×
Oct 27,2026
Four different numbers on the same datasheet are all called "width". Only one of them cleans metal. The other three sell machines.
That story repeats itself every month, and it is not a story about salesmanship. It is a story about a unit that almost nobody converts correctly: watts per square metre. Power is the ceiling of what a laser can clean. Width is the multiplier that decides how much of that ceiling you actually reach before your arm, your robot, or your fixture runs out of travel.
This article decodes the vocabulary, gives you the two equations that govern every quote you will ever receive, and ends with a four-question spec sheet you can take to any supplier — including us.
Before any arithmetic, you have to know what you are reading. A typical laser cleaning datasheet uses four different measurements, and in English translation they collapse into one word.
| Label on the sheet | What it physically is | Typical figure | Why it is not your cleaning width |
|---|---|---|---|
| Spot Ø | Focused beam diameter at the focal plane | 50–300 µm | It is a dot. Used to compute the impressive "power density" headline |
| Scan width | Length of the line the galvo draws at maximum amplitude | 10–300 mm | Measured on a bare plate with no overlap requirement |
| Cleaning width | Scan width × 0.6, single pass, flash rust | 15–100 mm | The marketing number. Assumes the easiest coating on earth |
| Galvo field | Maximum addressable area of the mirror set | 100–400 mm | A travel envelope, not a swath. Nothing is cleaned at the edges |
Everything below follows from these. Fluence (F) is the energy delivered per unit area, in joules per square centimetre. It is the only thing the rust responds to.
Read the first equation twice, because it contains the whole argument. For a fixed coating (fixed F) and a fixed travel speed (fixed v — set by your operator or robot, not by the laser), the power you need is strictly proportional to width. Double the line, double the watts. That is the honest trade.
Now read the second. Throughput contains no power term at all. Watts do not appear in it. A machine with twice the power and half the width produces exactly the same square metres — and that is the case most buyers accidentally order.
F is not a guess. It is measurable, and it is dominated by how much material sits between the beam and bare metal. The figures below are single-pass effective fluence — total energy per cm² including every pulse needed to clear the layer, which is why they are much higher than textbook ablation thresholds.
| Coating | Thickness | Effective F (J/cm²) | Passes at 0.15 m/s | Difficulty |
|---|---|---|---|---|
| Oil, grease, fingerprint film | <5 µm | 2–4 | 1 | Trivial |
| Heat tint / weld discolouration | <2 µm | 4–8 | 1 | Easy |
| Flash rust (days old) | 5–15 µm | 6–10 | 1 | Easy |
| Paint, single coat | 60–100 µm | 15–25 | 1–2 | Moderate |
| Powder coat | 80–120 µm | 30–45 | 2 | Moderate |
| Medium rust, outdoor 6 months | 20–60 µm | 35–45 | 1–2 | Moderate |
| Laminated rust / mill scale | 80–200 µm | 80–110 | 2–3 | Hard |
Table 2 — Effective single-pass fluence by coating. Multiply by 1.4 for a 45° incidence angle.
Notice the spread: a factor of thirty between an oil film and mill scale. This is why "what power do I need?" is an unanswerable question on its own, and why every credible supplier asks for a sample before quoting. Anyone who quotes without asking what is on the metal is quoting a number, not a machine.
Now put the equations to work at a realistic handheld speed of 0.15 m/s (9 m/min) and 50% overlap. The middle columns show the power required to hold threshold fluence at that width and speed; the last column is the resulting throughput.
| Line width | P for flash rust (8 J/cm²) |
P for medium rust (40 J/cm²) |
P for mill scale (90 J/cm²) |
Throughput @ 50% overlap (medium rust) |
|---|---|---|---|---|
| 6 mm | 72 W | 360 W | 810 W | 1.6 m²/h |
| 10 mm | 120 W | 600 W | 1,350 W | 2.7 m²/h |
| 16 mm | 192 W | 960 W | 2,160 W | 4.3 m²/h |
| 25 mm | 300 W | 1,500 W | 3,375 W | 6.8 m²/h |
| 40 mm | 480 W | 2,400 W | 5,400 W | 10.8 m²/h |
| 50 mm | 600 W | 3,000 W | 6,750 W | 13.5 m²/h |
Table 3 — Power required and throughput produced at 0.15 m/s. Round up to the nearest standard size, never down.
Here is the same arithmetic run the other way, against the machines people actually buy. "Idle watts" is the difference between what you paid for and what you can physically deliver into the work at 0.15 m/s on medium rust. It is not lost in the cable. It lands in your part as heat.
| Machine you bought | Line width | Power usable @ 0.15 m/s | Idle watts (share of total) |
|---|---|---|---|
| 1,000 W, 6 mm line | 6 mm | 360 W | 640 W (64%) |
| 1,500 W, 8 mm line | 8 mm | 480 W | 1,020 W (68%) |
| 1,000 W, 16 mm line | 16 mm | 960 W | 40 W (4%) |
| 2,000 W, 25 mm line | 25 mm | 1,500 W | 500 W (25%) |
| 3,000 W, 40 mm line | 40 mm | 2,400 W | 600 W (20%) |
Table 4 — Utilisation at a fixed travel speed. Two-thirds of a mismatched machine's output never reaches the metal.
The 1,500 W / 8 mm combination is the most common mistake in the market, and it is expensive twice over. First, you paid about $2,400 for watts you cannot use. Second — and this is the part buyers discover after delivery — 125 J/cm² into an 8 mm line on 2 mm sheet produces measurable distortion and a heat-affected zone of 0.4–1.2 mm, enough to blue the edges and force a re-work pass on anything cosmetic. Over-cleaning is a real failure mode. The best cleaning result sits just above threshold, not far above it.
Width also buys tolerance, and tolerance is what keeps a job profitable on the second and third pass. A hand-held head wanders. The question is how much of that wander shows up as a defect.
| Wander per pass | 6 mm line | 10 mm line | 25 mm line | 40 mm line |
|---|---|---|---|---|
| ± 1 mm | 17% stripe | 10% | 4% | 2.5% |
| ± 2 mm | 33% stripe | 20% | 8% | 5% |
| ± 3 mm | 50% stripe | 30% | 12% | 7.5% |
| Visual result | Visible rust lines | Tiger striping | Blends out | Invisible |
Table 5 — Uncovered stripe width as a share of the line, per single pass.
A 2 mm drift on a 6 mm line leaves a third of the surface untouched — a rust stripe that fails incoming inspection, and that your operator will chase for the rest of the afternoon at 30% of production rate. The same 2 mm drift on a 40 mm line is a 5% band that disappears in the overlap. This is the hidden labour cost of a narrow line, and it never appears in a quotation.
The last piece of jargon to dismantle. Sales engineers often suggest that any machine can produce a wider line by backing the head off the focal plane. True in geometry, false in physics: widening a round beam spreads the same photons over a larger area, and irradiance falls with the square of the diameter.
| Defocus from focus | Beam Ø | Relative irradiance | Result on a 40 J/cm² job |
|---|---|---|---|
| 0 mm | 8 mm | 100% | Bare metal, single pass |
| +3 mm | 11 mm | 53% | Two passes; slower than before |
| +6 mm | 15 mm | 28% | Polishes the rust; no removal |
| +10 mm | 21 mm | 15% | Heat tint only — a cosmetic pass |
Table 6 — Round-beam defocus. Galvo line width behaves similarly: widening amplitude at fixed spot size forces a higher scan speed and fewer pulses per spot.
Defocus is a legitimate tool for adjusting a line by 10–20% to fit a flange or a corner. It is not a way to turn a narrow machine into a wide one. Width is bought in optics — a longer focal length, a larger galvo aperture, a beam expander — and that is where the money goes.
Put a price on the argument. Both options below clean the same 600 m² of medium rust at the same 4.3 m²/h, because throughput is width × speed and both machines can hold 16 mm at 0.15 m/s.
| Option | Machine | Line | Throughput | Capex |
|---|---|---|---|---|
| A — wide line, matched power | 1,000 W | 16 mm | 4.3 m²/h | $8,500 |
| B — narrow line, double power | 2,000 W | 8 mm | 2.2 m²/h | $14,000 |
Table 7 — Two routes to the same nominal duty. Option B costs 65% more and delivers half the area rate at the speed a hand can hold.
Option B is not a bad machine. It is a depth machine — 2 kW through 8 mm gives 167 J/cm² of headroom, enough to take mill scale off in one pass, and that is worth real money on the right job. The error is buying depth when the job is measured in square metres. Match width to area, match power to coating thickness. If you can only move one number, move width.
Take these to any supplier, in this order. If a supplier cannot answer question 1 with a number, everything after it is decoration.
Power is what the laser can do. Width is what your workshop lets it do. The datasheet sells the first because it is a single number that sorts neatly into a comparison table; the second requires knowing your coating, your part, and your travel speed — which is precisely the work a buyer is paying a good supplier to do with them.
If you remember one line: watts set the ceiling, width sets the floor you actually reach. A 1,000 W machine with a 16 mm line will out-clean a 1,500 W machine with an 8 mm line on nine jobs out of ten — and the tenth job is the one where you needed depth, not area.
We publish the pairing matrix on every quote we send, with the coating and the travel speed written next to it, so you can check the arithmetic yourself. If the numbers do not hold at your width, we would rather you found out before the wire transfer than after.
500 W pulsed cleaner, 8 mm practical line on medium rust — the width and the watts are matched, which is the whole point.
EN
