248 Quality Assurance
Graphical Output: Six Graphs in All
The six items that we just mentioned might be separated into two categories
(see Figure 16.14). They are the health side and the troubleshooting side.
Destructive Test
When we are conducting a destructive testing, we are interested in a nested
approach of gauge R&R MSA. The output for a nested gauge R&R is similar
to the normal output and is shown in Figure 16.15.
Chart Output Interpretation
Now that we have identied the process of entering data and selection of the
appropriate analysis, and we have shown the typical outputs of the Minitab
®
software in reference to the MSA, let us attempt to interpret the results.
• Tall bar charts: Distinguishes the components of variation in per-
centages (%)
• Repeatability, reproducibility, and parts (want low-gauge R&R,
high part-to-part variation)
• R chart: Helps identify unusual measurements
• Repeatability/resolution (no outliers permitted)
• Xbar chart: Shows sampled process output variety
• Reproducibility/sensitivity (want similar patterns for each
operator)
The leading graphical indicators with which the evaluator should be con-
cerned the most are
• Bar charts for components (see Figure 16.16)
• Look at the R chart rst, then Xbar chart (see Figure 16.17)
• More R chart indicators (see Figure 16.18)
• ANOVA tabular output % (see Figure 16.19)
Explanation of Some of the % Numbers
% StudyVariation =
R&R
*100
If the gauge does not pass %Study, it cannot perform the job of process
control (it will not be able to adequately distinguish one part from another).
Anything less than 10% is very good. Anything less than 30% is acceptable.