GR&R (Gage R&R) study on Minimum Force requirement

[rrofkar - 2009]

Our auditor had a recommendation that we perform a GR & R study on a pull force testor that is used to determine if the terminal crimp on a wire can withstand a minimum force. Is this possible using the standard MSA system? I don't understnad, can anyone help me ?
Confused

 

[Arvind]

It is not clear whether repeat reading of force measurement are possible in your case. If yes, i.e. if the force measurement is non destructive type then Gage R & R is relatively easy.
First you need to confirm that basic ISO expectations of gage like calibration etc are met.
You may, then, need to take representative production multiple samples which are measured two or more times by more than one inspectors.
It is desirable that same samples are measured by both operators (crossed)so that a more powerful ANOVA method can be used instead of Xbar R.
Any available commercial software can then do the work.

In case of destructive tests, it is not possible to do multiple measurements therefore conventional software cannot give you gage R & R. In that case, you need to find variation within a batch and compare with variation between the batches.
Trust this clarifies.

 

[rrofkar - 2009]

Thanks for your reply! I have tried the software but can only find it availble in the bilateral specificatons. There is nothing that I can find to do a unilaterial specifcation. The limit for the Pull test on a wire crimp is a 25 pound minimum. In another proces the measurement is 1.36 mm nominal. No ranges or limits on either. Doing the long verision of the AIAG study I come out with % variations of Equipment of 48.35; Appraiser of 50.81; and Part of 71.28. These are way over what I understand in the acceptable limit of 30 as stated in the AIAG MSA manual. Still unable to get my head around how to measure for a unilateral specification. Any suggestions?

 

[Arvind]

Gage R & R

Do you have raw data in excel? I can comment more accuarately if you give the data.
For gage R & R calculation, we donot need specifications. We are basically comparing how variation in measurement system compares to variation in parts.
For comparing precision of measurement system to Tolerance specifications, we need tolerance specifications.
Of the two, gage R & R is a better metric.

Arvind

 

[rrofkar - 2009]

Thanks for your quick reply! I do have the info on EXCEL but at this point I don't know how to send it on this reply? This is a wonderful site but I have not mastered all the tools yet.
25 lbs force minimum
A - Dick B - Chris C - Dawn
1 26.9 27.7 28.9 26.5 26.2 28.9 29.8 28 28.9
2 26.7 25.5 30.1 25.3 27.3 29.8 27.3 29.5 32.6
3 25.5 27.8 35.9 28.6 35.6 31.2 27.7 28.9 29.2
4 26 31.5 25.9 26.5 31.2 33 32.3 30.9 36.1
5 25.3 30 32 26.4 28.2 31.2 31.6 28.5 29.2
6 26 29.9 28.9 26.2 27.6 29.3 30.1 31.3 29
7 26.8 27 26 25.8 28.2 29.1 27.7 29.8 26.8
8 25.3 28.5 26.8 25.8 26.2 25.7 25.6 26.3 26.9
9 26.2 26 26.9 26 31.9 33.5 27.3 28.2 29.3
10 27.7 25.2 26.9 29.5 30.2 31.9 31.5 33.3 28.7

 

[Arvind]

Gage R & R on Force gage

Please clarify whether 9 readings for one sample in which of the formats?
1) A A A B B B C C C
OR
2) A B C A B C A B C

 

[rrofkar - 2009]

Number 1 AAABBBCCC. 10 readings each/

 

[Arvind]

Gage R & R on Force gage

Here are the findings for the data you have given. The gage R & R ( % study)is unusally poor at 93 %. Here are the main reasons.

1) High variability in repeatabilty readings is responsible for poor gage R & R. If we split contribution of repeatability and reproducibility then repeatability is 95 % and reproducibility is 5 %. Therefore focus needs to be improving repeatability.

2) If we compare repeatability variation with part variation, we find that part to part variation is 4.1. Compared to this, part # 3 measured by same operator A shows variation of 10.4 which is very high

3) You may like to find root causes for such high variability

4) % tolerance doesnot exist for unilateral tolerances like your case.

Arvind

 

[rrofkar - 2009]

Thank you for your help and generosity in resolving this issue of mine. The MSA has areference that anything over 30 should be of concern, 20 to 30 may be acceptable, and 10 and under is acceptable. I assume this is in percent although I haven't been able to find that .