iSixSigma

control charts in textile industry

Six Sigma – iSixSigma Forums Old Forums General control charts in textile industry

Viewing 11 posts - 1 through 11 (of 11 total)
  • Author
    Posts
  • #40109

    Alson
    Participant

    I am doing my textile study and are now learning control charting techniques. Can anyone tell me that how can it be applicable in textile industry. Can anyone in this forum provide me good example. Thanks

    0
    #123565

    Alson
    Participant

    No input uptil now, can anyone help me out there.
     

    0
    #123603

    Ray The real deal
    Participant

    Alson,
    What are you measuring, the stength of the thread? length, width,? I’m not a textile expert, if you can provide much info, I can try to assist you. Not alone…Ray

    0
    #123636

    Alson
    Participant

    Ray,
    Here is what i have

    Sample No
    # of readings

    1
    2
    3
    4
    5

    1
    220.9
    225.1
    225.1
    222.9
    222.3

    2
    124.3
    123.6
    123.6
    127.1
    126.2

    3
    173.6
    173.5
    173.8
    173.1
    171.8

    4
    200.3
    198.9
    198.4
    197.7
    199.4

    5
    137.3
    138.3
    137.3
    138.5
    138.2

    6
    173.6
    173.5
    173.8
    173.1
    171.7

    7
    200.3
    198.9
    198.4
    197.7
    199.4

    8
    137.3
    138.3
    137.3
    138.5
    138.2

    9
    204.1
    200.2
    203.8
    205.7
    205.8

    10
    204.8
    204.4
    208
    208.5
    209.8

    11
    202
    199.4
    202.3
    202.1
    204.1

    12
    198.8
    200.1
    203.4
    204.5
    206.1

    13
    104.6
    105.4
    104.1
    102.9
    103.7

    14
    217.4
    215.1
    216.3
    217.3
    215.8

    15
    212.01
    212
    210.9
    213
    207.7

    16
    130.8
    127.7
    133.7
    127.9
    128

    17
    208.9
    208.3
    206.8
    205.9
    211.2

    18
    139
    137.8
    137.3
    138
    137

    19
    143.5
    143
    141.2
    142.1
    141.2

    20
    157.2
    158.4
    156.9
    157.9
    156.4

    21
    150.7
    147.5
    148.2
    146.5
    151

    22
    151.3
    154.1
    153.7
    153.3
    154.8

    23
    149.7
    151.6
    150.7
    149.1
    147.7

    24
    154
    151.8
    153.23
    155.8
    155.6

    25
    152.8
    153.4
    154
    155.9
    157
    All my readings are gm/m-square. What i am basically doing is that a fabric is coming from my client asking for giving him the report of the fabric in gm/m-square. Now what i am doing is that i am making five round samples from the fabric and see its weight from my machine. Then i take the average of my readings and send the average reading to my client. But the problem is that my company have thousands of clients and each one is sending me fabric of different construction. For instance the construction of one fabric is 10 x 16 / 80 x 43, which is basically dimensions in warp and weft side of the fabric. But next day my other client send me the fabric with construction of 60 x 60 / 127 x 79. I have thousands of data with me like this. The readings which i got from these two fabrics are entirely different. Like with the first one i got reading like 204.8, 204.4, 208, 208.5, 209.8 and from the other one i get 143.5, 143, 141.2, 142.1, 141.2
    So my question is: How can i use control charts to see if my process is stable or not although i know i am using fabric of different construction everyday.
    Thanks for any input in this regard

    0
    #123639

    DaveS
    Participant

    Alson –
    Try Googleing on “short run SPC”. Several good books.
    An excellent one also is called “Innovative Control Charting”, but it has a lot of diatribe in it for the authorssoftware systems as I remember,
    You can just view your samples as low run parts with different noiminals. Some of the techiques are called DNOM charts, Q charts and target value charts.
    Go for it!
    Dave

    0
    #123707

    Alson to Ray
    Participant

    i am waiting for your thread
    Regards

    0
    #123796

    Alson
    Participant

    i am waiting for your thread
    Regards

    0
    #123919

    Alson to Gabriel
    Participant

    Gabriel,
    can you help me with my first thread.

    0
    #123936

    Robert Butler
    Participant

    Let’s go back to the post where you listed some of your data.  In that post you said
     “fabric is coming from my client asking for giving him the report of the fabric in gm/m-square. Now what i am doing is that i am making five round samples from the fabric and see its weight from my machine. Then i take the average of my readings and send the average reading to my client. But the problem is that my company have thousands of clients and each one is sending me fabric of different construction. For instance the construction of one fabric is 10 x 16 / 80 x 43, which is basically dimensions in warp and weft side of the fabric.”
    My understanding of your problem is as follows:
    1. You really have thousands of clients.
    2. Their only concern is the weight per unit of area of the fabric.
      If this is true then what you need to do is look at the weight requirements for the fabric and the tolerances they have around those requirements.  Take this data and plot it – make a histogram of the targets and see what it looks like – is it a reasonably normal distribution, is it bi or multimodal, are most of the clients bunched together with a few outliers, etc. 
      Next take a look at your production history and do the same thing.  Then compare to two – does your production history span their distribution, is there a mismatch, etc.  Armed with this information you can start thinking about your process relative to customer demands and, if they are in agreement then we can talk further about how you might want to sample them.
      All of the above is predicated on the assumption that #1 and #2 adequately describe your problem.  If they don’t then you will need to give us more particulars before anyone can hope to offer you some meaningful advice. Specifically, you will have to provide a reasonable summary of the customer requirements and the concerns your measurements need to address.

    0
    #123939

    eliz
    Participant

    Agree with Robert,
    You need to define what the customers ‘care about’, what is it that is making them unhappy with your process, if its weight, (per unit / area etc) then follow Roberts posting. Is there a tol. for each customers spec, or is there a ‘generalized’ spec (which then would give you your extreme outliers) do you already have SPC in place – is it aligned to the ‘care abouts’. looking at your posting i believe you will get some extremely useful advice if you include more detail – you will probably get unwanted posts from the usual protaganists – but take Kens advice from an earlier posting….
    E

    0
    #123950

    Shrestha
    Participant

    I am guessing the problem because your objective of deoing the measurement and the therefore the metric is not defined.
    So: I am guessing that your customers pay per sq m of fabrci purchased. 
    Theortically therefore a supplier company could make undie profit by supplying sq mtrs of cloth that are thinner (and therefore lighter in weight per sq mts) for a fixed price per sq meter.
    Therefore while paying by sq Mtr the buyer could be buying sub standard cloth compred to the price agreed.
    The weight per sq met will depend upon lot of physical factors which must be well known in your industry:
    like density of thread, even moisture %, closeness of weave, etc etc.
    You will therfore clearly need physical specs and tolerances for each type of fabric specified and a tolerance.
    the customer might have a tolerance of weight/sq meter for the fabric – in that case you will need to take large sample readings from your production of a particular fabric and plot a control chart and calculate the average and sigma etc for each product run. This can be control charted easily and Cp etc be worked out determine if your process is 3 sigma or 6 sigma etc. You can then work to improve the variability if necessary.
    Six samples can be sent to the cust. but there will need to be a control chart for each product and standards when it runs.
    Hope this helps.
    N
      

    0
Viewing 11 posts - 1 through 11 (of 11 total)

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