This article is part of the CNC Tolerances & Quality Control Guide: GD&T, CMM, Cpk, PPAP on CNX Precision.
The choice between concentricity vs runout decides how a shop inspects your rotating features and what they cost. Both GD&T controls relate a cylinder to an axis, yet they behave very differently on the shop floor. Concentricity controls the median points of a derived axis and is notoriously difficult to measure. Runout controls surface variation while the part rotates and can be checked with a simple dial indicator. This guide defines each control, explains how each one is measured, and shows when to call each one out on a CNC drawing.
What Concentricity Controls
Concentricity demands that the median points of a controlled cylindrical surface fall within a tolerance zone centered on a datum axis. The tolerance zone itself is a cylinder, and every median point of the feature must sit inside it. Note what the control does not touch. It ignores the actual surface. A part can carry wobble, lobing, or other form error and still pass, as long as the midpoints of opposed cross sections land on the datum axis.
That indirect definition is the source of every practical problem with the control. The axis is a derived feature, computed from measured points rather than touched directly. The inspector must scan multiple cross sections, find opposed points, average them into medians, and then evaluate where those medians sit. Most designers discover this complexity only after the first inspection report comes back slow and expensive. ASME Y14.5 still defines concentricity, but the standard gives designers better tools for almost every real requirement.
What Runout Controls
Runout limits how much a surface moves while the part rotates about a datum axis. Circular runout checks each cross section independently, capturing eccentricity and roundness error in that section. Total runout sweeps the full length of the feature and also captures taper, straightness error, and axial wobble. The control acts directly on the surface the part actually uses, which is why it matches function so well on shafts, bearing seats, and sealing diameters. Designers favor it because the number they write on the drawing matches what the assembly experiences in operation.
Because runout folds form error into the result, it controls more than location alone. That is exactly what rotating assemblies need. A bearing does not care where the median points of a seat theoretically sit. It cares how much the contacting surface deviates during every revolution. The total indicator reading gives the inspector one number that maps straight to vibration, noise, and service life of the assembly.
How Each Control Is Measured
Runout measurement is refreshingly simple. The part rides on a spindle, vee blocks, or rollers that establish the datum axis, and a dial indicator touches the controlled surface. The inspector rotates the part one full turn and records the indicator swing. Circular runout repeats this at several cross sections, while total runout sweeps the indicator along the surface during rotation. Any capable inspection room performs these checks in minutes, and many CNC shops run them right at the machine. No software model stands between the part and the number, which keeps results easy to trust and easy to repeat.
Concentricity measurement works the other way. A coordinate measuring machine probes several cross sections of the feature, then computes median points from opposed readings and derives an axis from them. The result depends on sampling strategy, probe count, and evaluation software, so two labs can report different numbers for the same part. In the concentricity vs runout debate, measurement effort is usually the deciding argument before any technical difference enters the room.
There is a subtlety worth knowing. Runout results shift with the chosen datum setup, which makes datum selection a design decision rather than an afterthought. Concentricity results shift with the evaluation algorithm. One uncertainty comes from fixturing and is easy to agree on. The other lives in software and is hard to dispute.
Which Control Belongs on Your Drawing
Default to runout for rotating parts. Bearing seats, seal journals, couplings, and threaded features that must spin true all behave exactly as runout describes. The concentricity vs runout question resolves itself once you ask what the part does in service: if the surface contacts something during rotation, control the surface.
Reserve concentricity for the rare case where wall thickness symmetry truly matters and form error is either irrelevant or controlled elsewhere. Even then, many engineers reach for position of a diameter instead. Position defines a clear tolerance zone, measures directly from coordinates, and usually costs less to verify.
Whichever control you choose, keep the value realistic for the process. Runout callouts tighten with part diameter, length, and the number of setups involved. If you are unsure what your geometry allows, send the drawing to CNX Precision before final release. We will point out callouts that add cost without adding function and suggest values our machines hold all day.
Cost and Capability Considerations
Runout keeps cost down because it verifies fast. A dial indicator check takes seconds, supports in-process inspection at the lathe, and needs no expensive programming. Concentricity pushes cost up in the opposite direction. It requires CMM time, careful sampling plans, and a data evaluation that inspectors must document. For a part that runs in the thousands, that difference repeats every batch. Fast verification also simplifies first-article approval, because a customer can witness the same indicator check on the shop floor.
Capability follows the same split. Skilled turning centers hold tight runout routinely because the same rotation that defines the control also creates the feature. Concentricity adds no capability benefit, only verification overhead, and it invites disputes when measured values drift between labs. That is why the concentricity vs runout decision is as much a procurement decision as a design one. Clearer controls bring faster quotes, faster first articles, and fewer argument cycles over borderline parts.
Concentricity vs Runout: Common Questions
Why is concentricity so hard to measure?
Because it controls computed median points of a derived axis rather than a real surface. Inspectors must probe multiple cross sections, derive medians from opposed points, and evaluate them in software. Every step adds sampling uncertainty, which is why runout replaced concentricity in most industries.
Can runout replace concentricity on most parts?
For rotating parts, almost always. Runout controls the actual surface variation that bearings, seals, and couplings experience, and it measures in seconds with a dial indicator. That is why the concentricity vs runout debate usually ends with runout on the drawing.
Should we still specify concentricity in new designs?
Only when wall thickness symmetry around a datum axis is the true requirement and form error is handled separately. Many teams use position of a diameter instead, because it defines a clear tolerance zone and costs less to verify.