iSixSigma

Cutting Rate Test Question

Six Sigma – iSixSigma Forums Old Forums General Cutting Rate Test Question

Viewing 8 posts - 1 through 8 (of 8 total)
  • Author
    Posts
  • #47662

    howe
    Participant

    I am trying to test rotary cutting instruments for cut rate ( time to cut a given depth). I need to determine a statistically significant # of samples to test. Typically I test 30 samples, using 10 cuts each for longevity. Is this appropriate? Do I need more samples? How can I prove that 30 is enough? Please help… Thanks.

    0
    #159591

    Jonathon Andell
    Participant

    If you don’t know whether or not your process is statistically stable, any sampling scheme is at risk.
    I think you should think about establishing a rational subgrouping scheme. With a control chart you can estimate process capability and determine whether your process is stable. Ideally, capability is estimated using stable processes.
    Bear in mind that establishing an effective scheme of rational subgrouping is very much a non-trivial task in and of itself.

    0
    #159594

    Fontanilla
    Participant

    You should consider your Components of Variance as well.  Draw out the “tree” structure and then, look at what your subgrouping strategies will be.  The COV will tell you where your process is stable/unstable and will tell you which family of variation should be attacked first.

    0
    #159606

    ellzey
    Member

    J, how are you calculating process capability from a CC?
    D, once you diagram the hiearachal (sp) layout for your batch-sample-test sequence, what method is used to analyze COV? 
    Thanks!

    0
    #159703

    Jonathon Andell
    Participant

    Once I establish that the process is stable, I use the underlying data to do the usual stuff: histogram, verify normality, capability numbers, etc.
    If this response is too cryptic to suit you, let me know.

    0
    #159704

    ellzey
    Member

    Ok, so you are simply carrying the data you collected for your CC over into the normal tool set?Thanks!

    0
    #159714

    Jonathon Andell
    Participant

    Correct, with the caveat that the data should come from a statistically stable process. At the risk of preaching to the choir let me say this. Before crunching capability numbers, one always should:
    – Develop appropriate metrics
    – Validate the measurement system
    – Confirm statistical stability (stabilize the process if necessary)
    – For continuous data, test for normality; if necessary identify a suitable distribution
    It can be a fair amount of effort, but it’s all part of one big safety net.

    0
    #159717

    Jim Shelor
    Participant

    Mike,
    Just to make sure I have the setup correct.
    Do you mean for each cutting instrument you use 30 samples and make 10 cuts on each sample?
    Then you time the rate of each cut for each instrument to a predetermined depth?
    Respects,
    Jim Shelor

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

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