MSA (Measurement System Analysis) - Results and Actions Taken

[Black arrow]

Hi again,

I have another question concerning taken action after a MSA study. Following result came out:

Scenario:% Tolerance: 41%
% Study Var: 98%
ndc:1


A: How will you act if the measuring equipment is used only for product control situations?

B: How will you act if the measuring equipment is used for product control situations and SPC?

C: How will you act if the measuring equipment is used for product control situations and process control situations ( not like SPC with control limits, more guidance with lower level of statistic approach)

 

[Bev D]

*I* probably wouldn't use this system for either product acceptance or process control. Without seeing your data I can only guess (The AIAG/Mini-tab method are not the most intuitive for understanding your measurement system) but my first response is that you should investigate why your system is so poor. It may just be how you sampled the process or it may be real.

If you post your data and describe the characteristic and measuring system we can provide much more insight.

 

[Black arrow]

Hi again,

I have added my report here. The resolution on the instrument is 0.001 mm. I think that could be one problem with poor Ndc and % Study Var. I have also added a
% Process column which are based on earlier made Cap.study.

It is about cylindricity max 0.020 mm. Note: In "options" in Minitab i have entered LSL to 0 and USL to 0.020 and what i have recently been told that this is wrong when it concerns one sided tolerance. I should only enter USL= 0.020. But i don't think that will improve the result.

 

[Black arrow]

Hi again,

Do i have to send the report again? Seems to be some text problems in it.

 

[Bev D]

I'm sorry, but summary reports for GR&Rs are all but useless in diagnosing problems.

if you can post your raw data we will have a much better chance of understading and diagnosing the probelm....

 

[Black arrow]

You mean all measured values?

 

[Black arrow]

Hello,

Here are the measured values.

 

[Miner]

The major source of variation is in repeatability. Your resolution appears to be adequate. Since you are actually measuring the within product variation, this may be an issue of inadequate coverage.

How many points are you using to assess cylindricity? Have you tried increasing the number of points and repeating the study? Try increasing the number of points per circle and the number of circles measured.

 

[Black arrow]

Yes, the major source is repetability. I do understand the definition of repeatibility and reproduceability but have difficult to take actions from the result from them. Needs some help concerning that, if you have higher repetibility than reproducibility what causes it and what actions to be taken? and vice versa.

Ok, the resolution is ok. What do you mean with "you're actually measuring the within product variation, this may be an issue of inadequate coverage"?

Thanks, i will check how many point we're using to assess cylindricity and come back.

 

[Miner]

Within product variation can show up in either repeatibility or reproducibility and inflate the results of the R&R study. Since you are actually measuring the within product (form) variation, the better you measure it the less it will inflate your results.

If you measure cylindricity with six points (3 points each on 2 circles) your results would be likely to vary a lot and would inflate your results. If on the other hand you were to (warning: extreme example ahead) measure it with 900 points (30 points each on 30 circles), your results would very likely be much more consistent and would barely inflate your results.

I am not recommending that you follow the extreme example, but I do recommend that you experiment with the number of measurements per circle and the number of circles measured to see the impact on your MSA.