A theoretical investigation was performed to predict the maximum achievable efficiency of radial inflow turbines for different design conditions. The analytical tool used in the investigation is a computer code able to perform the contemporary optimization of the main design variables, in order to obtain maximum efficiency. Since the results are strictly dependent on the loss correlations, reliability of the efficiency predictions was tested at first by comparison with several test-cases available in literature: good agreement with experimental data was found, pointing to the validity of the results presented here.
A large number of cases were analyzed: the efficiency and the main design parameters, obtained after the optimization process, are presented for optimum specific speed. Turbine efficiency was found to be dependent both on compressibility effects, related to the volume expansion ratio, and on actual turbine size, accounting for geometric non-similarity effects. Influence of non-optimum specific speed is also discussed.
By means of similarity rules, the results enable turbine design to be performed in a simple way, for a variety of working fluids and design conditions.
