Quantitative evaluation of rail rolling contact fatigue crack using alternative current field measurement

The quantification of internal dimensions in rail rolling contact fatigue (RCF) cracks is essential for preventing catastrophic failures. A coupled forward-inverse computational framework is developed to correlate internal RCF crack parameters with alternative current field measurement (ACFM) signal features. Field measurements conducted across the entire rail service life are used to characterise operational parameter ranges. The forward model, based on Taylor expansion and coefficient vector reconstruction, is designed to map crack parameters to electromagnetic responses. For the inverse problem, graphical analysis and a uniformed multi-initial newton iteration (UMNI) algorithm are formulated to reconstruct crack dimensions. Validation is performed through finite element simulations, robustness tests with 1% and 5% noise, and experimental measurements. Effective quantification of ellipse ratio (ER) and internal angle (IA) is demonstrated, with maximum errors measured as follows: numerical simulations (ER: 0.15, IA: 1.50°), robustness testing (ER: 0.61, IA: 3.74°), and experimental validation (ER: 0.30, IA: 6.30°). The proposed framework is shown to enable accurate and robust determination of internal crack parameters from ACFM response features.