GreatLight CNC Machining Factory logo
CNC Machining
Rapid Prototyping
Materials
Industries
News
About GL

Get Instant Quote

Tube cutting setup

How to Measure and Set Up a Circular Tube Cutting Machine

This guide is for operators and process engineers who cut round tube, square tube and header rings on an intersection line machine. It covers the measuring steps, the parameters worth checking, and the errors that scrap the first part.

Ø20–Ø400 mm tubeWall 1–20 mmProbe first, then cutFirst-article check
Circular tube cutting machine setup guide for intersection line cuts
Quick answers

Key takeaways

Measure the tube, not the catalogHot-finished and welded tube runs 0.3–0.8 mm off nominal OD. Measure each batch before setup.
Chuck runout drives saddle errorKeep radial runout under 0.05 mm at the chuck jaws. Above 0.10 mm the saddle shifts visibly.
Probe four points, not oneA single touch finds the top of the tube. Four points at 90° find the real center and ovality.
Cut a test ring firstOne ring at the finished length tells you more than any screen value before you commit.
Log the offsetsWrite the X and Z offsets on the setup sheet. The next batch starts from that number.
Before you touch the control

What the measuring routine actually controls

An intersection line circular tube cutting machine does not cut a straight end. It cuts a saddle, a cross hole or a layered profile, and the shape is only correct if the machine knows where the tube axis really is. Nominal OD in the drawing is a starting point, not a measurement. Tube arrives with mill tolerance, weld seam height and a slight bend along its length.

The measuring routine has one job: convert the physical tube into a coordinate frame the controller can trust. That means finding the true center at the chuck, finding the true center at the support, and comparing the two. Any difference between them becomes an angular error in the cut path.

On a 500 mm long part, a 0.20 mm center offset at the support throws the cut plane off by roughly 0.02°. That sounds small. On a Ø150 mm tube it moves the saddle edge about 0.05 mm, which is already outside a ±0.05 mm fit for a welded joint.

So the setup order matters. Measure first, clamp second, probe third, cut last. Skipping a step and correcting with cutter compensation later works only on straight cuts. On a saddle, compensation shifts the whole curve.

  • 1
    Tube OD toleranceHot-finished round tube typically varies 0.3–0.8 mm from nominal. Welded tube varies less on OD but more on wall.
  • 2
    Wall thickness variationEccentric wall changes where the knife exits. Measure wall at four points, not one.
  • 3
    StraightnessA bow of 1 mm over 1,000 mm is normal for commercial tube. It shows up as center drift.
  • 4
    Surface conditionScale, rust and weld spatter change probe readings. Clean the touch area.
Tools and checks

Measuring the tube before it goes into the chuck

Use a digital caliper or a micrometer for OD and wall, and a laser or dial indicator for runout. Take OD readings at four points 90° apart, at three stations along the length: 50 mm from each end and at the middle. Record the largest and smallest values. The average becomes your working OD.

Wall thickness is the number people skip, and it is the one that causes the most rework. Measure at the same four points with an ultrasonic gauge. On drawn tube, wall can vary 8–12% around the circumference. If your cut depth is set from nominal wall, the knife will either leave a burr or dig into the support.

Check straightness by rolling the tube on a flat granite plate with a dial indicator on top. A 1,000 mm tube that bows more than 1.5 mm should be straightened or moved to a shorter cut. Long bowed tube will fight the chuck and the support at the same time.

Clean the area where the probe will touch. A 0.1 mm flake of scale reads as a 0.1 mm center shift. On stainless and aluminum, wipe with a lint-free cloth and check for burrs from the previous saw cut.

  • 1
    Working ODAverage of four readings. Use this in the controller, not the nominal value.
  • 2
    Wall rangeNote min and max. Set cut depth from the minimum value.
  • 3
    LengthMeasure to the reference end. A 2 mm length error moves every saddle along the axis.
Parameters

Cutting parameters that follow from the measurements

Once the center is known, the rest of the parameters are arithmetic. Cut depth per pass should stay between 0.15 and 0.35 × wall thickness for steel, and 0.25 and 0.5 × wall for aluminum. On a 3 mm wall steel tube, that means passes of 0.45 to 1.05 mm. Deeper passes push the tube away from the support and open the saddle.

Feed rate depends on the knife type and material. A carbide insert on mild steel runs 80–150 mm/min in the cut. Stainless 304 runs slower, 40–80 mm/min, because it work-hardens at the cut edge. Aluminum runs fast but needs clearance for chip evacuation, so keep the air blast on and check for built-up edge every 20 parts.

Spindle speed for tube cutting is usually lower than for solid bar. Most round tube in the Ø20–Ø150 mm range runs 200–600 rpm. Larger tube, above Ø200 mm, drops to 80–200 rpm. If you hear a ringing sound, the speed is too high or the support is loose.

Coolant matters more on the saddle than on a straight cut because the knife exits the wall at an angle. Flood coolant aimed at the exit point, not the entry. On stainless, use a high-pressure jet. On aluminum, use a mist or a light flood to keep chips moving.

  • 1
    Depth per pass0.15–0.35 × wall for steel, 0.25–0.5 × wall for aluminum.
  • 2
    Feed rate80–150 mm/min mild steel, 40–80 mm/min stainless 304, 200–400 mm/min aluminum.
  • 3
    Spindle speed200–600 rpm for Ø20–Ø150 mm, 80–200 rpm above Ø200 mm.
  • 4
    CoolantAim at the exit point. High pressure on stainless, mist on aluminum.
What goes wrong

Errors that show up on the first part

A saddle that is too deep on one side and too shallow on the other usually means the chuck center and support center disagree. Check the support height first. A 0.10 mm height error on a 500 mm span tilts the cut by about 0.01°, which is enough to see with a profile gauge.

A saddle that is correct in shape but off along the tube axis is a length error. The controller places the cut from the reference end. If the reference end was not faced square, the whole profile shifts. Face the reference end before measuring, or measure from a known shoulder.

Burrs on the exit side come from too much depth per pass or a dull knife. Reduce depth by 20% and check the insert. On stainless, a burr that looks like a smear means the feed is too low and the material is work-hardening.

Ovality in the finished cut, where the saddle is wider at one point than another, points to tube ovality that was not measured. The four-point probe catches this. Enter the average OD and the controller will still cut a round path, but the wall at the thin point will be smaller than expected.

  • 1
    Tilted saddleSupport height or lateral position is off. Re-probe both ends.
  • 2
    Shifted saddleLength value or reference end is wrong. Face the end and re-measure.
  • 3
    Exit burrDepth per pass too high, or insert worn. Reduce depth and inspect.
  • 4
    Uneven wallTube ovality or wall eccentricity. Map wall at eight points.
Setup sequence

Step by step: setting the circular tube cutting machine

  • 1
    Level and warm up the machineCheck the level of the bed with a 0.02 mm/m precision level at both ends. Run the spindle and axes for 15–20 minutes before measuring. Cold linear guides move 0.01–0.03 mm as they warm.
  • 2
    Load the tube and close the chuck lightlyInsert the tube into the chuck to a depth of at least 1.5 × OD, or 100 mm minimum. Close the jaws on low pressure first, around 30–40% of working pressure, so the tube can still be rotated by hand.
  • 3
    Find the center at the chuck with four probe touchesTouch at 0°, 90°, 180° and 270°. The controller calculates center X and Y. Repeat the cycle once. If the two centers differ by more than 0.02 mm, clean the jaws and probe tip and repeat.
  • 4
    Find the center at the support endMove the probe to the support position and repeat the four-point routine. Compare with the chuck center. Adjust the support height and lateral position until the difference is under 0.05 mm.
  • 5
    Enter the working OD and wallType in the averaged OD, the minimum wall, and the measured length. Do not leave the nominal values in place. The controller uses these to build the saddle curve.
  • 6
    Set axis offsets and confirm with a dry runRun the program with the cutter retracted and the feed override at 10%. Watch the Z axis track the saddle path. Any jump larger than 0.1 mm means a wrong center or a bad length value.
  • 7
    Cut a test ring at finished lengthCut one ring on scrap tube of the same batch. Measure the saddle with a profile gauge or a CMM. Check the wall at the thinnest point of the cut.
  • 8
    Lock and log the offsetsWrite X, Y, Z offsets and the working OD on the setup sheet. Include the date and the tube batch number. The next operator starts from your numbers.
Judgment table

Which setup method fits which tube

Pick the row that matches your tube and tolerance.

Tube conditionMeasuring methodSetup timeBest for
Drawn or cold-finished, OD within 0.1 mmSingle probe touch plus nominal OD5–8 minShort runs, loose fit
Hot-finished, OD varies 0.3–0.8 mmFour-point probe, working OD entered12–18 minSaddle cuts on welded joints
Welded tube with visible seamFour-point probe, seam rotated to a fixed angle15–20 minCross holes and layered profiles
Long tube over 1,000 mmFour-point probe at chuck and support20–30 minHeader rings, structural tube
Thin wall under 2 mmFour-point probe plus wall map at 8 points20–25 minAerospace and medical tube
Square or rectangular tubeProbe on two flats, derive center10–15 minFrame joints, not saddles

Measure the tube, then trust the numbers

A circular tube cutting machine is only as accurate as the center you give it. Four probe touches and a measured OD cost 15 minutes and decide whether the saddle fits.

FAQs

Questions we get from the shop floor

How often should the probe be recalibrated?

Recalibrate the probe tip against a known ring gauge at the start of every shift, or after any crash. A 0.01 mm drift in the probe shows up as a 0.01 mm center error.

If the machine runs the same family of parts all week, a daily check is enough. If it switches between tube sizes several times a day, check before each new family.

Can I use the nominal OD from the drawing?

Only for rough cuts where the fit is loose. For a welded saddle joint, the nominal OD is usually 0.3–0.8 mm off on hot-finished tube, and that error goes straight into the cut path.

Measure the batch and enter the working OD. It takes 12 minutes and saves the first part.

What tolerance can the machine hold on a saddle?

On a well-set machine, a saddle profile can hold ±0.05 mm on the cut edge and ±0.10 mm on the wall at the thin point. The limit is usually the tube, not the machine.

If the tube OD varies more than 0.5 mm in one batch, no setup can hold ±0.05 mm across the whole batch.

Do I need a CMM to check the first part?

No. A profile gauge, a digital height gauge and a caliper will catch most errors. A CMM is useful when the saddle is part of a welded assembly and the fit is critical.

For most shop-floor checks, measure the saddle depth at four points and the wall at the thinnest point. That is enough to know if the setup is good.

How do I handle a batch with mixed OD?

Sort the tube into groups by measured OD, in steps of 0.1 mm. Run each group with its own working OD. Mixing a 0.4 mm spread in one run will show up as a visible step between parts.

If sorting is not practical, set the working OD to the average and accept a wider tolerance on the saddle.

When should I stop adjusting and call for help?

If you have re-probed twice, checked the support, verified the length and the first part is still out by more than 0.15 mm, stop. Something mechanical is off: a loose jaw, a bent support pin or a worn ball screw.

Continuing to adjust offsets will hide the problem and scrap the next ten parts.

Send us your tube drawing

Upload a STEP file or a 2D drawing and we will review the saddle, the wall and the fit. Quotation and DFM analysis within 12 hours.

12-hour quote100% inspectionNo minimum order

Follow us

More from the shop floor

We publish setup notes, tooling trials and inspection data from the factory floor.

FacebookTikTokYouTubeLinkedInInstagramThreadsPinterest

Trusted by engineers and manufacturers worldwide

Tesla Ford Motor Company BYD Auto Denso Magna International Boeing Airbus Medtronic KUKA FANUC