Guide
Roundness measurement: when is a micrometer or V-block enough, and when do you need a roundness instrument?
Roundness measurement is the check of how far a cross-section of a turned or ground part departs from a true circle, and only a radial method, a rotating spindle or table recording the profile around a fixed center, measures it as the drawing standards define it. A micrometer or a V-block with an indicator is a useful shop screen, but a micrometer cannot see odd-lobed shapes and a V-block reading must be interpreted for the lobe count, so a tight circularity callout needs an agreed radial method.
By Saola Manufacturing engineering team · Published · Updated
This guide is about the measuring method, not the tolerance. What circularity and cylindricity control, and when each belongs on a drawing, is covered in circularity vs cylindricity. Here the question is practical: a part arrives with a roundness callout, and the supplier and the receiving inspector have to agree on how to prove it. The three common methods give different numbers on the same part, and the difference is largest on exactly the shapes turning and centerless grinding tend to produce.
On this page
What are the three ways to measure roundness?
Two-point (a micrometer across the diameter), three-point (the part rotated in a V-block under an indicator) and radial (a precision spindle or rotary table that records the profile around a center).
Mitutoyo’s measurement glossary describes all three. The diameter method takes roundness as half the difference between the largest and smallest diameter readings; the three-point method reads the indicator deflection as the shaft turns in a V-block; and the radius method collects the deviation around the whole circumference on a roundness measuring machine. It notes that neither of the first two can reveal the actual cross-sectional shape (Mitutoyo, Evaluation Methods for Roundness).
The radial method is the one the standards are written around. ASME B89.3.1, Measurement of Out-of-Roundness, assesses roundness from the radial deviations of a cross-sectional profile measured from a defined center, and deals mainly with precision spindle instruments and polar chart presentation. ISO uses the same principle in ISO 12181.
| Method | What it reads | Blind spot | Good for |
|---|---|---|---|
| Two-point (micrometer, bore gauge) | Diameter; roundness taken as half of max minus min | Odd-lobed shapes, which have the same width in every direction | Ovality from clamping or bending; size checks |
| Three-point (V-block and indicator) | Indicator swing as the part rotates | Lobe counts the block angle is not matched to; reading needs a lobe-dependent factor | Shop screening of lobed ground or turned parts |
| Radial (roundness instrument) | Full profile from a reference axis, with a chosen filter and reference circle | Results change with filter and reference circle if they are not stated | Acceptance of tight circularity callouts; disputes |
| Coordinate scan (CMM) | Points around a section, fitted to a circle | Sparse point counts miss high-frequency lobes | Roundness on larger prismatic parts measured with other features |
Why can’t a micrometer detect a three-lobed part?
Because a shape with an odd number of lobes has a constant width: a high spot on one side always sits opposite a low spot, so the two anvils read the same diameter at every angle.
The Precision Machined Products Association explains the geometry for centerless-ground bar: such “pseudo-rounds” have odd-numbered crests, and the miked dimension is the center-to-apex radius plus the center-to-low-spot radius, which stays constant as the part is turned. A two-plane gauge will not reveal the condition (PMPA, 2010). Mitutoyo’s roundness instrument literature makes the same point: diameter measurement with a micrometer cannot detect an odd-number lobing condition (Mitutoyo RA-10 catalog).
On turned parts the usual source is clamping. A thin-walled sleeve held in a three-jaw chuck is cut round while clamped and springs back into a three-lobed form when released. Centerless grinding produces odd lobing from the geometry of the wheels and blade. In both cases the micrometer says the part is perfect, the bearing or seal says otherwise.
The micrometer is still the right tool for ovality, the two-lobed shape from bending or a part clamped between two points. It measures that directly, and roundness is half the spread between readings.
How do you read a V-block roundness check correctly?
Treat the indicator swing as a signal that has to be scaled for the block angle and the lobe count, not as the roundness value itself.
With the part resting in the V and the indicator on top, every lobe that passes under the indicator also rolls over the two contact faces, and the movements combine. Depending on the lobe count and block angle, the indicator can exaggerate the out-of-roundness or almost ignore it. PMPA gives the rule for choosing the block: first use a standard 90° V-block to count the lobes, then use a block whose included angle is 180° minus 360° divided by the number of lobes, which gives 60° for a three-lobed part (PMPA). In that matched block the indicator swing on a lobed part is magnified, so the raw reading overstates the value a roundness instrument would report until it is corrected with a factor confirmed against a radial measurement.
Practical rules for using it as a screen:
- Write down the block angle with the result. A reading without it cannot be compared with anyone else’s.
- Count the lobes first, from the number of indicator peaks in one revolution.
- Use the V-block to compare parts against a master part that was measured on a roundness instrument, not to produce an absolute value.
- Keep the part pressed into the V and the contact faces clean; a burr or chip on a face shows up as a lobe.
What settings make a roundness instrument result repeatable?
The filter, the reference circle, the measuring height and the centering must all be stated, because the same profile gives different roundness values when any of them changes.
Roundness instruments typically offer filter settings of 15, 50, 150 and 500 undulations per revolution (UPR) and four reference circles: least squares (LSCI), minimum zone (MZCI), minimum circumscribed (MCCI) and maximum inscribed (MICI), as listed in the specification of Mitutoyo’s compact instrument (Mitutoyo RA-10 catalog). A low UPR filter shows only the gross shape; a high one lets turning marks and chatter into the result. Minimum zone matches the tolerance-zone definition of circularity and gives the smallest value; least squares is the common default in software and gives a slightly larger one.
Centering matters too. The same catalog explains that an offset between the turntable axis and the part axis distorts the measured form (limaçon error) and inflates the calculated roundness, which instruments correct within limits. On the drawing or the inspection plan, say: reference circle, filter, the axial heights of the sections, and whether the result is for one section or the worst of several.
Which method should the drawing or inspection plan call for?
Match the method to the tolerance and the failure mode: a micrometer for ovality on loose callouts, a matched V-block as an in-process screen, and a radial instrument for acceptance of tight circularity on bearing, seal and valve surfaces.
A simple decision path works for most turned parts:
- Only a size tolerance, no circularity frame: two-point checks are what the size tolerance means under ISO defaults; nothing more is required unless the drawing applies the envelope requirement.
- Circularity callout comfortably larger than the measuring uncertainty of a V-block setup: a matched V-block check in process, with a radial measurement on first articles to confirm the lobe count and the conversion.
- Tight circularity on a sealing or bearing diameter, or any dispute: a radial instrument with the filter and reference circle written on the drawing or in the inspection plan.
Whichever method is chosen, the gauge has to be good enough for the tolerance. The rule of thumb on gauge resolution versus tolerance is covered in the 10-to-1 gauge accuracy guide.
How should the RFQ state roundness, and how does Saola check it?
State the circularity value, the measuring method and its settings in the RFQ, so the method can be agreed at drawing review before the first part is turned.
Turned parts are machined at our lead CNC plant in Vietnam on CNC turning. The plant’s inspection equipment includes a 2D optical (vision) measuring instrument, a height gauge and a hardness tester, plus hand gauges; it does not have a CMM, and a roundness instrument is not on its equipment list. If your drawing carries a tight circularity callout, tell us the method you accept. At drawing review we confirm whether it can be verified with the plant’s equipment, or agree a method with you, for example a V-block screen correlated to an outside radial measurement on first articles. New parts get a first-article inspection report and every lot a dimensional inspection record. Send the drawing for a quote.
Frequently asked
Is roundness the same as circularity?
Yes. ISO calls the control roundness and ASME Y14.5 calls it circularity; the symbol and the meaning are the same. Both are judged one cross-section at a time.
Can a three-point bore gauge find lobing in a hole?
It sees some lobe counts that a two-point gauge misses, but like a V-block its reading depends on how the lobe count relates to its contact geometry. Use it as a screen, and confirm a tight bore callout radially.
Is runout the same as roundness?
No. Runout is measured relative to a datum axis, so it combines roundness with eccentricity. A part can have good roundness and poor runout, and the reverse is not possible on the same section. See the runout guide for how each is checked.
Why does my supplier’s roundness value differ from my lab’s?
Usually because the filter, reference circle or measured height differ, or one side used a V-block. Agree the settings in writing and measure the same marked part on both sides before arguing about the parts.
Sources
- ASME B89.3.1, Measurement of Out-of-Roundness, ASME, 1972 edition (reaffirmed)
- Evaluation Methods for Roundness, Glossary of Measurement, Mitutoyo America Corporation, 2025
- Compact Roundness Measurement ROUNDTEST RA-10, catalog E15019(2), Mitutoyo Corporation, accessed 2026
- 5 Things To Know About Triangle Effect, Pseudo-Roundness, and Lobing, Precision Machined Products Association, September 2010