2.4 Hydrogen-Induced Cracking (HIC)
Key Takeaways
- HIC requires THREE simultaneous conditions: susceptible microstructure + hydrogen + tensile stress
- Remove any one condition to prevent cracking — this drives all prevention strategies
- HIC is "delayed" — can appear 48–72 hours after welding as hydrogen diffuses and accumulates
- Main hydrogen sources: electrode/flux moisture, surface contaminants, atmospheric humidity
- Prevention: preheat, low-hydrogen consumables, proper storage, PWHT/post-heat, clean surfaces
- Codes may require 48-hour delay before final NDE to allow HIC to develop before inspection
2.4 Hydrogen-Induced Cracking (HIC)
Hydrogen-induced cracking (HIC), also called cold cracking, delayed cracking, or underbead cracking, is the most common and most dangerous cracking mechanism in structural steel welding. It can occur hours or even days after welding is completed, making it particularly insidious.
Three Required Conditions (All Must Be Present)
HIC requires the simultaneous presence of all three conditions. Remove any one, and cracking will not occur.
- Susceptible microstructure — typically martensite in the HAZ
- Sufficient hydrogen — diffusible hydrogen (>4–5 mL per 100g of deposited weld metal)
- Adequate tensile stress — residual stress from welding, plus any applied load
Sources of Hydrogen in Welding
| Source | Mechanism | Prevention |
|---|---|---|
| Moisture in electrode coatings | Cellulose and organic compounds in flux decompose | Use low-hydrogen electrodes; proper storage and reconditioning |
| Moisture in flux | SAW bonded/agglomerated flux absorbs moisture | Dry flux per manufacturer recommendations |
| Surface contaminants | Oil, grease, paint, rust contain hydrogen compounds | Clean joint surfaces before welding |
| Atmospheric humidity | Moisture in air near the arc | Shield the weld area; avoid welding in rain |
| Shielding gas moisture | Contaminated gas cylinders or wet hoses | Use dry gas; check dew point of gas |
Characteristics of HIC
| Feature | Description |
|---|---|
| Location | Typically in the HAZ (underbead), but can be in weld metal |
| Orientation | Usually perpendicular to the weld axis (transverse) |
| Timing | Delayed — can appear hours or days after welding |
| Temperature | Occurs below ~400°F (200°C) during cooling |
| Appearance | Tight, often microscopic cracks; may not be visible to naked eye initially |
Prevention Strategies
| Strategy | How It Works |
|---|---|
| Preheat | Slows cooling → less martensite; keeps temperature above hydrogen mobility threshold |
| Low-hydrogen consumables | E7018, E71T-8 minimize hydrogen input (<4 mL/100g) |
| Proper electrode storage | 250–300°F holding oven; recondition if exposed >4 hours |
| Post-weld heat treatment (PWHT) | Tempers martensite; drives out residual hydrogen |
| Post-heat (hydrogen bake-out) | Maintain 400–500°F for 1–2 hours immediately after welding |
| Joint design | Minimize restraint; allow for contraction |
| Clean joint surfaces | Remove moisture, oil, paint, rust before welding |
| Controlled interpass temperature | Maintain minimum temperature throughout multi-pass welding |
The "Delayed" Nature of HIC
HIC is called "delayed cracking" because hydrogen is mobile in the steel lattice and takes time to:
- Diffuse from the weld metal to the HAZ
- Accumulate at high-stress locations (grain boundaries, martensite interfaces)
- Build sufficient pressure to initiate a crack
This process can take 48 to 72 hours after welding. This is why codes require a 48-hour minimum delay before final NDE on certain joints — to allow any HIC to develop before inspection.
For the Exam: HIC is the #1 tested cracking topic. Remember the three required conditions: susceptible microstructure + hydrogen + tensile stress. All three must be present simultaneously. Remove any one to prevent cracking.
Which THREE conditions must ALL be present simultaneously for hydrogen-induced cracking to occur?
Why is hydrogen-induced cracking called "delayed cracking"?
Which of the following is the MOST effective way to reduce the risk of hydrogen-induced cracking?