Key takeaways
- Laser wins on mild steel and stainless up to 20 mm on cut speed, kerf and small-hole capability; waterjet wins only when the edge must carry no heat effect.
- A waterjet kerf of 0.8–1.5 mm costs 2–5 times the material allowance of a laser kerf, which shows up in the nest long before it shows up on the edge.
- A 1,200 mm cut through 20 mm plate at 200 mm/min consumes about 2.4 kg of garnet; the same cut on the 12 kW laser takes about 1.2 min and no abrasive.
- ISO 9013 quality classes apply to thermal cuts, so a waterjet part cannot be inspected against them unless the drawing says so explicitly.
- Choose waterjet for pre-hardened tool steel, titanium, composites and plate above 20 mm; choose laser for everything else on this page.
A buyer comparing laser cutting versus waterjet for sheet metal is usually asking the wrong question. Both processes will hold the outline. What separates them is the state of the 1 mm of material at the edge, the width of the material they remove, and the thickness where each stops being economical.
On a 3 mm CRCA bracket there is no contest: the laser is faster, uses less material and holds tighter holes. On a 60 mm pre-hardened die block there is no contest either, but the other way. Everything interesting happens in between, and most sheet metal procurement lives in the 2–12 mm band where the laser is the right answer and waterjet is a fallback for a handful of materials. The laser cutting and fabrication hub covers the rest of the laser route.
Here are the numbers that separate the two, the crossover thicknesses, and the cases where neither is the right process.
Laser cutting versus waterjet: what is the difference?
The laser melts and, on mild steel, partly burns the material. The waterjet erodes it with a stream of water carrying garnet abrasive at 3,000–4,000 bar. One is a thermal process with a heat-affected zone; the other is a mechanical one with no heat input at all.
That single difference drives everything else. Because the laser delivers energy through a focused spot, its kerf is narrow — 0.1–1.2 mm depending on thickness — and its cut is fast in thin material. Because the waterjet carries abrasive in a jet that spreads, its kerf is 0.8–1.5 mm and it cannot easily cut a hole much smaller than the kerf plus the taper.
The heat is the trade. A laser leaves a heat-affected zone of 0.05–0.5 mm, an oxide on mild steel and a chromium-depleted layer on stainless. A waterjet leaves the metallurgy untouched, which is why it is specified for pre-hardened tool steel, titanium, thick aluminium and anything that will be machined immediately after cutting.
The waterjet's weakness is geometrical. The jet loses coherence as it travels through the material, and the cutting front lags behind the nozzle, so the bottom of the kerf trails and widens. That produces taper and a slight curve on a contour cut at speed. Slowing the cut reduces both, which is why a waterjet job with a tight tolerance is a slow waterjet job. On thin sheet the laser's focused spot does not have this problem, and its taper stays in the hundredths of a millimetre.
Laser and waterjet compared on the numbers
Values below are for a production job in mild steel unless stated. The Ganesh column says what is available in our plant, because a comparison that ignores what the supplier actually runs is a comparison of brochures.
| Criterion | Fibre laser | Abrasive waterjet | Ganesh limit or note |
|---|---|---|---|
| Kerf | 0.10–1.20 mm, widening with thickness | 0.80–1.50 mm | Laser kerf compensated in the nest from a chart |
| Heat-affected zone | 0.05–0.5 mm | None | We cut to 20 mm plate on the 12 kW laser |
| Hole position tolerance | ±0.1–0.2 mm in 2 mm sheet | ±0.2–0.3 mm in 2 mm sheet | ±0.5 mm typical overall, tighter by agreement |
| Cut-face roughness | Ra 1.5–15 µm depending on gas and thickness | Ra 3–6 µm, uniform matte | Laser roughness varies most with gas choice |
| Mild steel maximum | 20 mm on the 12 kW source | 100 mm and beyond | Above 20 mm the profile leaves our plant |
| Stainless maximum | 20 mm | 100 mm and beyond | 304 and 316L both cut in nitrogen |
| Aluminium maximum | 20 mm | 150 mm and beyond | Aluminium above 20 mm is not a laser job |
| Small hole limit | About 1 × thickness | About 1.5 × thickness plus taper | Drilled holes below that |
| Taper | 0.01–0.05 mm per mm of thickness, both kerf walls together | 0.0025–0.01 mm per mm of thickness per side | Both matter on press fits, neither on covers |
| Consumables | Assist gas, lens and nozzle | Garnet, nozzle, water, pump seals | Gas cost is the largest running cost on stainless |
| Distortion | Low on thin sheet, present on long parts | Almost none | Long thin laser parts need a stress-relief nest |
The material row is the one that decides most enquiries. A waterjet kerf of 1.0 mm removes five times the material of a 0.2 mm laser kerf, and on a nest with hundreds of small parts that difference shows up as sheets, not as edge quality.
Where the two processes cross over
Thickness, edge preparation and abrasive consumption decide the crossover, and only the first is fixed by the machine.
The first is thickness. On our machines the laser stops at 20 mm plate. That is a hard limit and there is no workaround; a 25 mm profile needs a different process. Below 20 mm the laser can make the part; whether it should is the next question.
The second is how much edge preparation the part needs afterwards. Below 6 mm, a nitrogen-cut edge on mild steel or stainless goes straight to pretreatment. Above 8 mm, oxygen-cut mild steel carries an oxide that has to be blasted before coating, and that handling step can cost more than the difference between processes. A waterjet edge arrives with no oxide, but it also arrives with a matte surface that still has to be prepared before powder coat, so the saving is smaller than it first appears.
The third is abrasive consumption, which is what a waterjet buyer rarely sees in a quotation. Garnet runs at roughly 0.3–0.5 kg/min. A single 1,200 mm contour through 20 mm plate at 200 mm/min takes 6 minutes and consumes about 2.4 kg of abrasive. The same 1,200 mm contour on the 12 kW laser at 1 m/min takes 1.2 minutes and consumes no abrasive, only assist gas. That ratio, not the headline cutting speed, is what makes the laser cheaper per part on steel.
A 10 mm stainless flange on both routes
Take a 10 mm 316L flange with a 1,440 mm outline, cut in nitrogen. On the 12 kW laser at 0.8 m/min the cut takes 1.8 minutes. On an abrasive waterjet at 250 mm/min it takes 5.8 minutes and consumes about 2.3 kg of garnet. For 100 flanges the difference is roughly 6.6 hours of machine time and 230 kg of abrasive.
Now price the edge. The laser edge comes off with light tint on the exit face and, if the flange is visible, needs blasting. The waterjet edge is matte and tint-free but is not a finished surface either; it still needs preparation before powder coat. In most cases the 6.6 hours decides the question, and the waterjet case rests on the metallurgy rather than the edge appearance.
Where the laser wins on sheet metal
The laser wins on anything thin, anything with holes, and anything where the nest matters. On 2 mm sheet it holds hole position to ±0.1–0.2 mm and cuts small features a waterjet cannot attempt without losing the hole to taper. It nests tightly because its kerf is narrow, so a 1,250 × 2,500 mm sheet yields more parts.
It also wins on speed, and speed converts to capacity. A machine that finishes a nest in 20 minutes instead of two hours changes the delivery date, not just the price. For a buyer running scheduled releases against a production line, that is often the deciding factor rather than the rate per metre.
The limit is the edge. On 10 mm and above, the laser cut is a thermal cut and the part inherits the consequences: oxide, a small heat-affected zone and slightly reduced ductility at the edge. Laser cutting mild steel: thickness and edge quality sets out exactly what that edge looks like by thickness band.
Where waterjet is the only honest option
Waterjet is not a fallback for a supplier who lacks a laser. It is the correct process for pre-hardened material, titanium and thick aluminium, composites and laminates, and plate above 20 mm, and no laser setting reproduces any of them.
Pre-hardened material. A die block already hardened to 30 HRC cannot be laser cut without softening the cut edge, and the edge is where the die does its work. A waterjet cuts it cold.
Titanium and thick aluminium. Titanium is reactive at cutting temperature and needs inert assistance a fibre laser does not provide on thick sections. Aluminium above 20 mm reflects and conducts heat away faster than a laser can put it in.
Composites and laminates. A thermal cut chars the matrix and delaminates the edge. Nothing about a laser is right for these.
Plate above 20 mm. Past the laser envelope, waterjet or plasma is the route and the choice between them is about edge squareness and heat, not about the laser.
If the part is in one of those four groups, the comparison ends and the drawing should say waterjet. If it is mild steel at 6 mm, the comparison has already ended in the other direction.
When neither process fits
Some parts need a third answer, and it is worth saying so before a quotation is wasted.
Sheet below 0.8 mm with tight flatness and hundreds of small holes is usually better fine blanked or chemically etched than cut and flattened. Very high volume of a small, stable part is a press job, where a progressive die amortises the tooling and beats both processes on piece price. Parts that need a square, machined edge rather than a cut edge belong on a mill regardless of how the blank was made.
There is also the geometry case. A deep, narrow slot in 15 mm plate is difficult for both: the laser struggles to clear dross from the bottom, and the waterjet taper closes the slot at depth. Widening the slot, or accepting that it is machined after cutting, is cheaper than arguing about which cutter is better.
A hybrid route — waterjet the blank, then laser or machine the features — sounds efficient and usually is not. It adds a handling step, a second setup and a dimension chain between two processes, and the blank tolerance of the first process has to be tighter than the finished tolerance of the second allows. It earns its place only when the blank is very thick and the features are small.
How the two edges are inspected
A laser-cut part and a waterjet part cannot be inspected the same way, and a drawing that assumes they can will produce an argument at goods-in.
For a laser cut, the inspection is dimensional plus an edge assessment. Measure blank size and hole position from a single datum, then look at the cut face for dross, striations and tint. If the drawing cites an ISO 9013 class, the class fixes the perpendicularity and roughness limits for the thickness; without it, the inspector is judging appearance, which is not a measurement.
For a waterjet cut, the dimensional check is the same but the edge assessment is different. There is no tint and no dross to speak of; what varies is taper and the matte roughness the abrasive leaves. Measure taper with a micrometer across the thickness at a mid-contour point, not at a corner, because the taper changes as the jet enters and exits.
The specification that works for both is a roughness value, a perpendicularity limit and a statement of which faces are visible. That covers a laser edge, a waterjet edge and a part that moves between suppliers. Laser cutting tolerances for OEM parts gives the numbers we hold on the laser side.
What we cut, and what we hand back
We run a 12 kW fibre laser with a 40 ft × 16 ft bed, plus two 6 kW and two 3 kW machines for thinner and smaller work. That gives us mild steel, CRCA, stainless, aluminium, brass and copper up to 20 mm plate, 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, and we accept PDF, DXF, DWG and STEP.
We do not run an abrasive waterjet. That is a real limit, and it means a 40 mm stainless profile, a pre-hardened die block or a titanium bracket is not something we can quote as a cut part. We will say so at review rather than accept the order and subcontract it without telling you. The laser cutting capability page lists the envelope, and the parts we do build from cut blanks are on the custom parts page.
How to specify the cut on the drawing
Material and thickness, the edge requirement, the smallest hole and whether a heat-affected zone is allowed settle the process in four drawing lines:
- Material and thickness, with the grade designation.
- The edge requirement as a roughness and a perpendicularity limit, or a named ISO 9013 class if it is a thermal cut.
- The smallest hole and slot, with a note on whether they may be drilled.
- Whether the edge may carry a heat-affected zone, stated as yes or no rather than left to inference.
That last line is the one that decides between the two processes. A drawing that says "no thermal alteration at the cut edge" is asking for a waterjet, and it is better to know that before the nest is programmed. To compare cut cost against thickness on the laser route, use the laser cutting cost estimator; to place a part that sits outside it, send the drawing for review and we will confirm the route, including the ones we do not run.
Frequently asked questions
Is waterjet more accurate than laser cutting?
Not on thin sheet. A well-run fibre laser holds about ±0.1 mm on hole position in 2 mm steel, while an abrasive waterjet typically holds ±0.2 mm on the same part because of jet lag and taper. Waterjet gains an advantage on thick plate, where the laser taper and heat input grow.
Can a laser cut 25 mm mild steel?
Not on our machines. The 12 kW fibre laser is rated to 20 mm plate, and above that we would hand the profile to a waterjet or plasma route rather than force the cut. The iron worker shears and punches to 30–40 mm but does not cut profiles.
Why is a waterjet edge better for some parts?
It is cold. There is no heat-affected zone, no oxide and no change in hardness, so a pre-hardened die block or a titanium bracket keeps the properties it was bought for. On mild steel the cold edge is not worth the extra cost.
Does waterjet leave a taper?
Yes, usually 0.0025–0.01 mm per mm of thickness per side, which is 0.025–0.1 mm per 10 mm, depending on speed and abrasive flow. On 20 mm plate that is 0.05–0.2 mm of taper per side, which matters on a press fit and does not matter on a cover.
Which process is cheaper per part?
For thin steel at volume, laser. Above roughly 20 mm, or on material a laser cannot cut, waterjet is the only option and cost comparison stops. Between 10 and 20 mm the answer depends on how much edge preparation the laser part needs afterwards.
Sources
- ISO 9013:2017 — Thermal cutting: classification of thermal cuts; the scope covers oxy-fuel, plasma and laser, and other processes only by agreement · 3 Nov 2026
- ISO 2768-1 — General tolerances for linear and angular dimensions without individual tolerance indications · 3 Nov 2026
- IS 513 — Cold reduced carbon steel sheets and strips, the substrate most sheet metal nests are cut from · 3 Nov 2026
Not sure which cut your part needs?
Send the drawing with material, thickness and the edge requirement. We will tell you whether the part belongs on the laser or needs a route we do not run.
Send a drawing