Spindle runout control: when to check and what to consider normal

What spindle runout means in practice

Runout isn't visible only on the dial indicator. It shows up quickly on the part: sizes begin to "wander", the surface loses finish, and the cutter leaves a trace that is hard to mistake for normal cutting. On a turning operation this usually looks like: the diameter was set correctly, the program didn't change, yet the actual size goes plus and minus.

Checking spindle runout isn't about a number in a log. It helps understand why the machine produces scrap even when cutting conditions are correct, the tool is new, and the program is written properly. If the axis of rotation is uneven, the tool removes slightly different amounts of material on each revolution. That causes waves, tool marks and matte patches instead of a smooth surface.

Often you can hear the problem before a measurement confirms it. The machine starts to feel harsher, an extra hum or a slight shake appears that wasn't there before. Sometimes the operator notices the cutter "nicking" the metal at regular intervals. That's already more like a repeating deviation than a random glitch.

A one-off defect and a persistent fault behave differently. If one part is bad after a insert change, a hit on the workpiece or a poor clamp, and the next part is fine, the cause may not be the spindle. But if the mark on the surface repeats from part to part, sizes follow a similar pattern, and vibration persists at the same settings, that's a systemic sign.

Shops usually notice four things: unstable size, worse surface finish, a new rotating noise and repeatable cutter marks. For a shop that's not trivial. Even small runout accelerates tool wear, ruins fits and forces extra measurement time.

On CNC machines for metalworking, especially during finishing passes and on parts with tight tolerances, the problem becomes visible fastest. If a shaft's size still "wanders" after every correction, don't immediately blame the program. First check how the rotating assembly behaves under actual load.

When you shouldn't postpone a check

If the machine held size yesterday but the same batch is "floating" today, don't wait for scheduled maintenance. Runout rarely goes away by itself. The longer you run with it, the faster scrap, tool wear and risk of another impact increase.

Start by looking for obvious events. A tool hitting the workpiece, the jaws or the tooling is already a reason to check the spindle, even if the machine didn't trip and continued working. After an impact the misalignment can be very small, but enough to produce taper, surface waves or unstable size.

The same applies after replacing the chuck, collet, arbor or a long tool. A new component may be fine but sit slightly differently than the previous one. The greater the overhang, the more any small error shows. If quality dropped sharply after a change, check immediately rather than blaming a "difficult batch."

There is a less obvious signal: scrap appears on identical parts even though program, material and cutting modes didn't change. When the operator adjusts offsets more often and size still wanders, the cause may not be the cutter. In that case a runout check quickly shows where to look: spindle, tool or tooling.

Don't leave noise, heat or vibration "until the weekend." Uneven hum, rising temperature around the spindle, ripples on the machined surface, or streaks on a finishing pass are all reasons to stop and check basic things before the assembly gets extra load.

A separate case is transport, new installation and long idling. After a move the machine may have been hit in transit or lose installation precision. After downtime, dirt, oil residue or slight corrosion often remain on seating surfaces, in the chuck and on the taper. So perform a short check before the first serious batch after start-up.

If there was an impact, tooling changes, a sudden rise in scrap, a new noise or recent reinstallation after transport, it's worth spending 10–15 minutes on a measurement. That's cheaper than searching the whole chain later and ruining the next batch.

What to prepare before measuring

A precise measurement is often ruined not by the spindle itself but by small things: dirt on the taper, a long overhang or rushing when mounting the indicator. So checks start not with numbers but with preparation.

You need a simple kit: a dial indicator with a magnetic stand, a clean control arbor and a sheet for recording values. Record values by points immediately: at the spindle nose, closer to the middle of the arbor and at the end. Relying on memory usually causes mistakes.

Before measuring clean all contact surfaces. The spindle taper, spindle nose, arbor, chuck and seating areas must be free of chips, oily dirt and dried dust. Even a tiny particle can create a picture that looks like a failure while the assembly itself is fine.

It's useful to prepare two tooling setups for comparison: a short control arbor or short tool and the working tool with the usual shop overhang. The idea is simple. If runout on the short arbor is small but grows sharply on the long working tool, the cause is often in the tool, chuck or excessive overhang, not the spindle.

Don't take measurements "on the fly." Stop the machine safely, remove load, make sure the spindle isn't rotating, and decide the measurement sequence beforehand: where the indicator will be, which points to record and what you'll change for a control comparison. When the order is clear, there are fewer mistakes.

If several people service the machine, leave a simple recording form nearby. For CNC service work this helps a lot: one measurement rarely gives the whole picture, while several records over a week show whether the problem is growing.

Quick check sequence

You don't need a long procedure here. You need clear logic: check the spindle itself first, then move to the taper, the arbor and the tool. This makes it easier to identify where the error grows.

Before measuring remove chips, wipe seating surfaces and mount the indicator firmly. If the stand moves or the probe is at a bad angle, the numbers will contradict each other.

1. Place the indicator at the spindle nose and record a baseline value. This is the starting point.
2. Rotate the spindle by hand through a full revolution. Turn smoothly, without jerks, to avoid false needle jumps.
3. Move to the spindle taper. If the nose is calm but the taper shows higher values, check for dirt, burrs or taper wear.
4. Then check the arbor and finally the tool tip. The further from the spindle, the more any small error shows.
5. Remove the tooling, reinstall it and repeat the measurement. If the picture changes after reinstallation, the cause is often the seating, the arbor or the tool.

Don't fixate on one number. Look at the whole chain of measurements. It's useful to record minimum and maximum at each point: spindle nose, taper, arbor and tool tip.

A simple example. If the spread at the spindle nose is small, the taper value barely changes, but the tool tip shows a sharp deviation, the problem is most often in the tool or tooling. If the spread starts at the nose, don't postpone checking the spindle assembly.

How to separate spindle problems from tooling and tools

One measurement at the tool tip proves almost nothing. Large runout there is often caused by long overhang, dirty seating, a worn collet or the arbor itself. So look not at guesses but at comparisons across several points.

Where to start

Check runout as close to the spindle nose as possible or use a control arbor with small overhang. If the nose value is small but the tip value rises sharply, first evaluate overhang and assembly stiffness. A thin, long tool easily amplifies a small misalignment.

Then rotate the arbor in the spindle and measure again. If the maximum point moves with the arbor, the cause is usually the arbor or tool. If the maximum stays in the same position, suspect the spindle assembly.

The logic with the collet is the same. Fit a different collet, clean the seating, retighten and see if the picture changes — if it does, deal with the clamping, nut and contact cleanliness. A single chip or a tiny burr easily creates runout that's mistaken for a machine failure.

What to compare

Quick swaps on the same operation work best: another arbor with the same tool, less overhang, a different collet or chuck, measurements with the spindle bare and with a control arbor. This approach quickly isolates the source.

If the operation "floats" only with one arbor, the spindle is usually not to blame. If different tools and holders give a similar picture and runout is visible on the bare spindle, then check the spindle unit itself.

In practice this saves time. Shops often blame bearings immediately while the real cause is simpler: too much overhang or a bad collet fit. But when runout repeats across assemblies and doesn't depend on tooling, don't delay service.

What to consider normal without guessing

Don't take a norm from someone else's experience or a foreman's memory. First open the machine manual and check the spindle tolerance specified by the manufacturer. Then compare it with the part tolerance. If the part requires a very precise fit, even a "permissible" manual value may be too large for the job.

For runout control it's useful to look at several points, not just one number. A measurement at the base shows the condition of the spindle assembly. A measurement at the working tool length shows the picture under real working conditions. These two values almost never match, and that's normal.

If the indicator at the base shows 0.002 mm but at the end of the long tool it's 0.015 mm, the cause is not necessarily the spindle. Dirt on the taper, misaligned arbor, worn collet or the tool itself often create the problem.

A practical guide is simple. First check the machine manual. Then compare the baseline and working-length measurements. After that reinstall the tooling and see if the result changes.

Repeatability tells more than a single measurement. If after reinstalling the arbor you see roughly the same small deviation every time, it's usually safe to work with that. Large jumps of 0.003, 0.01 and 0.02 mm between attempts usually point to dirt, play or a poor fit.

A good sign is when values hardly change after removing and reinstalling tooling. A bad sign is when readings "wander" without an obvious reason. Then don't argue about tolerances — search the chain: spindle taper, arbor, collet, tool.

If runout grows from the base toward the tool tip, inspect the assembly mechanics. Most often the culprits are play, dirt on seating surfaces or a crooked tooling setup. The spindle itself can also be the cause, but suspect it after excluding the other options.

For a shop this approach is usually enough: normal spindle runout is the value that fits the machine manual, the part tolerance, and repeats from check to check.

Measurement errors that spoil results

Even a precise indicator won't save you if you measure on a dirty base. A thin chip, an oily dust film or a small burr on the arbor shifts readings. Then people look for a spindle fault while the real culprit is ordinary debris.

A common mistake is measuring at a long overhang and immediately suspecting the spindle. The further the measurement point from the seating, the more arbor misalignment, weak tooling and clamping issues show. To see the spindle condition, measure near the base first, then compare with the overhang.

With the indicator it's easy to overdo it. If you press the probe too hard, a thin arbor or tool can deflect slightly and you'll measure not the runout but the response to the force. The probe should contact firmly but without excess pressure.

Another trap is taking a reading at one point and stopping. A proper spindle check requires several positions around the circumference and, if in doubt, measurements in two zones: at the seating and along the length. One quick hand turn, one glance at the dial and a rushed conclusion aren't enough.

A good habit is simple: clean the taper, arbor and indicator tip first, check for impact marks and burrs, then measure at the base and at a greater overhang, and repeat the check with a different arbor or tool. Compare multiple measurements, not just one.

The last mistake is more common than you think: a technician uses one tooling setup, takes one measurement and pronounces the spindle defective. It's easy to confuse a tooling defect with an arbor, chuck or tool fault. If a second tooling gives a different result, look to the clamping chain, not the spindle.

In the shop this saves time and money. One extra repeat check takes only minutes but often removes a false alarm and helps decide whether the machine needs service or a damaged tooling should be replaced.

Example from the shop

On the turning line a strange defect appeared: parts came out tapered even though no one changed the program and cutting modes stayed the same. The batch had been fine, then sizes began to drift from part to part.

The operator first suspected the spindle. That's a common reaction: when taper increases many immediately think of bearing wear or a serious machine fault.

The first measurement reinforced the suspicion. The indicator was placed on a long tool with big overhang and the needle showed a noticeable shift. In that context it's easy to conclude urgent repair is needed.

But the foreman didn't stop at one reading. He removed the tool and checked runout closer to the spindle nose with little overhang. There the deviation was small and didn't look like the cause of the taper on the part.

They followed a simple chain. They cleaned seating surfaces, swapped the arbor, and reduced the tool overhang as much as the operation allowed.

The picture changed immediately. The repeat measurement showed calmer values and a trial part returned to size. The taper disappeared without changing the program or touching the spindle.

This case shows why spindle runout checks must not be done at random. If you measure only on a long tool it's easy to confuse a spindle problem with an arbor contamination, seating dirt or excessive overhang.

After that the shop adopted a simple rule: check at the spindle nose first, then on the arbor, and only then on the tool at working overhang. This order saves time and avoids sending the machine to service unnecessarily.

Short checklist and next steps

If runout checks show extra microns, don't rush to blame the spindle. The error often hides in a dirty seating, worn arbor or too much tool overhang. One calm repeat measurement is usually more useful than a hasty conclusion.

First clean the taper, spindle face, arbor and all base surfaces. Then check the spindle without tooling, measure the arbor and then the assembled tool. This shows at which step extra runout appears. Record not only numbers but conditions: measurement point, overhang, arbor type, date and the person who checked. Compare results with both the machine manual and the part tolerance. The machine may still meet its own norm while the part does not.

Keep a simple history of measurements. If the bare spindle is fine today but numbers grow a month later, you'll see a trend rather than a random event. This way wear is noticed before it turns into stable scrap.

When to call service

Call service if values wander from check to check, runout increases after warming up, a clean arbor produces one result and another immediately shifts the numbers. Another bad sign is when the indicator shows acceptable numbers yet the part still comes out rough or out of size.

For EAST CNC machines it makes sense to start diagnostics with the spindle unit, seating and tooling rather than guesses. EAST CNC provides a full cycle of work: from selection and supply to commissioning and service, so if you already have measurement records the check will be faster and more focused.