied extensively in low speed experiments. It has been assumed a redesigned low speed blade profile with a matched blade loading should be sufficient to scale the high speed OTL flow down to the low speed condition.
In this paper, the validity of this conventional scaling approach is computationally examined. The CFD methodology was firstly validated by experimental data conducted in both high and low speed conditions. Detailed analyses on the OTL flows at high and low speed conditions indicate that, only matching the loading distribution with a redesigned blade cannot ensure the match of the aerodynamic performance at the low speed condition with that at the high-speed condition. Specifically, the discrepancy in the peak tip leakage mass flux can be as high as 22.2%, and the total pressure loss at the low speed condition is 10.7% higher than the high speed case.
An improved scaling method is proposed hereof. As an additional dimension variable, the tip clearance can also be “scaled” down from the high speed to low speed case to match the cross-tip pressure gradient between pressure and suction surfaces. The similarity in terms of the overall aerodynamic loss and local leakage flow distribution can be improved by adjusting the tip clearance, either uniformly or locally. The limitations of this proposed method are also addressed in this paper.