Scaling criteria accuracy for turbine blade film cooling effectiveness at unmatched temperature ratio conditions

Cooling performance experiments for turbine blades are increasingly being conducted at low temperature and pressure in order to reduce the cost and difficulty of the experiments. However, the lack of similarity between scaling and real engine conditions leads to deviations in the experimental results. The application of a suitable scaling criterion can significantly reduce this deviation and promote the engineering of scaling experiments. This study explores a suitable criterion for scaling of film cooling when the coolant to mainstream temperature ratio for scaling conditions deviates from the datum condition, taking the 1st stage stator from the E3 aero engine as an example. This study focuses on the pressure side of the blade. The effect of the coolant to mainstream temperature ratio on the pressure side film cooling effectiveness was initially investigated based on both numerical and experimental methods. The scaling accuracy of six criteria was then compared based on numerical results, followed by an analysis of the reason leading to the scaling deviation. In our study the datum condition was selected to have an inlet pressure and inlet temperature of 1861 kPa and 1543 K, respectively. This operating condition corresponds to that of a real engine. Finally, the results obtained from the numerical method are validated experimentally. Both the numerical and experimental results indicate that the film cooling effectiveness on the pressure side decreases with an increase in the temperature ratio at low temperature ratio conditions, while the trend reverses at relatively high temperature ratio conditions. We found that the M criterion is suitable for scaling the downstream region of −60 % l/l0 (the downstream region of the last cooling hole row), while the most accurate scaling results in the region from the leading edge to −60 % l/l0 (the region with dense film cooling holes) can be obtained by applying the I criterion. The region near the leading edge (around cooling hole row 1) is difficult to scale. This is because the separation in the downstream region of cooling hole row 1 in the datum condition does not exist in any scaling condition.