Hello, Everyone!

First of all I'd like to thank you for this forum! It is really great - I have found here lots of useful information on ISO/TS (we're situated in Ukraine, and there are no companies certified to ISO/TS, so we have nobody to ask and receive an answer. Thank you once again!).

Now my collegues from Measurement have faced a problem and asked to seek for assistance here.
We have some experience in working according to MSA and in analysis of measurment system. But now we have faced an issue that we won't understand.
For example, there is a system that involves: 3 appraisers, electronic calliper (grating period is 0.01 mm), 10 parts and 3 trials to measure the same dimension of the part.
Variation from part to part makes up 0.06 mm maximum. And as a result GRP% makes up almost 40% and it means that the system is not good!
In the 3d edition of MSA there is only one example of gage repeatability and reproducibility. It involves the same quantity of appraisers, parts and trials. The variation there makes up 3.511 and GRP% is 26.68%. I.e. it proves to be the regularity as follows: the worse quality the parts are of, the better the system is!? And we don't receive a good system as a result, because all our parts are high-quality. Please, help to understand this!

Many thanks in advance!!

How do you use the gage (for inspection or process control/SPC)?

Use P/T Ratio if the gage is used to inspect to tolerance. Part variation has no effect on this metric.

Use %GRR if the gage is used for process control (SPC). Note that this metric is only valid if the variation of the ten parts has the same standard deviation as the standard deviation of the process.

A highly capable process requires a better gage for process control than a less capable process. If measurement variation is a large percentage of the observed variation, it will inflate the control limits used for SPC. This will give the appearance that the process hugs the central line, or exhibits runs on either side of the central line.

Hi Maria,
in my humble opinion I think you are right. The R&R analisys is process dispersion angaist measurement system variability. The greater is the first the better the ratio.

Of course, the test is as reliable as you are able to represent the process variability in the 10 pcs. test sample.

Anyway, if it is the first time you check your measurement system vs. this process, I would suggest you to run before Cg (gauge capability) and Cgk (gauge Cpk), in order to check the measurement variability also angaist specs. tolerance range.

Hope this helps

Thank you very much for your replies! I have transferred them to our Measurement specialists, I'm sure they will take advantage of them.

Thank you.

Maria