**Why the delay matters for QC.**
By the time a crack appears, the weld has already passed visual inspection, been released for the next fabrication step, and sometimes even passed RT or UT. The study confirms that delayed cracking occurs because hydrogen atoms move slowly through the steel lattice and accumulate in areas of high stress and strain over time, not during the initial cooling phase.
**What promotes HACC in multi-pass welding?**
The study identifies a critical relationship: multi-pass thermal history redistributes hydrogen. Each pass reheats previous weld metal, which drives hydrogen effusion outward. But the direction and concentration depend on three variables:
1. Number of weld passes — more passes create more redistribution cycles
2. Heat input — higher input drives hydrogen deeper into the HAZ
3. Interpass temperature — controlled interpass reduces the peak hydrogen concentration
From the paper: "Multi-pass welding promotes hydrogen diffusion and results in lower local concentrations due to these reheating effects. However, if interpass temperature is not controlled, hydrogen redistribution becomes unpredictable and local concentrations can exceed the critical threshold."
**The TEKKEN test gap.**
The standard cold cracking test (DIN EN ISO 17642 / TEKKEN test) uses a single-pass Y-butt weld geometry, which does not replicate the multi-pass redistribution behavior. The authors argue that a practice-oriented test method better reflects real production conditions for thick-section high-strength steel fabrication.
**Field implication.**
If you are welding S690QL or S960QL in bridge or crane fabrication, do not release the joint for NDT immediately. Allow a 48 to 72 hour hydrogen dwell before final UT or RT. This is not conservative. The study shows that delaying inspection catches cracks that would otherwise be missed.
Source: Miedlig, H., Wei, X., Zinke, M. et al., Welding in the World, 2026.