Frequency and Sample Size Requirements for an MSA Studies

[ppmartin]

Hi guys!
I got a doubt with this matter, I don't know if there is a frecuency and sample size for the MSA Studies.
For example: If a got a Control Plan with 2 variables like this:

Variable one - Caliper
Variable two - CMM

But I have two Calipers and 3 CMMs.

So, Do I need to make 5 GR&R Studies?
Like this:
1. Var 1 - Caliper 1
2. Var 1 - Caliper 2
3. Var 2 - CMM 1
4. Var 2 - CMM 2
5. Var 2 - CMM 3

Thank you in advance!

 

[howste]

Re: MSA Frequency and size doubt

MSA stands for Measurement System Analysis, so you are evaluating the "system" not each piece individually. If the calipers are the same type and CMMs are the same type, you may only need to do one study for each (two total). If they are different enough you should do each separately.

 

[ppmartin]

Re: MSA Frequency and size doubt

Thank you so much howste!
Now it's pretty clear to me

 

[amiuda]

Hi @howste !
How to Deal with the "Families" of gauges For instance for calipers, please.
If you do 1 caliper, and it is NOK what do you do to all calipers from that Family?

Thanks

 

[Ninja]

Howdy amiuda,

Note in Howste's post:

so you are evaluating the "system" not each piece individually.

You aren't testing the calipers, you are testing the USE of those calipers in your testing system. In fact, you are analyzing whether your test regimen is adequate. If a set of calipers and how you use them to measure results in a failed GRR...the problem is likely in your measuring system, not in the gage itself.
As such, your calipers may all be fine...but not adequate in your current technique to give reliable results. And as such, one fails, all fail.

Review the rest of the test regimen...how you USE the calipers...rather than trying to separate good/bad calipers.

HTH