Metrology is the science of measurement. In practical work, it helps answer three questions: what quantity are we measuring, how was the result obtained, and is the evidence good enough for the decision?
A measurement needs a defined purpose
Suppose a workshop checks a shaft diameter before assembly. Recording "25.00" is incomplete. The record needs a unit, the feature and location measured, an identified instrument and an appropriate method. Measuring a hot component immediately after machining may answer a different question from measuring the same component after it reaches the specified inspection conditions.
Start with the drawing or process requirement. Define the feature, measurement conditions and acceptance rule before choosing the instrument. Display resolution alone does not establish whether a tool is suitable.
Calibration, error and correction
Calibration establishes a relationship between reference values and instrument indications under stated conditions. It is distinct from adjusting an instrument and from deciding whether a product meets a specification.
In a hypothetical example, a reference has an assigned value of 25.000 mm and an instrument indicates 25.020 mm. With error defined as indication minus reference, the observed error is +0.020 mm. A corresponding correction would be -0.020 mm at that point under the applicable conditions. This example does not establish the instrument's behavior at every length or in every production setup.
Traceability connects results to references
Metrological traceability relates a measurement result to a specified reference through a documented chain of calibrations. The uncertainty contributed at each stage matters. A label or a laboratory name by itself does not provide the whole chain.
Traceability does not automatically make a measurement suitable for a particular tolerance. The uncertainty and conditions must also support the task. NIST makes this limitation explicit in its traceability policy.
Uncertainty and repeatability answer different questions
Repeating a measurement can reveal variation, but closely grouped readings may still share a systematic error. Measurement uncertainty describes the information available about the result; it is not the same thing as a known mistake or a product tolerance.
The uncertainty reported during calibration is one input to a user's measurement evaluation. Handling, temperature, alignment, the workpiece and the measurement method can add further effects during use.
What to record for the next decision
- Feature or quantity measured, unit and required limit.
- Instrument identity, relevant function and calibration record.
- Measurement method and conditions that affect the result.
- Readings, any justified correction and applicable uncertainty information.
- Decision rule, result and reviewer where required by the procedure.
These records make an unexpected result easier to investigate. They do not replace a validated method or an applicable customer, manufacturer or regulatory requirement.
Continue with the practical guides
- Reading uncertainty on a calibration certificate
- Calibration certificate acceptance checklist
- Planning dimensional calibration