MIG Welding Problems Chart: Causes, Fixes & Prevention

Porosity: Why Your MIG Weld Looks Like Swiss Cheese

You drag the wire brush across a fresh bead and your stomach drops. Instead of a smooth silver crown, the surface is pocked with tiny craters, like someone went at the weld with a needle. That’s porosity, and it’s the most searched MIG defect for a reason: it shows up fast, it looks terrible, and it can mean the joint is genuinely weak.

Porosity is trapped gas. As the puddle solidifies, gas dissolved in the molten metal — or swept in from the arc atmosphere — doesn’t have time to escape before the metal freezes, so it leaves cavities behind. Some break the surface where you can see them. Others sit under a thin skin of sound metal and only appear when you grind, bend, or X-ray the joint. That second kind is the one that should worry you.

Three Sources of Gas, One Diagnosis

Every porosity problem traces back to one of three gas sources. Figure out which one you’re dealing with and the fix narrows dramatically.

• Lost shielding: Air reaches the puddle. Hobart Brothers states the most common cause of MIG porosity is poor shielding of the weld pool, letting atmospheric gases contaminate it. Nitrogen and oxygen from air create the classic scattered pinholes.
• Surface contamination: Oil, paint, rust, moisture, and galvanized coating boil and decompose in the arc, dumping hydrogen straight into the puddle. Lincoln Electric points to base metal contamination as the leading cause, with gas coverage problems second.
• Wire and base chemistry: Some gas comes from the metal itself. Damp wire, high-sulfur steel, or a filler with weak deoxidation can produce porosity even when your gas coverage is perfect.

Aluminum makes this worse because its surface oxide layer absorbs moisture eagerly, so prep matters more. Our guide to aluminum cleaner for welding covers the degreasing and oxide removal steps that prevent this specific failure.

The Systematic Gas Check, In Order of Likelihood

Don’t start by tearing apart your machine. Work down this list, because the boring answers are usually the right ones.

1. Cylinder empty or valve half-closed. Check bottle pressure and open the valve fully. A partially cracked valve can starve flow mid-bead.
2. Flow rate wrong. ESAB cites a typical shielding gas flow rate of roughly 25 to 40 CFH and advises checking nozzle position and coverage when porosity appears. Too much flow is just as bad — a turbulent stream aspirates surrounding air into the shield.
3. Wrong gas for the process. C25 (75% argon / 25% CO2) suits short-circuit mild steel. 100% CO2 runs hotter and spatterier. Argon-rich mixes with 1–5% oxygen or CO2 are for spray transfer and stainless. Grab the wrong bottle and you’ll chase ghosts.
4. Drafts. An open shop door, box fan, or ceiling fan blows the shield away, producing porosity that appears and disappears depending on where you stand.
5. Clogged nozzle. Spatter bridging the gas ports is a top offender. The gas has nowhere to go but sideways.
6. Damaged O-ring or split hose. A cracked O-ring at the gun connection leaks quietly and consistently.
7. Excessive stickout. Push the gun too far from the work and the shield dissipates before it reaches the puddle. Keep stickout around 3/8 to 1/2 inch on solid wire.
8. Dirty base metal or wrong filler. Last on the list, but if the first seven check out, this is your culprit.

One more item worth knowing: Lincoln Electric notes that an oversized contact tip causes arc inconsistency, porosity, and poor bead shape. Match the tip to your wire diameter exactly.

For a deeper look at matching gas to process, our aluminum MIG welding guide walks through argon and helium blends in detail — the same selection logic applies to steel, just with different mixes.

The Filler Metal Chemistry Angle

Here’s the part most charts skip. Your wire is a deoxidizer, and some wires deoxidize better than others. Lincoln Electric notes that ER70S-3 contains about 0.45–0.75% silicon and 0.9–1.4% manganese, while ER70S-6 carries roughly 0.8–1.15% silicon and 1.4–1.8% manganese.

Those extra points of silicon and manganese are scavengers. They combine with oxygen in the puddle and float the resulting oxides out as slag, leaving cleaner metal behind. On dirty, rusty, mill-scale-heavy steel — or when you’re on the edge of your gas coverage — ER70S-6 buys tolerance that ER70S-3 doesn’t. On clean, well-shielded plate the difference is subtle. On real-world material, it’s often the fix.

Porosity in Specific Locations

Where the holes appear tells you what’s wrong.

The start-of-bead pattern catches people constantly. Gas surges when the solenoid opens, briefly over-pressurizing the line before it settles, so the first half-inch gets a chaotic shield. Half a second of pre-flow eliminates it. The end-of-bead pattern is simply losing coverage as you lift the gun; hold your position for a beat and let the crater fill.

One last check: a loose, painted, or rusty ground clamp produces an erratic arc that mimics gas trouble. Clean metal, clamped close to the work, every time. Welders who blame their technique for weeks often find the answer under the clamp.

And if you’re on aluminum and suspect the process itself is the problem, aluminum flux welding explains when a different route makes more sense than fighting gas coverage on thin sheet. The diagnostic sequence stays the same either way: verify gas, verify ground, verify the wire path, then adjust technique.

Excessive Spatter and an Unstable Arc: Fixing the Mess Before It Cools

Spatter is molten metal that leaves the transfer stream and never joins the puddle. The droplets cool in flight, land on your plate, your table, and your forearms, and stay there. Every MIG arc throws a few, so a light dusting is normal. The real signal is in the volume and the shape. A fine, even sprinkle that brushes off with a gloved hand means the process is roughly healthy. A halo of bb-sized balls, a bead with torn edges, and a nozzle caked in metal after four inches of weld means the arc is unstable and something upstream is feeding it bad information.

JASIC’s troubleshooting guide makes the key point well: excessive spatter is a symptom of an unstable arc or unstable metal transfer, usually from mismatched voltage and current, wire feed trouble, or a dirty surface. That reframes the problem. You are not fighting spatter. You are chasing the instability that produces it.

The Four Families of Spatter Causes

Electrical. Voltage that doesn’t match wire feed speed is the number one offender. Too low and the wire stubs into the plate, the arc shortens, and droplets blow out sideways. Too high and the arc wanders and metal sprays. Settings trapped in the globular transition zone produce the ugliest spatter, because molten balls bigger than the arc can control drop through it. Wrong inductance or arc-control settings on short circuit make it worse by changing how abruptly the current rises.

Gas-related. Too little flow leaves the puddle exposed. But ESAB notes that too much flow contributes to spatter too, because a high-velocity stream turns turbulent and drags air into the shielding envelope. An open shop door or a fan pointed at your bench does the same thing from a different direction.

Material-related. Mill scale, rust, oil, paint, and galvanized coating interrupt arc stability as they burn or vaporize. MillerWelds also flags contaminated or rusty wire itself, including residual drawing lubricant on a fresh spool, which is why wire storage matters as much as plate prep.

Mechanical. Excessive stickout is on Miller’s list for good reason: every extra fraction of an inch of wire outside the tip raises resistance, drops voltage at the arc, and destabilizes transfer. Add the wrong contact tip size or recess, a worn or egged-out tip, and a loose ground clamp, and you get an arc that changes character mid-bead.

Spatter Triage in Five Moves

1. Clean a test plate to bare metal. No mill scale, no oil, no coating. That removes half the variables in one step.
2. Set a known-good baseline for that thickness from a published parameter chart, not from whatever the last job left on the dials.
3. Adjust voltage in small increments, a half-volt at a time on a 240V machine, and weld a short bead at each setting.
4. Then adjust wire feed speed. Change both at once and you learn nothing.
5. Finally check stickout and travel angle. Hold roughly 3/8 inch of wire out, keep the tip perpendicular to the joint, and watch what the arc does at the end of the plate.

Push Versus Drag, and What Each Does to Spatter

Pushing, tilting the gun so it leads the puddle by about 10 to 15 degrees, throws spatter backward away from the puddle and gives you a clearer view of the arc. On thin sheet and aluminum, push is usually the better habit; the complete guide to aluminum MIG welding covers why heat control makes that angle almost mandatory there. Dragging, with the gun tilted back, concentrates heat, gives deeper penetration on thicker steel, and pushes spatter ahead of the arc where it can’t contaminate solidifying metal. Neither angle eliminates spatter on its own. Both keep the arc stable and consistent, which is the actual goal.

Anti-Spatter: Useful, With a Catch

Aerosol anti-spatter works and it’s fast. Water-based versions cost less per weld and don’t carry the solvent load into your shop air. Both leave a film, though. Overspray that drifts into the puddle can cause porosity, and residue inside a joint will wreck paint adhesion later, a real problem on anything headed to a powder coater. Spray the nozzle and surrounding plate, not the joint itself, and wipe the weld zone before the next pass.

When the Wire Is Flux-Cored

Flux-cored and metal-cored wires follow the same logic with one twist. MillerWelds advises that low voltage can produce excessive spatter with both types, and that increasing voltage as needed brings it back under control, the opposite direction from what many solid-wire welders assume. For self-shielded flux-cored wire, the same source recommends straight polarity (electrode negative) and a drag technique. Get the polarity wrong and no amount of knob-twisting will clean it up.

One last note on prep: if you’re welding aluminum and the spatter looks like popcorn, the surface is probably fighting you. Prepping aluminum properly, degreasing first and then removing the oxide layer, removes more arc instability than any setting change.

Undercut, Burn-Through, Lack of Fusion, and Other Structural Defects

Everything up to this point has mostly been annoying or ugly. This section is different. The defects below decide whether a part holds, fatigues, or fails, and several of them are invisible from the outside. Each one still leaves a signature: a look, a parameter cause, a technique cause, and a specific fix. Work backwards from that.

Michigan Tech’s welding discontinuities training material groups spatter, porosity, cold roll, undercut, lack of fusion, arc strikes, inclusions, and cracking as the core discontinuity categories, which is a useful reminder that “bad weld” is really a family of separate problems with separate answers.

Undercut: The Groove You Never Refilled

Undercut is a groove melted into the base metal at the weld toe that the puddle never filled back in. It shows up as a dark trench or shadow line running parallel to the bead, most often on the upper toe of a vertical or horizontal fillet.

• Parameter cause: too much voltage (long arc wandering and melting sidewall without depositing), or excessive travel speed that outruns the puddle.
• Technique cause: wrong gun angle, a wire too large for thin material, weaving too wide and too fast, and no pause at the toes where the metal actually needs to pile up.
• Fix: drop voltage in small increments, slow travel, hold a beat at each toe, and keep the gun at roughly 10–15 degrees of push or drag depending on position.

Burn-Through: When the Root Disappears

Burn-through is complete melt-through of the root — a hole, a hanging icicle of metal, or a sagging blob on the back side. It’s a heat-input problem before it’s anything else.

• Parameter cause: too much voltage or wire feed speed for the gauge, or travel so slow the arc sits in one place.
• Technique cause: an open root gap with no backing, poor fit-up, and no chill bar on thin sheet.
• Fix: reduce WFS first (it controls current), speed up travel, close the gap, and back the joint with copper, aluminum, or a ceramic backing tape. On 22-gauge sheet, drop to 0.023-inch wire before touching anything else.

Lack of Fusion and Incomplete Penetration

Lack of fusion is a failure to bond to the sidewall, the root, or a previous pass. Incomplete penetration means the arc never reached the root at all. Both look fine from above: a smooth, well-shaped bead hiding a cold joint underneath. ESAB describes lack of fusion as a planar defect that acts as a stress concentrator and often escapes visual inspection, requiring ultrasonic testing to find on critical work.

That planar geometry is why lack of fusion is more dangerous than porosity. Porosity is rounded and volumetric; a crack-like plane concentrates stress at its tip and can propagate under cyclic load. Causes stack up fast: too cold a setting, travel too fast for the puddle to wet the sidewall, wrong electrode angle, mill scale or oil in the joint, short-circuit transfer with insufficient heat input, and weld prep that leaves a gap the arc simply can’t reach. On aluminum, the oxide layer raises the melting point and makes cold lap even easier — the prep advice in this aluminum cleaner for welding guide applies directly.

Overlap, Cracking, and Arc Strikes

Overlap (cold roll) is filler metal that rolls over the base metal without fusing to it — a rounded lip at the toe with a visible line underneath. It’s a cold, fast-travel defect, usually combined with a drag angle that pushes metal ahead of the arc.

Cracking comes in crater, longitudinal, and hydrogen-assisted varieties. Crater cracks form at termination when the puddle shrinks and no fill was added; hydrogen-assisted cracking shows up hours later, driven by moisture, oil, or rusty wire feeding hydrogen into the arc. Arc strikes are their own category: a momentary strike outside the weld zone creates a quench-hardened spot that can initiate cracking later.

What “Acceptable” Actually Means: AWS D1.1 Visual Limits

AWS D1.1/D1.1M:2025 is the 25th edition, released in April 2025, and covers structural steel at least 1/8 inch thick with yield strength up to 100 ksi. Its visual criteria are the benchmark most North American structural work is judged against.

Two cautions. First, these limits shift with joint type, thickness, and stress category, so the table above is orientation, not a ruling. Second, your project’s WPS and the inspector’s reading of the applicable code govern — a chart cannot pass or fail a weld.

For a broader walkthrough of how these imperfections are classified, TWI’s review of geometric shape imperfections and Lincoln Electric’s MIG problems and remedies guide are worth keeping on the bench. When these defects show up on aluminum rather than steel, the same logic applies but heat input and cleaning matter even more — see the complete aluminum MIG welding guide.

Wire Feed Problems: Birdnesting, Burnback, Stubbing, and Chattering Guns

Most MIG troubleshooting happens at the arc, but a surprising share of bad welds starts three feet behind it, inside the drive rolls and liner. The feed system is the part of the machine nobody services until it fails, and yet it controls whether the arc ever sees a steady stream of wire. When the feed is inconsistent, every other setting on the front panel becomes a guess. That’s why Lincoln Electric ties faulty wire delivery — especially an oversized contact tip — directly to arc inconsistency, porosity, and poor bead shape.

Birdnesting: When the Wire Tangles at the Rolls

A birdnest is exactly what it sounds like: a coiled snarl of wire erupting between the drive rolls and the gun. The wire stopped moving forward, but the rolls kept pushing, so it buckled sideways.

• Drive roll tension too low — the rolls slip, the wire stalls, then grabs and folds.
• Wrong groove size — a .035 groove trying to feed .030 wire (or vice versa) never grips properly. Flux-cored wire also needs a knurled or specific flux-cored groove, not a smooth V-groove.
• Kinked, clogged, or wrong-length liner — a liner trimmed too long kinks at the gun; one full of grinding dust and wire shavings drags.
• Spool binding — a hub nut cranked down tight makes the roll fight the feeder.
• Bent wire — a spool that was dropped or a wire end left unclipped will snag on every wrap.

Aluminum makes this worse because the wire is soft and has almost no column strength. That’s why aluminum MIG welding usually calls for a spool gun or push-pull setup rather than a standard 10-foot torch.

Burnback: Wire Fused to the Contact Tip

Burnback is the tip welding itself to the wire. The classic cause is wire feed speed set too low for the voltage — the arc burns the wire back faster than the feeder replaces it, and the arc climbs into the tip. Other culprits: a contact tip recessed too far inside the nozzle, a tip that’s oversized for the wire (loose contact means arcing inside the bore), a poor ground that destabilizes the arc, or too little stickout so the tip sits in the heat.

Tip geometry matters more than most welders realize. MillerWelds notes that the wrong size or a worn contact tip, plus incorrect contact-tip-to-nozzle recess, can cause excessive spatter on its own.

Stubbing and Chattering: The Feed Path Fighting Back

Stubbing feels like the wire is poking the plate instead of arcing into it. Chattering is worse — you hear it before you see it. A chattering sound in the gun cable is a classic sign of a wire delivery problem in the liner or drive system: a worn or dirty liner, tip drag, a spool brake wound too tight, or drive roll pressure so high the wire deforms and drags through the liner.

The Five-Minute Feed Setup Checklist

1. Confirm the drive roll groove matches your wire diameter and type.
2. Set tension with the “pinch test” — drive roll tension should be just enough to feed wire without flattening or deforming it.
3. Trim the liner to the correct length with the gun cable held straight, then blow it out or replace it if it’s contaminated.
4. Adjust spool hub friction so the roll coasts to a stop rather than spinning free or dragging.
5. Fit a clean contact tip and put tip changes on a schedule instead of waiting for a failure.

Contact Tip Life: Worn Versus Egged Out

A worn tip has an enlarged but roughly round bore from normal arc erosion. An egged out tip has an oval, gouged hole, usually from burnback, a bent wire, or running the wrong diameter. Egged-out tips throw the arc off-center and produce wandering, spattery beads. Tip life depends on amperage, duty cycle, wire type, and whether you’re running flux-cored (shorter life) or aluminum (longer but fussier). Keep spares of the right size and check the bore against a fresh tip now and then.

Poor feed also shows up as a rough, inconsistent bead that gets blamed on technique. If the wire isn’t arriving at a constant rate, the puddle can’t be consistent — and no amount of gun-angle fiddling will fix it.

Gas, Polarity, and Consumables: The Setup Mistakes That Mimic Technique Problems

When a bead starts acting up, most welders blame their hands. They change the travel angle, slow down, speed up, and get the same ugly result. The real problem is often upstream: wrong gas, reversed polarity, a ground through paint, or a contact tip that should have been replaced two spools ago. These errors produce defects that look exactly like bad technique, which is why people chase the wrong variable all evening.

Shielding Gas: Read the Label, Then Read the Flow

Start with cylinder identification. Mixed-gas cylinders are labeled with the exact composition, not just “MIG mix.” C25 is 75% argon and 25% CO2; C20 is 80/20. Grab the wrong cylinder and arc behavior changes immediately. Your supplier’s handling guide is worth reading once, and a reputable supplier such as AllGas Welding Supply explains cylinder labeling and safe handling clearly.

A regulator and a flowmeter are not the same thing. The regulator drops cylinder pressure to a usable level; the flowmeter sets and indicates actual flow in cubic feet per hour (CFH). Ball-style flowmeters are most accurate mid-scale, which is why 15–30 CFH reads better than pinning the ball at either end. ESAB cites roughly 25–40 CFH as a typical MIG shielding gas flow range. In metric, 1 CFH is about 0.47 liters per minute, so 25–40 CFH is roughly 12–19 L/min.

More gas is not better. Excessive flow creates turbulence at the nozzle that can aspirate shop air into the shielding stream, producing porosity even with a full cylinder. Check the hose connections, O-ring, valve, and a spatter-packed nozzle before touching the flow setting.

Which Gas for Which Job

Gas mix	Typical use	What to expect
C25 (75% argon / 25% CO2)	Short-circuit MIG on mild steel	Stable arc, good bead appearance, moderate penetration
100% CO2	Short-circuit MIG on mild steel	Hotter arc, deeper penetration, more spatter, lower cost
Argon-rich mix with 1–5% O2 or CO2	Spray transfer on mild steel; some stainless	Smooth spray arc, low spatter, high heat input
98% argon / 2% CO2 or stainless tri-mix	Stainless steel MIG	Stable arc and corrosion resistance when matched to filler
100% argon	Aluminum MIG	Required for aluminum; poor for steel short-circuit transfer

100% CO2 behaves differently because it is an active gas that dissociates in the arc and reacts with the pool. The arc is harsher and spatter is heavier than with argon mixes, but penetration runs deeper and the gas costs less. It is a poor choice for spray transfer. Match the gas to the transfer mode, not just to the base metal.

Polarity: The Cable You Forgot to Check

Solid-wire GMAW normally runs electrode positive, or DCEP. Self-shielded flux-cored wire is the exception: it typically runs electrode negative, or DCEN, because the wire’s flux core supplies the shielding and the arc behaves differently. Gas-shielded flux-cored wire generally runs DCEP. Reversed leads give a weak, unstable arc, poor penetration, heavy spatter, and sometimes wire melting back into the contact tip. Confirm the leads match the process printed on the spool before you adjust anything.

Grounding: The Quiet Culprit

A bad ground creates an arc that surges and stumbles like a gas problem. Clamp directly to bare, clean metal as close to the weld as practical. Paint, rust, mill scale, and anodizing all add resistance. A loose clamp or a connection made through a painted trailer frame causes random spatter bursts and inconsistent penetration.

Consumables: Small Parts, Big Symptoms

The contact tip must match the wire diameter and sit at the correct recess for the nozzle. A worn tip loses electrical contact and causes stubbing, spatter, and birdnesting. An “egged out” tip is worse: the bore is oval, the wire wanders, and the arc never points where you think it does. Keep spare tips, nozzles, O-rings, and a diffuser on hand. A clogged nozzle bridges the gas ports and kills coverage; a hardened O-ring leaks quietly.

Liner choice matters too. Steel liners suit steel wire, nylon works for many softer wires, and Teflon is the usual choice for aluminum because it reduces drag. Setting up aluminum means the liner, drive rolls, and gun length all need to work together; our Aluminum MIG Welding guide walks through that setup. Store wire dry. MillerWelds lists contaminated or rusty wire as a cause of both porosity and excessive spatter, and no machine setting compensates for a corroded surface.

Stainless and galvanized jobs add a fume concern. OSHA notes that chromium in stainless steel converts to hexavalent chromium, Cr(VI), during welding, and regulates it under 29 CFR 1910.1026 and 1926.1126. The permissible exposure limit is 5 micrograms per cubic meter as an 8-hour time-weighted average. Source capture and ventilation belong in the plan before the arc starts.

Note: A $12 flowmeter and a $3 O-ring fix more welds than a $1,200 machine upgrade. Check the cheap setup layer first.

None of this replaces technique, but it clears the noise. Once gas, polarity, ground, and consumables are verified, every adjustment you make afterward actually tells you something.

Bead Shape and Distortion: Reading Convex, Concave, Ropey, and Banana Welds

A weld bead is a record of every decision your hand and machine made. Crown height, toe angle, ripple pattern, and bead placement all leave evidence. Stop judging the bead as “pretty” or “ugly” and start asking what the shape says about heat, speed, angle, and arc stability.

What Each Profile Is Telling You

A high, convex crown usually means the puddle froze before it could flatten. Common causes are voltage too low for the wire feed speed, travel too slow, or a gun angle that crowds metal toward the center. The bead looks cold and often sits on top of the plate instead of biting in. A concave or flat bead is the opposite: voltage high, travel fast, or both. It looks smooth until you notice washed-out toes and undercut starting. ESAB notes that travel speed and voltage changes affect bead profile and can create undercut or a convex bead (ESAB).

Ropey, stacked-coin beads signal inconsistency: wire feed speed and voltage fighting each other, travel surging, or a gun angle that keeps changing. A bead that wanders side to side is usually a control problem first, involving body position, work angle, or wire cast. Magnetic arc blow comes second. Deep craters mean you stopped the arc without filling it, which is a crack waiting to happen on a restrained joint.

Bead appearance	Likely cause	First fix
High, convex crown	Voltage low for WFS, travel too slow, crowding angle	Raise voltage slightly, speed up, check gun angle
Flat or concave with washed toes	Voltage high, travel too fast	Lower voltage, slow travel, inspect for undercut
Ropey or stacked-coin	WFS/voltage mismatch, inconsistent travel	Reset baseline settings, steady the travel
Wandering or wiggling	Work angle, body position, wire cast, arc blow	Reposition, shorten stickout, check ground
Deep end crater	No crater fill or abrupt termination	Pause, add wire, use crater-fill if available
Part hinges or bows	Excess heat input, poor sequence	Reduce heat, backstep, clamp or fixture

Travel Angle, Work Angle, and the Position Trap

Push or drag? For solid-wire MIG on mild steel, a slight drag of about 10 to 15 degrees generally improves penetration and flattens the profile. A push angle gives better visibility and a wider bead but can reduce penetration. Work angle matters just as much. On a fillet, 45 degrees splits heat evenly between plate and web. Tilt too far toward one leg and that leg gets undercut while the other gets cold lap.

Here is the trap. Run a 1/8-inch mild steel T-joint flat with settings that produce a beautiful bead. Now weld the same joint vertical-up with those exact settings. Gravity pulls the puddle down, the bead turns convex and cold-looking, and the toes may lack fusion. The machine did not change. The position did. Vertical-up needs less wire feed speed, a tighter arc, and a small pause at each toe.

Wire cast can also steer the bead. Wire with a strong natural curve makes the arc wander, and soft aluminum wire makes it worse. Our aluminum MIG welding guide covers feed setup and gun handling for that case.

Distortion Is a Bead-Shape Problem Too

Distortion begins with heat input, which is a function of voltage, amperage, and travel speed. That makes it the primary lever you control. Angular distortion hinges the plates upward along the weld. Bowing, sometimes called a banana weld or banana part, curves the whole assembly along its length. Clamping holds shape but does not remove residual stress. Balanced sequence, backstep welding, stitch welding, and cooling between passes work better than brute force.

TWI points out that linear misalignment in single-sided welds can cause lack of root or sidewall fusion and leave a sharp continuous imperfection along the higher weld toe (TWI). Fit-up distortion is therefore not cosmetic. It concentrates stress exactly where the weld should be strongest. Undercut deserves the same respect: TWI calls it probably the most common shape imperfection in arc and gas welding, and it can appear at the toes of every pass in multi-run work.

Short Note: A bead can look perfect and still fail inspection, just as an ugly bead on a non-critical cosmetic joint can be structurally sound. Shape is a clue, not a verdict. Acceptance depends on the code, joint, loading, and inspector.

The Complete MIG Troubleshooting Workflow: A 10-Step Diagnostic Sequence

Random knob-twisting is how a ten-minute repair turns into a two-hour one. Lincoln Electric, MillerWelds, Hobart Brothers, and ESAB all build their troubleshooting guides around the same principle: change one variable at a time and observe the result. MillerWelds and Hobart both stress that many common MIG complaints trace back to setup, technique, or parameters rather than a machine that needs replacing. So here is the sequence to run instead of guessing.

The Ten Steps, In Order

1. Stop and look at the whole bead. One porous spot means something different from porosity running the full length. Read the pattern before you touch anything.
2. Check the ground. Clamp on bare, clean metal close to the joint. A loose or painted connection produces an arc that behaves exactly like a gas fault.
3. Confirm gas. Cylinder level, valve fully open, flow reading in the working range, nozzle free of spatter bridging the ports, and no fan or open door pulling the shield away.
4. Confirm polarity and process. Solid wire runs DCEP; self-shielded flux-cored runs DCEN. Wrong polarity mimics a dozen other faults.
5. Inspect the feed path. Spool binding, drive roll tension and groove size, liner condition, contact tip wear. Soft wires like aluminum punish a lazy liner faster than steel does, which is why feed faults show up so often in aluminum MIG welding.
6. Measure stickout and check gun angle. Roughly 3/8 to 1/2 inch of stickout, and a consistent push or drag angle.
7. Clean a test coupon to bare metal. Same thickness and joint type as the real part. Mill scale, oil, and galvanizing will lie to you.
8. Establish a baseline. Set voltage and wire feed speed from a known parameter chart for that thickness, then judge the bead against it.
9. Change one variable at a time, in small increments. Weld a 3- to 4-inch test bead after each change. That bead is the fastest validation you’ll ever get, and it costs pennies.
10. Record what worked. Write it on the machine with a paint marker or in a notebook. Next month you’ll thank yourself.

Quick Note: A $10 notebook beats a $500 gadget for building diagnostic intuition. Log the symptom, the change, and the result, and within a month you’ll recognize patterns faster than any app can suggest them.

Worked Scenario: Porosity on Galvanized Square Tube

A hobbyist building a gate gets uniform porosity across the whole bead. Step 2 clears the ground. Step 3 shows 20 CFH and a clean nozzle. Step 4 confirms DCEP with solid wire. The reasoning chain points at contamination, not equipment: galvanizing in the weld zone is vaporizing into the puddle. Grinding the coating back an inch from the joint and improving ventilation resolves it. The chart said “porosity,” but the sequence found the source.

Worked Scenario: Spatter on 1/4-Inch Plate

A shop welder fights heavy spatter on quarter-inch plate despite textbook settings. Voltage and wire feed speed are correct on paper, so the electrical explanation fails. Step 2 is the turning point: the ground clamp is biting onto rust and paint, so current hunts for a path and the arc wanders. Cleaning the contact area and moving the clamp tight to the work settles the arc immediately. MillerWelds notes this exact trap in its weld defect troubleshooting guide.

Worked Scenario: Lack of Fusion on a Vertical-Up Fillet

A fabricator sees a crowned, cold-looking fillet with fusion missing at the sidewall. The reasoning chain runs: the bead sits proud, travel looks fast, and the machine was set for flat position. Vertical-up needs less wire feed speed and a slight weave to let the puddle wet into both plates. Slowing travel and adding a touch more voltage fixes the wet-in. Because lack of fusion is planar and subsurface, visual inspection alone cannot clear it — as MTU’s welding discontinuities training material explains, methods like ultrasonic testing are required. ESAB’s defect resolution guide makes the same point about process control.

Printable One-Page Summary

Symptom	Probable Cause	Quick Check	Corrective Action
Porosity	Lost shielding or contamination	Flow reading, nozzle, base metal	Restore gas coverage, grind clean
Heavy spatter	Bad ground or voltage/WFS mismatch	Clamp contact, arc sound	Clean ground, tune voltage
Birdnesting	Drive roll tension or liner	Feed wire by hand, inspect liner	Reset tension, replace liner
Burnback	WFS too low for voltage	Tip condition, stickout	Raise WFS, check tip recess
Undercut	Excess voltage or travel speed	Bead profile at the toes	Lower voltage, slow down
Lack of fusion	Too cold, travel too fast	Bead wet-in, position settings	Increase heat input, adjust angle

Run this sequence every time a bead misbehaves and the chart stops being a wall decoration. The steps don’t change; only the answer does.

Prevention, Safety, and Knowing When a Weld Is Actually a Defect

Work through ten sections of symptoms and a pattern appears. Almost none of those ugly beads came from a mysterious machine fault. They came from what happened before the arc started: a dirty joint, a tired contact tip, a ground clamp biting into paint. Prevention is the unglamorous half of troubleshooting, and the half that saves the 9 p.m. repair.

Prevention: Cheap Habits, Expensive Problems Avoided

Build these into your routine and the chart gets shorter every month:

• Clean the base metal first. Degrease, grind mill scale back to shiny steel, strip galvanizing from the weld zone. OSHA advises removing coatings like solvent residue and paint because heating them creates a toxic exposure, not just porosity. Aluminum needs its own prep, so read up on prepping aluminum before welding.
• Store wire dry. Open a spool, use it, bag it. Rusty wire feeds badly and welds worse.
• Change tips and liners on a schedule, not when they fail. Keep a spare nozzle, diffuser, and a few O-rings in the drawer.
• Log settings per job on a card taped to the machine, and keep the ground clamp healthy: clean jaws, tight spring, bare metal contact close to the work.
• Sanity-check the flowmeter against a second gauge once a season.

Still hunting for parameters? A solid reference like this practical MIG welding guide saves guesswork, and a good welding supplies buying guide keeps the right consumables on hand.

Safety: The Part That Follows You Home

Arc radiation is the obvious hazard and the easiest to dismiss. Never weld without a properly rated helmet, and cover exposed skin: a t-shirt and a short bead is how forearms get UV burn that feels like sunburn and behaves like it. Lens rating must match the process and the amperage.

Fumes are the slower hazard. Stainless and galvanized steel deserve the most respect: stainless welding can generate hexavalent chromium, and OSHA’s Cr(VI) permissible exposure limit is 5 µg/m³ as an 8-hour time-weighted average under 29 CFR 1910.1026 and 1926.1126. Manganese is the other one. CDC/NIOSH lists a manganese REL of 1 mg/m³ TWA and 3 mg/m³ STEL, an OSHA PEL of 5 mg/m³ as a ceiling, and an ACGIH TLV of 0.02 mg/m³ respirable. That gap between the numbers tells you how seriously to take ventilation.

Source-capture extraction at the arc is the first line of defense. General shop fans are not, since they move contaminated air around rather than out. Where ventilation can’t control exposure alone, respiratory protection belongs under a written program with fit testing and medical clearance. Scale matters: according to AIHA, more than 200,000 US workers and roughly 1.8 million worldwide weld as all or part of their job.

Quick Note: If you can smell the puddle through your helmet, the plume is reaching you. Reposition the extraction hood; don’t weld faster.

Discontinuity vs. Defect: The Judgment Call a Chart Can’t Make

A discontinuity is any interruption in the weld’s uniformity: a pore, a ripple, slight undercut, a bit of spatter. A defect is a discontinuity that exceeds the applicable acceptance criteria. Same feature, two very different verdicts, and the code decides, not your eye.

AWS D1.1 acceptance criteria depend on joint type, thickness, and whether the member is statically or cyclically loaded. A 1/32-inch undercut on a garage stiffener and the same undercut on a cyclically loaded bridge member do not get the same answer. This MTU overview of welding discontinuities and defects is a useful primer on the logic. On structural or pressure work, the answer lives in a qualified WPS executed by qualified welders and verified by an inspector. No chart, including this one, tells you whether a weld passes.

Note: Grinding a discontinuity smooth so it “looks better” doesn’t change what’s underneath. On load-bearing work, that’s cosmetically fixing an inspection result, not a weld.

The Five Highest-Leverage Fixes

1. Clean the metal and confirm the ground. Half the mystery disappears here.
2. Verify gas — level, valve, flow, nozzle condition — before touching a knob.
3. Match wire feed speed to voltage on scrap, not on the part.
4. Fix stickout and gun angle before chasing settings.
5. Change one variable at a time, and write down what worked.

Back to the Trailer Fender

Nine p.m., cold shop, cracked fender, and a chart that seemed to offer a single answer. It never did. What fixed that weld was a sequence: gas confirmed, ground cleaned, tip swapped, one knob moved a quarter turn, a three-inch test bead. The chart pointed at possibilities. The habit of changing one thing at a time turned a possibility into a weld you’d sign your name to.

Frequently Asked Questions

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

Poor shielding gas coverage is the leading cause, according to Hobart Brothers and Lincoln Electric. That usually means low or turbulent flow, a clogged nozzle, a leak in the gas path, too much stickout, or welding in a draft. Contaminated base metal and rusty wire are the next most common culprits.

Why does my MIG weld have spatter no matter what settings I use?

If parameter changes don’t help, stop adjusting the machine and check the physical setup. A dirty or loose ground clamp, excessive stickout, a worn or oversized contact tip, mill scale or oil on the plate, and gas flow that’s too high can all cause spatter that no voltage tweak will fix.

What causes wire to burn back into the contact tip?

Burnback usually means the wire feed speed is too low for the voltage, the contact tip is too recessed, or the tip is the wrong size or badly worn. A poor work lead connection can also cause it by making the arc unstable.

Should I use 100% CO2 or a C25 mix for MIG welding mild steel?

C25 (75% argon, 25% CO2) gives smoother arcs, less spatter, and better bead appearance in short-circuit transfer, which is why it’s the default for most hobby and general fabrication work. Straight CO2 is cheaper and produces deeper penetration but a harsher, spatterier arc and no spray transfer.

How much shielding gas flow should I run on MIG?

ESAB cites roughly 25 to 40 CFH as a typical working range for MIG, with many indoor short-circuit jobs running well around 20 to 30 CFH. Higher flow is not better: excessive flow causes turbulence that can aspirate air into the puddle and create porosity.

Why is my MIG wire birdnesting at the drive rolls?

Birdnesting is almost always a feed-path problem: too much or too little drive roll tension, the wrong groove size for the wire diameter, a kinked or clogged liner, a spool that’s binding, or wire that got bent. Fix the cause before you cut the wire out, or it will happen again on the next bead.

Can I fix undercut by running a second pass over it?

A cosmetic pass can hide undercut without restoring the lost base metal thickness, and it may add heat and distortion. The correct approach is to correct the cause (usually voltage, travel speed, or angle), then repair per the applicable procedure, which often means grinding and rewelding rather than covering it.

Is a weld with porosity automatically a defect?

Not automatically. A discontinuity only becomes a defect when it exceeds the acceptance criteria in the applicable code or specification. AWS D1.1, for example, sets limits by joint type, thickness, and loading, and rejects all cracks regardless of size.

Why does my MIG weld look great flat but terrible vertical-up?

Position changes how gravity acts on the puddle, so the same voltage and wire feed speed that work flat will often run too hot or too cold vertical-up. Vertical-up generally needs lower wire feed speed, slightly longer arc, and a weave or step technique to control the puddle.

Do I need a fume extractor for MIG welding in a garage?

Yes for any sustained welding, and especially for stainless, galvanized, or painted metal. OSHA sets a hexavalent chromium PEL of 5 µg/m3 as an 8-hour TWA, and NIOSH recommends keeping manganese exposure at or below 1 mg/m3 TWA. Source-capture ventilation at the arc plus open doors and a respirator where needed is the practical minimum.

TOPICS: MIG Welding Procedure Inspection

NH

Nicholas Harper

Editorial Team

Covering welding processes, equipment selection, and shop-floor practice for WeldZen — with every claim checked against the governing code.

Share this technical standard

COPY LINK

Related Studies

Field Study

Hot Plate Plastic Welding: How It Works and How to Do It

Oct 25, 2026 · 55 min

Buying Guide

Argon Welding Tank Sizes: Complete Guide to Choosing the Right Cylinder

Oct 25, 2026 · 44 min

Safety

How to Connect PVC Pipe to Copper Pipe Without Welding

Oct 25, 2026 · 51 min

Technical Dispatch

Get The Next Benchmark First

New field studies and calibration sheets, once a week. No spam.

Email address Subscribe