Abstract
Being immersed in aqueous solution with Cl-- ions, specimens made from a low alloy steel (A537cl1) or from a stainless steel (1Cr18Ni9Ti) were tested under cyclic load. The tests were interrupted on schedule to measure the diameters and depths of the pits on the surface of the specimens. Then, by using nonlinear regression method, these experimental data were expressed with five kinds of probability distribution functions. Among them log normal function was demonstrated to be the optimal distribution function with the highest correlation coefficient ( > 0.99). Moreover, the probabilistic distributions and their evolution behaviors seem to coincide with the stochastic model proposed by the authors and Pickering's evolution theory of pitting.