Argon Arc Welding Process: Ten Taboos to Keep in Mind
TIG welding punishes small mistakes faster than most processes. This guide lists ten taboos in the argon arc welding process, gives the parameter range that works instead, and explains which parts should never be welded this way. Written for engineers and shop planners who need sound welds the first time.
In this article
- 1
- 2
- 3
- 4
- 5
- 6
Key takeaways
Taboos 1 to 3: Polarity, Grind Angle and Electrode Size
Taboo one: do not run DC electrode positive on a general-purpose weld. In DC TIG the anode side carries far more heat than the cathode side. Put the torch on positive and the tungsten takes that heat load. It overheats, balls up and drops into the pool within seconds. Standard steel, stainless and titanium work runs DCEN, torch negative.
Taboo two: do not put a needle point on a high-current electrode. A thin, sharply tapered tip concentrates current density. The tip melts back, the arc column stretches along the cone, and the bead goes wide and shallow. Above roughly 100 A, knock the point back to a truncated tip of about 0.5 to 1.0 mm flat. The arc stabilizes and the penetration profile stays tight.
Taboo three: do not use one electrode size for every current. Match diameter to amperage: 1.6 mm for roughly 60 to 100 A, 2.4 mm for 100 to 180 A, 3.2 mm for 180 to 250 A. Push 200 A through a 1.6 mm tungsten and no grind angle will save it.
On AC, the balance control decides how much cleaning you get. More positive half-wave time breaks up the oxide on aluminium, but it also heats the tungsten. Set the smallest cleaning ratio that still produces a bright, oxide-free pool. On a rectangular-wave machine this is a dial setting, not a guess, and it changes with plate thickness and joint design.
- 1DCEN for steel, stainless, titaniumTorch negative keeps the heat in the part, not the electrode.
- 2Truncated tip above 100 A0.5 to 1.0 mm flat, not a needle.
- 3Size the electrode to the amps1.6 mm / 2.4 mm / 3.2 mm across the common range.
Taboos 4 to 6: Gas Flow, Nozzle Size and Arc Length
Taboo four: do not set gas flow without looking at nozzle diameter. Flow and cup size have a working pair. For manual argon welding, a flow of 5 to 25 l/min pairs with a nozzle of 5 to 20 mm. Low flow through a large cup produces a soft, floppy column that cannot push room air out of the way. High flow through a small cup produces turbulence that sucks air back into the shield.
Taboo five: do not run the arc long because it feels easier to see. Arc length sets both the bead width and the gas coverage. A long arc spreads the heat, widens the weld and lets the shield break down at the edges. Hold 2 to 5 mm for most manual work. If you cannot see the pool at that distance, change your helmet shade, not the arc length.
Taboo six: do not ignore electrode stick-out. The distance from the ceramic cup to the work controls arc length and gas coverage at the same time. Push the electrode too far out and the shield no longer reaches the pool. Bury it too deep and you cannot see the joint. Set stick-out so the tungsten tip sits about 3 to 6 mm inside the cup edge for a standard joint.
Draft and shop airflow matter here too. A fan or open door a few metres away will strip the argon trail off the cooling bead. The weld looks fine under the helmet and turns grey an inch later. If you cannot move the part, build a simple screen around the joint.
- 1Pair flow with cup size5 to 25 l/min with a 5 to 20 mm nozzle.
- 2Short arc, 2 to 5 mmLong arcs widen the bead and weaken the shield.
- 3Screen the jointDraft strips the trailing shield off the cooling bead.
Taboos 7 to 9: Travel Speed, Filler Feed and Arc Starting
Taboo seven: do not travel faster than the shield can follow. Travel speed is set by part thickness, welding current and preheat, and it has to deliver the penetration depth and bead width the drawing calls for. Run too fast and the shielding gas is dragged away behind the torch. The bead surface oxidizes and the edges undercut. Run too slow and you build a wide, cold, convex bead with poor fusion at the toes.
Taboo eight: do not stab the filler into the pool. The rod goes into the leading edge of the pool, not the arc. Dipping into the arc column freezes the pool, contaminates the tungsten and creates a cold lap. Feed the rod at a steady rhythm and keep the hot end inside the gas shield while you wait for the next dip.
Taboo nine: do not start the arc by scratching. Lifting the tungsten off the plate to strike an arc is unreliable and it burns the electrode tip. Touch-start leaves tungsten in the weld and forces a regrind every few strikes. Use high-frequency start or lift-arc on the machine. If the machine has neither, that is a reason to change the machine, not to keep scratching.
Crater treatment belongs in the same group. Terminate the current with a slope-down of about 1 to 3 seconds and add a final filler dip. Cutting the arc cold leaves a shrinkage crack sitting right at the end of the weld, which is where inspection looks first.
- 1Rod into the pool edgeNot into the arc column.
- 2Lift-arc or HF startScratch starting contaminates the electrode.
- 3Slope down 1 to 3 secondsFill the crater before the current stops.
Setting Up the Argon Arc Welding Process, Step by Step
Work through this order before the first bead.
- 11. Confirm the base metal and jointCheck the material certificate and the joint drawing. Clean 25 mm either side of the joint down to bare metal. Remove oil, oxide and any galvanized or painted layer.
- 22. Choose polarity and current modeDCEN for carbon steel, stainless steel and titanium. AC for aluminium and magnesium. Set the AC balance to the lowest cleaning ratio that still gives a bright pool.
- 33. Grind and size the tungstenGrind lengthwise, never across the electrode. Above 100 A leave a 0.5 to 1.0 mm flat tip. Size the electrode to the amperage band: 1.6 mm, 2.4 mm or 3.2 mm.
- 44. Set gas flow against nozzle sizePick a nozzle from 5 to 20 mm, then set flow from 5 to 25 l/min to match it. Purge the line for a few seconds before striking an arc.
- 55. Set arc length and stick-outHold 2 to 5 mm arc length. Set the tungsten tip 3 to 6 mm inside the cup edge so you can still see the pool and keep the shield over it.
- 66. Strike with HF or lift-arcStart on a scrap tab or the run-off plate, let the pool form, then move. Never scratch the tungsten on the part.
- 77. Weld with a steady travel speedLet the bead width and penetration set your pace. If the bead goes grey behind the torch, slow down or add a gas lens.
- 88. Slope down and fill the craterRamp the current down over 1 to 3 seconds, add one final dip, then hold the torch over the crater for a few seconds while the shield still covers it.
When Argon Arc Welding Fits and When It Does Not
Judged on thickness, material and production volume.
| Condition | Argon arc welding fits | Use another process |
|---|---|---|
| Wall thickness under 3 mm | Yes, full control of heat input | No, use laser or resistance welding |
| 3 to 6 mm | Yes, with a prepared V groove | Consider plasma or MIG for speed |
| Over 6 mm | Only with multi-pass and preheat | Submerged arc or EBW is faster |
| Aluminium sheet | Yes, on AC with cleaning balance | MIG for thick plate at volume |
| Titanium and Inconel | Yes, in a purged glove box | Furnace or EB welding for deep sections |
| High-volume small parts | No, cycle time is too long | Laser welding or brazing |
| Deep narrow grooves | Limited access for the torch | Laser or EB welding |
| Field repair on dirty steel | No, cleaning is not reliable | Stick or flux-cored arc welding |
Get the Setup Right, Then the Weld Follows
Most TIG defects trace back to polarity, gas coverage or arc length, not to operator skill. Fix those three and the rest of the argon arc welding process becomes predictable.
Common Questions
Why does my tungsten keep balling up on DC?
You are almost certainly on electrode positive, or the electrode is far too small for the current. Check the torch lead first, then check the diameter. A 1.6 mm tungsten will ball at 200 A no matter how you grind it.
On a machine with a fixed polarity switch, a reversed ground clamp does the same thing. Verify the work lead is on positive and the torch on negative before you change anything else.
How much argon flow do I actually need?
Match the flow to the nozzle, not to a number you remember from another job. A 5 to 20 mm nozzle works with 5 to 25 l/min. Start near the middle of that band and adjust in small steps.
If the bead goes grey a few millimetres behind the torch, increase flow slightly. If the pool gets turbulent or the arc wanders, you have gone too high and the flow is drawing air in.
Can I weld aluminium on DC?
Not in the usual sense. Aluminium forms an oxide layer with a melting point far above the base metal, and DCEN does not break it up well. You need the positive half-wave of an AC cycle for the cleaning action.
Thin aluminium can be welded on DC with helium and a lot of preparation, but the oxide has to be removed mechanically first and the window is narrow. For production work, AC is the practical answer.
What causes a grey or black bead?
Oxidation behind the torch. The usual causes are travel speed too high, gas flow too low for the cup size, arc length too long, or a draft pulling the shield away from the cooling bead.
Work through them in that order. Check the bead colour right behind the tungsten, not at the end of the weld, because that is where the shield is failing.
Do I need a gas lens?
It helps on joints with poor access and on titanium or stainless where any oxidation is a reject. A gas lens smooths the flow and lets you run a longer stick-out with the same coverage.
It is not a fix for low flow or a torn shield. If the bead is already greying, set the flow and travel speed first, then add the lens.
When should we machine a part instead of welding it?
When the joint is a cosmetic or structural feature that a single CNC setup can produce directly, welding adds a second process, a heat-affected zone and an inspection step. For low-volume brackets, housings and frame nodes, a machined part is often cheaper than a welded assembly.
We machine from one prototype to 10,000+ part runs, with tolerances to ±0.005 mm and 100% inspection before shipment. Send the drawing and we will run a DFM analysis with the quotation within 12 hours.
Send Us the Drawing, We Will Send Back a Plan
Upload a part and we will return a quotation with DFM analysis within 12 hours. Welded or machined, the decision gets made on the drawing, not on a guess.
12-hour quote100% inspectionNo minimum order quantity
More Process Notes
We publish setup notes, tooling trials and inspection data from the factory floor.
