Is a MSA (Measurement System Analysis) needed on these two tests

[optomist1]

Hi To All,

I could use some guidance regarding whether or not to perform an MSA in the following situation, the process I am investigating involves two separate inspections or tests, in the manufacture of non-production electrical harnesses:

1) a destructive tensile strength pull-test of a wire/terminal crimped connection; the test is performed with a digital force gauge with +/-0.5% accuracy, 0-100 lbs range and resolution to 0.1 lbs. The device comes with certificate of calibration traceable to NIST. Currently within calibration; annual re-calibration is required.

This device is being used to perform the above test; minimum spec. = 13 lbs

2) a simple point to point continuity test with a micro-processor based test set. This comes with a self-calibration/set of standards traceable to NIST, to perform annual in-house annual (minimum) re-calibration. Currently within calibration; the standard must be re-calibrated every two-years.

This device merely performs point-to-point continuity checks.

In both cases the tests are performed on "non-production" electrical harnesses.

Your input will be welcomed...thank you.

Regards,
Marty

 

[Jim Wynne]

Re: Is a MSA needed

Hi To All,

I could use some guidance regarding whether or not to perform an MSA in the following situation, the process I am investigating involves two separate inspections or tests, in the manufacture of non-production electrical harnesses:

1) a destructive tensile strength pull-test of a wire/terminal crimped connection; the test is performed with a digital force gauge with +/-0.5% accuracy, 0-100 lbs range and resolution to 0.1 lbs. The device comes with certificate of calibration traceable to NIST. Currently within calibration; annual re-calibration is required.

This device is being used to perform the above test; minimum spec. = 13 lbs

2) a simple point to point continuity test with a micro-processor based test set. This comes with a self-calibration/set of standards traceable to NIST, to perform annual in-house annual (minimum) re-calibration. Currently within calibration; the standard must be re-calibrated every two-years.

This device merely performs point-to-point continuity checks.

In both cases the tests are performed on "non-production" electrical harnesses.

Your input will be welcomed...thank you.

Regards,
Marty

How is it possible to manufacture (as you put it in the first sentence) "non-production" items? Can you clarify what's being made, and why?

 

[optomist1]

Re: Is a MSA needed

Hi Jim,

Thank you for the response...this may be semantics, I define non-production in this case as electrical cables/harnesses that a Tier 1 automotive supplier uses "to support" (test function) production. The cables/harnesses are not part of the finished consumer product.

Thanks for your response...great site....

Regards,
Marty

 

[Jim Wynne]

Re: Is a MSA needed

Hi Jim,

Thank you for the response...this may be semantics, I define non-production in this case as electrical cables/harnesses that a Tier 1 automotive supplier uses "to support" (test function) production. The cables/harnesses are not part of the finished consumer product.

Thanks for your response...great site....

Regards,
Marty

Assuming that you have to provide PPAP submissions for these things, your customer(s) will decide what the PPAPs must include. In general, the PPAP requirements are applied only to production parts (which is why the first "P" in "PPAP" stands for "Production"), but your customers might have other ideas.

 

[optomist1]

Hi Jim,

No we have no PPAP requirement; I am looking at this from an internal perspective; as to whether it makes "good sense" to perform an MSA even if there is no requirement.

Thanks again for your input....

Regards,
Marty

 

[Jim Wynne]

Hi Jim,

No we have no PPAP requirement; I am looking at this from an internal perspective; as to whether it makes "good sense" to perform an MSA even if there is no requirement.

Thanks again for your input....

Regards,
Marty

Whether or not you do it depends on your confidence in your measurement system(s) and the risk involved, if any--the likelihood that something bad will happen because of measurement error.

 

[optomist1]

Thanks Jim,

In this case it comes down to a judgement call based upon my process knowledge and that of the production manager.

Your input was a great help..

Regards,
Marty

 

[NumberCruncher]

Hi Optomist1

I won't be so presumptuous as to tell you whether you should or should not carry out an MSA (I'm not going to be responsible for your judgement call!). But here are some things to consider.

Which of these tests are used in production? Whichever is used, that is the one to concentrate on regarding a potential MSA. That is the one which is actually being used to check goods before they go out of the door. I assume it isn't the pulloff test as this is destructive and you wouldn't have any test harnesses for use!


Regarding the pulloff test.

Given the specification of the gauge, if it is correctly calibrated, it should be more than adequate for the job. With a resolution of 0.1 lb and a data range of about 16 to 36 lb, you can make measurements to less than 1% of reading.

This is a destructive test. As you can't measure the same terminal twice, an MSA study has to make an approximation.

"Analysis of these systems [ie non replicable or destructive] will depend on whether a homogeneous set of parts (small part to part variation) can be found to represent a single part"

See page 153 of the MSA manual (4th edition)

Where are you going to find your homogeneous sample of crimped connections? You would need to be able to confidently say that "All of these connections will fail at about 25 lb force, all of those will fail at about 30 lb force.." and so on.

If you intend to use pulloff force as a measure of

I think that the most that you could determine with MSA would be reproducibility. Do all of your staff carry out the test to the same level. Is the reproducibility of this test important to know?


Regarding the continuity test.

I will guess that this is the actual test that is used in production as it is non destructive.

You said in a previous post that one of your staff was not carrying out the crimping correctly (paraphrasing wildly, but I think the gist of the statement is there). Possibly your testing method was not good enough to pick up the bad crimps. There was electrical continuity, but a weak mechanical connection.

An MSA study is certainly a possibility.

A final, probably fruitless thought

Does your continuity tester just give a pass/fail or does it give you an actual reading in ohms?

If it gives you a reading in ohms, do the weakest connections give the highest resistance? I doubt that you will find any difference in most crimps. But you got one crimp in your last set of data that was very low. Did this have a significantly higher electrical resistance?

I suspect a connection would have to be very bad before it shows up as a high resistance joint, so bad that it would be obvious without testing. But who knows?

NC

 

[optomist1]

Good Day NC,

Thanks for the thoughtful, thorough response......

Tensile Test
It seems as though the general de facto guide for "test resolution" is 10:1, so if I am measuring (or the specification) to 1 pound then the test device or instrument must measure or resolve to 0.1 lbs. Makes sense.

Great point regarding attempting to conduct a MSA on a destructive test....the instrument is properly calibrated and traced to NIST....I'll leave this dog for another day.

Continuity Test
The typical resistance threshold for a "pure cable assembly", no components is 5 ohms; although the device can be programmed to output the actual resistance reading, it typically reads out a pass/fail based on the user programmed 5 ohm threshold. Regarding a possible difference in resistance readings.....the visually "bad crimp" aside, I doubt it. Although this may be worth further investigation.

In both cases the devices are properly calibrated and traced to NIST; no MSA for me. But the "Must An MSA Performed" exercise or question must be asked and answered.

Thanks again for the thorough thought provoking response.....

Regards,
Marty