BACK TO BASICS
The effect of faulty measurement on previously produced products
by Scott Stamm
A measuring instrument, gage or device failed recalibration, and you have been asked to determine the influence on previously produced products. Where do you start, and what do you do?
To begin with, understand why this could be a problem. If a gage or instrument is out of calibration that was used to measure a product, all products made since the last known successful calibration are suspect and may need to be recalled.
Additionally, ISO 9001:2000 and ISO 9001:2008 (clause 7.6) and the U.S. Food and Drug Administration’s quality system requirements, among other standards and regulations, require determining an out-of-calibration gage’s or instrument’s influence on previously produced products.
Step by step
Start investigating the potential harm to previously produced products. First, determine if any actual products using this measurement device have been made since the last known successful calibration. In large organizations with many SKUs, not all products are regularly manufactured. Check this before investigating how the measurement device failed and its influence on previously produced products.
Next, while reviewing the influence on previously produced products, determine how the device was discovered to be out of calibration. Was the out-of-calibration status determined because of a scheduled recalibration, or was it the result of a specific event, such as dropping the gage on the floor?
If it was the result of a sudden event, it is possible the gage or instrument was not used after the event and there is no harm to products. This may seem obvious, but sometimes in large organizations, measuring devices are sent to the metrology area for inspection and recalibration with no explanation.
For the third step, carefully review how the gage or instrument is used and the characteristics it measures. This is especially true for functional gages that are specifically designed to measure a single aspect of a product. Intimately understanding how the measuring instrument, gage or device is failing is necessary, as it could lead to insights about the influence on previously produced products.
For example, point micrometers occasionally fail a calibration check because the tips have worn off the points. The failure mechanism is that the point micrometers understate the actual measurement. As such, any harm to products measured is to the high side of the tolerance.
If the dimension in question has bilateral tolerances, then the minimum tolerance side of the dimension is not in question—only the maximum tolerance. Look to subsequent dimensional checks, such as an optical comparator or overlay check, that may inspect the maximum dimension. In addition, compare the tolerances of the dimension checked with the gage with the amount the gage is out of tolerance. It is possible that even though the gage is out of calibration, the amount is still within the tolerances.
Taking it a step further
Additionally, check to see if the measurement device’s failure is in the actual use range. For example, torque wrenches are usually calibration-checked at the low, middle and high end of their scale in clockwise and counter-clockwise directions. If the torque wrench only failed calibration in the highest range in the counter-clockwise direction but is used only in the lowest range in the clockwise direction, there is no harm to previously produced products.
Another method used to understand any possible harm is by partial vendor product processing. The vendor may duplicate the failed measurement or a proxy of the measurement. In addition, if your organization has a receiving inspection function, the actual failed measurement or a proxy of the failed measurement may be performed upon receiving the product in-house.
When all else fails, another possible technique may help determine any harm to previously produced products: Examine an approved first article layout (FAL) of the supplied material, such as forging, casting or pre-finishing component. If the FAL is approved and proven, and nothing has changed since its approval, it may be possible to substitute the failed measurement using the FAL as a proxy or determine the minimums and maximums of a dimension in question.
This information should not be construed as ammunition to avoid a product recall, as a recall is sometimes necessary and appropriate. Rather, it is intended to help an examiner ensure a full investigation is completed.
- ASQ, ANSI/ISO/ASQ Q9001-2000, Quality management systems—requirements, 2000.
- International Organization for Standardization, ISO-9001, Quality management systems—requirements, 2008.
- U.S. Food and Drug Administration, 21 CFR 820—Quality system requirements, 1996.
Scott Stamm is a senior quality engineer at Zimmer Inc. in Warsaw, IN. He earned an MBA from the University of St. Francis in Fort Wayne, IN. He is a senior member of ASQ and is a certified calibration technician, quality manager, mechanical inspector, reliability engineer, biomedical quality auditor, Six Sigma Black Belt, quality engineer and quality auditor.