Aluminium conducts heat away from the arc faster than steel does, and it does so in every direction. The practical result is that the puddle is fed by the arc but drained by the surrounding metal, so the same arc that would overheat a stainless seam barely wets an aluminium one. Rising current is the instinct, and it works right up to the point where the part is hot everywhere and the weld runs away.
The better mental model is mass. Thick aluminium sections sink heat and need a lot of it; thin sections reach temperature almost instantly once the part is warm. The welder is managing the temperature of the whole part, not just the pool, and that is a different discipline from steel work.
Aluminium wears an oxide skin that melts at a far higher temperature than the metal underneath. So the arc can be melting the parent while the skin on top is still solid, which produces a puddle that refuses to wet out and a bead that sits on the surface like a blob of solder on a dirty joint.
That skin also re-forms instantly. There is no window in which you can clean the metal and then weld it at leisure in open air. Whatever the process does about the oxide, it has to do it during the weld, not before it.
Alternating current solves the oxide problem by reversing polarity many times a second. In one half of the cycle the arc puts heat into the work; in the other half it lifts the oxide off the surface and throws it clear. That second half is usually described as the cleaning action, and it is the reason aluminium TIG is done on AC rather than on direct current.
The trade is real: the cleaning half puts less heat into the joint than the welding half. So an arc with a heavy cleaning bias is cool and wide, and an arc with almost no cleaning bias is hot and prone to inclusions. The balance is the single most useful control on an aluminium machine.
The balance setting shifts the proportion of time spent in each half of the cycle. More cleaning gives a wider, whiter, etched zone either side of the bead and a shallower penetration profile. Less cleaning gives a narrower bead with deeper penetration and a greater risk of trapping oxide in the root.
Material condition drives the choice. Cast aluminium and heavily oxidised sheet want more cleaning; clean, newly cut extrusion wants less. Thick sections want penetration, thin sections want the arc to stay wide and cool. There is no universal setting, only a setting that suits the part in front of you.
The visible result is a frosted band along both toes of the bead. That band is normal and expected. A complete absence of it usually means the oxide was not removed and is now inside the weld.
Many aluminium alloys crack if they are welded without filler, because the composition at the centre of the pool falls into a range that is weak while it is still semi-solid. Adding the right filler changes the composition of the pool enough to move it out of that range, and the cracking stops.
That is why autogenous welding, joining two pieces with no filler at all, is a bad habit on aluminium even where the fit is perfect. It looks tidy and it produces a seam that can split along its centreline days later, especially if the part is later machined or heat cycled.
Aluminium tacks shrink hard. A part tacked at both ends and left to cool will pull, and the seam that was tight at the bench will be open by the time the weld reaches the middle. Tacks placed too far apart let the joint move; tacks placed too close together lock in stress.
The start of the seam is the other weak point. Aluminium is cold at the beginning and hot at the end, so a bead welded at constant speed will be cold and narrow at the start and wide and hot at the finish. Backing off the current as the part warms, or running a short starting ramp, keeps the bead consistent along its length.
Thin aluminium sheet is the hardest common case. The edge of a sheet has less mass to sink heat, so it reaches melting point before the middle of the seam, and the bead that is perfect in the centre runs off the edge as a ragged melt-back.
Techniques that help are all about giving the edge somewhere to dump heat: a chill bar behind the joint, a backing plate of steel or copper, or a heat sink clamped alongside. Filler added early, before the edge starts to collapse, is also a form of protection, because it gives the arc something to melt that is not the parent.
Aluminium welds are judged on appearance more than steel welds, because the metal is usually visible and because the etched cleaning band reads as either craftsmanship or sloppiness depending on how even it is. A consistent ripple pattern and a uniform frost line are what buyers look for.
None of that proves the joint. A beautiful bead can hide incomplete fusion at the root, and aluminium is good at hiding it because the oxide film can sit undisturbed at a fusion face while the cap looks perfect. Appearance is a starting point, not a verdict.
Aluminium parts are often quoted from a drawing with no note about alloy or temper, and that single omission decides how difficult the job really is.
Aluminium is a smaller part of our output than titanium or stainless, but the same standards apply. You can see the aluminium work on our aluminium TIG welding pages, the pipe side under exhaust pipe work and exhaust pipe sections, and fabricated assemblies under fabricated components. Stainless work is described under TIG stainless steel welding, and titanium under titanium TIG welding parts.
