CMM Inspection and First-Article Reports

Start with the characteristic, not the machine

A CMM gathers coordinates from a manufactured surface and uses them to evaluate features. NIST emphasizes an important limitation: geometric tolerances can concern an entire surface, while a measurement samples only part of it. Point selection and mathematical evaluation therefore influence the result. More output does not automatically mean more useful evidence. [1]

Begin the inspection plan with a list of drawing characteristics. Identify the feature, applicable tolerance, datum references where required, and the final condition in which it must be inspected. Then select a method that can access and evaluate that requirement. Some characteristics may need a different instrument or a complementary check.

For quoting, send the released drawing and model, part envelope and weight, material condition, quantity to be inspected, and required reporting format. Include fixtures or restraint requirements when the design specifies them. This gives the supplier something concrete to review before promising a turnaround or inspection package.

Make alignment and setup repeatable

ZEISS's inspection guidance stresses consistent revisions across CAD, drawings, and inspection plans, appropriate fixturing, and management of environmental influences. A setup should support the intended measurement without introducing avoidable deformation or instability. [2]

Ask the inspection team to describe the alignment and the surfaces used to establish it. For a datum-based requirement, confirm that the evaluation follows the drawing's references and governing rules. A convenient best fit to a model should not be silently substituted for the required reference system.

Clean and identify the part before the agreed inspection. Preserve the relationship between the part identifier and the report. If the part is repositioned, include enough setup information to explain how separate measurements relate. For repeat orders, retain the approved program revision and documented changes so results remain understandable across production runs.

Discuss difficult access early. A deep bore, obstructed face, thin wall, or large fixture can affect the practical inspection arrangement. Machine travel alone does not answer whether a particular feature can be reached and evaluated in the required condition.

Sampling and fitting require deliberate choices

NIST's CMM uncertainty report identifies influences including probe approach, point distribution, selected algorithms, fixturing, temperature, and part behavior. A fitted geometric element is an interpretation of sampled data; an unsuitable sampling strategy can miss relevant form variation. The report is a technical basis for planning, not a universal point-count recipe. [3]

For each demanding feature, ask why the proposed coverage is appropriate. Should the inspection sample more than one section? Is scanning or discrete probing suitable for the requirement and available system? How will edges, interruptions, or inaccessible regions be handled? Record exceptions explicitly rather than letting a software default decide acceptance.

Do not create a rule that every bore needs the same number of points. A repeatable measurement of the wrong geometric question remains the wrong answer. The designer, supplier, and receiving team should agree on the characteristic and evaluation basis before comparing results.

Calibration supports the result; it does not replace it

Metrological traceability belongs to a measurement result. NIST describes a documented calibration chain to a reference, with uncertainty contributions throughout that chain. It also cautions that calibration of an instrument alone does not establish traceability for every result, and traceability alone does not prove fitness for a particular purpose. [4]

For a critical acceptance decision, ask whether the measurement process is suitable for the tolerance being evaluated. The answer should consider the actual task and its conditions, not only the instrument's displayed resolution or a calibration sticker. NIST's work on complex instruments specifically distinguishes CMM calibration information from the additional factors needed for a task-specific uncertainty statement. [5]

Agree in advance how results near an acceptance limit will be handled, including any contractually required decision rule. If two laboratories disagree, preserve the original data and compare setup, temperature conditions, algorithms, and uncertainty information. Repeating a measurement without examining those differences may leave the underlying disagreement unresolved.

Define the report before the first part

A practical report should be easy to reconcile with the order. Request the part and drawing revision, relevant part or lot identification, characteristic identifiers, units, nominal requirements, measured results, and the agreed acceptance presentation. Add setup or method information where it is needed to interpret the results. Specify the electronic format if your receiving system requires one.

First-article inspection has a broader process and documentation purpose than producing a CMM printout. IAQG's 9102 material addresses a standardized FAI process; its revision-labeled guidance discusses planning for characteristics that may not be measurable after later operations. Establish the invoked revision and customer requirements rather than assuming every dimensional report is a complete FAIR. [6] [7]

ATM's shop-supplied equipment register lists Giddings & Lewis Cordax RS-50 and Cordax 1000 CMMs, along with comparator and height-measurement equipment. That listing is not a published measurement-uncertainty guarantee or a commitment to inspect every part envelope. Confirm your job's access, setup, method, and deliverables. [8]

Review ATM's machining and inspection capabilities, then compare drawing requirements with quality-system and first-article planning. The Machining Knowledge hub brings these related decisions together.