Multi-objective optimum design for low-pressure axialflow fan on multi-points

This paper presents a multi-objective optimization method for a low-pressure axial-flow fan on four operating points through hybrid optimization algorithm based on three-dimensional computational fluid dynamic simulation technique. The multi-objective optimization approach coupled with Kriging approximation model that build the relationship between the design parameters and target variables was proposed for reducing many time-consuming CFD calculations during the process. The distribution of chord length, camber length and stagger angles along the radial direction were parameterized by the polynomial function, Gaussian function and exponential function separately, and eight parameters related were selected as the design variables to generate sample points by Latin hypercube experiment design method to construct the approximation model in the design space for optimization after CFD simulations. Simulation analysis was implemented to obtain objective functions by applying finite volume method to disperse the three-dimensional RANS equations, which were solved by segregated and implicit method with SST turbulence model. The average weighted total pressure rise and efficiency over four operating conditions were selected as objectives. The maximum torque, the minimum efficiency and the minimum total pressure were constrained by requirement specifications. The redesigned fan has the 21% improvement in the pressure rise compared with the required total pressure rise and gains a wider and more stable working range, also the maximum efficiency is higher than the original fan.