The first-article report reads "conforming," yet a measurement taken elsewhere comes back out-of-tolerance. Most of the time no one altered the number — the measurement conditions were simply different. Take a common case: for a thin-wall part, the supplier measured with a CMM under constant temperature, taking measuring points from the drawing datum, and the report read conforming; the customer re-inspected on the shop floor with conventional gauges, the part was deformed by clamping, and the datum was taken from a different edge — the result was out-of-tolerance. Neither side performed the measurement incorrectly; the problem was that one common set of conditions had never been agreed. (Scenario is a generic illustration, not referencing any specific project.)
Inspection data is not "read off" — it is produced under a set of conditions:
Observed value ≈ true dimension + the deviations jointly introduced by equipment, environment, datum, clamping, and acceptance rules.
So disputes are rarely about whether the data is true or false, but about conditions not being written into the agreement. The 5 variables below decide whether that value on the report can be trusted.
# | Variable | What it affects | What to ask the supplier for | What happens when conditions differ |
① | Measurement equipment & method | CMM, in-machine inspection, and conventional gauges each have their applicable objects and uncertainty | Report states the equipment & measurement method | The same part yields different conclusions on equipment of different accuracy |
② | Gauge calibration | A gauge outside its validity period gives a conclusion with no basis | Calibration records & validity period | Data cannot be traced; re-inspection has nothing to compare against |
③ | Measurement environment | Temperature & isothermal state affect the dimensions of metal parts and gauges | Ambient-temperature record; part fully temperature-stabilized | Thermal expansion & non-equilibration, especially pronounced on thin-wall parts |
④ | Datum & measuring-point definition | Datum, measuring points, and clamping define "where is being measured" | Drawing datum, measuring-point definition & clamping method | Not measuring the same location, so the numbers naturally differ |
⑤ | Acceptance rules | Which drawing revision, which key-dimension list, full or sampling inspection | Revision number, dimension list, acceptance criteria | Different drawing revisions give different pass criteria |
(General metrology & acceptance practice, not a national standard and not mandatory; specifics are subject to the bilateral technical agreement and the drawing. The table contains no tolerance values.)
Clamping deformation: clamping force itself is an error source; agree on the support method and keep measuring points away from deformation-prone zones.
Measurement timing: dimensions change after demoulding with cooling, moisture absorption, or stress relief; agree on the conditioning time and state.
Contact vs. non-contact: probing force and point-taking differ, especially on soft parts; run a correlation before switching methods.
(General process experience, subject to material-supplier data and the technical agreement.)
A first article is a single-point result; mass production looks at the trend. To make data comparable over time, achieve the "three consistencies": the same datum & measuring points, the same type of measurement method, and the same set of acceptance rules. When the gauge, the operator, or the environment changes, keep a comparison sample and run a correlation first before proceeding. Otherwise the "dimensional drift" on the trend chart may simply be a changed measurement method — and modifying the mould at that point is entirely inadvisable.
First check the drawing revision & key-dimension list — are we judging the same thing;
Then check datum, measuring points & clamping — are we measuring the same location;
Then check equipment, calibration status & environment — are we measuring under the same conditions;
Only when the first three are consistent do we discuss re-measurement, re-inspection, or concession acceptance — and write the conclusion into the record.
Field | Purpose |
Mould no. / cavity no. | Link to per-mould history |
Part no. & name | Identify the measured object |
Drawing revision | Basis of judgment |
Key-dimension list & tolerances | What to measure, judge to what |
Measured values | Result |
Equipment & calibration status | Conditions traceable |
Environmental conditions | Conditions reproducible |
Inspector, date & conclusion | Responsibility & effective timestamp |
Before mould making, write the key-dimension list, datum & measuring-point definitions into the technical agreement, and produce a co-signed controlled dimension list — don't wait until the first-article report to discuss "where should be measured";
Require the report to attach equipment, calibration status, environmental conditions & drawing revision, and standardize the report template — changing format every batch means historical data can't be linked up;
In mass production, fix measuring points for trend records, retain comparison samples, and produce a dimensional trend chart — change the method and the trend breaks.
One line: The number on the report is the result; the conditions that produced the number are the evidence. Agree conditions first, judge numbers second; with conditions absent from the agreement, "conforming or not" can always be argued either way.
Jingmu outputs first-article full inspection and inspection records per the agreed datum & measuring points, archived against the mould number; when the revision or measurement method changes, it supports correlation and record updates.
13724506670 
