Are you certain about your measurement uncertainty?

When ordering goods and services, it is the norm to enter into some kind of contract review, or at least as a minimum, have an expectation of the minimum deliverable quantity, quality and specification.

Do you apply similar criteria when engaging suppliers for calibration and preventative maintenance work?

Let’s look at a typical example. You have an ageing profile projector, you “need” an annual certificate of calibration to show to your BSI auditor, thus remaining compliant for your QMS. You make turned parts for the aerospace industry and check a particularly tricky part on the Shadowgraph, which has a 2mm radius against a shoulder. This dimension has a tolerance of 0.050mm, or a couple of thou in old money. How accurately do you need the profile projector to be? If you carry out a simple gauge capability study on that instrument with a 10x lens fitted, you would typically find you are already using up around 20-30% of your permissible limit. This reduces your wriggle room further. You now have 70% of 0.050mm left.

When you arrived and put the kettle on in the morning, the inspection area was 15 degrees, Jim then comes in and sticks the heater on full, it soon climbs to a balmy 25 degrees. The change in temperature of both the measuring instrument and component degrade your measuring capability further by another 10%. The typical measuring uncertainty employed when calibrating a profile projector is around 5-10um, so that’s another 20% of your tolerance. Some providers will offer 10um and some 5um or less. You have now lost 50-60% of your overall limits, just in uncertainty. In other words when you measure the radius at 2mm dead, it could be 2.025mm or + 1 thou, or under by 25um.

The better or lower that calibration uncertainty, the more certainty you have in the result and the greater control you get back when making corrections in your manufacturing processes. Ask your calibration provider what their measuring uncertainty is……………….