Performance Improvement of an Industrial Axial Turbine: Two-Stage Power Turbine Aerodynamic Optimization

Abstract

Abstract For high-aspect-ratio turbine blades, profile loss is dominant and its reduction is important for improving performance. In this paper, a two-stage low-pressure turbine (power turbine) is optimized in two steps: cycle optimization and aerodynamic optimization. In the cycle optimization step, the thermodynamic boundary condition of the turbine is modified. In the second step, which is the main focus of this study, the aerodynamic improvement of a turbine using a two-dimensional optimization technique is performed. The Genetic Algorithm (GA) is used in the optimization process to identify variables that satisfy a defined objective function which is subject to some constraints. The Artificial Neutral Network (ANN) is introduced to correlate variables with the defined objective functions and constraints. Maximum velocity and its location on the blade surface determine transition location and diffusion factor which directly affect total pressure losses. Therefore, in the optimization process, maximum velocity and its location are controlled and defined as objective functions. Profile shape at any given radius is represented with the camber-thickness method. The flow solver used for the aerodynamic analysis of profiles is the cascade analysis code MISES. After improving the performance of vanes and blades’ 2D sections, to be confident for performance enhancement of final power turbine ANSYS CFX is used. Finally, the aerodynamic characteristics of the power turbine using optimized geometries are discussed and compared to the original one. The results demonstrate a 1.19 percent improvement in power turbine efficiency at the same pressure ratio.