iSixSigma

Capability analysis without lower limit

Six Sigma – iSixSigma Forums Old Forums General Capability analysis without lower limit

Viewing 5 posts - 1 through 5 (of 5 total)
  • Author
    Posts
  • #40416

    Patrick
    Participant

    Hi there…
    Can we use the capability analysis with attribute data or variable data, although without Lower Limit?  Let’s say I’m using the lead time to measure a process. The upper limit would be 20 min. and the perfect time would be 0 min, so in that case, there is not lower limit.
    Thanks in advance.
    Patrick Robillard
     

    0
    #125329

    Darth
    Participant

    Of course you can.  In many cases though, both attribute and variable will have natural limits of zero. For example, for time it would be zero and for defects it would be zero. But, double check and see if there any assumptions about the shape of the distribution required to do capability analysis that might affect your answer.

    0
    #125331

    Patrick
    Participant

    After all measure, I appears that the distribution is from an Normal shape.
    Am I going to use the traditional X-LSL/3s, with the the LSL as 0?
    Patrick
     

    0
    #125333

    Mikel
    Member

    No LSL, but a hard limit of 0 (minitab will not assume anything below 0)

    0
    #125341

    BTDT
    Participant

    Dr.Darth:We use food product weights and volumes to show that specification limits are usually not one sided even though they appear so. In general, this can be resolved by considering both the internal and external customers.I like to talk about Coca-Cola bottling, because most people (on this planet) are familiar with the fizzy sugar water. Each independent bottling plant buys very expensive concentrate from Coca-Cola(Atlanta) and must make a profit by using the concentrate efficiently. Each plant distributes product to the consumers.The most important CTQs for customers are colour, fizzyness and fill height. Taste is third or fourth on the list. The product must match the customer expectations of ‘a full can’. We can spend a few minutes talking about conducting a destructive Gauge R&R about ‘full cans.’ Measuring the fill heights at about 1,425 cans per minute is an interesting technical problem. Industrial Dynamics makes a Filtec machine whereby gamma radiation from an Americium-241 pellet is blasted through the top of each and every can to detectors on the far side of the can as they go down the track. The Filtec is synchronized with a kicker that can fire out underfilled cans. When the filler goes on the fritz, the kicker sounds like a machine gun. External customer:The department of consumer and corporate affairs represents the consumer, and will conduct random audits of food products to ensure that the published weight or volume on the front of the product is an accurate representation of the material inside the package. The tolerance specifications are public knowledge and defined in legislation. The punitive damages can be considerable for underweight product. An underfilled can is 0.205 gm too light. This is a combination of a practical limit of detectability and ‘reasonable doubt’ for the interpretation of the weight or volume on a printed product label.Internal customer:Management will try to minimize the degree of overfills because it is a financial loss. A very overfilled can might leak and cause problems with the packaging, but this is very infrequent. An overfilled can is 0.215 gm too heavy.Every half hour, the operator will send an standard underfilled can(red), and a standard overfilled can(green) through the Filtec to test the limits. The best is both cans get kicked out. The operators all know that a failed red can (underfill) is far more serious than a failed green can (overfill) and will run the filler accordingly.In general for food products, the target mean is infrequently halfway between the two limits and is set at the best balance point between the projected losses due to litigation on the low side and projected losses in production on the high side.You might see the common theme for conducting gauge R&R in the middle of the afternoon. Everyone gets so jazzed up on sugar and caffeine, they can stay awake for the rest of the day when we do the ANOVA stuff.Cheers, BTDT

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

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