iSixSigma

Quality Testing Frequency

Six Sigma – iSixSigma Forums Old Forums General Quality Testing Frequency

Viewing 6 posts - 1 through 6 (of 6 total)
  • Author
    Posts
  • #48669

    David Hart
    Participant

    Anyone know a simple formula to calculate a quality testing frequency?  Here’s the situation–We manufacture approximately 30 reels of paper each day.  A sample is taken from each reel produced.  Is there a data driven approach to determing the frequency for which I would run various quality tests?  Currently, we are testing every 4th reel and then of course increasing the frequency based on our results.  A Green Belt wants to determine the optimum testing cycle versus using the “every 4th reel” standard has no scientific basis.

    0
    #165327

    ibike365
    Participant

    To determine the proper frequncy of sampling you need to consider a couple of things:
    1. Are there any stratifications or cyclical effects in the product that are known?
    2. What amount of difference are you looking for in your results (what is important to the customer)
    3. What is the standard deviation of the measured variable.
    If you can answer these things then Minitab (I suppose Jump can do this but I dont have expereince with Jump) can help you determine a statistically appropriate number of samples to gather. An improtant point is the amount of difference. This is practical difference, not statistical difference. For instance if you are measuring for thickness and the customer cares about a difference of 2.5 millimeters in thickness, then 2.5 is the practical difference you are wanting to see in your samples. That knowledge, combined with the standard deviation, will give you an appropriate number of samples needed in order to see that difference. The other question regarding stratification or cycles is needed to determine the sampling plan. if there are changes in the process throughout the production cycle, say from a cleaning process where the process equipment is clean at the start of a cycle but gets dirty as the process operates, then you would want to intentionally sample at various points of the cycle between cleanings in order to show the full range of variation. Hope that helps.
     

    0
    #165346

    AlexanderSO
    Participant

    I think you can use the MIL STD to define the sampling of your process, in the MIL STD you can choose from a wide range of sampling options based on the acceptance that you requiered to guarantee you quality level.

    0
    #165451

    Michael Schlueter
    Participant

    Hi David,
    There is an approach driven by economic considerations to establish a justified balance between frequency, desired quality level, cost of measurement, cost of adjustment, frequency of adjustement and the like.
    It usually leads to different frequencies for different quality measurements and to significant economic improvement. It assumes, you can readjust your process (known handle) once a quality characteristic drives off-target.
    Are you interested in more details? If so, please post your email address and reply to this post.
    Thanks and kind regards, Michael Schlueter

    0
    #165455

    David Hart
    Participant

    Definitely interested in additional details ([email protected])

    0
    #165503

    Michael Schlueter
    Participant

    Hi David,
    Ok, thank you. Please allow me a few days to compile and prepare material for you.
    Kind regards, Michael Schlueter

    0
Viewing 6 posts - 1 through 6 (of 6 total)

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