Hi,
TS16949:2002 Clause 7.6.1 says: This requirement (Measurement system analysis) shall apply to measurement systems referenced in the control plan. One of our projects document (APQP) references 50 calipers in the Control Plan. And our consultant persuaded us to analyse EVERY caliper involved measurement system…. :bonk: Gosh, we are really unable to afford it. Is it necessary?
Thanks

This shall be applied to measurement systems and not to the each measuring instrument referenced in the control plan....In my opinion we can have MSA(as per 3rd edition) done on each type of instruments. If someone can throw more light...

vasu

No this is not necessary - your consultant is incorrect. As long as your calipers are all the same type of gage - you need only one gage r & r. Your registrar should know that and if they insist, I would take it to the top for interpretation issues. My knowledge in calipers is that there arent many different types. You may want to do a separate one on a manual caliper and an electronic caliper.

Hi,
TS16949:2002 Clause 7.6.1 says: This requirement (Measurement system analysis) shall apply to measurement systems referenced in the control plan. One of our projects document (APQP) references 50 calipers in the Control Plan. And our consultant persuaded us to analyse EVERY caliper involved measurement system…. :bonk: Gosh, we are really unable to afford it. Is it necessary?
Thanks

Until I know the facts I would have to side with your consultant.
A measuring sysrem not is not just the gage only, but also includes the operator.
If the 50 calipers are used one at a time by the same operator then yes one MSA is acceptable. However if the 50 calipers are all used at the same time by 50 different operators then you would need to do an MSA on each one.

Yes, Sam you are right and I agree with you(the system of measurement shall consist with gage, product,characteristic,environment,operator,method ect.)the calipers are used at different time by 50 different operators on different characteristics...does it means that we have to do MSAs on each caliper??

Sam - by your logic I would need an R&R for every gage/employee/weather/characteristic combination possible. I really do not believe that is the intent of MSA's

If so, I need to hire lots and lots of folks to do all my MSA's! :bonk:

It has been quite awhile since I've worked in production automotive, but I seem to remember QS-9000 requiring GR&R only for gages requiring a "set up", and that hand tools, such as dial/digital calipers and micrometers were not included - that calibration controls were sufficient.

I tend to agree with "GoKats78", especially if operators/inspectors are changing frequently. That could get to be a real mess. I guess it all comes down to risk vs. benefits, and how well you are at reasoning it out with your consultant.

I've got more to say - but I gotta go....

Sam - by your logic I would need an R&R for every gage/employee/weather/characteristic combination possible. I really do not believe that is the intent of MSA's

If so, I need to hire lots and lots of folks to do all my MSA's! :bonk:

True. But then, what sense would it make to have only conducted one MSA while there are 49 others that have not been conducted. That says to me you can only be certain that 1/50th of your product/process is acceptable.

If the same two or three operators are going to use all 50 calipers then by all means do one MSA. That will be accdeptable as long as the calipers have met the calibration criteria. However, if you have 100 people that are going to use the 50 calipers then you would need to conduct 50 MSA's.

I don't understand why you would need to hire additional people.

Here's what I always understood:

We are going to measure the thickness of a sheet of friction material - it is raw (soft) and has not been pressed yet. We decide to use micrometers - so we take a micrometer that is known to be calibrated, get three operators together, and have each operator (all using the same mics) to measure 10 spots three times each (3 operators, 10 samples measured 3 times each by each operator). Say we come up with an unnacceptable GR&R results - the variation in measuring results deems the METHOD unnacceptable. So - we decide that the micrometer is not repeatable or reproducable when dealing with non-inspection employees (skill level) because the "feel" of the gage is so critical when dealing with the soft material (it's too easy to squeeze). Now we have choices - do we limit the inspection of this characteristic to Inspection Personnel only? - or do we use perhaps a snap gage that anyone can use regardless of their skill level?

If you've verified the inspection method, any micrometer that is in calibration will be acceptable to use. It's not about this particular micrometer, and this particular micrometer and this particular micrometer - it's about using a micrometer for measuring this characteristic. Back when I used to do GR&R for automotive companies we did not do it for every gage - it's not really about the gage itself, but the type of gage (or measuring instrument) you've chosen for a particular characteristic and the ability for anyone who has to measure said characteristic will get the same measurement as the next person. Things may have changed...or maybe we were wrong back then - but our customers were happy. :bigwave:

Cari

YOU ARE RIGHT.....I think the intent is more of designing a measurement system exclusive to each measurement characteristic that requires to be verified during the production....it helps more in preventing a failure on account of a wrong measurement system.

Also see: http://Elsmar.com/Forums/showthread.php?t=1819

The standard says "each type of measuring and test equipment system."
The normal interpretaion is type refers to the equipment, digital caliber one type, analog caliber another type.
But the system can include the operator then the number of permutations becomes impossible.
I think that there are 2 aspects:
1) in relation to the tool then MSA with the usuall operators, I have requested at PPAP R&R from the actual designated operators and not the quality people who normally do this to get it done.

2) The operators should as part of their training and qualification participate in R&R and this should be the proof of the effectiveness of their training

2) The operators should as part of their training and qualification participate in R&R and this should be the proof of the effectiveness of their training
Howard,
I like this thought. It is so simple yet I never thought of it thanks. A training record proven effective by its MSA being conforming or not. Thank you!!

Mark

Hi,
TS16949:2002 Clause 7.6.1 says: This requirement (Measurement system analysis) shall apply to measurement systems referenced in the control plan. One of our projects document (APQP) references 50 calipers in the Control Plan. And our consultant persuaded us to analyse EVERY caliper involved measurement system…. :bonk: Gosh, we are really unable to afford it. Is it necessary?
Thanks


So is one MSA for 50 calipers enough? Or any rule of thumb of randomly picking how many calipers to do MSA?

So is one MSA for 50 calipers enough? Or any rule of thumb of randomly picking how many calipers to do MSA?

tohbin

I ran into simular situation and so far the following solution has been accepted but I suppose someone will object eventually

One: Perform Standard GR&R using operators/inspectors normally doing work to prove the measurement system.
Two: Take appraiser with least range variation and perform a study with one operator X three calipers X 10 parts. Enter into the calculation which effectively reverses the equipment and appraiser values. This should "prove" if the equipment is consistent enough to be interchangable.
Three: Each Operator that might be used must have passed GR&R in each "measurement system" as part as part of their training effectness.

Maybe someone else in the cove can tell me where it will fall down but up to now as a policy its been accepted for simple hand held instruments,.

MSA 3 Chap 1 Section C Page 23

Martin

So is one MSA for 50 calipers enough? Or any rule of thumb of randomly picking how many calipers to do MSA?

The requirement of the MSA is; a gage and operators,)2/3 recommended. If the same operators from the initial study will use the 50 calipers, then one MSA is sufficient. However, if other operators are going to use the caliper then additional MSA will be required.

I still feel that this entire MSA evaluation has to find it's way into the production system more from the FMEA's rather than the Control Plans (though the standard explictly mentions the source as Control Plan), I think it is appropriate to investigate at a FMEA stage whether a measurement system could influence the product / process quality of a characteristic being addressed and set targets for the required repeatability and reproducibility, which could be subsequently identified in the Control Plans. Only such characteristics could be put on a regular schedule for peridoic MSA evaluation and corrective action as well as re-visited whenever FMEA's are revised.

We evolved a similar system in my previous organization which was both productive and acceptable to our registrars. More so, our Customers appreciated it.

Larry Barrentine recommends the following short-cut method where there are a large number of one type of gages used - such as Micrometers: (Concepts for R&R Studies, 2nd Ed. ASQ Press)

Use 4 operators and 10 samples. The key is that each operator uses a DIFFERENT micrometer. The reproducibility estimate will also include instrument/calibration effect combined with operator effect. If normal R&R calculation results are acceptable, you have qualified 4 micrometers in one study. If not, do R&R on each instrument.

The intent of MSA is to evaluate the width variation ( which extent the measurements can vary from person to person and in number of trial ) and Location variation ( how far the actual value departuring from it's reference value )

It is good enough to conduct MSA for a family of instruments as it cover person to person variation ( Gage R&R ), time variation ( Stability ) ....

This MSA applied to the measurement systems specified in Control plan.

We have never done MSA studies on every gauge, all of our calipers, micrometers, indicators are the same, so one MSA study out of each "family" will do. We have passed many TS audits and customer (GM, Chrysler, Valeo, Behr, etc...) audits this way and it has never been a problem. :biglaugh: