Aluminium Is Not A Harder Steel

Aluminium Is Not a Harder Steel

The question arrives at every exhibition in some version of the same words: can this machine weld aluminium? It is a reasonable question and it is the wrong one to start with, because it treats aluminium as steel with a different number in the programme. It is not. Two of its properties are working against you at the same time, and the second one is the one that catches people out.

A machine that welds 2 + 2 mm steel perfectly well may be unable to make a usable weld in 2.5 + 2.5 mm aluminium, not because it is a poor machine but because the job has moved outside what it can deliver in the time available.

An aluminium panel welded along its flange, seen close. Bright, continuous, and no help at all in telling you whether there is porosity under it.

Two Numbers, Both Against You

Mild steel conducts electricity at roughly a tenth of the conductivity of copper. Aluminium conducts at roughly six times the rate of mild steel. In a spot weld that difference is not a small correction: the current that enters the sheet spreads sideways instead of concentrating at the interface, so the current density where you need it falls away. Aluminium therefore needs substantially more current than steel of the same thickness, not a little more.

The second number is thermal conductivity, and it is the harder one. Aluminium conducts heat at roughly four times the rate of steel. Heat generated at the interface leaks into the sheet as fast as it is generated, so lengthening the weld time does not build a bigger nugget; it builds a bigger heat-affected zone, softens more of the parent metal, and pushes the nugget outward until it expels at the edge. The window is short and it is high: a lot of current, delivered quickly, under more force than steel would need.

The Oxide Is Not Dirt

Aluminium carries an oxide layer that forms within minutes of any cleaning and melts at around 2,070 °C, far above the melting point of the metal underneath. It is also an electrical insulator. In resistance welding this means the contact resistance at the interface is not a property of the alloy but of whatever film is on it that hour, which is why two identical parts welded on two different days can need different settings.

There is no way to design around this. Either the current density is high enough to break the film as the weld starts, or the film is removed mechanically or chemically immediately before welding, and then it starts growing back. This is the single most common reason for a plant reporting that aluminium "welds fine in the morning and badly after lunch".

What Has to Change on the Machine

Three things, and a fourth that is often forgotten. The transformer needs headroom: the current required is high enough that a machine sized for steel at the same thickness will sit at its limit, with the current waveform collapsing before the weld is finished. The control needs a fast rise and a defined up-slope, so the current arrives before the heat has left, and a forging pulse after the current stops, so the nugget solidifies under pressure rather than shrinking away from the interface.

Force needs to be higher, for the same reason: the sheets must be held in intimate contact before the current arrives, because there is no time to establish contact once it has. And the electrodes need to be a different material. Copper alloys with aluminium readily, so a standard copper electrode pits, picks up aluminium and sticks within a few hundred welds. Dispersion-strengthened copper or a tungsten insert buys life, but it does not buy immunity, and the dressing interval on aluminium is measured in hundreds of welds rather than thousands.

A press-type spot welder at a customer in Malaysia, set up to run 2 + 2 mm steel and 2.5 + 2.5 mm aluminium on the same machine, on two schedules that share nothing.

The Enquiry We Got Wrong

A customer sent 2.5 + 2.5 mm aluminium sheet and asked whether our machine could do it. We took the schedule we were running on 2 + 2 mm steel, raised the current, and welded a set of coupons. The surfaces looked right: even dents, a light straw tint at the edge, no expulsion. We sent photographs and a short report.

The customer peeled them. Seven of ten left a button; three left a hole. The three that failed were the ones in the middle of the sheet, where the heat had furthest to travel and the current had the most aluminium to spread through. Our diagnosis, after the fact, was three faults at once: the weld time was too long because we had kept the steel timing, the force was too low to hold the sheets during a weld that short, and the electrode face was larger than the nugget could fill, which lowered the density exactly where we needed it.

The machine was capable. The programme was a steel programme with the current turned up, which is precisely the mistake the question invites.

The same trap exists on the laser side, in a different form. At a wavelength of about one micrometre, aluminium reflects most of the beam until a keyhole opens, so the power density needed to start is higher than for steel, and the moment it does start the reflectivity drops and the process runs away. Aluminium also dissolves hydrogen and does not reject it on solidification, so porosity forms from the inside; the oxide contributes oxygen, and the 6xxx alloys are prone to hot cracking along the centreline. A bright, smooth, continuous bead on aluminium proves that the beam was stable. It does not prove the seam is sound.

When the Two Sheets Are Not the Same Thickness

Aluminium is far less forgiving of a thickness mismatch than steel. The heat generated at the interface has to be balanced by each sheet conducting its share away, and because aluminium conducts so well the thicker sheet behaves as a heat sink. Weld 3 mm to 1 mm and the interface cools from the thick side while the thin side is already at melting point: the result is a surface mark, an expulsion, and no nugget in the thin sheet.

Where the combination cannot be avoided, the working answers are a smaller electrode face on the thin side so that the current density is higher there, a shorter time still, and, in the worst cases, a change to the design so that the thin sheet is not one of the two carrying the current. The last of those is a conversation to have while the drawing is still a drawing, because at that point it is free.

Surface Preparation Is a Clock

On steel, cleaning is a step: do it and it stays done for the shift. On aluminium it is a clock. Brushing or chemical cleaning resets the oxide to a thickness that will weld, and then it grows back, faster in humid air and faster still in a workshop with coolant mist in it. A part cleaned at eight in the morning is not the same part at four in the afternoon.

The practical answer is to clean at the station, not in the stores, and to set the schedule for the worst condition the part will be in rather than the best. It costs a little current and saves the argument about why the last batch of the day was the one that failed.

What to Put in the Enquiry

An enquiry that says "aluminium" is asking for a guess. These five items turn it into a quotation that will still be right when the machine is installed.

• Alloy and temper, not just the word aluminium: 1050, 5754 and 6082 are three different jobs.
• The thickness pair, and whether either sheet is an extrusion or a casting.
• How the surface will be prepared, and how long before welding.
• Whether the visible face matters, because that decides which side the mark goes on.
• The volume, and how often electrodes can be changed without stopping the line.

The honest summary is that aluminium is not a harder version of the same job. It is a different job that happens to use the same word, the same machine frame and, often, the same operator, who will tell you within a week which of the two they would rather run.

PDKJ continues to provide customized laser welding and resistance welding solutions based on different materials, workpiece structures, welding requirements, and production applications.

Related Products

• Spot Welding Machines
• Laser Welding Machines

A row of machines on our assembly floor. The same frame can be built for steel or for aluminium; the difference is in the transformer, the control and the tooling.

Contact PDKJ

Send us your part drawing or a sample, and our engineers will run a welding feasibility trial on your own material and recommend the equipment configured for your application.