|
This thread is carried over and continued in the Current Elsmar Cove Forums
|
The New Elsmar Cove Forums
|
The New Elsmar Cove Forums
![]() Measurement, Test and Calibration
![]() Difficulties Complying to 4.11
|
| next newest topic | next oldest topic |
| Author | Topic: Difficulties Complying to 4.11 |
|
Weyli Lurker (<10 Posts) Posts: 5 |
Hi, We have just gone through our 1st ISO9002 surveillance audit. During which the auditor found out that we are still not fully comply with 4.11. NCs and observations given are that we have not done the below:- 1. Calibration Report does not contain before and after calibration data. 2. Calibration Report does not contain Tracebility info to National or International standards. The cal report states only the national / international agency that the calibration standard is traceable to, but the auditor was looking for the actual standard (ie, info including the serial number etc) that is used by the international agency to calibrate the calibration standard used by my calibration vendor. 3. Q:How to determine the calibration frequency of IMTE? The auditor told me that through time, from the differences between the before and after calibration data collected, I can determine the calibration frequency for the IMTE. How do I go about doing that? Is there any standard formula that I can use? 4. We did not specify the environmental conditions to calibrate our IMTE in. 5. We should separate the IMTE master list 'Calibration Required' category into 'Calibration Required' and 'Verification Needed'. Reason given is take a Block Gage for example, when I send out for calibration it is actually sent out for 'verification'. Q:Is this separate tracking list a common practice? 6. Q:For 4.11.2(f), how to determine how much data I need to refer back if a certain IMTE is out of cal? Thank you for taking time to read this long posting. Any suggestions will be appreciated. Weyli IP: Logged |
|
Lassitude Forum Contributor Posts: 89 |
quote:Please define initial performance spec. I will comment on each of the other questions however it sounds as if no one at your facility has a good grasp of calibration. You need before and after data to assess the instrument over time. quote:At the very least you should be ready to call the lab that does your calibrations to get them to FAX you that information. I see this one go both ways - some auditors want it on the cal cert while others will give you a bit more slack. This should not be a problem for your cal lab anyway, so requesting it should not be a problem. quote:There is no formula per se. This is in part why you want before and after calibration data. Calibration cycle is determined by many things: Frequency of use, calibration results (history), manufacturers recommendation, etc. Here again, the issue is having someone who understands measurement systems. Looks like you're in for an education.
quote:Again, understanding of measurement systems and devices comes into play. Typically an instrument is calibrated at room ambient temp and humidity, but you have to be able to discuss what is done and why. In your example, the question becomes does is your multimeter linear and without bias from -10C to 40C? Check the manual which came with it. Use the manufacturers accuracy statement for uncertainty (+/- x volts or whatever at y degrees F). Typically there will be no significant bias between say 40F and 120F but you should be ready to state that and how you know that. State that you want it calibrated at room ambient.
quote:This horse shit and rhetoric. All you have to do is recognize what is calibratable and what is not. I personally have never seen two lists but I've never gotten into a fight over calibration vs verification. In part because most 'calibrations' are nothing more than verifications even when the equipment is calibratable. If you have a calipers you do not really send it for calibration - you send it for verification. It may not need to be calibrated - or it may. By default, you should expect the device to be in calibration when it goes in for 'verification of calibration status'. If you are calibrating an instrument every time it is in for 'verification of calibration status', you have a problem. Shorter cal cycle may be the answer, a sturdier instrument may be the answer, but other than fixtures you should not see many actual calibrations. The expectation is you will send it for verification of calibtration status and then, if it is out of calibration, it is evaluated (Is it broken? Is it worn out? Is it used to harshly? Etc.) for a solution. Calibrate it? Repair it? Buy a more appropriate device? Shorten calibration cycle? This person is using semantics and, to be honest, you need to learn MSA fundamentals so you can 'fight'. Two lists? Bah! It's pretty basic knowing what can be calibrated and what cannot. quote:You have to understand containment. Do a search here in the forums for CONTAINMENT and then come back and ask more questions after you read through a few hits. Also see: [This message has been edited by Marc Smith (edited 03 August 1999).] IP: Logged |
|
TheOtherMe Forum Contributor Posts: 64 |
Variations on the theme: Subject: Re: Q: Compliance to 4.11 /Li/Stein From: Philip Stein >From: Neoh Wey Li This is a mysterious requirement from your assessor. You must verify that your calibration vendor is acceptable, either because they are accredited to Guide 25, or because you audit them, or something, but there is NO requirement that you re-inspect IMTE when it is returned from service/calibration. How would you know how to do that? - you sent it out because you don't have the facilities or capability of doing it in-house. > Yes and no. It is best to have the data. A traceable calibration has both a numerical value and a statement of its uncertainty. It would be good to have the actual number determined by the cal lab. > The assessor is overreaching. There is no requirement beyond a statement of traceability and the aforementioned data value and uncertainty value. > See national Conference of Standards Laboratories procedure RP-1. http://www.ncsl-hq.org > You must calculate or have measured for you the performance of the instrument over all expected environmental conditions. Usually, you can refer to the manufacturer's specifications for extended ranges of environment and then calibrate at one temperature, but you must have a statement in your procedures that indicates that an evaluation of instrument performance at all expected temperatures has been done and shown to have negligible effect on your results at the required level of accuracy. No. > You must have a written procedure that says what you will do, and then you must do it. There are no requirements as to what must be done. Philip Stein IP: Logged |
|
TheOtherMe Forum Contributor Posts: 64 |
Subject: Re: Compliance to 4.11 /Li/Sugiarto Date: Tue, 3 Aug 1999 12:45:18 -0600 From: ISO Standards Discussion From: Welly Sugiarto > NCs and observations given are that we have not done the below:- One problem on relying with calibration vendor's conclusion is that they may not use the same standard as you are. Sometime this can work against your advantage. For example, you may use a caliper to measure your product, the range of measurement is from 5cm - 10cm. With tolerance of 1 mm. Your calibration vendor may measure your caliper according to manuf. specification (which is much more stringent than what you use it for) and conclude that the caliper is out of spec, whereas in your case, the caliper is perfectly acceptable to be used. I have an experience in calibrating weighing instrument, the calibration body says the equipment is ok, however when we cross check to their data we found that the equipment has 4% deviation which is not acceptable. Hence, relying on calibration vendor may not be a good idea. They can make mistake, too. After all, you are the one using the equipment, not them (calibration vendor) > 2. Calibration Report does not contain before and after calibration 1. Data can indicate whether the instrument is out of calibration or not prior to adjustment. Thus, you can then go back to your inspection record to make sure that your previous inspection (using this instrument)is valid. (4.11.2f) 2. Data can indicate if the calibration is performed accordingly. I had an experience in assiting my client to calibrate a thermometer, when we received the thermometer back we were quite surprise, because the data indicate that the calibration vendor only check the range of temperature around room temperature, whereas we also use it for our refrigerator (around freezing temperature). Our PO clearly states the range the thermometer is used for. Again, calibration vendors can make a mistake. 3. Data can indicate more precisely what range of measurement your instrument has problem. One time, we received a calibration result in which the data showed the instrument was not precise for a specific range but ok on the others, so we still use the instrument but put a warning not to use it on a certain range. I think, you have a right to ask for calibration data, if your vendor can't give you one, I think you better look for someone else. Believe me, having the actual data is to your advantages. > 3. Calibration Report does not contain Tracebility info to National or > 4. How to determine the calibration frequency of IMTE? The auditor told Refer to ISO 10012-1:1992 Annex A: Guidelines for determination of confirmation intervals for measuring equipment > 6. We should separate the IMTE master list Calibration Required > 7. Lastly, for 4.11.2(f), how to determine how much data I need to refer > Thank you for taking time to read this long posting. Any suggestions are Welly If you need more information, please do not hestitate to contact me directly at: [email protected] I used to work as an ISO 9000 consultant for several years before coming back to school pursuing my MBA/MSE degree IP: Logged |
|
TheOtherMe Forum Contributor Posts: 64 |
The voice of 1 more expert: Subject: Re: Q: Compliance to 4.11 /Li/Andrews From: [email protected] >WL Neoh's inquiry and my responses; >NCs and observations given are that we have not done the below:- Response #1 - Acceptance criteria needs to be stated; however, it may be stated exactly as you did above. That is acceptable - provided that is what you do and what works for you. >"2. Calibration Report does not contain before and after calibration Response #2 - The calibration reports need to show the before and after conditions. If no adjustments were made it should indicate such. >"3. Calibration Report does not contain Tracebility info to National or Response #3 - No it does not. If the auditor wishes that information they may contact your calibration lab. The lab should have that information (you may wish to verify that they indeed do). >"4. How to determine the calibration frequency of IMTE? The auditor told Response #4 - This is not required by ISO 9001 or ISO 9002. It is only required that you calibrate the equipment at regular intervals - how you arrive at those intervals is up to you. >5. We did not specify the environmental conditions to calibrate our IMTE Response #5 - Calibration houses usually calibrate equipment at set environmental conditions (eg; 68 deg. F at 40-60% rel. h). Unless you are doing very close tolerance work this method of calibration is adequate. If all of your calibration work is outsourced, a sentence stating that the calibration environment utilized by your calibration house is an acceptable environment is sufficient. If you do some in-house calibration work another statement about acceptable room temperature and humidity ranges is sufficient (just make sure that you have a (calibrated) instrument to verify the conditions). >6. We should separate the IMTE master list Calibration Required Response #6 - He/She is off on this one. Ask for where that is required by 4.11 of ISO 9002. Calibration/Verification - the debate between the usage of the two words can be quite protracted (see list archives >7. Lastly, for 4.11.2(f), how to determine how much data I need to refer Response #7 - This is a tough one. Unless you have a system that allows for part/tool traceability you are going to have a tough time figuring out EXACTLY what parts were inspected with what instrument. You do need to have defined some system for verifying the previous inspection results - to the "greatest practical extent". This one will require some thought on your end. WL - from the overall tone of the above items it seems that the auditor brought up some valid points; however, there are also a number of items where he/she was off the mark and perhaps was interjecting their own bias into the audit. Hope my responses are somewhat useful. I'm sure you will receive others from the list members and perhaps all together will provide some useful insights. Good Luck, Ethan Andrews IP: Logged |
|
Marc Smith Cheech Wizard Posts: 4119 |
Subject: Re: Q: Compliance to 4.11 /Li/Randall Date: Mon, 9 Aug 1999 11:49:41 -0600 From: ISO Standards Discussion From: [email protected] > From: Neoh Wey Li Did you document the above explanation in your procedure(s)? If not, then your auditor may have a valid point. While there is no specific requirement for you to inspect your IM&TE upon return, you must still examine the calibration report to determine whether the IM&TE was found in or out of tolerance. > 2. Calibration Report does not contain before and after calibration The calibration report must include the calibration "results" (not necessarily before and after "point" data). If the initial performance test found the instrument within spec., then it should say so (this is required by most calibration quality system standards). > 3. Calibration Report does not contain Traceability info to National or The calibration report is not required to contain any more than a statement of traceability. Your calibration subcontractor should maintain a "record of use" for each instrument calibrated (provided that a calibration quality system standard was invoked). The "record of use" will provide detailed evidence of traceability (one assumes that you verified this when you qualified the calibration subcontractor). If you have a large enough sample size you can statistically monitor the number (or percentage) of instruments of a given type that are found in tolerance (2 sigma is typical depending upon the critical nature of the measurements made). The calibration industry will refer to this a K factor. As does Mr. Stein, I also would refer you to NCSL RP-1. > 5. We did not specify the environmental conditions to calibrate our IMTE Most calibration reports include the environmental conditions at the time of calibration -- however, most electronic instrumentation is not significantly affected by typical factory floor temperature variations. Dimensional tools on the other hand can is affected by temperature which can be compensated for through "correction factors" (as Mr. Sugiarto points out). > 6. We should separate the IMTE master list Calibration Required In my experience it is not a common practice to separate these devices into separate lists. > 7. Lastly, for 4.11.2(f), how to determine how much data I need to refer The standard does not dictate "how" you must comply with this element. What is the worse thing that can happen if defective product leaves your facility? If a safety related issue could arise, then I would say that you should have complete traceability OR take additional measures to mitigate this possibility (e.g., maintaining a 4:1 or greater test accuracy ratio between you IM&TE and product being measured). One would assume that in this situation there would be requirements for traceability and recall of defective product should the need arise. However, if you're manufacturing small electronics or film for commercial use (and no one is going to die or be injured) then I would perform an impact analysis only (since no traceability is required). In other words, the impact analysis would tell the auditor that defective product "may" have left the facility, but you're not going to do anything about it -- other than perhaps adjust the calibration interval of the instrument (but even this is not required). Where traceability is not utilized, this requirement (4.11.2f) often becomes little more than a paperwork exercise. Overall, the best suggestion that I can make is for you to (1) make sure that your calibration subcontractor knows what they're doing (a lot of them don't). Really qualify the calibration lab -- invoke ISO 25 or ANSI/NCSL Z540-1. If possible, try to use a calibration lab that is either accredited to ISO 25 (by NVLAP, A2LA or NQA-USA) or at least ISO 9000 registered. (2) After you're satisfied with the qualifications of your cal lab, call your calibration subcontractor and talk about your calibration program. Most will be willing to help you develop your in-house program because they know that you will come to rely on them for all of your calibration needs. Allow them to work with your people to setup your calibration recall system. If they know their business, they should certainly know how to address calibration recall! BTW, I noticed that a couple of respondents mentioned ISO 10012-1. It is my understanding that this standard is scheduled to be canceled when ISO 17025 replaces ISO Guide 25 (later this year). I hope that this helps, Richard C. Randall IP: Logged |
|
Marc Smith Cheech Wizard Posts: 4119 |
From: Carlos Pereira da Cruz Subject: Re: Q: Compliance to 4.11 /Li/Randall/Cruz
ISO 10012-1 will stand, ISO 17025 wil replace EN 45001 Carlos IP: Logged |
|
Weyli Lurker (<10 Posts) Posts: 5 |
Hi all, Thank you for providing me with very helpful information. I got to know from one of our calibration vendors that according to ISO/IEC G25, Accuracy is not required to be included in a calibration report. In ISO10012 -1, 4.8 also does not state Accuracy as one of the items required in cal reports. Does this mean I can set my own acceptance criteria for after cal IMTE (4.11.2(C)) and not neccessary to include Accuracy as acceptance criteria as long as Uncertainty is stated(4.11.1)? And ISO10012-1 4.8(f) "the designated limits of permissible error". How do I interpret this? Thank you very much. IP: Logged |
|
Marc Smith Cheech Wizard Posts: 4119 |
quote:Sounds like it to me.
quote:I don't have 10012-1 here and I hate to comment out of context. If I can dig up a copy in the next day or two I'll get back on this. Or if you want to post more of 4.8 so I can see the context. IP: Logged |
|
Weyli Lurker (<10 Posts) Posts: 5 |
Hi Marc, Thank for the reply. ISO10012-1 Calibration results shall be recorded in sufficient detail so that that the traceability of all the measurement can be reproduced under conditions close to the original conditions, thereby facilitating the resolution of any anomalies. The recorded information shall include: Thanks. IP: Logged |
|
Marc Smith Cheech Wizard Posts: 4119 |
You've got me there. Do you have a calibration cert you can look at? f) the designated limits of permissible error These are the 2 issues. Sorry I can't help further. We need an expert here. IP: Logged |
|
Weyli Lurker (<10 Posts) Posts: 5 |
I have a cal cert of 25.000mm ring gage. The cert states, (I wonder this is the permissible error that the guide is refering to? Or is this the accuracy of the gage?) And the uncertainty is +/- 0.0001mm at 95% confidence level. Weyli IP: Logged |
|
Marc Smith Cheech Wizard Posts: 4119 |
I believe that's it. IP: Logged |
All times are Eastern Standard Time (USA) | next newest topic | next oldest topic |
![]() |
Hop to: |
Your Input Into These Forums Is Appreciated! Thanks!
