3.4 Incomplete Fusion and Incomplete Joint Penetration
Key Takeaways
- Incomplete fusion = no bond at a fusion interface (sidewall, interpass, or root face)
- Incomplete joint penetration = weld metal does not extend through the joint root of a groove weld
- IJP is a defect only in CJP groove welds; in PJP welds limited penetration is by design
- Both are planar, crack-like discontinuities — read the welding symbol/WPS before judging the root
- UT is generally superior to RT for these tight planar flaws because of beam orientation
Two Planar Discontinuities That Are Often Confused
After cracks, incomplete fusion and incomplete joint penetration are the most serious discontinuities, because both are planar — they create flat, crack-like, unfused interfaces that act as severe stress concentrators. The CWI exam loves to test the distinction, because the everyday terms ("lack of fusion," "lack of penetration") are loose, while AWS A3.0 is precise. The key mental hook: incomplete fusion is about bonding at an interface; incomplete joint penetration is about how deep the weld reaches into the joint root.
Incomplete Fusion (IF / Lack of Fusion, LOF)
" There is a physical, unbonded separation even though the area may look filled. It can occur anywhere a fresh deposit should have melted into solid metal, and because the unfused surfaces are nearly in contact it behaves like a crack under load. A common practical clue: a bead that looks full and well-shaped on the surface can still hide sidewall or interpass lack of fusion beneath it, so VT cannot be trusted to clear it — the planar gap is internal. This is why incomplete fusion is one of the discontinuities most likely to escape detection and cause in-service failure if the wrong NDE method is used.
| Type | Where | Description |
|---|---|---|
| Sidewall (groove-face) LOF | Between weld metal and the prepared groove face | Arc didn't melt the sidewall |
| Inter-run (interpass) LOF | Between adjacent weld beads/passes | New pass didn't fuse to the previous bead |
| Root LOF | At the root face | First pass didn't fuse the root face |
Causes of incomplete fusion
| Cause | Mechanism |
|---|---|
| Insufficient heat input / low amperage | Not enough energy to melt the adjoining surface |
| Improper electrode/torch angle | Arc force directed away from the sidewall (a top sidewall-LOF cause) |
| Travel speed too fast | Surface doesn't reach melting before the pool passes |
| Excessive weave width | Arc dwells at center, leaving sidewalls cold |
| Magnetic arc blow | Arc deflected off the intended fusion line |
| Surface oxide / contamination / mill scale | Oxide barrier prevents fusion even under the molten metal |
| Cold lap (puddle ahead of arc) | Molten metal flows onto unmelted base, common in vertical-down |
Incomplete Joint Penetration (IJP)
Per AWS A3.0, incomplete joint penetration is "a joint root condition in a groove weld in which weld metal does not extend through the joint thickness." In other words, the weld did not reach as deep into the root as the joint design requires. It is judged against the joint type: in a complete-joint-penetration (CJP) groove weld the weld must fuse through the full thickness, so any IJP is a defect; in a partial-joint-penetration (PJP) groove weld a designed, limited penetration is intentional, so unwelded root below the specified depth is normal and not a defect.
Causes of incomplete joint penetration
| Cause | Mechanism |
|---|---|
| Root opening too small | Arc cannot reach the root |
| Root face (land) too thick | Excess metal blocks penetration |
| Insufficient amperage / heat input | Not enough energy at the root |
| Electrode too large for the groove | Can't physically access the root |
| Travel speed too fast | Inadequate root heating |
| Groove angle too narrow | Restricts arc access to the root |
| High-low (misalignment) at the root | Offset root faces prevent full fusion |
The Critical CJP vs. PJP Distinction
| Joint | Full penetration required? | Is IJP a defect? |
|---|---|---|
| CJP | Yes — weld extends through full thickness | Yes, any IJP is rejectable |
| PJP | No — only a specified depth (effective throat E) | No — limited penetration is by design |
This is why the inspector must read the welding symbol and the WPS/drawing before judging root penetration: the same radiographic image of an unfused root is a reject on a CJP weld and acceptable on a PJP weld. Note also that IJP and root LOF often appear together on a radiograph as a straight, dark line along the root — distinguishing them matters for root-cause and repair, but both are planar and both are treated seriously.
Detection
Both IF and IJP are subsurface planar flaws, so VT alone usually cannot find them (unless they break the surface). Ultrasonic testing (UT) is the most reliable method because these flat reflectors return strong echoes when the sound beam strikes them near-perpendicular; radiographic testing (RT) can find IJP (a sharp straight line) and root LOF but may miss tight, favorably-oriented sidewall LOF because the planar gap can be too thin to register on film. This orientation sensitivity is a frequent exam point.
Exam essentials: Incomplete fusion = no bond at an interface (sidewall, interpass, root). Incomplete joint penetration = weld doesn't reach through the root of a CJP groove. IJP is a defect only in CJP welds. UT is generally superior to RT for these tight planar flaws.
What is the PRIMARY difference between incomplete fusion and incomplete joint penetration?
Incomplete joint penetration is a rejectable defect in which type of groove weld?
A welder is consistently producing sidewall lack of fusion in a groove weld. Which corrective action is MOST likely to fix it?