Gas pockets trapped in the solidifying weld metal. It is the most common weld rejection on the shop floor — and the most preventable, once you check gas and cleanliness in a fixed order.
Porosity is a cavity-type discontinuity: bubbles of gas trapped inside the weld metal as it solidifies. The gas is generated in the arc and pool faster than it can escape to the surface before the pool freezes.
Porosity rarely fails a weld on its own — but it reduces the effective cross-section, can be the surface symptom of contamination that also causes cracking, and it is the number-one reason welds get rejected in RT. It is also a diagnostic gift: the pattern of porosity tells you which side of the process is leaking.
Gas pockets, gas holes, pinholes, wormholes, worm tracks, blowholes, gas porosity.
The pattern of porosity is a clue to its source — read it before you change anything.
Every porosity source is one of three things: a gas generator (contamination, moisture), a gas path failure (shielding, stick-out, leaks), or a gas escape failure (cooling too fast).
Porosity is a solubility story. Molten steel can dissolve a surprising amount of gas — hydrogen and nitrogen in particular — but the solubility collapses when the metal freezes. Hydrogen solubility in steel drops to roughly one-fortieth of its liquid value at the freezing point. When the pool starts to solidify, the dissolved gas is pushed out of solution faster than it can diffuse to the surface and escape. The result: bubbles, and if the pool freezes before they rise out, those bubbles are trapped as porosity.
Three gases dominate:
The most common source. Comes from moisture in the air, on the plate, in the wire, or in flux — and from oil, grease, and paint. It produces fine, even pores and the long worm-track pipes under the slag.
Forms when oxides (rust, mill scale) meet carbon in the molten pool. It makes coarse, irregular blowholes, often at the root of the weld. Clean metal prevents it.
Signals air getting into the arc — failed shielding, leaks, excessive arc length, wind. Produces a fine, uniform distribution and can also cause cracks in some steels. If you see uniform pores, suspect air before moisture.
Escape matters as much as generation. A fast-freezing pool — high travel speed, thick section, cold plate — traps gas that a slower pool would release. That is why the same contamination can produce nothing on warm, thick plate and visible pinholes on a cold, thin one.
Run the porosity check in this order and you will find the cause nine times out of ten before you change a single setting.
Set the flow per the process (MIG/FCAW typically 30–45 CFH, TIG 15–25 CFH with a gas lens). Listen for hissing leaks; soap-test hoses and fittings; check the nozzle is clean and unrestricted. Wind screens below 3 mph of breeze for gas-shielded work.
MIG: 10–15× wire diameter (e.g. 3/8–1/2" on 0.035" wire). Self-shielded FCAW: 3/4–1-1/4". TIG: tungsten sticks out 2–3 electrode diameters. If the shield can't reach the arc, you get air.
Degrease, then mechanically clean — wire brush or grind — until bright metal inside the joint. Strip paint and mill scale back from the weld zone. Galvanized steel: remove the zinc around the joint and ventilate.
Keep wire spools covered and dry. Re-dry low-hydrogen electrodes per the manufacturer's spec. Warm cold plate to keep condensation off the joint — preheat when specified. Dry the area before welding on damp days.
TIG: pre-flow and post-flow 1–2 seconds (2–3 for stainless and aluminum). MIG: don't strike the arc on dirty surface; start ahead of the joint or on a scratch plate. Fill the crater before stopping.
Don't dial up the gas flow to fix porosity. Above the optimum range, turbulent gas drags air into the arc and makes porosity worse. Flow is not a cure-all — it is one link in the gas path.
When porosity is already in the weld, the repair is the same regardless of cause — but only after you have fixed the cause.
| Step | Action | Why |
|---|---|---|
| 1 | Find the cause and fix it before touching the weld. | Re-welding with the same cause reproduces the defect — deeper. |
| 2 | Locate the extent — visual for surface pores, RT or UT for internal. | Internal porosity can extend past the visible area. |
| 3 | Grind or gouge out to sound metal, plus a margin. | Pores hide in the walls of the cavity; excavate past them. |
| 4 | Verify removal — VT, PT, or MT on the excavated groove. | Never re-weld over unconfirmed porosity. |
| 5 | Re-weld to the qualified WPS; clean the groove first. | The repair pass must meet the same standard as the original. |
Never "burn over" porosity to hide it — the gas is still there, now buried deeper where it is harder to find and worse for the joint. Confirm the repair scope against the applicable code before starting.
Each process has its own porosity signature. Read yours against the list.
Number one: gas coverage. Check CTWD, nozzle blockage, contact tip wear, and wind in that order. Short-circuit transfer with low voltage and high stick-out is especially vulnerable. Dirty base metal is the second cause.
Almost always technique or prep: arc length too long, no gas lens on aluminum, short post-flow on stainless, or contaminated filler and base metal. On aluminum, check that the arc starts clean and pre-flow is set.
Moisture: damp 7018 is the classic — re-dry per spec (typically 650–700°F for 1 hour, once). Also long arc length, wrong polarity, and wet base metal. Cellulosic electrodes (6010/6011) tolerate dirty joints but not moisture in the coating.
Self-shielded wire run with the wrong polarity (needs DCEN) or too-short stick-out is instant porosity. Gas-shielded wire run without wind protection. Either family: damp wire, moisture in the core.
Wet or contaminated flux is the main source — recycled flux mixed with rust or fines. Also voltage too high, which draws air through the flux blanket. Store flux dry and follow re-dry procedures.
Plasma gas or shielding gas contamination, worn nozzle orifice, and insufficient shielding flow. On aluminum, oxide on the joint face and insufficient cleaning action produce root porosity.
Check gas first, then these — in this order.
| Parameter | Check | Typical Fix |
|---|---|---|
| Gas flow (CFH) | Set at regulator, verified at nozzle; no leaks | MIG/FCAW 30–45, TIG 15–25; below spec → reseat/repair gas path |
| Stick-out / CTWD | 10–15× wire Ø (MIG); 3/4–1-1/4" (self-shielded FCAW) | Reset to spec; watch the nozzle for spatter buildup |
| Polarity | DCEP for gas-shielded MIG/FCAW; DCEN for self-shielded | Wrong polarity → violent arc, spatter, pores |
| Arc length | TIG/stick: short, stable arc | Long arc sucks air — shorten, steady the hand |
| Travel speed | Pool trails the arc, no undercut, no freeze-in | Too fast → gas trapped; slow to let bubbles escape |
| Pre-flow / post-flow | TIG: ~1–2 s before, 1–2 s after (more for stainless) | Add gas dwell so start and crater stay covered |
| Preheat | Joint above dew point, dry | Cold, wet plate condenses moisture — preheat when specified |
Use the Gas Flow Calculator to set flow and bottle life, and the Heat Input Calculator when travel speed is part of the equation — heat input controls how long the pool stays liquid.
The gas path is the first place porosity lives. Walk it from cylinder to arc.
Check the regulator and flowmeter against a calibrated gauge, the hoses for cuts and loose fittings, the torch for a worn gas diffuser or o-ring, and the nozzle for spatter buildup. Replace worn contact tips — an erratic arc sucks air into the shield. Keep wire spools sealed and electrodes in an oven or rod box.
Porosity is found visually at the surface — and by RT or UT where the code requires internal examination.
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 porosity — answered straight.
Trapped gas in the freezing weld pool — most often hydrogen from moisture or contamination, plus air leaking into the arc through bad shielding, leaks, long stick-out, or wind. Cleanliness, gas path, and stick-out fix the large majority of cases.
In order: check gas flow at the regulator and nozzle, leak-check the hoses, set stick-out to 10–15× wire diameter, replace a worn contact tip, clean the base metal to bright metal, and block the wind. Change one variable at a time and weld a test bead between changes.
No. Grind or gouge to sound metal and confirm removal before re-welding. Welding over porosity buries the gas pockets deeper and guarantees a hidden rejection on the next inspection — usually RT.
No. Internal porosity shows up only on radiography or ultrasonic testing. A bead can look clean and still be rejected — which is why root and fill passes get NDT on critical welds.
Starts: gas hasn't stabilized, the arc strikes on cold or contaminated metal. Ends: shielding leaves before the crater solidifies. Use pre-flow and post-flow on TIG, start on clean metal, and fill the crater with a deliberate stop — not an abrupt lift.
Not necessarily. Above the optimum, turbulent gas pulls air into the arc and makes porosity worse. If flow is already in spec, the cause is almost always elsewhere — cleanliness, stick-out, leaks, or wind.