1. Fixture geometry, measured against the drawing you approved
A CMM report on the fixture itself: datum faces, pin positions, pad heights, parallelism between faces. The reference is the approved drawing, not a verbal specification, which is why the drawing has to be signed before machining starts.
Pass criterion: every dimension inside its stated tolerance, with the actual value recorded, not just a tick. A report that only says “OK” tells you nothing about margin.
2. Clamp force per station
A hydraulic fixture has a pressure gauge, but pressure is not force: force depends on the cylinder area and the lever ratio of the clamp, and on multi-station fixtures the stations may differ. Measure with a clamp force gauge at each station.
Pass criterion: measured force within ±10 % of the design value at every station, at the working pressure stated in the schematic. If one station is 30 % low, a seal is leaking or a sequence valve is set wrong, and both will get worse.
3. Repeatability after re-mounting
This is the test most suppliers skip and the one that decides whether the fixture is useful in production. Machine a test feature, remove the fixture from the table, put it back, machine again, and measure the difference between the two results.
Pass criterion: 0.01 mm or better with a fixed zero-point interface; 0.02 mm with T-slot keys, which is the practical limit of key location. If the fixture is designed for 0.005 mm repeatability, this test is where it shows.
4. First article on a real part
Fixture geometry is not part geometry. A fixture can be perfect and still produce a bad part, because the clamp load distorts the workpiece. Measure the part, not only the fixture, with the part clamped at working pressure.
Pass criterion: every drawing dimension inside tolerance, and the distortion pattern documented — where the part moves when clamped, and by how much. That record is what lets you adjust the process six months later without starting from zero.
What to do when a measurement fails
Separate the three possible causes before accepting a fix: the fixture is wrong, the machine is wrong, or the process is wrong. If the fixture passes geometry and force but the part fails, the machine or the programme is the suspect, and no amount of fixture rework will fix it. A supplier who immediately agrees to rework without this separation is guessing.
| Check | Method | Pass criterion |
|---|---|---|
| Fixture geometry | CMM against approved drawing | All dimensions in tolerance, actual values recorded |
| Clamp force | Force gauge at each station | Within ±10 % of design value |
| Repeatability | Machine, remove, re-mount, machine | ≤ 0.01 mm zero-point, ≤ 0.02 mm T-slot |
| Hydraulic tightness | 1.5× working pressure, 10 min hold | No pressure drop, no seepage |
| First article | Real part, clamped at working pressure | All drawing dimensions in tolerance |
| Documentation | Report package | Geometry report, force log, trial-cut video, schematic |
Questions
Should I insist on a CMM report for a simple manual fixture?
Who pays for third-party inspection?
What if the fixture passes but the part is out at one dimension?
How long should acceptance take?
Published by Liuzhou Dongben Machinery Equipment Co., Ltd. · More notes