October 26, 2026 · Technical Guide

Laser Tube Cutting Tolerances: What You Can Hold, and What You Cannot

A fiber laser tube cutting machine with a ±0.03 mm positioning spec will not give you ±0.03 mm parts. On the finished tube, plan for ±0.1 mm on cut length and hole position, ±0.1° on rotation and 0.1–0.3° on end squareness. The rest of the error comes from the tube you feed in and how you clamp it, not from the machine you bought.

I quote tube lasers most weeks, and this is the first number buyers argue about. They read ±0.03 mm on a datasheet, then send a drawing marked ±0.05 mm on everything, then wonder why the first article fails incoming inspection. Here is where the numbers actually come from.

What does a ±0.03 mm positioning spec actually mean?

It describes where the cutting head goes, under test conditions. Cold machine, 20 °C, one axis at a time, no load on the chuck, measured with a laser interferometer. It is a machine tool number, and it is honest as far as it goes.

What you measure on the part is a different quantity. The part goes through a rotary chuck, a support system, a focal position that drifts with lens temperature, a kerf that changes with gas pressure and a tube that was never perfectly straight to begin with.

Datasheet figure	Typical value	What it means on a real tube
Positioning accuracy	±0.03 mm	head position vs. command, cold machine, no load
Repeatability	±0.02 mm	same program, same tube, back-to-back
Rotary resolution	0.001°	smallest commanded step, not the angular error you get
Cut length on part	±0.1 mm	what your customer measures after the part cools
Hole position on round tube	±0.1 mm axial, ±0.1° around	depends on datum, clocking and tube straightness
End squareness	0.1–0.3°	wall thickness and gas pressure dominate

The gap between the first row and the last three is the whole subject of this article. Nothing is wrong with the machine. The machining system is bigger than the machine.

Which tolerances should you specify separately?

Tube laser cutting does not have one tolerance. It has at least six, and they behave differently. A drawing that says "all dimensions ±0.1" tells the supplier nothing about which one you actually need.

Characteristic	Realistic laser result	Main driver
Cut length (end to end)	±0.1 mm up to 3 m, ±0.2 mm on 6 m	thermal expansion, tube temperature
Hole diameter	±0.1 mm on 1–6 mm wall	kerf width, focus, gas
Hole position, axial	±0.1 mm	feed calibration, tube end reference
Hole position, angular	±0.1° to ±0.3°	clocking datum, chuck slip, seam
End squareness	0.1–0.3°	wall thickness, cut strategy
Bevel / mitre angle	±0.5°	nozzle tilt, 3D head capability

Angular position is the one buyers forget. On a round tube you have a degree of freedom a flat sheet never has: the feature can rotate around the axis. Two holes can be exactly 120 mm apart and 20° off the clocking position. If the drawing does not define a clocking feature — a seam direction, a flat, a reference hole — nobody can inspect it and nobody can hold it.

Why does the tube eat more tolerance than the machine?

Three reasons, and all three scale with tube size.

Thermal expansion. The controller cuts to the drawing at the tube temperature, not at 20 °C. Carbon steel grows 11.7 µm per metre per °C, stainless about 16.5, aluminium about 23. On a 6 m tube that adds up fast.

Material	Expansion per m per °C	6 m tube, +10 °C	6 m tube, +20 °C
Carbon steel	11.7 µm	0.70 mm	1.40 mm
Stainless steel 304/316	16.5 µm	0.99 mm	1.98 mm
Aluminium 6061/6063	23.0 µm	1.38 mm	2.76 mm

This is why a shop that cuts 6 m aluminium extrusion in an unheated bay in winter and a warm bay in summer sees the same program produce parts 1–2 mm different. It is not the machine losing calibration. It is the material doing what the tables above say it does.

Chucking and support. A self-centering chuck on a 20–300 mm machine clamps with enough force to deform 1.5–2 mm wall tube by a few tenths of a millimetre. Cut a slot while the tube is squeezed and the slot closes or opens when the chuck releases. Support spacing does the same thing vertically: unsupported span on a 6 m tube sags and the end squareness changes with it.

Kerf and gas. Kerf on 1–3 mm stainless with nitrogen runs roughly 0.2–0.4 mm, and it is not constant across the wall. Hole size and slot width follow the kerf, so a hole dimension that is tight to the drawing will move when you change gas pressure, focus or nozzle.

What in the raw tube throws the cut off?

More than the machine does, in most shops.

EN 10219 cold formed hollow sections allow straightness deviation of 0.2% of total length. On a 6 m tube that is 12 mm of bow before you cut anything. Wall thickness tolerance is commonly ±10%, so a 3 mm wall can be 2.7 mm. Twist, out-of-roundness and weld seam height all add their own contribution.

Precision tube under EN 10305-1 is a different product: tighter wall tolerance, controlled straightness, defined surface condition. If your part needs ±0.05 mm on hole position, the cheaper fix is often precision tube, not a more expensive machine. I have seen a buyer spend USD 30,000 extra on a higher-spec machine to fix a problem that USD 300 per tonne of cold-drawn tube would have solved.

Ask two questions before you blame the laser. What standard is the tube supplied to, and has anyone measured the tube you are cutting right now?

Round tube or square tube: where does accuracy drop?

Round tube. One continuous curved surface, no natural datum. Angular position is the weak point, and the weld seam is the usual clocking reference — which only works if the seam is straight and consistently placed. Seam wandering is common in ERW tube.

Square and rectangular tube. Better for fixturing because flats exist, worse for consistency because corner radius and twist vary. Hole position measured from a corner inherits the corner radius tolerance. A hole 10 mm from a corner on 60×60×3 tube can move 0.2 mm just from radius variation, with the machine doing nothing wrong.

Open profiles. L-angle, C-channel and custom sections flex. They need support closer together and lower chuck pressure, and long parts are where this shows up. If the profile is thin-walled and long, expect the tolerance to follow the fixture, not the spec sheet.

End squareness and weld fit-up: where the millimetres come from

Squareness is usually quoted in degrees, and degrees sound small until you convert them to a gap.

Out-of-square	Ø50 mm tube	Ø100 mm tube	Ø200 mm tube
0.1°	0.09 mm	0.17 mm	0.35 mm
0.3°	0.26 mm	0.52 mm	1.05 mm
1.0°	0.87 mm	1.75 mm	3.49 mm

A TIG root on 3 mm wall wants a gap under about 1 mm. On a Ø200 mm tube, one degree of out-of-square eats the whole allowance on one side and closes the other. That is where the fitter starts grinding, and grinding is where the hours go.

Laser cutting on 3–6 mm wall normally lands in the 0.1–0.3° band when the tube is straight and supported. If your joint needs better, ask for a facing operation. Cutting a bit longer and machining the face is usually cheaper than fighting the machine.

How do you check accuracy before you accept the machine?

Do not accept a demo cut. Ask for a sample test on your own tube, at production speed, with a written report. Here is the test I recommend to buyers.

Three sizes, on the machine you are buying: Ø48×3 round, 60×60×3 square, Ø114×6 round
Ten parts of each, first part and last part measured separately
Cut length with a caliper or height gauge, recorded per part
Hole position measured axially from the cut end and angularly from the clocking feature
Squareness checked with a machinist square and feeler gauge, or on a granite plate
Wall thickness measured on the incoming tube, not assumed
Report states room temperature at the start and end of the run

The last two items are what separate a real report from a sales sheet. If nobody recorded the tube temperature and the incoming wall, the numbers cannot be compared with your production floor.

Run the same test twice, two hours apart. If cut length drifts on the second run, you have found the thermal effect and you know what to plan for in summer.

What should go on the drawing and the RFQ?

Datum: the tube axis plus one end plane, stated explicitly
Clocking feature: seam, flat, reference hole or a marked face
General tolerance for non-critical dimensions, and a short list of critical ones
Which standard the tube is supplied to (EN 10219, EN 10305-1, ASTM A500, or your own spec)
Whether holes are clearance, locating or weld-prep — they carry different tolerances
A STEP model for multi-face parts, because a 2D drawing hides the third face
Secondary operations expected after cutting: facing, deburring, machining, bending

Suppliers quote to what is written. A blank general tolerance block means someone will pick a number for you, and it will be the loose one.

Five mistakes I see in tube laser tolerance discussions

Treating the machine spec as the part tolerance. ±0.03 mm is a head position number, not a part number.
"All dimensions ²0.1 mm" with no datum and no clocking. Nobody can inspect it, so nobody can hold it.
Blaming the machine for the tube. Measure straightness, wall and twist of the incoming bundle first.
Ignoring tube temperature. A 6 m aluminium tube moves 1.4 mm per 10 °C. That is bigger than every other error combined.
Buying accuracy you do not need. If the hole is for an M8 clearance bolt, ±0.1 mm is generous. Save the tight tolerance for the features that locate.

None of this is a reason to avoid laser tube cutting. The market is moving this way for good reason — tube and pipe laser cutting was valued around USD 1.45 billion in 2025 and is forecast at USD 2.57 billion by 2035 at a 5.9% CAGR, while average installed power on new tube machines went from about 2 kW in 2020 to about 4 kW in 2026. The machines are getting better. The tube is still the tube.

Frequently Asked Questions

What tolerance can a fiber laser tube cutting machine hold?

On a production tube, plan on ±0.1 mm for cut length and hole position, ±0.1° for rotation and 0.1–0.3° for end squareness. The machine itself positions to ±0.03 mm, but the finished part adds thermal expansion, chuck clamping and the straightness of the tube you feed in. Our fiber laser tube cutting machine datasheet lists ±0.03 mm positioning accuracy, and we quote part tolerances separately for that reason.

How square is a laser cut tube end?

Usually 0.1–0.3° on wall thickness up to about 4 mm, which on a Ø100 mm tube is a 0.17–0.52 mm gap across the face. One degree would be a 1.75 mm gap, and that is where weld fit-up starts to fail. If your joint needs better than 0.1°, plan a facing or machining step after cutting.

Does wall thickness change the tolerance I can hold?

Yes. Thin wall moves, thick wall slopes. Below 2 mm the tube deflects under chuck pressure and the cut edge pulls, so support spacing and clamping pressure matter more than the machine spec. Above 10 mm the kerf tapers and the bottom edge lags, so squareness and hole size drift. The 3–6 mm range is the easy one.

Can a tube laser hold hole positions as well as a CNC machine?

Not quite, and you rarely need it. A tube laser lands holes within about ±0.1 mm on a round tube and ±0.1° around the axis; a CNC mill holds ±0.02 mm. Use the laser for clearance holes, slots and weld prep, and leave bearing bores or dowel fits for a second op. Deciding this at drawing stage saves the argument later.

Why does my cut length drift during a shift?

Heat. A 6 m carbon steel tube grows about 0.70 mm for every 10 °C it warms up, stainless about 1.0 mm and aluminium about 1.4 mm. If the machine is calibrated cold at 07:00 and the shop warms up by lunch, the same program cuts a different length. Measure the first part, then re-measure after two hours of running.

Sources: FANY LASER fiber laser tube cutting machine datasheet, 2026 (positioning accuracy ±0.03 mm, round tube 20–300 mm, up to 16 mm carbon steel at 6 kW); the six-characteristic tolerance framework follows the tube laser cutting tolerance guide published by Lumenfuture (2026); tube and pipe laser cutting market size USD 1.45 billion (2025) rising to USD 2.57 billion (2035) at 5.9% CAGR per DataMintelligence market report (2026); Bystronic / marketSTEEL reporting USD 1.5 billion (2024) to USD 2.8 billion (2033) at 8.1% CAGR; IndexBox (July 2026) forecast of 7–9% CAGR with average installed power rising from 2 kW in 2020 to 4 kW in 2026; EN 10219-2 tolerances for cold formed hollow sections (straightness 0.2% of length, wall ±10%); EN 10305-1 for cold drawn precision tube. Thermal expansion coefficients are standard values for carbon steel, austenitic stainless and 6xxx aluminium. Part tolerance figures are FANY LASER field observations from 2026 installations, not a guarantee. Confirm the tolerance on your own tube and drawing with the supplier before you order.

Written by David Chen, Sales Engineer at FANY LASER. Connect on LinkedIn for tube laser specs, tolerance reviews and sample test reports.

Have a tube drawing with tight tolerances on it? Send the drawing, the tube standard and one sample length, and we will tell you which features a laser tube cutting machine can hold and which ones need a second operation. Talk to our engineers here.