A supplier sends you a first article inspection report: three forms, forty-odd rows of dimensions, every result marked "PASS". Your purchase order says drawing revision C. The report header says revision B. Do you read the measurements?
No. A first article inspection report is only evidence about the configuration it names. If the part number, drawing number or revision on the report does not match what you ordered, nothing below that header is verifiable — the numbers may be excellent and still describe a part you never bought. That single check takes thirty seconds and should happen before you open the dimensional sheet.
A first article inspection report is the documented verification that a first production part, made with production tooling and processes, meets every requirement on the released drawing.
This walkthrough goes one level deeper than our overview, First Article Inspection: What to Require and How to Review It. Here you get what belongs in each field of Form 1, Form 2 and Form 3, one clearly labelled illustrative part carried through all three forms, a deliberately flawed entry per form with the rejection rationale, and a review sequence you can run on any submission.
About the format and the field names
AS9102, published by SAE International, is the aerospace first-article report format. Buyers in automotive, medical and industrial programmes also use the three-form structure as a first article inspection report (FAIR) template, because it forces accountability for the part, the material and processes, and every drawing characteristic separately. Fields below are described by name, not by number — confirm the exact field numbering and edition against the copy of the standard your contract invokes, and against the customer-specific template if one applies.
Table Of Contents
- What Each of the Three Forms Is Accountable For
- Form 1: Part Number Accountability, Field by Field
- Form 2: Product Accountability — Materials, Processes, Testing
- Form 3: Characteristic Accountability, Row by Row
- Worked Example: One Illustrative Part Through All Three Forms
- The Buyer's Review Sequence
- Full FAI or Partial FAI: When to Demand a New Report
- What NICE Rapid Provides With a First Article Submission
- FAQ
What Each of the Three Forms Is Accountable For
Each form answers a different question, and a defect in one cannot be compensated by strength in another. Perfect dimensions do not rescue an unidentified material lot.
| Form | Question it answers | What you audit it against |
|---|---|---|
| Form 1 — Part Number Accountability | Is this the right part, at the right revision, from the right source? | Purchase order, drawing title block, approved supplier list |
| Form 2 — Product Accountability | Were the right materials and special processes used, and were required tests run? | Drawing notes, material and process specifications, certificates |
| Form 3 — Characteristic Accountability | Was every drawing requirement measured, with a suitable method, and did it conform? | Ballooned drawing, tolerance callouts, raw measurement records |
Read them in that order. Form 1 establishes identity, Form 2 establishes provenance, Form 3 establishes conformance. Reversing the order wastes hours auditing dimensions on a report you should have returned unread.
Form 1: Part Number Accountability, Field by Field
Form 1 is the report's identity page. Every field on it should be traceable to a document you already hold — the PO, the drawing, or your supplier approval record.
Part number, part name and drawing number with revision level
Enter the part number exactly as it appears on the PO and the drawing title block, including suffixes and dash numbers. The drawing number and its revision level are separate entries from the part number: a part can stay at the same number while the drawing moves from B to C. Record the drawing revision, and where the drawing carries them, the revision date and any change notice or ECO reference.
It is worth checking three things together: the revision on the report, the revision on the ballooned drawing attached to the report, and the revision on your PO. All three must agree. In our pre-production review of drawings, tolerances and inspection expectations, configuration identity is the first thing we settle: if the report names a revision you did not order, there is nothing further to assess until it is corrected.
FAI type: full or partial, and the reason for a partial
The report should state whether this is a full FAI (every characteristic verified) or a partial FAI (a defined subset re-verified). A partial FAI is not automatically suspect — but it is only meaningful with three things attached: the reason for the partial, the identifier and date of the previous full FAIR it builds on, and the list of characteristics actually re-verified.
A partial marked without a reason, or referencing a prior report you cannot locate, hides the scope question entirely. Ask which characteristics were carried forward and why they were unaffected by the change.
Manufacturing source, purchase order and organisation identifiers
Record the manufacturing source — the facility that actually made the part, by name and location — plus its supplier code where your system assigns one, the PO number and line, and the organisation that prepared the report. Where a subcontractor performed machining, molding or a special process, that relationship should be visible rather than implied.
The failure mode to look for: the report names the trading entity or the group headquarters, while the parts came from a different plant. Source changes matter because they change process capability, and a report that obscures the source cannot support a source-change decision later.
Assembly component lists and traceability
For an assembly FAIR, the component list should enumerate every part number in the assembly with its own revision and the identifier of its own FAIR — detail parts do not inherit acceptance from the assembly report. Each sub-tier report should be locatable by number and date, and each should itself pass the Form 1 checks above.
An assembly Form 1 listing components without revisions, or without pointers to their detail reports, leaves you unable to say which configuration was built. Treat blank component revisions as an open item, not a clerical one.
Signature, date and the completeness statement
The report should carry the name and signature (or controlled electronic equivalent) of the person accountable for it, the date, and the reviewer/approval entries your contract requires. An unsigned report is a draft.
Form 1 red flags
- Drawing revision on the report differs from the PO or from the attached ballooned drawing.
- "Partial" selected with no reason, no prior FAIR reference, or no list of re-verified characteristics.
- Manufacturing source given as a group name rather than the producing facility.
- Assembly components listed without revisions or without links to their own detail FAIRs.
- No signature or date, or a date preceding the drawing revision it claims to verify.
Form 2: Product Accountability — Materials, Processes, Testing
Form 2 lists what the part is made of and what was done to it. Each line is a claim; the value of the form lies entirely in the evidence sitting behind it.
Raw material specifications and material certifications
For each material, record the specification and grade/temper/alloy designation as called out on the drawing, the material form (bar, plate, resin grade, sheet), the supplier or mill, and the certificate identifier with its lot, batch or heat number. Then attach the certificate.
A material certification is useful when it lets you tie three things together: the specification and revision it certifies against, the composition or property results, and the lot identifier that also appears on the traveller for the parts inspected. "Aluminium 6061" in a field, with no certificate attached, tells you nothing you can audit. Neither does a certificate for a different heat number.
Where a plastic resin is specified, the same logic applies to grade and colourant/masterbatch identification and the lot on the supplier's certificate — a generic "PC" entry does not establish that the specified grade was molded.
Special processes
Special processes — heat treatment, plating, anodizing, conversion coating, welding, sterilisation and similar operations whose result cannot be fully verified by inspecting the finished part — need their own lines: the process specification with its revision, the class/type/grade where the spec defines them, the processor's name and location, and the certificate number.
Two questions decide whether an entry is real. Does the certificate reference the specification and revision on your drawing, not just a family name? And does it cover the lot or batch that the inspected parts came from? A process certificate that names neither is not evidence of compliance with your requirement.
Functional testing and required test reports
Where the drawing or specification requires functional or performance testing — pressure/leak, torque, electrical, hardness, coating thickness, adhesion — Form 2 should identify the test specification, the acceptance criteria, the sample identity, the actual result and the report reference. "Tested — OK" is not a result.
An empty testing field is ambiguous in the worst way: it can mean the test was not required, or that it was required and not performed. Insist that non-applicable entries be marked as such with the reason, so silence is never an option.
What vague Form 2 entries hide
| Entry as submitted | What it fails to establish | What to request |
|---|---|---|
| Material named without spec revision | Which edition of the specification was met | Certificate citing spec and revision |
| Certificate with no lot/heat/batch number | That the certified material is the material in these parts | Lot-linked certificate plus traveller reference |
| Process listed as "anodize" | Class/type and thickness requirement compliance | Process certificate citing spec, revision, class |
| "Functional test: pass" | Acceptance criteria and measured value | Test report with criteria, result, sample ID |
| Blank process or test fields | Whether the requirement was waived or missed | Explicit N/A with reason, or the missing evidence |
Form 3: Characteristic Accountability, Row by Row
Form 3 is where most of the audit time goes, and where a report can look complete while leaving requirements unverified.
The ballooned drawing and one row per characteristic
The submission should include a ballooned drawing: the released drawing with a unique number beside every requirement. Every balloon must be accounted for on Form 3 — at least one row per characteristic, with extra rows or continuation sheets where several features or samples are reported. The audit test is that no balloon is missing, not that the counts match exactly. That includes dimensions, tolerances, GD&T feature control frames, surface finish callouts, and drawing notes — material notes, break-edge notes, marking requirements, general-tolerance title-block callouts.
Notes are the most commonly dropped items, because they are text rather than numbers. It is worth counting balloons yourself before reading results: if the drawing has 34 balloons and Form 3 has 29 rows, five requirements were never verified, and the report's overall "PASS" is meaningless. Missing rows hide unverified requirements more effectively than bad numbers do, because nothing on the page looks wrong.
Requirement, tolerance and measured result
Each row should restate the requirement as drawn — nominal plus tolerance, or the tolerance zone for a geometric callout — then the actual measured result. Two habits make results auditable:
- Record the value, not a verdict. "0.06" against a 0.10 position tolerance is evidence; "OK" is an opinion. For attribute requirements, record what was observed against the stated criterion.
- Keep units and precision consistent with the callout. A tolerance stated to three decimals should not be reported to one; rounding can turn a marginal result into an apparent pass.
For a geometric tolerance, the row should also make the datum reference frame explicit — the same datums, in the same order, as the feature control frame. A position result measured from a different alignment is a different measurement, even when the number looks acceptable. For how alignment and datum setup show up in the raw report, see our quality inspection operation.
The measurement method column
Every row should name the method or equipment used: CMM, optical comparator, height gauge, micrometer, bore gauge, surface roughness tester, visual per stated criterion. This column is where you catch the quietest failure in a FAIR — a method that cannot resolve the tolerance it claims to verify.
A vernier caliper resolving 0.02 mm cannot verify an 18 µm band, and no hand tool can establish a position tolerance referenced to three datums. A mastered bore micrometer or bore gauge with a stated uncertainty appropriate to the band can be acceptable — the test is the instrument's resolution, uncertainty and calibration against the tolerance, not whether it is hand-held. Ask what the instrument's resolution and uncertainty are relative to the tolerance band, and ask for the calibration record. We deliberately give no universal ratio rule here: the acceptable relationship between measurement uncertainty and tolerance is a decision to settle against the measurement-system guidance your contract invokes, and it should be agreed before inspection, not argued after a borderline result.
Traceability to the CMM report
A CMM-measured row should be traceable to a specific raw record: report number, part/serial identifier, program or routine name and revision, date, and the characteristic identifier matching the balloon. Ideally the balloon number appears in the CMM report itself, so the mapping is mechanical rather than reconstructed by hand.
When a Form 3 row says "CMM" but no CMM report accompanies the submission, the dimensional results are unverifiable assertions. Request the raw report — not a retyped summary — and check that the datums, units and evaluation settings match the drawing.
Worked Example: One Illustrative Part Through All Three Forms
Illustrative example — not a customer part
The part, numbers, certificates and report identifiers below are invented for teaching purposes. They are not a NICE Rapid customer part, a real FAIR, or a record of any actual inspection outcome. Dimensions and tolerances are plausible for a small machined component and are used only to show what a correct entry looks like.
Illustrative part: mounting bracket, machined from aluminium bar, 60 × 40 × 8 mm, with a locating bore, a bolt-hole pattern and a milled sealing face. Part number MB-4402-01, drawing DWG-4402 revision C. Ten ballooned characteristics, including two notes.
Form 1 — correct entries and one flaw
| Field (by name) | Correct illustrative entry |
|---|---|
| Part number | MB-4402-01 |
| Part name | Mounting bracket |
| Drawing number / revision | DWG-4402 / C |
| PO number | PO-77310, line 2 |
| FAI type | Full |
| Reason for partial | N/A — full FAI |
| Manufacturing source | Producing facility named, with city and country |
| Report prepared by / date | Named inspector, dated after revision C release |
Deliberately flawed entry: drawing revision recorded as B while the PO and attached ballooned drawing are revision C.
Buyer's rejection rationale: return the report before reviewing any measurement. Revision B predates the change to the locating-bore position tolerance (balloon 5), so the results — however good — do not address the configuration ordered. Request a reissued FAIR against revision C, with a ballooned revision-C drawing and results for the changed characteristics.
Form 2 — correct entries and one flaw
| Field (by name) | Correct illustrative entry |
|---|---|
| Material specification | Aluminium alloy and temper exactly as the drawing note states, with spec revision |
| Material form / supplier | Extruded bar; named mill |
| Material certificate / lot | Certificate number tied to heat lot H-2291, matching the traveller |
| Special process | Anodizing per the drawing's cited specification, revision and class |
| Process certificate | Processor named; certificate cites spec, revision, class and the same lot |
| Functional test | Not required by drawing — marked N/A with reason |
Deliberately flawed entry: special process recorded as "anodize — done", with no specification, no class and no certificate attached.
Buyer's rejection rationale: the drawing calls out a specific process specification and class, which controls coating thickness and therefore the fit of the locating bore after finishing. An unspecified process entry cannot demonstrate compliance, and the post-finish bore result on Form 3 cannot be interpreted without knowing the coating condition. Hold acceptance pending a lot-linked process certificate citing spec, revision and class.
Form 3 — correct rows and one flaw
All ten balloons appear, one row each. Values in mm unless stated.
| Balloon | Requirement as drawn | Measured result | Method / record |
|---|---|---|---|
| 1 | Overall length 60.00 ±0.15 | 60.04 | CMM, report CMM-1187, part FA-01 |
| 2 | Overall width 40.00 ±0.15 | 39.96 | CMM, report CMM-1187, part FA-01 |
| 3 | Thickness 8.00 ±0.10 | 8.02 | Micrometer, cal. record MG-042 |
| 4 | Locating bore Ø12.000 +0.018/0 | 12.009 | CMM, report CMM-1187, part FA-01 |
| 5 | Bore position ⌀0.05 M to A|B|C | 0.031 (at Ø12.009; bonus 0.009; allowed ⌀0.059) | CMM, report CMM-1187, datums A|B|C |
| 6 | Bolt pattern 4× Ø4.50 ±0.10 | 4.52 / 4.48 / 4.55 / 4.51 | CMM, report CMM-1187 |
| 7 | Sealing face flatness 0.03 | 0.018 | CMM, report CMM-1187 |
| 8 | Sealing face Ra 1.6 µm max | 1.1 µm | Roughness tester, cal. record SR-011 |
| 9 | Note: break all sharp edges 0.2–0.5 | Edges 0.3, verified visually and by profile | Comparator + visual per stated criterion |
| 10 | Note: mark part number per drawing note | Marking present, legible, correct location | Visual per stated criterion |
Deliberately flawed entry: balloon 4 (Ø12.000 +0.018/0, an 18 µm band) recorded as "12.01 — vernier caliper".
Buyer's rejection rationale: the method cannot credibly resolve an 18 µm tolerance band, so the recorded value does not establish conformance even though it falls inside the limits. Request re-measurement on an instrument whose uncertainty is appropriate to the band, with the calibration record and the raw measurement report — and confirm whether the bore was measured before or after anodizing, since the drawing's dimension applies to the finished condition.
Balloon 5 carries the maximum-material modifier (⌀0.05 M to A|B|C), so the permitted position tolerance is not fixed at 0.05: the bore earns bonus tolerance as its actual size departs from its maximum-material limit of Ø12.000. At the illustrative measured size of Ø12.009 (balloon 4) the bonus is 0.009, giving an allowed zone of ⌀0.059, so the recorded 0.031 conforms with the bonus applied. A defensible row states the actual size used, the bonus and the total allowed tolerance next to the measured deviation, and states the measurement condition: because the drawing dimension applies to the finished part, the bore result must be taken after anodizing, not before.
Note what the correct rows share: a restated requirement, a numeric result at the callout's precision, a named method, and a pointer to a record you can request. That combination is what makes a FAIR auditable rather than merely tidy.
The Buyer's Review Sequence
Run these steps in order and stop at the first hard failure. Returning a report early is cheaper than auditing it twice.
- 1
Match identity — part number, drawing number and revision on Form 1 against the PO and the attached ballooned drawing. Any mismatch: return the report.
- 2
Confirm scope — full or partial. If partial, obtain the reason, the prior full FAIR reference and the list of re-verified characteristics before going further.
- 3
Count balloons against rows — every balloon on the drawing, including notes and GD&T frames, be accounted for by at least one Form 3 row, with extra rows or continuation sheets where several features or samples are reported. List the missing balloon numbers explicitly in your response.
- 4
Check evidence behind Form 2 — material and process certificates cite specification and revision, and carry the lot/heat/batch that matches the inspected parts. Required tests show criteria and actual results.
- 5
Test method suitability — for each tight-tolerance or geometric characteristic, confirm the recorded method can resolve the band, and ask for calibration records where it is doubtful.
- 6
Trace results to raw records — pull the CMM report and match part identifier, datums, units and characteristic IDs to the Form 3 rows. Retyped summaries are not raw records.
- 7
Read the marginal results — values near a limit deserve attention even when marked pass; ask how uncertainty was handled and whether the process is centred.
- 8
Close the loop in writing — issue accept, accept-with-open-items, or reject, naming the specific fields and balloon numbers involved.
What a defensible FAIR looks like
- Form 1 identity matches the PO and the attached drawing revision exactly.
- Every Form 2 line has attached evidence citing specification, revision and lot.
- Every balloon, notes included, has a Form 3 row — no characteristic is missing.
- Every result is a number at the callout's precision, with a method that can resolve the tolerance.
- CMM rows trace to a raw report with matching datums and part identifier.
- Partial FAIs name their reason, their parent report and their re-verified scope.
Full FAI or Partial FAI: When to Demand a New Report
A partial FAI is a scoped re-verification, so the question to ask is always the same: could this change have affected any characteristic that is not being re-measured? If yes, the carried-forward results no longer describe the current process.
Use this as a buyer's decision aid rather than as a quotation of the standard — the authoritative trigger list, including any production-lapse provision, should be read in the edition your contract invokes.
| Change | Reasonable buyer position |
|---|---|
| Drawing revision affecting form, fit, function or any toleranced characteristic | New full FAI, or a partial covering every affected characteristic plus those sharing the same datums or process step |
| Manufacturing source or facility change | New full FAI — process capability travels with the equipment and the shop |
| Process change (different machine type, tooling, mold, or process route) | New full FAI unless you can justify that unaffected characteristics remain demonstrably unchanged |
| Special-process processor change | Re-verify process-dependent characteristics and require new lot-linked process certificates |
| Material grade, supplier or lot change | Re-verify material-dependent characteristics; require new material certification |
| Cosmetic or documentation-only drawing change with no requirement altered | Partial FAI limited to the changed items, with the reasoning recorded |
| Long gap since the last production run | Confirm the lapse provision in the invoked edition and in your customer flow-down before accepting a partial |
Write the expectation into the PO. "FAIR per the invoked first-article standard, full FAI on source or process change, partial only with written justification and buyer approval" removes most of the argument later.
What NICE Rapid Provides With a First Article Submission
On our side of the table, first-article work runs through the same staged controls as production. Incoming materials and components are checked against specification at IQC before production starts; IPQC monitors during production; MQC is a further defined inspection stage; QA verifies finished parts against the agreed customer requirements before delivery; and QE completes the five stages, each with documented inspection procedures.
Dimensional verification is performed on our Tango 564 coordinate measuring machine (CMM), supporting first-article inspection (FAI), dimensional inspection, measurement reporting and production-run inspection for parts with defined inspection requirements. CMM reports, material certifications and full inspection documentation are provided on request — which is exactly the raw evidence the review sequence above asks you to demand. Our quality inspection operation documents the CMM and the five QC stages.
The practical implication for a buyer: specify the reporting expectations before production, not after the parts arrive. Balloon the drawing, state which characteristics need CMM measurement, state the datum scheme, and say which certificates you need attached. Agreeing the inspection scope up front is what makes the resulting report reviewable in one pass — whether the parts are CNC machined or injection molded.
FAQ
What is the difference between Form 1, Form 2 and Form 3 in a first article inspection report?
Form 1 accounts for the part itself — part number, drawing number and revision, FAI type, manufacturing source and, for assemblies, the component list. Form 2 accounts for materials, special processes and required testing, with certificates behind each entry. Form 3 accounts for every ballooned drawing characteristic, with the requirement, the measured result and the measurement method. Each form answers a separate question, so a strong Form 3 does not compensate for a weak Form 1 or 2.
Can I accept a partial first article inspection report?
Only when the change could not have affected the characteristics being carried forward, and when the report names its reason, references the previous full FAIR, and lists what was re-verified. Design, process and manufacturing-source changes generally warrant a new full FAI or a carefully justified partial scope. Check the trigger provisions in the edition of the standard your contract invokes rather than relying on a supplier's summary.
How many rows should Form 3 have?
At least one row per ballooned characteristic on the released drawing — including GD&T feature control frames, surface finish callouts and drawing notes, not just dimensions — with extra rows or continuation sheets where several features or samples are reported. Count the balloons yourself and check that every balloon appears before reading any results; the counts need not match exactly. A shortfall means requirements were never verified, which an overall "pass" statement will not reveal.
The Form 3 result is inside tolerance but measured with a caliper. Should I accept it?
Not without checking whether the method can resolve the tolerance band. A vernier caliper resolving 0.02 mm cannot verify an 18 µm band, and no hand tool can establish a position tolerance referenced to three datums; a mastered bore micrometer or bore gauge with a stated uncertainty appropriate to the band can be acceptable. The test is resolution, uncertainty and calibration against the tolerance, not whether the tool is hand-held. Ask for re-measurement with a suitable instrument, its calibration record and the raw measurement report.
Is a first article inspection report the same thing as a PPAP submission?
No. A first article inspection report verifies that a first production part meets every drawing requirement, in a structured three-form format. PPAP is the automotive production part approval process, a broader submission package with its own defined elements and levels. Many suppliers reuse FAI dimensional data inside a PPAP package, but the two are not interchangeable — confirm the required elements against the AIAG manual edition your customer specifies.
What should I ask for if the FAIR arrives without a CMM report?
Request the raw CMM report, not a retyped summary: report number, part or serial identifier, program name and revision, measurement date, units, and the datum alignment used. Match those against the Form 3 rows. Without that record, CMM-measured results are assertions you cannot verify.
Agree the inspection scope before production
Send your drawing, tolerance requirements and reporting expectations with your enquiry — the quality team reviews tolerances, datum schemes and inspection documentation before production starts. Include STEP or IGS files with quantity, material and finish requirements. You can also email rfq@nice-rapidtooling.com or review our quality inspection capability first.
Send drawings for reviewSources and how to verify
The format, the form titles and the definitions above are normative claims from published standards, not our interpretation. Verify them at the publisher, in the edition your contract invokes:
- AS9102 — SAE International (the standard's listing on the SAE International site): confirms the first-article inspection requirement and its three-form structure. It does not confirm any customer template's field numbering — confirm the edition your contract invokes.
- PPAP (Production Part Approval Process) — AIAG (the AIAG PPAP manual page): confirms that PPAP is a separate automotive approval package with defined elements and submission levels. It does not define the FAI forms — confirm the edition your customer specifies.
- ISO 9000:2015 — ISO (quality management fundamentals and vocabulary): the reference for terms such as a process whose output cannot be fully verified by subsequent inspection. It does not designate which of your processes are special — confirm the edition your contract invokes.
Scope note: field names above describe the AS9102-style three-form structure as buyers commonly apply it; they are not quoted field numbers or clause text. Confirm numbering, edition and any full-versus-partial trigger provisions against the copy of the standard your contract invokes. The worked example is illustrative and does not represent a customer part or an actual inspection record.
