Runout Requirements for CNC Turned Parts: Design and Inspection Guide

Author：admin Views：61 Source：This Site Time：2026-10-16

Summary：

Runout is one of the most common callouts on a turned part drawing, and one of the most frequently misunderstood at the RFQ stage. If you are buying a shaft, spool, bushing, or connector body, the practical question is simple: runout tells you how much the surface wanders relative to a datum axis while the part is rotated, and it only means something when that axis matches the way the part actually turns in service. This guide covers circular versus total runout, the datum axis, setup, and inspection.

Why functional rotation comes before the callout

Runout is a control that only exists in relation to rotation. Both circular and total runout are measured with the part rotated 360° around a datum axis, and the tolerance is the total variation of the surface as it turns. That means the callout is meaningful only when the axis you rotate about is the same axis the part will rotate about in the finished assembly.

Consider a bearing journal on a motor shaft. In service, the journal turns inside a bearing, so the bearing seat is the natural rotation reference. If the drawing instead references a feature that happens to be easier to machine, you can pass inspection and still get vibration in the field, because you measured wobble about the wrong axis. The designer’s job is to ask, “What actually turns, and about what?” and then choose the datum that represents that functional rotation.

This is the difference between runout and a purely form-based control. Runout is a control of how much one feature can vary with respect to another feature (the datum) when the part is rotated around the datum axis — in plain terms, the amount of “wobble” in a feature as it rotates. Circularity, by contrast, controls only form and is independent of any datum.

Datum axis and how it is established

The datum axis is not a physical line drawn on the part; it is a theoretical axis derived from a datum feature. When a cylindrical datum feature is referenced, the datum axis is established by engaging all the surface elements of that feature — effectively the axis of the smallest circumscribed (for an external feature) or largest inscribed (for an internal feature) cylinder that contacts the high points of the datum surface.

In practice the datum axis is realized on the shop floor by the fixturing used to hold the part during inspection: a collet, a chuck, V-blocks, or bench centers, after which the part is rotated around this axis while variation is measured. for a buyer is that the inspection setup is part of the specification. Two inspectors using different fixtures can produce different numbers on the same part if the datum feature is not cleanly established.

That is also why very tight runout on a datum feature that is short, interrupted, or poorly formed is risky. A short datum establishes the axis over a small contact band, so small surface imperfections translate into larger axis tilt. If your function demands a tight runout, make sure the datum feature is long and clean enough to define the axis reliably.

Circular versus total runout

These two callouts look similar but control different things, and confusing them is a common source of cost and dispute.

Circular runout is a two-dimensional control. It is evaluated one cross section at a time: the part is rotated a full turn about the datum axis, the indicator reading is taken at that single axial location, and the total variation of the surface elements at that section must fall within the tolerance. Each cross section is inspected independently, with no comparison between different cross sections. The tolerance zone is two concentric circles, coaxial to the datum axis, lying on a plane perpendicular to the datum axis, and the tolerance value is the radial distance between them.

Total runout is the three-dimensional extension of the same idea. The indicator is swept along the surface while the part rotates, so the entire surface — not just one cross section — must fall inside the tolerance zone. The total runout tolerance zone is two concentric cylinders, coaxial to the datum axis, and all surface elements must fall within that zone; the tolerance value is the radial distance between the two cylinders. Because it simultaneously limits location, orientation, and form of the whole surface, total runout is a stricter control than circular runout. If the size tolerance is tighter than the runout tolerance, size controls the form of the feature.

The buyer-facing rule of thumb: specify circular runout when you care about wobble at discrete planes (for example, two bearing seats that must each sit true), and specify total runout when the whole surface must stay coaxial (for example, a long sealing journal or a precision spigot). Total runout is harder to hold and costs more, so do not upgrade the callout without a functional reason.

Manufacturing setup and where runout comes from

On a CNC lathe or Swiss-type machine, runout is rarely a single error — it is the sum of several setup and process conditions. Understanding them helps you predict capability before you commit to a tolerance.

Workholding and chucking. If the part is held on a surface that is not concentric with the functional axis, every feature machined from that setup inherits the offset. Re-chucking between operations is a common source of “surprise” runout on turned parts.
Spindle and tool-holder condition. Machine-tool spindle error and tool-holder runout add directly to the part’s measured runout. A worn collet, a damaged taper, or chips on a locating face can each contribute a few microns, or far more.
Datum quality from earlier operations. The datum feature itself was machined earlier; if it was turned against a poor reference, the part carries that error forward. Establish the datum early and protect it through subsequent operations.
Thermal and clamping distortion. Thin walls and slender shafts can deflect under clamping or grow with heat, so runout measured in a free state may differ from runout in the fixture.
Material stability. Stress relief and heat treatment can move a part after machining, changing runout later. Sequence and fixturing must account for this on parts with tight coaxiality needs.

Practically, the machinist controls runout by keeping the functional axis coincident with the workholding axis — turning critical diameters in one setup, or using a qualified datum and soft jaws — and by verifying setup with an indicator before running the batch. Runout capability is therefore a function of process planning, not just machine accuracy.

Inspection method for runout

Runout is one of the few GD&T controls that can be verified with simple shop-floor instrumentation, and that is a real advantage for both maker and buyer.

Establish the datum axis. Mount or support the datum feature so the part rotates about the derived datum axis — using a collet, chuck, V-blocks, or bench centers, as appropriate to the datum feature’s form.
Zero the indicator. For a drop or dial indicator, ensure the contact is normal to the surface and perpendicular to the axis of rotation. Position the stem so readings are taken radially.
Rotate a full revolution. For circular runout, rotate the part one complete turn at a single axial position and record the total indicator variation (maximum minus minimum reading). Re-zero the indicator before moving to a new axial position, so one cross section is never compared with another.
Sweep along the axis for total runout. For total runout, rotate the part while translating the indicator along the surface, capturing the maximum and minimum deviations of the entire surface. The total spread must not exceed the callout.
Compare to tolerance and record. If the gauge variation does not exceed the runout tolerance, the feature is in spec at that location or over that surface.

For higher volumes or complex geometry, a CMM can sample many points around the axis and reconstruct both circular and total runout from the point cloud, reducing operator dependence and producing a traceable record. Whatever the method, the measurement is only as valid as the established datum axis.

Specifying runout on a drawing: practical guidance

A well-specified runout callout answers four questions without ambiguity: which surface, which datum, which type (circular or total), and what value.

Reference a functional datum. Choose the feature that represents the part’s real rotation, and make it long enough to establish a stable axis.
Match the type to the function. Use circular runout for plane-specific wobble and total runout for whole-surface coaxiality. Do not default to total runout everywhere.
Be realistic about capability. Runout is a cumulative error that includes machine, tooling, and workholding contributions. Very tight values on a long or slender turned part may require special processes rather than a standard lathe setup.
State the inspection conditions. Note whether runout is verified in a fixture, in free state, or on a CMM, because clamping and thermal state affect the result.

Frequently asked questions

Is runout the same as concentricity? No. Runout controls surface variation during rotation and is measured directly; concentricity is a more complex derived-axis control. Runout is simpler to inspect and better aligned with functional rotation.

Can circular runout be looser than total runout? They are independent callouts with different tolerance zones. Circular runout limits each cross section; total runout limits the entire surface. Where both appear, each must be satisfied.

Why does runout change after heat treatment? Heat treatment and stress relief can redistribute residual stresses and move the part, so a value measured before treatment may not hold after.

Does runout control size? No. Runout is a geometric control relative to a datum axis; size is controlled separately by the dimension’s limits.

Summary

Runout is a rotation-based control: circular runout limits the variation of individual cross sections about a datum axis, while total runout limits the variation of the entire surface with two coaxial cylinders. Both depend on a well-chosen datum that reflects the part’s real rotation, and both are shaped by workholding, tooling condition, and the quality of the datum feature established in earlier operations. Inspection is straightforward with indicators and fixtures, or CMM sampling for higher confidence. Specify the datum around functional rotation, match circular versus total runout to the job, and agree on inspection conditions, and runout becomes a reliable, verifiable requirement.