C ORROSION is one of the degradation mechanisms in aerospace structures [1]. Among various types of corrosion, pitting corrosion associatedwith the dissolution ofmetal is caused by the breakdown of the passive film on the metal surface and is known to be one of the major damage mechanisms affecting the integrity of many aerospace structures. Pitting corrosion is a complex process, and a fundamental aspect of pitting corrosion failure mechanisms is that they tend to initiate at the micro/nanostructure level [2]. The details of the mechanisms vary with material composition, electrolyte, and other environmental conditions [3]. It is well known that pitting corrosion has a strong effect on the fatigue life of aluminum alloys used in aircraft structures [4–6]. Prediction of pitting corrosion damage is therefore very important for the structural integrity of aerospace materials and structures. Although pitting corrosion has been studied extensively over the past two decades, computational modeling of pit initiation and growth is still open to investigation. Several pitting corrosionmodels exist in the literature [7–13] and are mostly empirical, mechanistic, deterministic, or phenomenological in nature. Pitting corrosion involves repassivation, mass transport, electrochemistry, and other mechanisms that operate at different length scales, thus making modeling a difficult problem. Recently, Burstein et al. [14] demonstrated that pit nucleation occurs at the microscopic level and is random in nature and that somemetals showpreferential sites of pit nucleation. Even though there are experimental studies aimed at understanding the basic mechanisms, the observed complexity is very difficult to interpret, with many controlled variables affecting the pitting process. In general, the corrosion damage modeling should involve not only physicochemical and environmental factors, but also various parameters, random in nature. Therefore, a more realistic computational corrosion damage model should integrate various parameters that are random in nature from electrochemistry, materials science, and probability and statistics. In contrast to mechanistic approaches, computational modeling of pitting corrosion based on local rules with evolving patterns may open up the possibility of getting insight into pitting corrosion from a different point of view. Recently, Pidaparti et al. [15] developed a two-dimensional model for studying pitting corrosion growth based on the cellular automata (CA) approach. However, they did not consider the initiation phase of pitting corrosion in their model. The main objective of this work is to develop an approach to modeling three-dimensional surface-corrosion-damage initiation and growth using specific rules governing the electrochemical reactions in a cellular automata environment. The results of corrosion damage growth obtained from the three-dimensional model are compared with the experimental data obtained from the Center for Materials Diagnostics at theUniversity ofDaytonResearch Institute.
Search from 250 M+ research papersSearch from 250 M+ research papers
Jun 17, 2017
J Jung + 3
Open Access