What a pressure film calibration certificate should tell you

A pressure film arrives with a datasheet that says ±5 % of full scale, and a certificate that repeats it. Neither tells you whether the number applies to the way you are loading the film. That is the problem with most calibration paperwork: it documents a test that may not be yours.

Five things a usable certificate states

Loading geometry: point load or area load. Reference standard, and when that reference was itself calibrated. Uncertainty, with a coverage factor. The temperature at which the calibration was done. And which sheet — a serial number, or "a representative sample". A certificate with a percentage and none of these is a claim dressed as a measurement.

Loading geometry is where most of the error lives

A film characterised by pressing a small indentor onto it under a universal testing machine is characterised for a point load. Press the same film between two flat plates over half a square metre and the physics has changed: load spreads through the layer instead of concentrating under an indentor, the sealed edge takes a share of it, and the sheet's own flexural stiffness redistributes pressure near its boundary. Every number can be right and the application still be wrong. This is the same geometry problem that makes inserting a film into a clamped joint change the interface it measures.

For surface measurement the calibration has to be done with area loading — a known pressure applied across the whole sensing area, usually by a pressure bag or a calibrated press with a rigid platen, with the film supported the way it will be supported in service.

Per sheet, or per batch

Two films from the same production run are not the same film. Sensing-layer thickness, particle distribution and adhesive thickness all vary within a tolerance that is generous compared with the accuracy you are trying to claim. Per-sheet calibration removes that term from the budget. Batch calibration is defensible if the batch spread has been measured and is smaller than the uncertainty you are willing to accept — note that this is a statement about a measurement that was performed, not a promise that the spread is small.

Traceability, in practice

Traceability means an unbroken chain of comparisons back to a national standard, each link carrying its own uncertainty, combined as a root sum of squares and usually quoted at a coverage factor of two. In China that chain normally runs through a provincial or municipal institute of metrology and up to the national institute.

One practical detail is worth knowing before booking a slot: the verification regulations in force are written for pressure gauges and transmitters, not for flexible arrays. There is no standard procedure for a 490 × 520 mm sensing sheet. What an institute can do is calibrate the reference transducer and the loading fixture; the film is then compared against that chain, and the report describes the setup as much as the device. Bring the fixture, agree the loading plan in advance, and expect a report in which the interesting part is the method section.

The check that costs nothing

Between formal calibrations, run a single-point check before each campaign: one known load, one recorded deviation, one line in the file. It catches the failures that actually happen — a creased sheet, a damaged seal, a connector that has been pulled, a channel that has drifted — and it takes ten minutes. A certificate dated eleven months ago catches none of them.

What our certificate does and does not say

Ours is a factory calibration at room temperature, with area loading, recorded per sheet. It is not a third-party verification. If a project needs one, we either accompany the sheet to an institute or release the factory record under NDA. The distinction matters when a customer's quality department asks for "a calibration certificate": the document they are expecting and the document we issue can carry the same title and a different weight.

Four questions worth asking any supplier

Was the calibration done with point loading or area loading? What is the reference, and when was it last calibrated? What is the expanded uncertainty, at what coverage factor? And is this certificate for this sheet, or for a sample from the batch? The answers sort suppliers faster than any datasheet comparison.

Every argument about whether a pressure map is right eventually reduces to this: compared against what, under what loading, at what temperature, with what uncertainty. The map is only as good as that chain. The film is the easy part.