A Systematic Literature Review of Vibration-Induced Fatigue Assessment in Ship Structures: Publication Trends, Research Themes, Bibliometric Insights, and Future Research Directions (2002–2026)

Abstract

One of the main causes for deterioration of marine structures and reduction of their service life is the fatigue by vibration. Fatigue damage accumulates over time due to cyclic loading from ocean wave and machinery excitation, hydroelastic response and operational vibration in service, with implications for structural integrity, operational safety and maintenance efficiency. There has been a dramatic increase in the research on this problem in the last two decades. It now includes conventional fatigue assessment, numerical simulation, structural health monitoring, machine learning and more recently, Digital Twin technology. A broad synthesis of the development of this literature is still missing. This study fills the gap with a systematic literature review (SLR) based on the bibliographic metadata of 75 scientific publications published from 2002 to 2026. This review is reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) guidelines to ensure transparency and reproducibility of the review process. Instead of full-text findings, the study synthesizes metadata and focuses on the publication characteristics, classification of themes, methodology development, bibliometric patterns, and emerging directions of research. The literature is classified into eight research themes based on the analysis: (1) fatigue assessment methods, (2) wave-induced fatigue, (3) hydroelasticity and hull girder vibration, (4) vibration-induced fatigue of local structural components, (5) numerical simulation and finite element analysis, (6) structural health monitoring, (7) artificial intelligence and machine learning applications, and (8) fatigue prediction based on Digital Twin. The results indicate a transition from deterministic and physics-based fatigue assessment to intelligent, data-driven and real-time monitoring approaches. Recent research has applied machine learning, sensor-based monitoring, and virtual models of ship structures. Few studies combine these into one predictive maintenance framework. Several gaps remain: real-world operational ship data is scarce, physics-based models are rarely integrated with artificial intelligence, no comprehensive Digital Twin framework exists for fatigue prognosis, multisensor data fusion is limited, and validation against long-term onboard monitoring data is rare. These gaps clearly indicate opportunities for future work on intelligent, explainable, and adaptive fatigue prediction systems that can support predictive maintenance and decision making in maritime engineering. The results offer a snapshot of the state of vibration-induced fatigue research to researchers, naval architects, classification societies and industry practitioners and provide a foundation for future work towards Digital Twin-enabled structural health monitoring and remaining useful life prediction in ship structures.