iSixSigma

DOE with Discreet data

Six Sigma – iSixSigma Forums Old Forums General DOE with Discreet data

Viewing 5 posts - 1 through 5 (of 5 total)
  • Author
    Posts
  • #53072

    kd
    Participant

    Hi All,
    I’m facing problem in generating Transfer Function with one discreet & one continuous variable (2 factors). Could any one suggest me with the expertise & experience in DOE with discreet variables?
    Problem-
    Factors
    a, Diameter of Holes
    b, Number of holes
    2K full factorial runs with above parameters. My main objective is to reduce the number of holes but do not want to sacrifice the performance. The holes were located circumferentially and equally spaced.
    Question-
    Whether developed Transfer function with above factors (equally spaced) will be validate for hole located circumferentially and not equally spaced?
    Thanks for helping
    KD
     

    0
    #187792

    Remi
    Participant

    Hai KD,
    the correct answer is: it depends.
    If y = air usage, it does not matter how the holes are placed. The air usage is the Area of all the holes together.
    If Y = cooling strength on a small area, then situation with holes all near the centre gives a different result than holes on the border.
    So the question is: what is your Y=CTQ and how is it influenced by the X’s. The function itself is just a mathematical model for this relation.
    Remi

    0
    #187793

    kd
    Participant

    Hi Remi,
    Thanks for you reply, my CTQ is flow in the holes. In my case I cannot bring the holes near to center so I fixed one diameter.
    Thanks
    Deevakar

    0
    #187808

    Robert Butler
    Participant

      It sounds like you actually have three variables
      1. Hole Diameter
      2. Radial position of hole(s)
      3. Location of hole(s) along radius
        According to your initial post you have fixed #3 at an arbitrary value and therefore are assuming hole location relative to center doesn’t matter.  I understand the reasoning behind the choice but in addition to assuming location from center doesn’t matter you are also assuming a pattern of various holes at different radii from the center is equal to a pattern of holes at a fixed radius. I don’t understand how this could be the case.
      As for #2 – if you use radial location then you could model the effects of symmetric and non-symmetric hole distribution (for example 4 holes evenly spaced 90 degrees apart vs. 4 holes 30 degrees either side of the 0-180 axis)
      Even with only. #1 and #2 there are a huge number of possible combinations.  If you have some kind of prior knowledge about hole sizes and hole patterns and their performance you might be able to use this infomation to guide you with respect to identifying combinations that would reduce the choices for experimental combinations to something more manageable.

    0
    #187909

    kd
    Participant

    Hello Robert,
    I think you got it right! And thanks for your comments.
    #1(hole Diameter), #2 (no. of holes) is basically the major affecting factor and that’s the reason I screened out other to save time (analysis time) and right now (as you said) I’m struggling to develop the transfer function so that it will be common for both symmetric & non-symmetric #3 (radial location) but it seems I cannot develop common transfer function without including #3
    Which one you form below run you would prefer
    1, 1-level full factorial
    2, Response surface method
    Thanks for helping
    Deevakar
     

    0
Viewing 5 posts - 1 through 5 (of 5 total)

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