Home   |   Technical Articles

Technical Articles

Gauge Wear as a Control Chart: Planning Rechecks by Insertion Count, Not by Calendar

2026-11-18

Calendar-based rechecking is a habit, not a control strategy. A gauge used twenty times a month and a gauge used two thousand times a month both get their annual sticker, and only one of them is anywhere near its wear limit. Wear is caused by rubbing, so the honest axis is the number of insertions.

The bias is predictable in direction. A go gauge that wears smaller starts passing parts that used to fail. A no-go gauge that wears smaller starts rejecting parts that used to pass. Both errors look like normal variation to the operator.

Setting the axis

Start with the tolerance that matters: the difference between the go and no-go limits, which for a socket gauge is often a few hundredths of a millimetre. A common planning rule is to recheck when the gauge has consumed ten percent of that band, or every 1,000 insertions where the usage rate is unknown.

CheckpointInsertions since last checkWhat is measuredAction if exceeded
Baseline0 (new or after refurbishment)Diameter at three heights, two directionsRecord as the reference value
Routine1,000Same six readings, plus lead-in chamferRe-measure monthly instead of quarterly
Warning5,000Same, plus surface roughnessOrder a replacement, keep using with tighter sampling
Limit10,000 or 10 % of band consumedFull dimensional reportWithdraw from production

The numbers are planning values, not a standard: what makes them work is that they are written down before the first reading is taken. A limit chosen after the fact is a justification, not a control.

What to plot

Plot the deviation from the baseline, not the absolute diameter. Absolute numbers hide the trend in the rounding of the instrument. A control chart with a centre line at zero and limits at plus and minus the wear allowance will show a worn gauge as a one-way staircase, while a chart with limits at plus or minus the full tolerance will show nothing at all until it is too late.

  • One point per check, worst of six readings. Averaging six readings hides the ovality that matters.
  • Different charts for go and no-go. They wear in different places, at different rates.
  • Mark every refurbishment. A re-polished gauge is a new baseline, and mixing the two series fakes a step change.
  • Record who measured. A 2 µm step between operators is usually a technique difference, not wear.

Why the calendar still has a role

Insertion count misses two things: corrosion and accidental damage. A gauge stored in a humid cabinet can lose its surface finish without being used, and a dropped gauge can go out of round in one event. That is what the annual (or six-monthly) full recheck is for. The counting chart handles wear; the calendar handles incidents.

Both need a measurement that is traceable to the same reference, at the same 20 °C, with the same instrument. A gauge whose history is a set of numbers from three different micrometers has no history at all. The dimensional side of this is covered in our socket and plug gauge range, and the handling rules that keep a gauge in calibration longer are in the technical notes.

Standards referenced: IEC 60884-1 Annex A (gauges), ISO/IEC 17025 clause 6.4 (equipment), ISO 14253-1 for decision rules.

Turning the chart into a purchasing decision

A gauge that reaches its wear limit after 8,000 insertions at a plant running 2,000 insertions a month needs replacing every 4 months. That number, not the annual calibration date, is what belongs in the maintenance budget.

Two gauges in rotation, one at 4,000 and one at 0, smooth the replacement cost and keep a spare available.

The rate matters as much as the limit. A plant running 2,000 insertions a month wears a gauge by about 0.003 mm per quarter on the major diameter; against a 0.01 mm allowance that is 3 quarters of useful life, which is why the 1,000 insertion checkpoint finds the trend long before the annual certificate does.

close
Scan the qr codeClose
the qr code