The Gage Repeatability and Reproducibility study helps you to determine how much of the observed process variation is due to measurement system variation. It breaks down the overall variation into variation due to part-to-part, repeatability and reproducibility. The ANOVA method is commonly used as it also breaks fown reproducibility into operator and operator by part interactions. This tool can be used on crossed Gage R&R studies at which each part is measured by each operator several times.
The following are some guidances to set up a Gage R&R study.
Parts measured should cover the entire range of possible values to get a better estimate of part-to-part variation
If you can run limited number of tests, it is better to test more parts than take more measurements on the same part to get a better chance of detecting linearity issues
Ensure complete randomization of the measurement process. Repeating measurement of the same part by the same operator too often may produce over-optimistic estimates.
If the variation among operators for the same part is large, you may need more operators to get a better estimate of reproducibility
At the Excel Menu (For Excel 2007, go to Add-ins first)
Choose ProcessMA > Quality Tools > Gage R and R > Gage R and R
In Measurement, select the column containing measurement data (Numeric)
In Part, select the column containing part indicators (>=2 distinct values)
In Operator, select the column containing operator indicators
In Method of Analysis, choose the method to use for analysis. The Xbar and R method distinguishes the observed process variation into part-to-part, repeatability and reproducibility components. The ANOVA method is similar but it also breaks down reproducibility into operator and operator by part components
In Study Variation, enter a value to be used to calculate Study Variation. The default is 6, which is the number of standard deviations needed to capture 99.73% of your process measurements (Optional, Numeric, >0)
In Process Tolerance, enter the tolerance range (USL - LSL) if known (Optional, Numeric)
Click OK
Three operators took measurements on 10 parts, twice on each part. You want to determine how much of the observed process variation is due to measurement system variation.
Open data worksheet by choosing ProcessMA > Tools > Data
Choose ProcessMA > Quality Tools > Gage R and R > Gage R and R
In Measurement, select AQ - Measurement
In Part, select AR - Part
In Operator, select AS - Operator
Click OK
By looking at %VarComp, the variation due to differences between parts (Part-to-Part) contributes 94.1% to the overall variation. It is much larger than that of the Total Gage R&R at 5.9%. The same observation can be made by looking at the Component of Variation Chart. This indicates that measurement system error is small. From the By Part Chart, the variation of the red line shows that there are significant differences between parts. On the other hand, from the By Operator Chart, the red line is relatively flat, which indicates that the differences between operators are small. In addition, most of the points in the Xbar Chart for Measurement by Operator are out-of-control, indicating that variation is mainly due to differences between parts.
Copyright © ProcessMA