iSixSigma

Control limits

Six Sigma – iSixSigma › Forums › Old Forums › General › Control limits

Viewing 5 posts - 1 through 5 (of 5 total)
  • Author
    Posts
  • March 11, 2008 at 2:26 pm #49559

    Whitehurst
    ★ 10 Years ★
    Participant

    Hello – I have a query about control limits.  I understood that they were set at +/- 3 x Std Dev from the Mean value of a data set.  My question is, does this only apply when there are no special causes?  I realise that you shouldn’t use the control  limits for an unstable process, but I was wondering if there is a different calcuation to produce the control limits for unstable processes.  It is especially noticeable with the X bar chart output from GRR assessments.
    Thanks for your help
    Joe

    0
    March 11, 2008 at 3:49 pm #169532

    aiman farid
    ★ 10 Years ★
    Participant

    Hi Joe,
    Control charts are used for stable processes. If you have a special cause point, you have to investigate for the reasons.  If you find out causes for that point, erase it completly from the data set and recalculate the limits. By this way, you are not allowing special causes variation to affect the value of control limits.  But, if you do not find out reasons, calculate the limits including that point considering it as a part of general causes variation (0.3%)
    I hope this is helpful
    Aiman

    0
    March 11, 2008 at 4:04 pm #169534

    BC
    ★ 20 Years ★
    Participant

    Joe,
    The xbar chart in Gage R&R is SUPPOSED to be “out of control”.  The control limits in the gage R&R xbar chart represent measurement variation.  The xbar values represent part-to-part variation.  You want the measurement variation to be much smaller than part to part variation.  Hence, a large proportion of points should be outside the control limts.
    The Gage R&R xbar chart is a special use of this chart.  It isn’t meant to show out of control or special causes or anything like that; only the relationship between measurement variation and part to part variation.  Even so, the measurement system must be stable before calculating control limits from it.
    BC

    0
    March 11, 2008 at 4:16 pm #169536

    Whitehurst
    ★ 10 Years ★
    Participant

    Thanks BC – I knew it was supposed to be out of control but I didnt know the control limits were based on the measurement variation – All becomes clear!
    cheers

    0
    March 26, 2008 at 5:05 am #170061

    hitesh
    ★ 10 Years ★
    Participant

    hi Joe,
    very first purpose of having the Control Limits is to determine whether process is Stable or not.
    Presence of Special Causes in the process which is determined after bringing Control Limits in frame of control Charts indicates Instability of the process & need for analysis to identify causes of these special causes of variation & take an improvement actions thereon.

    0
  • Author
    Posts
Viewing 5 posts - 1 through 5 (of 5 total)

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