Design Optimization of Centrifugal Pump Impeller Geometry for Maximal Efficiency and Cavitation Resistance With CFD-FSI Analyses: Full Review

Abstract

Centrifugal pumps play a fundamental role in the industrial, agricultural, and municipal fluid transport systems, since the impeller geometry is associated with hydraulic efficiency, cavitation resistance, and structural stability. In contrast, classical impeller optimization is based on decoupled CFD and finite element analysis (FEA), typically assuming rigid blades rather than accounting for fluid-structure interaction (FSI). Although these models can perform well in simulations, they may underestimate cavitation, blade deformation, and efficiency losses under actual operating conditions, particularly at low flow rates. We review the latest developments in centrifugal pump impeller optimization with specific attention to different mathematical methods, including parameters and approaches in the domain of designing surfaces, computational modelling, and multi-objective optimization. Among recent proposals on centrifugal pump impeller optimization, we have proposed a transition from direct CFD to a two-way coupled FSI solution and the use of artificial neural network (ANN) representation models to mitigate computational costs. As the research literature shows, coupled fluid-structure optimization is still limited, with no recognition of blade deformation effects at cavitation onset and rigid-blade surrogate training being ineffective. A systematic and ANN-driven method to optimize the dynamics in the steady-state CFD-FSI coupling system is suggested to simultaneously maximize hydraulic efficiency and preserve the integrity of the impellers in terms of both their hydraulic operating efficiency and the net positive suction head (NPSHr). We conclude with a brief analysis that transient FSI modelling is required, and multi-fidelity surrogates need to be developed to demonstrate the operation of coupled impellers.