Acoustic emission from cracking in a weldment subjected to non-uniform residual stresses

Cracking in a welded structure can cause catastrophic failure, and crack growth monitoring is thus an important and challenging task in the quality assurance of weldments during and after welding processes. In this research, the acoustic emission generated from a cracking source in a weldment has been investigated theoretically and experimentally.;In the theoretical analysis, a general analytical expression for the relationship between acoustic emission surface motion and crack opening displacement was derived based on the analogy of earthquake problems in seismology and crack opening in fracture mechanics. A crack-tip equation of motion was then developed for a rectangular, through crack due to a non-uniformly distributed residual stress in a weldment. The dependence of dynamic stress intensity factor and dynamic fracture toughness on crack speed was considered, as well as the effects of a bluntness parameter and the difference between initiation toughness and arrest toughness. Finally, the surface motion was obtained as a function of instantaneous crack length. In the time and frequency domains, closed-form, far-field surface motion for the constant crack speed model has been compared with the numerical simulation for the variable crack speed model. Some valuable trends have been observed in the simulation of surface motion using the variable crack speed model.;In the experimental investigation, acoustic emission signals have been recorded and analyzed during and after gas metal arc welding. In addition to the study of cold cracking of a weldment in air, hydrogen embrittlement was induced using the electro-chemical method to cathodically charge weldments with hydrogen in order to accelerate cold cracking. The spectrum of the corresponding acoustic emission signal has been compared with theoretical results. Reasonable agreement was obtained.;The outcome of the study will improve our understanding of the generation of acoustic emission by cracking in weldments and enhance the capability of pattern recognition analysis in the frequency domain.