Probabilistic model based on Markov chain for the assessment of localized corrosion of stainless steels

Pitting corrosion depends on a variety of factors related to the metal (chemical composition, differences in the metallurgical structure, inclusions), the environment (chloride content, pH, temperature, differential aeration) and the geometry of the system. Due to their unpredictable occurrence, localized corrosion events can not be explained without using a proper statistical method. In this work a probabilistic approach based on Markov chains for the assessment of pitting and crevice corrosion initiation is proposed. Formally, a Markov chain is characterized by a set of states and a transition probability matrix. The model calculates the probability to have pitting (and vice versa to maintain a stable passive condition) involving a large number of operating parameters related to both metal and environment. Experimental tests were carried out to validate the model which requires more accurate investigations.