NDC (Number of Distinct Categories) - Long Gage R&R Study in MSA 3rd Edition
What's really sad is that the quality profession continues to promote practices that are invalid. We have ceased thinking. The games that are played to get an 'approval' divert effort from true quality improvement.
I have begun to appreciate how Deming felt.
E
elgracias8581
Please help why my NDC is 1 and my GR & R not acceptable, using Gage R&R Study (Nested) for Destructive test. See attached Data
Looks like major typo in your data. One point at 0.44???
Even correcting the typo to 6.44, doesn't help. there is a serious repeatability issue, and it is not caused by inadequate resolution.
Can you provide additional information on what is actually being tested? With a destructive test, you might be seeing the inherent variability of the material being tested.
Can you provide additional information on what is actually being tested? With a destructive test, you might be seeing the inherent variability of the material being tested.
E
elgracias8581
We were pull test the 24awg wire using push pull gage
This type of test has a variety of repeatability problems. Trying to utilize Gage R&R on physical testing is a nightmare. I will give you some tips.
At you measuring ultimate, catastrophic failure? If so, put your pen down and walk away. It will never repeat. You may have to use "surrogate" measures to verify the tester. If it chart records, use yield instead of ultimate failure. It is far more repeatable in materials.
One approach I used was to use trilene fishing line at various test strengths as a surrogate material. It provides - by design - specific variation that your tester should be able to prove it can tell the difference between. You might not have those kinds of samples with your own material. I was working in a small range of pull strengths, so I could easily find line in my range. But, if you can use it in your tester (and you will need to learn the "trilene knot" to keep it from slipping), then input historical variation for your PV, it may work fine.
If your wire is stranded, try using different number of strands to determine of the tester can see the difference.
I hurts my brain when people want to rubber stamp Gage R&R on physical testing, but that rubber stamping is chronic, especially in automotive.
Last edited by a moderator:
Depending on the make and model of the gage, this may be a portion of the repeatability issue. However, a large portion will likely be due to the variation in tensile strength of the material, variation in the cross-sectional area of the wire and surface flaws, such as nicks.
After reading Bob's post, I recalled a white paper from Instron that addresses these issues nicely and offers an alternate approach to the R&R.
Similar threads
