Roundness harmonic analysis helps a buyer see the shape hidden behind one circularity result. A measured trace shows radial variation around a feature; its spectrum organizes that variation by undulations per revolution (UPR). Reviewing both prevents a low-detail pass/fail number from masking a lobe pattern, setup artifact, or reporting mismatch that matters to the drawing and application.
What harmonic analysis adds to a roundness report
A roundness value reduces a circular profile to one result against a chosen reference and evaluation method. That result is useful, but it does not describe where deviations occur around the circumference or whether they are concentrated in a repeated pattern.
Harmonic analysis starts with the same measured circular trace and separates its repeating content by order. An order is the number of undulations per revolution (UPR): a second-order component repeats twice around the feature, while a third-order component creates three repeating high-and-low regions. The spectrum displays the amplitude associated with each order. Digital Metrology describes harmonic analysis as decomposing measured roundness data into component frequencies and reporting their amplitudes. Digital Metrology
For CNC part approval, the important question is not “what order caused the part?” A spectrum is evidence for discussion, not proof of a production cause. Its practical value is comparative: it lets the buyer, supplier, and quality team determine whether they reviewed the same profile information under the same measurement conditions.
A lobe is a recurring departure from a perfect circle. “Lobing” is commonly used for the overall repeated pattern, although a real trace can contain several harmonic orders at once. A profile dominated by a three-lobe component can look materially different from one dominated by a higher-order texture-like pattern, even where a single overall roundness result is similar.
The trace, reference circle, and spectrum answer different questions
Treat the report as three connected views, not as interchangeable outputs.
The polar or Cartesian trace plots radial deviation around the measured section. It shows whether a deviation is isolated, broadly oval, repetitive, or interrupted by a local event. It also makes angular orientation available for comparison when the reporting setup preserves a meaningful angular reference.
The reference circle is the mathematical baseline from which radial deviations are calculated. A least-squares circle is one common baseline: NIST describes fitting equally spaced trace points to a circle to estimate noncircularity from the difference between each measured distance and the fitted radius. NIST Roundness Measurements Other roundness evaluations may use different reference concepts. The drawing, inspection plan, or applicable standard must decide which evaluation is required.
The spectrum converts the trace into amplitude by harmonic order. It helps answer: which repetitions are prominent? It does not replace the trace, because it may omit orientation information, local defects, and the visual context needed to judge whether comparison conditions are genuinely equivalent.
Phase is the angular location of a harmonic component relative to the selected zero reference. It can be valuable when comparing repeat measurements or multiple parts, but only if the angular datum, part orientation, and reporting convention are documented. A spectrum without a reliable shared angular reference should not be used to argue that two patterns occurred in exactly the same physical location.
Why equal overall roundness can hide a different problem
Consider two hypothetical bore reports that show the same overall roundness result under an agreed evaluation. Part A has a smooth trace dominated by a low-order, three-lobe component. Part B has many smaller, closely spaced deviations distributed around the profile. The overall number alone does not tell the reviewer which part has which distribution.
That distinction can matter when a mating condition, sealing behavior, balancing concern, downstream measurement method, or drawing-specific harmonic requirement makes certain profile patterns more relevant than others. It can also matter during supplier dispute resolution: a replacement part may match the headline roundness number while clearly failing to match an approved trace-and-spectrum signature.
The correct buyer response is not to assume that one pattern is always worse. Ask whether the drawing defines harmonic limits, a functional reason for a particular restriction, or only overall roundness. If no harmonic requirement exists, do not retroactively convert an exploratory spectrum into an undisclosed acceptance criterion. Instead, use the information to decide whether the drawing and control plan should be clarified for future lots.
Separate setup eccentricity from component form
A report can contain variation created by the part, the fixture, centering, alignment, measurement system, or a combination of these. Eccentricity during setup may introduce a strong low-order appearance, but a low-order pattern is not automatically setup error. Conversely, correcting centering does not establish that every remaining lobe belongs solely to the component.
This distinction is especially important with a three-lobe pattern. Three lobes can be a useful clue that prompts review of workholding, manufacturing sequence, or measurement setup. It is not proof that a three-jaw chuck created the condition, and it should not be used to assign fault without corroborating evidence.
Ask for controlled comparison rather than causal storytelling. Repeat the measurement after documented re-fixturing where appropriate; compare another section or orientation if the inspection plan allows it; and preserve the same reference-circle method, filtering, and amplitude convention. NIST notes that a single roundness trace can contain both spindle and workpiece error, while multiple traces with rotation are used in high-precision work to separate those effects. NIST Roundness Measurements That background does not turn every production dispute into a high-precision study; it explains why one trace alone may not settle attribution.
Report fields buyers should review before accepting a spectrum
A spectrum without its measurement context is incomplete. Request the report package or exported data that makes these items clear.
Trace section and feature identity
Confirm the drawing revision, part identifier, feature identifier, measured cross-section, and whether the result concerns an ID, OD, or another circular feature. For a stepped or long feature, the axial location of the section can be decisive. A report should make it possible to connect the trace to the inspected feature without inference.
Datum, setup, and alignment
Record how the part was located, what surface or feature established orientation, and whether the measured axis was aligned to a specified datum scheme or a measurement-derived axis. This is not paperwork for its own sake. It tells reviewers whether two reports are truly comparable and whether phase can be interpreted.
Angular sampling and captured data
Ask how the angular profile was acquired and whether the report preserves the raw trace or exportable numerical data. Sampling must be suitable for the harmonic range being evaluated, but there is no universal point count that fits every component. The specified range, feature size, evaluation objective, and measurement method govern that decision.
Filtering and passband
Filtering can attenuate some content and change the relative prominence of the remaining components. Review the declared filter type, cutoff or passband, and whether the displayed trace and spectrum use the same conditioned data. Do not prescribe a universal UPR filter. A buyer should require the filter settings defined by the drawing, internal inspection plan, customer requirement, or agreed approval procedure.
Harmonic range and omitted orders
State the UPR range shown in the spectrum and any orders excluded from evaluation. Low-order terms may be treated differently from higher-order content depending on the reporting convention. A cropped spectrum can make a profile look simpler than it is, while an excessively broad display can distract from the range actually relevant to the requirement.
Amplitude convention
This field deserves explicit attention. Harmonic amplitude can be reported as mean-to-peak or peak-to-valley, and those are not interchangeable. Digital Metrology specifically cautions that harmonic tolerance interpretation depends on matching the amplitude convention used in the requirement and report. Digital Metrology
Before comparing supplier data to an internal limit, verify units, radial versus diametral presentation, amplitude convention, and whether the value is a single harmonic amplitude, a band result, or a derived summary. A buyer should not divide, double, or otherwise convert a value unless the governing specification explicitly supports that conversion.
A practical decision framework for report review
Use the following sequence to decide what action is proportionate.
- If the drawing specifies only overall roundness and the supplied report follows the agreed method, use the harmonic plot as diagnostic context, not an unapproved rejection rule.
- If the drawing names a harmonic order, UPR band, filter, or amplitude limit, verify every reporting convention before making an acceptance decision.
- If the component has a functional concern but no harmonic requirement, hold a joint engineering review. Decide whether to add a future drawing note, a controlled master trace, or a defined report format.
- If reports conflict, compare raw trace data and report settings before debating suspected machining causes.
- If the feature is critical enough that measurement-system contribution could control the decision, escalate the metrology plan rather than relying on a single screenshot.
Best-fit versus minimum-zone evaluation and CMM-derived circularity can be relevant to the broader inspection plan, but neither should distract from the harmonic review. First establish what trace was captured, how it was referenced and filtered, and which spectrum convention was applied.
Numbered workflow for approving a harmonic report
- Confirm the contractual basis. Identify the drawing revision, feature callout, acceptance rule, and whether harmonic limits are explicitly required.
- Verify report identity. Match the part, serial or lot reference where applicable, section location, ID or OD, and inspection date record to the submitted part documentation.
- Review the trace before the spectrum. Look for repeat patterns, isolated departures, clipping, missing regions, and a scale that could conceal meaningful differences.
- Check setup information. Review datum strategy, fixturing, alignment, centering approach, and angular zero reference. Mark any missing information as a comparison limitation.
- Reconcile processing settings. Compare angular sampling, filtering or passband, reference-circle method, retained harmonic range, units, and amplitude convention with the requirement.
- Evaluate the spectrum against the correct requirement. Compare only the specified orders or bands and only after confirming that both limit and report use the same convention.
- Preserve approval evidence. Retain the signed report, trace image or data export, spectrum, settings, revision, and disposition so a later reorder has a defined baseline.
- Close the loop on ambiguity. If the data reveal a functional concern not controlled by the drawing, document a proposed future requirement rather than treating an informal expectation as a hidden acceptance limit.
Hypothetical supplier-dispute example
A buyer sees a stronger third-order peak on a replacement sleeve report, although both lots show similar overall roundness. Before attributing the difference to machining, compare the measured section, setup, raw trace, filter, UPR range and amplitude convention. If the difference persists under equivalent conditions, assess it against the stated requirement or request engineering review. If it disappears when reporting settings match, document the reporting mismatch and retain the reconciled reports. Neither screenshot alone establishes a manufacturing cause or authorizes an unstated rejection limit.
Compact report-review checklist
- Drawing revision and feature section are identifiable.
- Overall roundness evaluation method matches the agreed requirement.
- Trace is included, legible, and paired with the spectrum.
- Datum, setup, alignment, and angular reference are stated.
- Angular sampling and data retention are documented.
- Filter or passband is declared.
- UPR display range and exclusions are clear.
- Units, radial/diametral basis, and amplitude convention match the limit.
- Raw trace and spectrum settings are available for a dispute comparison.
- Approval record preserves the report configuration, not only the headline value.
For drawing-defined components, SUUXIANG can review the inspection information that should accompany a roundness requirement during DFM and quotation discussion. For related callout context, see GD&T Inspection. To align a report format, trace retention, and critical-feature documentation before production, discuss your project.
Buyer FAQ
Is UPR the same as the number of lobes?
A dominant harmonic order describes how many times that component repeats per revolution. Real profiles can contain multiple orders, so the visible shape is not always represented by one lobe count alone.
Can a part pass roundness and still need harmonic review?
Yes. A pass/fail overall result and a harmonic distribution answer different questions. Harmonic review is appropriate when the drawing, function, approval history, or dispute process requires it.
Does a three-lobe spectrum prove a chuck problem?
No. It is a clue for investigation, not proof of a particular workholding or machining cause. Compare controlled measurements and setup information before attributing cause.
Which UPR range should be specified?
There is no universal range. The functional requirement, drawing, feature geometry, measurement method, and agreed inspection plan should define the relevant range.
What should an RFQ include when harmonic control matters?
Include the drawing revision, controlled feature and section, required overall roundness method, harmonic orders or bands if applicable, amplitude convention, filtering/passband, report format, data-retention expectation, and required approval documentation.