Three Dimensional Optimum Design of a Steam Turbine Stage With NURBS Curves

Abstract

Progress in the computer performance has enabled automatic optimization of the three dimensional shape of turbine blades with a large number of large-scale CFD (Computational Fluid Dynamics) calculations. This paper presents an advanced aerodynamic optimization system for turbine blades. The system can automatically find improved blade shapes that give better aerodynamic performance in a turbine stage and reduces human efforts to generate blade shape data, computational mesh, CFD input data etc. The system consists of three parts; parameter updating part, blade shape generation part, and aerodynamic performance evaluation part. In the parameter updating part, users can choose DOE (Design of Experiment) or evolutionary optimization method such as GA (Genetic Algorithm), ASA (Adaptive Simulated Annealing), etc. to define the parameters in each step. The shape generation part changes the blade shapes using NURBS curves whose control point parameters are defined in the parameter updating part. Three-dimensional CFD grid is automatically generated for the changed blade shapes and steady CFD calculation is used to evaluate the aerodynamic performance of the changed blades in a turbine stage. Stagger angle distribution in the radial direction was thought as one of the important design parameters of turbine blades because it determines the flow pattern in radial direction. Then it was chosen as an optimized parameter with NURBS curves in this system. First, DOE was used for the human optimization, in which the parameter range for the advanced optimization was estimated and the best shape obtained was used as the initial shape for the evolutionary optimization to explore better blade shape parameters. Stage loss of an exhaust stage of IP (Intermediate Pressure) turbine which contains relatively high aspect ratio was chosen as the objective. In spite that such kind of stage was considered not to be sensitive to three dimensional stacking, the results showed good performance enhancement.