1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.
Dismiss Notice
You must be a registered member in order to post messages and view/download attached files in this forum.
Click here to register.

Cpk & Ppk Target

Discussion in 'Capability - Process, Machine, Gage …' started by Pongsakorn, Feb 19, 2016.

  1. Pongsakorn

    Pongsakorn Active Member

    Joined:
    Aug 10, 2015
    Messages:
    72
    Likes Received:
    16
    Trophy Points:
    7
    What should be the proper Cpk and Ppk Target.
    QS9000 and VDA 6.3 requirement is totally different.

    QS9000 recommend Ppk >1.33 and Cpk >1.67 whereas
    VDA 6.3 require Ppk >/=1.67 and Cpk >/=1.33.

    In my opinion, I tend to agree with QS9000 since the Cpk usually
    greater than Ppk. I do not understand how VDA 6.3 determined such target.

    This is from QS9000
    upload_2016-2-19_18-30-31.png

    This is from VDA 6.3
    upload_2016-2-19_18-47-32.png
     

    Attached File(s): 1. Scan for viruses before using. 2. Report any 'bad' files by reporting this post. 3. Use at your own Risk.:

  2. ncwalker

    ncwalker Well-Known Member

    Joined:
    Sep 21, 2015
    Messages:
    261
    Likes Received:
    168
    Trophy Points:
    42
    Location:
    North Carolina
    I think VDA 6.3 has "short term" and "long term" swapped in their text. I agree with QS.
     
    Pongsakorn likes this.
  3. Bev D

    Bev D Moderator Staff Member

    Joined:
    Jul 30, 2015
    Messages:
    605
    Likes Received:
    663
    Trophy Points:
    92
    Location:
    Maine
    The VDA doesn't have short term and long term swapped. they are using different formulas and data collection / sampling structures than QS is.
    both are correct given their definitions. Although some would argue (and theoretically correctly) that QS is wrong because you cannot have a chronically unstable yet predictable process...

    And of course the targets are atmospherically derived based on an erroneous assumption that the Normal distribution exists for everything, there actually IS data beyond +/-3 SD and that one can actually use a continuous data distribution to predict a categorical result AND that the specification limits actually are predictive of failure.

    ...so both are wrong since all capability indices are useless abominations on the face of the earth. but hey, that's just my opinion based on 35 years of continually improving quality without EVER calculating one of the darn things.
     
    Bob Doering, Pongsakorn and Candi1024 like this.
  4. Bev D

    Bev D Moderator Staff Member

    Joined:
    Jul 30, 2015
    Messages:
    605
    Likes Received:
    663
    Trophy Points:
    92
    Location:
    Maine
    I have posted a brief article on these definitions in the resources section...
     
  5. Bob Doering

    Bob Doering Member

    Joined:
    Jul 30, 2015
    Messages:
    37
    Likes Received:
    27
    Trophy Points:
    17
    Also, QS9000 is obsolete, but when utilizing these indices you MUST understand their statistical limitations. These limitations are nicely laid out in AIAG PPAP 4th edition section 2.2.11.5 Note: The above mentioned acceptance criteria (2.2.11.3) assume normality and a two-sided specification (target in the center). So, first you need to find out if your distribution and specification apply before even bothering. Then, recognize that even if they do, their value is marginal as Bev D. indicated.