TS 16949 Clause 7.6.3.2 - Use of External Calibration Laboratories

[PurduePete]

Using External Calibration Labs

What is required per the standards most automotive companies follow, i.e. QS9000 / ISO/TS?

Per ISO/TS 16949

Clause 7.6.3.2 mandates that any IMTE sent out to an external lab must be sent to an ISO 17025 accredited lab (or approved by the organization's customer) and be on their scope. However, it doesn't explicitly say the calibration shall be "accredited". Is this an obvious assumption that all should make? I do not like making any assumption when preparing for an audit. Many cal labs offer differing levels of a calibration; w/out data, w/ data, accredited, non-accredited. What's worse, they offer these at different prices.

What should the QS9000/TS 16949 organization request of their cal lab to ensure compliance to the standard?

 

[howste]

I'm not sure exactly what the difference between accredited and non-accredited calibrations are. If the calibration lab itself is accredited, I'm assuming they would stick with approved methods of calibration regardless of how much you paid.

Here's what I am sure of:

1) As you stated above, the lab needs to be accredited to ISO 17025 (or customer-approved)
2) There definitely needs to be data. 7.6.2 of TS 16949 says that records need to "provide evidence of conformity."
3) There needs to be traceability per the requirements of 7.6a

Basically 7.6.2 gives the big list of things that cal. records must include. Some of this must be provided by the outside lab, while others (such as assessments of the impact of out-of-specification conditions) need to be recorded internally.

 

[Craig H.]

PerduePete:

First, let me welcome you to the Cove!!!

I hope this is on-topic, but we got stung 2 external (9001) audits ago when it was discovered that, although we had used an outside calibrator that was ISO 17025 registered, their scope did not cover the equipment they calibrated for us. Further, the report was sub-par, and I didn't catch it because I was unfamiliar with the particular device.

So, its a good idea to check their scope, and not just rely on the "we're 17025" statement. I am sure you calibration folks out there must check this all the time...

Craig

 

[D.Scott]

Welcome to the Cove.

The comments made are absolutely correct. If you send any IMTE to an outside lab, the lab must be accredited and the calibration they are doing must be listed in their scope.

Dave

 

[PurduePete]

Re: Using External Cal Labs...

I'm not sure exactly what the difference between accredited and non-accredited calibrations are. If the calibration lab itself is accredited, I'm assuming they would stick with approved methods of calibration regardless of how much you paid.

I am not sure but here is what I believe to be the difference...

Accredited Certificate:
Calibrated by an ISO 17025 accredited company. Calibrated using the methods and procedures audited and validated as technically competent to produce the lowest uncertainty. The certificate must bear the logo of the accreditation body and is "traceable" (see the link for what "traceable" means to A2LA).

https://www.a2la2.net/policies/tracepol.pdf

Non-Accredited:
The lab could be accredited or not. If the lab is accredited, provides a gage verification against standards that may or not be "traceable" (per the terms defined in the link), does not provide uncertainty statements, does not have the logo on the certificate, may use amended procedures that are less expensive to perfrom in order to lower calibration costs.


The way I understand it is the commercial cal labs can provide their customer with whatever type of cal is requested via contract review, for instance:

1. Non-automotive industry (appliance, medical, etc.) customers that simply are getting their internal gages checked periodically for quality assurance sake, not due to an international standard. They do not need data, "traceability", uncertainty values, etc. They probably have never heard of ISO 17025 and wonder why their calibration costs would go up.

2. To keep costs down, the Tier 1,2,or 3 supplier may be asking for lower costs while the cal lab is raising them due to the high costs of implementing and paying for their accreditation. So, different "levels" of calibration are offered at different prices.


So I ask again, when contracting to a commercial lab and they ask you, "What type of calibration do you need, accredited or ...(some other variation)?" What am I going to be audited for having? If the lab I choose is accredited and the gage I'm sending them is on their scope of accreditation, is that good enough? Or do I need to pay for the full "Accredited Calibration" bearing the accrediting logo and in most cases at 50%-200% higher costs?

Thanks to those of you that have tried to help me thus far. The standard seems a bit vague here and I am hoping to get some feedback from others that have gone through QS9000 audits. What exactly do the auditors ask for? I realize this is but one element of the standard but it seems like an easy target for auditor dings!

 

[Ryan Wilde]

Re: Using External Cal Labs...

Hi all,

I work for an accredited lab (I won't mention who), and I can tell you, there is very different cost involved between accredited, non-accredited, and "standard commercial cal".

A standard commercial cal is just that - You get a sticker and a certificate that says it meets specs. No data, no bells, no whistles, no datasheet formatting, no quality check of the data by a separate inspector.

A non-accredited with data involves all test results reported to you, which must be checked by a separate inspector for validity and let's face it, to make sure the tech didn't put too many zeros, or show too few digits. That costs money, especially on manual calibrations. It can raise my cost to provide the service by 50-75%, just in typing and proofreading time.

An accredited calibration takes that a step further. Instead of measuring a point once, I may have to measure it 10 times to get an idea of the repeatability. After that, I have to calculate the uncertainty for that specific point. If it is a multi-range item, that can mean literally hundreds of statistical calculations for a single piece being calibrated. It easily doubles, and triples, the time required to calibrate an item. This, on top of the fact that accreditation itself cost several thousand dollars per year raises the price I must charge to maintain a semblance of a profit margin.

We do charge more for accredited calibrations than non-accredited. It is also by far our lowest profit margin that we earn, because if we tried to keep our profit margin on the cost of providing accredited calibrations, your calibration costs would go through the roof, which we do try to control, believe it or not.

Now, as to your question on whether you have to request accredited calibrations or not, that is outside of my arena of expertise. You have received some good comments on that, so I guess you have to make up your own mind on that. It does seem a bit ambiguous, though, and open to interpretation.

Ryan

 

[Montse]

Confuse about 7.6.3.2 17025 or acceptable via customer approved


According with 7.6.3.2 External Laboratory...The lab must be approved by the customer (with appropriate documented evidence) or accredited to ISO/IEC 17025 (or national equivalent).

I have an equipment that was calibrated by an external laboratory with recomendation of the OEM. However, the external laboratory is not accredited to ISO/IEC 17025 or national equivalent; only have the ISO 9002 certification.
We have thought to send them a survey calibration ( or audit them) in order that they prove us that the lab accomplishes with the Scope Lab. Could be a right procedure to approve them?

If we take the decision to send to the OEM, they do not have ISO/IEC 17025 certification, only have ISO 9002. Could be enought that they send us evidence that have a scope of Lab (by a survey or to audit them?

Sorry but I am confuse right now about the "or".

Tks in advance

 

[Hershal]

The question of accredited calibration is one that you must resolve with your TS registrar.

Ryan is completely correct about the steps involved in having a calibration that gets the logo on the cert. Accrediting bodies like ours spend a lot of time during an accreditation assessment ensuring that the techs can do the very things Ryan describes. That does add cost.

I can't speak for A2LA or NVLAP although I believe they would echo our position, but we consider the technical part to be absolutely crucial to the accreditation. The down side is the cost.

TS requires an accredited lab, and the natural assumption is that a cert with the logo is the evidence of a fulfillment of that requirement. I hope we will see the day when ALL calibration comes under the accredited scope of a lab and there is NO extra charge for a cert with the logo. That will only happen when the clients make an accredited calibration table stakes instead of an additional service.

With respect to manufacturers, many of them....even in the test equipment world.....are not accredited. As with accredited calibration, the client base must drive the accreditation of manufacturers. Fluke and Agilent are good examples of accredited labs. I can name many others who are not.

If an audit is made of the manufacturer, remember to take along a Metrology professional who is trained and experienced in the discipline(s) of interest, to accomplish the technical portion of the audit. If the TS requirement of an accredited lab is carried to the logical extent of including an audit of a non-accredited lab or manufacturer, then the audit must meet the requirements of ISO/IEC 17025 or the U.S. version ANSI/ISO/IEC 17025:2000. Otherwise, it is not equialent to an accreditation.

Hope this helps.

Hershal