Multi-objective optimization on internal cooling strategies for gas turbine blade leading edges

Abstract

In this work, multi-objective optimization on internal cooling strategies is performed for gas turbine blade leading edges, aiming to maximize the heat transfer rate and minimize the friction loss. Inspired by previous advanced internal cooling concepts (swirling cooling, impingement cooling and double impingement cooling), four geometrical parameters are determined as design variables, including the deformation length of impingement chamber, the horizontal position and incline/compound angles of nozzles. A series of validated numerical simulations based on Reynolds Averaged Navier-Stokes equations are performed to provide sufficient training source and generate accurate surrogate model. Two cooling strategies are optimized by genetic algorithm, and their heat transfer and flow characteristics are compared with the previous cooling methods. By the optimized structures, the area with high Nusselt number enlarges, and the heat transfer rate distributes more uniformly. Compared to swirl cooling, the optimized strategies can obtain a heat transfer enhancement of 7.5% to 15.5% with a friction loss reduction of 6.7% to 0.5%, when Reynolds number is 18500. Correspondingly, the comprehensive aerothermal efficiency is improved by at least 15.8%. Moreover, the effects of the four design variables are discussed, and the design guidance of internal cooling for blade leading edges is provided thereby.