Ordinary state-based peridynamic framework for pit-to-crack activation, corrosion–fatigue, and crack-fluid pressure in marine structures

The article reports a compact bond-topology verification benchmark and a direct mechanical validation and an externally benchmarked environmental rate-transfer assessment for an ordinary state-based peridynamic (OSB-PD) framework for marine corrosion-fatigue. First, a linearized two-dimensional plane-stress C(T) model is evaluated without mechanical fitting against published S355G10+M back-face-strain calibration relations and the ASTM C(T) stress-intensity relation. Across 12 values evaluated from the published experimental calibration relations, back-face strain gives a mean absolute percentage error of 8.06% and R 2 = 0.891; the stress-intensity benchmark gives 1.96% and R 2 = 0.993. Second, environmental-transfer parameters are fixed from one 9 kN, 0.2 Hz seawater condition, and five other load-frequency conditions are predicted without retuning. Across 85 common-support holdout points, the log10 (da/dN) RMSE is 0.140 and 95.3% lie within a factor of two. Across the full 110-point holdout domain, RMSE is 0.1753, 95.45% remain within a factor of two, and common-support coverage is 77.27%. The compact benchmark separately exercises pit-to-crack activation, corrosion-fatigue, Morison forcing and a reduced crack-internal pressure state. The supported scope covers cracked-specimen elastic mechanics, crack-driving-force scaling and an externally benchmarked corrosion-fatigue rate-transfer closure. Pit nucleation, pit-to-crack timing, CIPP, the Sleipner illustration and field-scale offshore prediction remain outside the validated claim.