MIG Welding Porosity: Causes, Fixes, and Prevention

You lay a bead that looks perfect, then the grinder hits it and the surface opens up into a cluster of tiny craters, like the weld was carbonated. You change the gas bottle, swap the wire, and burn three more coupons — and the pinholes come back in the same spot. The frustrating truth is that MIG welding porosity is almost never one single fault; it is a chain of small failures that only shows up after the metal freezes.

Why Your MIG Welds Keep Coming Out Porous (And Why It’s Rarely the Machine)

Dale had a Tuesday he would rather forget. He was running 1/8-inch mild steel tube on a handrail job, and the beads looked clean enough to ship. Then the grinder touched them, and the surface opened into a scatter of pinholes. He changed the gas bottle. He swapped the wire spool. He bumped the flow from 20 to 30 CFH, then 35, and watched a full cylinder disappear into test coupons that still hissed at him. By afternoon he had rebuilt the ground clamp, cursed the machine, and was no closer to an answer. The actual fault turned out to be a fifteen-dollar hose that had once been used for air tools.

That story matters because it shows where most welders look first and where the problem usually lives. Porosity is gas that was dissolved in, or mechanically trapped by, the molten pool and could not escape before the metal froze. TWI’s technical summary describes it plainly, and AWS A3.0 defines porosity as “cavity-type discontinuities formed in weld metal by gas entrapment during solidification.” The entire event happens in a narrow window: the moment the pool is liquid and gas solubility is high, followed by the instant it solidifies and solubility drops sharply. Whatever gas is still in the metal at that instant becomes a void.

There are only two ways gas gets there, and almost every cause on a troubleshooting chart fits one of them:

• Gas absorption. Hydrogen, nitrogen, and oxygen dissolve into the liquid pool from moisture, coatings, atmosphere, or contaminated filler. As the metal freezes, it cannot hold them anymore.
• Mechanical entrapment. Turbulence, drafts, coating breakdown, or an open root literally pulls gas into the pool and freezes it in place before it can bubble out.

Here is the part that stings: the machine is usually the last thing to blame. Amperage, voltage, and wire speed matter, but most porosity traces back to shielding, cleanliness, or consumables. One widely cited industry troubleshooting guide estimates that roughly 90 percent of porosity can be prevented when causes are systematically eliminated instead of guessed at. And unlike stick or flux-cored welding, MIG/MAG has no slag to shield or clean the pool. All the protection you get comes from the gas envelope and the deoxidizers alloyed into the wire, so there is no margin for a leaky fitting or an oily joint.

What follows is a full path through the subject: the metallurgy of trapped gas, pattern-based diagnosis, material-specific traps like aluminum (the aluminum MIG welding guide covers those in depth), code acceptance limits, proper repair, and a prevention routine you can run every morning. Miller’s defect troubleshooting reference is a useful companion along the way.

By the end, you will not have a list of things to try. You will have a diagnostic sequence: a way to read the bead, isolate one variable, prove the fix on a coupon, and write down what worked so the same pinholes never cost you another bottle of gas.

What Porosity Actually Is: A Metallurgical View of Trapped Gas

Forget the torch for a second and look at the weld from the metal’s point of view. At arc temperature, that molten pool is a solvent. Steel and aluminum in the liquid state will happily dissolve hydrogen, nitrogen, and oxygen in quantities that would be unthinkable once the metal is solid. Then the arc moves on, the pool cools at hundreds of degrees per second, and the solubility of those gases falls off a cliff. Everything that was dissolved a moment ago suddenly has to go somewhere. It either bubbles out through the surface of the pool before the metal freezes, or it freezes in place as a void. That is porosity in one sentence: gas that ran out of time.

Solubility Versus Temperature: Why “Welding Hotter” Doesn’t Fix It

In steel, hydrogen and nitrogen solubility drops sharply as the metal passes from liquid to solid — the solid lattice simply has no room for those atoms in the same concentration. In aluminum the collapse is even more dramatic, which is one reason aluminum is so unforgiving about moisture and contamination. The excess gas has to escape, and escape is a race against solidification.

This is exactly why turning up the heat is such a poor fix. A hotter, longer arc gives the gas more time to rise, sure — but it also widens the pool, lengthens the arc, and pulls more atmosphere into the shielding envelope. You have not removed the source of the gas, you have just given it a bigger door. If the hydrogen is coming from oil in a lap joint, no amperage setting on earth will burn it out; it will just dissociate and re-enter the pool. The gas has to be kept out in the first place, or the surface has to be dry and clean enough that there is nothing to release. (For the specific material quirks, the complete practical guide to aluminum MIG welding is worth a read alongside this section.)

The Morphology Families and What Each One Tells You

Pores are not interchangeable. Their size, shape, and distribution are a fingerprint of where the gas came from. Industry guidance generally sorts them into distributed porosity, surface-breaking pores, wormholes or elongated piping, and crater pipes — and experienced inspectors read that pattern before they touch a dial.

Type	Appearance	Typical location	Most likely cause
Uniformly distributed fine porosity	Small, evenly scattered round pores	Throughout the bead	Shielding gas contamination, wrong flow, air entrainment
Clustered porosity	Localized group of pores	One patch along the bead	Local contamination or a gas surge at the start
Linear porosity	Pores in a straight line	Along a root gap, lap, or coating edge	Gas released from the joint interface
Wormholes / piping	Elongated, tunnel-like voids	Deep in the bead	Gross contamination, thick coatings, heavy galvanizing
Surface-breaking pores	Open craters on the face or root	Weld surface	Gas escaping late, or grinding into subsurface voids
Crater pipe	Funnel-shaped cavity	End of the bead	Shrinkage at arc termination, lost post-flow

How Porosity Actually Hurts a Weld

Porosity is not just cosmetic. Each pore removes load-bearing metal, so the effective cross-section of the joint shrinks. Worse, a round void in a stressed section is a stress raiser: load lines crowd around the pore edge, and those edges become preferred crack initiation sites under fatigue or impact. In corrosive service, pores act as traps for moisture and chlorides, undermining the corrosion resistance the alloy was chosen for. In pipe, tanks, and pressure work, a chain of connected pores is simply a leak path waiting for pressure.

Detection matters here too. TWI’s technical guidance on porosity notes that radiography is generally the more effective method for finding and characterizing porosity, though small pores are difficult to resolve in thick sections.

Discontinuity Versus Defect: The Distinction That Matters

Here is where a lot of shop-floor arguments go wrong. A discontinuity is any interruption in the expected structure of the weld. A defect is a discontinuity that exceeds the acceptance criteria of the applicable standard. Porosity is only a defect when the code says it is. Under BS EN ISO 5817, acceptance is defined by quality levels B, C, or D, while AWS D1.1 and ASME IX set their own limits — and the project specification always governs. The Welder’s breakdown of weld porosity facts is a good sanity check on how those limits get applied in practice.

So the metallurgy gives you the mechanism, and the morphology gives you the source. Neither one tells you whether to repair until you know what standard you are welding to.

The Three Gas Families Behind MIG Porosity — and Where They Come From

Most porosity checklists fail at the bench because they mix symptoms, causes, and cures into one long list. A more useful way to think about MIG porosity is to ask one question: which gas is creating the void? Nearly every case comes back to one of three families. Once you know the family, the likely source and the first test become much easier to narrow down.

Hydrogen: The Quiet Contaminant

Hydrogen is the most damaging and most overlooked of the three. It does not need a leaking hose or a windy doorway. It rides in on moisture, grease, oil, rust, paint, cutting fluid, humid air, and condensation trapped between lap or fillet surfaces. It also comes from undried consumables, including flux-cored wire and any electrodes or fluxes stored in the same humid shop. At arc temperature, those compounds break down and release hydrogen, which dissolves readily in the molten pool. As the metal freezes, its ability to hold hydrogen collapses, and the excess has nowhere to go except into voids.

In carbon steel the result is often fine scattered porosity or wormholes. In high-strength low-alloy steels such as A514, A588, and A709, the risk is bigger than a few pinholes: hydrogen can cause delayed cracking hours or days after the weld cools. That is one reason low-hydrogen practice, dry consumables, and preheat matter on those grades. Aluminum is another special case because hydrogen solubility drops sharply on freezing, which is why clean, dry preparation is so critical. Our aluminum MIG welding guide covers that preparation routine in detail.

Nitrogen and Oxygen: The Atmosphere Inviting Itself In

Nitrogen and oxygen porosity is almost always atmospheric. The shield failed, not the metal. Common entry points include a loose connection, a cracked hose, a damaged O-ring, too little flow, too much flow, a nozzle held too far from the work, or a fan or open bay door pulling the envelope away. The weld pool then absorbs nitrogen and oxygen from the air, and both can end up as porosity or as oxide inclusions that interfere with fusion.

The threshold is surprisingly small. As little as 1 percent air entrainment in the shielding gas can cause distributed porosity, and above roughly 1.5 percent it tends to produce gross surface-breaking pores. One percent sounds trivial until you do the math: at 20 CFH of shielding gas, 1 percent is only 0.2 CFH of air, a tiny leak or a brief turbulence event. That is why a gauge can look normal while the weld still fails. For a structured look at these mechanisms, TWI’s overview of porosity and its prevention is a useful reference.

Carbon Monoxide: The Steel-Specific Reaction

In steel, oxygen that is not tied up before solidification can react with carbon to form carbon monoxide. The gas forms late in freezing, and the bubbles can become trapped as internal pores. Filler metals are designed to prevent this through deoxidation. Silicon and manganese, and in some wires additional deoxidizers, combine with oxygen and keep it out of the CO reaction. That is the practical difference between a general-purpose ER70S-3 and a higher-deoxidation ER70S-6. If the filler has too little deoxidizing capacity for the steel condition, oxygen stays free and CO porosity becomes more likely. Lincoln Electric’s MIG problems and remedies guide is a good next step on filler and process matching.

Gas family	Typical source	Typical appearance	First corrective action
Hydrogen	Moisture, oil, grease, rust, paint, humid air, condensation in joints	Fine scattered pores, wormholes, occasional delayed cracking	Dry and degrease the joint, clean the wire, control moisture, use preheat where specified
Nitrogen and oxygen	Leaks, low or excessive flow, long stickout, drafts, damaged seals	Uniform fine porosity or larger surface-breaking pores	Letest and flow verification, correct nozzle distance, and draft control.
Carbon monoxide	Insufficient deoxidation in the filler, oxidized steel, poor gas coverage	Internal pores near the centerline, scattered porosity	Verify filler classification and condition, clean the joint, confirm gas coverage

Why “More Gas” Can Make It Worse

When the Pores Do Not Fit One Family

Mixed or intermittent porosity usually means more than one source is active. A humid morning, a worn O-ring, and a slightly long stickout can combine into a pattern that no single fix explains. That is why welders who change three things at once never learn what actually worked. They may stop the pores for a day and still have no idea which change did it, so the problem returns when conditions shift. Work through the families in order, change one variable, run a coupon, and write down the result. That is how a long checklist becomes a usable decision tree.

Shielding Gas Delivery: Flow Rate, Turbulence, Leaks, and Drafts

If porosity is trapped gas, the delivery system earns the first look. Not the machine, not the wire lot — the plumbing between the cylinder valve and the molten pool. In most shops, this is where the fix actually lives, and it is also where welders burn the most time chasing the wrong thing. The gas has one job: arrive at the pool as a slow, continuous, uncontaminated blanket until the metal freezes. Everything in this section is about protecting that blanket.

Flow Rate: Finding the Window

Typical GMAW shielding gas flow is about 30–40 CFH, with 35–45 CFH a common general recommendation. Small tapered nozzles on root passes often run better at 25–30 CFH, because the tighter orifice produces a more coherent column at lower volume. Flows of 50–60 CFH at the nozzle are generally excessive — and this is the single most common error in the shop, because “more gas” feels safe. It isn’t. Past a certain velocity the gas jet becomes turbulent and aspirates surrounding air into the plume, so over-shielding produces the same porosity as under-shielding.

Published research on DH36 steel found that 9 L/min of pure argon was enough for sound welds while 6 L/min produced clear porosity, with 12 L/min often used in production as a safety margin — useful evidence that the correct window is narrower than most people assume and that a modest, verified flow beats a big number (Journal of Materials Processing Technology).

Two things change what that window looks like at the arc. First, nozzle diameter: the same CFH forced through a smaller bore exits at higher velocity and entrains more air, while a larger bore delivers the same volume more gently but is easier for drafts to strip. Second, distance: the farther the nozzle sits from the work, the more the envelope has to travel through moving air.

A Regulator Is Not a Flow Meter

This trips up experienced welders constantly. A two-gauge regulator shows cylinder pressure and delivery pressure. Delivery pressure is not flow — it tells you nothing about how many cubic feet per hour are reaching the nozzle. You need a ball-type flowmeter or a calibrated flow gauge, and it must be matched to the gas. CO2 behaves differently from argon mixes because it is stored as a liquid; CO2 flowmeters are calibrated specifically for CO2, and at high flow rates or in a cold shop the regulator body can frost and give erratic readings. If your CO2 flow gauge is frosting heavily, back the flow down or add a heater — don’t crank it open.

Leaks, Worn Parts, and the Daily Leak-Down Test

Most delivery faults are mechanical and cheap to fix. Run a leak-down test every morning: open the cylinder, pressurize the line for 15–20 seconds, shut the cylinder valve, and watch the high-pressure dial. A slow coast-down means gas is escaping somewhere — usually a hose, a fitting, or an O-ring. Pay attention to the O-rings at the whip-to-feeder connection, which crack and harden with age and heat. Check for hoses that are pinched under the cart or cracked at the strain relief. And watch for the classic: a gas hose previously used for air tools, now quietly delivering oil mist straight into your shielding stream.

At the torch end, inspect the nozzle for spatter bridging and the diffuser for plugged holes, both of which distort the envelope even when flow at the regulator looks perfect. Set contact-tip recess so the tip sits at or slightly inside the nozzle face, keep wire stickout no more than 1/2 inch beyond the nozzle, and hold the gun within roughly 5–15 degrees of perpendicular. Push technique generally gives better gas coverage than drag, because the nozzle leads the pool and pre-floods the joint. Pre-flow should be long enough to purge the line before the arc starts, and post-flow long enough for the pool to freeze — short post-flow is a leading cause of crater pipes. A sticky gas solenoid produces the maddening kind of porosity that appears and disappears between beads; if the pattern is truly random, suspect the solenoid.

Drafts: The Invisible Culprit

Wind is not only an outdoor problem. AWS D1.1 guidance advises against GMA welding when wind speed exceeds 5 mph at the joint, and 5 mph is barely perceptible — the kind of air movement produced by an overhead door left cracked open, a pedestal fan 25 feet away, or a ventilation system pulling across the bay. A fan that feels like nothing on your face can strip a shielding envelope off a bead on an otherwise calm day.

Symptom-to-Adjustment Table

Symptom	Likely cause	Adjustment
Fine scattered pores across whole bead	Flow too low, or air entrainment	Verify flow at nozzle, then set 30–40 CFH
Pores worse after turning gas up	Turbulence from excessive flow	Reduce to the low end of the range
Pores at bead start only	Line not purged	Increase pre-flow
Crater pipe at bead end	Post-flow cut short	Extend post-flow until pool freezes
Pores only on one side of bead	Gun angle or draft from that side	Square the angle, block the airflow
Random good/bad beads	Sticky solenoid or leaking O-ring	Leak-down test, inspect solenoid

How to Tune Flow Correctly

Start at the low end of the window — around 30 CFH for a standard nozzle — run a test coupon on scrap from the same material, and inspect. If the bead still shows contamination, add 5 CFH and repeat. Stop as soon as the bead runs clean. That approach finds the minimum effective flow, which is the flow least likely to create turbulence, and it gives you a documented number you can return to. Aluminum is less forgiving here because of its oxide and thermal behavior, so the same discipline pays off in the aluminum MIG welding guide. For a broader checklist of defects and their remedies, Lincoln’s MIG problems and remedies reference is a solid bench companion.

Surface Prep and Contamination: The Fix Most Welders Skip

If you could fix only one variable in your MIG setup, make it cleanliness. Lincoln Electric identifies improper surface condition, including oil, rust, paint, or grease, as the most common cause of weld porosity in GMAW. That ranking matters because cleaning is the variable you fully control. You cannot always change the wire lot or the weather, but you can decide whether the joint goes under the arc dry, bare, and free of oxide.

Contamination rarely announces itself. A film of coolant or a fingerprint looks harmless until the arc turns it into gas. By then the pool is freezing and the bubble has nowhere to go. TWI’s guidance on porosity makes the same point: preventing voids starts with controlling what enters the pool, and the joint surface is usually the source.

The contamination stack, in the order it appears

Most contaminated joints carry more than one layer. Work through them in the order a part actually gets handled:

• Mill scale and rust. These oxides do not melt at MIG temperatures. They sit between the pool and base metal, blocking wetting and trapping moisture underneath.
• Cutting oils and coolant. Sawing and drilling fluids vaporize and dissociate into hydrogen at arc temperature. The molten steel absorbs that hydrogen, then rejects it as it freezes.
• Shop dust. Dust is hygroscopic, so it holds moisture against the joint and adds another gas source.
• Fingerprints and sweat. Salt and water remain on the surface. A clean-looking fingerprint can produce a surprising cluster of pores.
• Anti-spatter spray or gel. Many products leave silicone or oil behind. Burned into the pool, they generate gas and can interfere with fusion.
• Paint and primer. Coatings vaporize ahead of the arc. The thicker the film, the more gas the pool must shed before it solidifies.
• Galvanizing. Zinc melts at about 420°F and boils near 1,665°F, far below arc temperature, so the coating flashes to vapor almost instantly.
• Adhesive residue. Labels, tape, and glue break down into gas, often right where two plates meet.
• Moisture. Dew, rain, and condensation between faying surfaces can stay trapped after the visible water dries. The arc then turns it into hydrogen and steam.

Weldable primers are not a free pass. Most manufacturers publish a maximum dry-film thickness, commonly around 20 micrometers, or about 0.8 mil. Above that limit, porosity becomes likely, especially in processes other than shielded metal arc welding. Check the product data sheet, measure the film, and remove the coating from the weld zone if you cannot confirm the thickness.

A cleaning sequence that works

Order matters as much as effort:

1. Degrease first. Wipe or spray with a suitable solvent and use clean, lint-free rags. Do not flush oil into a lap joint.
2. Remove scale and oxide mechanically. Grind, wire brush, or use a dedicated abrasive until you reach bright metal. Use stainless brushes and discs reserved for each material.
3. Degrease again. Abrasives can push oil and debris into the surface. A second wipe removes what the first pass spread around.
4. Weld within a short window. Freshly cleaned steel begins to oxidize and pick up moisture again. Clean only what you can weld the same shift.

Aluminum needs its own chemistry and tools; our guide to aluminum cleaner for welding covers that prep in detail. For steel, the rule is simpler: never use the same grinding disc on aluminum and steel, and keep grinding dust away from a cleaned joint.

Preheat as a moisture-control tool

Preheating is usually discussed as a cracking preventive, but it is also a practical drying step. Raising the joint to roughly 200–225°F drives off surface and joint moisture and gives you a useful test: if porosity disappears after preheating, moisture was contributing. Heat evenly, measure with a contact thermometer or temperature crayon, and do not confuse a warm surface with a clean one.

What clean actually means

Wire, Liners, and Consumables: Hidden Sources of Porosity

Ask a welder what they checked before chasing pinholes and you will hear about gas flow, voltage, and maybe the ground clamp. Almost nobody says the liner. Yet the liner is a ten-foot tube of shop air that your wire drags through on its way to the pool, and on a humid Monday morning it can be delivering a fine film of oil, dust, and condensation directly into the arc. The consumables are where porosity hides because they degrade slowly and invisibly, and they are the last thing anyone suspects.

Filler Metal Deoxidation: Reading the Last Digits

Every ER70S designation tells you how much deoxidation insurance the wire carries. ER70S-3 contains roughly 0.45–0.75 percent silicon and 0.9–1.4 percent manganese. ER70S-6 raises that to about 0.80–1.15 percent silicon and 1.4–1.8 percent manganese, which is why it tolerates light mill scale and moderate rust far better. ER70S-2 goes further: it is a triple-deoxidized wire with aluminum, zirconium, or titanium added alongside silicon and manganese, and it exists for genuinely dirty or rusty plate.

Those elements scavenge oxygen before it can combine with carbon to form CO at solidification, which is one of the mechanisms behind internal porosity described in TWI’s overview of porosity. Moving up the ladder is a legitimate fix when you are welding through a coating you cannot fully remove, or when the base chemistry is unusually rimmed or high-sulfur. It is a band-aid when the real problem is a wet joint or a leaking hose. A deoxidized wire will not save you from hydrogen, and it will not save you from air pulled in through a torn O-ring.

Wire Storage: The Spool Is a Sponge

Solid wire picks up surface moisture and flash-rusts in a humid shop, and that rust is iron oxide plus water sitting inches from the arc. Cardboard spools are worse than plastic because the cardboard itself absorbs moisture and holds it against the wire. Keep spools in a sealed cabinet or their original bag with desiccant, and never leave one hanging on the feeder over a damp weekend with the gun lying on a concrete floor.

Flux-cored wire is a different animal. Its flux core is hygroscopic by design, so storage requirements are stricter, and most manufacturers publish limits on exposure time and re-drying. A spool that has been open in a humid shop for weeks will produce porosity no amount of gas tuning can fix.

The Wire Path: Liner, Drive Rolls, Tips, and Nozzles

Contaminated gun liners collect grease, oil, dust, and dirt from the shop, then feed that mixture to the pool a few granules at a time. Stainless and high-nickel wires are especially susceptible to picking up those contaminants, and the damage shows up as scattered fine porosity plus a dull, gray bead. Check for the wrong liner size or material, kinked whips, and drive roll tension set so tight that it flattens the wire and flakes its surface. A worn contact tip lets current wander, which changes the arc without changing your settings. Spatter packed into the diffuser distorts the gas stream just as effectively as a draft.

Symptom	Likely Cause	Fix
Fine scattered pores, dull bead	Contaminated liner or oily wire	Replace liner, wipe wire, check spool storage
Erratic arc, popping feed	Crushed wire or worn contact tip	Reset drive tension, fit new tip
Pores plus poor wetting on scaled plate	Under-deoxidized filler	Step up to ER70S-6, clean the joint
Sudden porosity after a spool change	Rusted or damp wire	Fresh sealed spool, test coupon

When to Suspect the Gas Supply Itself

If the wire and liner check out, look upstream. A defective gas solenoid at the wire feeder or a damaged O-ring at the whip connection can introduce air into the shielding stream, and both fail intermittently, which is why the porosity comes and goes. A swapped or mislabeled bottle, or a supplier dew point problem, will do the same thing across every joint in the shop. Prove it the cheap way: fit a known-fresh cylinder, run a two-inch bead on clean scrap, and compare. If the pores vanish, the fault was the supply, not your technique. Lincoln Electric’s MIG problems and remedies chart is a good cross-check for symptoms that survive a cylinder swap.

Build two habits: wipe the wire with a clean lint-free rag every time you load a spool, and log liner changes on the feeder so you replace them on schedule instead of after a reject. Consumables are cheap; the rework loop they cause is not. For wires chosen for tougher alloys, our guide to alloy steel welding electrodes covers matching filler to base metal without overshooting on deoxidation.

Material-Specific Porosity: Aluminum, Stainless, Galvanized, and Primed Steel

Porosity on mild steel is usually a prep problem. Here it is a chemistry problem: aluminum hides behind an oxide film, stainless drinks hydrogen, galvanized steel boils a coating beside the arc, and primed plate hides a vapor source under clean metal.

Aluminum: The Oxide Film and the Hydrogen Cliff

Aluminum breaks the most rules at once. Its oxide layer melts near 3,700 F (about 2,040 C); the aluminum beneath melts around 1,220 F (about 660 C). The arc can collapse base metal while the oxide skin is still solid, so the pool never wets out and gas trapped under the crust becomes a pore.

That oxide will not burn off under the arc, so remove it first: degrease, chemical clean, brush with a stainless brush that has never touched steel, then scrape. Keep a grinding disc reserved for aluminum, because one that has cut carbon steel embeds iron particles that contaminate the surface and become corrosion sites.

Hydrogen is the other half of the problem. Liquid aluminum dissolves a great deal of it; solid aluminum holds almost none, so as the pool freezes, solubility collapses and hydrogen that cannot escape becomes a pore. Moisture, oil, condensation in a lap joint, and humid shop air all feed it.

Three habits follow. Prep and weld in the same session, because a surface cleaned last night re-oxidizes overnight. Run a push angle, match argon flow to your nozzle, and keep the gun close. Keep the arc short, since aluminum pulls heat away fast. For machine settings, read our complete practical guide to aluminum MIG welding, and how to prep aluminum right covers cleaners and brushes. aluminum boat welding handles hull joints; aluminum flux core welding covers the wire-feed-only route.

Stainless Steel: Hydrogen Sensitivity and Fume You Cannot Ignore

Stainless is quieter about porosity, which is why it gets missed until a leak or bend test finds it. Hydrogen is the principal gas again, and the proven prep is degrease, wire brush, degrease again, because brushing pushes oil and dust into the surface as easily as off it.

Gas choice shapes the risk. Tri-mix and argon-rich mixes stabilize the arc and improve wetting, while 75/25 argon-CO2 is a mild-steel choice that brings oxidizing and carbon-picking potential to stainless. Nitrogen arrives the way oxygen does: turbulence, a worn nozzle, excess stickout, or a fan moving air across the pool.

Stainless adds a hazard the others do not: hexavalent chromium fume, which OSHA regulates at a permissible exposure limit of 5 micrograms per cubic meter as an 8-hour time-weighted average. General ventilation is often not enough in a confined area, so use local exhaust at the arc and check 29 CFR 1910.252 for your situation.

Galvanized and Zinc-Coated Steel: Boiling Coatings and Trapped Vapor

Zinc boils far below the steel melting point, so galvanized coating turns to vapor the instant the arc reaches it, and some goes through the pool. The result ranges from scattered fine pores to large blowholes, often ringed by a yellowish-white oxide deposit.

Grind the coating back from both faying surfaces and a margin around them, back side included. Zinc oxide fume causes metal fume fever, so work upwind and use extraction. Geometry decides a lot: welding one side of a T joint leaves the far coating intact, but welding both sides traps vapor under the vertical member with nowhere to vent except through your first bead. Grind both and sequence so gas can escape.

That is why a galvanized job can pass a butt weld and fail every fillet.

Primed and Painted Steel: Thin Coatings, Big Problems

Weldable primers exist, and many manufacturers publish a maximum coating thickness, often a few thousandths of an inch, beyond which porosity becomes likely. Check the data sheet rather than assuming any grey primer is weld-through; too thick or wrong chemistry and it vaporizes into the pool.

Joint design amplifies it. A butt joint with a slight gap lets vapor escape upward along the seam. A T joint gives vapor a pocket; a fillet welded on both sides gives it two chances to be trapped. Where the spec allows, grind back at minimum on the member welded second.

High-Strength and Low-Alloy Steels: When Porosity Becomes Cracking

On mild steel, hydrogen porosity is cosmetic and a little section loss. On high-strength and low-alloy steels it initiates cracks: the same moisture and oil diffuse into the heat-affected zone and cause delayed cracking hours or days later. That is why low-hydrogen practice and preheat are standard, with dry consumables, controlled storage, and a procedure qualified for the material.

Deoxidized fillers help with oxygen-related porosity but cannot rescue a wet joint or damp wire. Our guide to alloy steel welding electrodes explains the classifications, and Alloy 20 welding procedures shows how strict prep and storage get on higher-alloy grades.

The Material-by-Material Shortcut Table

Material	Dominant gas	Cleaning	Key precaution
Aluminum	Hydrogen	Degrease, brush, scrape	Remove oxide; weld soon after cleaning
Stainless steel	Hydrogen, nitrogen	Degrease, brush, degrease	Correct gas; chromium fume extraction
Galvanized	Zinc vapor	Grind coating from both sides	Fume control; sequence to vent
Primed steel	Vaporized coating	Grind back where allowed	T joints trap worse than butt joints
High-strength steel	Hydrogen	Dry wire, clean surfaces	Low-hydrogen practice, preheat

One habit ties this together: run a two-inch test bead on scrap from the same lot before you touch the part. Each material fails in its own signature way, and the coupon tells you which one you are facing in under a minute.

Reading the Pattern: How to Diagnose Porosity Like an Inspector

Porosity rarely announces its cause, but it almost always leaves a signature. A welder who turns knobs after every pinhole is guessing. An inspector reads location, distribution, timing, and depth, then tests the most likely source. That shift from pore count to pattern is the fastest way to stop chasing your tail. It works on steel and on aluminum, where hydrogen sensitivity makes pattern reading even more valuable; our aluminum MIG welding guide covers those material-specific traps.

Start With Four Questions

Before touching the regulator, ask:

• Where are the pores? At the start, middle, or end? On the root, the face, or one side?
• How are they distributed? Scattered, clustered, in a line, or worm-shaped?
• When do they appear? First inch, last inch, only on the second pass, or only in the afternoon?
• Are they surface-breaking or subsurface? A light grind tells you more than another random test bead.

Those answers narrow the cause to a short list. Keep this matrix near the bench.

Pattern	Most likely cause	First check
Porosity at bead start	Gas surge, solenoid delay, long stickout before the arc stabilizes	Pre-flow setting and line purge
Porosity at bead end	Crater pipe or post-flow cut short	Pause and feed the crater; extend post-flow
Along one side of the bead	Gun angle, arc blow, or poor joint geometry	Angle, work clamp position, fit-up
Root only	Open root drawing air; no backing or purge	Root gap, backing gas, joint preparation
Second side of a T joint	Coating or primer trapped in the faying surface	Grind coating back before welding
Only in the afternoon	Drafts, rising humidity, fan or open bay door	Shield the zone; verify flow at the nozzle

Inspection: Match the Tool to the Question

Visual inspection is first and cheapest. Surface-breaking porosity can be found with penetrant testing, or magnetic particle testing on ferrous metal, which also catches some near-surface indications. Subsurface porosity needs radiography or ultrasonic inspection. For detecting and characterizing porosity, radiography is normally more effective than ultrasonics, particularly in thinner sections. Destructive bend and macro tests belong to procedure qualification, where you need to see the actual cross-section.

Isolate One Variable, Then Prove It

Change one thing, run a coupon from the same material heat, and record the result before changing anything else. If you alter gas flow, wire, and cleaning method together, a clean bead tells you nothing about which change worked. Keep a simple weld log with gas flow measured at the nozzle, wire lot, material heat number, cleaning method, and ambient conditions. Mixed or intermittent porosity usually means more than one source is active, and the log is how you separate them.

Worked Example: Porosity Only on Galvanized T Joints

A shop reported porosity exclusively on the second side of galvanized T joints. The first side ran clean; the second side showed clustered pores along the toe. Here is the sequence that confirmed the cause:

1. Visual inspection confirmed clusters on the second side only, with no porosity on the first side.
2. A coupon of clean mild steel, same wire and settings, welded without porosity. That largely cleared the gas supply and wire.
3. The galvanized joint was degreased and wire-brushed, but the zinc coating was left in place. It still foamed at the toe.
4. The coating was ground back about an inch from the weld zone on both faying surfaces, then cleaned and re-welded. The porosity disappeared.
5. The fix was repeated on a production joint with a controlled root gap. It stayed clean.

The lesson: the location pointed to trapped zinc vapor, not a mysterious gas-flow fault. Pattern first, then one controlled test, then a written result. That is how you diagnose porosity like an inspector instead of repairing the same joint twice.

Codes, Acceptance Limits, and Repair: When Porosity Actually Matters

A pore is not automatically a failure. This is where welders get themselves into trouble in both directions: one pinhole sends somebody grinding out a perfectly acceptable joint, while a bead full of wormholes gets shipped because “it’s just porosity.” Both reactions skip the only question that matters — what does the governing code actually allow?

That distinction is worth stating plainly. An imperfection is any interruption in the weld’s uniformity. It becomes a defect only when it exceeds the acceptance criteria in the applicable standard. TWI’s guidance on porosity makes the same point: acceptance depends on the specification, not on how the bead looks to the naked eye.

What the Major Codes Actually Allow

You don’t need to memorize every clause, but you should recognize the shape of these rules. Here is a representative summary:

Standard	Reported porosity limit	What it means on the floor
AWS D1.1	Porosity shall not exceed 3/8 inch in any linear inch of weld, and shall not exceed 3/4 inch in any 12-inch length.	Scattered small pores may pass; aligned or heavy concentrations usually won’t.
ASME B31.1	Porosity dimensions limited to no greater than 3/16 inch.	Pore size matters more than count in many piping situations.
API 1104	Porosity limited to 1/8 inch diameter.	Pipeline work tends to be strict on individual pore size.
ISO 5817	Quality levels B, C, and D, with limits tightening as the level rises.	Level B is the most demanding; D is the most permissive.

These are summaries, not substitutes for the standard. The project specification, drawing notes, and the engineer’s acceptance criteria always govern. A code may permit a pore that your customer’s spec rejects outright.

When Passing Visual Still Isn’t Good Enough

Some joints deserve a higher bar than the minimum. Pressure-retaining welds can leak through a pore that passes dimensional limits. Cyclic loading turns pore edges into crack initiation sites. Corrosive service loses protection where a pore interrupts the surface. Leak-tightness requirements and appearance-critical work may reject porosity that is structurally harmless. In those cases, the service condition — not just the code table — should drive the decision.

Repairing Porosity the Right Way

Never weld directly over porosity. It usually makes things worse by trapping more gas and creating a larger, deeper void under a fresh cap. The correct sequence is to grind or gouge out the affected area to sound metal, verify that removal is complete, re-prepare the joint, and re-weld to the qualified procedure. Localized defects call for localized excavation. If porosity is widespread, remove the entire weld and re-prepare the joint. For deep excavation, carbon arc gouging is often faster than grinding, provided the surface is cleaned afterward.

The economics are brutal but predictable. Rework consumes labor hours, wire, gas, and inspection time, and it pushes the schedule. A reject caught at the bench costs a few minutes. The same reject found at final inspection costs disassembly, re-inspection, paperwork, and possibly a missed delivery. The Fabricator’s overview of weld porosity makes the case that prevention is far cheaper than repair.

Documentation, Traceability, and Repeat Offenders

Record what you did: weld map, joint ID, WPS number, filler lot, gas flow, and repair history. When the same joint keeps producing porosity, that is rarely an operator problem. It points to the procedure, the gas supply, the joint design, or the prep method. Treat it as a process investigation, not a personal performance review.

Before you repair any code weld, talk to the inspector or engineer. Get the acceptance criteria in writing, agree on how the defect will be excavated, and confirm what inspection is required before and after re-welding. That conversation takes ten minutes. Guessing can cost a shift.

A Prevention System That Sticks: Daily Rituals, Settings, and the Mistakes to Stop Making

Diagnosing porosity is satisfying. Preventing it pays the bills. The shops that almost never fight it are not staffed by welding mystics; they run a boring, repeatable routine that catches small failures before they ever reach the arc. Here is how to build that routine, whether you weld two hours a week or eight hours a shift.

The Ten-Minute Morning Setup

Ten minutes before the first production bead buys back hours of grinding and rework. Run it in the same order, on every machine, every day:

1. Leak-down test. Pressurize the gas system for 15–20 seconds, shut the cylinder valve, and watch the regulator dial. A slow coast-down means gas is escaping between the bottle and the gun.
2. Nozzle and diffuser. Look through the nozzle against a light. Spatter bridging the diffuser ports starves one side of the pool and shows up as porosity along a single edge.
3. Contact tip. Check for an ovalized bore, spatter buildup, and correct recess inside the nozzle. A worn tip gives erratic current and an unstable arc.
4. Flow verification at the nozzle. Use a flowmeter, not the regulator dial. A small-diameter wire on thin gauge typically runs near 25 CFH; heavy plate with large-diameter flux-cored wire needs considerably more.
5. Wire spool and liner. Look for rust bloom, a kinked whip, and drive roll tension that is crushing the wire instead of feeding it.
6. Material cleanliness. Confirm the joint is dry, bare, and oxide-free before you strike an arc, not after you see pinholes.
7. Test bead. Two inches on offcut from the same heat, so today’s gas, wire, and prep prove themselves before you commit a production joint.

Parameter Discipline: Where the Settings Actually Fit

Settings are not a magic number copied from a chart. They are a set that matches wire diameter, amperage, voltage, and travel speed to the joint, thickness, and position. Too much voltage lengthens the arc and widens the gas envelope, which feels forgiving until the arc starts pulling atmosphere and moisture into the pool. Too little voltage gives spatter, a stubbing arc, and cold lap at the toes. Excessive electrical stickout is just as damaging: it overheats the wire, destabilizes transfer, and can dissolve moisture straight from the air into the pool.

When something goes wrong, resist moving three dials at once. Lincoln Electric’s MIG problems and remedies guidance and Miller’s defect troubleshooting hub both work best when you change one variable and confirm the result on scrap.

The Mistake List, and What to Do Instead

Nearly every recurring porosity problem traces back to one of these eight habits. Each has a cheap correction.

Mistake	Why it backfires	Do this instead
Cranking gas flow to cure porosity	Excess flow creates turbulence that aspirates air into the shield	Verify flow at the nozzle, start at the low end, increase only if contamination persists
Changing three variables at once	You never learn which one fixed it, so the fault returns	One change, one coupon, one written note
Welding over contamination because “it burns off”	Oil, paint, and zinc vaporize into the pool instead of vanishing	Degrease, grind to bare metal, degrease again
Skipping the root purge	An open root pulls atmosphere straight into the back of the weld	Purge or back with shielding gas, and bump the trigger after any line break
Using one grinder disc for aluminum and steel	Steel particles embed in aluminum and contaminate every following pass	Dedicated discs and stainless brushes per material — see this aluminum cleaner and prep guide
Ignoring humidity and drafts	A bay door or shop fan strips the shield on a calm-looking day	Check airflow, morning dew, and afternoon humidity swings
Trusting a regulator as a flow meter	Inlet pressure is not flow rate at the nozzle	Use a calibrated flowmeter at the gun
Never documenting what worked	The same fault costs you a full shift next month	Keep a shared troubleshooting log

On aluminum specifically, consumable discipline goes hand in hand with the settings covered in this practical aluminum MIG welding guide.

People, Process, and the Log That Pays for Itself

Systems outlive individuals. Train new hands on the morning routine instead of letting them inherit bad habits, and qualify procedures so everyone welds to documented parameters rather than personal preference. Then keep one shared log: gas flow, wire lot, material heat number, cleaning method, ambient conditions, outcome. When porosity appears, the log usually narrows it to one or two suspects in minutes. The fabricator’s checklist of possible causes is far easier to work through when you already know what changed since the last good weld.

The Printable Checklist

• ☐ Leak-down test passed (15–20 seconds, no dial drop)
• ☐ Nozzle, diffuser, and contact tip clean and correct
• ☐ Flow verified at the nozzle with a flowmeter
• ☐ Wire spool sound, liner clean, tension correct
• ☐ Joint dry, bare, oxide-free
• ☐ Two-inch test bead passed on same-heat scrap
• ☐ Drafts and doors checked
• ☐ Log entry written

Back to That Tube on the Bench

Remember the fabricator burning a whole bottle of gas on 1/8-inch tube, convinced the machine was at fault? The fix was never one heroic adjustment. It was a dry joint, a seated O-ring, a nozzle half-blocked for a week, and a flow rate that crept up until the shield blew itself apart. Porosity is a system problem, and it yields to a system fix. The habit that matters most: change one variable, prove it with a coupon, write it down. Do that, and pinholes stop being a mystery and become a checklist item you already cleared before the arc ever started.

Frequently Asked Questions

What is the most common cause of porosity in MIG welding?

Loss of shielding gas coverage is the most common single cause, usually from incorrect flow rate, a leak, a spatter-clogged nozzle, too much stickout, or a draft. Dirty base metal runs a very close second, especially oil, rust, paint, and moisture that release hydrogen at arc temperature.

Can I weld over porosity to fix it?

No. Welding over porosity usually traps more gas and can push the defect deeper or spread it along the bead. The correct repair is to grind or gouge out the porous metal to sound material, confirm removal, re-prepare the joint, and re-weld to the qualified procedure.

Does increasing gas flow fix porosity?

Only if the flow was genuinely too low. Once flow is above the correct range, extra gas creates turbulence at the nozzle that aspirates surrounding air into the plume, which can cause the very porosity you are trying to eliminate. Verify flow at the nozzle with a flow meter, not at the regulator.

Why does porosity only appear at the start of the weld?

A start-of-bead problem usually means the gas line was not purged, the pre-flow is too short, or the nozzle was still full of air when the arc initiated. Bumping the trigger for a second or two before welding, and adding a small pre-flow delay, typically solves it.

Why do my aluminum MIG welds pit even after cleaning?

Aluminum is highly susceptible to hydrogen porosity because its hydrogen solubility drops sharply as the metal freezes. Clean with a dedicated stainless brush and separate abrasives, avoid hydrocarbon solvents that leave residue, use the correct argon flow, push the gun, and keep the arc length short and stable.

Is a small amount of porosity acceptable?

It depends entirely on the governing code and the service of the part. AWS D1.1, ASME B31.1, API 1104, and ISO 5817 each set different limits, and a project specification can be stricter. Cosmetic porosity may pass; porosity in a pressure or fatigue-critical weld often will not.

Can humidity cause MIG welding porosity?

Yes. High humidity promotes condensation on cold plate, inside lap joints, and between a backing bar and the base metal, and it can be drawn into the pool as hydrogen. Preheating the joint to roughly 200–225 F to drive off moisture is a practical test and often a fix.

How do I tell the difference between porosity and a gas pocket from a coating?

Look at location and pattern. Coating-driven gas tends to produce elongated wormholes or concentrated clusters near the coated surface, often in T joints or on the second side of a fillet. Distributed fine porosity across the whole bead points more toward shielding gas contamination.

Should I change my wire to fix porosity?

Only after confirming the gas and surface prep are correct. A more heavily deoxidized wire such as ER70S-6 or ER70S-2 can help with marginal plate condition, but it masks contamination rather than removing it and may not suit the procedure or code.