Variable Gage R&R - Two operators instead of three

[BSwan]

Good afternoon,

Quick questions. We are doing variable gage R&R's on our attribute fixtures and using 2 operators, 10 parts, 3 trials.

I understand the range method is 2 operators, 5 parts, 1 trial.

I think we are going above and beyond that and we can determine some the causes of our variation with the method we are using.

I'm concerned during audit time that people may be looking for a third operator. Right now our only customer is general motors and our IATF is coming up in a couple months. Has anyone ever had any issues with this concern? Thoughts on our method?

 

[AMIT BALLAL]

re: Variable Gage R&R - Two operators instead of three

Good afternoon,

Quick questions. We are doing gage R&R's on our attribute fixtures and using 2 operators, 10 parts, 3 trials.

I understand the range method is 2 operators, 5 parts, 1 trial.

I think we are going above and beyond that and we can determine some the causes of our variation with the method we are using.

I'm concerned during audit time that people may be looking for a third operator. Right now our only customer is general motors and our IATF is coming up in a couple months. Has anyone ever had any issues with this concern? Thoughts on our method?

TBMK, Range method you have specified is used for variable studies, not for attribute.

Sent from my CPH1821 using Tapatalk

 

[BSwan]

re: Variable Gage R&R - Two operators instead of three

Its an attribute gage with tons of fixed stab pins, but we are using the SPC bushings for our gage R&R data.....so its essentially a variable study. Attribute gage with some added SPC points. Variable Gage R&R study. Bad clarification on my part.

 

[Miner]

Since you are using 10 parts and 3 replicates, you are complying with this portion of the standard AIAG approach to an R&R study. The only difference is in your use of 2 vs. 3 operators.

There are two possible scenarios.

1. There are only 2 operators that ever perform this measurement in actuality. In this scenario, you are fully justified in only using two operators (the same two) in your study, and the results are reflective of reality.
2. The other scenario is that the two operators are representative of a larger group of operators that could perform this measurement. In this scenario, you run a risk that two operators MAY underestimate the amount of Reproducibility variation in your system. At a minimum, the degree of uncertainty in the estimate is greater. This would be more difficult to justify.

Regarding"I understand the range method is 2 operators, 5 parts, 1 trial.", where have you seen this? 5 parts is next to useless, and 1 trial will not estimate repeatability.

 

[AMIT BALLAL]

Regarding"I understand the range method is 2 operators, 5 parts, 1 trial.", where have you seen this? 5 parts is next to useless, and 1 trial will not estimate repeatability.

This method is specified in AIAG MSA 4th edition manual on page 102. But this method is to quick check change in GRR / to get overall picture of the measurement system (as specified in this manual), not to determine repeatability and reproducibility. I also don't find it useful.