Reasons for Porosity in Welding: 12 Causes and Fixes

The bead looked perfect until the grinder hit it—and suddenly the weld was a honeycomb of tiny pin holes. Three hours later, the part was cut out, the joint re-prepped, and a frustrated welder was asking the same question every fabrication shop eventually asks: where did all that gas come from? The answer is rarely one single culprit, and that is exactly why porosity keeps coming back.

Why Porosity Keeps Showing Up in Otherwise Good Welds

The job had been running fine all week. Same machine, same wire, same operator, same settings written on the whiteboard. Then Monday morning, the first three parts came off the bench looking tight and uniform — until someone wiped the cap with a rag and found a neat line of pinholes running along the toe of the weld. Nothing had changed. That is the detail that makes porosity so maddening: it appears in welds that were, by every visible measure, going well right up until they weren’t.

The consequences stack up quickly. The weld gets gouged out, the joint is re-prepped, the part is re-welded, and the whole sequence is re-inspected — often while a schedule slips and a customer waits for an answer. On a certified job, a failed NDT report drags in concessions, paperwork, and questions nobody enjoys answering twice. Understanding the real reasons for porosity in welding is the only reliable way out of that loop, because the pinholes are the last link in a chain that almost always started somewhere else entirely.

What Porosity Actually Means

Strip away the drama and the definition is straightforward. AWS A3.0 defines porosity as “cavity-type discontinuities formed in weld metal by gas entrapment during solidification.” In shop language: gas was dissolved in the molten pool, and it did not escape before the metal froze around it.

Those cavities are generally spherical, which is why codes tend to treat them as less severe than cracks or lack of fusion. A round hole does not concentrate stress the way a sharp-tipped crack does. But “less severe” is not “harmless.” Every pore removes cross-section, and clusters of them reduce fatigue life and give corrosion a place to start. Porosity can sit on the surface or hide entirely beneath it, and it shows up as scattered pinpoints, tight clusters, aligned lines, or elongated wormholes.

Here is the distinction that trips up more welders than any other: porosity is a discontinuity, not automatically a defect. A discontinuity is any interruption in the expected structure of the weld. A defect is a discontinuity severe enough to be rejectable under the applicable code. AWS D1.1 and similar standards accept limited porosity rather than banning it outright, which means “there is porosity in this weld” and “this weld is rejectable” are two very different statements. Confusing them leads to unnecessary grinding on one hand and missed rejections on the other.

The Reasons for Porosity in Welding, Mapped Out

Because the causes are so widely distributed across the process, this guide is organized as a diagnostic path rather than a list of tips. First comes the metallurgy: how gas dissolves in molten steel or aluminium, why solidification forces it out, and why cooling rate and pool size change the result. Next is the framework that carries the rest of the article — the three gas sources behind nearly every case: hydrogen, air, and trapped process or reaction gases.

From there the coverage gets specific. You will find process-by-process causes across SMAW, GMAW, FCAW, GTAW, and SAW, followed by the contamination chain that runs through consumables, storage, and surface prep. Then come the equipment and setup failures that quietly steal shielding gas, the pattern-reading skills an inspector uses to localize a problem, and a plain-language summary of what the codes actually permit. The final sections turn all of it into a prevention checklist and dismantle the myths that keep shops repeating the same expensive mistakes. Aluminium gets extra attention throughout, since its oxide layer and hydrogen behaviour make it a repeat offender — the prep side alone is worth reviewing in a dedicated guide to aluminum cleaner for welding.

This is written for everyone who has stared at a pinhole and wondered where it came from: the hobby welder running a small MIG machine in a garage, the pipe hand working to a procedure on a cold morning, the fabrication shop chasing a run of failed visual inspections, and the inspector who has to explain a rejection without starting an argument. Each of them sees a different part of the same problem. The welder sees the arc, the shop sees the schedule, and the inspector sees the code. The gas source sits somewhere between all three.

What Porosity Actually Is: Gas Solubility, Solidification, and Trapped Bubbles

Porosity looks random. Metallurgically, it is remarkably orderly: gas dissolves in the liquid pool, the pool freezes, and the gas has nowhere to go. TWI describes porosity as the absorption of nitrogen, oxygen, and hydrogen in the molten weld pool, which is then released on solidification and trapped in the weld metal.

That definition matters because it moves the conversation away from bad luck and toward solubility. At welding temperature, liquid metal can hold far more gas in solution than the same metal can hold once solid. Steel and nickel behave this way. Aluminium is extreme: hydrogen solubility in aluminium drops by roughly a factor of 20 as the metal solidifies, as TWI’s aluminium porosity guidance explains. The last liquid to freeze is therefore the most dangerous place in the weld.

From Dissolved Gas to a Visible Pore

Gas does not spontaneously boil out of perfectly clean liquid metal. Bubbles need a starting point, and in a weld pool they find one easily. Bubbles nucleate preferentially on existing substrates such as dendrite boundaries, second-phase particles, or oxide films rather than in a homogeneous liquid. That is one reason dirty aluminium and heavily deoxidised steels seem to produce pores so readily: the pool is full of convenient nucleation sites.

Once a bubble forms, it grows as surrounding gas diffuses into it. Then comes the race that decides whether you see a pinhole or a sound weld. The bubble must rise through the liquid and break the surface before the advancing solidification front overtakes it. If it escapes, the gas leaves the weld. If the freezing front catches it, the bubble is locked in as a pore.

Cooling rate is the referee. Faster cooling shortens the escape window, so more gas is trapped. Increasing heat input or preheat slows solidification and gives dissolved gas more time to leave the pool. Larger pools are a trade-off: they contain more dissolved gas and bubbles have farther to travel, but their thermal mass also keeps them molten longer. Which effect wins depends on the gas source, turbulence, and joint geometry.

The Main Gas Families

• Hydrogen. Moisture in electrodes and flux, hydrocarbons such as oil and grease, and hydrated oxide layers all feed hydrogen into the arc. Aluminium’s oxide layer is especially good at holding moisture.
• Nitrogen and oxygen. Air entrainment from poor shielding, leaks, drafts, or excessive stickout introduces both gases. Nitrogen can form porosity in steel; oxygen reacts with the pool and contributes to oxide inclusions and gas evolution.
• Process and reaction gases. CO₂ shielding gas breaks down in the arc and can react with the pool. Zinc vapour from galvanised steel, sulfur in free-machining steel, and primer breakdown products can all add gas. Shielding gas turbulence itself can shear air into the stream.

Aluminium welders meet all three families at once, which is why surface preparation deserves as much attention as torch settings. Our guides to aluminium MIG welding and aluminium cleaner for welding cover the practical prep routine in detail.

Reading the Shape of Porosity

Morphology is a clue, not a verdict, but it narrows the search quickly.

Morphology	What it looks like	What it often suggests
Scattered or distributed	Fine pores spread across the bead	Gas coverage, gas quality, or general contamination
Clustered	A local group of pores	Localised contaminant, arc interruption, or a dip in shielding
Linear or aligned	Pores in a line along the joint	Air drawn along a gap, fusion-line contamination, or a root defect
Piping or wormhole	Elongated, tunnel-like cavities	A large volume of gas, typically gross surface contamination or excessively thick primer or coating
Crater pipe	A hole at the end of the weld	Poor crater fill or insufficient post-flow
Surface-breaking	Pinholes open to the surface	Gas escaping late, contamination at the start, or a unstable pool surface

Once you can picture the bubble racing the freezing front, porosity stops being mysterious. The next step is identifying which gas is winning the race.

The Three Gas Sources Behind Nearly Every Porosity Problem

Here is the mental shift that separates a welder who fixes porosity from one who fights it for a week: stop asking “what setting is wrong?” and start asking “which gas is this?” Every pore in a weld is a fingerprint of something that should not have been in the pool. Nearly all of them trace back to just three sources.

Name the gas, and the search narrows fast. This framework is the backbone of every diagnosis in the rest of this article.

Hydrogen: The Most Common Uninvited Guest

Hydrogen is the smallest atom in the shop and the hardest to keep out. It dissolves readily in molten steel, stainless, and aluminium, then comes out of solution as the pool freezes. When it cannot escape, it becomes bubbles.

The usual hiding places:

• Moisture inside electrode coverings and flux that were never dried or stored properly
• Hydrated oxide layers on aluminium, which release hydrogen at arc temperature
• Rust and rust-stained plate
• Oil, grease, cutting fluid, paint, marker, and adhesive residue
• Humidity in the shop air, especially on damp mornings and near open doors

Aluminium is the extreme case. Published aluminium welding research has shown that as little as 2 ppm of hydrogen in the molten aluminium, or roughly 250 ppm of hydrogen in the shielding gas, can be enough to produce porosity. The TWI guidance on aluminium porosity describes the same mechanism from a practical angle.

This is why prep matters more on aluminium than on almost any other metal. A dedicated stainless brush and a proper solvent wipe are not cosmetic steps; they are gas control. If you are unsure which products and sequence to use, this guide to aluminum cleaner for welding walks through the full prep routine.

Air: Nitrogen and Oxygen Sneaking Past the Shield

Air is roughly 78% nitrogen and 21% oxygen, and neither belongs in the arc. Air gets in when the shield is lost: flow set too low, flow set so high that the stream turns turbulent, a leak in the hose or torch body, a draft pushing the envelope away, or stickout stretched too far for the gas to cover.

The numbers here are sobering. According to TWI’s technical reference on porosity, as little as 1% air entrainment in the shielding gas can cause distributed porosity, and more than 1.5% typically produces gross, surface-breaking pores. You cannot see 1% air in a clear gas stream, which is exactly why this failure mode fools so many people.

Trapped Process and Reaction Gases

Some porosity is manufactured by the weld itself. The arc creates gas faster than the pool can release it, and the result looks identical to a shielding fault.

• Zinc. Galvanized coatings and zinc-rich paints vaporize at arc temperature and generate copious fume and gas. Entrapment is especially likely in T-joints fillet welded on both sides, where the second pass traps what the first one released.
• CO₂ breakdown. Carbon dioxide shielding dissociates in the arc. Most of it recombines, but a turbulent or unstable arc can leave gas behind in the pool.
• Sulfur. Free-machining steels with high sulfur content can generate porosity through sulfur-based reactions in the pool.
• Primers. Low-zinc “weldable” primers are usually fine below the manufacturer’s maximum thickness, often around 20 µm. Thicker layers promote porosity, especially in processes other than SMAW.

Matching the Source to the Signature

Gas source	Typical appearance	First corrective action
Hydrogen (moisture, oil, hydrated oxide)	Scattered fine pores, often clustered near the fusion line	Degrease, remove oxide, dry electrodes and flux
Air (nitrogen and oxygen)	Distributed porosity across the bead; gross surface-breaking pores above 1.5%	Check flow at the nozzle, fix leaks, shorten stickout, block drafts
Zinc vapor or paint	Large irregular voids, wormholes, heavy fume	Grind back the coating before welding
Primer breakdown	Fine to medium pores in the cap and along edges	Reduce primer thickness or remove it at the joint
CO₂ dissociation or sulfur reactions	Internal porosity with a relatively clean-looking surface	Review gas mix, arc stability, and base metal chemistry

Three questions, then, at every failure: is there a hydrogen source on this joint, is the shield actually reaching the pool, and is the weld generating its own gas? Answer those and you are already ahead of most of the porosity in welding.

Process-by-Process Causes: SMAW, GMAW, FCAW, GTAW, and SAW

Every welding process fails in its own dialect. A pinhole pattern that means “wet electrode” on a stick weld may mean nothing on a TIG bead, and the GMAW habit that ruins a clean joint has no equivalent in submerged arc. This breakdown points you to the most likely mechanism for the equipment in your hand.

One principle ties all five together: the more a process relies on liquid shielding, the less forgiving it is of dirt. Slag-generating SMAW and FCAW tolerate surface contamination better than GMAW because slag helps clean the metal surface, as Lincoln Electric notes. GMAW has only the deoxidizers alloyed into the wire working for it.

Shielded Metal Arc Welding (SMAW): Moisture and Arc Length

Stick welding’s biggest enemy is water. Electrode coverings are hygroscopic and pull moisture from humid air; that moisture breaks down in the arc and releases hydrogen into the pool. Poor baking, sloppy storage, or an opened can in a damp shop causes pinholes before the second arc.

• Electrode moisture: low-hydrogen rods need baking and hot storage; a rod left on a bench overnight is no longer low-hydrogen.
• Long arc length: excess voltage and a wandering arc pull air into the stream and burn off the covering’s protection.
• Rusty or scaled plate: mill scale and rust carry moisture and oxygen into the pool.
• High-sulfur steel: free-machining grades generate gas and provoke porosity plus hot cracking.
• Slag confusion: deep surface pinholes may be trapped slag or arc strikes. Chip them out and look again before blaming gas.

Check rod condition first. It is the cheapest fix and the most common cause.

Gas Metal Arc Welding (GMAW): The Least Forgiving Process

MIG has no slag to absorb impurities and no flux to scavenge oxygen. Protection comes from deoxidizers alloyed into the wire: silicon, manganese, and traces of aluminium, zirconium, or titanium. Wire choice matters. ER70S-6 carries roughly 0.8–1.15% silicon and 1.4–1.8% manganese, the highest deoxidizer levels among common ER70S carbon steel wires. ER70S-3 sits around 0.45–0.75% silicon and 0.9–1.4% manganese, making it noticeably less forgiving of dirty plate. Switching to S-6 often separates porosity from a clean bead.

• Shielding loss: low flow, high-flow turbulence, leaks, or a clogged nozzle.
• Wrong gas: 100% CO2 on stainless, or an argon mix on thick plate where short-circuit transfer cannot penetrate.
• Dirty wire: rust, drawing lubricant, or dust on the spool.
• Feed problems: worn drive rolls, a kinked liner, or the wrong liner size create erratic feed and irregular porosity. Aluminium MIG shares this mechanism, where drive-roll slip and poorly wound spools are documented by TWI.
• Excessive stickout: every extra millimetre of wire beyond the nozzle is unprotected and lengthens the path for air.
• Short-circuit transfer on thick plate: cold, fast-freezing puddles trap gas that never had time to rise out.

Flux-Cored Arc Welding (FCAW): Moisture Inside the Wire

Flux-cored wire carries a core that behaves like a sponge for humidity. Wire left on the machine or stored in an unheated container absorbs moisture that vaporizes in the arc and shows up as scattered porosity. This is the biggest FCAW cause and the easiest to overlook because the wire looks dry.

Gas-shielded and self-shielded wires differ. Gas-shielded FCAW still needs proper flow and wind protection, so it inherits most GMAW shielding problems. Self-shielded wire generates its own gas and tolerates wind better, but is more sensitive to voltage and damp flux. Low voltage is a classic mistake: the arc becomes short and stubby, the flux does not fully react, and gas gets locked in. Inconsistent feed from excessive roll pressure or a worn liner produces the same stop-start turbulence seen in MIG.

Gas Tungsten Arc Welding (GTAW): Contamination and Shielding Loss

GTAW porosity usually comes from contamination or lost shielding, not the power source. Check the tungsten, gas path, and part before suspecting the machine.

• Contaminated tungsten: dipping the electrode in the pool or touching filler to the tip deposits junk that spits into the weld.
• Insufficient trailing shield on stainless: the hot bead oxidizes behind the torch, and lost protection pulls gas into the pool.
• Torch angle: too much tilt breaks the gas envelope and lets air curl into the arc.
• Gas lens problems: a cracked lens, wrong-size cup, or clogged screen produces uneven coverage.
• Filler handling: wire wiped with a dirty rag, or not wiped at all, drags oil and moisture into the pool. Use a clean, lint-free cloth and a suitable solvent, then let it dry.

For aluminium, the hydrated oxide layer is a hydrogen reservoir, so pre-cleaning matters more than any setting. Our guide to the right aluminum cleaner for welding covers brushes and solvents that keep that layer from forming again before you strike the arc.

Submerged Arc Welding (SAW): Flux Is the Whole Story

SAW hides the arc under granular flux, so most porosity traces back to that flux. Damp flux releases hydrogen and oxygen; recycled flux that picked up dirt, rust, or moisture does the same. Coverage depth matters too: too thin a layer exposes the arc, while excessive depth causes poor gas escape and surface pitting.

High current adds nitrogen pickup from poorly protected arcs, and the deep, fast-freezing pool gives bubbles little time to escape. SAW is usually thick-section work, so internal porosity is expensive to find and remove. Store flux dry, dry it before use, and never sweep reclaimed flux off a dirty floor back into the hopper.

Material-Specific Cases: Aluminium and Stainless

Aluminium deserves special mention. Its oxide layer melts at a higher temperature than the base metal and holds moisture. Hydrogen solubility is high in the melt and drops sharply on freezing, so scattered or clustered porosity is common, especially with erratic MIG feed. For setup, see our complete practical guide to aluminum MIG welding; for flux-cored options, this piece on aluminum flux core welding explains what works.

Stainless steel is sensitive to hydrogen and nitrogen and needs adequate back-purge on pipe and thin sheet. Poor purge leaves oxygen and nitrogen inside the joint, which shows as root porosity. Trailing shields on the cap side protect against oxidation, not porosity, but a broken shield often signals both.

Process Comparison: Signature and First Check

Process	Typical porosity signature	First thing to check
SMAW	Scattered fine pores, often along bead center	Electrode storage and baking record
GMAW	Fine scattered pores, sometimes clustered at starts and stops	Gas flow at nozzle and wire feed consistency
FCAW	Scattered pores across the whole pass	Wire moisture and spool storage
GTAW	Surface pinholes, clusters at restart points	Tungsten condition and gas path integrity
SAW	Internal porosity, often found only by RT or UT	Flux dryness and coverage depth

Treat this table as a starting point, not a verdict. The next sections explain why these checks keep appearing and how to close the gap between what the settings say and what the arc actually sees.

Consumables, Storage, and Surface Prep: The Contamination Chain

Porosity rarely starts at the arc. It starts when a rod is opened, a spool unwrapped, or a sawn tube left sitting overnight. Base metal, filler, gas, flux, coatings—each is a link in a chain, and one damp link is enough to push gas into the pool. Here is the chain to walk before you touch a single dial.

Base Metal: Where Moisture and Oil Hide

Most welders clean what they can see. Gas comes from what they cannot. Cutting oils used to saw tubing or drill holes are a classic shop contaminant, and they creep along the cut edge and into the joint gap where the arc will eventually burn them into hydrogen and carbon-bearing gas. Rust, mill scale, and paint behave the same way: they are not just dirt, they are chemical feedstock.

Two hiding places catch experienced hands. The first is the faying surface between lapped plates, where moisture sits in the gap and flashes to vapor as the heat arrives. The second is between a butt joint and its backing bar—often the dampest, least-visible surface in the whole assembly. On aluminium, the hydrated oxide layer holds water at the molecular level, which is why a dry-looking extrusion can still produce porosity. Proper prepping matters more than most people expect; our guide to aluminum cleaner for welding walks through the sequence step by step.

Filler Metal and Flux: The Storage Problem

Low-hydrogen SMAW electrodes are only low-hydrogen while they are dry. They are typically held in ovens around 250–300 °F (120–150 °C) and must stay protected once removed—sealed containers, portable quivers, and a short working window. A rod left on a bench overnight has already begun re-absorbing moisture from the air, and no amount of arc-time skill will fix that. If you are unsure how a given electrode class should be stored and handled, our all-position welding rod guide covers the practical basics.

Wire has the same weakness in a subtler form. Filler left exposed in an open workshop can absorb moisture into its oxide layer, which is precisely why aluminium wire should stay packaged until the moment of use—a point reinforced in AWS’s guidance on welding aluminium. Flux-cored wire picks up moisture through its flux core, and submerged-arc flux is worse still: damp flux, recycled flux, or a shallow flux layer all invite porosity. Wire feeders and liners also deposit lubricant and debris onto the wire, so a wipe-down before the drive rolls is cheap insurance. On the aluminium side, our aluminum MIG welding guide covers wire feed and cleanliness in more depth.

Shielding Gas and Cylinder Condition

Gas quality is an input like any other. Moisture can ride along in poor-quality gas, and a cylinder with a damaged valve or a contaminated outlet is a genuine hazard, not just a porosity risk—never attempt to use one, and never oil a regulator fitting. A trustworthy supplier removes most of this variable, which is why the source matters; see this AllGas welding supply guide for what to look for.

Coatings and Primers

Weldable primers are only weldable within limits. Thickness above the manufacturer’s maximum, often around 20 µm, is a recognized cause of porosity, especially outside SMAW. Galvanized zinc, residual paint, and shop primers all break down in the arc and release gas. The Fabricator’s overview of weld porosity is a useful sanity check on how much coating is too much.

Surface Prep in the Right Order

Sequence matters, because the wrong order just spreads contamination around.

1. Degrease first—before grinding, so you are not driving oil into the surface.
2. Remove coatings—grind back primer, paint, and zinc to bare metal.
3. Remove oxide—a clean, dedicated stainless brush or carbide tool, never one that has already chewed on carbon steel.
4. Weld within a defined window—freshly cleaned surfaces re-oxidize and re-absorb moisture, so clean-and-weld in the same shift.

One safety warning worth repeating: Lincoln Electric cautions that chlorinated degreasers such as trichloroethylene near the arc can react with the arc and produce toxic gases. The OSHA welding safety fact sheet covers fume and chemical exposure in more detail. Degrease, let it dry, and keep the solvent well away from the arc.

Equipment and Setup Failures That Steal Your Shielding Gas

If the filler metal and the part are clean, the next suspect is the hardware. Shielding gas only does its job when it reaches the pool as a smooth, continuous envelope. Between the cylinder valve and the arc, it passes through a regulator, flowmeter, hose, solenoid, torch body, liner, diffuser, contact tip, and nozzle. Every connection is a possible leak, and every restriction is a possible source of turbulence. The frustrating part is that the regulator can show a perfect reading while the nozzle is starved or quietly breathing air.

Follow the Gas Path From Cylinder to Arc

A loose fitting or cracked hose downstream of the solenoid often pulls air into the stream rather than leaking gas out. The same happens with a flattened torch O-ring, a split liner, or a damaged torch body. Inside the torch, spatter can block part of the diffuser or bridge the nozzle opening. The flowmeter still reads correctly because it is measuring gas upstream of the problem, not coverage at the pool.

This is why a gauge reading is not proof of shielding. A damaged contact tip recess, a partially blocked nozzle, or a leaking gas hose can reduce coverage without moving the regulator needle. Check the whole path, not just the number on the flowmeter.

Flow Rate: Too Little and Too Much

Low flow is the obvious problem. Excessive flow is the one that catches experienced welders. Blasting gas through the nozzle creates turbulence, and that turbulence can draw nitrogen and oxygen from the surrounding air into the weld area. In other words, turning up the gas can make porosity worse.

Set flow according to the process, nozzle size, and working conditions, then verify it at the nozzle with a flow tester. If the gas is still turbulent, check for restrictions, a worn diffuser, or a nozzle that is too small for the job. [Miller’s GMAW troubleshooting guide](https://www.millerwelds.com/en-us/resources/knowledge-hub/mig-welding/how-to/troubleshooting-weld-defects){target=”_blank” rel=”noopener”} lists inadequate shielding, wrong gas, dirty wire, dirty workpiece, and excessive stickout among the primary causes of porosity, which is a useful reminder that gas flow alone is never the whole story.

Nozzle, Contact Tip, and Stickout

Nozzle condition matters more than most operators realize. Spatter buildup disturbs the gas flow and can partially close the opening. A nozzle that is too small for the current and joint configuration gives poor coverage. A recessed contact tip increases the distance between the gas exit and the pool, while a tip that protrudes too far can be damaged or fouled.

Miller recommends keeping the nozzle roughly 1/4 to 5/8 in. (6–16 mm) from the workpiece and the wire no more than 1/2 in. (13 mm) beyond the nozzle. Excessive stickout stretches the shielding envelope and exposes the molten metal to air. On aluminium, where the oxide and hot metal are especially sensitive to contamination, setup discipline is even more important; our Aluminum MIG Welding: The Complete Practical Guide covers the practical details.

Drafts, Fans, and Wind

Shielding gas is fragile. An open shop door, a ceiling fan, a cooling fan aimed at the weld, or a ventilation system can strip the envelope away without making a sound. AWS D1.1 requires protection from wind for gas-shielded processes, with a commonly enforced limit around 5 mph (2.2 m/s). If you can feel a breeze on your face, assume the pool can feel it too. Screens and temporary tents are cheap compared with cut-out and rework.

Polarity, Cables, and Work Clamp Placement

Electrical problems are secondary porosity contributors, but they can mimic gas faults by destabilizing the arc. Verify polarity for the process and wire being used, inspect cable connections for heat damage or looseness, and place the work clamp on clean, bare metal close to the joint. A poor work connection causes erratic current, inconsistent penetration, and an arc that is harder to shield.

A Troubleshooting Sequence That Saves Time

Work from the cheapest and most likely checks toward the most invasive:

1. Clean the nozzle and inspect the contact tip.
2. Check gas flow at the nozzle with a flow tester.
3. Inspect hoses, fittings, O-rings, and the torch body for leaks or damage.
4. Verify the gas type and flow rate against the WPS.
5. Reduce stickout and correct torch angle.
6. Shut down fans, close doors, and erect a wind screen.
7. Replace the liner, diffuser, or torch components if the problem persists.

Finally, hold the gun over the bead until the crater solidifies. That small habit helps prevent crater pipe and end-of-weld porosity, and it costs nothing but a second of patience.

Reading the Pattern: How to Diagnose Porosity Like an Inspector

Finding porosity is one thing. Figuring out why it’s there is what stops a shop from repeating the same mistake. An inspector doesn’t just count pinholes; they read the weld for clues. Porosity location within the bead, whether root, mid-wall, or cap, narrows the likely gas source dramatically. Add pore shape and timing to that, and the culprit usually surfaces fast.

Pattern Recognition: Let the Bead Tell the Story

Not all porosity looks the same, and each form points toward a different gas source. Start with location, then shape, then when it appears.

• Scattered fine porosity across the whole bead: a general loss of gas coverage or poor gas quality. Think low flow, contaminated shielding gas, or a small leak in the delivery system.
• Linear porosity along the fusion line: air being drawn into the joint. Common with tight fit-up, poor cleaning, or insufficient purge on a root pass.
• Clustered porosity in one spot: a concentrated contaminant such as oil in a crevice, a patch of rust, or an arc interruption that let atmosphere in.
• Wormholes (piping porosity): elongated cavities that can run through the weld, usually from gross contamination or thick coatings like heavy mill scale or zinc on galvanized steel releasing vapor into the pool.
• Surface pinholes at the start of a bead: gas purge timing. The shielding gas hasn’t finished displacing air from the torch.
• Surface pinholes at the crater: poor crater fill or short post-flow, leaving the puddle exposed as the arc dies.

These clues matter most on tight-tolerance work. Intricate repairs like alloy wheel welding, or precision jobs such as aerospace TIG welding, leave little room for interpretation because the acceptance limits are strict.

Inspection Methods and Their Practical Limits

Once the pattern is clear, confirm how deep the voids go. Each method sees something different.

• Visual inspection: AWS notes that visual inspection by trained inspectors catches the majority of defects that would otherwise be found later. It’s the cheapest option, but it cannot reliably reveal small or fully buried pores. A bright light and a magnifier help.
• Penetrant testing (PT): good for surface-breaking porosity. Dye seeps into open pores and shows after cleaning. Affordable, but useless if the void is closed to the surface.
• Magnetic particle testing (MT): best for ferrous metals. Effective on surface and near-surface porosity in steel, and it can flag indications the eye misses. Like PT, it won’t reach buried pores.
• Radiography (RT): the usual choice for internal porosity. TWI notes radiography is normally more effective than ultrasonics for detecting and characterizing porosity, though small pores in thick sections stay difficult. It maps size and distribution well.
• Ultrasonic testing (UT): another subsurface option. Operator-dependent, and weaker than RT at characterizing very small round pores, but faster over large areas in skilled hands.

The Inspector’s Diagnostic Workflow

Guessing and diagnosing are separated by discipline. Work through this sequence in order, changing one thing at a time. It holds up on a carbon steel shop weld or a specialized job like aluminum boat welding, where one pinhole becomes a leak.

1. Section a sample. Cut and polish a cross-section of the suspect weld instead of troubleshooting an intact joint. It reveals the true location, root, mid-wall, or cap, that surface inspection can’t.
2. Map the location and pattern. Note where the pores sit, their shape, and how widespread they are. A single cluster near a tie-in reads differently than a line running along the fusion zone.
3. Check the WPS. Compare actual process, parameters, gas, filler, and joint prep against the Welding Procedure Specification. Deviations are often subtle: a little extra stickout, or a gas blend that isn’t what was qualified.
4. Isolate one variable. Start with the cheap checks. Verify flow at the nozzle, not the regulator. Inspect hoses for cracks, look for drafts, confirm filler storage and surface prep. Test after each change before moving on.

Diagnosis is about precision, not perfection. Read the pattern first, confirm it with the right inspection method, and work systematically. Do that and you stop chasing symptoms and start eliminating the real gas source.

Acceptance Criteria, Rework, and Repair: What the Codes Actually Allow

When porosity is discovered, the debate should never be, “Does this look bad to me?” The real question is, “Is this acceptable to the code for this exact joint?” Porosity is judged against a code or standard, not personal preference. What may be permitted on a simple shop support could be completely rejectable on a pipe line, a pressure vessel, or a cyclically loaded structure. In short, acceptance all comes down to loading, joint type, and the inspection method used to find it.

How Different Codes View Porosity

While every code has its own fine print, there are a few widely referenced limits that fabrication shops run into on a daily basis. According to AWS D1.1 (Structural Welding Code – Steel), visible piping porosity is not a free pass. For most welds where it is permitted, the sum of visible piping porosity diameters 1/32 in. or greater must not exceed 3/8 in. in any linear inch, and must not exceed 3/4 in. in any 12 in. length. With stricter rules applied to cyclically loaded structures, these allowances tighten even further. Most importantly, AWS D1.1 prohibits visible piping porosity in complete-joint-penetration (CJP) groove welds in butt joints transverse to tensile stress, with no room for it in those highly stressed areas.

Outside of structural steel, pressure piping holds even tighter tolerances. ASME B31.1 acceptance criteria commonly limit porosity dimensions to no more than 3/16 in. For pipeline work, API 1104 commonly limits porosity to no more than 1/8 in. in diameter. These numbers may seem small, but for high-pressure or high-cycle service, they are there for a reason.

Repair: Removing the Gas, Not Hiding It

Once a weld has been deemed rejectable, the goal is to remove the porosity to sound metal, not blend it away. Grinding the cap may make a pinholed bead look acceptable to the naked eye, but subsurface porosity will remain and can quickly become a fatigue initiation point. For this reason, AWS guidance states repairs of porosity typically consist of removal by gouging or grinding, weld repair, and re-inspection by the appropriate NDT method. Methods such as air arc welding can be useful for excavation on larger repairs, however, care must be taken to not gouge deeper than necessary.

The process is straightforward, but has no shortcuts. First, the defective area must be excavated until only sound, clean metal remains. Next, the joint must be re-prepared to fit the qualified procedure. From there, the repair must be re-welded to an agreed, qualified WPS using compatible consumables. For jobs utilizing alloy steel welding electrodes or other specialty fillers, extra attention must be paid to preheat and interpass requirements to avoid adding new issues to the repair.

Finally, re-inspection by the applicable NDT method must be performed before the weld is considered acceptable once more. If porosity is widespread throughout the weld rather than a small, isolated cluster, the entire weld should be removed, the joint re-prepared, and re-welded to the agreed procedure. Attempting to chase pinholes from one end of a bead to the other is almost always more costly in time than removing and redoing it correctly.

The Commercial Reality of Rework

Repair is a normal part of code work, however, repeated repairs come with consequences. Many codes place limits on repair frequency for the same weld location, and excessive gouging can change the cross-section of a joint enough to reduce toughness. Beyond code compliance, each repair cycle means lost production time, added NDT, and increased scrutiny from the client or inspector.

With that being said, a well-planned repair performed to the WPS will always be the correct route over forcing an over-tolerance weld to pass. The key is restraint: fix the root cause that created the porosity in the first place, or the repaired weld will fall victim to the exact same gas entrapment all over again.

The Prevention Playbook: A Practical Checklist That Stops Porosity

Porosity is rarely caused by carelessness. More often it comes from a process with no repeatable routine. Shops that almost never scrap a weld are not staffed with superheroes. They check the same things in the same order, before the arc starts rather than after the grinder exposes the damage.

Pre-Weld Checklist: Ten Minutes That Saves a Cut-Out

1. Read the WPS. Confirm the gas, flow rate, polarity, filler, amperage range and joint prep match the plate. If the WPS does not match reality, stop and clarify it.
2. Confirm base metal identity. A length of A36 mixed into hardenable plate behaves differently. Spark test, PMI or check the paperwork.
3. Degrease, then remove coatings and oxide. Oil, paint, mill scale, zinc and hydrated aluminium oxide are gas sources waiting for heat.
4. Check filler lot and storage. Sealed packaging, correct oven for low-hydrogen electrodes, no rusted wire, no damp flux.
5. Purge gas lines. Crack the valve, run a few seconds to clear air, and confirm the cylinder is the gas you want.
6. Verify flow at the nozzle with a flow tester, not the regulator. That is the only reading the pool cares about.
7. Inspect nozzle and contact tip. Clean, round, right size, correct recess, no spatter bridged across the diffuser.
8. Check the work clamp. Clean metal, close to the joint, secure. A wandering return path causes erratic arc behaviour.
9. Confirm wind protection. Screens up, doors closed, fans off. If you can feel air moving, it is too much.

Two items do the most work for gas-related porosity. Nozzle flow testing belongs in the equipment section, but a regulator gauge shows pressure, not delivery. Surface cleanliness is where aluminium welders lose their shirt, so read how to prep aluminum right before the next job.

During-Weld Checklist

• Hold stickout steady; long stickout loses gas coverage fast.
• Keep travel speed and work angle consistent; both change pool shape and how long gas has to escape.
• Listen to the arc. Hissing, popping or sputtering means gas turbulence or loss before pinholes appear.
• Do not whip out of the crater. Taper off and hold post-flow, or fill it deliberately.
• Clear spatter from the nozzle before it redirects the gas stream.

Post-Weld Checklist

Inspect the weld before cleaning it. Wire brushing hides fine surface-breaking pores. Mark suspect areas with a paint stick, then log gas, flow, settings, filler lot and ambient conditions. That log turns a bad day into a solvable problem, not a mystery.

Environmental Control

Screens and tents are close to mandatory on gas-shielded work. Humidity matters too: a damp shop on Monday morning gives different results than the same shop on Thursday afternoon. Where permitted, preheat and controlled cooling help, and the TWI guide to preventing porosity is a useful backstop.

Welding Aluminium Without Pinholes

Aluminium is unforgiving because its oxide layer holds moisture. Remove the oxide with a stainless brush kept only for aluminium; one that has touched carbon steel deposits the contamination you are trying to eliminate. Wipe with acetone, use dedicated tools and push the torch rather than dragging it. A filler with adequate deoxidiser, matched to how clean the base metal is, is a recognised porosity control. The AWS guidance on welding aluminium covers filler selection, and TWI’s aluminium porosity FAQ is a solid companion.

Welding Galvanized Steel

Zinc boils far below the arc temperature, and that vapour has to go somewhere. AWS guidance recommends removing zinc from the weld area or using techniques that volatilise it ahead of the pool. Point-of-welding fume extraction protects the welder, and the OSHA welding safety fact sheet is worth reading before you start. Grind back the coating, keep ventilation tight to the arc and slow down enough to let vapour clear.

Symptom-to-Fix Quick Reference

Symptom	Most likely cause	First fix
Fine scattered pores	Gas coverage or quality	Check flow at nozzle, then cylinder
Porosity at weld start	Gas not purged to nozzle	Add pre-flow
Porosity at end or crater	Post-flow too short	Extend post-flow, fill crater
Clustered pores in one area	Localised contamination	Grind and re-prep that section
Wormholes or piping	Coating, grease, gap contamination	Strip joint to bare metal
Porosity on aluminium only	Oxide layer or wrong filler	Dedicated brush, deoxidised filler
Porosity on galvanized only	Zinc vapour	Grind coating back first

One last point, and one most supervisors skip: consistency is a skill, not a personality trait. The same welder on the same machine can produce a clean bead on Tuesday and a porous mess on a humid Monday morning when the doors were open and the rods sat out all weekend. Build the checklist, use it every time, and you stop relying on luck. The Lincoln Electric MIG troubleshooting guide makes a good laminated companion for the wall.

Common Mistakes, Myths, and What Porosity Really Costs You

By now the causes are mapped. So why does the same pinhole problem come back six months later in the same shop? Almost always because a myth got in the way of the fix. The welder chased the wrong variable, “fixed” the symptom, and moved on. Here are the beliefs that keep porosity alive, the habits that feed it, and the honest price you pay for both.

Six Myths That Keep Porosity Coming Back

• “More gas is always better.” Past a certain flow, shielding stops behaving like a blanket and starts behaving like a jet, dragging room air into the stream. TWI’s porosity guidance lists excessive gas flow as a cause, not a cure. Turn the flow down before you turn it up.
• “Porosity is only cosmetic.” It isn’t. Every pore removes effective cross-section, and each one acts as a stress concentrator that eats into fatigue performance. Surface-breaking porosity also opens a path for moisture and corrosive media, which is why stainless and aluminium welds with open pores pit long before the surrounding metal does.
• “Grinding the cap fixes it.” Grinding hides porosity and thins the joint at the same time. You have traded an inspectable discontinuity for an uninspectable one with less metal holding the load.
• “Stainless doesn’t rust, so contamination doesn’t matter.” Stainless resists rust because of a passive oxide layer. Carbon steel dust, grinding sparks from a mild steel job, and chloride residue all break that layer down, and the resulting pits look nothing like the clean weld the drawing promised.
• “If it looks good, it passed.” Subsurface porosity is invisible by definition, and codes rely on NDT rather than appearance to judge internal discontinuities. A beautiful cap tells you almost nothing about what is underneath it.
• “The machine is faulty.” Occasionally true, usually not. Nine times out of ten the input was wrong: damp wire, a leaking hose, a contaminated part, or gas that never reached the pool.

The Mistakes That Turn One Bad Weld Into a Bad Week

The single most common diagnostic failure is changing three things at once. New gas, new wire, scrubbed the plate, and turned the voltage down, all before the next arc. It worked, but you learned nothing, and the fault will return the moment conditions drift back. Isolate one variable per test weld and the answer stays with you.

• Welding over porosity without excavation. The gas source is still in the joint. You will get the same defect again, usually bigger.
• Using the wrong filler for the base metal. A mismatch changes fluidity, gas solubility, and how the pool freezes, all at once.
• Reusing flux without drying it. Recycled flux carries moisture and fines straight back into the arc.
• Ignoring humidity. A muggy morning changes how much water sits on the plate, in the wire, and on the electrode coating.
• Skipping the WPS because “it’s just a small job.” Small jobs fail inspection the same way large ones do.

Aluminium is where these habits bite hardest. The oxide layer that protects the metal is also the layer that holds moisture, so a casual wipe-down isn’t prep. The cleaning steps in our aluminum cleaner for welding guide exist precisely because this metal punishes shortcuts.

What Porosity Really Costs You

The visible bill is the cut-out: grinding back to sound metal, re-prepping the joint, re-welding, re-inspecting. Then there’s the schedule slip while the next operation waits on a joint that should have been finished yesterday. The invisible bill is worse. Repeated rework drains the morale of a crew that starts second-guessing every bead, and it quietly teaches people to hide defects instead of reporting them.

On a certified job, the stakes climb. A failed visual or NDT result can trigger additional inspection, a concession request to the client, or a repair procedure that has to be qualified and documented. Suddenly a defect that measured a few thousandths of an inch is being discussed in a meeting with the customer’s engineer. That is the real price: not the electrode you burned, but the confidence you spent.

The Welder Who Can Name the Gas Source Wins

Almost every case of porosity traces back to process control: moisture, air, contamination, or a delivery path that quietly stopped shielding. The welder who can point at the bead and say “that’s hydrogen from the plate, not my gas flow” fixes it in twenty minutes. Everyone else spends the afternoon adjusting a regulator that was never the problem. Learn the patterns, isolate one variable, respect the prep, and the reasons for porosity in welding stop being a mystery and become a checklist you can run in your head before you strike an arc.

Frequently Asked Questions

What is the most common cause of porosity in welding?

Loss of adequate shielding gas is the single most common cause, usually from drafts, leaks, excessive flow turbulence, or too much stickout. Surface contamination—oil, moisture, rust, paint, or hydrated oxide—is a very close second and dominates in aluminium welding.

Does porosity always mean the weld must be rejected?

No. Most structural codes accept limited porosity based on size, distribution, joint type, and loading. AWS D1.1, ASME B31.1, and API 1104 each set different limits, and cyclically loaded or complete-joint-penetration welds face stricter rules. The applicable code and the inspector’s method decide.

Can porosity be fixed without cutting the weld out?

Local porosity can often be removed by grinding or gouging to sound metal, re-welded, and re-inspected. If the porosity is widespread or extends through the thickness, the entire weld should be removed and the joint re-prepared.

Why do I get porosity only at the start or end of a weld?

Start-of-weld porosity usually means the gas had not fully purged the torch and line yet, so pre-flow is too short or the nozzle was positioned before flow stabilized. End-of-weld porosity is typically crater pipe from extinguishing the arc too quickly; use a crater-fill or slope-down function and hold the gas post-flow.

Why is aluminium so prone to porosity?

Hydrogen solubility in molten aluminium is roughly 20 times higher than in solid aluminium, so dissolved hydrogen must escape during freezing or it forms pores. Aluminium’s oxide layer also holds moisture, and its high thermal conductivity speeds solidification, shortening the escape window.

Does more shielding gas prevent porosity?

Not necessarily. Excessive flow creates turbulence at the nozzle that can aspirate air into the shielding envelope, which increases porosity. Set flow within the WPS range and verify it at the nozzle, not just at the regulator.

Can I weld over porosity to fill the pinholes?

That is a bad practice. Welding over porosity usually traps the same gas source under new metal and often makes the defect larger. Proper repair requires removing the affected metal to sound material before re-welding.

How do inspectors find porosity that I cannot see?

Surface-breaking pores can be found with penetrant testing or magnetic particle testing. Buried porosity requires radiography or ultrasonic testing, with radiography generally better at characterizing pore size and distribution.