1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.
Dismiss Notice
You must be a registered member in order to post messages and view/download attached files in this forum.
Click here to register.

Calculating K1, K2, and K3 for 30 part sample

Discussion in 'Gage R&R and MSA - Measurement Systems Analysis' started by S1D3K1CK, Jan 22, 2021.

  1. S1D3K1CK

    S1D3K1CK Active Member

    Joined:
    Nov 5, 2018
    Messages:
    78
    Likes Received:
    28
    Trophy Points:
    17
    Location:
    The Central Valley
    Hello, our customer requires us to do MSA's for PPAP and E-buy product at the rate of 10% of the E.A.U. or 30 parts, whichever is least. We have a format that was created prior to my start in Quality, that goes to 10 parts, and now I am the lead in Quality due to personnel retiring. I, never having any Quality background prior to my employment here, am not familiar with calculating any of the formulas required for a complete overhaul of a Gage R&R form. I have done some reading in the AIAG manuals and have created a new format but ran into an issue. The format I created is set up for a 30 part or fewer sample data collection but am stumped at completing it with gathering the K3 values up to 30 parts. Is this achievable or am I wasting time here?
     
    S1D3K1CK, Jan 22, 2021
    #1
  2. Miner

    Miner Moderator Staff Member

    Joined:
    Jul 30, 2015
    Messages:
    577
    Likes Received:
    493
    Trophy Points:
    62
    Location:
    Greater Milwaukee USA
    This has been extensively discussed without resolution in the Elsmar Cove forums. The easiest way around this is to use an ANOVA approach instead of the Xbar/Range approach, or to calculate the standard deviation directly rather than using k#*Range.
     
    Miner, Jan 22, 2021
    #2
    Bev D and S1D3K1CK like this.
  3. S1D3K1CK

    S1D3K1CK Active Member

    Joined:
    Nov 5, 2018
    Messages:
    78
    Likes Received:
    28
    Trophy Points:
    17
    Location:
    The Central Valley
    I did some research over the weekend and found helpful software for this. QI Macros has a lot of detail covering almost everything you can think of, with training on some subjects. The software uses ANOVA for their GAGE R&R. I am not familiar with ANOVA but hopefully, this software can guide me and anyone else looking for help.
     
    S1D3K1CK, Jan 25, 2021
    #3
  4. Bev D

    Bev D Moderator Staff Member

    Joined:
    Jul 30, 2015
    Messages:
    606
    Likes Received:
    664
    Trophy Points:
    92
    Location:
    Maine
    My response from Elsmar to the same question:

    With a large number of repeated measurements you shouldn't use the range. you need to use the standard deviation.
    1) the range tables do not go beyond n=25 (commonly available tables do not go beyond n=10.) because the formula for d2 is horrendous.
    2) the standard deviation is MUCH more precise at large sample sizes (n>3-5) because it uses all of the data and the range only uses the min and max number.

    Of course to do this without statistical software you need to understand the formulas - you can't treat Gauge R&R as a black box. Actually you shouldn't treat any statistical approach like a black box.
     
    Bev D, Jan 25, 2021
    #4
    S1D3K1CK likes this.