When To Calibrate - To Calibrate or Not To Calibrate

[Marc]

Date: Mon, 30 Aug 1999 08:05:19 -0400
From: Jack Gale
To: 'Greg Gogates'
Subject: RE: To Calibrate or Not To Calibrate

Douglas,

Our lab has been providing some useful tips to our customers for several years. We give them a series of questions to ask about the equipment and if any of the answers are "Yes", they should calibrate. We ask the questions from a user's point-of-view, not necessarily the lab.

You should calibrate IF:

1) The instrument is called out in a process FMEA or Control Plan;
2) any equipment used in incoming, in-process, or final inspection;
3) used in statistical process or quality control (SPC/SQC);
4) any equipment used to perform lengthy or costly machine setup (time and scrap);
5) any situation where the failure costs are too high (safety equipment);
6) equipment used during costly maintenance (basic costs and downtime);
7) equipment used in R&D where design decisions are made;
8) used during failed or returned material analysis, where critical determinations are made;
9) Is someone being paid to look at this instrument and act upon the reading provided?

Each organization should review these questions, determine the applicability and the potential impact of non-calibrated equipment and then publish these to the shop floor. You need the instrument owners onboard so that new equipment doesn't hide until audit day. If you have the shop floor looking out and getting new equipment in the system, you'll be much better off.

Once this is done, there is one left for the "Lab":

10) Equipment used for calibration/traceability for equipment addressed by items 1-9 above.

Jack Gale
ASQ-CQE
Essco Cal Lab

[Marc]

Date: Mon, 30 Aug 1999 06:05:02 -0700 (PDT)
From: James England
Reply-To: micky
To: Greg Gogates
Subject: Re: To Calibrate or Not To Calibrate

Doug,

The best guidelines for deciding what should be "Calibrated" vs "Calibration Not Required" In my opinion, is as follows:

If you use the instrument to measure, verify, and document compliance to a customer specification, the instrument used for this purpose must be calibrated, and placed on a recall cycle. This applies to ISO and QS-9000 registration. You may determine the accuracy and calibration frequency. Naturally you should follow guidelines established within the Metrology community in establishing these.

The "grey" areas seem to cause the most confusion and this is to be expected.

The single largest grey area in aproduction environment seems to be the process monitoring and control equipment. I would like to suggest you read Appendix C-1 of NCSL Recommended Practice RP-3, Calibration versus the Maintenance Process, for a well thought out approach to this subject. While Calibration and Maintenance are not the same, they cannot be 100% segregated and must work together to accomplish the intent of the standards.

I would be happy to fax a copy if you do not have one.

One might consider the following:

If the user of the instrument could "unknowingly" be collecting false data (quantitative) due to drift and or malfunction of the instrument, then the instrument MUST be certified on a cyclic schedule.

If the operater could detect an abnormal condition and take action to correct it, then you might not need to calibrate that instrument. You must document the reason. I suggest you run all such documentation through engineering for their concurrence.

Only instruments that might normally be calibrated, but due to the way they are used, are not calibrated need to be labeled as such, document this.

Example (simple) A Multimeter used only for continuity testing would normally require calibration but there is no quantitative data produced and only a pass fail indication, while this is a test, it does not require the use of a "certified" instrument. You are not required to calibrate this instrument on a cyclic basis.

I hope this helps, at least give you a starting point. Document, Document, and fulfill the "intent" of the standards and you'll do fine!!

James England
Metrologist,
Goodyear Tire & Rubber Co.