Precision Lathe Work Tolerances: How to Read Runout on Turned Shafts Before Your Bearing Won't Fit?

Precision Lathe Work Tolerances: How to Read Runout on Turned Shafts Before Your Bearing Won't Fit?

Precision Lathe Work Tolerances: How to Read Runout on Turned Shafts Before Your Bearing Won't Fit?

a multi-step shaft on a CNC lathe with a dial indicator measuring runout at a bearing seat

You just finished machining a shaft. The diameter checks out perfectly. But then you go to press the bearing on — and it binds, tilts, or simply refuses to seat. Sound familiar? This is one of the most common and costly surprises in precision machining. The part looks right on paper, but it still fails at assembly. The culprit, almost every time, is a misunderstood or missing runout tolerance on the drawing.

Quick Answer: What Are Precision Lathe Work Tolerances for Runout?

Runout tolerance controls how much a turned surface deviates from a perfect axis of rotation. It is not the same as diameter tolerance. Even a perfectly sized shaft can fail assembly if its surface wobbles, tilts, or shifts away from the datum axis.

Tolerance Type What It Controls Typical Value (Precision)
Circular Runout Each cross-section, one at a time 0.005 – 0.015 mm
Total Runout Entire surface (all cross-sections + straightness) 0.008 – 0.020 mm
Bearing Seat Diameter Size of the journal Per bearing class (e.g., js5, k5)
Shaft Concentricity Axis-to-axis offset between steps 0.005 – 0.010 mm

Values shown are common ranges for precision CNC turning. Always verify against your bearing manufacturer's fit table.

So if diameter tolerance is not the whole story, what is? The answer lies in geometry — specifically, how the surface behaves around the axis, not just at a single measured point. To truly understand why your bearing won't seat, you need to understand runout: what it is, how it is called out on a drawing, and how it differs from simple size control. Let's break it down step by step.


Table of Contents

  1. What Does Runout Actually Control on a Precision Lathe Drawing?
  2. Circular Runout vs Total Runout: Which One Does Your Shaft Actually Need?
  3. How Do You Keep Multiple Bearing Seats Coaxial on a Multi-Step Shaft?
  4. How Do You Measure Runout Correctly and Specify It Without Creating Impossible Requirements?
  5. Conclusion

What Does Runout Actually Control on a Precision Lathe Drawing? 

Every machinist has seen it: a small circle with an arrow on a drawing, pointing at a turned surface, with a tiny number like "0.010" beside it. That symbol is the GD&T runout callout — and it is telling you something that no diameter dimension can. It is telling you that the surface must stay within a specific band as the part rotates about a datum axis. Miss this, and your assembly will fight you.

What Does the Runout Symbol Mean?

  • The circle-with-arrow symbol = runout (circular or total)
  • The datum letter (e.g., "A") = the axis everything is measured from
  • The tolerance value = the full band the surface must stay within during one full 360° rotation
  • Runout includes both out-of-roundness and eccentricity in one control
  • Runout does not control size — diameter tolerance is a separate callout

In short: Runout asks, "Does this surface stay centred on the axis as it spins?" Diameter tolerance only asks, "Is this surface the right size?" You need both — but they answer completely different questions.

Layer 1 GD&T runout callout — precision turned shaft (ASME Y14.5-2018) A Ø40h6 Ø30h6 Tolerance zone t = 0.010 mm Tolerance zone t = 0.010 mm 0.010 A Feature control frame Dial indicator (TIR reading) 360° rotation Datum axis A (true axis) Controlled surface Bearing seat A (Datum feature) Bearing seat B (Controlled feature) Journal Ø22h6 Key Datum axis A — common reference axis for all runout callouts on this shaft Tolerance zone — annular band t = 0.010 mm (0.0004 in) width, centred on datum axis Shaft body — Ø40h6 (seat A / datum), Ø30h6 (seat B, controlled), Ø22h6 (journal) 0.010 A Feature control frame: circular runout symbol | tolerance value | datum reference Dial indicator probe — placed ⊥ to surface; TIR read over full 360° rotation Typical precision bearing journal runout: 0.005–0.020 mm. h6 tolerance on Ø30: 0 / −0.013 mm (ISO 286-1) Per ASME Y14.5-2018 / ISO 1101:2017 / ISO 286-1

Understanding GD&T runout for turned parts starts with recognising the difference between size and geometry. A shaft journal can be perfectly round, perfectly sized — and still be eccentric to the bearing bore axis. That eccentricity creates a tight spot on one side of the bearing race and a loose spot on the other. The result: premature wear, noise, heat, and eventual failure.

There are also two directions of runout to consider. Radial runout acts perpendicular to the axis — this is the classic wobble that throws a bearing off-centre. Axial runout acts parallel to the axis — this causes a face or shoulder to rock rather than sit flat, which matters enormously for thrust bearings and precise axial location. Both are controlled by the same symbol; the direction depends on which surface the callout arrow points at.

This is why relying on diameter tolerance alone is never enough. Diameter confirms the journal fits the bore in terms of size. Runout confirms the journal sits true in the bore in terms of geometry. A high-quality CNC turning machining service will always control both — but only when the drawing asks for it.


Circular Runout vs Total Runout: Which One Does Your Shaft Actually Need? 

This is the most common point of confusion on turned-part drawings. Engineers sometimes use circular runout when they need total runout — and vice versa. The difference seems small on paper, but it has a big impact on what your machinist must achieve and how your inspector must measure it.

Circular Runout vs Total Runout — Side-by-Side Comparison

Feature Circular Runout Total Runout
Scope One cross-section at a time Entire surface length
Catches out-of-round? ✅ Yes ✅ Yes
Catches eccentricity? ✅ Yes ✅ Yes
Catches taper / straightness? ❌ No ✅ Yes
Measurement method Single probe position, rotate 360° Probe sweeps full length while rotating
Tighter to achieve? Less demanding More demanding
GD&T symbol ↗ (single arrow circle) ↗↗ (double arrow circle)

Simple rule: Use circular runout when each zone of the shaft is independent. Use total runout when the whole surface must behave as one perfectly cylindrical zone around the datum axis.

Layer 1 Circular runout vs total runout — tolerance zone comparison (ASME Y14.5-2018) Circular runout ↗ Total runout ↗↗ Probe fixed at one position Part rotates 360° t = 0.008 mm Checks: out-of-round + eccentricity Does NOT catch: taper, bow, straightness error One independent band per cross-section 0.008 A = circular runout callout t = 0.015 mm Probe sweeps full length while part rotates Part rotates 360° + probe traverses axially Checks: out-of-round + eccentricity + taper + bow Controls the ENTIRE surface as one cylindrical zone All points on surface must fall within one band: t 0.015 A = total runout callout Circular runout Total runout Scope Single cross-section at a time Entire surface length (all cross-sections combined) Catches Out-of-round, eccentricity Out-of-round, eccentricity, taper, bow, straightness error Misses Taper, bow, overall straightness Nothing — most comprehensive surface control Typical value 0.005–0.015 mm (precision bearing journal) 0.008–0.020 mm (stricter if surface length > 1.5× dia.) When to use Short journal, taper not a concern Long surface, multi-step shaft, taper must be controlled Measurement Probe fixed axially, rotate 360°, read TIR Probe traverses full length while rotating, read max TIR ASME Y14.5-2018 / ISO 1101:2017

The circular runout vs total runout choice comes down to one question: does taper or bow in the surface matter for your assembly?

For a simple bearing journal where the bearing width is narrow and the journal is short, circular runout is usually enough. It confirms that each slice of the surface stays within its band. But if the journal is long, or if the bearing spans multiple shoulders, or if you are controlling lathe runout tolerance across a surface that also has taper concerns — then total runout is the correct call.

Total runout essentially creates a perfect cylindrical tolerance zone around the datum axis. Every point on the surface — across its full length and all 360° — must fit inside that zone. This is a stricter requirement and it is harder to achieve in production. However, for high-speed spindles, precision gearbox shafts, and industrial machinery drive components, it is the right control to use.

A practical guideline: whenever a drawing calls for precision CNC turning services tolerances on a surface longer than roughly 1.5× the diameter, consider whether total runout is actually what the function demands. If the surface needs to be truly cylindrical — not just round at each cross-section — write total runout on the drawing.


How Do You Keep Multiple Bearing Seats Coaxial on a Multi-Step Shaft? 

A shaft with one bearing seat is relatively straightforward. A shaft with two, three, or four bearing seats — each at a different diameter, separated by shoulders and undercuts — is where multi-step shaft coaxiality becomes genuinely difficult. Every machining setup, every re-chucking, and every tool change is an opportunity to introduce an axis offset between steps.

Top 5 Causes of Coaxiality Error on Multi-Step Shafts

  1. Re-chucking without a common datum reference — Every time you move the part, you re-introduce chuck runout and jaw seating error.
  2. Worn or inaccurate spindle bearings — Even a small amount of spindle runout correction left unaddressed multiplies across the shaft length.
  3. Soft-jaw bores not concentric with the spindle axis — Poorly bored soft jaws force eccentricity into every part they hold.
  4. Excessive stock removal in separate setups — Removing large amounts of material in op 20 springs the part away from its op 10 geometry.
  5. Datum confusion on the drawing — When the drawing references different datums for different steps, machinists cannot hold a single consistent axis.
Layer 1 Multi-step shaft — GD&T datum A + total runout callouts on bearing seats B and C (ASME Y14.5-2018) A Ø40h6 (0 / −0.016) Ø30h6 (0 / −0.013) Ø22h6 (0 / −0.013) Bearing seat A (Primary datum) Bearing seat B (Controlled — total runout) Bearing seat C (Controlled — total runout) 0.010 A Seat B total runout callout 0.012 A Seat C total runout callout t = 0.010 mm t = 0.012 mm A Datum axis A All three seats share datum axis A — coaxiality controlled by total runout callouts 360° rotation Why total runout — not circular runout — for a multi-step shaft? Total runout controls the entire surface of seats B and C as a single cylindrical zone relative to datum axis A. This catches taper and bow across the full journal length — critical when bearing width spans the whole surface. Circular runout would only check individual cross-sections and miss accumulated straightness error between steps. Feature Seat A (datum) Seat B (controlled) Seat C (controlled) Nominal dia. Ø40 mm Ø30 mm Ø22 mm Tolerance class h6 (0 / −0.016 mm) h6 (0 / −0.013 mm) h6 (0 / −0.013 mm) GD&T role Datum feature (reference axis) Controlled feature Controlled feature Runout type N/A (is the datum) Total runout ↗↗ Total runout ↗↗ Runout tol. to A 0.010 mm (0.0004 in) 0.012 mm (0.0005 in) Controls Sets axis for all callouts Eccentricity + taper + bow vs axis A Eccentricity + taper + bow vs axis A Coaxiality range Reference 0.005–0.010 mm (precision apps) 0.005–0.012 mm (precision apps) Measurement Mount on centres or precision chuck Probe sweeps full length — max TIR Probe sweeps full length — max TIR Process risk Re-chucking destroys coaxiality Machine all seats in one setup if possible Machine all seats in one setup if possible ASME Y14.5-2018 / ISO 1101:2017 / ISO 286-1

Controlling shaft concentricity tolerance across multiple steps starts on the drawing, not on the machine. The most important decision is datum selection. Choose one surface — typically the primary bearing seat, the centre bores, or the ground journal — and make everything else reference that same axis. When all runout callouts share a common datum, the machinist and inspector both have a single, unambiguous reference.

For bearing seat runout tolerance on precision shafts, many manufacturers work to values in the 0.005–0.015 mm range for individual journals. The coaxiality between two seats — meaning the axis offset from one journal to another — is often held to 0.005–0.010 mm in demanding applications such as automotive drivetrain and gearbox assemblies.

On the shop floor, the most reliable way to hold coaxiality is to machine all critical diameters in a single setup between centres. When that is not possible, ground soft-jaws bored at production pressure, combined with a proven CNC turning inspection services protocol to check each feature before the part leaves the machine, are the next best options. Re-chucking is the enemy of coaxiality — minimise it wherever your process allows.


How Do You Measure Runout Correctly and Specify It Without Creating Impossible Requirements? 

Even a perfectly machined shaft will fail inspection if runout is measured incorrectly. And even a perfectly written drawing will cause production headaches if the tolerance values are tighter than the process can realistically achieve. Both measurement discipline and sensible DFM (design for manufacturability) matter equally here.

How to Measure TIR (Total Indicator Reading) on a Turned Shaft — Step by Step

  1. Set up the datum correctly. Mount the part on the defined datum — live centres in the centre bores, a precision mandrel, or a ground steady-rest journal. Do not simply chuck the part and assume the chuck jaw is the datum.
  2. Zero the dial indicator. Place the probe perpendicular to the controlled surface. Use a high-resolution indicator (resolution ≤ 0.001 mm for precision work).
  3. Rotate the part one full 360°. Watch the full swing of the needle. The total movement — from lowest to highest reading — is the TIR.
  4. Record TIR, not radius. TIR is the full band, not a radius. A TIR of 0.010 mm means the surface is within a 0.010 mm wide annular zone around the datum axis.
  5. For total runout: Move the probe slowly along the full length of the surface while rotating the part. The maximum TIR across the entire sweep is the total runout value.
  6. Check multiple positions. For long journals, check at both ends and the middle.
Layer 1 TIR measurement — dial indicator on precision shaft between centres (ISO 1101 / ASME Y14.5-2018) Granite surface plate Headstock (live centre) Tailstock (dead centre) Magnetic base adj 0 5 10 5 0.001 mm 360° rotation TIR = 0.010 mm Max reading Min reading Position 1 Position 2 Position 3 Step A Ø40h6 Bearing journal Ø30h6 Step C Ø22h6 Dial indicator Resolution: 0.001 mm Ruby probe tip 60° centre Datum axis A TIR measurement procedure (per article Section 4.2 / ISO 1101:2017) 1 Mount shaft on datum — live centre (headstock) and dead centre (tailstock), 60° cone angle 2 Zero the dial indicator — place probe perpendicular (⊥) to surface, resolution ≤ 0.001 mm 3 Rotate shaft slowly one full 360° — observe full needle swing (do not stop early) 4 Record TIR = max reading − min reading (full band, not radius). TIR ≠ radius error 5 For total runout: sweep probe along full journal length while rotating — record max TIR across sweep 6 Check at 3 positions: both ends + midpoint. Precision target: TIR 0.005–0.020 mm for bearing journals ASME Y14.5-2018 / ISO 1101:2017 / ISO 866 (centre holes)

The most common drawing mistake is referencing the wrong datum. If the drawing calls runout back to a datum that the machinist cannot easily fixture from — such as a small undercut groove or an unreferenced theoretical axis — the callout becomes practically unmeasurable. Always reference runout to a surface that is both functional (the real mating surface in the assembly) and fixtureable (a surface the machinist can actually hold or reference in the measuring setup).

The second most common mistake is specifying tolerances that are tighter than necessary. A good guideline for working with a high precision lathe machining supplier: ask what the process naturally achieves before you write the drawing requirement. Most well-maintained CNC turning centres with live-tool capability hold circular runout in the 0.005–0.010 mm range reliably. Asking for 0.002 mm without a clear functional reason adds cost and lead time without adding reliability.

Finally, always consider spindle runout correction as part of your supplier qualification. A supplier who monitors spindle condition, performs regular compensation, and integrates precision CNC turning services with in-process gauging will consistently deliver parts that pass first-article inspection — rather than parts that need re-work or concession.


Conclusion

Contents of Conclusion

Runout tolerance is not an optional detail on a turned-part drawing — it is the control that determines whether your bearing seats, spins true, and survives its design life.

Here is what to take away from this guide:

  • Diameter tolerance and runout tolerance are separate controls. A part can be in size tolerance and still fail assembly due to runout.
  • Circular runout controls each cross-section independently. Use it when taper and bow are not concerns.
  • Total runout controls the entire surface. Use it when the function demands a truly cylindrical surface relative to the datum axis.
  • Multi-step shafts need a single, shared datum. Reference all bearing seat callouts to the same axis to avoid coaxiality errors across steps.
  • Measure from the correct datum, using TIR methodology. Probe placement, fixturing, and rotation all affect the validity of your runout measurement.
  • Specify tolerances that match both the function and the process. Tighter is not always better — work with your machining supplier to align drawing requirements with production capability.

A bearing that fits first time, runs quietly, and lasts its full service life is not luck — it is the result of a drawing that controls the right things, and a machining partner who understands how to hold them.

Whether you are designing a gearbox shaft, a spindle assembly, or a precision instrument component, getting runout right at the drawing stage is always faster and cheaper than sorting it out at assembly. When in doubt, ask your supplier early — and choose a partner with proven CNC turning capability and documented inspection processes.


External Links Recommendation

[Lathe runout tolerance][^1]
[Precision lathe work tolerances][^2]

[Circular runout vs total runout][^3]
[GD&T runout for turned parts][^4]

[Bearing seat runout tolerance][^5]
[Shaft concentricity tolerance][^3]

[^1]: Industry publication providing practical guidance on controlling runout in lathe operations, including achievable runout tolerances (0.0005" with collet chucks, 0.001-0.002" with three-jaw chucks) and workholding selection strategies


[^2]: Machinery supplier explaining precision lathe tolerance capabilities including Tsugami and Nexturn Swiss machines achieving ±0.005mm tolerances for medical, aerospace, and electronics applications with discussion of influencing factors.

[^3]: Academic textbook chapter explaining circular runout as a 2-dimensional control of circular elements (roundness + coaxiality) versus total runout as a 3-dimensional composite control of roundness, straightness, angularity, taper, coaxiality, and surface profile for complete surface control [citation:1].

[^4]: Technical comparison detailing circular runout (controls circularity + concentricity, measured at individual cross-sections) versus total runout (controls cylindricity + concentricity, measured across entire surface) with practical guidance on ease of manufacturing and inspection for turned components [citation:3].

[^5]: SKF bearing manufacturer specifications detailing required ISO tolerance grades for bearing seats: shaft seats require IT6 dimensional tolerance with IT5 total run-out tolerance; housing seats require IT7 dimensional tolerance with IT6 total run-out tolerance, with specific examples including a 6030 bearing calculation [citation:2].

[^6]: Peer-reviewed research confirming that geometrical characteristics including concentricity significantly affect rotating shaft vibration, establishing benchmarks for tolerance optimization to maintain component performance [citation:5].

Leave a comment

What are you looking for?