Stamped Automotive Parts: A 7-Point Incoming Inspection Checklist
-
The 7-Point Checklist
-
1. Check the part number and revision level before touching a gauge
-
2. Match the material certificate to the heat number
-
3. Measure the features that affect function, not just the overall dimensions
-
4. Look at edges and surfaces with a 2x loupe and your fingers
-
5. Verify PPAP status and open deviations
-
6. Do a trial assembly with a real mating part
-
7. Lock down traceability and retain a sample
-
1. Check the part number and revision level before touching a gauge
-
Final notes
I'm a quality/compliance manager at NTN. I review stamped automotive parts before they go to customers—roughly 200 part numbers a year. In 2024, I rejected about 4% of first production runs. The annoying part: most of those weren't exotic metallurgy or impossible tolerances. They were missing revision levels, unreadable traceability marks, or burrs on a sealing surface.
This checklist is for receiving inspectors, quality engineers, and anyone who has to approve a batch of stamped parts for a vehicle program. If you're dealing with brackets, heat shields, flanges, yokes, or fuel filter brackets, these are the checks I'd run before you sign off. It's seven steps, in the order I use them.
The 7-Point Checklist
1. Check the part number and revision level before touching a gauge
Part number alone is not enough. Stamping dies get modified. Engineering changes happen. I've seen 3,000 brackets get rejected because the drawing said Rev C but the tool was still Rev B. Same part number, same general shape. The only giveaway was a missing hole in a mounting tab.
I didn't always check revision levels. I do now. That rejection cost us a redo, a late delivery, and a very quiet meeting with the program manager. The lesson: five minutes of verification beats five days of correction.
Also check the physical marking. If the drawing says the part carries a brand or traceability mark, make sure it's there and legible. With our stamped parts, the NTN logo is not just branding—it's a route back to the batch and tool. If a logo or lot code is partially worn, treat it as a red flag, not a detail.
2. Match the material certificate to the heat number
An MTR with the right grade means nothing if the material on the floor is from a different heat. Match the heat number on the certificate to the heat number on the coil, pallet, or bundle tag. It's a 30-second check that I used to skip. Now it's the second step.
Stamped parts are sensitive to material thickness and properties. On a 2.0 mm bracket, a 0.1 mm thickness shift can be within typical tolerance but still change springback and hole alignment. The drawing and the material certificate should agree.
For general tolerances, many drawings reference ISO 2768-1:1989. But for structural or safety parts, don't rely on general tolerances—verify the special characteristics on the drawing.
3. Measure the features that affect function, not just the overall dimensions
Overall length and width are easy to check, but a part fails in holes, slots, flanges, and flatness. A wheel bearing flange can look fine on the outside and still have a bolt circle that's off by 0.2 mm. That shows up as a bearing that doesn't seat evenly, which later becomes noise or premature wear.
Use a CMM or hard gauges for hole center distances, edge margins, and flatness. And measure from the datum structure on the drawing. If Datum A is the large face, don't measure from the bottom edge because it's easier. I've seen parts pass an internal bench check and then fail assembly because the wrong datum was used.
4. Look at edges and surfaces with a 2x loupe and your fingers
Burr height is one of those things everyone says they check, but too many accept. A stamped heat shield for a car cylinder head might look like a simple cover. But a small burr on a mounting edge can cause a buzz at idle when it touches a neighboring component. On a sealing surface, a burr is a leak path.
Use a burr gauge if the drawing specifies a limit. If it doesn't, get a sample from the supplier and document it as your acceptance standard. Don't assume no specification means no control. (I've learned this from a returned part that looked fine in a photo and failed in my hand.)
Also watch for die marks and scratches. A sudden change in surface appearance often means the die was polished out of cycle, or the supplier moved the job to another press. That's worth a question.
5. Verify PPAP status and open deviations
For production automotive parts, there should be a PPAP behind the part. Per the AIAG Production Part Approval Process manual (4th edition), a Level 3 submission includes a full dimensional report, material test report, capability study, and process flow. That's not just vendor paperwork; it's your baseline for what an approved part looks like.
Ask about temporary deviations. A deviation without an expiration date is a process change in disguise. If a supplier says 'the customer approved it,' get the approval in writing.
Here's something vendors won't tell you: after the first order, production tends to drift toward the cheapest setup. Under IATF 16949, moving a die to a different press can require a new run-at-rate or capability study. Part of receiving inspection is making sure the process you approved is the process they're running.
6. Do a trial assembly with a real mating part
This is the step most people skip. Individual dimensions can all be within tolerance and still stack up wrong in an assembly. You have to physically see the part with its mating components.
Example: if you've ever searched 'where is the fuel filter located' and found a stamped bracket under the car, you know the bracket needs to hold the filter securely. If the hole spacing is off, the strap doesn't pull tight, and the filter moves. The gauge can pass; the car tells you the truth.
For driveshaft components, trial assembly matters even more. A flange from NTN Driveshaft Inc. might have perfect hole diameters on paper, but if the pilot bore and the bolt circle don't line up with the output shaft, it will bind. No CMM report catches that. A ten-minute assembly test does.
I learned this the hard way. Early in my career, I assumed 'same CAD model' meant identical results from two suppliers. Didn't verify. Turned out one supplier had a springback compensation step and the other didn't. The mating gap was 2 mm. Now I never assume the CAD file tells the whole story.
7. Lock down traceability and retain a sample
After you approve a batch, keep one physical sample with the CMM report, surface results, and a photo of the approved mark. Label it with the lot code, date, and inspector name. If a field issue appears nine months later, this sample is the difference between a one-hour answer and a one-week investigation.
Traceability marks are especially important for safety-related parts like wheel bearing flanges, steering yokes, and brake brackets. If the lot code is missing, treat the lot as unapproved. Perfect dimensions don't matter if you can't trace the material and tooling history.
Final notes
The common mistakes I see on incoming inspection are the same every time: skipping the revision check, accepting a material cert without matching the heat number, measuring only overall dimensions, ignoring burrs on edges that 'nothing touches,' treating PPAP as a file so it stays locked, and never trying the part in a real assembly.
Bottom line: five minutes of verification beats five days of correction. I've had to own that mistake myself. The checklist looks simple, and it is. That's sorta the point.
Leave a Reply