New to Metrology and told to complete Gage R&R with AIAG compliant software

B

Brackish

Hello All.

Long story short. I have been a CMM arm operator for the past year and now I have to run the Metrology department. But this has come at the worst possible time.
I have been asked to find an AIAG compliant software to complete the Gage R&R and MSA study for TS 16949.

It is to my understanding that there is not a program for the GR&R but an AIAG compliant template (.xls) and then there is software for the MSA (minitab?).

The way I have perceived what has been asked of me is that I need to complete the GR&R and then run the MSA on the results. So from that I think I need the template and then to run the GR&R's on our equipment then complete an MSA? What has really got me worried is that we are on re run from the Auditors in 1 month.

Any advice would be greatly appreciated so I can approach my Manager with confidence of pending failure!

Regards,
Brackish
 
B

Brackish

Re: New to Metrology and told to complete Gage R&R

Thank you Miner. :thanks:

The info you have provided is a great starting point. I tried to explain it to my manager but time is what we do not have. Far as I can see I can not and certainly should not rush this. 1 month is not enough to correct all actions from the audit report including GRR's for all measuring equipment.
 

Golfman25

Trusted Information Resource
Re: New to Metrology and told to complete Gage R&R

My advice is to start drinking. :)

There are some commercial packages out there which will do the Gage R&R. We use gagepack, which also handles our calibration.

Good luck. You're going to need it.
 

Hershal

Metrologist-Auditor
Trusted Information Resource
Re: New to Metrology and told to complete Gage R&R

Others more qualified have commented on the GR&R, so I won't.

However, for metrology, you need uncertainty analysis. It is one of the two components of traceability, the documented, unbroken chain of comparisons being the other. GR&R by itself will not provide uncertainty. However, it provides a solid beginning for the Type A (aka Random) contributions. You still need the Type B (aka Systemic) contributions to complete uncertainty calculations.

To help with the uncertainty, you should be able to download TN 1297 from NIST. It is a free download.

Hope this helps.
 
B

Brackish

Re: New to Metrology and told to complete Gage R&R

Hello again!

Ok so I have Miner's GRR template and I understand how to conduct a GRR on a gage block with a vernier and micrometer but what about a CMM protable arm?

Do TS16949 auditers requre all GRR's on production parts only? Our production parts all have varying tolerances and the accuracy of a portable can vary within its volumetic range.

(I say I understand but I probably do not:lol:)
 
G

Grimaskr

Re: New to Metrology and told to complete Gage R&R

Do TS16949 auditers requre all GRR's on production parts only? Our production parts all have varying tolerances and the accuracy of a portable can vary within its volumetic range.

I'm not familiar with CMM portable arms, but here's some general advice regarding production parts:

1) I'd start with the product that is most challenging for your system, perhaps one utilizing the most volume of the arm's reach?

2) I'd use production parts, wherever possible. One common issue is not having production parts that vary enough across your spec.

I think the rule of thumb is that your parts should represent at least 85% of your spec and can go up to about 115% of your spec. You want to ensure your system measures accurately throughout the range of your spec.
 
B

Brackish

Re: New to Metrology and told to complete Gage R&R

Ok that is food for thought!

I have a part or 2 in mind but the different inspection stages in the program have varying tolerances, according to customer spec, so I would in theory choose say one stage that has 10 points that are to the same tolerance. Example: contact probe points on an A surface that span from the top of the part to the bottom? The said area has a generous tolerance of +- 1.0mm but the riv nuts have a tolerance of +- 0.25mm. The CMM arm is acurate to 0.057mm (the arms accuracy we tend to say is closer to 0.090mm factoring in the operator and environment) also the more the arm is at its extremities the less accurate they become so that means I should be able to have an accurate measurement system but will see poor precision from the gage in the results?

I have asked for Minitab 17 but it does not look like I will get it so it is the hard route all the way!
 
B

Brackish

Re: New to Metrology and told to complete Gage R&R

Hi All,

Spent the day reading through the the AIAG 4th MSA manual and Miner's Blog!

If possible could someone help me with:

Parts... 10 parts, 3 trials by 3 operators. But with regards to the parts do we measure all features relating to drawing/ customer requirements but using the lowest tolerance and then randomly choose the dimensions to input into the GRR template? Do they have to be the same part?


Validation of the software. This was one of our non conformance's from the audit. I have read through Miner's blog and tried to understand the Ford validation but to validate the software is it as simple as input the data from the Ford document into the GRR template? Or do I have to provide some form of reasoning and method as to how I came to the conclusion that the template is AIAG compliant? Miner's template has the readings from the manual and shows results.

I apologize if my questions are confusing.

Kind regards,

Brackish.
 
G

Grimaskr

Re: New to Metrology and told to complete Gage R&R

I've never had an auditor expect a measurement system analysis to cover every dimension of every part that's being measured by the gage.

Assuming the gage is used the same way with the same procedure by all operators for all parts, I usually aim for my most challenging measurement on my most challenging part. If I pass that, I feel good about my system. If I don't pass that first attempt... I dig deeper and start troubleshooting.

For instance, choose one product (maybe a big piece that forces you to go to the outside of your measurement volume). Then choose one dimension, usually something small with a tight customer spec range. Then get 10 of those parts and measure that one dimension 3 times each with 3 operators.

The key I mentioned above is making sure those 10 parts represent the full range of your spec, or you may fail because you don't have much part-to-part variation in the analysis. If your spec is "5.0mm +/- 0.5mm", try to include a part around 5.5mm and a part around 4.5mm in your 10-part trial.
 
Top Bottom