Hi, ;)
This is my first time asking for your help so I hope I'm giving you enough information for you to understand my problem. We are presently trying to certifiy to AS9100. Our calibration is mostly done in-house and we do not have specific tolerances for each of our instruments. We are a custom machine and fab shop and have over 3000 instruments of about 50 different types. I was wondering if anyone knows of a standard that could help me figure out what the acceptable tolerance for each instrument is. I am also searching for a standard that can give me the acceptable environmental conditions required for my metrology lab, i.e. temperature, humidy, lighting etc. Thanks,

Veronique

Well, hello there! Welcome to the Cove!:bigwave:

Are the instruments currently being calibrated? Hopefully, previous calibration certificates would help identify tolerances. Also, hopefully the departments using the instruments would have some assistance.

As a default, you can start with mfg. specifications. Also, if you have access to industry specifications like GIDEP, they will be of assistance.

A good standard may be ISO17025. Basically, you will want to provide an environmentally controlled facility to perform the calibration work. The temperature, humidity, vibration, etc. should be controlled to the point that it does not create undue influence with the calibration. That is not only the Unit Under Test, but also the standards used.

If you would like, aside from any responses here, the search function may be of assistance. We have some Covers who have provided invaluable input on contructing, building calibration labs, control, establishing tolerances, etc. That may be a good option for you.

How long has your area been doing internal calibration work? Has it been done to any other specification/requirement? That can also gauge how much work your procedures and stuff may need.

Hope this helps.

Thanks, Brad - nice advice, especially about the GIDEP information. This is a relatively overlooked/unknown source.

The issue of controlling the environment isn't strictly required. It is important to know, through monitoring, what the conditions (temp/humidity in this case) are and their effects on the cal. result. You don't necessarily have to provide env. controls, simply to decide what to do, should the environmental conditions exceed the spec - like suspending any cal. work.

The issue of controlling the environment isn't strictly required. It is important to know, through monitoring, what the conditions (temp/humidity in this case) are and their effects on the cal. result. You don't necessarily have to provide env. controls, simply to decide what to do, should the environmental conditions exceed the spec - like suspending any cal. work.

Good point, Andy. :agree1:That probably needed to be clarified.

This sounds like a Hershal issue. He's travelling (and inside good old Arkansas as I type this). Hershal will probably drop in to visit the Cove tonight and will most likely jump in.

Another source to look at is NCSLI.....they have Recommended Practices (RPs) that will provide guidance on environmental conditions, calibration cycle establishment, and other laboratory management and technical issues, even one on preparing for accreditation.....

Also as has been mentioned, GIDEP is a good source for procedures and related.....

Hope this helps.

So if I get this right, there is nothing absolutely defined for calibration tolerances and environmental conditions in calibration labs that I have to respect. All I have to do is justify that what we are doing for the moment respects our customers' and manufacturers' specifications. Correct me if I am wrong. Thanks alot for your help guys.

Véronique
:thanks:

So if I get this right, there is nothing absolutely defined for calibration tolerances and environmental conditions in calibration labs that I have to respect. All I have to do is justify that what we are doing for the moment respects our customers' and manufacturers' specifications. Correct me if I am wrong. Thanks alot for your help guys.

Véronique
:thanks:

Veronique:
If I read your post correctly, it's not quite like that. There are some well-established principles of tolerances on equipment that you will be calibrating, so you have to show that your calibration methods are capable of meeting those tolerances - this is why, for example, measurement uncertainty is important.

If we misled you on the environmental side, a point of clarification is that you don't have to invest in a complex control of the temperature and humidity, if you can show that you monitor conditions, that you can show what effect it has on the calibration result and, then, that you take action(s) when the conditions begin to have some quantifiable effect - like if the temperature got above a certain limit, based on a standard of 68F, let's say.

C'est tout!

Ca Va?

Oui, Merci !

Spot on, Andy!

Verod, the accuracy of the instruments being checked is important. You are wanting to verify the equipment is operating properly, and work correctly for your needs.

As to the calibration lab, what Andy is saying is you are not mandated to go out and spend a fortune on developing some state-of-the-art lab. What you are required to do is control for environmental factors that may be significant (if there are any). Hence, Andy's point about setting up a recorder.

for example, if you are calibrating micrometers/calipers, then the gauge blocks (and Unit Under Test) will be affected by temperature and humidity. However, depending on the tolerance/classification, you may be OK with your current situation. If you already have a room that is monitored and the fluctuations are within the requirements for that class; you're fine. If it is not, then accomodations are in order to remove undue influences. Make sense? And since there are so many different influences, and so many different tolerances/classifications, it's just not accurate to make blanket statements about requirements.

It's less of a Shou Shalt..., and more of a getting control of your process, and assuring that all significant variables (noise) are controlled (minimized) when performing calibrations.

Is there any standard that can help me with tolerance/classification of our instruments? Or should I base myself on manufacturing requirements only?

I have seen several sources for tolerances:

1. Manufactures stated tolerance
2. Customer supplied tolerance
3. Industry specific tolerance
4. Calibration industry accepted tolerance (GIDEP)
5. Process tolerance

With your scope, surely you have something established already.:)

Just some ideas for proceeding:

1. Start with mfg. tolerances for the instruments and see how they do.
2. Develop matrix for the different classes of instruments. NOTE: This can get tricky segregating the instruments and keeping up with them. However, it may become a necessary evil.
3. Go find each instrument and develop an instrument sheet for each. This will assure it exists, assure it's labeled properly, and you can capture a use tolerance for it at that time. Very time consuming, but may be necessary if control has been lost on the current instrument system.

ASIDE: I am getting a bit concerned for your situation; but it may be due to my lack of understanding. You indicate that you have been doing calibrations, yet I get the impression you have not identified any tolerances. Is that accurate? How have calibrations been performed?

Is there any standard that can help me with tolerance/classification of our instruments? Or should I base myself on manufacturing requirements only?

One thing to consider is the tolerances of the parts you are checking with the gages in question. You will want to have about a 10:1 accuracy ratio between your gage and part. As an example, if your parts have a tolerance of +/- .5mm, then your gage should be at least accurate to +/-.05mm. However, you don't need gages too accurate for your process. Don't use a gage accurate to .001mm for the example I just gave. That's going too far. Initial gage costs are higher and you have a greater likelyhood that the gage won't meet the tight tolerance.

Ideally, gages and equipment would be obtained according to needs based on this 10:1 ratio. Since you've already got them, I would just verify that the gages you are using are more accurate than your part tolerances - i.e. 10 times as much. Manufacturers like Mitutoyo state accuracies for most of their items. For other makers, it may not be so easy.

There are tolerances written in standards for items like thread plugs... and those should be followed. But for devices like calipers and micrometers, set the tolerances based on your part tolerances (making sure that the gage can meet that tolerance in the first place!).

Hi, ;)
This is my first time asking for your help so I hope I'm giving you enough information for you to understand my problem. We are presently trying to certifiy to AS9100. Our calibration is mostly done in-house and we do not have specific tolerances for each of our instruments. We are a custom machine and fab shop and have over 3000 instruments of about 50 different types. I was wondering if anyone knows of a standard that could help me figure out what the acceptable tolerance for each instrument is. I am also searching for a standard that can give me the acceptable environmental conditions required for my metrology lab, i.e. temperature, humidy, lighting etc. Thanks,

Veronique


ISO/IEC 17025 and if you can find a a book by clifford w. kenndy this takes you through the process on gage calibration for equipment and lab is to be maintain at 68 degrees.

ISO/IEC 17025 and if you can find a a book by clifford w. kenndy this takes you through the process on gage calibration for equipment and lab is to be maintain at 68 degrees.

Books by Clifford W. Kennedy (http://www.alibris.com/search/books/author/Kennedy,%20Clifford%20W).

Stijloor.

ISO/IEC 17025 and if you can find a a book by clifford w. kenndy this takes you through the process on gage calibration for equipment and lab is to be maintain at 68 degrees.

Yes, the book Inspection and Gaging is pretty good. Mr. Kennedy is deceased and there are a lot of editions out there (I've got 6th Ed. from 1987) Seems like only quite old ones available on Amazon.