First Article Inspection (FAI) — How Do You Run the FAI Process, What Do You Measure, and What Is the Acceptance Criterion for Every Dimension?
Last updated: 2026-09-03 · Reading time: about 6 min
First Article Inspection is the formal measurement of the first 3–5 consecutive shots from a stabilized mold. The acceptance criterion is blunt: every dimension on the drawing must fall inside its stated tolerance, measured and documented before the mold is released to production.
See also related article.
Key facts
<>
>
Warpage — Part deformation after ejection caused by uneven shrinkage. Amorphous resins warp less (0.4–0.7% shrinkage). Semi-crystalline resins warp more (1.0–2.5% shrinkage). Fix: uniform wall thickness, balanced cooling, optimized gate location, adequate packing pressure. See also related guide.
What is the FAI process — and what are the steps?
Four steps turn a fresh mold into an approved one. Each protects the measurement from a different source of error.
- Stabilize the process. Run 20–30 shots so mold temperature and melt temperature stop drifting. Lock and record injection speed, injection pressure, melt and mold temperature, packing pressure and cooling time — the FAI is only valid on a repeatable process.
- Collect 3–5 consecutive shots. Not random parts from different runs. Consecutive shots prove repeatability: if dimensions swing shot to shot, the process is not stable and the mold or settings need adjustment first. Label each part with shot number, date, time and mold number.
- Condition the parts. Semi-crystalline resins (PP, PA, POM, PBT) keep shrinking as crystallization completes — 48–72 h at room temperature, or anneal at 80–120 °C for 2–4 h. Amorphous resins (PC, ABS, PMMA, PS) stabilize in about 24 h. Measuring too early invalidates the report.
- Measure every dimension on the drawing. Overall size, hole positions, boss heights, wall thickness, flatness, parallelism, concentricity and surface finish. Record the measured value against nominal, tolerance and deviation for each one.
What are the measurement instruments — and when do you use each one?
Pick the instrument by the tolerance you must hold. A caliper cannot resolve a ±0.05 mm callout, and a CMM is overkill on a ±0.5 mm one.
| Instrument | Accuracy | What it measures | When to use |
|---|---|---|---|
| Digital calipers | ±0.02 mm | Overall dimensions, accessible wall thickness | Non-critical dimensions at ±0.1 mm or larger |
| Micrometer | ±0.005 mm | Wall thickness, small diameters | Critical dimensions at ±0.05 mm |
| Height gauge | ±0.01 mm | Step heights, boss heights, 1D hole positions | Dimensions referenced from a datum surface |
| Optical comparator | ±0.005 mm | Small features, thread profiles, radii | Features too small for calipers or micrometers |
| CMM | ±0.002 mm | Complex 3D geometry, true position, flatness, parallelism, concentricity | All critical dimensions and GD&T callouts |
The CMM is the arbiter. It measures any point on the part surface in 3D and computes true position, flatness, parallelism and concentricity from the point cloud. When a caliper reading disagrees with the CMM, the CMM wins — and its raw coordinate data is what settles a dispute with the customer.
“We do not release a mold on a good-looking part. Every FAI we ship carries a signed dimensional report — each drawing dimension, the instrument used, the measured value and the deviation from nominal. When a customer questions a number, we hand over the CMM data file, not an opinion.”
— Project Engineer, aCatMould
Frequently asked questions
>
What is the acceptance criterion — and what happens if a dimension is out of tolerance?
Every dimension must sit inside its drawing tolerance — out by even 0.01 mm fails that callout.
The fix depends on direction: an oversized feature means removing steel, an undersized one means weld or a new insert.
A modification runs $500–5,000 and 1–4 weeks, then the mold is re-tried and re-measured. A complex mold with 100+ dimensions often needs 2–3 FAI rounds before everything passes.
What is the difference between FAI and PPAP — and do you need both?
FAI is the dimensional measurement of the first parts. PPAP is the wider automotive package (AIAG) that wraps the FAI report with a process flow diagram, PFMEA, control plan, material certifications and sample parts.
FAI is a subset — the dimensional report inside the PPAP. If your customer requires PPAP, the FAI feeds it; if not, the FAI alone supports mold approval. FAI lands at T1 or T2; PPAP closes after the run-at-rate.
How do you handle shrinkage variation — and when should you measure the parts?
Shrinkage variation is the most common FAI failure. Measure only after shrinkage stabilizes — about 24 h for amorphous resins, 48–72 h for semi-crystalline, or anneal at 80–120 °C for 2–4 h to accelerate it.
Record the conditioning time and temperature in the report — without it the FAI is not valid. The spread across the 3–5 parts should stay within 20% of the total tolerance; wider than that means the process is not stable.
Key takeaways
- FAI is the formal measurement of 3–5 consecutive shots from a stabilized, documented process — never random parts.
- Measure every drawing dimension with the right instrument; the CMM (±0.002 mm) is the reference and settles any dispute.
- Condition parts first — 24 h amorphous, 48–72 h semi-crystalline — then pass every dimension or modify the mold and re-try.
The FAI is the most objective gate in mold approval — the CMM report does not argue. It is also the baseline for production: if parts drift out of tolerance later, you compare against the original FAI to trace the cause, whether that is wear, a process change or a material change.
Sources
SAE AS9102 — Aerospace First Article Inspection standard
AIAG — Production Part Approval Process (PPAP)
ISO 20457:2018 — Plastics moulded parts: tolerances and acceptance conditions
Zeiss — industrial metrology and CMM measurement
Mitutoyo — precision measurement instruments
Sources: Plastics Technology · Plastics Technology Online · Society of Plastics Engineers (SPE) · ThomasNet