THIS IS A 2 PART QUESTION.
1) I am have a dissagrement with my boss over my cal due date. My computer software gives a due date as mm/dd/yy, however I would like to state in the procedure that the gauge is not past due until the last day of the month. Everyone outside of my plant says this is how they do it. He insists we would not pass our qs audit with the procedure written this way. He claims that a gauge due on the 1st of the month would still be overdue by 30 days?
2) How can I control or justify changing calibration frequency's ? Is there specific statistical controls or test to do this without great time investments in reserching history files? I am refering now to mechanical gauges(ie. calipers,mic's,etc.)

Scott,

I have answers to your questions. So will others. http://www.qs9000.com/ubb/smile.gif

As long as your procedure says that the day is unimportant and that calibration is due sometime within the month listed and only considered late when that month is passed (wording better than this is highly recommended http://www.qs9000.com/ubb/smile.gif LOL) than QS will accept it as long as you do it that way and you can prove it is appropriate.

If your program is set for mm/dd/yy why not try to always use the same /dd/ for all instruments. or see if you can change the format to mm/yy.

This is allowed by QS. Stating that the day listed is a target but it is not considered late until after the last day of the month works.

As for justifying the change, look at your operations:
Is it critical that certian equipment be calibrated on a precision (exact) time schedule or only that the schedule must be accurate (on or about). Is it cost effective? Do you have the resources? Is it efficient? Is there waste as a result? Is the time schedule practical? does it promote quality or make the cost of quality too high? and lastly is it really necessary to calibrate at the frequency currently implemented? For this last question, use history data and mfg'r recommendation as a guide.

I hope this was some help.

Scott,

I agree with Leslie. I would suggest the mm/yy (or yy/mm) format. As far as justifying the changes, I am sure your document control procedure allows for this kind of flexibility. You are not changing the frequency, just the details. The yy/mm format that I used in a MIL-I-45208 environment was fine with the GSA auditors (surely QS assessors are not as bad as the GSA guys). I would drop the /dd/ from the date format, audit your documentation to ensure that it is up to date, and carry on.

I am not a QS specialist, but I know that this method would work for an ISO style of assessment. Justify and document; you should be fine.

Best Regards,
Don

Scott,
1) We state in our procedure that the calibration is due sometime during the month noted as the due "date". So far, it has passed 2 surveillance audits and 3 different auditors with no problems.
2)As far as changing the calibration frequency, we base it on historical data (we have checked this gage once a month and have never found a problem - we can change it to every 6 months.) We do not mess with frequencies on items that measure critical parameters in our system (such as chemistries and temperatures).
I hope this helps.
Jennifer

Scott -
We do the same thing as Jennifer's company, including using historical data to change cal frequencies, and we are QS certified.

I know we had a calibration frequency debate going in a recent thread, but the search engine didn't find it - so I'm putting this here. If anyone knows where the other cal frequency thread is, let me know and I'll link these.

-----snippo-----

Date: Thu, 9 Sep 1999 10:33:25 -0700
From: Bruce Mayfield
To: Greg Gogates
Subject: Survey Results

To learn more about how other companies, organizations and our customers handled calibration intervals I set up a Focus group to discuss two questions

These were:

1. Does your company use the recommended interval on the Calibration Certificate to determine your calibration interval?

2. Does this change cause you any problem with your subcontractors not providing a recommended calibration interval in compliance to the requirement?

Here are the results I obtained through the sample of respondents I received:

Demographically here is how the group broke down:

Government Internal Service Company
Labs Company Labs Testing/ Calibration Labs

13.7% 48.3% 37.9%


The focus group contained organizations from the USA and other countries:


USA Other

78.6% 21.4%

Question #1. 27.5% responded Yes they used the recommended intervals.
They also stated that they used them (about 50% of the
respondents) as the introductory interval and then adjusted
the interval to meet their internal requirements. About 50%
of those responding Yes, represented facilities out side of
the USA. The balance, 72.4%, developed their own
internal calibration intervals and did not use the
recommended intervals from other companies/organizations.


Question #2. 17.2% responded Yes that it would cause a problem. In
reviewing the respondents 50% (8.5%) of those identifying
Yes were out side of the USA. About 50% represented the
USA. Again the balance of the respondents 82.7% stated
that it would not make any difference at all.

In summary, if I take all of the comments and numbers as a whole, in only a minority of cases do individual organizations need or use externally derived recommended intervals. Most of those that do, use them as a starting basis for determining their internal derived interval. Those that do use recommended intervals, have not developed the means to properly determine intervals. Those individuals responding from countries that are not as industrialized as the USA, I can understand why there would be problems with this concept. For those organizations in the USA deriving intervals solely based on someone's else's concept of intervals are not in step with concepts outlined in the various guides and standards.

I also want to thank all of those who responded and to Greg Gogates for his patience in forwarding many of the e-mails over to me.

Bruce Mayfield
Telogy

Help!

I am rather new to the area of Gage R&R, Does QS9000 via MSA set a guideline on how often something needs to have Gage R&R studies done? I have read through each....maybe I missed it...

Or is this based on usage or %R&R values

Thanks for your help!

I am not aware of QS requiring specific calibration period.
It is your decision based on how often you use the gauge, how critical is the feature measured.....
What could help as well would to do a wear sudy on the gauges based on at least 3 previous calibration results.My company has just been certified to the TS16949 standard, and the auditor was quite picky on the MSA studies, how they are performed and why.
Hoping this will help a bit, good luck!!!
Audrey :) :)

This helps......however our gauges dont consist of parts but rather titration test, analytical testing....

I dont know if that matters....thanks for your help!

I love this Forum:cool:

Is it possible / acceptable (to auditors) to perform condition based calibration rather than having a fixed frequency ?

I think the Stability test in MSA may be a good method to determine if the instrument is deteriorating and needs calibration.

- Atul.

Originally posted by Yasupawoman

does QS9000 via MSA set a guideline on how often something needs to have Gage R&R studies done?


I don't think there are any definitive guidelines for R&R frequency.

You should do R&R :

1. Whenever there is a significant change in the measurement system characteristcs ... instruments, methods, operators, etc

2. (according to some) after every calibration !!

3. Whenever your customer requires it.

-Atul

I think what you would have to convince an auditor is that you have a defined confidence that the measurement equipment will perform measurements within some acceptable limits of error/confidence (either tolerance limits or statistically based (control charts, X-Bar - R, etc..)) for some definable length of time.

The standard method (in the calibration world) is, based on historical data, to set an appropriate interval based on the equipments historical ability to remain within a range of tolerance limits.

In the semiconductor world it is quite common to have golden units which check upper/lower control limit confidence, etc. The golden unit has to be well characterized, and meet the stability requirements of the MSA, etc.. Then the MSA is performed at some defined interval.

One of the challenges of replacing a defined calibration interval method with an MSA method is that you still need to have a defined confidence of how long the equipment remains within acceptable CPk or whatever method you use.

Whether this is a better method or not will depend on the specifics of your application. There are many circumstances where a calibration at defined interval is simpler.

I do not believe in fixed frequency calibrations (in general), as they tend to overlook variables of use, environment, equipment degradation. Having a "fixed frequency" for calibration will over the long term, be likely to do one of two things. It will either overuse resources (cal interval too short) or inflict product risk due to unit drifting out of tolerance too often (cal interval too long). For critical applications, it is more reliable to either define a confidence level and adjust intervals as needed to the confidence level, or use MSA (a lot of work in places where it is not required).

Using the stability test in MSA would still need to be done at some defined intervals. The defined interval may well end up being the same as the calibration interval you would arrive at through regularly scheduled calibrations, and may not buy you any resource savings. If you could tell me a little more about what test equipment you are referring to, I might (or might not) be able to give a little further comments.

Stability Test does involve obtaining measurements at a regular frequency. But the period here is much shorter, something like one reading per day or shift. Then you plot these readings on a chart (a simple XBar-R or X-MR chart) whose control limits are previously determined through a benchmark study. You can catch out-of-control and trend situations and decide whether an instrument is drifting. This way statistical stability over a period of time can be monitored. When you do calibrate an instrument, conduct a benchmark study again to determine new control limits and start over...

What I wanted to say is that the QS certified people must already be doing stability tests as part of QS compliance. It should be possible to carry out condition based calibration and justify it.

The question still remains unanswered for

1. Instruments not covered in Control Plan, (and hence not coming under MSA but still need calibration) and
2. Non-QS companies

??

- Atul

PS: er.. yes, Jerry. I am referring to instuments such as verniers, micrometer, dials, etc, typically in an auto/engg industry. I have no knowledge of calibration requirements in chemical and other indutry. There as you say, periodic calibration may be a simpler option.

Just a few comments which you may or may not agree with.

1) As long as your procedures state that all calibrations are valid until the end of the month I cannot see a problem. (Say what you do and do what you say).

2) If you always have 'as found' results on items that require calibration if you calibrate late (it happens) and these 'as found' results are in spec you have no problem. A instrument does not magically go out of spec just beacuse you have passed the date specified on its calibration label.

You may get pulled up because you have not kept to your schedule but if the 'as found' results are in spec you do not have to worry about a potential product recall.

How about this train of thought?

Do an initial R&R and use ongoing Bias, Linearity, and Stability as the drivers for determining the frequency of R&R studies.

Just my opinion, but bias, linearity, and stability give me better confidence in a gage than R&R.

The train has left the station:bigwave:
Does R&R using ongoing bias, linearity and stability and better confidence in a gage than R&R determine the end of the month calibration date on my caibration? Just asking
:smokin:

Bias test is done at a fixed 'master' value, linearity is bias study over the range of the instrument. In essence these are 'static' tests, that help you determine 'limitations for use', if any.

Stability test confirms statistical stability over a period of time. I would trust stability test to determine if instrument is drifting and hence due for calibration.

R&R is a different ball game. The idea is to determine errors produced due to factors like instrument, measurement methods, operators, environment to decide if the instrument is indeed good enough to control the process it measures. As such, any time any of these factors change, you should do an R&R test, like if you change the fixtures, introduce a new instrument / operator, repair/calibrate existing one, use the instrument to measure some other part,..... Also you can find out if a certain operator needs to be trained.

Hope the train hasn't left the tracks! ;)
- Atul.

Atul,

Train is still on the tracks, very well said!

Originally posted by SCOTT SNYDER
THIS IS A 2 PART QUESTION.
1) I am have a dissagrement with my boss over my cal due date. My computer software gives a due date as mm/dd/yy, however I would like to state in the procedure that the gauge is not past due until the last day of the month. Everyone outside of my plant says this is how they do it. He insists we would not pass our qs audit with the procedure written this way. He claims that a gauge due on the 1st of the month would still be overdue by 30 days?

This is what my question was about, boys. Let's not miss the forest because of the trees:smokin:
However, I defer to the previous posts regarding this issue to answer my question. There are some good ones. So in keeping in line with question 2, would the following apply to my calibration frequency for a 1" Micrometer with a +/-.002 tolerance on my bar stock parts?
Process control for positional specifications can be separated into the X,Y,&Z components of the measurement. It can be very beneficial to the process owner to identify the source of variation. Especially if the ratio of tolerance consumed to tolerance specified is high. Monitoring the resultant diameter of position deviation adds little value to the process owner when one process delivers a position deviation that is scattered in all four quadrants and another with an equivalent magnitude of deviation emanates from a single point in one quadrant. If the ratio of tolerance consumed to tolerance available is low it may not be worth the effort to monitor the coordinates separately.
To measure process capability of geometric specifications with variable tolerances (MMC-bonus) one can monitor the residual or remaining unused value of tolerance rather than the amount consumed. For each measured feature just add the bonus to the tolerance specified and subtract the amount of positional tolerance consumed. Rather than comparing the mean and standard deviation to a "variable USL" the mean and standard dev of the residual is compared to zero. Cpk of the residual is derived from the Cpl. The variation in feature size when merged with that of position changes the distribution. The standard deviation typically increases but so does the separation of the mean of the residual from zero (consistent with the magnitude of the bonus). It is important that feature size is "in-control" as well as the coordinates of feature position when predicting capability from the residual.
One enormous benefit of doing process capability from the residual is the realization the dynamic relationship between size and position. One should not necessarily target feature size at its specification midpoint! A positional tolerance that is specified as "Zero @ MMC" where the amount of tolerance is fully dependent upon size probably demonstrates that relationship the best:biglaugh:
:smokin:

Marc, I remember that thread well. I was the one that started it, but I can't find it either.

I do remember I wanted to put a +/- 2 days tolerance on the calibration due date to compensate for weekends. The majority of the replies gave me grief for that, but it seems times sure have changed after reading the current replies.

My only advice is if the calibration is due, do it. There shouldn't be any reason not to.

Ken,

You posted:

----------------------------------------------------
My only advice is if the calibration is due, do it. There shouldn't be any reason not to.
----------------------------------------------------

I love it, realism at its best!:bigwave:



:cool: :cool: :cool:

Say what you do, do what you say. Part 2,you have calibration records, so based on that past history lengthen or shorten the interval.

Are there any guidelines or thumb-rules to decide how much you can lengthen or shorten calib intervals? How much historical data would be considered 'sufficient' to take this decision?

- Atul

{edit}
OOPS! Sorry, Hit the submit reply button too soon ! In fact there are some good threads here..

http://Elsmar.com/Forums/showthread.php?t=1108
http://Elsmar.com/Forums/showthread.php?t=1147
{/edit}

None that I know of. How comfortable are you with the intervals. A lot will depend on the # of instruments, how often you calibrate now, also the amount of training people have on the use and care of instruments, the environment in which used, severity of use, etc.

In a previous place of employment, working under MIL-Q-9858 and MIL-STD-45662, we had many cylindrical and thread plugs, as well as Threaded Master plug gages for setting threaded ring gages. Do they still allow people to "set" threaded ring gages? That's how long I've been out of it. Anyway, there were many gages that were only used for a specific part. The gages may be used once every two years, or more. When they came back from the Cal Lab, they were coated with a plastic substance. If we pulled the gage when it was due and it still had an unbroken coating, I simply extended the due date another 6 months. Our resident DPRO, you old timers know who they were, had no heartburn with it. We just entered in the gage record, "Gage not used since last calibration". Our DPRO Representative pulled a half dozen gage ident numbers from the master list every month and looked at a few of those "coated" plugs very closely. Never had a problem. We had some gages in inventory that had not been used for 4 years. We would then "de-activate" it and store it away. When a need arose for it, we just "activated" it and began the countdown from that day for 6 months. It was so very simple. There were over 500 gages in the system and it just made sense to make up a rule that reduced the cost of calibrating plugs that did not need it. :ko: :biglaugh:

We do something very similar with Type R thermocouples used in very high temperature processes. Most metrology folks skilled in temperature disciplines know that under ambient conditions, thermocouples don't change very much. It is a pretty common practice (in many companies) to require only initial calibration on many thermocouples for that reason.

However, using a type B, R, S or other noble metal/high temp thermocouple is an entirely different scenario. When Platinum/Rhodium alloys are left in an environment of up around 1000 - 1200 degrees Celsius for nearly 24 hours a day, it is a matter of months until that thermocouple drifts down below our lower tolerance limit. This is due to precipitation of Rhodium oxides. The rhodium vaporizes off the thermocouple, and the material characteristics change.

As we fabricate thermocouple assemblies in quartz and ceramic assemblies, we make up extras and store them in a cabinet in our environmentally controlled lab area. When the user picks the thermocouple up, we start the cal cycle (length varies with users needs).

A calibration interval is the period of time within which (under set conditions)a measurement device (system, etc.) can be expected to measure within a specified tolerance to a specified confidence level. If you have some reasonable common sense data, ANY measuring device should be able to be recalled in that manner (in theory).

There are some more complexe electronic instruments that sitting on a shelf ages them just as quickly (or nearly as quickly) as when the user has them, and others where they do not. So each case needs to be weighed on its own merits.

I think in the mechanical (dimensional) world, most of the tolerance drift (thermal expansion coefficient excepted) is due to wear and tear. And so most mechanical instruments properly stored and/or protected (i.e.: oil coating, plastic coating, tamper evident seal on storage container, etc.) would seem to lean toward the latter method.

I can't state strongly enough that each case must be adjudicated on its individual merits. :biglaugh:

As far as electronic gaging goes, sitting on the shelf is quite often worse for drift. Capacitors "dry up" when not used, potentiometers get dirty when not used, and specialized circuits that are designed to have constant operation become unpredictable.

Case in point: Our reference voltage sources, which had been in storage since some "wise" person decided that they weren't needed. It is now been 4 months and a few thousand dollars later, and in another couple of months, they might even be useable. They haven't quite 'stabilized' yet, had their drift characteristics fully charted, etc.

Case 2: High accuracy (8-1/2 digit) multimeters have an internal calibration. Regardless of use, I have to perform that calibration regularly to keep the meter accurate. Not doing it for over 90 days is catastrophic to the accuracy, and a full recalibration is required before I can use it at its designed accuracy.

In other words, if it is electronic, you had better prove your drift, and the effects of non-use prior to extending cal intervals.

Physical gages, such as torque wrenches, also suffer from non-use. Physical joints tend to settle into a place if not regularly exercised.

Dimensional gages are the exception. 90-95% of the relaxation of the metal occurs in the first two years after the original hardening (according to my source at NIST that I would not dare doubt). After that, variation due to thermal effects are sinusoidal rather than drift-related (I just call it "slop" ;) ).

Oh, and to fuel the Month/Year fire again - try it with an assessor that has a clue. By definition (including that accepted by ISO) the word "Date" means day/month/year. If you receive accredited certificates, they will have the full date, or their accrediting body will have a problem with it. A2LA and NVLAP actually spell it out in guidance documents that it MUST be in a day/month/year format. The easiest way to push it to the end of the month is to add a clause in your Quality Manual that states something to the effect that "Calibrations may be extended up to X days in the same calender month to facilitate production. Any non-conformance noted during calibration at that time follows the normal product recall procedures." As long as you can afford the relative risk (which is quite often extremely small), go for it.

Ryan

Again I say, as was stated before, why look for a way out when you can do it correctly in the first place. Anybody that sets up gage system that is rife for areas of non-compliance should be reprimanded or lowered in grade.

A date is a date. Tell you girlfriend/boyfriend that you will pick them up Saturday at 8:00 pm +'- 2 days. Good luck finding another date.

Let's write a procedure that says we will calibrate this gage every February 29th. That would go over like a lead balloon, by the time the leap year came around the gage has probably already been stolen or trashed.

Back to the point, if you have a date, meet it! If the date is not met, then **** starts to run uphill for awhile. The tech says the boss said do it later, the bosses boss asks why it wasn't done and yells at the Tech. The Tech procedes to raise his middle finger and tell his boss to calibrate this.

Bosses boss askes why no calibration has been done, boss says "Tech was not a team player", had to let him go.

Etc.........................................................................

Say what you do, do what you say, and realize your not trying to fool the registrar, your trying to improve the company!!!!!!!!!!!!

Marc, where's that soapbox I sent you!

> Marc, where's that soapbox I sent you!

Always close by, Al! :thedeal: Look down. You're standing on it.

I just noticed I posted as "unregistered". Oops, not trying to hide folks!:)

Al,

You took the scenic route, but I totally agree with you. (if that was you!)
I don't understand why someone would want to put off what's due today.

Al,

It has been my experience that more energy is spent on trying to figure out ways to circumvent rules than there is energy spent on following them. I wish it weren't so, but you are a rare breed these days, the one that doesn't fight the winds of change, but uses it to bring improvement to his field. But my guess is that you have to fight to carry out your good intentions. I know that most companies I've been in dedicate an incredible amount of effort to trying to get away with things.

So when you start your own company, give me a call, you have yourself a metrologist.

Ryan

Ryan,

You have a deal!

Yes it is sometimes hard to do things the proper way without stepping on too many toes, I guess that over the years I have come to realize that if you"creme over" one requirement the rest are soon to follow.

I guess that is why I am so popular at times and that a few people get nervous when I show up at their office or cubicle with my clipboard in hand.

Just a short story:

Mostly I work alone without management. If I send a memo to a department manager and I get no response I resend the memo to the Manager and the President. If no response, I send a third memo to the same requesting immediate action. After the President keeps seeing by memos he usually calls the Manager into the office and has a word with him/her. I don't know what they discuss but I assume either the manager has been told to cooperate for the sake of the system or cooperate because He/She is getting sick of receiving my memos abd tells the Manager to handle it.

Either way, my job is done. And everybody loves me!!!!;)

See http://Elsmar.com/Forums/showthread.php?t=1147

Originally posted by Ryan Wilde
It has been my experience that more energy is spent on trying to figure out ways to circumvent rules than there is energy spent on following them.

While I agree that a lot of companies do not want to stay the course, rules should not always be the end all. We must constantly look for alternative methods to accomplish the same thing. I used to resent others trying to find an easier way. We were comfortable doing things the same old way. Change is unsettling. But, if there is no valid reason for not making the change, you have to roll with it. Not that this is the case, as described in your post. Let me borrow the soap box and quote some Machiavelli that I have posted on the wall in my work space: "For the initiator has the enmity of all who would profit by the preservation of the old system and the merely lukewarm defenders in those who would gain by the new one".

I've had my eyes opened a few times by people who viewed long standing practices with a new set of eyes. Management will always look for a quicker, cheaper way to do things. And, sometimes they are right.:rolleyes: :smokin:

Energy,

I agree with you, finding a better way is always a good thing. I was more refering to the "bending" of rules that we cannot change, such as those written into those ISO Standards that seem so popular these days. I've written a few manuals, and have received a lot of backlash for writing in requirements that we didn't previously have. I usually replied with "We have to meet this requirement, and if you can come up with a better way while still meeting the requirement, tell me about it and I will write it in." I love change, which tends to irk a few peers, especially those that have worked in the same place, doing the same job, for years.

What I meant to say, I guess, was that I've seen an awful lot of energy (no relation) dedicated toward not meeting a requirement while trying to give the appearance that we are. Oftentimes, the energy (no relation) spent fighting what must be done, regardless of method, was greater than the energy (no relation) required to meet the standard.

Oh, and these days I seem to be Management, so I actually do try to find quicker, cheaper ways to meet requirements. But I am rarely right ;) :o

Speaking of drifting...

O.K. Sometimes I (we) have problems staying the course. Take away two stars, do not pass go and do not collect $200:biglaugh: :smokin:

Why don't you set your calibration dates so that 6 months or a year (whatever the tool frequency is set for) so that the due date always falls within the first few days of a new month. Then follow your original process of calibrating only at the end of each month. Now you will not miss a calibration date (i.e. boss and auditor should be happy) ... you will just be calibrating early and ahead of schedule ... what a concept. Just a thought.