Conduct a Gage R&R Study to evaluate your measurement system.
When you measure, three factors come into play: Part variation, appraiser variation, and equipment variation.
You want most of the variation to be between the parts, and less than 10% of the variation to be caused by the appraisers and equipment.
For more info, see: www.qimacros.com/gage-r-and-r-...
Download the 30-day trial at: www.qimacros.com/trial/30-day/
This video shows you how to perform a Gage R&R study, using the QI Macros for Excel.
One of the causes of error in a given system can be your measurement system, and to deal with measurements (what's called measurement systems analysis), you'll want to start to learn how to use the Gage R&R Chart.
Now, the idea behind Gage R&R is that you have gages, and you have people using those gages called appraisers.
The idea is that you want the variation among your parts to be the biggest cause of variation. Note how appraiser one puts the gage on the device and appraiser two puts it on a different way.
So, you want to be able to conduct a trial to find out how good you are at measuring things. To do that, you need two appraisers, appraiser one and appraiser two, and at least two trials and preferably at least ten parts.
What I've done is brought up an example of that. Here you can see an example of the Gage R&R Chart filled in-this is from the automotive industry action group’s third edition manual using their data.
You can see we have these different parts and three different trials and three different appraisers.
You can also, if you have a reference value for each one of these, you could put that in here, but you can see there's a lot of variation in here (the different kinds of parts).
Now, what you want to do then is scroll down and look at Gage R&R. There's several different pieces in here. There's the equipment variation, so how much does the gauge contribute to the overall variation in the system; there's appraiser variation, this is how much does each person contribute the variation.
Then when you come up with the calculation based on those two things, it gives you how much or what percent of the total variation is due to the equipment and the appraiser.
QI macros will calculate for you that if the percentile is less than 10%, then you're probably pretty good; if it's between 10 and 30%, you'll get a message that says this gage system may be acceptable.
It'll also try and figure out whether you need to retrain your operators or whether you need to maintain or redesign your gage, and that will start to tell you whether or not your measurement system is good enough.
If you scroll down even further, you'll see a plot of each part by appraiser. So, you get some idea whether somebody, a couple of them here looked like they're pretty close and one person looks fairly low. Maybe we just have to train that one appraiser to bring some consistency out among all of our appraisers.
This will also show the range in variation. You see that this particular person here had a pretty consistent track record, but this person was measuring all over the board.
Maybe this is something we could track is find out why this appraiser is doing things a little bit differently.
That’s how you do a variable analysis on your gages, and how you measure things using the Gage R&R Chart.
In other sections, we're going to talk about how to use something called linearity and the attribute gage worksheet to do attribute studies. So, look for those as well.