Main CNC Tool Setting Technology for Five-Axis Work

• 1
  Key takeaways
• 2
  What main CNC tool setting technology actually controls
• 3
  Offline presetting vs on-machine measurement
• 4
  Why five-axis work raises the bar on tool data
• 5
  Thermal drift and when to re-measure
• 6
  When tight tool setting is worth the cycle time
• 7
  Step by step: setting a tool on a five-axis center
• 8
  Which tool setting method fits the job

Tool setting is the act of telling the control where the cutting edge is. On a three-axis mill that means a length offset and a radius value. On a five-axis center it also means the pivot distance, the tool vector, and the relationship between the rotary table center and the spindle. Every one of those numbers ends up in the same coordinate chain, so an error anywhere in the chain shows up on the part.

The main CNC tool setting technology in modern shops splits into two families. Offline presetting measures the tool on a bench unit and transfers the numbers to the control. On-machine setting measures the tool on the spindle using a probe, a laser, or a dial indicator against a known surface. Both work. The choice depends on how tight the tolerance is and how long the tool sits between measurement and cut.

For work held to ±0.005 mm, the chain matters more than any single step. A tool preset to 0.002 mm on a bench can still cut 0.015 mm off if the spindle grew after a cold start. That is why we treat setting as a sequence, not a single measurement.

• 1
  Length offsetSets Z position of the tip. Most common single source of depth error.
• 2
  Radius or diameter offsetControls wall position on contours and pocket sides.
• 3
  Tool vectorDirection of the axis in five-axis work; errors tilt the cut.
• 4
  Pivot distanceDistance from spindle gauge line to rotary center; a fixed machine value that must be calibrated.

An offline presetter measures length and diameter with a fixed reference and stores the values against a tool ID. It is fast, it keeps the spindle free, and it is repeatable to roughly 0.002–0.005 mm on a good unit. The weakness is time. A tool measured at 08:00 and loaded at 11:00 carries whatever thermal state the machine reached in between.

On-machine measurement closes that gap. A spindle probe or a non-contact laser measures the tool in the same thermal state it will cut in. Laser systems handle small tools well, down to Ø0.5 mm, and they do not touch the edge, so coated tools keep their coating intact. Touch probes are slower but work on any tool that can reach the probe stylus.

Manual setting with a dial indicator or a shim still has a place: one-off jobs, tools too large for the presetter, or a quick check when the probe is out of service. It depends on the operator, and repeatability lands near 0.01–0.02 mm. That is fine for roughing and not fine for a finishing pass on a sealing face.

We run all three at GreatLight depending on the job. A 10,000-part run gets laser setting on every tool change. A single prototype might get a dial indicator and a witness cut. The method should match the tolerance, not the habit.

• 1
  Offline presetterBest for large tool libraries and pre-staged setups; watch thermal lag.
• 2
  Laser on machineBest for small and coated tools; no edge contact.
• 3
  Touch probeWide tool range, slower cycle, needs a clean stylus.
• 4
  Manual indicatorFallback for oversize tools and one-off work; ±0.01–0.02 mm.

On a three-axis cut, the tool tip stays normal to the part. On a five-axis cut, the tool tilts. Once it tilts, the radius offset no longer acts in a single plane, and the length offset acts along a moving vector. A 0.01 mm length error on a tilted tool can shift the contact point by more than 0.01 mm on the surface, depending on the tilt angle.

The pivot distance is the other piece. It is the distance from the spindle gauge line to the center of the rotary trunnion. If that number is wrong by 0.02 mm, every tilted cut is wrong by roughly that amount, and the error grows with tilt angle. It is a machine calibration value, not a per-tool value, but it belongs in the same verification routine.

Rotary table center position matters too. On a Ø400 mm rotary table, a center error of 0.01 mm shows up as a 0.01 mm shift on any feature machined in a rotated orientation. For parts with features on four or five faces, that error stacks with tool offsets. Check both, or spend the afternoon chasing the wrong one.

• 1
  Tilt amplifies length errorVerify length at the actual cutting tilt, not at zero tilt.
• 2
  Pivot distance is machine-wideRecheck after any crash, spindle service, or rotary swap.
• 3
  Rotary center driftConfirm with a test bar or a known ring before a multi-face job.

A spindle grows as it warms. On a machine that has been idle overnight, the first 60–90 minutes of cutting can move the tip by 0.01–0.03 mm in Z even when nothing else changed. This is not a setting error. It is a setting that was correct at 07:30 and stale by 09:00.

The practical answer is a warm-up cycle followed by a re-check. Run the spindle at a moderate speed for 20–30 minutes, then probe the tool again and compare. If the delta is under 0.005 mm, the offsets can stay. If it is larger, update them and log the change.

Coolant temperature matters as much as spindle speed. A chiller holding ±1 °C keeps the machine frame stable. A chiller drifting 4 °C across a shift will move the part as much as the tool. If your tolerance is ±0.005 mm and your chiller swings, fix the chiller before you rewrite offsets.

Ambient swings in the shop do the same thing on a slower scale. A door left open in winter, or a roof that heats up in the afternoon, will move a large part more than a small one. For parts near the 4,000 mm travel limit, measure the workpiece temperature before the finishing pass.

• 1
  Warm-up first20–30 minutes at cutting speed before the first finish pass.
• 2
  Re-probe at 60–90 minutesUpdate offsets only if delta exceeds 0.005 mm.
• 3
  Watch the chillerHold ±1 °C for tight-tolerance work.

Not every part needs laser setting and a witness cut. A bracket with a ±0.1 mm profile tolerance can be set with a touch probe and shipped. Spending 20 minutes per tool on a job like that adds cost with no benefit.

The math changes when the tolerance tightens and the part value rises. A five-axis aerospace housing with a ±0.005 mm bore and a sealing face justifies the full sequence: warm-up, laser measurement, pivot check, witness cut. One scrapped housing costs more than a week of extra setting time.

There is also a geometric signal. Parts with features on three or more faces, deep pockets with thin walls, or surfaces cut at steep tilt angles carry more setting risk, because every offset error compounds across orientations. Those are the jobs where the extra checks pay back.

We make the call per job and tell the customer which method we will use before cutting. If a customer wants the cheaper route on a tight part, we say so and note the risk. That conversation is cheaper than a rework cycle.

• 1
  Worth itMulti-face features, tight bores, sealing faces, high part value.
• 2
  Not worth itLoose profile tolerances, simple two-axis work, roughing passes.
• 3
  BorderlineSingle-face parts at ±0.02 mm; a probe check is usually enough.