Design Optimization of Cryogenic Pump Inducer Considering Suction Performance and Cavitation Instability

Abstract

This paper presents the result of design optimization for three-bladed pump inducer using a three-dimensional (3-D) inverse design approach, Computational Fluid Dynamics (CFD) and DoE (Design of Experiments) taking suction performance and cavitation instability into consideration. The parameters to control streamwise blade loading distribution and spanwise work (free vortex or non-free vortex) for inducer were chosen as design optimization variables for the inverse design approach. Cavitating and non-cavitating performances for inducers designed using the design variables arranged in the DoE table were analyzed by steady CFD. Objective functions for non-cavitating operating conditions were the head and efficiency of inducers at a design flow (Qd), 80% Qd and 120% Qd. The volume of the inducer cavity region with a void ratio above 50% was selected as the objective function for inducer suction performance. In order to evaluate cavitation instability by steady CFD, the dispersion of the blade surface pressure distribution on each blade was selected as the evaluation parameter. This dispersion of the blade surface pressure distribution was caused by non-uniformity in the cavitation length that was developed on each inducer blade and increased when the cavitation number was reduced. The effective design parameters on suction performance and cavitation instability were confirmed by sensitivity analysis during the design optimization process. Inducers with specific characteristics (stable, unstable) designed using the effective parameters were evaluated through experiments.