What specifically must be calibrated per ISO 9001:2000?

[freelovefest]

The only time equipment needs to be calibrated under 7.6 is if it is used for the basis of acceptance or rejection. Do you use it to say yes or no with?

The other things that might need to be calibrated are process equipment used for special processes, or those requiring vaildation. If you use the temperature say in a plating process to ensure the bath is at a specific temp to ensure proper plating, then the temp. gauge would need calibration in that it is enuring a special process meets required results.

Does your process need validation? Is that why that gauge is there? To show you the temp. since it is a required process parameter? Is this temp called out in a work instruction? Chances are if you have a temp. on a peice of process equipment, it falls under the guidelines in 7.5.2, not 7.6

 

[Jim Wynne]

The only time equipment needs to be calibrated under 7.6 is if it is used for the basis of acceptance or rejection.

This is wrong. The standard says that calibration must be done when it's necessary "...to ensure valid results" when valid results are needed "...to provide evidence of conformity of product to determined requirements."

It might not be necessary to measure the parts (the product) in order to "provide evidence of conformity." Note also that it doesn't say proof of conformity, it says evidence of conformity. Thus a measuring device that is not necessarily used to make accept/reject decisions but is used to provide evidence of conformity, has to be calibrated.

 

[freelovefest]

This is wrong. The standard says that calibration must be done when it's necessary "...to ensure valid results" when valid . Thus a measuring device that is not necessarily used to make accept/reject decisions but is used to provide evidence of conformity, has to be calibrated.

I have to disagree, anything not being used to accept or reject would be considered process equipment, not inspection equipment.

You would need something calibrated to ensure valid results if it were part of a process where you can not verify that you are meeting requirements, otherwise it is used for inspection, which is to accept or reject.

If you are saying all process equipment needs calibration, that is very false. If someone has a foot counter that tells how many feet they are running an hour and uses that for improvement data, it in no way affects their inspection or their process, and does not need calibration.

I think we agree, we are just getting caught up in semantics, in that any measuring equipment you are using would have to be inspection equipment, or process equipment needed to show a process to be capable.

Everything else is truly "for reference only", unless you have an example? I can not think of anything else that would need to be calibrated, save inspection equipment, and validated process equipment, which needs calibration to ensure valid results. All monitoring equipment needing calibration is either for inspection, or process control in the case of a validated process, unless you have an example? I can't think of what else would need to be calibrated?

 

[Hershal]

One rule of thumb I used when I was at my previous employer (as many as 15 plants, and several were 9K) was this.

If there is a quantity provided by the instrument that means something, then calibrate the instrument.

If the instrument provides a quantity that means nothing, get rid of it.

If it is solely an indication or similar (similar may be like paint samples to judge color), don't calibrate it because there is no value in it.

When I took the occasional CAR, I required the external auditor to provide reasonable evidence that in fact the item required calibration, unless it was obvious. An example of that is a measuring tape.

The advantage I had is that it is EXTREMELY rare to find a 9K auditor who is also a metrology professional, so he has little or no technical basis for arguing.

Hershal

 

[jrcook5]

What SPECIFICALLY must be calibrated?

One rule of thumb I used when I was at my previous employer (as many as 15 plants, and several were 9K) was this. Hershal

Doesn't a "rule of thumb" muddy up the waters again when discussing specifically what must be calibrated. I have been trying for years to find a fact based definition of what must be calibrated as required by standards such as 9k and CFR 820.72. Discussions such as this one get close when a few contributors detail their understanding as to what is required to be calibrated per the standards as measuring devices used to accept or reject product or measuring devices on a validated process. We start to get meaningful dialog and then before a definitive criteria can be agreed upon somebody comes out and says we should calibrate if it affects quality or form, fit, or function or some other incomprehensible statement such as the following quote. Sorry, if my tone may seem unacceptable to some.

If there is a quantity provided by the instrument that means something, then calibrate the instrument.
If the instrument provides a quantity that means nothing, get rid of it. Hershal

What if maintenance uses the device for troubleshooting???

When I took the occasional CAR, I required the external auditor to provide reasonable evidence that in fact the item required calibration, unless it was obvious. An example of that is a measuring tape.

Am I missing something? This is not obvious to me?

The advantage I had is that it is EXTREMELY rare to find a 9K auditor who is also a metrology professional, so he has little or no technical basis for arguing.

Are we arguing to win or to understand the requirements and implement the best possible system?

 

[mjones2]

I agree with Hershel

I really liked your answer - this is a no nonsense approach. Everyone struggles with the answer - does it affect the quality -

What happens if the gage is removed - gets to the bottom of it - fast.

MJones

 

[Hershal]

JR Cook,

I take no offense to the tone, it is a struggle.

As for troubleshooting, the quantity then means something, otherwise you have no idea if a part is good. Calibrate it, though you may be able to go to a longer cycle, based on use.

Measuring tapes (in my opinion) GENERALLY do not require calibration, and unless there is a SPECIFIC reason, I consider it a waste of time, money, and effort. There may well be similar instances in any organization.

As for the arguing part, I would sometimes take CARs that were unjustified.....in those cases, I made the auditor justify the CAR.....and if they could not qoute "chapter and verse" or had a solid technical reason for the CAR, they had a choice: remove it or leave a copy of the complaints/appeals procedure because I assured them that I would be filing a formal complaint immediately for an unjustified CAR.

Hope this helps explain a little better.

Hershal

 

[Helmut Jilling]

puzzling

It is puzzling to me why this seems so complicated. A lot of comments in this thread do not quote the standard, but rather ideas, past practices, opinions, etc.

The ISO based standards don't talk about process gages, references gages, doing inspections, etc. They simply say, (paraphrase) if it is important that you know the results are accurate and reliable, then the gage has to be validated (calibrated).

As I write this, my watch says it is 6:53 pm. But, it could be off by 2, 5, 10 or more minutes.

If it is important that I know that the time is accurate, so I can make important decisions, then I have to calibrate my watch. If it is not important to be accurate, then my occasional check against the weather channel suffices.

It doesn't matter whether I label it "reference" or whatever. It matters whether accurate readings are needed.

So, if a dimensional measurement determines whether the product is accepted or rejected, accurate results are required, so the gage needs to be calibrated.

If an accurate process temperature reading is important (heat treat), then calibration is required. If it is not important, it may be wrong, but it doesn't matter. You could just as soon cover it up, because it is not known to be accurate.

If a scale, or linear counter is measuring product, and has to be accurate, then it has to be calibrated.

If a setup guy is setting fixtures to certain dimensions, then it has to be calibrated.

As the standard says, If it is necessary that "results are valid," then it has to be calibrated. This seems pretty clear, if you stick to the actual words.

 

[freelovefest]

If a setup guy is setting fixtures to certain dimensions, then it has to be calibrated.

As the standard says, If it is necessary that "results are valid," then it has to be calibrated. This seems pretty clear, if you stick to the actual words.

I have to agree, things are certainly getting complicated, but when you are auditing/being audited, it is important to have a good understanding of things to know how to proceed with people that may demand you to do something that should not be required of you, and that may represent a significant cost to your company's bottom line, like calibrating set-up tools.

While I agree this is a good idea, and not really an unreasonable request, the tools for set-up most likely would not require calibration in that they are not ensuring valid results, the inspection equipment used to measure the output of the operation they are setting up would do that. Other wise set-ups of machines would turn into some insane complicated process, and I have seen auditor's try to make some companies do this which was largely to the detriment of the bottom line, and in no way a requirement.

I would most likely inspect anything coming off the line after setting up. If I didn't inspect it after set up, then it would have to mean I could not verify I met requirements, in which case the process I am setting up would need to be validated to show it meets requirements.

I really have to stick to the point, and I can think of no exeption, that only things used for inspecting( saying yes or no), or ensuring a process meets the set requirements (in place of an inspection since one would be impossible) need to be calibrated.

If anyone has an exception to this let me know? What else would require calibration? Is there any equipment some one can think of requiring calibration that is not an inspection tool, or required in a validated process? Any input would be appreciated in that this is how auditors learn I think, discussion from examples.

 

[Helmut Jilling]

I really have to stick to the point, and I can think of no exeption, that only things used for inspecting( saying yes or no), or ensuring a process meets the set requirements (in place of an inspection since one would be impossible) need to be calibrated.

If anyone has an exception to this let me know? What else would require calibration? Is there any equipment some one can think of requiring calibration that is not an inspection tool, or required in a validated process? Any input would be appreciated in that this is how auditors learn I think, discussion from examples.

If you are using a micrometer to measure something, you must be assuming it is accurate or there would be no point in using it. It would be a waste of time.

My thinking is not limited to the minimum requirements I can get away with. We have to consider what is best for the business.

For example, most setup gages are usually simple gages. If so, calibrating them is not a burdensome cost. Why would you want to spend 4 hours setting up a die, only to find that the setup is wrong when you inspect the first parts. Compliant, perhaps, but how outdated?

Now, if the gage is expensive to calibrate, then use your judgement. I encourage my clients to do what makes sense for the business, not just the minimum the standard allows. Why should we focus on the standards anyway?

There are many examples. If we adjust a process based on temp gages, they should be accurate, they should be validated.

If we measure fixtures in the tooling dept., they should be accurate, they should be validated.

If we use the results of a counter to know when to stop a process, it should be accurate, it should be validated.

If we use parallel gages to setup a fixture, they should be accurate, they should be validated.

As the standard says, if the results have to be valid, it should be calibrated. My clients all do this without any issue, difficulty, or arguement, because it makes sense. If there is a particular gage where it doesn't make sense, then it might not be required. Use some judgement and common sense.

You don't calibrate because it is a gage, you calibrate when you need to rely on the results of the reading. Don't change the meaning of the standard.