Two coordinate measuring machines — a ZEISS CONTURA and a TESA MICRO-HITE 3D — cover the whole flow: first article, in-process control and the dimensional report that ships with the batch.
A figure without an uncertainty is not a measurement but an opinion. Every instrument has an uncertainty of its own, and it is subtracted from the tolerance it verifies: the band in which a part may be declared good is narrower than the band on the drawing. Choose a tolerance and an uncertainty and see what is left.
Calculation per ISO 14253-1, with a symmetrical expanded uncertainty. The values are your own choice; they are not company figures.
A part measured exactly at the tolerance limit may in fact lie outside it — by as much as the instrument’s uncertainty. ISO 14253-1 settles this with a guard band: only parts inside the tolerance by at least the measurement uncertainty are accepted. Rigour has a price: part of the tolerance is lost for good, and it is lost from the supplier’s side.
Shop convention asks the instrument to be several times more accurate than the tolerance it verifies — ratios between 4:1 and 10:1 are the ones usually cited. Below 4:1 the measurement consumes so much of the tolerance that good parts are rejected and borderline parts accepted. The figure above shows exactly where you stand.
A serious dimensional report does not give the measured value alone: it gives what it was measured with, at what temperature and with what uncertainty. Without those three, the figure cannot be compared with a measurement made at your end — and when the two disagree, there is nothing to settle it on.
The make, model and declared uncertainty of our coordinate measuring machines are not confirmed in this site’s data. Until they are, they appear described generically, and the demonstration here is about the relationship between tolerance and uncertainty, not about our instruments.
Every dimension on a drawing refers to the part at 20 °C — that is the convention set by ISO 1. Under it, temperature is not a laboratory detail: at close tolerances it shifts the result more than the measuring machine does.
| Material | Expansion α [µm/m·K] | Drift at 200 mm and +5 K |
|---|---|---|
| Unalloyed steel | 11.5 | 11.5 µm |
| Austenitic stainless (1.4301) | 16 | 16 µm |
| Brass | 19 | 19 µm |
| Aluminium | 23 | 23 µm |
| POM | 110 | 110 µm |
Reference values for the materials we machine. The drift is a simple product: ΔL = α × L × ΔT. At 200 mm and 5 K above the reference temperature, the drift in micrometres equals the coefficient itself — a rule of thumb you can check in your head. The coefficient follows the structure rather than the family name: ferritic and martensitic stainless steels expand almost like unalloyed steel, not like 1.4301.
A machine shop is not a metrology room. Five degrees above 20 °C is an ordinary summer day, and on a 200 mm aluminium part that moves the dimension by 23 µm. A tolerance of ±0.02 mm is a 40 µm window: more than half of it is spent on temperature before the machine has made any error at all.
The scales of a measuring machine are referenced to 20 °C. If the part and the reference expand alike and sit at the same temperature, the two elongations subtract and the result stays close to correct even at 25 °C. With aluminium on the same machine the difference remains: 23 minus 11.5, that is 11.5 µm over 200 mm. The drift does not disappear; it halves.
After cutting, a part is warmer than the room, and a dimension taken straight away measures the leftover process heat rather than the part. Equalising takes time, and the time depends on mass and material — which is why it belongs in the flow as a step, not in an eyeball judgement on each piece.
If the tolerance drops below 0.02 mm over lengths beyond 100 mm, or the material is aluminium, brass or plastic, the measuring temperature has its place on the drawing. Written there, the measuring condition becomes part of the agreement and both sides arrive at the same figure. Left unwritten, both measure correctly and arrive at two.
Description and park composition confirmed by METALPET. Envelope figures are confirmed against the machine plate before quoting.
Coordinate measuring machine for incoming, in-process and final inspection, with repeatable measuring programmes.
Three-dimensional coordinate measuring machine for accurate checking of dimensions and geometry: points, diameters, distances, angles, positions, flatness, perpendicularity and other geometrical characteristics. By comparing results against the drawing or the CAD model it supports incoming inspection, intermediate checks and final inspection. Measuring programmes bring traceability and repeatability.
Confirmed by METALPET. Description and park composition confirmed by METALPET. Envelope figures are confirmed against the machine plate before quoting.
to be confirmed The measuring volume and probe type of the ZEISS CONTURA. Make and model are confirmed by the client; the figures have not been supplied yet.
Description and park composition confirmed by METALPET. Envelope figures are confirmed against the machine plate before quoting.
Coordinate measuring machine for quick checks near the shop floor: validating parts and setting up processes.
Three-dimensional coordinate measuring machine for quick, accurate checking of parts in the inspection area or close to production. It determines dimensions, positions and geometrical deviations by probing part features along the three axes. The operating interface makes both single measurements and repeated inspection sequences straightforward: releasing parts, setting up processes and confirming conformity.
Confirmed by METALPET. Description and park composition confirmed by METALPET. Envelope figures are confirmed against the machine plate before quoting.
to be confirmed The measuring volume of the TESA MICRO-HITE 3D. Make and model are confirmed by the client.
Send us your drawing. We respond with a quotation within 48 hours, and your data stays confidential.
Send us your drawing