What problem it addresses
A calibration gives the state of the standard at one moment. Between calibrations it can be damaged, drift or be affected by a change of procedure. The control chart monitors that continuity with real-use data, without waiting for the next calibration.
How it is built
- Choose a stable check artifact (it can be the working standard itself or a control item measured with it).
- Measure it under repeatable conditions and regularly, with the same procedure.
- Calculate the center line with the mean of an initial history, along with the standard deviation of the process.
- Set the limits: by convention, a warning limit and an action limit, expressed as multiples of the standard deviation on each side of the center line.
- Plot each new measurement in chronological order.
How to read the signals
- A point outside the action limit: investigate before continuing to use the standard.
- Several consecutive points on the same side of the center line: they suggest a level shift.
- A continuous upward or downward trend: it indicates drift.
- Greater spread: points further apart point to noise, contamination or wear.
What the chart does not do
It does not prove that the value is correct, only that it is behaving as it did before. If the standard underwent a change just before the history began, the chart will treat it as normal. For this reason the center line is checked against the value on the latest certificate.
Additional uses
The accumulated standard deviation feeds the repeatability and stability component of the uncertainty budget. The drift slope, if stable, helps decide the interval between calibrations or apply a correction over time.
What an engineer sees in practice
A simple graph with dates and values, the center line and two pairs of limits. Discipline matters more than sophistication: always measure the same way, record who measured and under what conditions, and define in advance what to do when a limit is exceeded.