Measurement Uncertainty and Traceability
A number that has no known uncertainty isn’t engineering data. A measurement result only becomes meaningful when its uncertainty is evaluated alongside the part’s tolerance: if the measurement system isn’t sufficiently capable relative to that tolerance, the result alone doesn’t support a reliable conformance decision. Common practice in measurement system analysis (MSA) is that measurement uncertainty should be a small fraction of the tolerance band — typically a 10% target, with 20–30% as an upper limit.
Temperature is one of the biggest factors driving uncertainty. The reference temperature for dimensional measurement per ISO 1 is 20°C; since aluminum’s coefficient of thermal expansion is roughly 23 µm/m·K, a one-meter aluminum part changes size by about 115 µm for a 5°C deviation. That’s why environmental control in precision measurement isn’t optional — it’s a requirement. Measuring equipment accuracy is defined through ISO 10360 acceptance and periodic verification tests; a measurement result needs to be traceable to national/international standards. We measure under controlled conditions, with equipment whose calibration is traceable.