Discontinuity
Underbead crack
A crack lying in the heat-affected zone beneath a sound-looking weld bead, produced by hydrogen trapped in hard martensite that cracks under residual stress after the joint has cooled.
What it looks like
- Invisible from the surface, since the crack sits under the bead in the base metal
- Roughly parallel to the fusion line, following the hardened band of the heat-affected zone
- Revealed on a macro-etched cross-section as a fine line just outside the weld deposit
- Sometimes linked to a toe crack that has grown until it breaks through to the surface
Causes
- Hydrogen introduced by damp electrode coverings, wet flux, or moisture and oil on the joint
- A hardenable base metal with enough carbon and alloy content to form martensite on cooling
- Fast cooling from inadequate or absent preheat in thick or cold sections
- Tensile residual stress from restraint concentrating in the hardened zone
- Welding in high humidity or on condensation-covered steel without drying the joint first
Prevention
- Use low-hydrogen consumables and control their exposure time out of the holding oven
- Preheat and maintain interpass temperature to slow cooling and give hydrogen time to diffuse out
- Apply post-weld hydrogen bake-out on heavy restrained joints in hardenable steel
- Clean and dry joint surfaces thoroughly, since surface moisture is as much a hydrogen source as a damp rod
How it is found
- UT
- The crack lies below the surface, so ultrasonic angle-beam scanning is the practical field method for finding it at all.
- MT
- Subsurface cracks close to the surface can produce a broad, fuzzy particle indication as the field leaks around them.
- RT
- Only detects the crack when the beam happens to lie along its plane, so a clear radiograph does not by itself clear a joint of underbead cracking.
How inspectors think about it
- This flaw is dangerous precisely because a joint can look perfect on the surface while carrying a crack underneath, so material and procedure control matter more than post-weld looking.
- Cracking may not appear for many hours after welding, which is why inspection of susceptible steels is commonly delayed rather than done straight off the torch.
- Because the crack is in base metal rather than weld metal, an inspector considers whether the base metal chemistry and preheat regime were suited to each other in the first place.
General reasoning only. The accept/reject limits for your work come from the code or specification governing the job, which this site does not reproduce.
Related discontinuities
Cold crack (hydrogen-induced cracking)A crack that appears after the weld has cooled, sometimes many hours later, when dissolved hydrogen collects at defects in a hard microstructure and residual stress finishes the job.Toe crackA crack starting at the junction of the weld face and the base metal, where the abrupt change in section concentrates stress and the heat-affected zone is at its hardest and least forgiving.Transverse crackA crack running across the weld axis, at right angles to the direction of travel, caused by longitudinal shrinkage stress in a weld that is stronger or harder than the metal around it.Lamellar tearingA step-like separation in the base metal below a weld, where rolled-in inclusion layers pull apart because the joint loads the plate through its thickness rather than along its rolling direction.
Can you name it on sight?
The gauge trainer's identification mode shows you a weld and four names. Reading about porosity and picking it out of a line-up are not the same skill.