Corrosion damage and long-term performance evolution of coastal engineering structures: mechanisms, challenges, and protection

Abstract

Focusing on the corrosion damage and long-term performance evolution of coastal engineering structures, this paper presents a comparative review of four structural systems – steel, reinforced concrete (RC), steel-reinforced concrete (SRC), and concrete-filled steel tubular (CFST) structures – under marine environments, highlighting their distinct degradation characteristics and common challenges. Chloride-induced electrochemical corrosion is identified as the common starting point for performance degradation, while each system exhibits distinct features: steel structures by stress-corrosion-fatigue coupling; RC structures by rust expansion-cracking feedback; SRC structures by interface bond degradation; and CFST structures by tube corrosion and confinement loss. Research advances are categorized into multi-field coupled degradation models, probabilistic performance assessment, and intelligent protection technologies. Studies indicate that the coupling of chloride ingress with carbonation, freeze-thaw, wet-dry alternation, and fatigue loading drives a cross-scale degradation chain from material deterioration to system reliability decline. Recent progress in multi-field modeling, data analytics, and artificial intelligence, coupled with high-performance protective materials and intelligent maintenance, has enabled a paradigm shift from periodic inspection towards predictive maintenance. However, current research faces core limitations: insufficient understanding of time-varying degradation mechanisms under multi-factor coupling, incomplete analytical theories for complex loads and multi-hazard sequences, and a lack of unified synergy among protective materials, design methods, and maintenance standards. These constraints hinder engineering applications. Future research should prioritize time-varying constitutive models under multi-physics coupling, intelligent prediction and digital twin methods fusing monitoring data with machine learning, and unified design and assessment frameworks for whole-lifecycle performance, facilitating the transition from experience-based protection towards a predictable, designable, and manageable durability assurance system.