NDC (Number of Distinct Categories) - Long Gage R&R Study in MSA 3rd Edition

Would you care to give me an example, our push pull gage is up to 30kgs.

Your product test range is about 17 lbs, if it was reported in kg. I would get 10# 12#, 15#, 17# and 20# test monofilament. I would then run your gage r&r on each # test line as your "sample". Because of the nature of this line, you will have to secure it to keep it from slipping, and that depends on what kind of jaws you have on the machine. This will give you very good repeatability and reproducibility exercise of the machine. I would repeat the test 9 times for each material, using the first 3 as appraiser 1, second 3 as appraiser 2 and the last 3 as appraiser 3. That way appraiser to appraiser comparison will see if there is any deterioration of R&R over time during the test. If you have historical pull test data for that material, use that as your PV instead of the "sample PV". It is an option in Minitab - the "preferred" option, actually.

What Bob is recommending is very similar to that recommended by Instron. Instron would recommend that you use tension springs in the same range as your product testing, but the concept is the same. Specifically, to perform the R&R study on a surrogate with an insignificant level of variation in properties.

I've followed the concept but the result is still fail the value of NDC is still 1 and the %GR &R is quite high. What i did is i secure the wire on vise grip at 100mm length and pull the wire with the length of 50mm.

Please help to adjust/correct my data to have acceptable NDC and %GR&R. See attached data
Need to submit this report on May 16, 2018.

The x-bar chart by operators tells you based on the variation you presented, the gage can't tell the parts apart. You can't "fix the data to make it pass". The ndc of 1 is very accurate.

What did you do differently from your first study? Did you use the trilene monofilament or the springs?

On my first study i'll let the crimped wire to break/loosed, (gather the date when its break/loosed)

The second one i'll control the displacement of the crimp wire, initial setting is 100mm length then pull about 50mm and record finding.
Note: the crimp wire did not break or loose.

Returning to initial question - can anybody explain/clarify in details based on practical examples - what does NDC mean? From this and other threads the “physical meaning” of this coefficient is not clear. Unfortunately, it is very difficult to understand the explanation defined in MSA_Reference_Manual_4th_Edition (pages 45-49)!
I know that NDC value represents the number of groups measurement tool can distinguish from the data itself. The higher this number, the better chance the tool has in discerning one part from another (Understanding "Number of Distinct Categories" in Your Gage R&R Output). But what does "groups" mean? How can I calculate/define the group quantity? Any examples, please.

At a high level, any dimension within a single group is indistinguishable from another due to measurement variation. You can distinguish a measurement in one group from a measurement in another group. I wouldn't worry too much about understanding the details about ndc as it is 100% redundant with the %Study Variation metric and adds no additional value.

Thanks, Miner! But the terms "high level" and "group" are not clearly defined also. May be you can provide the example?