What Causes Porosity in MIG Welding? Causes & Fixes

A fabricator I once read about ran an entire shift of perfect-looking MIG beads on a batch of bracket assemblies, only for the grinder to reveal a row of tiny gas pockets the moment the first weld was cut open. Nothing on the machine had changed overnight — but the shop door had been left open because it was a hot summer morning, and a fan 25 feet away was quietly blowing the shielding gas off the puddle.

Porosity in MIG Welding: The Short Answer, and Why Pores Form

The fabricator’s story from the start of this article is more common than most shops care to admit. A full shift of beads that looked textbook on the surface, a batch of brackets stacked and ready, and then the first cut-open revealing a neat row of tiny gas pockets. The machine settings hadn’t drifted, the wire was the same spool as yesterday, and the welder’s hands hadn’t changed. A fan 25 feet away and an open shop door did all the damage. That is porosity in a nutshell: the weld looked fine, but gas got into the puddle and froze there.

The short answer

Porosity is trapped gas. AWS A3.0, Standard Terminology for Welding, Cutting and Allied Processes, defines porosity precisely as “cavity-type discontinuities formed in weld metal by gas entrapment during solidification.” Read that once more and the whole problem makes sense: the metal was liquid, gas dissolved into it, and the metal froze before the gas could escape.

Three gases matter: nitrogen, oxygen and hydrogen. All three dissolve readily in molten metal, and when the puddle cools they come back out of solution, much like carbon dioxide escaping a warm soda when you crack the seal. If the surface of the weld is still liquid, the bubbles rise and pop, and you never know they were there. If the surface has already crusted over, the bubbles have nowhere to go. They stay.

The numbers are tighter than people expect. TWI reports that as little as 1% air entrainment in the shielding gas is enough to cause distributed porosity, and more than 1.5% produces gross, surface-breaking pores. One percent of a 25 CFH flow is a teaspoon of air leaking in through a bad O-ring, a porous hose or a draft crossing the arc. You will not see it. You will only see the result.

What MIG welding is, for the newcomers

MIG is gas metal arc welding, or GMAW: a continuously fed solid wire electrode melts into the joint while a shielding gas, delivered through the same gun, blankets the puddle from the atmosphere. That blanket is the only thing standing between white-hot liquid metal and the 78% nitrogen in the air around you. Pull the blanket aside for a fraction of a second and the puddle starts drinking. The process is forgiving and fast, which is exactly why porosity shows up so often in production work: it is the defect most sensitive to everything happening around the arc rather than to the arc itself.

The three families of cause

Almost every porosity problem ever chased on a shop floor falls into one of three buckets:

• Shielding problems — gas flow too high or too low, drafts, leaks, plugged nozzles, gun held too far back, empty or contaminated cylinders.
• Contamination — oil, paint, primer, glue, cutting fluid, moisture, rust, mill scale and zinc on the workpiece, plus dirty wire, a worn liner or a gas line that once carried compressed air.
• Parameters and technique — voltage and wire feed mismatched, excessive stick-out, travel speed too fast or too slow, torch angle that scoops air in from behind, open root gaps and joints the nozzle simply cannot cover.

How this article is organized

We start with the physics — why gas becomes insoluble exactly at the worst moment — then walk the three cause families in the order a welder should check them: gas delivery first, workpiece second, wire and hardware third, settings and technique fourth. From there we compare how steel, stainless and aluminium behave differently (our Aluminum MIG Welding: The Complete Practical Guide goes deeper on aluminium-specific shielding), look at what inspection methods and codes actually accept, and finish with a ten-minute diagnostic workflow you can run before you strike the next arc.

Expect practical numbers throughout: real CFH ranges, real stick-out distances, real cleaning methods such as acetone wipes and dedicated stainless brushing — the kind of prep covered in Aluminum Cleaner for Welding: How to Prep Aluminum Right. Where a claim rests on a standard or a study, you will get the citation, including OSHA’s welding guidance on ventilation and fumes, which matters more once you understand that porosity and fume exposure share a root cause: the shielding blanket isn’t working.

By the end, you should be able to look at a porous bead, name the likely family of cause in under a minute, and fix it without turning the gas flow dial up “just in case” — the single most common mistake shops make, and the one we will dismantle in the final section.

The Science of a Pore: How Gas Gets Trapped in the Weld Pool

A pore is not a hole that opens up in the weld. It is a bubble that was there first, and the metal froze around it. Nothing collapses, nothing drills in. The gas arrives while the steel is liquid, the surface skins over, and the bubble is simply caught mid-escape. Once you picture porosity that way, most of the causes in this article stop feeling random.

Solubility: Gases Go Into Liquid Metal Easily, and Come Out Badly

Hydrogen, oxygen and nitrogen dissolve into molten metal much the way carbon dioxide dissolves into an open can of soda. Liquid steel holds a surprising amount of all three. Liquid aluminium holds even more hydrogen, as Hobart’s aluminium porosity guide explains: hydrogen solubility jumps dramatically the moment aluminium passes its melting point, then plummets the instant it solidifies. That sudden loss of capacity is why aluminium punishes a sloppy technique so much faster than steel.

As the puddle cools and the crystalline structure forms, the dissolved gas is rejected from the solidifying metal, exactly like salt being pushed out of freezing ice. If the surface is still open, the bubble rises and escapes and nobody ever knows it happened. If a thin crust has already formed across the top, the bubble is sealed inside. The TWI technical note on weld porosity describes this rejection process as the fundamental mechanism behind every form of porosity.

Nucleation: Where a Bubble Actually Starts

Dissolved gas does not spontaneously form a bubble in the middle of perfect, clean liquid. It needs something to nucleate on: a tiny oxide particle, a speck of dirt, a rough spot on the joint face, or an inclusion. This is a practical clue worth remembering. A weld that shows porosity on an impeccably clean joint is telling you the gas arrived from the arc or the shielding envelope, not the base metal.

Whether a nucleated bubble escapes depends on a race between the bubble’s rise and the solidification front chasing it down. Buoyancy pushes it upward; arc forces and puddle circulation shove it sideways or pin it in place; travel speed determines how quickly the freeze line moves across it. Win the race and the pore vanishes unnoticed. Lose it and you get a trapped void.

Why Pores Are Round — and Why Wormholes Aren’t

Surface tension pulls a free bubble into a sphere, so a single trapped gas pocket reads as a round, isolated hole in the finished bead. Wormhole or piping porosity is different: it forms when gas keeps feeding a bubble from below while the freeze front advances, drawing the bubble out into a long tail aligned with the direction of travel. Welders see it most often when the source is intermittent or concentrated, and codes treat it more harshly than scattered round porosity for exactly that reason.

The Arc That Inhales: Plasma Instability in Consumable-Electrode Processes

MIG welding is not a sealed system. The arc is a column of plasma that occupies a cavity in front of the advancing puddle, and that cavity is continuously fed by shielding gas. Research from MIT’s Eagar group found that in GMAW, arc interruptions lasting only a few milliseconds can extinguish the plasma entirely and aspirate large amounts of atmospheric gas into the arc cavity. The puddle then freezes with that air mixed in. You can read the original work in Eagar’s published paper on shielding gas behaviour.

What interrupts an arc? A wire feed stumble, a worn or misaligned contact tip, short-circuit transfer cycling too aggressively, sudden stickout changes, or a wanderingly inconsistent hand speed. Every one of those shows up later as porosity, and The Fabricator’s list of porosity causes flags erratic arcs as one of the most common culprits.

What Deoxidizers Do — and What They Cannot Do

Solid GMAW wire contains deliberate deoxidizers, principally manganese and silicon, and these come solely from the steel used to make the wire. In flux-cored wire, they come largely from the core ingredients instead. Their job is to grab free oxygen and form solid oxides such as MnO and SiO₂ that stay dispersed as fine inclusions rather than growing into gas pockets. They are a genuine safety margin, but a limited one: deoxidizers cannot neutralise nitrogen swept in by a draft, nor hydrogen released from wet, greasy or coated metal. That is why proper aluminium cleaning and sound shielding still matter even with premium wire.

In short, porosity is a supply problem. Gas either comes in faster than the puddle can reject it, or it arrives after the surface has closed. Every fix in the sections that follows traces back to one of those two moments.

Cause 1: Shielding Gas Problems (Flow, Drafts, Leaks and Clogs)

Where the Shielding Gas Actually Comes From (and Where It Goes Missing)

Shielding gas is the simplest part of a MIG setup and the most common source of porosity. Its whole job is to blanket the arc and molten puddle from the atmosphere, because at arc temperature nitrogen and oxygen from ordinary air dissolve into liquid metal almost instantly. As TWI’s technical overview of weld porosity explains, those absorbed gases then come back out of solution as the puddle freezes, leaving round pockets behind.

Everything in this section is about protecting that blanket. Start at the cylinder: an empty or nearly empty bottle delivers mostly the pressure of a warming cylinder rather than a steady gas stream, and the last few hundred psi of a mixed cylinder can drift slightly off ratio. A contaminated cylinder, a regulator frozen with moisture, or a bottle refilled from a suspect source will put porosity in every bead you lay. If a porosity problem starts the day after a bottle swap, swap the bottle back first, and use a supplier who will verify purity and dew point on request, such as a reputable welding gas supplier.

Flow Rate: Too Much Is as Bad as Too Little

The instinct when porosity shows up is to crank the flow meter, and that makes things worse. Excessive flow becomes turbulent at the nozzle exit, and turbulence pulls surrounding air into the stream, contaminating the very shield you are trying to protect. Too little flow, on the other hand, simply fails to exclude air in the first place.

Realistic numbers vary with process, wire diameter, nozzle size and gas mix, so treat these charts as a starting point:

Source or Situation	Typical Recommended Flow
General GMAW / FCAW guidance (The Fabricator)	35–45 CFH
0.023-in. wire on thin gauge	About 25 CFH
Short-circuit transfer, light plate (common machine charts)	20–30 CFH
Spray transfer, heavier plate	35–50 CFH
Drafty or outdoor work	Modest bump per manufacturer, not double

Compare a few charts, such as those compiled at GoWelding’s MIG settings charts, and you will see the same pattern: smaller wire and smaller nozzles want less gas, not more. Manufacturers like ESAB and Miller both warn that flow increases should be modest and paired with draft control, because shielding is a gas blanket, not a jet.

Nozzle geometry matters here too. A tapered 3/8-in. nozzle flowing the same volume as a full 5/8-in. nozzle forces gas through a smaller exit at higher velocity, producing a turbulent stream that entrains air. If you switched to a small tapered nozzle for tight access and porosity followed, that is your culprit.

Leaks, Kinks, Clogs and Restricted Hoses

Between the regulator and the nozzle sits a chain of parts, and any weak link bleeds shield or admits air:

• Kinked, punctured or crushed gas hoses, often where the lead drags across the floor or passes through a doorway.
• Damaged or missing O-rings at the feeder connection and the gas nozzle connection, the two most common leak points.
• A plugged nozzle: spatter buildup inside the diffuser or nozzle turns clean laminar flow into a lopsided, turbulent mess.
• A sticking gas solenoid that opens late or closes early, so gas arrives after arc start or leaves before the puddle solidifies.
• Hoses previously used as air lines, sometimes retaining oil mist inside.

Diagnose it properly with a leak check: pressurize the hose with the nozzle blocked and watch the gauge, then soap-test every fitting. Hobart Brothers’ problem-solving guide for reducing weld joint porosity also recommends the old cup-and-damper test at the nozzle, which confirms actual output rather than trusting the flowmeter setting.

Nozzle Distance, Stand-Off and Gun Angle

Even with correct flow, technique can defeat the shield. Holding the nozzle more than about 3/4 in. from the puddle lets the gas cloud disperse before it reaches the weld. An excessive drag or work angle pushes gas off the back side of the puddle and lets air in from the uncovered side. On outside-corner joints the puddle is fully exposed and the nozzle cannot wrap gas around it, so those joints are naturally porosity-prone and want conservative parameters and tighter stand-off. Miller’s breakdown of common MIG defects lists these same technique faults. Aluminium is the unforgiving case; the full workflow lives in our aluminium MIG welding guide.

Drafts: The Invisible Porosity Generator

Open bay doors, ceiling fans, extractors and even a welder walking past can bend the gas column mid-pass. Draft porosity has a signature: it appears on one side of the shop, worsens in summer when doors stay open, and improves the moment you hang a curtain. Fixing it is cheaper than any equipment change, which is why it comes first on the troubleshooting checklist in this article’s workflow section.

Cause 2: Dirty Metal, coatings and Moisture on the Workpiece

Gas shielding gets the blame for most MIG porosity, but the second-biggest culprit sits right under your nozzle: the metal itself. Oil, grease, paint, primer, glue, cutting fluid, sweat, rust, mill scale and zinc coatings all decompose or vaporize at arc temperature, and the resulting gas has nowhere to go but straight into the molten puddle. This kind of contamination is especially punishing in solid-wire GMAW, because there is no flux or slag blanket to absorb the reaction the way there is in stick or flux-cored welding.

What happens to a contaminated joint at arc temperature

The arc core runs in the range of several thousand degrees, well past the point where organic compounds stop being solids or liquids and become a burst of hydrogen, carbon monoxide, carbon dioxide and water vapor. A thin smear of drawing oil can release a surprising volume of gas in the fraction of a second the arc passes over it. The puddle absorbs that gas, and if it freezes before the bubble can rise to the surface, the pore is permanently locked in place.

Common workpiece contamination sources and what they release:

• Oil, grease and cutting fluid — hydrocarbons that crack into hydrogen and carbon gases.
• Paint, primer and adhesive — organic binders that burn off rapidly, often with a visible puff of smoke at arc strike.
• Sweat and skin oils — the fingerprint problem; handling a cleaned joint with bare hands puts moisture and salts right back on the surface.
• Rust and mill scale — hydrated iron oxides that release water vapor as they break down, which is why heavily rusted plate often welds worse than clean hot-rolled steel.
• Zinc galvanizing — vaporized metal that reacts with the puddle and porosifies a wide band either side of the bead.

Moisture, condensation and the dew point

Water is a subtler enemy than oil because you often can’t see it. Any time cold plate is brought into a warm, humid shop, its surface temperature sits below the surrounding air’s dew point, and moisture condenses directly onto the steel. The plate looks dry, feels dry to the back of your hand, and still welds full of pores. The same thing happens to torch nozzles, stored fittings and anything that came in from an unheated trailer on a muggy morning. Keep material stored indoors, wipe it down before fit-up, and give cold stock time to acclimate. Where moisture pickup is critical, preheating the joint area to around 200–225°F drives off surface condensation before the first pass.

Guidance on porosity prevention from manufacturers such as Hobart Brothers consistently puts joint cleaning alongside gas delivery as the first two checks, and TWI’s technical notes on porosity list oil, grease, paint and rust among the standard causes of gas-pore formation.

Coatings: galvanizing, paint and weldable primers

You can’t always remove a coating, so the industry developed weldable primers designed to burn off cleanly. The catch is thickness: apply more than the manufacturer’s maximum, typically around 20 microns on the vertical member of a T joint, and porosity becomes likely, because the primer trapped in the corner has more material to decompose and a harder time escaping. Structured joints are the worst case, since in a T-joint or lap joint the coating, moisture and cleaning residue get physically entrapped in the root and driven straight into the puddle. Where possible, grind or blast the coating back a couple of inches from both sides of the joint line. On galvanized work, weldable primers or grinding the zinc back to bare steel along the seam are the standard options; welding straight through heavy galvanizing is asking for a porous bead.

Cleaning methods that actually work

Match the method to the contaminant:

1. Solvent wipe with acetone or a dedicated degreaser for oil, grease and adhesive. Wipe, don’t spray, and use a clean rag that you replace often.
2. Wire wheel or flap disc for mill scale, rust and paint on steel. Dedicate the wheel to steel only.
3. Stainless steel brush, used only on aluminium, to break up the hydrated oxide layer that forms moments after cleaning. Never touch an aluminium brush to steel or the brush becomes a contamination source itself. The full prep sequence is covered in Aluminum Cleaner for Welding: How to Prep Aluminum Right.

Contaminated aluminium is a special case because the oxide layer actively absorbs moisture from the air. Clean it immediately before welding, keep it dry, and degrease with a dedicated aluminium cleaner rather than a shop rag that has seen everything else. If you’re troubleshooting recurring problems on aluminium specifically, the diagnostic approach in Aluminum MIG Welding: The Complete Practical Guide walks through the same order of checks.

One more habit worth building: anti-spatter sprays and gels are themselves contaminants if they creep into the joint. Keep them off the weld zone, because a perfectly shielded arc over a chemically dirty groove still makes pores.

Cause 3: Contaminated Wire, Dirty Liners and Faulty Gas Hardware

When gas flow reads correctly and the joint looks clean, porosity can still rear its head. More often than not, the fault isn’t at the puddle, but upstream along the path that delivers shielding gas and filler wire to the arc. A break anywhere in that chain can allow air in or restrict coverage, introducing nitrogen and oxygen that freeze into pores.

Contaminated Filler Wire

Wire may arrive clean, but it doesn’t stay that way for long. Greasy gloves, paint pens, tape, solvent-soaked rags, or dropping a spool onto the shop floor all transfer hydrocarbons, cutting fluid, or grinding dust to its surface. As that wire is pulled through the liner and into the arc, those contaminants break down and release gas directly into the weld puddle.

The fix is simple, but easy to overlook. Store spools off of the floor in a dry area, keep labels on the flange rather than the wire, never wipe wire with a dirty rag, and pull a foot or two of fresh wire through the liner after any suspected contamination. If a spool has been exposed to the outdoors or prolonged dampness, treat it as suspect.

Dirty, Worn or Oversized Gun Liners

A gun liner has two jobs: guiding the wire and acting as a low-friction bearing surface. With time, wear or a mismatch to wire diameter can cause the wire to shave fine metal particles inside the gun. Those fines increase drag, create arc instability, and can intermittently expose the puddle to air. Improper drive roll pressure makes this worse, as too much pressure can scrape and deform the wire, while too little causes slipping and stuttering.

As pointed out by The Fabricator’s Consumables Corner, these consumables related issues are a frequent source of porosity. Replace liners on a schedule rather than a breakdown, match the liner bore to the wire diameter, and use the mark-the-wire test to set roll pressure without crushing it.

The Gas Line and Hardware

Gas hoses take a lot of abuse. Dragged across hot plate, run over by carts, or left coiled for years, they can crack, kink, or collapse. Rubber hoses that were once used for shop air can also carry residual oil and moisture that vaporizes into shielding gas. In some cases, as The Fabricator has documented, stored hoses have even developed blockages from debris. Additionally, sticking or slow gas solenoids can delay coverage at arc start, porosifying the first half inch of a weld.

Testing the Delivery System

1. Feel the flow: With gas flowing and no arc, hold a hand or business card at the nozzle to confirm actual output.
2. Soap test: Mix dish soap with water and dab it onto every fitting, hose end, regulator connection, and feeder solenoid while gas is on. Bubbles indicate a leak.
3. Inspect seals: Check O-rings at the feeder and gun-to-feeder joints. Hardened or nicked O-rings are one of the most common hidden leaks.
4. Check for restriction: Disconnect the hose at both ends and blow through it. Any noticeable restriction points to kinks, collapse, or debris.
5. Verify at the gun: Confirm flow rate with a calibrated flow meter at the gun, not just at the regulator gauge.

Gas Purity and Dew Point

Sometimes, the gas itself is to blame. Contaminated cylinders, mix-ups, or moisture pulled in through a leaky regulator will ruin a weld no matter how clean the prep is. For low-porosity aluminium work, Hobart specifies argon purity of 99.997% with a dew point of −76°F or lower. Anything wetter than that can condense on cold aluminium and introduce hydrogen into the puddle. For more on getting that surface clean, see Aluminum Cleaner for Welding: How to Prep Aluminum Right. Always ask your supplier for a certificate of analysis and stated dew point.

When chasing porosity from this category, stick to this order: verify gas at the nozzle, confirm flow at the gun, soap test the hardware, swap the liner, replace any suspect wire, then re-weld. More often than not, that’s all it takes.

Cause 4: Wrong Machine Settings and Bad Welding Technique

Gas and dirt get the blame for most porosity, but the machine dial and the hand holding the gun are responsible for a surprising share. Bad settings never add gas or contamination to the puddle; they change how the arc behaves and how much shielding actually reaches the molten metal. That is why a welder can fix a leaking hose, scrub the plate spotless, and still cut open a bead full of holes.

Voltage: Too Much Turbulence, Too Little Arc

Voltage controls arc length, and arc length controls how much of the arc column is exposed to the air around it. Set it too high and the arc gets long and soft, the pool boils and swirls, and the extended arc draws atmospheric nitrogen and oxygen down into the melt — one of the classic parameter-driven causes of porosity. Penetration becomes inconsistent while the puddle grows wide and restless.

Set voltage too low and the wire stubs into the work instead of melting off cleanly. Arc starts become erratic, the arc cuts out and restarts, and each interruption momentarily sucks normal air into the shielding envelope. A mismatched pair — high wire feed speed with low voltage, or the reverse — produces the same stuttering arc. If the bead looks like a stack of cold lumps or spits and pops on start, the settings, not the gas, are the suspect.

Contact-Tip-to-Work Distance: The Setting That Drifts

For short-circuit transfer, contact-tip-to-work distance (CTTWD) should generally run 3/8 to 1/2 inch. Inside that window the arc is stable and the nozzle is close enough to keep the gas envelope tight over the puddle.

Push the stickout longer and two things go wrong at once. First, more of the wire resists before the arc, so effective amperage drops and the wire melts at a different rate than your settings assume — the machine has effectively drifted out of calibration without anyone touching a knob. Second, the shielding gas leaves the nozzle and disperses before it reaches the pool. Longer stickout also puts the arc column further from the shielding stream, giving air more opportunity to be absorbed. Settings charts assume a correct stickout; hold it constant and the numbers on the chart mean what they say.

Travel Speed and Torch Angle

Travel too fast lays a thin, narrow bead that the gas stream cannot blanket evenly — the shield trails behind and the trailing edge of the puddle is briefly exposed. Travel too slow does the opposite: the pool widens, gets turbulent, and holds heat long enough for gas to be driven in and then frozen. Both extremes produce holes; only the bead shape tells them apart.

Angle matters just as much. A GMAW gun held more than roughly 5 to 15 degrees off perpendicular spreads the gas stream sideways and starts drawing atmosphere in from the opposite side. Work-angle discipline is one reason aluminum MIG work is so unforgiving: the puddle freezes fast and has no time to let trapped gas escape.

Joint Geometry: Root Gaps and Outside Corners

An open root gap acts like a suction channel. As the arc moves, the shielded zone shifts and air is pulled through the gap from the back side of the joint, entering the pool from underneath where no nozzle can protect it. Tack the gap shut, add a backing, or narrow the fit-up before blaming the cylinder.

Outside-corner joints are the mirror-image problem: the nozzle physically cannot sit over the puddle without riding on the joint edge, so the gas stream splits around the corner. Drop to a smaller tapered nozzle, get closer, or bevel the edge so the nozzle can cover the fusion zone.

Correction Steps You Can Run in Two Minutes

1. Pull the stickout back to 3/8–1/2 in. and trim the wire end.
2. Compare voltage and wire feed speed to the door chart for that material thickness.
3. Reset gun angle to within 15 degrees of perpendicular.
4. Slow down or speed up until the bead width stays consistent.
5. Close or back up any root gap that lets you see daylight.
6. Run three inches of bead and re-inspect before committing to the whole joint.

Note on ground clamp placement: Clamp the return lead to clean metal close to the joint, never to a rusty flange, a painted edge, or the far end of a large assembly. A poor ground forces current through an unstable path, causes stray arcing at the clamp, and can destabilize the arc enough to make the puddle wobble and admit air. Move the clamp and the porosity often disappears with it.

Material Matters: Porosity in Steel, Stainless and Aluminium MIG Welding

Porosity is always trapped gas, but the gas behaves differently depending on what you are melting. The same draft that leaves a harmless pinhole in mild steel can open a line of wormholes in aluminium, and a shielding gas mix that runs beautifully on carbon steel can produce a sooty, cavity-riddled bead on stainless. Knowing how each material behaves tells you which cause family to investigate first.

Carbon Steel: Mill Scale, Rust and the CO2 Question

Mild steel is the most forgiving MIG substrate, mostly because the arc itself is somewhat self-cleaning. The standard 75/25 argon-CO2 blend (C25) gives a stable, low-spatter arc, while straight CO2 runs hotter and deeper. That extra heat disassociates the CO2 molecules, and the released oxygen reacts with carbon, manganese and silicon in the puddle rather than with nitrogen from the air. The resulting oxides are largely captured by the slag or float out, which is why many fabricators find 100% CO2 more porosity-resistant on rusty or mill-scaled plate, at the cost of a rougher, spatter-heavy bead.

Mill scale, rust and moisture are the usual steel culprits. Welding over damp or contaminated plate drives hydrogen and steam straight into the pool, and cutting fluids left in a laser-cut edge are a classic single-joint porosity source. The Fabricator’s roundup of porosity causes repeatedly flags oil, moisture and paint as top offenders.

Where the shielding story changes is flux-cored work. With self-shielded FCAW, the flux does the protective job, and Hobart’s technical guidance explains that 100% CO2 improves porosity resistance because the high arc temperature disassociates the molecules, letting controlled oxygen react with other elements and trap impurities in the slag. That slag cover also physically blocks nitrogen pickup, which is why FCAW tolerates outdoor drafting far better than short-circuit GMAW.

Stainless: Sensitization Hides Behind the Pores

Stainless MIG porosity is less about the metal rejecting gas and more about the extras: moisture in tri-mix gases, oil on the back of sheet, and heat tint that signals the joint was not fully shielded. Losing argon coverage while welding stainless lets chromium oxidise and can pull the heat-affected zone into sensitization, where chromium carbides precipitate and corrosion resistance drops. A porous stainless weld is therefore also a compromised stainless weld. Flow rates matter here because tri-mix gases (argon, CO2 and sometimes helium or oxygen in small percentages) are heavier and more sensitive to turbulence than pure argon.

Aluminium: The Porosity Magnet

Aluminium is the hardest of the three for one simple reason: liquid aluminium absorbs hydrogen far more readily than liquid steel, and the hydrated oxide layer sitting on the surface happily supplies it. Every scrap of condensation, every fingerprint, every smear of cutting oil becomes a bubble the moment the arc passes. This is why Aluminum Cleaner for Welding: How to Prep Aluminum Right matters so much in practice: acetone or a dedicated aluminium degreaser first, then a dedicated stainless steel brush used on aluminium only.

Technique differs too: aluminium wants a push technique with MIG, pulling drags soot and oxide across the leading edge of the puddle. Argon-helium mixes add penetration and fluidity on thicker aluminium, but helium’s rising tendency makes gun angle and nozzle distance even more critical. For a full walkthrough, see Aluminum MIG Welding: The Complete Practical Guide.

Moisture control is where the numbers get concrete. Hobart and ESAB both specify an argon dew point of −70°F or lower to stop moisture condensing on cold stock, and Hobart’s dew point chart notes that at 75% relative humidity, metal no colder than about 7.2°F below the air temperature will not collect condensation. In practice, aluminium plate carried from a cold yard into a warm shop wets up within minutes.

Quick Comparison

Material	Typical Gas	Porosity Trigger	Best Defence
Carbon steel	C25 or 100% CO2	Mill scale, rust, oil, damp plate	Grind clean, dry stock, let CO2 clean the arc
Stainless	Argon-CO2-O2 tri-mix	Moisture, back-side contamination, lost coverage	Purge, control flow, protect heat tint
Aluminium	100% argon (or Ar-He)	Hydrated oxide, condensation, oils	Degas, stainless-brush only, push technique
Outdoor/drafty	Self-shielded FCAW	Wind stripping shielding gas	Flux cover instead of gas

When the work moves outside or near an open door, most fabricators abandon gas-shielded MIG and switch to self-shielded flux core, because the flux-generated slag blanket resists nitrogen and oxygen pickup that no cylinder flow rate can overcome. If you are weighing that switch, Aluminum Flux Core Welding: What Actually Works in 2026 covers where the process genuinely delivers and where it still struggles.

Matching material, gas and environment is the first diagnostic step. Before you touch a single valve, know which metal you are welding and what its shielding requirements are.

How Porosity Is Detected and What the Codes Actually Allow

The tricky part about porosity is that it rarely announces itself. A bead can look smooth, evenly rippled and perfectly capped, then fail the moment it is cut, etched or radiographed. That is why inspection is a separate discipline from welding, and why the gap between “looks fine” and “meets the code” decides a lot of jobs.

Visual Inspection First, Then the Surface Methods

Nearly every examination starts with the eye and a low-power magnifier. Visual inspection (VT) catches surface-breaking pores, craters, undercut and spatter, and it is fast enough to run on every inch of weld. But VT has a hard limit: it cannot see inside the weld, and a pore opening onto the surface looks identical to a pinhole caused by arc blow or a smear of oil until someone probes it.

For surface-breaking discontinuities the next step is dye penetrant testing (PT) or magnetic particle testing (MT). Liquid penetrant works on any nonporous material, which makes it the default for aluminium and stainless: a coloured or fluorescent dye wicks into any pore opening, then a developer pulls it back out so the flaw blooms visibly on the surface. Magnetic particle testing only works on ferromagnetic steel, but it shows near-surface indications without consumables left behind on the part. Both methods answer one question well: does gas break the surface? Neither can tell you how deep the pocket goes.

Radiography, Ultrasonics and Why RT Wins for Porosity

To characterise subsurface gas porosity, nothing beats radiographic testing (RT). An X-ray or gamma source passes through the weld, and every gas void shows up as a dark, rounded spot on the film or digital detector. Radiography gives the inspector the size, shape, distribution and depth of each pore in a single image, which is exactly the information acceptance criteria are written around. As TWI’s technical notes on weld porosity describe, pores can be isolated, clustered, aligned or linear, and RT is the method that reliably distinguishes those patterns.

Ultrasonic testing (UT) can find larger voids and is useful where no radiation source is practical, but scattered small pores tend to produce a messy signal that is hard to size accurately. In practice, UT is used for volumetric flaws while RT characterises porosity; magnetic particle and penetrant remain the tools for anything that breaks the surface, and visual inspection under good lighting is still the first gate on every weld.

What the Codes Actually Allow

Here is where welders get surprised: codes permit porosity. The goal is never zero voids, it is limited, well-distributed voids that do not meaningfully reduce section or initiate a crack. The common numbers look like this:

Code / Authority	Typical porosity acceptance limit
AWS D1.1 (structural steel)	Rounded porosity not exceeding about 3/8 in. in any linear inch and 3/4 in. in any 12-in. length of weld; piping porosity checked separately in Table 6.1
ASME B31.1 (power piping)	No single pore with any dimension greater than 3/16 in.
API 1104 (pipeline welds)	No individual pore exceeding 1/8 in. in diameter
Iowa DOT and similar DOT specs	Porosity classified as scattered, clustered or linear, with limits per category and no allowance for clustered or aligned pores in critical zones

Those numbers come from real published specifications: Iowa DOT’s Engineering Limits of Reinforcement and Repair is a good example of how state agencies sort porosity by distribution rather than size alone. Notice the pattern — pipeline work is strictest on individual pore diameter, power piping on absolute dimensions, and structural steel on cumulative size per inch of weld.

Why Round Porosity Is the Forgiving Defect

Round pores are the least damaging discontinuity a weld can have, and the reason is mechanics. A sharp crack tip concentrates stress at its tip radius; a spherical void has a blunt geometry that blunts the advancing crack front, redistributing stress instead of focusing it. In other words, a well-dispersed field of round pores can arrest a crack rather than drive it, which is why codes tolerate them while rejecting a single sharp lack-of-fusion line. Porosity still costs you fatigue strength and section, so it is not harmless — but it fails a lot more gently than a crack.

Whatever method you use, know which code governs before you strike the arc. An acceptance criterion applied after the fact is not quality control; it is a gamble. The OSHA welding and cutting standard, OSHA FactSheet 3647, also reminds us that inspection runs alongside proper ventilation and fume control, since porosity-driven corrections like higher gas flow and enclosed booths change the air the welder breathes.

Fixing and Preventing Porosity: A Step-by-Step Shop Workflow

When porosity shows up, resist the urge to start turning knobs. Guessing costs you wire, gas and a pile of scrap. Instead, run the same short diagnostic routine every time, in the same order, and you will usually find the culprit in under ten minutes. The sequence walks from the puddle backwards to the gas cylinder, which is exactly the direction the problem travels.

The Ten-Minute Diagnostic Sequence

1. Confirm gas is actually leaving the nozzle. Pull the trigger away from the work for two or three seconds and hold a hand near the nozzle. No felt flow means you have a solenoid, hose or cylinder problem, not a welding problem. Check the regulator gauge and make sure the cylinder valve is open and the contents are not effectively empty.
2. Check the flow rate with a flowmeter, not by ear. Most GMAW setups want shielding flow typically between 25 and 40 CFH, adjusting upward only for drafts and downward for small nozzles. More is not better: excessive flow pulls air in around the nozzle through turbulence. As a practical starting point, MIG settings charts give you sane figures for wire diameter and gas mix.
3. Look at the nozzle. A nozzle packed with spatter narrows the gas shield and disrupts flow. Clean it with a nozzle reamer or pliers, replace it if it is cratered, and confirm your contact tip recess and gun-to-work distance match the machine chart.
4. Block the drafts. Shut the bay door, reposition the fan, or put up a welding screen. Even a mild breeze several feet away can strip the shield off a puddle, and the effect is worst on the light, slow-moving shield of an aluminium weld.
5. Inspect hoses, fittings and O-rings. Look for kinks, scorch marks, punctures and cracked seals at the feeder and gun connections. A pinhole leak draws air into an otherwise healthy gas line.
6. Clean the joint. Solvent-wipe oil and grease, remove paint and primer, and knock off loose mill scale. Aluminium needs dedicated mechanical cleaning with a stainless brush, and the prep chemistry is covered in Aluminum Cleaner for Welding: How to Prep Aluminum Right.
7. Check the wire and liner. Wipe the incoming wire, look at the spool for paint, tape or oil, and check drive-roll pressure and liner wear. Wire that grinds fines off a dirty liner arrives at the puddle already contaminated.
8. Verify parameters against the door chart. Voltage too high, wire feed too low, travel too fast, excessive stick-out and steep torch angles all feed atmospheric gas into the pool. Correct these before touching anything else.
9. Control moisture. Store electrodes and consumables dry, keep the shop dehumidified where possible, and preheat cold stock that has sat outside. If condensation is the suspect, preheating the joint to roughly 200–225°F is a practical field test: weld the joint hot, let it cool, then re-inspect. If the porosity disappears, your material was wet, not your setup.
10. Re-weld correctly. Only strike the arc again once the cause is removed. Re-running over a porous bead without fixing the source simply traps a second layer of gas.

Repair Rules: What Comes Out and What Stays

Once porosity is found in a finished weld, removal of rejectable porosity is normally done by localized gouging or grinding; if porosity is widespread the entire weld must be removed and re-welded. Mechanically, that means grinding the pore out until clean, sound metal is visible, then re-welding into sound material. Air Arc Welding and Carbon Arc Gouging is the standard tool for opening up deep or extensive porosity quickly.

Not every pore needs removal. Codes such as AWS D1.1 tolerate a measured quantity of scattered, rounded porosity, and chasing fully code-compliant welds can remove more sound metal than the defect ever threatened. In non-structural architectural and ornamental work, welders sometimes back-fill pores for appearance, and sealed porosity is occasionally accepted in leak-critical applications only where the governing specification allows it. Know the standard you are working to before you pick up the grinder, and follow the inspection requirements documented in your Aluminum MIG Welding: The Complete Practical Guide or equivalent procedure paperwork.

Prevention Habits That Actually Hold Up

The shops that stop fighting porosity treat it as a system, not an accident. Prevention starts with a written WPS and a qualified welder, because a procedure that specifies gas, flow, cleaning method, joint prep and parameters removes the guesswork that causes most defects. Hobart Brothers’ problem-solving guide makes the same point: fix the joint prep and the gas delivery before you change the machine.

Environmental control matters just as much. Keep gas cylinders upright and secured, source gas from a supplier who will certify purity and dew point such as Airgas Welding Supplies, store wire indoors, and keep the shop dry and draft-free. ESAB’s Quality Matters column repeatedly traces recurring porosity back to these unglamorous fundamentals rather than machine settings.

Finally, train for it. Welders who understand why gas gets trapped diagnose faster than welders following a checklist, and that is the real purpose of the ten-minute routine above. For anyone building toward formal qualification, the same reasoning underpins what is taught in an Aluminium Welding Course: What You Learn, Costs & Careers. Prevention is a habit loop, not a one-time fix, and it is always cheaper than a second pass.

Common Porosity Mistakes, Troubleshooting Table and Key Takeaways

What separates a shop that fights porosity daily from one that rarely sees it is usually not better equipment. It is avoiding the same handful of mistakes over and over. Here are the ones that keep showing up.

The Six Mistakes Shops Keep Repeating

1. Cranking the gas flow up to fix everything. This is the reflex answer to any porosity problem, and it often makes things worse. Past roughly 40–50 CFH the shielding stream turns turbulent and starts aspirating air into the puddle, which is exactly the porosity you were trying to cure. Diagnose before you twist the regulator.
2. Blaming the cylinder when the joint was dirty. Swapping gas bottles is expensive troubleshooting that rarely pays. Oil, cutting fluid, primer and mill scale release far more gas at arc temperature than most cylinder problems do. As an industry roundup notes, contamination and shielding issues dwarf nearly every other category. Check the metal first, and work through The Fabricator’s porosity cause list when you are stuck.
3. Overusing anti-spatter sprays and gels. A light film on a nozzle is fine. A soaked joint is not. These compounds are hydrocarbons that boil straight into gas at arc temperature. Wipe off the excess, keep spray off the weld zone, and never use gel inside a root or T-joint.
4. Welding cold stock straight from outside. Steel pulled from a 35°F trailer into a humid shop condenses an invisible layer of water on the surface. Let it warm and dry before it hits the fixture, as discussed in the weld porosity overview from The Welder.
5. Ignoring seasonal patterns. Porosity that appears in July and vanishes in October is environmental: doors open, fans running, humid air. If your defect rate tracks the weather, your shielding is losing to airflow, not to the machine.
6. Applying the wrong repair. Surface pores ground shallow and re-welded over without full removal simply trap the same gas again. Cut it out properly, as covered in the repair workflow above.

Symptom-to-Cause Troubleshooting Table

Symptom	Most likely cause	First check
Porosity only in flat position	Travel speed too fast on open root, or puddle sagging through a gap	Slow down, widen bevel slightly, reduce heat
Porosity only in vertical or overhead	Technique manipulation breaking the gas envelope; puddle turbulence	Narrow the weave, keep nozzle close, reduce stickout
Porosity only in summer	Drafts from open doors or fans; humid air on cold stock	Shield fans, confirm the joint is dry, verify flow at the nozzle
Porosity only on aluminium	Hydrated oxide layer and condensation; wrong cleaning tools	Fresh stainless-wire-brush the joint immediately before welding; see Aluminum Cleaner for Welding: How to Prep Aluminum Right
Sudden onset after a wire change	Contaminated spool (paint, tape, glue, oil), or new wire diameter without liner or roller change	Wipe wire with solvent, inspect liner, re-set drive-roll pressure
Porosity along an entire seam	Gas supply problem: empty cylinder, kinked hose, bad O-ring, plugged nozzle, stuck solenoid	Soap-test fittings, check nozzle flow, clean or replace the nozzle
Pores only at weld starts	Insufficient preflow, cold joint, or arc starting on a dirty tack	Clean tacks, add gas preflow time, warm the start
Intermittent popping arc with pores	Voltage/WFS mismatch, erratic feed from a worn liner or bad drive rolls	Re-check the settings chart, replace liner and contact tip

Conclusion: What Causes Porosity in MIG Welding

So, what causes porosity in MIG welding? Gas. Specifically hydrogen, oxygen and nitrogen that get into the molten puddle from three places: shielding that fails to hold, metal that arrives dirty or wet, and wire or settings that let atmosphere in. Clogged nozzles, summer drafts and a sticky solenoid all reduce to that one sentence.

Work the problem in this order:

• Confirm gas is actually reaching the puddle, at 20–30 CFH for most short-circuit work
• Eliminate drafts and airflow across the joint
• Clean the base metal and let cold stock reach shop temperature
• Inspect wire, liner, contact tip and drive rolls
• Verify voltage, wire feed speed and stickout against the door chart
• Control moisture and dew point, using 200–225°F preheat where materials call for it
• Remove accepted porosity fully before re-welding, or leave it alone per code

Then keep it gone with weld procedure discipline: qualified procedures, welder qualification to the applicable code, and documented cleaning standards for every joint. For aluminium-specific practice, our aluminum MIG welding guide walks through the same workflow for nonferrous work.

Porosity hides well. It looks like a clean bead until it is cut, bent or put into service. The shops that catch it early treat gas shielding as a system: cylinder, hose, liner, nozzle, puddle, environment. Fix one link, and the pores disappear.

Frequently Asked Questions

What is the number one cause of porosity in MIG welding?

Poor shielding gas coverage is the leading cause, because even about 1% air mixing into the argon/CO2 shield is enough to produce distributed porosity. Leaks, clogged nozzles, wrong flow rate and shop drafts are the usual ways that air gets in.

What gas flow rate prevents porosity in MIG welding?

Most shops get good results between 25 and 40 CFH for GMAW. Too little flow leaves the puddle exposed, while too much (50–60 CFH) creates turbulence that sucks atmospheric air into the stream — so more is not better.

Why does my MIG weld have porosity only when welding outside or near a door?

Air movement displaces the shielding gas, which is heavier than air and only protects locally. Fans, open bay doors and even a breeze a few miles per hour can strip the shield, which is why the problem appears intermittently with no settings change.

Can dirty metal really cause porosity?

Yes. Oil, grease, paint, rust, mill scale, cutting fluid and especially galvanized coatings release large volumes of gas when they hit arc temperature. Thorough solvent cleaning, grinding or wire-wheeling the joint before welding removes most of this risk.

Why is aluminium MIG welding so prone to porosity?

Hydrogen is far more soluble in molten aluminium than molten steel, and hydrated aluminium oxide plus any surface moisture releases hydrogen as the metal melts. Control it with meticulous cleaning, dry stock stored above dew point, and pure low-dew-point argon.

Is porosity a weld defect that must always be repaired?

Not always. Codes such as AWS D1.1, ASME B31.1 and API 1104 set size and distribution limits, so small round pores may be acceptable. Rejectable porosity must be removed by grinding or gouging and re-welded, and widespread porosity means the whole weld comes out.

How do I find out if my weld has subsurface porosity?

Surface pores can be caught with visual inspection, dye penetrant or magnetic particle testing, but internal pores need radiography or ultrasonic testing. Radiography is normally the most effective way to detect and characterize porosity.

Does turning up the shielding gas fix porosity?

Only if low flow was the actual cause — and turning it up too far usually makes things worse by creating turbulence. Diagnose in order: gas delivery and nozzle first, drafts second, joint cleanliness third, then wire, liner and machine settings.