Author:

Keywords:

Science & Technology, Physical Sciences, Electrochemistry, Stainless steel, Cyclic voltammetry, Pitting corrosion, Survival analysis, CORROSION BEHAVIOR, LOCALIZED CORROSION, STAINLESS-STEEL, EQUATIONS, INITIATION, GROWTH, PITS, 02 Physical Sciences, 03 Chemical Sciences, 09 Engineering, Energy, 34 Chemical sciences, 40 Engineering, 51 Physical sciences

Abstract:

The Pitting Resistance Equivalent Number (PREN) is widely applied to assess pitting corrosion resistance of stainless steel, despite containing only a few compositional elements. A reformulation of the PREN was established by statistical analysis of a public pitting potential data set. Statistical pitfalls were identified and properly addressed by multiple imputation and principles from survival analysis, such as censoring and mixed effects. This statistical framework allowed for simultaneous inclusion of surface pretreatment and test environment next to alloy composition in an updated PREN. Comparison of Weibull and non-parametric baseline hazard functions indicated that pitting potentials might not be Weibull distributed.