Construction and Application of a Likelihood Function for Pitting Corrosion with Confidence Bounds for Structure Integrity

The impact of pitting corrosion on structural safety operation and aging control of corroded structure is essential for maintenance and replacement schedule at the right time, from both a safety and economic point of view. This paper focuses on statistical analysis of pitting growth rate and lifetime predictions of structure integrity for pitting corrosion initiating at structure surface. An improved new flexible stochastic methodology based on heterogeneous Poisson process is developed to simulate of pit initiation, growth rate, and death to forecast maximum pit depth impacted structural integrity. The pitting corrosion prediction model is based on data obtained on a 304 stainless steel under atmospheric corrosion conditions. The model fits the data very well, and estimates of the maximum pit depth for future times are reasonably accurate. Specifically, the construction of a nominal 95% confidence interval that covers the future maximum depth with high probability, including accounting for various errors in the measurement and other statistical random errors.