A numerical study on the effect of pit’s size with a random geometry created by cellular automata on corrosion fatigue life prediction

نوع: Type: Thesis

مقطع: Segment: masters

عنوان: Title: A numerical study on the effect of pit’s size with a random geometry created by cellular automata on corrosion fatigue life prediction

ارائه دهنده: Provider: Farzaneh soltani vahed

اساتید راهنما: Supervisors: Dr. Rahman Seyfi , Dr. AmirHossein Mahmoudi

اساتید مشاور: Advisory Professors:

اساتید ممتحن یا داور: Examining professors or referees: Dr. Hashem Mazaheri , Dr. Sadegh Ghorbanhosseini

زمان و تاریخ ارائه: Time and date of presentation: 2026

مکان ارائه: Place of presentation: 55

چکیده: Abstract: Localized corrosion and fatigue are among the most critical factors that reduce the safety and service life of metallic structures operating in corrosive environments. Corrosion pits create local stress concentrations on the material surface, which significantly contribute to fatigue crack initiation and the reduction of structural fatigue life. In many numerical studies, pitting corrosion is commonly represented by simplified idealized geometries such as spherical pits. However, real corrosion pits are typically irregular and geometrically complex, and therefore cannot be accurately described by such simplified models. The main objective of this research is to investigate the influence of realistic pit geometry on the stress field and fatigue life prediction, and to evaluate the accuracy of simplified geometric models. In this study, two-dimensional random corrosion pits were generated using a cellular automata method in MATLAB. The generated geometries were then transferred to Abaqus, where finite element analyses were performed under identical loading and boundary conditions. For comparison purposes, models containing spherical pits with the same characteristic sizes were also created and analyzed. After obtaining the stress fields, fatigue life prediction was carried out using FE-SAFE, and the results of the spherical and random pit models were systematically compared. The results showed that corrosion pits induce significant stress concentration around the corroded region, and these areas represent the most critical locations for fatigue crack initiation. It was also found that, within the investigated size range, changes in the size of spherical pits had a relatively limited influence on the maximum stress and fatigue life. In contrast, random pits produced much more severe and localized stress concentrations due to their irregular boundaries, local curvature variations, and sharp geometric features. The comparison demonstrated that the use of realistic random geometries provides a more conservative and more representative prediction of corrosion–fatigue behavior, whereas simplified spherical models tend to overestimate the fatigue life.