What kerf width should a fiber laser cutting machine produce?

For a standard fiber laser cutting machine with a 1.2 to 1.5 mm nozzle, kerf width normally falls between 0.1 mm and 0.3 mm on thin sheet and widens to roughly 0.5 to 0.8 mm on 20 mm and thicker plate. Kerf width is set by beam quality, focus spot size and nozzle diameter, not by laser power alone. If a part needs a tight kerf for nesting, ask the supplier for the measured kerf at your working thickness instead of the brochure number, because ISO 9013:2017 covers laser cuts only from 0.5 mm to 32 mm, and tolerance classes change fast above 20 mm.

Does the assist gas change cut quality and defects?

Yes, and more than most operators expect. In a 2026 study on 09G2S structural steel cut with an 8 kW fiber laser, compressed air assist left an uneven cutting surface, an oxidized layer up to 200 microns thick and martensite in the near-surface zone, while oxygen assist produced an even, clear edge with a homogeneous heat affected zone and lower oxide content. Carbon steel is normally cut with oxygen, stainless steel with high pressure nitrogen, and compressed air is a cost compromise that accepts a rougher edge.

Technical Guide September 10, 2026

Laser Cutting Defects: How to Fix Dross, Burrs and Taper

Q: How do I fix burrs on laser cut parts?

Burrs are the hardened edge of dross that did not fall off. Run a fingernail along the bottom edge of the part: if it catches, the cut lost energy near the exit. Reduce cutting speed by 5-10% and cut one test part, then check focus position and nozzle centering. Do not raise gas pressure first. Published tests show increasing assist gas pressure from 0.3 bar to 1.5 bar raised cut surface roughness from 1.264 to 2.979 microns, so more gas often makes the edge worse, not better.

Q: Why is my laser cut edge tapered?

A laser kerf always widens toward the bottom of the cut, so a small taper is normal. Taper grows with material thickness and with cutting speed. In a 2024 study on S355JR steel, the smallest kerf taper was 0.027 mm with a 0.192 degree taper angle on 4 mm plate, but taper rose to 0.104 mm with a 0.747 degree angle when cutting speed was increased from 2900 to 4500 mm/min. On 6 mm plate the average kerf taper reached 0.199 mm. If taper is suddenly worse than usual, check nozzle centering, focus position and height control before changing the program.

Almost every laser cutting defect comes back to five things: cutting speed, focus position, assist gas pressure, nozzle condition and material flatness. Dross, burrs, taper, wide kerf and rough striations are all symptoms of those five. In a 2025 five-factor study on stainless steel, laser power and focus position together explained about 84% of dross formation, while gas pressure explained only 2.7%. That ordering matters. This guide shows what each defect looks like, what actually causes it, and which setting to change first.

The Six Laser Cutting Defects You Will Actually See

After a few hundred service calls, the same six problems come up again and again. Everything else is a variation of them.

Dross — molten metal stuck to the bottom edge of the part.
Burrs — hardened dross ridges that catch your fingernail and need grinding.
Taper — the kerf is wider at the bottom than at the top.
Kerf width drift — the cut is wider or narrower than the nest assumed.
Rough striations — drag lines on the cut face, heavy near the bottom.
Excess heat affected zone and discoloration — a wide oxidized band that shows up in welding and coating later.

The habit that causes most of the trouble is changing several settings at once. You get a better part, but you no longer know which change fixed it, and the next job starts from zero. Change one thing per test cut.

Dross in Laser Cutting: Soft Dross vs Hard Dross

Dross is metal that melted but never left the kerf. The important question is how it feels, because soft dross and hard dross point in opposite directions.

Soft dross — grey blobs you can knock off with a scraper. That is too much heat in the cut. The usual causes are cutting speed too low, power too high, or focus sitting too deep into the plate.

Hard dross — sharp, tenacious, and it takes a grinder. That is the cut losing energy before it reaches the bottom. The usual causes are speed too fast, power too low, focus too high, or assist gas pressure too low.

A recent five-factor study on stainless steel 201 ran 32 test cuts and ranked the influence of each variable on dross area: laser power about 45%, focus position about 39%, cutting speed 7%, pulse frequency 6.4%, assist gas pressure 2.7%. Power and focus are where the leverage is. If someone reaches for the gas regulator first, they are working on the smallest lever on the machine.

One more clue worth knowing: if the dross appears on one side of the part only, stop changing the program. One-sided dross is almost always a nozzle centering, alignment or gas symmetry problem. Cutting speed will not fix it.

Burrs: When Dross Hardens Instead of Falling Off

Burrs and dross are the same material at different temperatures. Dross that cools slowly stays soft. Dross that solidifies fast as a thin ridge becomes a burr.

Here is a quick test any operator can do without instruments. Run a fingernail along the bottom edge of the part. If it catches, the process ran short of energy at the exit. Reduce cutting speed by 5-10% and cut one test part before touching anything else.

What I would not do is raise the gas pressure to "blow the burr off". A 2024 study on S355JR steel measured surface roughness rising from 1.264 to 2.979 microns when assist gas pressure went from 0.3 bar to 1.5 bar. More gas gave a worse cut face, not a cleaner one. Gas pressure is a fine-tuning knob, not a cure.

If deburring is a fixed cost in your shop, it belongs in the machine decision. A well set 6kW fiber laser cutting machine cutting 3-6 mm steel with correct focus and clean nitrogen usually leaves a part you can send straight to the press brake. Two or three operators on a grinder every day means the parameters are wrong, not that the material is bad.

Taper and Kerf Width: What Counts as Normal

A laser kerf is never perfectly parallel. It widens toward the bottom of the cut, so some taper is built into the process. What you control is how much.

Quality measure (S355JR, 4 kW fiber laser)	4 mm plate	6 mm plate
Best kerf taper achieved	0.027 mm	0.15 mm
Best taper angle	0.192°	0.716°
Average kerf taper across tests	0.056 mm	0.199 mm
Heat affected zone depth	26-40 µm	155-231 µm
Dimensional deviation (best)	0.225 mm	0.096 mm

Four findings in that data are worth keeping in mind. Taper roughly tripled from 4 mm to 6 mm plate, which tells you thick plate needs more attention on focus. Cutting speed had a bigger effect on taper than laser power: pushing speed from 2900 to 4500 mm/min raised taper from 0.027 to 0.104 mm on the same power setting. High assist gas pressure also hurt geometry, with kerf taper reaching 0.337 mm at 4 bar. And the heat affected zone on 6 mm plate was about six times deeper than on 4 mm plate, which matters if the part gets welded or coated.

For reference on what to write on a drawing: ISO 9013:2017 classifies thermal cuts and covers laser cuts from 0.5 mm to 32 mm with defined tolerance classes. If your customer sends a drawing with an ISO 9013 class on it, ask which class before you promise a price. Holding a fine class on 25 mm plate is a different job from cutting the same part to a general tolerance.

Rough Striations: Compare Ra, Not Opinions

Roughness is the one defect where numbers beat adjectives. A portable roughness tester costs less than a day of scrap, and it settles arguments between shifts.

From the same 2024 test series: raising laser power from 2.8 kW to 4 kW reduced average surface roughness by 0.653 microns on 4 mm plate. Raising cutting speed from 2900 to 4500 mm/min reduced roughness by 0.765 microns. Raising gas pressure did the opposite, making the surface rougher.

Striations that open up toward the bottom of the cut mean the beam lost useful energy or the melt stopped being ejected as the cut went deeper. Check focus, standoff and gas delivery before you slow the machine down. Slowing a cut that is already losing energy at depth usually just widens the kerf and adds heat damage.

Assist Gas Changes the Defect List

The gas you choose does not only change speed. It changes which defects you get.

In a 2026 study on 09G2S structural steel cut with an 8 kW fiber laser, compressed air assist left an uneven cutting surface, an oxide layer up to 200 microns thick and martensite forming in the near-surface zone. Oxygen assist on the same material gave an even, clear edge, a homogeneous heat affected zone and lower oxide content. The air cut was cheaper per meter and more expensive per good part.

That does not make air wrong everywhere. A separate study on S235 structural steel found air assist produced the best results when cutting small features and thin sections. The rule I give customers is simple: oxygen for carbon steel, high pressure nitrogen for stainless and aluminum where the edge will be visible or welded, and compressed air when the part gets painted or ground anyway. Our assist gas comparison has the cost numbers per meter if you want to check the trade-off against your own gas price.

The Diagnostic Order That Saves Time

Most tuning time is wasted on the wrong variable. Work in this order and you will find the cause faster than by adjusting speed and gas together.

Stop if something is broken. After a head crash, a loose nozzle, a cracked optic, a gas leak or unstable height control, no parameter change is valid. Fix the hardware first.
Check material and flatness. A bowed sheet changes nozzle standoff as the head travels. If the defect follows the high spot, it is the plate, not the program.
Check the nozzle. It should be clean, round, correctly sized and centered on the beam. A partially blocked nozzle distorts the gas jet even when the gauge reads normal.
Run a focus test. Keep material, gas, nozzle, power, speed and standoff fixed and move only the focus through its documented range. Judge penetration, taper and dross, not just the top kerf width.
Then adjust speed or power, one at a time. State the expected result before the test cut. If reducing speed does not improve bottom penetration, the problem was never energy.

One practical note on where these defects bite hardest: thick plate and high power machines. A 20 kW fiber laser cutting machine cutting 40 mm carbon steel has less margin for a dirty nozzle than a 3 kW machine cutting 2 mm sheet, simply because the same error is applied over a much deeper kerf. If you are moving into thick plate, budget for a spare nozzle set, a lens inspection routine and an operator who understands focus.

If you want the wider picture on parameters and cost, our price and operating cost guide covers electricity, gas and consumables per hour for 1500W to 20kW machines.

Frequently Asked Questions

What causes dross in laser cutting?

Dross is molten metal that was not blown out of the kerf before it solidified. Soft dross means too much heat, so speed is too low, power too high, or focus sits too deep. Hard dross means the cut ran short of energy at depth, so speed is too fast, power too low, focus too high, or gas pressure is low. In a 32-run study on stainless steel 201, laser power and focus position explained about 45% and 39% of dross area, while gas pressure explained only 2.7%. Check power and focus first. If your dross is one-sided, check nozzle centering instead of the program.

How do I fix burrs on laser cut parts?

Burrs are hardened dross that did not fall off. Run a fingernail along the bottom edge: if it catches, the cut lost energy near the exit. Reduce cutting speed 5-10% and test one part, then verify focus and nozzle centering. Do not raise gas pressure as a first move. Tests on S355JR steel showed roughness rising from 1.264 to 2.979 microns as gas pressure went from 0.3 to 1.5 bar, so extra gas commonly makes the edge worse. If deburring is a daily cost, it is worth checking what a properly set fiber laser cutting machine leaves behind at your thickness before you accept grinding as normal.

Why is my laser cut edge tapered?

Some taper is normal because the kerf widens toward the bottom of the cut. Taper grows with plate thickness and with cutting speed. On S355JR steel, the best result on 4 mm plate was 0.027 mm taper with a 0.192 degree angle, and the average across tests was 0.056 mm, rising to 0.199 mm on 6 mm plate. Increasing speed from 2900 to 4500 mm/min pushed taper from 0.027 to 0.104 mm on the same power. If taper suddenly got worse, look at nozzle centering, focus position and height control before rewriting the cutting program.

What kerf width should I expect from a fiber laser cutter?

On a standard machine with a 1.2-1.5 mm nozzle, kerf width is usually 0.1-0.3 mm on thin sheet and widens to roughly 0.5-0.8 mm on 20 mm and thicker plate. Kerf is set by beam quality, focus spot size and nozzle diameter rather than laser power alone. For nesting work, ask the supplier for the measured kerf at your working thickness instead of the catalogue figure. Note that ISO 9013:2017 only covers laser cuts from 0.5 mm to 32 mm, and achievable tolerance classes loosen quickly above 20 mm.

Does assist gas change cut quality and defects?

Yes, more than most operators expect. A 2026 study on 09G2S steel cut with an 8 kW fiber laser found compressed air left an uneven surface, an oxide layer up to 200 microns thick and martensite near the surface, while oxygen gave an even, clear edge with a homogeneous heat affected zone. Carbon steel is normally cut with oxygen, stainless steel and aluminum with high pressure nitrogen, and compressed air is the cost compromise for parts that get painted or ground later. See the nitrogen vs oxygen vs air comparison for cost per meter.

Chasing dross and burrs every shift?

Send us your material, thickness, gas setup and the defect you keep seeing. FANY LASER builds fiber laser cutting machines from 1500W to 6000W and supplies nozzle, lens and focus-check consumables with every machine. We will come back with a parameter starting point for your thickness and a factory direct price. Contact our sales team and we reply within one business day.

Request a Quote

Sources: Physics-informed optimization of dross formation in fiber laser cutting, stainless steel 201, five-factor Central Composite Design (2025); Cepauskaite and Bendikiene, Effect of Fiber-Laser Parameters on Cutting Accuracy of Thin and Thick S355JR Structural Steel Plates, Metals 14(6):723 (2024); Kombayev et al., Influence of cutting parameters and assist gases on the laser cutting quality of 09G2S structural steel (2026); Olmez et al., Laser cutting of simple and complex geometries in structural steel, International Journal of Advanced Manufacturing Technology (2026); ISO 9013:2017 Thermal cutting - Classification of thermal cuts; ADH Machine Tool laser cutting troubleshooting guide (August 2026).

Written by David Chen, Sales Engineer at FANY LASER. Connect on LinkedIn for laser cutting parameter notes and machine selection data.