Quality Magazine

Other Dimensions: Calling the Shots

October 24, 2008

One of the good things about getting items calibrated is the information the process provides. The bad news is some folks don’t know how to deal with that information. Far too many buyers of calibration services only scan the reports and, in the absence of red flags, file them in case an auditor wants proof of calibration.

Whether you calibrate your own equipment or send it out, someone has to review the data in the reports. Such a review could indicate that while the gage or instrument is within your acceptance limits at this time, it is getting close to exceeding them, thus triggering a reduction in the time for the next calibration. It could mean you would be wise to downgrade this item so it continues in service for lower tolerance work.

If the item is beyond acceptable limits, it has to be withdrawn from service and the potential impact of this fact evaluated. The biggest problem in all of this: Who determines what is acceptable? All too often, folks expect the calibration source to make the call-an expensive and risky business. Others apply the ‘new’ gage or instrument tolerance as the acceptance criteria, which is equally problematic.

As an example, let’s take the trusty outside micrometer that the manufacturer claims is accurate within 0.00005 inch, or 0.001 millimeter. This is the brand-new, out-of-the-box accuracy. If you decide this will be your criteria for acceptance when it is calibrated later, you have now boxed yourself into a corner where only ‘new’ will do. A bit of wear leading to the instrument being accurate to, say, 0.000l inch, or 0.002 millimeter, may be more than suitable for your work. However, if you use the ‘new’ tolerance as the basis for acceptance, a perfectly good instrument gets thrown out.

Assuming your calibration source has all the new product specifications for equipment and you allow them to make accept/reject decisions on your behalf, mayhem and misery are sure to follow. First of all, they will apply new tolerances to used equipment, which, as noted earlier, wastes money. It also could trigger a product recall when one was not needed. More importantly, in many cases, they don’t really know what standard the instrument was made to in the first place.

The reason for this is the multinational nature of the manufacture of such equipment. Most micrometers sold in North America today are imports that may not match American standards. Many are private label products that may not even meet the standards of the country in which they were made. Accepting or rejecting used instruments or gages using new tolerances is bad enough, but if the wrong standard is being referenced it’s even worse.

The only one who should call the shots when it comes to tolerances on used equipment is the person or company using it. They alone know the application the instrument or gage is being used for and how critical it needs to be for that application or their work in general.

Since space is limited, I won’t get too deeply into one of the thornier issues in this discussion: the application of the calibration sources’ measurement uncertainty in making the accept/reject decision.

Most companies ignore the uncertainty attached to a given calibration unless it appears to be way out of line. They use the laboratory’s reading as the decision point to keep things simple. This can be a dangerous thing to do when the ‘reading’ is on the accept/reject borderline. When the uncertainty is factored in, an item that was accepted could just as easily have been over the line. In a similar situation, a reading that was just outside the borderline may trigger an unnecessary product recall.

If you are intent on having an outsider calling the shots on equipment suitability, make sure your corporate lawyer agrees to it. And be prepared to answer a lot of questions when a quality auditor starts reviewing your calibration records.

The targets should be created based on your needs rather than new product standards that may not apply to the situation at hand. After all, targets are not worth anything unless you have a reasonable chance of hitting them.