Does aluminium need to be preheated before welding?

Usually not. Aluminium does not harden and does not suffer hydrogen-induced cold cracking like steel, so there is no minimum preheat temperature. Preheating is used for specific purposes: to remove moisture, to improve the cold start, to balance large differences in thickness or to slow down cooling on thick parts. For aluminium, maximum values apply, because too much heat costs strength.

This is the most important difference from steel. There you calculate how hot a component must be at least, for example with the preheat temperature calculator according to EN 1011-2. For aluminium, EN 1011-4 specifies how hot it may get at most.

When is preheating aluminium useful?

EN 1011-4 gives four reasons: removing moisture before welding, for example on site; avoiding imperfections at a cold start; balancing the heat where very different thicknesses are joined; and moderating the cooling rate when welding thick parts.

  • Moisture: cold parts mist over in warm workshop air. Brief warming dries the joint, see Condensation. According to the ASM handbook, brief heating to no more than 120 °C is enough for this.
  • Cold start: aluminium conducts heat around three times as well as steel. At the cold start of the weld the heat flows away so quickly that lack of fusion occurs and the gas has too little time to escape. Against porosity, too, the standard therefore recommends more heat input or preheating.
  • Differences in thickness: the thick part draws the heat away, the thin part overheats. Preheating the thick part compensates for this. A practical rule of thumb is a wall thickness ratio of no more than 1:3.
  • Thick parts: on large cross-sections, preheating slows down the rapid cooling that would otherwise lead to lack of fusion and porosity.

Preheating can often be avoided altogether. According to EN 1011-4, a shielding gas of argon and helium or pure helium can replace the preheating that would otherwise be needed, because helium brings the heat into the workpiece more effectively; more on this under Gas shielded welding.

Video (in German): welding thin to thick aluminium, when heat conduction becomes a trap.

How high may the preheat and interpass temperature be?

According to EN 1011-4, no more than 120 °C for non-heat-treatable alloys, 120 °C preheat and 100 °C interpass temperature for 6xxx alloys, and 100 °C and 80 °C respectively for 7xxx alloys. The interpass temperature applies at the start of each new run.

Base metal Maximum preheat temperature Maximum interpass temperature
Non-heat-treatable: 1xxx, 3xxx, 5xxx, AlSi castings, AlMg castings 120 °C 120 °C
Heat-treatable: 6xxx, AlSiMg castings, AlSiCu castings 120 °C 100 °C
Heat-treatable: 7xxx 100 °C 80 °C

Recommended maximum values according to EN 1011-4:2001-02, Table 1. The values are intended as a guideline and may be specified differently by contract; they then belong in the procedure qualification.

  • The heating time must be as short as possible. Holding at temperature for a long time can partially anneal work-hardened alloys and over-age heat-treated alloys.
  • In 5xxx alloys with more than 3.5 % magnesium and in 7xxx alloys, phases can precipitate that increase susceptibility to exfoliation and stress corrosion cracking in certain service environments.
  • For heat-treatable alloys, according to the ASM handbook, the highest strength is achieved with an interpass temperature of no more than 65 °C. Letting the part cool between runs therefore gains strength.

Higher values from practice for thick components

For thick components, higher preheat temperatures are also quoted in practice, but with a limited preheat time. An example from a presentation by Air Liquide (Cerkez Kaya, 19th ÖGS workshop 2018):

Material Thickness Maximum preheat temperature Maximum preheat time
AlMgSi over 20 mm 180 °C 60 min
AlSi1MgMn over 20 mm 200 °C 30 min
AlZn4,5Mg1 over 16 mm 140 °C 30 min
AlMg4,5Mn0,7, AlMg3 over 16 mm 150 to 200 °C 10 min

These values are well above the maximum values of EN 1011-4. They require the stated minimum thickness and a strictly limited time at temperature and are not a standard. Anyone wishing to use them should specify them by contract and demonstrate in the procedure qualification that the required strength is achieved.

Why does too much heat cost strength?

Because the heat-affected zone beside the weld loses its strength condition. Work-hardened alloys become soft, heat-treated alloys over-age. Every degree of preheat and every interpass that is too hot widens this zone.

  • Non-heat-treatable, work-hardened alloys such as 5083-H lose their work hardening after just a few seconds above 345 °C, according to the ASM handbook over 25 to 38 mm beside the weld. Welded in the soft O temper, by contrast, they retain almost their full load-bearing capacity: in a study on 13 mm thick 5083-O, the MIG weld reached a tensile strength of 308 MPa, the base metal around 333 MPa.
  • Heat-treatable alloys such as 6082-T6 over-age in the heat-affected zone, fastest between 290 and 425 °C. In the same study, MIG welds on 6082-T6 reached only 232 MPa instead of around 300 MPa, and the fracture was in the heat-affected zone. Preheating, not cooling between runs and slow weaving reduce this value further, until a procedure qualification may fail.

The weld itself reacts too. In the research project AiF 12.239 N, a backing heated to 40 °C made no difference. At 85 °C the welds became wider and the penetration deeper, with a tendency to burn through. Anyone who preheats has to adjust the parameters accordingly.

More on the differences between the alloy groups and on the choice of filler metal: Differences from steel.

How do you preheat correctly and how do you measure?

Evenly, as briefly as possible and with the temperature measured. Suitable methods are a furnace, electric heating mats, induction or hot air. According to EN 1011-4, the temperature is checked with contact thermometers, thermochromic paints or temperature-indicating crayons, the interpass temperature as close to the weld metal as possible.

Preheating with a flame is common but tricky with aluminium. The flame itself brings water vapour: according to Air Liquide around 31 % with propane-oxygen and around 3.6 % with acetylene-oxygen. On cold metal it first condenses as water. Anyone working with a flame heats until the surface is dry and avoids excess oxygen, which according to EN 1011-4 promotes oxide inclusions.

Between runs, clean each bead with a stainless steel brush before the next one is laid on top. EN 1011-4 recommends this as well.

Frequently asked questions

From what thickness does aluminium need to be preheated?

There is no fixed limit. Whether to preheat depends on thickness, thickness ratio, environment and shielding gas and is specified in the welding procedure specification. With argon-helium mixtures it can often be omitted.

How hot may 6082 get during welding?

According to EN 1011-4, no more than 120 °C when preheating and 100 °C between runs. The best strength is achieved with an even lower interpass temperature, according to the ASM handbook no more than 65 °C.

What is the interpass temperature?

The temperature of the welded joint at the start of each further run. It is measured as close to the weld metal as possible and should not exceed the maximum values.

Why does aluminium not need a minimum preheat temperature like steel?

In steel, preheating prevents hardening and hydrogen-induced cold cracking. Aluminium does not harden during welding. Here it is only about moisture, heat balance and fusion, and too much heat harms the strength.

About the author

Robert Lahnsteiner, Ing., is an International Welding Engineer (IWE) and managing director of MIGAL.CO GmbH. After training in telecommunications and electronics at HTL Braunau, he spent eighteen years in welding technology at Fronius from 1983. Since 2001 he has run MIGAL.CO, the aluminium welding wire manufacturer in Landau an der Isar, together with Martin Erl. His focus is aluminium filler metals, wire manufacturing and wire feeding, and the monitoring of welding processes. Within DVS, the German Welding Society, he has served for around 30 years on technical committee FA03 “Arc welding” and chairs working group 2.4.4 “MIG welding of aluminium”.

More about the author · LinkedIn