Guide

Welding Defects: Causes and Fixes

Most weld defects trace back to four things — wrong heat, a contaminated joint, lost shielding, or bad technique — and each defect points at a specific one of them.

A weld defect is a symptom, and each one points at a fairly short list of causes. Learning to read them backwards — from what the bead looks like to what the machine or the joint was doing wrong — is faster than adjusting settings at random.

Porosity

Gas trapped in the solidifying weld, showing as round holes on the surface or scattered voids inside. It is the most common defect and almost always a shielding or cleanliness problem.

Shielding gas lost to wind, a blocked nozzle, a leaking hose, or flow set too low.
Contamination on the joint: oil, paint, rust, galvanising, cutting fluid or moisture.
Damp electrode coating or damp flux-cored wire.
Arc length too long, which lets air into the puddle.
Gas flow set too high, which creates turbulence and pulls air in — a counterintuitive but real cause.

Worm tracks — long narrow surface channels rather than round holes — nearly always mean moisture. On stick welding that points at electrodes that have been out of the oven too long.

Undercut

A groove melted into the base metal along the toe of the weld and left unfilled. It creates a sharp stress concentration and is a common rejection point on inspected work.

Amperage too high — the usual cause.
Travel speed too fast to fill the melted edge.
Excessive weave width, or not pausing at the toes on a vertical-up weave.
Wrong electrode angle, throwing the arc at one member.
Arc length too long.

Lack of fusion and lack of penetration

Lack of fusion means the weld metal did not bond to the base metal or to the previous pass. Lack of penetration means the weld did not reach the root. Both are serious — they are planar defects that behave like cracks under load, and neither is always visible from the surface.

Amperage or heat input too low for the thickness.
Travel speed too fast, so the arc rides on the puddle instead of the base metal.
Short-circuit MIG on thick plate — the classic cold lap scenario, and why many codes restrict it.
Root opening too small or the land too thick on a groove joint.
Slag or previous-pass contamination not cleaned out.
Wrong electrode for the job: shallow-penetration rods like E6013 and E7024 cannot make up for a joint that needs a digging arc.

Slag inclusions

Non-metallic slag trapped in the weld, usually between passes. Only affects processes that produce slag — stick, flux-cored and submerged arc.

Slag not fully removed between passes.
Travel speed too slow, letting the slag run ahead of the puddle.
Welding vertical-down with a heavy-slag electrode such as E7018 — gravity puts the slag in front of the arc.
Poor bead profile leaving deep valleys the next pass cannot clean out.
Wrong gun or electrode angle, pushing rather than dragging a flux-cored wire.

Cracking

Cracks fall into two families that have entirely different causes and fixes.

Hot cracking

Occurs while the weld is still solidifying, often as a centreline crack down the middle of the bead. Caused by low-melting-point films at the grain boundaries, deep narrow bead profiles, high restraint and, in stainless, insufficient delta ferrite. The fixes are a wider, shallower bead profile, lower heat input, and the correct filler — which is why E309L exists for dissimilar joints and why ER4043 is chosen over ER5356 on crack-sensitive aluminium.

Cold cracking

Also called hydrogen-induced or delayed cracking. It appears hours or even days after welding, usually in the heat-affected zone. It requires three things at once: diffusible hydrogen, a susceptible hardened microstructure, and tensile stress. Remove any one and it does not happen.

Use low-hydrogen consumables and keep them dry — this is the whole reason rod ovens exist.
Preheat to slow the cooling rate and give hydrogen time to escape.
Maintain interpass temperature; do not let the joint go cold between passes.
Reduce restraint where the design allows it.
On thick or high-strength steel, post-weld hydrogen bake-out may be specified.

Burn-through

The arc melts a hole straight through. Almost always too much heat for the thickness, but joint geometry contributes.

Reduce amperage, or step down an electrode or wire size.
Increase travel speed, or use stitch and skip welding to let the plate cool.
Use a smaller root opening, or add a backing bar.
Weld away from the edge on thin material, and watch corner joints — there is little metal there to absorb heat.

Spatter

Not a structural defect, but a cost: cleanup time and wasted filler. Usually too much voltage or too little, wrong gas, contaminated base metal, or too long a stickout on wire processes. CO2 shielding spatters more than an argon mix by nature.

Distortion

Weld metal shrinks as it cools, and that shrinkage pulls the joint. Angular distortion closes a single-V butt joint; longitudinal shrinkage bows a long fillet. Control it with balanced double-sided preparations, backstep and skip sequences, presetting the parts open, clamping and fixturing, and above all by not depositing more weld metal than the joint actually needs.

Reading a defect backwards

What you see	Look first at
Round holes in the bead	Shielding gas, joint cleanliness, arc length
Long surface channels (worm tracks)	Moisture in the electrode or flux
Groove along the weld toe	Amperage too high, travel too fast
Weld sitting on top without wetting in	Heat input too low, travel too fast
Crack down the bead centreline	Hot cracking: bead profile, filler chemistry, restraint
Crack appearing a day later	Cold cracking: hydrogen, preheat, consumable storage
Grey non-metallic material in the weld	Slag not cleaned between passes
Hole through the plate	Too much heat for the thickness

Frequently asked questions

What causes porosity in welding?

Gas trapped in the solidifying weld. The usual sources are lost shielding gas (wind, blocked nozzle, low flow, leaking hose), a contaminated joint, damp electrodes or flux-cored wire, and too long an arc.

What is the difference between lack of fusion and lack of penetration?

Lack of fusion means the weld metal did not bond to the base metal or the previous pass. Lack of penetration means the weld never reached the joint root. Both behave like cracks under load and both are frequently invisible from the surface.

Why did my weld crack a day after welding?

That is classic hydrogen-induced cold cracking. Hydrogen from moisture or contamination diffused into the heat-affected zone and, under residual stress in a hardened microstructure, initiated a crack hours later. Low-hydrogen consumables kept dry, plus preheat, are the prevention.

What causes undercut?

Most often amperage set too high, compounded by travel speed that is too fast to refill the melted edge. Excessive weaving and a poor electrode angle also contribute.

How do you stop burn-through on thin metal?

Reduce amperage, step down to a smaller electrode or wire, speed up travel, use stitch or skip welding to let the plate cool, and add a backing bar if the joint allows it.