Multi-Objective Optimization of a Centrifugal Pump Impeller Using Volumetric Minimum Pressure as a Cavitation Indicator

This study aims to compare the effectiveness of sequential and multi-objective optimization methods in improving centrifugal pump performance and evaluate the use of volume fraction-based minimum pressure as an early indicator of cavitation. Simulations were performed using a single-phase CFD approach to analyze flow characteristics, efficiency, pressure distribution, and the relationship between the local minimum pressure (Pmin) and the net positive suction head required (NPSHr). The results show that the volumetric minimum pressure at a fraction of 0.11% of the impeller volume has a good correlation with NPSHr with a value of R2 = 0.7392, thus being able to represent the cavitation tendency stably. The conversion of Pmin to pressure head (HPmin) shows a consistent trend with NPSHr against flow rate variations, although the absolute values are different. Quantitatively, the sequential optimization method produces an efficiency of 80.25, corresponding to an increase of 0.25 percentage points, and a minimum pressure of 53.03 kPa, representing an increase of 5.32 kPa. In contrast, multi-objective optimization achieves a higher efficiency of 82.20%, representing an improvement of 2.20 percentage points, while the minimum pressure decreases to 41.54 kPa, corresponding to a reduction of 6.17 kPa. These results highlight the trade-off between hydraulic efficiency and minimum pressure characteristics. The proposed volume fraction-based pressure indicator provides an empirically calibrated screening metric for cavitation-related assessment within the investigated design space, while the multi-objective optimization framework enables systematic evaluation of competing hydraulic performance objectives.