How to braze aluminium with a torch: rods, flux and flame control

A good air torch - a 1300°C-class cartridge head or a propane torch - brazes small aluminium work: sheet, tube, brackets, small repairs. With the right low-temperature rod that is honest, useful work, and it is the normal way these rods are used.

The honest limit is mass and structure. Thick sections sink heat faster than an air flame supplies it, and anything structural or load-critical wants oxy-fuel or TIG, not a cartridge torch. Tell us the job and we will say which side of that line it sits on - that judgement call is free, and getting it wrong is not.

Three properties make aluminium the awkward one. Its oxide skin melts far above the metal itself, so a surface that looks ready is still wearing a barrier the rod cannot wet. It gives no colour warning - steel glows before it goes, aluminium looks identical at 300°C and at collapse. And it conducts heat so well that the whole part, not the joint, is what you are really heating.

Every step in this guide exists to answer one of those three: cleaning and flux for the oxide, temperature reading by the flux rather than the metal for the missing colour change, and patient all-round preheating for the conductivity.

Torch brazing aluminium runs on low-temperature aluminium brazing rods - filler made to flow at the heat an air torch actually makes. This is a different family from the rods an oxy-fuel rig uses, and the rod matters more than the torch: a good rod on a modest flame beats a wrong rod on any flame.

Some rods carry their own flux in the core, others want a separate flux - find out before the bench, not during it. We confirm flux and diameter per job in the quote, precisely because the right answer depends on your metals and your joint. For the wider selection logic across aluminium, copper and steel, see how to choose a brazing rod.

Degrease first, then mechanically clean the joint faces until they are bright - a stainless wire brush kept for aluminium only, never the brush that has been near steel or brass. Cross-contaminated brushes smear other metals into the surface and the wetting suffers for it.

Clean, flux, braze, in one sitting. Freshly cleaned aluminium starts growing its oxide skin again immediately, so the surface you cleaned an hour ago is not the surface you cleaned.

Use a soft, slightly bushy flame rather than a hard hissing one, and keep it moving. The job is to bring the whole joint area up together - a flame parked on one spot just makes one hot spot surrounded by metal that is stealing the heat.

Because the metal will not change colour, read the flux instead. Flux dries, then melts, then turns active and glassy - that glassy stage is the metal telling you it is in the working range. On a windy day or a thick part, shield the work and preheat longer; a cartridge head with a built-in air-mix fan holds its flame noticeably steadier in a draught, which matters when the margin between ready and ruined is this narrow.

Preheat around the joint, not on it. When the flux goes glassy, touch the rod to the joint face - not to the flame. If the metal is ready, the rod melts from the metal's own heat and wets out along the seam; draw the rod along and let capillary action carry the filler into the fit.

If the rod does not melt on contact, the part is not ready - lift the rod away, keep the flame moving, and test again a few seconds later. The one thing you never do is melt the rod in the flame and drip it on: that is glue, not brazing, and it will fail the first time the joint is asked to work.

The rod balls up and sits there: the base metal is too cold, or the oxide won - preheat more, or clean again. The base sags or collapses: too hot - aluminium gives no warning, so next time read the flux and hold the flame further back. The rod melts but will not wet: a cleaning or flux failure, back to the brush. A porous, grey joint: the flux was overheated and spent before the rod arrived.

Every one of those is recoverable knowledge, and every one of them is why the first joint you braze should be a scrap piece, not the part.

Mixed joints - aluminium to copper is the one everyone asks about - are possible in some combinations and a poor idea in others. The answer depends on both metals, the joint shape and what the joint has to survive, and the rod choice changes with the pair.

Tell us both sides of the joint and we will confirm the rod before you order. That is a two-line email via the contact page, and it is cheaper than a drawer of the wrong rods.

Aluminium rewards rehearsal. Run the rod on offcuts of the same metal and thickness until the timing is in your hands - what glassy flux looks like, how the rod takes, how close collapse comes. Ten minutes of scrap teaches more than any guide, this one included.

That is the thinking behind the head-and-test-rods pack: the 1300°C EN417 head with aluminium and copper rods alongside, so the flame gets proved on your own joints before any bulk order. Once you know the rod you want, the brazing rods are sold on their own.