Gage R & R - Do we specifically have to do Gage R&R on every gauge?

[Dawn]

Do we specifically have to do Gage R & R on every gauge to be in compliance with QS? Is it customer control plan driven? Everyone I talk to tells me a different story. Why doesn't QS specify instead of playing games?
Is this standard something anyone wants to play games with?

 

[Marc]

Rule of thumb is if it's on the control plan, you have to evaluate it. R&R is only one way but is what most folks use.

 

[Christian Lupo]

Actually QS does specify, that all measurement systems in the control plan must be evaluated using the MSA or an equivalent evaluation system.

 

[Marc]

Christian and I are saying the same thing. I used the 'rule of thumb' statement because you may find other instruments you want to do an analysis on but as a minimum you must address those on the control plan.

 

[Arvind]

If you are OK with process capability then Gage R & R doesnot matter

Think of it from the first principle why we do gage R & R before evaluating process capability.
Gage R & R will only reduce the process capability compared to ideal state. Therefore when capability is low, you need to know whether low capability is a result of process common cause or poor measurement system!
If you donot know exact answer, you may end up in spending money on process whereas you should have spending on measurement system.
However, if your process is robust and if you have a proof of satisfactory performance from customer point of view, you donot need to do gage R & R just for the sake of it.

It is good idea to use engineering judgement rather than checking the box approach.

Arvind

 

[Caster]

Do we specifically have to do Gage R & R on every gauge to be in compliance with QS? Is it customer control plan driven? Everyone I talk to tells me a different story. Why doesn't QS specify instead of playing games?

The standard is pretty clear. I call this Snakes & Ladders.

See how quickly it creates several hundred pages of reading in 3 books?

Measurement System Analysis - 4.11.4



....This requirement shall apply to measurement systems referenced in the customer approved Control Plan (see 4.2.3.7). The analytical methods and acceptance criteria used should conform to those in the Measurement Systems Analysis reference manual [e.g. bias, linearity, stability, repeatability and reproducibility studies).....

The Control Plan - 4.2.3.7

Refer to the Production Part Approval Process manual.

PPAP Manual v3 page 8 para I.2.2.10 MSA Studies see also MSA manual

It is understood that each "type" of gage mentioned on the CP is to be studied.

We do not do MSA on every Go No Go plug gage (thousands and thousands). We just do by type.

This has survived 3 different registrars and PPAP submissions are clearly indicated as "type".

Good luck, nothing is a frustrating as trying to "lawyer" the 600 plus pages of QS 9000.

 

[martin elliott]

Playing the game or common sense

Of course don't forget MSA 3 Pg 23 which notes "common sense is the guide in any case" - always assuming your common sense is the same as the auditor/customer!!!!

In my opinion It does not say use GR&R in all cases, it says use the appropriate MSA technique for the situation. This could be as little as calibration trend analysis, to the full box of MSA toys.

I have found to date (famous last words), by specifing the MSA considered to be appropriate during APQP, and I use an extra column on my version of the Control Plan to record this, none yet have questioned me too deep to raise a NCR.

Big problem is rewriting historic Plans just to show this as I've only got nine years to retirement.

Martin in the sunny UK

 

[Joe_winter]

MSA just for equipments for calibration in PPAP?

as Arvind said, MSA is just for defects analysis but not mistake-proofed.
in PPAP cl.I.2.2.10, for all equipment used for new or modified gages, , do you think so??
another question, R&R is facilitated, but bias, linearity, stability studies are all difficult to get. These are also required? no, my point of view. Are agreed upon?

 

[Marc]

Also see: http://Elsmar.com/Forums/showthread.php?t=10270