Statistical analysis of AA-1050 localized corrosion in anhydrous ethanol

• Peculiarities of Al alloy temperature-induced localized chemical corrosion are discussed. • Pre-processing procedure of experimental data is developed. • Statistical extreme value analysis is employed for characterization of pit initiation and growth. • The results allow prediction of expected maximum pit sizes after corrosion reaction. Temperature-induced localized chemical corrosion of AA-1050 in anhydrous ethanol is examined experimentally at 105 °C and 110 °C and the generated data is analyzed based on statistical methods. Initiation times are described by a Weibull distribution whereas pit maxima are proven to be drawn from Gumbel parent distribution which is confirmed by modified Anderson-Darling test. Expected values and Gumbel parameters are correlated to reaction times in order to predict maximum pit sizes. Moreover, alcoholate corrosion reaction mechanism is postulated based on published literature and our observations. The statistical methodology can presumably be applied to other fuel-material combinations.