Aluminium sheet laser cutting: process, gas and edge quality

Aluminium sheet laser cutting works on a different balance from steel. About 88% of the 1,070 nm beam is reflected at room temperature, and the sheet conducts heat away at roughly 205 W/m·K, four times the rate of mild steel. The cut happens because a focused 12 kW spot reaches around 38 MW/cm², which is enough to break the oxide and start the melt before the heat escapes.

That single fact explains most of the decisions on an aluminium part. The gas choice, the focus position, the pierce time and the edge quality all follow from a material that reflects the beam and pulls heat out of the kerf. A drawing that treats aluminium like CRCA will produce a part with a rough edge and a hole that tapers, and the supplier will be blamed for a material property.

The numbers below are the ones we cut to on a 12 kW fibre laser with a 40 ft × 16 ft bed. They are process values for 5052 and 5754, with notes where 1050, 6061 or 5083 behave differently.

Why aluminium sheet laser cutting needs power density rather than raw speed

The beam has to raise the surface past the oxide melt at about 2,070 °C before the metal underneath reaches its own melting point of 660 °C. On steel the oxide is thin and breaks down easily; on aluminium the oxide is tenacious and reforms within seconds, so the pierce is the hardest part of the cycle.

Power density is the lever. A 12 kW beam focused to a 0.2 mm spot covers 3.14 × 10⁻⁴ cm², which gives 38 MW/cm². Halve the spot to 0.1 mm and the same power gives four times the density, which is why lens choice and focus position change the cut more than a speed change does. The trade is depth of focus: a tight spot loses focus through 6 mm of sheet, so thick aluminium is cut with a larger spot and lower density, and the edge quality falls with it.

Reflectivity is not a safety story on its own at these power levels, but it does affect the pierce. A cold, polished surface reflects more than a rough one, which is why the first pierce of a cycle takes longer than the rest and why a pierce point placed on a scrap tab costs less time than a pierce in the middle of the part.

Focus position

For 1–3 mm aluminium the focus sits at or just below the top surface. For 4–8 mm it moves 1–2 mm below the surface, so the beam is widest at the top and narrowest near the bottom of the kerf. That keeps the kerf open at the exit and reduces the dross that otherwise attaches to the underside.

Get the focus wrong and the symptoms are readable. Focus too high gives a wide top kerf and heavy dross; focus too low gives a narrow kerf that the assist gas cannot clear, and the cut stalls halfway through the thickness.

Nitrogen, air or oxygen: the assist gas decision

Nitrogen at 12–16 bar produces a clean, oxide-free edge and is the default for any part that will be seen, welded, anodised or painted. It is also the slowest and the most expensive gas option per metre cut.

Compressed air at 6–10 bar is cheaper and faster on thin sheet, and it leaves a light grey oxide on the cut face. That is acceptable on an internal bracket, a stiffener or a part that will be powder coated after a blast. It is not acceptable on an edge that will be TIG welded without cleaning, because the oxide sits exactly where the arc needs clean metal.

Oxygen is not used on aluminium sheet. The exothermic reaction adds heat, but it also produces a heavy oxide and dross that has to be ground off, and on thin sheet the reaction is difficult to control. The apparent speed gain turns into a finishing operation.

Speed, kerf and thickness limits

The table is a working set for 5052 with nitrogen. Speeds are straight-line values on a 12 kW source; a nest with many small contours runs slower because acceleration and pierce time dominate.

Two limits are worth stating plainly. Holes below one material thickness taper, because the beam enters at an angle and the gas cannot clear the bottom; below half the thickness the hole becomes a pierce mark with a ragged exit. And a cut edge is not a machined edge: it carries a striation pattern and a heat-affected zone of 0.05–0.2 mm, which matters if the part is a bearing surface or a sealing face.

Kerf is not scrap to be ignored. On a nest with 10 mm webs, a 0.3 mm kerf on each side of every contour removes about 3% of the sheet area on a dense layout, and that is material the quote has to carry.

Alloy and temper change the cut more than the gauge

The same thickness of two aluminium alloys is two different cutting jobs. 1050 and 1100 are commercially pure, soft and highly reflective, and they produce a sticky dross that has to be removed by hand; they are used for trims and covers where strength does not matter. 5052 and 5754 carry 2.5–3% magnesium, cut with a clean edge and anodise evenly, which is why most enclosure and panel work lands on 5052-H32.

6061-T6 cuts well but the heat-affected zone loses its age hardening. Within about 0.2 mm of the cut, yield strength falls from the 240 MPa temper value toward the annealed 55 MPa, so a hole placed close to an edge in 6061-T6 is a hole in soft metal. That is a design consequence, not a cutting defect, and it is why a structural bracket in 6061 needs its edge distances checked rather than assumed.

5083 is the difficult one. At 4–4.5% magnesium the kerf is rougher, the dross is heavier and the cut needs more power and a slower pass. It is specified for marine and welded structures, and above 6 mm it is usually a waterjet part. Temper matters as well: soft O-temper sheet can gum on the bed and lift, while H32 and T6 cut cleanly at the speeds in the table.

Edge quality: what to put on the drawing

"Clean edge" is not a specification. ISO 9013 classifies thermal cuts by perpendicularity and roughness against thickness, and a class number on the drawing tells the inspector what to measure instead of what to admire.

For aluminium sheet, class 2 is a realistic callout on visible edges up to 3 mm, and class 3 is realistic at 6 mm. Class 1 is a precision cut that usually means a slower pass, a different gas setting and a higher piece price, and it is rarely needed on a panel that will be powder coated.

Three checks cover most of the requirement:

• Dross. None on the underside on a class 2 edge. Dross that brushes off is acceptable on a hidden face; dross that has to be chipped is a process problem, not a finishing problem.
• Roughness. Ra 1.6–3.2 µm on nitrogen-cut 2–3 mm sheet. An air-cut edge typically lands at Ra 4–6 µm plus an oxide film.
• Perpendicularity. Within 0.1 mm over the thickness on 3 mm, widening to 0.25 mm at 8 mm. A tapered hole is a hole that will not take a pressed-in insert.

Deburring is a separate decision from cut quality. A laser edge on aluminium has a fine, sharp burr on the exit side that is a handling hazard on thin sheet, and the deburring and edge quality guide covers when a tumbler, a brush or a hand pass is the right finish.

Nesting and the real cost driver

A worked nest shows where the money goes. Take a 2 mm 5052 blank at 800 × 400 mm and a 2,500 × 1,250 mm sheet.

Three blanks across at 400 mm plus four 10 mm webs is 1,240 mm, which fits the 1,250 mm width. Three down at 800 mm plus four webs is 2,440 mm, which fits the 2,500 mm length. That is nine parts on the sheet, 2.88 m² of 3.125 m², or 92% utilisation before the edge margin.

The cut length is the perimeter plus the holes: 2,400 mm around the blank and about 126 mm for four 10 mm holes, so 2,526 mm per part. At 7.5 m/min that is 20.2 seconds of cutting, and nine parts is 3.0 minutes, plus about two seconds of piercing per part and a toolpath that never runs at full speed. Call it 3.3 minutes a sheet.

Gas follows the time. A 1.5 mm nozzle at 14 bar flows around 30 Nm³/h, so 3.3 minutes consumes about 1.65 Nm³ of nitrogen for the sheet. On thin sheet the gas is a small share of the cost; on 6 mm it is a large one, because the same area takes four times as long. The laser cutting cost estimator works the split from thickness, nest and gas, and the nesting guide covers the layout rules that buy back material.

Utilisation is where a quote is really won. Moving from 92% to 85% on this part adds 0.26 m² of scrap per sheet, which is 8% more material for the same output. On aluminium, where the material is the dominant cost line, that is larger than any plausible speed gain.

What we cut aluminium on

Aluminium runs on the 12 kW fibre laser with the 40 ft × 16 ft bed, which takes a 2,500 mm blank and its full nest without a splice, and on the 6 kW and 3 kW support machines for thinner gauges and smaller nests. The bed size matters on aluminium because long canopy and enclosure panels are cut as one piece, so there is no joint to seal and no mismatch at a butt line.

The laser is rated for plate up to 20 mm, but the honest limit for a coated or anodised aluminium part is lower. Above 8 mm the edge class drops, the deburr time climbs and the taper on holes becomes a design constraint. When a part sits above that line we say so and quote the waterjet or sawn route instead, which is slower to set up but produces a square edge.

Aluminium is one of the named materials in the laser cutting capability, alongside CRCA, stainless, brass and copper, and the aluminium cutting guide covers the alloy-specific behaviour that this process article does not repeat.

What to put on an aluminium cutting drawing

1. Alloy and temper, such as 5052-H32 or 6061-T6. The temper changes both the cut and the formability afterwards.
2. Thickness with tolerance, because 2 mm nominal sheet can arrive at 1.9 mm and the nest and the bend allowance both move.
3. Edge class as an ISO 9013 class, on the edges where it matters rather than as a blanket note.
4. Assist gas, if the edge will be anodised, welded or left visible.
5. Hole sizes below 1t, flagged so the pierce can be sequenced or the hole moved to a drill.
6. Grain direction and the visible face, because aluminium scratches and the protective film has to go on the right side.
7. Deburr and finish sequence, including whether anodising happens before or after any forming.

When laser cutting is the wrong route for aluminium

Laser cutting is the wrong route above about 10 mm, where the edge class and the taper both fail the drawing, and on 5083 plate above 6 mm, where the magnesium content makes the kerf rough and the dross stubborn. Waterjet cuts those thicknesses with a square edge and no heat-affected zone, at a slower rate that the part's value usually justifies.

It is also the wrong route for a mirror-finish anodised sheet that will be handled after cutting. Every nest contact and every sheet separation mark shows through a bright anodised finish, and the part needs a protective film, a clean bed and a packing method that the laser shop has to plan for.

And it is the wrong route for a high-volume simple blank. A 200 mm washer at 100,000 pieces a year belongs on a press tool, where the piece price is a fraction of the laser time and the edge is sheared rather than melted. The laser cutting versus stamping comparison sets out where that crossover sits.

Send the DXF with the alloy, temper and edge class through the RFQ route and the quotation will state the gas, the nest and the ISO 9013 class it is priced to. The rest of the cutting library sits under the laser cutting and fabrication hub, including the tolerance guide that tells you what to hold on the drawing and what to leave to the process.