A piece of the tungsten electrode in the weld — the signature TIG and plasma defect. Tungsten is denser than the steel around it, so the particle sinks and stays. It shows up bright white on radiography, and it never goes away by itself.
A tungsten inclusion is a particle of the tungsten electrode embedded in the weld metal during TIG (GTAW) or plasma (PAW) welding. The electrode tip breaks off — a dip, an overheated tip, a contaminated electrode — and the particle lands in the pool.
Tungsten is a hard, brittle metal roughly 2.5 times denser than steel, so the particle sinks into the weld and stays there. It weakens the section, acts as a stress raiser, and can nucleate cracks under load. Codes reject internal inclusions on radiography — which is why "it's only a speck" doesn't survive an RT film.
Tungsten inclusions, tungsten pick-up, tungsten contamination, W inclusion.
Most tungsten inclusions are invisible from the outside — they live inside the weld and only appear on film. Here's what you're looking for.
Tungsten inclusion is an electrode problem, full stop: the tip touches something, overheats, or sheds. Three routes — contact, current, contamination.
The electrode tip runs hot by design and lives a whisker from the pool. Three ways it becomes a defect: one, you dip — the tip touches the puddle or the filler rod and a piece snaps off into the pool. Two, the current is too high for the electrode size and the tip melts, dropping droplets into the weld. Three, the tip is contaminated — oxidized or picked up base metal — and sheds particles steadily as you weld.
Whatever the route, the physics then do the rest: tungsten is about 2.5 times denser than steel, so the particle sinks into the liquid pool and is buried as it freezes. On radiography it appears as a bright white indication — the dense metal absorbs more radiation than the surrounding steel. What's left is a hard, brittle inclusion in the section: a stress raiser that can nucleate cracks under fatigue or restraint.
Prevention is electrode management: right size, right grind, right distance — and the discipline to stop when the tip misbehaves.
Pick the right tungsten size for the current — the amperage guide has the table. Grind the tip correctly: flat tip for DC, balled for AC. Store tungsten clean, away from contamination.
Keep a clean distance between tip and pool. Never touch the filler into the tungsten. Keep stick-out around 2–3 electrode diameters. On aluminum, a gas lens keeps the shield stable.
Inspect the tip. Discolored, contaminated, or wrongly balled — stop and re-grind before the next pass. Use a foot pedal or current ramp to avoid a molten tip at starts and stops.
Welding on with a contaminated tip is the classic mistake — you shed particles for the rest of the run. The moment the tip changes color or touches the pool, stop and re-grind.
The repair is excavation, not hoping. Tungsten doesn't dissolve and it doesn't float out.
| Step | Action | Why |
|---|---|---|
| 1 | Locate the inclusion — RT if it's internal, VT if it broke the surface. | You dig to the deepest particle, and you can't guess depth. |
| 2 | Grind or gouge to the deepest tungsten, plus a margin. | Particles hide in the sides of the cavity, not just the bottom. |
| 3 | Re-weld to the qualified WPS. | The repair pass must meet the same standard as the original. |
| 4 | Re-inspect — RT where the code requires it. | One leftover particle is a new rejection. |
The particle is dense and can sit below where you think you're digging. Verify removal on film before the repair is signed off — grinding blind and re-welding over a hidden particle is the most common failed repair.
Two processes carry a tungsten electrode in the arc — the causes differ, but the electrode is the common thread.
Dipping is the most common route — tip into the pool, or filler fed into the tip. Then: high amps on an undersized electrode, wrong tungsten type, and no gas lens on aluminum. The classic fix is tip discipline: distance, stick-out, and re-grind.
Electrode set-back — the recess inside the torch — set wrong, the electrode overheats and erodes. Add electrode wear over time and contaminated shielding gas. Set-back is a torch setup number, checked before you strike.
Three parameters write the story: current against electrode size, tip geometry, and stick-out. Check them in this order.
| Parameter | Check | Typical Fix |
|---|---|---|
| Current vs. tungsten size | 1/16" 2% thoriated ≈ up to 150 A DCEN; larger sizes for more | Oversize the electrode or lower the current — never run a melting tip |
| Tip geometry | Flat tip for DC, balled for AC | Re-grind to spec before welding |
| Stick-out | 2–3 electrode diameters | Shorten it; long stick-out puts the tip in the hot zone |
| Shielding gas | Correct gas, adequate flow, gas lens on aluminum | A stable shield means a clean tip — and a clean tip doesn't shed |
| Contamination | Tip color and surface | Discolored tip — re-grind before the next weld |
The Amperage Guide has the current-to-tungsten-size tables, and the Filler Metal Selector covers filler choices that keep your feed hand clear of the tip.
The torch stack decides inclusion risk — from the electrode material to the collet that holds it.
Torch setup decides inclusion risk: collet and collet body condition, gas lens fit, and torch seals. Consumables: tungsten types — thoriated, ceriated, lanthanated — each with its own current capacity and grind requirement, plus collets, collet bodies, and gas lenses sized to the electrode.
Tungsten inclusion is an internal defect on a process that looks clean on the surface — the film tells the truth.
Acceptance depends on the code and project: AWS D1.1, ISO 5817, API 1104 and project specifications each set their own limits for this discontinuity. Check the governing document before judging a weld acceptable — there is no universal pass/fail rule.
The questions welders actually ask about tungsten inclusion — answered straight.
Under most codes, yes. Tungsten is a hard, brittle inclusion that weakens the section and can nucleate cracks; codes reject internal inclusions on radiography. Grind to the deepest particle, re-weld, and verify removal.
Usually stick-out or torch control: too much stick-out puts the tip in the hot zone, and feeding filler too far forward brings it into the tip. Shorten stick-out to 2–3 electrode diameters, keep the filler out of the arc, and steady the torch hand.
By current. As a guide, a 1/16" 2% thoriated electrode runs up to about 150 A DCEN — the amperage guide has the full tables for every type and polarity. Oversize the electrode when in doubt; an electrode that never runs hot never melts.
Yes — any TIG or plasma weld on any material: steel, stainless, aluminum, nickel alloys. The contamination is either embedded particles or surface tungsten smears, and both are unacceptable.
No. It is a hard, brittle, non-fusing inclusion that reduces section strength and can start a crack under load. On radiography it is a bright indication any inspector will flag — remove it.