iSixSigma

Gage RR average chart control limits

Six Sigma – iSixSigma › Forums › Old Forums › General › Gage RR average chart control limits

Viewing 9 posts - 1 through 9 (of 9 total)
  • Author
    Posts
  • October 22, 2007 at 7:03 pm #48475

    Ropp
    ★ 10 Years ★
    Participant

    Hey,
    What what is the correct calculation for UCL and LCL on an average chart for a Gage R&R. I’m fairly new to doing these studies and need a little help. I thought that the standard SD would work but it seems that the control limits on these charts are much tighter, and the goal is to have points outside of these limits. I gather it’s a test to see normal random variation from part to part.

    0
    October 22, 2007 at 7:16 pm #163498

    Erik L
    ★ 10 Years ★
    Participant

    Dave,
     
    The control limits in the average chart are set by the standard formula:
     
    CL=Xbarbar +/- (A2*Rbar)
     
    Remember that the resulting space between the LCL and UCL is a reflection of the variation that we expect from the measurement process and then we add part-to-part variation on top of this ideal estimate of measurement variation.
     
    Regards,
    Erik

    0
    October 22, 2007 at 8:03 pm #163499

    Ropp
    ★ 10 Years ★
    Participant

    Thanks Erik,
    I know this is a silly question and I probably should know, but what is A2?

    0
    October 22, 2007 at 8:35 pm #163501

    Erik L
    ★ 10 Years ★
    Participant

    Dave,
     
    Not silly at all.  A2 is a control chart constant that is impacted by the rational subgroup size that you use in the creation of the control chart.  It is capturing numerous terms (e.g. 3, d2, ãn) from the following:
     
    ƒÊ +/- 3ƒÐ
     
     
    Xbar +/- 3 Ŝxbar
                        ãn    
     
    Xbar +/- 3 Rbar  1 
                       d2    ãn
     
    Xbar +/- A2* Rbar
     
    Hope that helps with where it came fromc
     
    Regards,
    Erik

    0
    October 22, 2007 at 8:40 pm #163502

    Erik L
    ★ 10 Years ★
    Participant

    Hmmm,
    Not good this is suppose to be the sqrt of n=ãn 
    ƒÊ=mu
    ƒÐ=sigma 
    I would have thought it would have converted from Word better….

    0
    October 22, 2007 at 9:27 pm #163505

    Ropp
    ★ 10 Years ★
    Participant

    Ahh its one of those tables you gotta look up in the MSA book ;0) just needed the direction thanks a bunch, I assume if I use 10 Parts, 3 appraisers, and 3 iterations the subgroupsize is 3, therefore the A2 value I need to use is 1.023.

    0
    October 22, 2007 at 9:32 pm #163506

    Erik L
    ★ 10 Years ★
    Participant

    You got it…I guess this is the proverbial telling of how the watch was built, vs telling you about the watch!!! 

    0
    October 22, 2007 at 9:37 pm #163507

    Ropp
    ★ 10 Years ★
    Participant

    Yes, I probably should crack that book open a little more than I do currently, thanks Erik

    0
    October 22, 2007 at 11:30 pm #163516

    Ropp
    ★ 10 Years ★
    Participant

    Check me out,
    http://killgorack.hightechhangout.com/images/gr&r-_screenie.jpg

    0
  • Author
    Posts
Viewing 9 posts - 1 through 9 (of 9 total)

The forum ‘General’ is closed to new topics and replies.