Effect of corrosion morphology on the mechanical strength and ductility of structural steel: Experiments and simulations

Corrosion represents a significant form of degradation in steel bridges, profoundly affecting surface morphology, cross-sectional area, and mechanical integrity. In experimental settings, researchers often simulate prevalent types of corrosion, such as uniform and pitting corrosion, through artificial techniques like milling or drilling. Although these simulations can effectively replicate strength reduction, they often fail to accurately reflect the extensive loss of ductility. This study presents the findings from monotonic tensile tests conducted on dog bone specimens fabricated from A572 structural steel that underwent natural corrosion processes. The as-received steel samples, with the mill scale intact, were subjected to a corrosive environment under controlled conditions, including temperatures of 26 °C, 40 °C, and 50 °C, and chloride concentrations of 0%, 1%, 2%, and 3%. Over a period of 16 months, samples were periodically retrieved at 4, 9, and 16 months to establish three distinct corrosion levels. Both pitting and uniform corrosion were observed. The experimental results indicated a nearly linear increase in weight loss over time, irrespective of temperature, while the severity of pitting increased nonlinearly. A significant reduction in elongation was observed, which was attributed to sample irregularities, even in cases of uniform corrosion—phenomena that artificial methods do not replicate accurately. Additionally, a systematic finite element modeling approach was employed, tailored to each corrosion type, to simulate the observed mechanical degradation. Model inputs included weight loss, total corroded surface area, degree of pitting for pitting corrosion, and an irregularity factor (IR) for uniform corrosion. The calibration of the IR showed that it was minimal for samples corroded at higher temperatures and relatively larger for those corroded at lower temperatures, indicating that smoother surfaces at elevated temperatures will indeed have higher ductility as shown experimentally. The finite element model demonstrated a strong agreement with experimental data relevant to a wide range of strength and ductility reductions. Collectively, experiments and simulations emphasize that corrosion-induced reductions in member strength could be predicted by assuming a reduced cross-sectional area as a function of weight loss. However, reductions in ductility cannot be accurately predicted without considering the irregularity in the corrosion morphology.