Stochastic Modeling of Corroded Surface Features of Structural Steel Under Sulfate Attack

Abstract

Steel structures exposed to the sulfate corrosive environment for a long time will inevitably suffer random corrosion damage, which will lead to uncertain degradation in the mechanical properties of materials and structures. The safety and reliability assessments of corroded steel structures largely depend on the quantification of corrosion characteristics. In order to investigate the corrosion features of structural steel under sulfate attack, six batches of accelerated corrosion experiments were conducted on 18 steel specimens. The surface morphologies of all corroded coupons were first measured by a 3D surface profilometer, and then the surface parameters were calculated/extracted and analyzed by a self-written analysis algorithm to clarify the distribution characteristics and evolution laws of corrosion depth, pit depth, and pit shape. The results revealed that the corrosion form of structural steel under sulfate attack was a rather uneven general corrosion, which exhibited the intersection of general corrosion and pits. The corrosion depth obeyed a normal distribution, and its average value, standard deviation, and the power spectrum peak increased as the corrosion age increased. The depth and aspect ratio of corrosion pits were both in accord with the lognormal distribution and showed the increasing and decreasing trend with the corrosion time going on, respectively. Additionally, the probabilities of different pit shapes in different depth ranges were obtained, in which cone pits accounted for the largest proportion, and the pits gradually changed from cylindrical and hemispherical to conical with the increase of pit depth. In the end, the random models for corrosion depth and pits were established based on the statistical results, which realized the reconstruction and random modeling of corrosion characteristics of structural steel under sulfate attack.