4.2 Weld Metal and HAZ Discontinuities
Key Takeaways
- Lack of fusion and incomplete penetration are planar bonding failures—often better detected by UT than by RT when tight and unfavorably oriented
- Porosity is volumetric and gas-related; slag inclusions track flux processes and incomplete interpass cleaning
- Undercut and underfill are geometric profile defects; VT is primary, with codes limiting depth and length
- Hot (solidification) cracks differ from cold/hydrogen/HAZ cracks in timing, location, and morphology—critical for method timing and technique
- Process→flaw mapping (e.g., SMAW→slag, GTAW→tungsten, damp electrodes→hydrogen cracks) is a core Basic exam skill
4.2 Weld Metal and HAZ Discontinuities
Quick Answer: Weld discontinuities are process-linked flaws in weld metal, fusion line, or HAZ—lack of fusion, incomplete penetration, porosity, slag, undercut, underfill, cracks (hot, cold/hydrogen, HAZ), incomplete sidewall fusion, and tungsten inclusions. Detection depends on morphology: volumetric voids favor RT; planar cracks and LOF often favor UT; surface openings favor VT, PT, and MT.
This section is high-yield for the Basic exam: questions frequently present a process or morphology and ask for the discontinuity name, cause, or best method. Pair this material with Section 4.1 process knowledge.
Lack of Fusion and Incomplete Penetration
Lack of Fusion (LOF)
Lack of fusion is failure of weld metal to fuse completely with base metal or with prior weld beads. Forms include:
- Sidewall lack of fusion — bead does not wet/fuse the groove face.
- Interpass lack of fusion — successive beads do not fuse to each other.
- Lack of fusion at root or toe — incomplete bonding at critical stress raisers.
Causes: low heat input, wrong travel angle, excessive travel speed, improper joint prep (rust, paint, mill scale), wrong torch/gun angle, or short-circuit modes on heavy sections. LOF is typically planar and may be tightly closed—RT contrast can be poor if the plane is not aligned with the radiation path. Angle-beam UT oriented to reflect from the unfused plane is often more reliable for critical planar LOF.
Incomplete (Lack of) Penetration
Incomplete penetration means the weld did not reach the required depth or root fusion for the joint design (e.g., root faces not consumed in a single-V groove intended as CJP). It leaves an unfused root plane that acts like a crack under load. Distinguish design-intent PJP (partial penetration allowed by drawing) from rejectable incomplete penetration on a CJP joint. VT of the root (when accessible), RT, and UT root scans are common detection approaches.
Porosity
Porosity is trapped gas voids in solidifying weld metal—typically rounded or elongated (“wormhole” / piping porosity). Sources:
- Contaminants (oil, rust, moisture, paint) on joint or wire.
- Moisture in electrode coatings or flux.
- Inadequate or turbulent shielding (wind on GMAW/GTAW).
- Excessive arc length or wrong gas.
- Base metal chemistry (e.g., high sulfur can contribute to gas issues in some systems).
| Porosity form | Appearance | Common context |
|---|---|---|
| Scattered | Random round voids | General contamination or gas issues |
| Clustered | Local group of voids | Local contamination or stop/start |
| Linear / aligned | Chain along axis | Solidification path gas |
| Piping / wormhole | Elongated tunnels | Gas evolution during freeze |
| Surface-breaking | Open pores on face | VT/PT accessible |
RT is excellent for volumetric porosity (dark spots on film/digital). UT may lose back reflection or show discrete signals. Surface-open pores: VT/PT. Acceptance codes often limit size, spacing, and accumulated length.
Slag Inclusions
Slag inclusions are nonmetallic solid material trapped in the weld—almost always associated with flux-bearing processes (SMAW, FCAW, SAW) when slag is not fully removed between passes or is poured over by the next bead. Morphology is often elongated or irregular, following bead boundaries or interpass lines.
Detection: RT (slag often less dense than steel → dark indications; shape more irregular than gas porosity); UT (reflectors along fusion or interpass planes). Prevention is process control and thorough slag removal—NDT finds what cleaning missed.
Undercut and Underfill
Undercut is a groove melted into the base metal at the toe (or root) of the weld and left unfilled. It is a stress raiser and fatigue initiator. Causes: excessive current, long arc, wrong travel angle, or too high travel speed. VT is primary; depth/length limits appear in AWS and fabrication codes. MT/PT may also show undercut edges if they trap particles/penetrant, but measurement is visual/mechanical.
Underfill (or insufficient throat/face reinforcement below flush when flush is required) means the weld face or root is below the adjacent surface or below required reinforcement. It reduces section thickness. VT and weld gauges assess profile. Underfill is geometric, not a crack, but can be rejectable under workmanship standards.
Cracks — Hot, Cold, HAZ, and Hydrogen
Cracks are the most severe class of weld discontinuity. Subdivide by mechanism and location.
| Crack type | When formed | Typical location | Drivers | Notes for NDT |
|---|---|---|---|---|
| Solidification (hot) crack | During solidification (liquid films present) | Weld centerline, crater, interdendritic | High S/P, deep narrow beads, high restraint, concave beads | Open or tight; RT/UT/MT/PT depending on surface |
| Liquation crack | HAZ partial melting of segregates | HAZ near fusion line | Susceptible alloys, high heat input | Often fine; surface methods + UT |
| Hydrogen-assisted (cold) crack | Hours after cool-down (delayed) | HAZ (often toe/root), sometimes weld | Diffusible H + hard microstructure + stress | Delayed MT/PT critical; UT for subsurface |
| Reheat / stress-relief crack | During PWHT or high-temp service | Coarse HAZ, notches | Creep-strengthened alloys, restraint | Process history matters |
| Lamellar tearing | During welding of restrained T/corner joints | Base metal under fusion line, rolling plane | Through-thickness strain + poor Z-direction ductility | UT from accessible face; plate quality |
Hot vs Cold — Exam Distinctions
- Hot cracks: form while metal is hot; often centerline, may show oxidized fracture surfaces; linked to solidification chemistry and bead shape (deep/narrow vulnerable).
- Cold/hydrogen cracks: form near ambient after hydrogen diffusion and residual stress act on susceptible (often martensitic) microstructures; delayed hours to days; common at toes and roots of hardenable steels with moisture-bearing consumables and low preheat.
Level III procedures for hardenable steels may require inspection delay (e.g., 24–48 h) before final MT/PT so delayed cracks have time to open. Immediate inspection after welding can miss hydrogen cracks that form overnight.
Incomplete Sidewall Fusion and Related Planar Flaws
Incomplete sidewall fusion is LOF along the groove face—classic in narrow grooves, steep bevels, or when the arc is directed to the center rather than washing the walls. On RT, a straight dark line along the fusion boundary may appear if gapped; tight LOF may be invisible. UT with beams aimed at the sidewall is preferred for critical service.
Related planar issues:
- Cold lap — weld metal rolls over base metal without fusion (often toe).
- Unfused root face on CJP joints — incomplete penetration family.
Tungsten Inclusions
Tungsten inclusions occur almost exclusively with GTAW (and related tungsten-electrode processes) when the electrode touches the pool or filler, or when current is excessive for electrode diameter. Tungsten particles are high density: on radiographs they appear light/bright (opposite of porosity’s dark spots). They may be discrete spots or clusters. UT can also detect metallic inclusions. Prevention: proper electrode grind, stick-out, and technique—not “more amperage.”
Process → Flaw Mapping (Critical for Basic Exam)
| Process / condition | Higher-risk discontinuities |
|---|---|
| SMAW, poor interpass cleaning | Slag inclusions, interpass LOF |
| SMAW/FCAW, wet electrodes, hardenable steel, low preheat | Hydrogen cold cracks in HAZ |
| GMAW short-circuit on thick groove | Lack of fusion, cold lap |
| GTAW electrode dip | Tungsten inclusions |
| Contaminated joint, wind on gas shield | Porosity |
| Excess current / long arc | Undercut, spatter, burn-through |
| High restraint, high S/P, concave crater | Hot cracks, crater cracks |
| SAW multipass, incomplete flux/slag control | Slag, LOF at bead boundaries |
| High heat input on susceptible alloy | Liquation cracking, wide HAZ issues |
| Restrained T-joint on poor Z-ductility plate | Lamellar tearing |
Matching NDT Methods to Weld Flaws
| Discontinuity | Preferred methods | Caveats |
|---|---|---|
| Porosity | RT primary; UT supplemental | Surface pores: VT/PT |
| Slag inclusion | RT, UT | Irregular shape vs round porosity on RT |
| LOF / incomplete sidewall fusion | UT angle beam often best; RT if open gap | Tight planar LOF may miss on RT |
| Incomplete penetration | RT, UT root techniques; VT if root accessible | Know CJP vs PJP design |
| Undercut / underfill | VT (+ gauges) | Profile, not volumetric |
| Surface cracks | MT (ferro), PT (nonferro), VT | Delayed exam for hydrogen cracks |
| Subsurface cracks | UT; RT if sufficiently open/oriented | Orientation vs beam/film critical |
| Tungsten inclusion | RT (bright spots); UT | GTAW history is the clue |
| Lamellar tearing | UT in base metal under weld | Plate quality + joint design |
Magnetic particle (MT) detects surface and near-surface cracks on ferromagnetic welds—excellent for toe cracks and undercut-related openings after proper magnetization direction (field roughly perpendicular to expected crack).
Liquid penetrant (PT) serves nonferromagnetic alloys (stainless, aluminum, nickel) and any nonporous metal for surface-breaking cracks and open pores; surface must be clean and free of coatings that block penetrant.
Radiographic testing (RT) images volumetric density differences through the joint—porosity, many slag pockets, tungsten, and some incomplete penetration. Sensitivity depends on thickness, energy, geometry, and technique quality. Tight cracks and LOF parallel to the beam can be missed.
Ultrasonic testing (UT) finds planar flaws when the beam intersects them—LOF, cracks, incomplete penetration. Procedure design (angles, skips, calibration, DAC/TCG) is Level III territory. Austenitic welds need specialized techniques due to noise and anisotropy.
Visual testing (VT) remains mandatory baseline: profile, undercut, underfill, arc strikes, spatter, overlap, and obvious surface cracks.
Exam Scenarios
- Radiograph shows rounded dark spots scattered in weld metal → porosity (gas), not slag (more irregular) or tungsten (bright).
- Bright white spots on RT of GTAW root → tungsten inclusions.
- Crack at HAZ toe 36 hours after welding quenched-and-tempered steel with damp cellulosic electrodes → hydrogen-assisted cold crack → delayed MT.
- Centerline crack in deep, narrow single-pass weld during cooling from solidification range → hot/solidification crack.
- Straight reflector along groove face on angle-beam UT; RT almost clean → likely tight sidewall LOF.
Study Focus
Separate volumetric (porosity, slag) from planar (LOF, cracks); hot vs cold crack timing and location; RT bright vs dark (tungsten vs porosity); process tables until mapping is automatic.
On a radiograph of a GTAW root pass, several small, sharply defined light (high-density) spots appear in the weld metal. Which discontinuity is most consistent with this indication?
A planar lack of fusion along the groove sidewall is tight and unfavorably oriented for radiography. Which method is generally more effective for detection when access allows?
Which combination most strongly indicates hydrogen-assisted cold cracking rather than solidification hot cracking?
Incomplete removal of flux residue between SMAW passes is most likely to produce which discontinuity in the finished multipass weld?