Does anybody can give a help about GR&R

[jackylpt]

It is the result of GR&R:
Gage R&R %Contribution
Source VarComp (of VarComp)
Total Gage R&R 0.236594 38.81
Repeatability 0.228623 37.50
Reproducibility 0.007971 1.31
operator 0.007971 1.31
Part-To-Part 0.373007 61.19
Total Variation 0.609601 100.00

Study Var %Study Var
Source StdDev (SD) (5.15 * SD) (%SV)
Total Gage R&R 0.286410 2.50501 62.30
Repeatability 0.278146 2.46245 61.24
Reproducibility 0.089281 0.45979 11.43
operator 0.089281 0.45979 11.43
Part-To-Part 0.710743 3.14533 78.22
Total Variation 0.780770 4.02096 100.00

Question:
I want to know which factor which I can use it to judge the Gage?
"study var", "%SV" , " %Contribution " or others?

1. according to " %Contribution" , it is 38.81%, then the gage is bad,
2. accroding to " study var,stdDev(SD)", it is 28.64%, means gage can be accepted.
3. according to "%SV" , it is 62.30%, also means gage is bad.

because I found the book of "Implementing six sigma" say to use "% contribution", my MBB let me use " study var,stdDev(SD)", and my training document say to use "%SV", now I am confusing.

By the way , there is probably a error about the data of this result, but I really just want to know which one I can use to judge the gage?

Does anyone can help me? thanks

Michael

 

[Jim Wynne]

Have a look at this similar thread:
Gage R&R - Indices - My list - What do you think? Any suggestions?

 

[Miner]

It is the result of GR&R:
Gage R&R %Contribution
Source VarComp (of VarComp)
Total Gage R&R 0.236594 38.81
Repeatability 0.228623 37.50
Reproducibility 0.007971 1.31
operator 0.007971 1.31
Part-To-Part 0.373007 61.19
Total Variation 0.609601 100.00

Study Var %Study Var
Source StdDev (SD) (5.15 * SD) (%SV)
Total Gage R&R 0.286410 2.50501 62.30
Repeatability 0.278146 2.46245 61.24
Reproducibility 0.089281 0.45979 11.43
operator 0.089281 0.45979 11.43
Part-To-Part 0.710743 3.14533 78.22
Total Variation 0.780770 4.02096 100.00

Question:
I want to know which factor which I can use it to judge the Gage?
"study var", "%SV" , " %Contribution " or others?

1. according to " %Contribution" , it is 38.81%, then the gage is bad,
2. accroding to " study var,stdDev(SD)", it is 28.64%, means gage can be accepted.
3. according to "%SV" , it is 62.30%, also means gage is bad.

because I found the book of "Implementing six sigma" say to use "% contribution", my MBB let me use " study var,stdDev(SD)", and my training document say to use "%SV", now I am confusing.

By the way , there is probably a error about the data of this result, but I really just want to know which one I can use to judge the gage?

Does anyone can help me? thanks

Michael

The data that you provided appears to be from Minitab, which will also provide a %Tol column when you enter specification in one of the options tabs. The answer to your question depends on what you are trying to determine from the gage R&R in the first place. Lets take the three columns individually:

% Tolerance
This compares the gage variation to the tolerance and answers the question "Is the gage good enough to distinguish variation within the tolerance range?" This is where the < 10% is good, 10 - 30% may be acceptable and > 30% is no good criteria is typically used. If this % get too large, you increase your Alpha and Beta risks of passing bad parts and rejecting good parts.

% Study
This compares the gage variation to the process variation and answers the question "Is the gage good enough to distinguish variation in the product itself?" This becomes very important when trying to determine the impact of making process changes introduced by DOEs, Six Sigma studies, etc. It also has an impact on capability studies performed with the gage because it will inflate the observed variation. I have not seen good criteria established for this ratio, but this "roughly" corresponds to Wheeler's Discrimination Ratio (minimum of 4), so I would recommend a maximum of 25%. In reality, the smaller the change that you are trying to detect, the smaller this % needs to be.

% Contribution
This simply shows the relative contribution of each component of variation including the part variation. It shows you which to focus on first. One key item to look at is the part variation %. If it is low compared to the gage, look at whether your parts were representative of the full range of process variability. If it is representative, you probably have a gage issue.

In summary, use the ratio appropriate to the question that you want answered.

 

[jackylpt]

Thanks so much for JSW05 and Miner's kind help.