Theoretical and experimental analysis of the correlation between magnetic memory signals and cumulative ductile damage of butt weld under low cycle fatigue

Welded structures experiencing superimposed cyclic loads with large plastic strain during a strong earthquake can easily lead to ductile or fatigue fracture. A newly developed self-magnetic flux leakage (SMFL) nondestructive detection technology named metal magnetic memory (MMM) is a valid approach to monitoring the development of fatigue damage in ferromagnetic materials. To investigate the correlation between MMM signals and cumulative ductile damage for butt weld under low cycle fatigue (LCF), combining strain-based JilesAtherton hysteresis model, magnetic charge model, and cumulative damage mechanics model, the theoretical relations between normal component HSF(z) of MMM signals and cumulative ductile damage D for single V groove butt weld were established in this paper. Relevant monotonic tensile tests and strain-controlled LCF tests were conducted, and the HSF(z) signals on the surface of welding specimens under different cyclic loads were measured. Comparing with experimental results verified the feasibility and accuracy of the theoretical relations. The results indicated that the peak-valley gradient Kp-v on the HSF(z) signals curve was decreased monotonously as the cumulative ductile damage D increased and the rate of decrease also diminished gradually. A possible exponential descending tendency of Kp-v with increasing D was demonstrated. Besides, the influences of weld width and weld reinforcement on the correlation between Kp-v and D were discussed theoretically in detail. This research provided a potential possibility for the assessment of the cumulative ductile damage at the welding joints under LCF based on MMM technology.