In a previous publication, the use of a simple two dimensional finite difference model to predict hydrogen distribution in a simulated part wall pipeline girth weld repair was demonstrated. The validity of the model prediction was assessed using slow bend tests where the deflection to failure was an indirect indicator of the local bulk hydrogen concentration at the critical site for delayed cracking. In the present investigation, the same hydrogen diffusion model has been used to predict the maximum time for delayed cracking in a simulated girth weld repair for two ambient temperatures, (20° and −10°C). These predictions are compared with the delayed cracking times in “constant deflection” tests performed at the selected ambient temperatures.

The results obtained showed that there was some scatter in the relationship between the experimentally applied deflections and the delay time to crack initiation. The delay time predicted by the model for the local hydrogen concentration to reach its maximum value at the crack initiation site, however, was a function of the ambient temperature and these agreed reasonably well with the maximum delay times to cracking observed in the experiments.

This content is only available via PDF.