Need Assistance Interpreting Gage R&R Data - Stylus profilometer

[Kevin Malone]

Hi Folks,

Here is a measurement study I performed on a stylus profilometer using (6) parts, (3) operators and 5 measurements per part.

Parts 5 and 6 are calibration artifacts. Part 5 was certified as 8636A +/- 40A.

Parts 1- 4 are representative of the parts that we will measure during manufacturing. The operators must interact with the system in order to level the data and extract the thickness of the metal step.

Any insights on what the data shows? I have JMP, but am just getting started in gage R&R.

Thanks,

Kevin

 

[James Oldham]

Kevin,

"Parts 5 and 6 are calibration artifacts..." - are you saying these are used to calibrate the gage, or are known "standards"? If so, I'd use these two parts to estimate the accuracy of the measurement device rather than use them in the gage study - just an opinion.

For the other 4 parts, run them in the Gage R&R in JMP. From these results, what do you think it's telling you? Not trying to be too snooty, just want to see if you've actually run the data in JMP. Maybe post the JMP results (preferably in a picture form so others can view it) and ask a specific question about the analysis.

 

[Atul Khandekar]

Parts 5 and 6 are calibration artifacts. Part 5 was certified as 8636A +/- 40A.

Parts 1- 4 are representative of the parts that we will measure during manufacturing. The operators must interact with the system in order to level the data and extract the thickness of the metal step.

Any insights on what the data shows? I have JMP, but am just getting started in gage R&R.

Hi Kevin, Welcome to the Cove!

You need to take 6 samples of the same part from your production. You have a part that measures 900 something and another that measures around 9000. Make sure you are not mixing parts with different 'nominal' specs.

I did not run the data through any software, but if you are using this data, my guess is that you will see very low GRR, high part variation(PV) and a very high nDC value.

 

[DZ-Jang]

Kevin ,
I have join the G R&R work for a few days in our company , and have done some analysis on parts thickness .But in our study we never measure or calculate any "Calibration artifacts" .
I`d like to know why u count the parts 5 and 6 in ,would u want to compare it with the other nominal parts ,but as Atul says it seems they r not under the same control standard .
we use ms excel or the software Minitab to calculate .