SINTS Blog Quality and Verification

How to Review First Article Samples for Metal Parts

A first article is a measurement exercise, not an inspection. Here is what to verify, which reports carry real information, and how to close out a discrepancy.

Universal testing machine in the SINTS inspection laboratory

First article samples decide whether a program moves to production. They are also, too often, reviewed by holding the part next to a drawing and judging whether it looks right. Appearance tells you very little about a sintered metal part: the dimensions that matter are the ones inside bores, on datums and across fits.

This guide covers how to structure a first article review for MIM and PM parts — what to measure, what to ask for in writing, and how to handle the results that fall outside tolerance.

Start with the drawing, not the part

Before measuring anything, confirm which revision of the drawing the supplier built to. Samples that are dimensionally correct against an obsolete revision are a recurring and entirely avoidable source of confusion.

Then read the datum scheme. MIM and PM parts are measured from datums established after sintering, and the drawing should define them. A first article report that records dimensions without stating how the part was fixtured is not verifiable — you cannot tell whether a reading reflects the part or the setup.

Finally, identify the feature set that actually matters. Typically this is a short list: the fits, the bores, the sealing or bearing surfaces, the datums and any dimension whose deviation would cause an assembly or field failure. Everything else is secondary, and treating it as equally important wastes inspection budget and delays the review.

What to measure, what to leave

A first article report should cover every dimension on the drawing once, but the interpretation should be tiered.

How to tier a first article review
TierWhat it coversAcceptance approach
CriticalFits, bores, sealing faces, bearing journals, datumsCompare against drawing tolerance feature by feature; require a capability statement, not just a reading
SignificantAssembly clearances, mating dimensions, wall sectionsVerify against tolerance with the process capability in mind
GeneralCosmetic and non-functional dimensionsConfirm within general tolerance note; do not over-interpret single readings

A single sample tells you whether the part is in tolerance. It does not tell you whether the process is capable of staying there. Where a critical feature sits mid-tolerance, that is reassuring. Where it sits on the tolerance limit, ask what the producer expects the variation to be across a run — a first article that only just passes is a forecast of rejects.

Check the material, not only the dimensions

Dimensional conformance says nothing about whether the part has the properties the application needs. Three checks cover most of the risk.

  • Density. Usually measured by the Archimedes method and reported as a percentage of theoretical density. MIM normally reaches 95–99%; a figure well below the specification indicates a furnace or feedstock problem that will affect strength, sealing and magnetic response.
  • Hardness. Confirms that the specified heat treatment was applied and produced the intended result. A hardness reading also flags an unintended condition — 17-4PH that was never aged, or a martensitic grade that was over-tempered.
  • Chemical composition. Supplied by the powder producer and confirmed on the material certificate. This is the only practical way to detect a grade substitution, and it is worth checking when the application is sensitive to alloy content.

Density is the underrated first-article check. It is quick, it correlates with mechanical and magnetic performance, and it is the earliest indicator that the sintering process was not where it should be. If a part is dimensionally correct but low on density, the dimensional pass is not good news — it means the shrinkage was compensated for the wrong reason.

Review fit and function together

A dimension can be inside tolerance and still not work. This happens most often at fits and at features that interact with a mating part. Where the design permits, ask for samples to be checked against the mating component rather than measured in isolation — an assembly test catches form and position problems that a dimensional report, taken feature by feature, does not.

The same applies to appearance. If the part is visible to the end customer, inspect the finish under the lighting conditions the customer will see, not under a bench lamp. Cosmetic disagreements at first article review are common and are much cheaper to settle with a photograph and an agreed standard than with a rejected lot.

Reports worth requesting

  • Dimensional first article report — every drawing dimension, with the drawing revision, measurement method and datum reference stated.
  • Material certificate — alloy grade, heat or lot number, and chemical composition from the powder supplier.
  • Density report — measured value and method, usually Archimedes.
  • Hardness report — the scale used, the readings and the test locations.
  • Process capability data where the program requires it — readings from a run, not from a single sample, which is what actually supports a capability claim.
  • Inspection plan — what will be checked in production, how often, and with what method.

Where a formal submission is required, these documents form the backbone of a production part approval package. Building them into the first article review is easier than assembling them afterwards.

When results fall outside tolerance

This is normal, and it is more informative than a clean first article. What matters is the cause and the response.

  • Establish whether the deviation is a measurement or a part problem. A dimension measured differently between the two parties is common, especially on features that are difficult to fixture. Confirm the method before accepting a deviation.
  • Identify whether the cause is tooling, process or material. A dimensional offset across many features usually means shrinkage compensation; a deviation on one feature usually means a tool correction; a density or hardness miss means a process parameter.
  • Ask what the correction will be and whether it requires tool work. Some corrections are process adjustments that cost a run; others require a mould change with a lead time.
  • Get the corrected plan in writing, with a date. The point of a discrepancy report is to close it, not to record it.

Closing out from sample to production

Production approval is not the same as passing a first article. Before releasing a program, confirm four things: that the first article is approved against the current drawing revision, that density, hardness and material records are on file, that the production inspection plan matches the features that actually matter, and that both parties agree on the process window that will be held.

That last item is the one most often skipped. A first article produced under special attention is not evidence that routine production will match it. Agreement on the parameters that will be monitored — sintering profile, feedstock lot control, tool maintenance intervals — is what makes the first article a predictor rather than an isolated good result.

Conclusion

A first article review is a structured comparison between a part, a drawing revision and a specification. Dimensions are the visible part, but density, hardness and material records are what confirm the part is made of what it claims to be.

The most valuable output of the process is not an approval signature. It is a clear understanding of which features carry the function, how much margin they have, and what will be monitored in production. Programs that establish that at first article stage tend to run without surprises; programs that do not tend to rediscover it at a rejected lot.

Related component families

These pages cover the component families where first article requirements tend to be most demanding.

Have a drawing to discuss?

Send it to our engineering team. We review process fit and identify the information needed for a quotation.

Send Drawing for Review

Frequently Asked Questions

Practical answers for OEM sourcing and engineering teams.

What should I check on a first article?

Measure every drawing dimension against the stated datum scheme, but interpret the results in tiers: critical features such as fits, bores and bearing journals first, then assembly dimensions, then general dimensions. Alongside dimensions, verify density, hardness and the material certificate. A part can be dimensionally perfect and still fail in service if density or heat treatment is wrong.

Which reports should I request with first article samples?

Ask for a dimensional first article report that states the drawing revision, measurement method and datums; the material certificate with alloy grade and chemical composition; a density report; a hardness report with scale and test locations; and the production inspection plan. Where the program requires formal approval, these documents form the basis of the submission package.

How do I close a PPAP for a metal part?

Confirm that the first article is approved against the current drawing revision, that density, hardness and material records are on file, that the inspection plan covers the features that actually matter, and that both parties agree on the process parameters that will be held in production. Agreement on those parameters is what makes the first article predictive rather than a one-off good result.

What if first article samples are out of tolerance?

First establish whether the difference is a measurement issue or a genuine part deviation, because differing measurement methods are a common cause on features that are hard to fixture. Then identify whether the cause lies in tooling, process or material. Ask what the correction is, whether it requires tool work or only a process adjustment, and get the corrected plan and date in writing.

Does a passing first article mean production will be consistent?

Not on its own. A first article produced with special attention is not evidence that routine production will match it. Consistency comes from agreeing which parameters will be monitored during production — sintering profile, feedstock lot control, tool maintenance intervals — and from having capability data from a run rather than readings from a single sample.