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

V

Vatsa

I am convinced about your information. I want one more clarification. Once I tried GRR and I got NDC equal to 5 and GRR 6.41. The next time I tried with the same instrument. I got NDC less than 3 and GRR 6.41. How do you conclude or what corrective action do we take.
YKT:

As far as I can understand, this number is related with the "practical resolution" or the "practical discrimination" of a measurement instrument. The concept is not very difficult to grasp, but trying to explain WHY is calculated in the way that it is calculated, is another thing...

Here is my understanding with regard to NDC index:

Measurement Instruments have physical limitations to "discriminate" between different parts (i.e. give different values of a dimension for parts that are actually different). An example: The discriminative power of the scale that my wife has in the kitchen is 5 gr. By using this device it will be impossible for me to "discriminate" between packs of flour wheighing 1.010 gr, 1.007 or 1011 gr (eyeballing & rounding is not allowed!). All these three packs will be considered as having the same weight when actually it is not true. The three packs are in "the same category". In this way we start to talk about "categories"

The NDC tells you how many categories your measurement system (MS) is able to diferenciate considering your process variation.

If your NDC = 1 then your MS is useless.
If the NDC = 2 then your MS can only distinguish between two categories ("small" and "big", for example). (useless too)
If NDC = 3 then you have small, medium and big, and so on...

Automotive industries recomend a value equal or greather than five, if I remember well... The greatest the NDC the better.

Of course the term "category" is not physical or tangible. It is an "abstract category".

I hope this helps a little...

Regards.
 

bobdoering

Stop X-bar/R Madness!!
Trusted Information Resource
I am convinced about your information. I want one more clarification. Once I tried GRR and I got NDC equal to 5 and GRR 6.41. The next time I tried with the same instrument. I got NDC less than 3 and GRR 6.41. How do you conclude or what corrective action do we take.

The answer is in observing the measurement process. Did anything change? Did you use the same parts? Were the same locations used on the parts? If not, the part variation may change from GRR to GRR. Were different people used? Is the measurement process affected by the person doing it? Also look into the type of variation that is participating the most. Is it within- (equipment) or between- (appraiser) the biggest participant in the error? If you have the AIAG MSA 4th edition book, look at pg 52-56. There are excellent lists of possible causes. You might even need to consider stability - although that is rare in short-term studies. Look at these sources of variation, and determine if any can be better controlled.

Do not look at the data itself to tell you the answer. It can't.
 

Miner

Forum Moderator
Leader
Admin
I am convinced about your information. I want one more clarification. Once I tried GRR and I got NDC equal to 5 and GRR 6.41. The next time I tried with the same instrument. I got NDC less than 3 and GRR 6.41. How do you conclude or what corrective action do we take.
Was the GRR %SV or %Tol? ndc and %SV are dependent on the sample variation. %Tol is not. If your sample variation changed from study to study, ndc and %SV will change, while %Tol would remain fairly constant.
 

EBO000001

Starting to get Involved
Hello to all,

For a Caliper with 0.01 mm resolution we can obtain a NDC equal or above 5 NDC? I use a RR Study, but the NDC is always 1.

Thanks

Edgar
 

bobdoering

Stop X-bar/R Madness!!
Trusted Information Resource
Without actually seeing your data, the key factor is the variation of the samples presented to the study. If the readable resolution and error cannot tell the parts apart, you will not get high ndc.
 

EBO000001

Starting to get Involved
Thanks.

We make the study R&R with this data:

10 parts (same batch) + 3 reading each part + 3 appraisers

The results:

%GRR=9,49

%PV= 6,99

NDC= 1

Edgar
 

Miner

Forum Moderator
Leader
Admin
This is still insufficient information. Please attach your data and analysis.
 

Bev D

Heretical Statistician
Leader
Super Moderator
your tolerance is .6 (+/- .3)
The standard deviation of your parts is >.02
you have almost no variation in your parts. this is why you have an NDC of 1 but a decent %GRR

try plotting your data against the spec limits - even a simple histogram will show you the issue.

you need better variation in your parts OR a better analysis method...
 
Top Bottom