A universal phenomenological model for cavitation erosion based on a modified Weibull distribution

Cavitation erosion accumulation consistently manifests a characteristic S-shaped progression. This study establishes a modified four-parameter Weibull cumulative distribution function (CDF) as a robust phenomenological model to capture this universal temporal pattern. The model demonstrates superior accuracy, achieving a near-perfect fit ( R 2 = 0.9997 ) for standard ASTM G32 Nickel 200. Its robustness is further confirmed across the comprehensive International Cavitation Erosion Test (ICET) database, encompassing results from 15 laboratories using 7 test methods on 6 distinct materials. Beyond superior descriptive capability, the model exhibits a unique diagnostic function: its deliberate non-convergence for an outlier dataset serves as an objective indicator of anomalous test conditions or data quality issues. For engineering practice, the model offers immediate utility. The accompanying R code facilitates rapid implementation, and its parameters directly feed into a predictive maintenance framework, enabling quantitative material selection and service life prediction. This work bridges the gap between high-fidelity simulation and practical engineering decision-support.