Molten metal globules thrown out of the arc and welded to the plate around the bead. Cosmetic on its own — but a reliable early warning that the arc is unstable and a parameter problem is building.
Spatter is a surface discontinuity: droplets of molten metal ejected from the arc that land on the plate or the weld and freeze there. It is the visible by-product of an unstable arc — most often in MIG, FCAW, and stick.
Spatter rarely fails a weld by itself, but it never shows up without a reason. It costs cleanup time, hardens the surface where it lands, can pit thin material — and the same instability that throws it is often producing porosity or undercut on the same bead. Read it as a symptom.
Weld spatter, molten metal globules, flying sparks buildup.
The spatter pattern tells you which side of the process is off — look at where it lands and how heavy it is.
Every spatter source is an instability: the arc cannot transfer metal smoothly, so the pool throws it out.
In short-circuit transfer the wire physically touches the pool, the current spikes, and the bridge of molten metal ruptures. Done cleanly, the droplet transfers smoothly. When the voltage is too low, the short-circuit is prolonged and the rupture is explosive — a shower of droplets is blasted sideways and lands on the plate. That is the classic MIG spatter signature.
The other causes work the same way: wrong polarity and worn contact tips make the arc wander and re-strike violently; contamination boils and pops; too much stick-out makes the wire whip. Whatever the source, the ejected droplets freeze where they land — and the same instability that threw them is robbing the pool of heat, which is why spatter so often travels with porosity and undercut.
A quiet arc is the whole game. Set it up for smooth transfer and spatter mostly disappears on its own.
Clean the metal to bright, choose the right gas for the wire, fit a clean contact tip, and apply anti-spatter to the nozzle and the plate around the joint.
Balance voltage and wire feed until the arc sounds steady and quiet; keep the correct polarity and hold a consistent stick-out. A hiss, not a pop.
Knock spatter off the nozzle and tip before it chokes the gas flow — a nozzle full of spatter is a porosity machine.
Anti-spatter is a bandage, not a fix. If you are spraying heavily to live with spatter, the arc is telling you something — fix the voltage, wire feed, polarity, or gas first.
Cleanup is simple — the cause-fix is the real work. Do the cause first, then the plate.
| Step | Action | Why |
|---|---|---|
| 1 | Fix the cause before cleanup — the same settings will throw the same spatter. | Cleanup without a cause-fix is doing the job twice. |
| 2 | Chip the large globules with a chipping hammer or pick. | Big drops come off cleanly once broken loose. |
| 3 | Wire-brush or grind the rest; on aluminum, sand or grind carefully. | Spatter embeds in soft metal — forcing it gouges the plate. |
| 4 | Check for arc pits under the spatter and grind out any deep ones. | A pit in thin material is a stress riser that cleanup alone will not fix. |
| 5 | Re-check the surface against the spec — some surfaces require spatter-free welds. | Coated and machined surfaces are often held tighter than the weld itself. |
Spatter removal on aluminum is the trap: grinding hard drags embedded globules across the plate and creates new pits. Remove it patiently, then check the area for arc damage before judging the weld acceptable.
Each process has its own spatter signature — and its own number-one fix.
The big one: voltage too low for the wire feed, or the wrong gas for the wire. Short-circuit transfer should be quiet — if it pops and spits, raise the voltage a step at a time and rebalance the wire feed with it.
Polarity is the FCAW trap — self-shielded wires run DCEN, gas-shielded wires DCEP. Wire them wrong and the arc sprays. Also check stick-out: self-shielded wants a long one, gas-shielded a short one.
High current and a long arc throw slag and metal; damp rods sputter. Drop the amps a touch, hold a tight arc, and keep the rod angle right — stick spatter is mostly technique and rod condition.
Genuine spatter is rare in TIG — the electrode is non-consumable. If the weld looks spattered, the usual suspects are a contaminated tungsten or filler rod; the defect to look for is tungsten inclusion, not spatter.
Voltage and wire feed are the first two knobs on every spatter job. Check them as a pair, not alone.
| Parameter | Check | Typical Fix |
|---|---|---|
| Voltage | Arc pops and spits; rough, harsh sound | Raise voltage until the transfer is smooth; rebalance wire feed with it |
| Wire feed speed | Wire drives into the pool faster than it melts | Lower WFS or raise voltage together — never one alone |
| Polarity | DCEP for MIG and gas-shielded FCAW; DCEN for self-shielded | Set per the wire manufacturer — wrong polarity sprays |
| Gas mix | Wrong mix for the wire or material | 75/25 argon/CO₂ or 100% CO₂ per the wire spec |
| Contact tip | Worn, wrong size, or dirty | Replace; match the tip to the wire size |
| Stick-out | Too long makes the wire whip | 10–15× wire Ø for MIG; per the wire spec for FCAW |
Use the Welding Calculator for a starting voltage and wire-feed balance for your wire size, and the Gas Flow Calculator to verify the shield is actually covering the arc — a gas problem shows up first as spatter.
The spatter hardware chain runs from the power source to the contact tip — every worn link adds instability.
Check arc control on the power source, the drive rolls and liner on the feeder, and the contact tip and nozzle on the gun. Consumables covers the contact tips, nozzles, and anti-spatter products that keep the chain clean.
Spatter is found by eye — but its causes are found by listening to the arc.
Acceptance depends on the code and project: AWS D1.1, ISO 5817, API 1104 and project specifications each set their own limits for this discontinuity. Check the governing document before judging a weld acceptable — there is no universal pass/fail rule.
The questions welders actually ask about spatter — answered straight.
A little spatter is normal in short-circuit MIG and in stick — it comes with the process. Heavy spatter is not: it signals an unstable arc and is a warning to fix settings. Codes and project specs often require spatter removal on surfaces that will be coated, machined, or inspected — check the governing document.
New machines ship with generic defaults that rarely match your wire size, gas, and material. Check polarity first, then confirm the contact tip matches the wire, the gas is the right type and flowing, and the preset voltage and wire feed match the wire you actually bought.
Used on the nozzle and the plate around the joint, no — it is routine. The risk is over-application: spray that reaches the joint or the wire can contaminate the pool and cause porosity. Spray sparingly and keep it off the joint faces.
Aluminum MIG is sensitive to the transfer mode: it needs the right voltage and inductance for the wire size, 100 percent argon — never CO2 mixes — and a clean, oxide-free surface. Aluminum is also soft: spatter embeds, so the same settings that pit steel will damage aluminum plate.