Test Equipment, Calibration, and Measurement Uncertainty
This domain links equipment selection, calibration, and uncertainty. Calibration establishes the relationship between an instrument's reading and a reference standard; it does not, by itself, adjust the instrument. Adjustment changes the output; verification confirms it still meets requirements. Understand traceability - an unbroken chain to a national/international standard. Know as-found versus as-left data and the concept of calibration interval (set by stability, usage, and criticality). Out-of-tolerance findings trigger an impact assessment on prior results.
Measurement uncertainty is the heart of this domain. Sources include the reference standard, resolution, repeatability, environmental effects (temperature), operator, and fixturing. Combine them (commonly root-sum-square for independent sources) to get combined standard uncertainty, then apply a coverage factor k (k=2 ≈ 95% confidence) for expanded uncertainty. Traps: (1) Reporting a single reading without uncertainty. (2) Confusing resolution with accuracy - a fine display does not mean an accurate gauge. (3) Forgetting guard-banding: when uncertainty is large relative to tolerance, decide using a decision rule (simple acceptance vs. guard-banded acceptance). (4) Treating calibration as a one-time event.
Rule of thumb: expanded uncertainty U = k·u_c; report it with the result and state the decision rule. Worked example: tolerance 10.00 ± 0.05 mm; measured 10.048 mm; expanded uncertainty U = 0.010 mm (k=2). Under simple acceptance the part passes (10.048 < 10.050), but the upper limit minus U = 10.040, so under guard-banded acceptance it is indeterminate - flag it and apply the documented decision rule rather than silently passing it.