The paper proposes a method for evaluating the reliability of models for estimating the energy losses in the blade rows of axial turbines, based on the statistical analysis of the experimental data deviation from the calculated. It was shown that these deviations are subject to the normal distribution law and can be described by mathematical expectations μΔξ and standard deviation σΔξ.
The values of profile losses were calculated by five well-known models for 170 different axial turbines cascades, representing the diversity of turbines used in aircraft gas turbine engines. The findings were compared with experimental data. Comparison results were subjected to statistical analysis. It was found that the best model to describe the profile losses in axial turbines is model that has been developed in Central Institute of Aviation Motors (Russia). It allows the calculation of profile losses deviating from the actual values of losses by −8±84% with a probability of 95%.
Taking into account the mentioned statistical criteria, a new equation was proposed based on the analysis of the profile losses nature and using mathematical optimization techniques. This equation makes possible to define the profile loss of axial turbine more accurate than the investigated models. It allows the calculation of profile loss values in the axial turbine that differ from the actual value by 10±61% with a probability of 95%. The proposed new equation takes into account more geometric and operational factors affecting the value of losses.