RECENT DEVELOPMENTS AND CHALLENGES IN FRACTURE MECHANICS–BASED FATIGUE LIFE PREDICTION

Abstract

This systematic review critically evaluates the advances and limitations of fracture mechanics–based fatigue life prediction approaches within the context of structural integrity assessment. Over the past five decades, fracture mechanics has evolved from a linear analytical framework into a multidisciplinary domain that integrates theoretical formulations, experimental validation, computational modeling, and data-driven analytics. Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, this study systematically reviewed and synthesized 214 peer-reviewed research papers published between 1972 and 2024. The analysis reveals that significant progress has been achieved in extending classical Linear Elastic Fracture Mechanics (LEFM) and Elastic–Plastic Fracture Mechanics (EPFM) toward nonlinear, probabilistic, and multi-scale modeling approaches that capture the complex behavior of materials under cyclic loading. Computational developments—particularly finite element analysis (FEA), extended finite element methods (XFEM), cohesive zone modeling, and continuum damage mechanics—have enabled detailed simulation of crack initiation and propagation under mixed-mode and variable amplitude stresses. Experimental advancements such as digital image correlation, acoustic emission monitoring, and thermographic analysis have enhanced the precision of crack growth measurement, enabling robust model calibration and validation. Moreover, the integration of Structural Health Monitoring (SHM), digital twin architectures, and machine learning has established a new paradigm for real-time fatigue prediction and adaptive maintenance in high-reliability structures. Despite these achievements, the review identifies persistent challenges including data inconsistency in fatigue thresholds, theoretical limitations under large-scale yielding, computational cost, and inadequate coupling of environmental and microstructural effects. The findings underscore the need for standardized testing methodologies, open-access fatigue databases, and harmonized probabilistic frameworks to improve predictive accuracy and cross-material applicability. Overall, this review concludes that fracture mechanics–based fatigue life prediction remains an indispensable yet continuously evolving foundation for structural integrity assessment, bridging traditional mechanics with digital innovation to ensure safer, more reliable, and sustainable engineering systems.