13.2 Datum-Axis Simulation and Rotating Parts
Key Takeaways
- Runout is measured by rotating the part about a simulated datum axis, so the quality of datum-axis simulation directly governs the repeatability and meaning of every runout reading
- When two coaxial datum features (e.g. two bearing journals) are referenced, the datum axis is the common axis established by simulating both features simultaneously - not the axis of either feature alone
- Common setups include between centers, V-block, collet/chuck, and precision arbor; each injects its own error (center-hole form, V-block 60-degree averaging, chuck runout, arbor-to-feature fit) into the reading
- A poor datum feature - out-of-round, worn, undersized - inflates the measured runout of the controlled feature because its errors appear in the rotation, falsely rejecting good parts
- Per ASME Y14.5-2009, datum features used for runout should themselves be functional, accessible, and of sufficient size; a small or low-quality datum feature makes runout readings non-repeatable
13.2 Datum-Axis Simulation and Rotating Parts
Quick Answer: Runout is measured by rotating the part about a simulated datum axis and reading an indicator on the controlled surface. The physical setup that simulates that axis - between centers, V-block, collet, chuck, or arbor - injects its own error into the reading. The Senior-level trap: a worn, out-of-round, or undersized datum feature makes the controlled feature appear to have more runout than it really does, falsely rejecting good parts.
Why Datum-Axis Simulation Is the Whole Game
A runout tolerance is defined against a datum axis, but in practice the inspector does not have an abstract axis - they have a physical setup that approximates one. The simulated datum axis is the axis about which the part actually rotates during inspection. If that simulated axis differs from the true datum axis defined by the drawing, the difference shows up as indicator movement and is indistinguishable from real runout of the controlled feature.
This is why ASME Y14.5-2009 stresses that datum feature quality is critical for runout. The datum feature is the physical surface the setup contacts; its form errors, wear, and size all enter the rotation. A runout reading is only as good as the datum feature that produces it.
Single vs Common Datum Axis
When a runout control references a single cylindrical datum feature (e.g. a bore labeled A), the datum axis is the axis of that feature's true geometric counterpart - the axis of a perfect cylinder that contacts the feature. The part is typically mounted on an arbor or in a collet that simulates A.
When the control references two coaxial datum features (e.g. two bearing journals labeled A and B, as A-B), the datum axis is the common axis established by simultaneously simulating both features. The standard setup is to place each journal on a V-block or roll-rest, or to turn the part between centers if the journals have center holes. The common axis is NOT the axis of A alone or B alone - it is the axis defined by the pair, which averages out some individual feature error but is sensitive to the geometry of both.
Common Setup Methods
| Setup | How the datum axis is simulated | Typical error introduced | Best for |
|---|---|---|---|
| Between centers | Lathe centers in center holes at each end | Center-hole roundness and angle; worn centers | Shaft-shaped parts with machined centers |
| V-block / roll-rest | Two V-blocks contacting two journals | 60-degree V averages roundness; sensitive to part size | Finished journals, inspection room |
| Collet / chuck | Grips one datum feature directly | Chuck jaw imbalance; collet runout; gripping distortion | Parts with one dominant datum feature |
| Precision arbor | Arbor pressed or slipped through a bore datum | Arbor-to-bore fit clearance; arbor runout | Pulleys, gears, ring-like parts |
| Surface plate + rolls | Part rests on rolls; indicator on a stand | Roll roundness; plate flatness; setup geometry | Large or heavy rotating parts |
The Trap: Poor Datum Feature Inflates Runout
Consider a shaft with two bearing journals (A and B) that establish the datum axis, and a mid-shaft gear seat controlled by 0.05 circular runout to A-B. Suppose journal A is worn and out-of-round by 0.04.
When the part is set up on two V-blocks simulating A and B, the worn A journal causes the part's rotation to wobble by roughly its own form error. If the gear seat is actually perfect (zero runout to the true A-B common axis), the indicator still reads about 0.04 FIM simply because the simulated axis is wobbling. A part that should pass at 0.05 reads 0.04 - it passes, but barely. If the wear were 0.06, the same perfect gear seat reads 0.06 and fails - a false reject.
This is the core Senior trap. The measured runout of the controlled feature equals the real runout of that feature plus the contribution of datum-feature error entering through the setup. You cannot measure runout more accurately than the datum feature permits.
Remedies in Practice
- Use the functional datum - the surfaces that actually locate the part in assembly (e.g. the bearing seats) are usually the best datum features because their error directly represents service behavior.
- Specify datum feature controls - put a cylindricity or circularity callout on the datum feature itself so its form error is bounded.
- Pick a larger datum feature when possible - a larger feature averages out local form error and gives a more repeatable axis.
- Restore the datum feature before re-inspection if it is worn.
What the Senior Exam Tests Here
Questions typically pair a drawing with a runout callout and a description of the setup, then ask what is wrong or what the reading represents. The answer almost always involves one of:
- The datum feature is worn/out-of-round, inflating the reading (false reject)
- The setup uses the wrong feature as the datum (e.g. gripping a non-datum surface)
- A single datum feature is used where the callout requires a common axis from two features
- The simulated axis does not match the datum reference frame on the drawing
Recognize the pattern: a runout reading is only credible if the datum axis is simulated from the features the drawing names, and those features are of controlled quality.
A shaft's mid-shaft gear seat is controlled by 0.05 circular runout to datum axis A-B (two bearing journals). The gear seat is actually perfect, but journal A is worn out-of-round by 0.04. When inspected on two V-blocks simulating A and B, what reading is most likely and what is the result?
A drawing applies a runout tolerance referenced to datum axis A-B, where A and B are two coaxial bearing journals. Which setup correctly simulates the required datum axis?
Which practice most directly reduces the risk that worn datum features inflate runout readings and cause false rejects during inspection?