everybody
i met a supplier that first she told me the attribute date (such as Go/ No Go gauge) can not do the MSA study, after i show the AIAG to her, then she said to me for all attribute Gauges one MSA study is enough!:biglaugh:

she means all Go/ No Go gauge is similar, so they need do a MSA study!
anybody can tell me do we need do separate attribute MSA study for maybe each Go/No Go gauge?

thanks very much!:applause:

I can't imagine doing one MSA study to cover a group of gages. One gage can be within specs while another is not. I do not understand the idea of group GR&R. :confused:

The main purpose we do the MSA is to segregate the measurement variation and the actual process variation from the observed variation in the line.

We can study what ever we want in the MSA like stability, linearity, bias or GR&R on the go no go gage but it is depended to its intend use or application in the line.

We might need to understand on how the design team set the gap allowance to control or detect our non conformance product. Are they using the worst case analysis? What is the gage limitation? Do we have any alpha or beta risk when we use the gage?

After we understand and have the clear cut on its limitation then only we can evaluate its efficiency of the gage by using the attribute GR&R. Beside that any cosmetic reject can be caused by the application of the gage? Any problem on the ergonomic issue when doing over time by an operator?

I believed if we have above outcomes in the validation study then we can qualify the gage and release to the line and keeping monitoring on its worn out time or stability.

I can't imagine doing one MSA study to cover a group of gages. One gage can be within specs while another is not. I do not understand the idea of group GR&R. :confused:

When I have a dozen or better pin/plug gages for a part, I pick the tightest toleranced feature and perform a Gage R&R on that pin/plug gage. That covers all the others in my thinking. 'In or Out of specs" is taken care of through the calibration process. This is strictly an economic choice of my company.

Functional type gages (MMC pins) are a different matter. They are typically dedicated and Gage R&Rs are routinely done on those.

:2cents:

A very good question, and I don’t thing there is a definitive answer, it all depends on the gauge, part, process and company.

It has been my insistence that all these ‘monitoring’/ ‘understanding the process’ control methods must not take over the business to such a degree it affects the bottom line, or that it takes over from what the business is about. . . making money!

With regards to the question, you shouldn’t be doing, in my opinion, a Gauge R&R on every single measuring device within a device type its counter productive, and frankly stupid. For instance, the company I’m currently work for has over 300 0 - 25mm micrometers that are all used to measure various parts, you think I’m going to instruct them to do a Gauge R&R for each micrometer, I don’t think so, it would not benefit them in the slightest. I would however, instruct them to do a Gauge R&R for each part.

At the end of the day, providing the gauge is fit for purpose then you have be able to group them, there is no other feasible alternative. Having said that this instruction is based on micrometer type devices, I wouldn’t group CMM’s laser trackers and the like.

Another thing, yes it is true your trying to separate measurement variation from the process, but what your really doing is making sure the final result is not influenced by the gauge itself, albeit accuracy/ repeatability, the operator influence, the environment (is the guy stretching over to take a measurement and only takes a quick rough measurement) and so on. It’s the full measurement process not just the gauge.

Rergards

Generally:

1. A critical measurement step in the process (for monitoring, SPC...)
2. Final inspection prior to shipping to the customer
3. Where customer demand MSA.

even if they are gages of the same type / similar. (my 2c)