PREDICTING CAVITATION EROSION ON TWO-STAGE PUMPS USING CFD

Abstract

Cavitation is a common problem that occurs in pumps which reduces its useful life and bring increased operating costs to the user. A study of cavitation erosion on a two-stage centrifugal pump has been carried out using Computational Fluid Dynamics (CFD). Most cavitation studies on pumps have been focused on modelling the severity of cavitation; specifically, on understanding its visual effects and performance penalties. Few works have been carried out to predict the most erosion-sensitive areas inside a pump. The focus of this study is on modelling the permanent damage that would be caused by cavitation and to identify specific areas within the pump which are most susceptible to erosion. The model is first validated against experimental data from another work. Once the simulation has been successfully calibrated, the cavitation simulation is carried out again with the subject pump. Not only does this work extend the findings previous works by predicting cavitation erosion on a two-stage pump, but the pump rotation speed is also varied to observe how the erosion-sensitive areas on the pump changes as a result. A specific focus on the Gray Level Method is carried out to predict the erosion damage on the pump. This technique is chosen as it has been experimentally proven with single-stage radial pumps, using specialized CFD code. It is found that the algorithm used to predict erosion when applied with commercial CFD packages, are useful in distinguishing areas inside the pump which are most vulnerable to erosion damage. The Scherr-Sauer cavitation model coupled with the κ-ω SST turbulence model have been used to run the cavitation simulations.