# Circular runout

Interpreting datum-related rotational variation at individual circular elements and distinguishing it from roundness and total runout.

Reviewed: 2026-11-01

Tags: parts-and-design, specifications, metrology

Evaluate radial circular runout at each specified circular section about the established datum axis. The datum setup affects the result.

An ideal coaxial taper with radius increasing from 10.000 to 10.030 mm has **zero circular runout at each station**. Its whole-surface radial range is **0.030 mm**. Keep sectional and whole-surface results separate.

KEYENCE's introductory example applies a **0.03 mm** radial circular-runout limit to any measurement plane perpendicular to the datum axis during one revolution. Its total-runout example instead covers the entire cylindrical surface. These are explanatory examples, not recommended tolerances for a shaft, seal journal or bearing seat. [KEYENCE, Run-out Tolerance](https://www.keyence.com/ss/products/measure-sys/gd-and-t/type/run-out-tolerance.jsp)

## Establish the governing specification first

Identify the drawing/model's governing system and edition before interpreting its symbols, datum indications and other requirements. ASME's current public listing identifies **Y14.5-2018 (R2024)** and says it replaces Y14.5-2009. ISO's listing identifies **ISO 1101:2017**, edition 4, confirmed in 2022, covering the symbol language and interpretation rules for geometrical specifications. Their public summaries do not establish that every detailed rule is interchangeable. [ASME Y14.5 listing](https://www.asme.org/codes-standards/find-codes-standards/y14-5-dimensioning-tolerancing/2018), [ISO 1101:2017 listing](https://www.iso.org/standard/66777.html)

Retain the characteristic symbol, tolerance value, applicable surface/extent and complete datum-reference sequence. The controlled feature and datum feature are different roles. If the referenced axis cannot be established unambiguously from the product definition, resolve that ambiguity before assigning an acceptance result.

## Circular runout, circularity and total runout

PTC's **Creo Elements/Direct 20.8** GD&T documentation describes circular runout using full indicator movement on an independent circular element over 360°. Its circularity page instead defines a zone between two concentric circles without a datum reference. Its total-runout page extends the indicator evaluation across the entire feature. These are vendor explanations of the characteristics, not replacements for the governing standard. [PTC, Circular runout](https://support.ptc.com/help/creo_elements_direct/r20.8.0.0/advanced_documentation/GDT_advisor/circular-runout.html), [Circularity](https://support.ptc.com/help/creo_elements_direct/r20.8.0.0/advanced_documentation/GDT_advisor/circularity.html), [Total runout](https://support.ptc.com/help/creo_elements_direct/r20.8.0.0/advanced_documentation/GDT_advisor/total-runout.html)

| Characteristic | Reference and extent | Important distinction |
| --- | --- | --- |
| Circularity | Individual circular element, no datum | Form alone does not locate the element relative to a functional axis |
| Circular runout | Individual circular element relative to datum-derived rotation | Includes the combined effect of surface form and its relationship to the datum axis |
| Total runout | Whole indicated surface relative to datum-derived rotation | Also constrains variation between sections within the controlled surface |

Do not interpret radial circular runout as an independent diameter tolerance. It does not by itself specify the required journal size. Nor should total runout be reported by separately zeroing each section and taking only the largest local circular-runout value: that discards differences between sections.

Mitutoyo's measurement guide also distinguishes radial and axial circular runout. A radial indication checks variation toward/away from the axis; an axial indication acts parallel to the datum axis. State the contact direction and surface being evaluated rather than assuming that “runout” always means a radial journal check. [Mitutoyo, Quick Guide to Measurement, p. 61](https://www2.mitutoyo.co.jp/eng/useful/E11003/pageindices/index63.html)

## The datum setup is part of the measurement

A datum is theoretically exact; the physical datum feature has real form variation. PTC distinguishes a datum from the datum-feature simulator, such as a suitable mandrel or gauge surface contacting the feature to establish the simulated datum. Therefore, the spindle axis, chuck axis and drawing datum axis are not automatically the same thing. [PTC, GD&T glossary](https://support.ptc.com/help/creo_elements_direct/r20.8.0.0/advanced_documentation/GDT_advisor/glossary.html)

Choose support and alignment that realize the specified datum system and agreed inspection method. Check for contamination, burrs, seating errors, clamping distortion and unintended axial movement. Do not center the controlled journal independently until its runout disappears if doing so replaces the required datum relationship.

For computational measurements, preserve the datum construction, sampled data and evaluation settings needed to understand the result. A best-fit circle of the measured section can be useful for circularity, but substituting that section's fitted center for the specified datum axis changes the circular-runout question.

## Indicator evaluation and sampling

For a simple radial indicator setup, keep the probe at a defined axial station, establish stable contact in the intended direction and observe a complete revolution. The indicated range is `maximum reading − minimum reading`. Zero selection does not change that range. Keep indicator resolution, calibration, contact force, stylus alignment and setup stability appropriate to the tolerance.

Check enough sections and angular coverage for the actual feature. A finite sampling plan provides evidence at its sampled locations; it does not automatically reveal every local defect on the surface. For axial face runout, identify the evaluated track/radius and any requirements for coverage of the face.

Document the angular/axial sampling and any filtering or data rejection. A smooth plotted trace or a small displayed value can result from evaluation settings that omit relevant variation. Investigate anomalous spikes before classifying them as dirt, damage or removable noise.

## Original calculations and their limits

**Ideal eccentric circle:** assume a perfectly circular radial section of radius 10 mm whose center is offset **0.008 mm** from the datum axis, with no tilt, form error or setup error. Its maximum and minimum radial distances from that axis are **10.008 mm** and **9.992 mm**. The ideal circular runout is therefore `10.008 − 9.992 = 0.016 mm`, twice the center offset.

Against an illustrative 0.020 mm tolerance, that geometry is within the numerical limit by 0.004 mm. Real measurements also contain form and setup effects, so dividing every runout result by two does not generally recover a true center offset. The numerical margin is not a measurement-uncertainty allowance or a complete conformance decision.

**Perfect coaxial taper:** assume a surface has perfectly circular sections centered on the datum axis, with radius increasing from **10.000 to 10.030 mm** along its length. At any fixed station, rotation produces zero ideal radial variation, so local circular runout is zero. Across the whole surface, the radial range is **0.030 mm**. This shows why circular runout alone does not control that between-section taper; size and other applicable requirements still need evaluation.

These examples are original idealized geometry calculations, not measured part results or tolerance recommendations.

## Reporting and acceptance

A useful report records the part/revision, controlled surface, tolerance, datum realization, measurement direction, section locations, method and observed ranges. Keep section-specific values when they matter instead of reporting an unexplained “TIR.” The acronym alone does not identify whether the evaluation was circular or total runout.

Apply the agreed decision rule with the measurement uncertainty where required. Instrument resolution is only one contributor: datum simulation, spindle error, probe alignment, repeatability, temperature and part deformation can affect the result. If a value changes materially after reseating, address the setup contribution before treating the change as a part correction.

Select the characteristic for function. A datum-related seal or rotating interface may need runout control; local roundness alone answers a different question. Conversely, avoid imposing total runout merely because it sounds stricter: verify whether the added whole-surface constraint is required and whether it is inspectable.

## Source editions

ASME and ISO edition details above come from their public listings, not reproduced normative clauses. PTC links are explicitly versioned 20.8. The cited KEYENCE and Mitutoyo explanatory pages do not state a technical revision date; their copyright or catalog identifiers are not treated as editions of ASME or ISO rules.

## Related reference

[Dimensions, tolerances and specifications](index.md)