G
gavinqce
Any clues on how to complete gage R&R when you have only a single sided tolerance?
Thanks
Gavin
Thanks
Gavin
How to complete gage R&R when you have only a single sided tolerance
The method is the same, but you should use %GRR against Total Variation (TV).
If you must use P/T Ratio, there are a couple options.
If your tolerance is a maximum such as maximum flatness, it is legitimate to use zero as a minimum because flatness cannot be less than zero. However, the farther from zero your maximum gets, the less likely that this is a realistic method.
In this case, or in the case where the spec is a minimum, you may use a calculation similar to the Cpk calculation where you use the difference between the part averages and use 3 sigma instead of 6 sigma.
I will warn you in advance that there will be a lot of disagreement on this issue.
If you must use P/T Ratio, there are a couple options.
If your tolerance is a maximum such as maximum flatness, it is legitimate to use zero as a minimum because flatness cannot be less than zero. However, the farther from zero your maximum gets, the less likely that this is a realistic method.
In this case, or in the case where the spec is a minimum, you may use a calculation similar to the Cpk calculation where you use the difference between the part averages and use 3 sigma instead of 6 sigma.
I will warn you in advance that there will be a lot of disagreement on this issue.
gavinqce said:
If you take a minute an look at the 'Similar Threads' list at the bottom of this thread page you will find several discussion threads about gage R&R when you only have a single sided tolerance. A Search in the forums will also reveal a number of existing discussion threads about single sided tolerance gage R&Rs.
Y
Yew Jin
hi guys,
Please give me some input regarding the format that can be used for the one sided specification for GR&G study. Please see the attachment.
Thanks and regards
YJ
Please give me some input regarding the format that can be used for the one sided specification for GR&G study. Please see the attachment.
Thanks and regards
YJ
L
lee01
Okay, firstly I would highly recommend you get rid of any excel spreadsheets for GRR, I found roughly 6 different GRR’s forms and all gave different results and as the AIAG doesn’t recommend any spreadsheet per say, well you have no correct Spreadsheet or bad Spreadsheet.
Secondly, I have let to see a spreadsheet that will do single sided tolerances. When you enter the product (feature) tolerance you enter the total tolerance (+/-0,010mm would be entered as 0,020mm in the relevant cell) and the calculations are straight forward.
However if the same tolerance is used (+0,020mm) the spreadsheet automatically thinks you mean +/-0,010 and so it will calculate readings accordingly. So lets say on average, readings are mid limit to the 0,020mm single sided tolerance (perhaps how you would like them) the spreadsheet things the results are actually on the top limit, so you end up with really bad GRR results.
In Minitab, you have the option to enter single sided tolerances and the results are massively different as Minitab is calculating the readings based on single sided tolerance.
I had the same problem when choosing a suitable method of hole position measurement, as hole position GD&T callouts are diametric and not the previous standard of +/-. I couldn’t find a measurement device that had a good enough GRR result until I used Minitab. Minitab is my best tool and its soooo cheap to buy!
Regards
Lee01
Secondly, I have let to see a spreadsheet that will do single sided tolerances. When you enter the product (feature) tolerance you enter the total tolerance (+/-0,010mm would be entered as 0,020mm in the relevant cell) and the calculations are straight forward.
However if the same tolerance is used (+0,020mm) the spreadsheet automatically thinks you mean +/-0,010 and so it will calculate readings accordingly. So lets say on average, readings are mid limit to the 0,020mm single sided tolerance (perhaps how you would like them) the spreadsheet things the results are actually on the top limit, so you end up with really bad GRR results.
In Minitab, you have the option to enter single sided tolerances and the results are massively different as Minitab is calculating the readings based on single sided tolerance.
I had the same problem when choosing a suitable method of hole position measurement, as hole position GD&T callouts are diametric and not the previous standard of +/-. I couldn’t find a measurement device that had a good enough GRR result until I used Minitab. Minitab is my best tool and its soooo cheap to buy!
Regards
Lee01
J
jrubio
Why do not you use the Delphi´s Templete?
Reinventing the wheel?....
It has this option.
T
Teknow
Why do not you use the Delphi´s Templete?
Reinventing the wheel?....
It has this option.
Any chance of posting a link to this template?
J
jrubio
A christmas present following AIAG mandates. a few templetes Press one side button.
The Delphi R&R do not include the one sided as I thought.
My appologies.
The Delphi R&R do not include the one sided as I thought.
My appologies.
Last edited by a moderator:
Thanks jrubio.
Truly a great Christmas present.
Wish you a Happy Christmas and a New Year.
L
llapanowski - 2007
The spreadsheet jrubio posted R&R 1 Sheet.xls is outdated. It is not using the correct Kx factors, at least the ones that are in the MSA 3rd edition. According to the 3rd edition you should use the new kx factors and when solving based on tolerance divide the approriate metric by 6*Tolerance. Of course this is for a bilateral tolerance. I'm still trying to find what to do with unilateral tolerances.
Similar threads
