Laser Cutting Mild Steel: Thickness and Edge Quality

A 3 mm CRCA bracket leaves the laser with a 0.3 mm kerf and an edge that goes straight into pretreatment. The same profile at 12 mm leaves a 0.8 mm kerf, a 10–40 µm oxide layer and a striated face that powder coat will telegraph through. Thickness does not just slow the machine; it changes what the cut edge is made of.

Laser cutting mild steel is a specification problem before it is a machine problem. Assist gas, the kerf allowance in the flat pattern, the smallest hole worth drawing and the edge acceptance criteria all move with thickness, and they move at different rates. The laser cutting and fabrication hub holds the rest of this cluster; this page covers the thickness bands and the edges they produce.

By the end you should be able to read a thickness off a drawing, state the kerf and roughness you will accept, and write the edge note that stops a 10 mm part being quoted and inspected as though it were 2 mm.

How laser cutting mild steel changes above 4 mm

The cut changes mechanism at roughly 4 mm. Below it, a 1.07 µm fibre beam melts the steel and a gas jet blows the melt out. Above it, oxygen assist starts an exothermic reaction with the iron, and that reaction supplies a large share of the energy needed to keep the kerf open. The beam no longer works alone.

That is why thick mild steel is cut with oxygen and thin mild steel is usually cut with nitrogen. Oxygen buys depth and speed. It pays for them with an oxidised edge, a wider heat-affected zone and a rougher face. Nitrogen shields the melt, so the beam has to do all the work; the edge stays bright and the kerf stays narrow, but the power required climbs steeply with thickness.

Kerf widens for a second reason. The focused spot stays around 0.1–0.3 mm across, but the melt front at the bottom of a thick section is wider than at the top because the gas jet loses coherence as it travels down the kerf. The result is a taper of a few hundredths of a millimetre per millimetre of thickness, and a kerf measured at the top face that understates what the part actually loses.

Material condition matters as much as thickness. CRCA sheet arrives cold reduced and oiled, so it cuts predictably and the oil burns off at the entry point. Hot-rolled pickled sheet carries a scale that varies across the coil, absorption changes where the scale is heavy, and a 6 mm hot-rolled nest can show a band of deeper striations along one edge. If the part is hot-rolled, specify edge roughness only where it is visible, because holding a CRCA edge standard on scaled material buys nothing but sorting.

The thickness, kerf and edge table

These are the bands we quote against. Kerf is measured at the top face, and roughness is on the cut face at mid-thickness. Both map to ISO 9013 quality classes rather than to a single pass or fail number, which is why the drawing note matters more than the table.

Read the last row as a limit, not as an alternative. A 25 mm profile is outside our laser envelope, and shearing plate to 40 mm does not produce the same part.

Choosing between oxygen and nitrogen assist

Use nitrogen below 4 mm. Use oxygen above 6 mm. Between 4 and 6 mm the finishing route decides, and the laser will do either.

Below 4 mm the power penalty for nitrogen is small and the edge quality gain is large: a bright, oxide-free face that takes powder coat after pretreatment with no mechanical preparation. Above 6 mm the penalty dominates. Nitrogen cutting needs the beam to supply the whole melt energy, and the required power rises faster than thickness because the gas jet also has to clear a longer kerf. On a 12 kW source, 10 mm in nitrogen is slow enough that oxygen becomes the production choice, and the edge is dealt with downstream.

What an oxidised edge does to a coating

Powder coat does not hide a rough substrate. On a 60–80 µm film, a striated face carrying 10 µm of oxide shows as a mottled panel under raking light, and the oxide is a weaker bond surface than blasted steel. A cross-hatch adhesion test per ISO 2409 can still pass on a badly prepared edge, which is why appearance rather than adhesion is usually what rejects the part. The failure mode and its fixes are set out in powder coating laser-cut edges.

There are two honest routes. Cut thin visible panels in nitrogen and skip the problem. Or cut thick parts in oxygen, blast the edge, and accept the extra handling. Treating the oxide as cosmetic is the route that produces a rework claim three weeks later.

Holes, slots and narrow webs

A hole smaller than about 1.2 × thickness in oxygen-cut mild steel is where most rejections start. The pierce takes longer than the contour, the gas jet has nowhere to go, and the hole comes out tapered or blocked with dross. Below that ratio, either open the hole up or plan to drill it after cutting.

Webs between parts follow a simpler rule: keep them at least equal to thickness, and 1.5 × thickness if the skeleton has to survive handling to the end of the nest. A 2 mm web in 6 mm plate will break on the slats and take the nest with it. Deburring and edge quality after laser cutting covers what to do when the web or the pierce has already gone wrong.

Pierce, lead-in and the first millimetre of the cut

Every contour starts with a pierce, and the pierce is where thickness turns into cycle time. On 2 mm sheet the pierce is a fraction of a second and barely visible. On 16 mm it is a pause of several seconds that ends in a crater wide enough to sit outside the finished profile. That is the rule: pierce on a scrap bridge or outside the part, then lead in on an arc of 2–3 mm so the contour starts at full speed and full pressure.

A nest that pierces every part individually wastes the advantage of common-line cutting. Two adjacent parts that share a cut line need one pierce between them, not two, and on 12 mm plate with 40 small parts that difference is a large part of the cycle. It also produces a better edge, because the shared line is cut once rather than entered twice.

Where a part has no scrap to pierce into, the pierce has to be softened: reduced power, a dwell, or a small pre-drilled hole. A 4 mm pre-drilled pierce hole in 12 mm plate costs a drilling operation and saves several seconds per part. On a 500-piece order it is worth doing; on a 20-piece order it is not.

The machines these thickness bands come from

The 12 kW fibre laser has a 40 ft × 16 ft bed, which is what makes 6,000 mm canopy panels and rail-length sections a single-piece cut rather than a splice. Two 6 kW and two 3 kW machines take the thinner and smaller work, so a 1.5 mm cover does not wait behind a 16 mm base plate. We cut mild steel, CRCA, stainless, aluminium, brass and copper, up to 20 mm plate on the laser, with 30–40 mm plate work handled on the iron worker by shear and punch.

Typical sheet metal tolerance is ±0.5 mm, tighter on critical features by agreement. Kerf is compensated in the nest from the chart for that material and thickness, so the flat pattern you approve is the one that gets cut. Files are accepted as PDF, DXF, DWG or STEP, and the laser cutting capability page lists what each machine holds.

What thickness does to the nest and the shop floor

Thickness changes handling before it changes cutting. A 2,500 mm × 1,250 mm sheet weighs 24.53 kg per millimetre of thickness at 7,850 kg/m³, so a 3 mm sheet is 73.6 kg, a 6 mm sheet is 147 kg and a 10 mm sheet is 245 kg. Two people can turn a 3 mm sheet on a table. They cannot turn a 10 mm one, so the nest has to be planned around a lifter.

The same arithmetic drives microjoints. A 300 × 200 mm part in 2 mm sheet stays put on friction alone. At 8 mm the same part weighs 3.77 kg, and a 0.5 mm microjoint will not hold it through a fast contour. We size microjoints at roughly 0.5 × thickness for parts under 2 kg and 1 mm or more above that, placed on an edge that will be hidden or trimmed.

Flatness is the third consequence. Oxygen cutting puts more heat into a thick section than nitrogen does into a thin one, and a long, narrow part will bow toward the cut side. On 10 mm and above, leave a skeleton strong enough to resist that pull, or cut the part with a stress-relief orientation rather than tiling it edge to edge.

What a first-off should measure

A first-off checks the features the next operation consumes, not the whole drawing. For a cut blank that means blank size, two hole positions from a single datum, kerf on a scrap leg, cut-face angle and dross on the bottom face. Flatness goes on a surface plate only if the part is long or thin enough to bow.

Kerf is worth measuring because it is the assumption everything else rests on. If the real kerf is 0.1 mm wider than the value in the cut chart, every hole comes out 0.1 mm larger and every outside profile 0.1 mm smaller, because the tool path is offset by half the kerf on each side. A 20 mm hole becomes 20.1 mm and a 20 mm boss becomes 19.9 mm. That is enough to fail a press fit and not enough to notice on a cover.

Record the result against the nest, the material heat number and the gas. If a later batch of the same part drifts, the record tells you whether the material changed or the process did. For repeat parts, the laser cutting tolerances for OEM parts set out which features can be held at ±0.5 mm and which need a note.

What to put on the drawing

Six fields carry the whole thickness decision:

1. Material grade and actual thickness, not a nominal with a wide band.
2. Kerf allowance, or a note that the flat pattern accounts for it.
3. Edge acceptance: roughness, dross permitted or not, oxide permitted or not.
4. Minimum hole and slot size, with a note on whether they may be drilled instead.
5. Finishing route, because it decides the assist gas.
6. Any bend line within 2 × thickness of a cut edge.

That last field matters more than it looks. A cut edge carries a heat-affected zone of 0.05–0.5 mm depending on gas and thickness. On 2 mm CRCA it is irrelevant. On 10 mm plate, a bend line inside that zone can crack before the bend reaches 90°, because the affected material has less ductility than the parent plate. Move the bend out, or cut that edge with a different gas, or accept a larger inside radius. The distance rules are worked through in hole-to-bend distance in sheet metal design.

When laser cutting mild steel is the wrong route

It is the wrong route above 20 mm, below 0.8 mm, and at volume. Above 20 mm the profile is outside our laser envelope and belongs with a plasma or oxy-fuel profile cutter, which we do not run. Below 0.8 mm the handling and dross control cost more than the part is worth. At steady volume above roughly 10,000 pieces a year, a blanking or progressive die beats a laser on piece price once tooling is amortised, and laser cutting versus stamping sets out where that crossover sits.

The route is also wrong when the drawing demands ±0.2 mm hole position across a 4,000 mm panel. That is not a laser tolerance on a large bed; it is a drill-and-fixture tolerance, and asking for it on the cut file only moves the cost to the inspection bench.

For a first estimate of cut cost against thickness, the laser cutting cost estimator works from perimeter, thickness and gas. For a firm number, send the drawing with the six fields above through the quote request route and we will quote the gas choice along with the price.