A compact and accurate empirical model for turbine mass flow characteristics

Abstract

The model of turbine mass flow characteristics is vital to simulate equipment and systems having a turbine, such as renewable energy power systems, air cycle refrigeration systems, power plants, and turbine engines. Existing empirical and partly empirical models of turbine mass flow characteristics need to be improved either to increase the prediction accuracy and extrapolation performance or to reduce complexity. A new empirical model describing the turbine mass flow performance map, also called a mean value model, is developed through extensive computing tests using curve fitting and nonlinear regression software. Measured data of a turbocharger turbine and a simple air cycle machine (ACM) turbine are used for the model building. The proposed model is highly compact and accurate, its predictions agree with measured data very well, and it has excellent extrapolation performances as well. The mean absolute relative error is 1.38% for the simple ACM turbine and 0.91% for the turbocharger turbine. Comparison with the best existing model shows that the new model reduces the mean absolute relative error by about 40%, and is much easier to use and much more compact.