Radial Inflow Turbine Design Based on Residual Entropy Method for Ultra-Low Pressure Ratio Conditions

To achieve space-saving, noise reduction, and lower direct emissions in marine diesel engines, it is necessary to position the exhaust pipe below the water surface. This paper proposes a new method for designing low pressure ratio variable geometry turbines to increase the maximum tolerable back pressure and enhance the engine safety of underwater exhaust diesel engines. Firstly, to solve the problem of poor convergence and accuracy of the traditional turbine design method under low pressure ratios, a novel turbine performance design method called the residual entropy method is introduced. This approach transforms the turbine performance prediction problem into solving for the intersection of the state entropy and the loss entropy curves, which corresponds to the root of the residual entropy. Secondly, the accuracy and convergence of the proposed model are validated. The results indicate that the main source of prediction errors in models based on ideal gas properties is loss entropy, particularly beyond a Mach number of 0.1. The residual entropy design method outperforms commercial software approaches in terms of convergence and accuracy. Experimental validation further confirms that the efficiency prediction error does not exceed 4.5%. Finally, a genetic algorithm is used for turbine design optimization to reduce turbine size while maximizing efficiency. The findings indicate that the velocity ratio is the key factor influencing turbine efficiency, with peak efficiency achieved at a velocity ratio of approximately 0.63 under low pressure ratios. After optimization, turbine efficiency reaches 83.2%, while maintaining a size comparable to the original design. The turbine efficiency shows a negative feedback relationship with the pressure ratio. The proposed new method is also applicable to the design of various types of turbines.