iSixSigma

Regarding DOE, Each Factor at Different Level..

Six Sigma – iSixSigma Forums Old Forums General Regarding DOE, Each Factor at Different Level..

Viewing 8 posts - 1 through 8 (of 8 total)
  • Author
    Posts
  • #50038

    Ravi Kumar B
    Participant

    Hello,
           Please tell me, how to perform esign Of Experiment for 3 factor and each factor is having at different levels. For Example, factor A is having 4 levals, factor B is having 2 levels, and factor C is having 3 levels.
         I know how to prepare for factors at same level. Please tell me, how to prepare test domain or OA table for factors each at different level.
     
    Thanks,
    Ravi Kumar B

    0
    #171866

    BC
    Participant

    Ravi
    It’s the same setup:  run all combinations.
    A

    0
    #171867

    BC
    Participant

    Ravi
    It’s the same setup:  run all combinations.

    0
    #171868

    BC
    Participant

    OK, I will try this again, and be careful about what buttons my fingers touch.
    You do the same with multi-level DOEs as you do with 2- or 3-level DOEs.  You just run every combination.  For example:
    A         B           C
    1           1            1
    1            1            2
    1            1            3
    1            2            1
    1            2             2
    1            2             3
    2            1              1
    2             1            2
    2             1             3
    and so on until until you get the 24th run:
    4             2             3
    There are two places where multi-level DOEs can be difficult:  analysis and fractionating.  Analysis should be automatic with Minitab but if you try to do analysis by hand, it can be a mess (analysis of anything but 2-level DOEs is difficult to do by hand).
    Fractionating is a real mess too, and there are different techniques you can try.  But with 4, 2, and 3 levels (24 runs total), I’d stick with a factorial.
    BTW, to set this up in Minitab, use DOE > Factorial > General Full Factorial.  You can specify any number of levels for any number of factors.

    0
    #171874

    Venugopal G
    Member

    Ravi,
    I have created the design using DOE>Factorial Design>Generate Full Factorial Design. I am placing here the complete design ( I hope u know how to create and u might have created by this time) Hope this is useful for you. If possible can u give me the results of the experiment for practising this kind of mixed DOE. I have practised different DOEs like factorial/taguchi/shainin, but never comed across mixed one and designs at different levels(like this).
    Kindly post the results if possible only.

    StdOrder
    RunOrder
    PtType
    Blocks
    A
    B
    C

    14
    1
    1
    1
    3
    1
    2

    18
    2
    1
    1
    3
    2
    3

    3
    3
    1
    1
    1
    1
    3

    5
    4
    1
    1
    1
    2
    2

    11
    5
    1
    1
    2
    2
    2

    8
    6
    1
    1
    2
    1
    2

    24
    7
    1
    1
    4
    2
    3

    17
    8
    1
    1
    3
    2
    2

    23
    9
    1
    1
    4
    2
    2

    16
    10
    1
    1
    3
    2
    1

    1
    11
    1
    1
    1
    1
    1

    4
    12
    1
    1
    1
    2
    1

    9
    13
    1
    1
    2
    1
    3

    13
    14
    1
    1
    3
    1
    1

    21
    15
    1
    1
    4
    1
    3

    15
    16
    1
    1
    3
    1
    3

    10
    17
    1
    1
    2
    2
    1

    2
    18
    1
    1
    1
    1
    2

    6
    19
    1
    1
    1
    2
    3

    7
    20
    1
    1
    2
    1
    1

    22
    21
    1
    1
    4
    2
    1

    20
    22
    1
    1
    4
    1
    2

    12
    23
    1
    1
    2
    2
    3

    19
    24
    1
    1
    4
    1
    1
    G. Venugopal.

    0
    #171880

    Eric Maass
    Participant

    Hi G,
    It would be a bit easier to follow if you had these in runs in standard order rather than run order (where the order is randomized), but this looks like a full factorial for 2x3x4 = 24 runs.
    I am pretty sure that I could quickly show you how to do this in 12 runs rather than 24 runs…but you would have to analyze it as a custom design.  I think I know how to do this in the Minitab environment, but it’s sort of tricky, so I’ll only tell you if you are really interested.
    (Okay, scroll down if you are interested…at the risk of perhaps getting confused)
     (Basically, it involves designing a fake version of the experiment – with 3 factors at two levels, half fraction = 2^(3-1) = 4 combinations, 3 replicates = 4×3 = 12 runs…then the three replicates become factor C, one of the factors becomes factor B with 2 levels, and Factor A becomes the combination of the other two factors, where -1 -1 becomes A=1, -1 +1 becomes A=2, +1 -1 becomes A=3, and +1 +1 becomes A = 4.  Then, you bring it back into Minitab for Analysis by choosing customer design)
    Best regards,Eric Maass

    0
    #171907

    Manee
    Participant

    L16 OA is better created by rdExpert

     


    1
    1
    1
    1

    2
    1
    1
    2

    3
    1
    2
    3

    4
    1
    2
    1

    5
    2
    1
    1

    6
    2
    1
    2

    7
    2
    2
    3

    8
    2
    2
    1

    9
    3
    2
    1

    10
    3
    2
    2

    11
    3
    1
    3

    12
    3
    1
    1

    13
    4
    2
    1

    14
    4
    2
    2

    15
    4
    1
    3

    16
    4
    1
    1
    Manee
     

    0
    #172256

    Eugene Lanning
    Participant

    OK on the DOE under discussion.
    In general is it correct to say that you picked the half fraction because 2^3/2 = 4, the highest level in the DOE?
    Is it correct to say that the number of replicates was picked to be the same as the odd level factor.  Is that a general strategy or just a concidence of the DOE considered?
    Now it is not general that one factor was chosen to be the values MiniTab edited, as that would not be correct for a DOE that was a 3x3x5 design.  Perhaps the method only works when one of the factors has two levels.
    For a 3x3x5 design, is a person just out or is there a broader technique that would work?

    0
Viewing 8 posts - 1 through 8 (of 8 total)

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