Rapid performance prediction model of axial turbine with coupling one-dimensional inverse design and direct analysis

Abstract

The reasonable and rapid prediction of turbine performance based on design targets such as mass flow rate and pressure ratio is indispensable in the preliminary and conceptual design stage for gas turbine engines in the aviation and power generation industry. However, the lack of detailed geometrical and aerodynamic parameters in this stage significantly challenges the reasonable evaluation of turbine characteristics. The present work proposes a rapid prediction method of axial turbine performance by coupling the inverse design and direct analysis. Firstly, according to the one-dimensional mean line design theory with empirical loss and deviation correlation models, the basic aerodynamic and geometrical parameters are quickly obtained by solving the inverse problem to fulfill the design targets such as mass flow rate and pressure ratio with optimized efficiency. Then, by adopting the loss and deviation correlation models in the off-design conditions, the turbine characteristics in a wide range of operating conditions are obtained by solving the direct analysis problem. Thus, the reasonable and rapid prediction of the turbine characteristics can be realized by coupling the inverse design with direct analysis. Based on the built prediction model, the effect of various deviation models on the predicted turbine characteristics, which was not systematically investigated in previous studies, is clarified by comparing the predicted performance with published test results. Finally, an optimization method of the empirical coefficients in turbine loss and deviation models based on the Genetic Algorithm (GA) is introduced. This method employs optimization theory and existing experimental data to improve the prediction accuracy. The results show that the averaged prediction discrepancies of output powers and mass flow rates in a wide range of operating conditions are no more than 2% and 2.5% for test cases after the optimization of model coefficients. Therefore, this optimization method is expected to be used for updating the empirical coefficients to improve the model prediction accuracy, accompanying the evolution of turbine design strategy at present or in the future.