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.

Gage R&R according to customer spec

Discussion in 'Gage R&R and MSA - Measurement Systems Analysis' started by laffranchinis, Sep 10, 2020.

  1. laffranchinis

    laffranchinis Member

    Joined:
    Jan 30, 2020
    Messages:
    12
    Likes Received:
    2
    Trophy Points:
    2
    Hi Everybody,

    I am a rookie in MSA. I just run few of them (both internal and for customers) but I still have some doubts regarding the correct way to run a Gage R&R when we have a tolerance given by our customers, and I hope you can help me to dissolve these doubts.
    Here below two approaches to run a MSA:

    1) Sample selection: samples taken on purpose over the entire tolerance field given by customer.
    --> We consider the %Study Var to assess if the Measurement System is OK for the intended application.

    2)sample selection: random from our process. In Minitab we will insert manually the customer tolerances.
    --> We consider %Tolerance to asses if the Measurement System is OK for the intended application.

    Which one is a more correct approach? With the first case, for some processes, oft is very difficult to prepare samples scattered over the entire tolerance field given by the customer.


    BR
    Samuele
     
    laffranchinis, Sep 10, 2020
    #1
  2. Miner

    Miner Moderator Staff Member

    Joined:
    Jul 30, 2015
    Messages:
    577
    Likes Received:
    493
    Trophy Points:
    62
    Location:
    Greater Milwaukee USA
    Samuele,

    Neither one is correct. First decide how the gage will be used (inspection to a tolerance, SPC or both).

    If the gage will be used for inspection to a tolerance, you may select the parts as you wish, because the variation of the parts will not affect the results of the correct metric, which is % Tolerance. Ignore % Study Variation and ndc, because they are irrelevant in this case.

    If the gage will be used for SPC, you should either randomly select the parts from actual process variation and used % Study Variation or ndc, or you should use the variation from a capability study/control chart and use % Process Variation (preferred). Ignore % Tolerance in this case.

    If the gage will be used for both SPC and inspection, you should either randomly select the parts from actual process variation and used % Study Variation or ndc, or you should use the variation from a capability study/control chart and use % Process Variation (preferred). In this case you will also consider the results of % Tolerance.
     
    Miner, Sep 10, 2020
    #2
  3. laffranchinis

    laffranchinis Member

    Joined:
    Jan 30, 2020
    Messages:
    12
    Likes Received:
    2
    Trophy Points:
    2
    Hi Miner,


    As usual, thanks a lot for your quick reply! this really help me to understand how to run such gage R&R in a correct way. I do not know why, but on the web you find many times statements that the samples should be have the characteristic to be measured scattered over the tolerance field.


    Samuele
     
    laffranchinis, Sep 10, 2020
    #3
  4. Miner

    Miner Moderator Staff Member

    Joined:
    Jul 30, 2015
    Messages:
    577
    Likes Received:
    493
    Trophy Points:
    62
    Location:
    Greater Milwaukee USA
    There is a lot of misinformation spread over the web. Some of it is from people that honestly don't know better passing on what they were told, and some of it is from people that know selecting samples in that manner will usually make their metrics appear better than they would otherwise look.
     
    Miner, Sep 10, 2020
    #4
    Bev D likes this.
  5. kyleruppelt

    kyleruppelt Member

    Joined:
    Oct 30, 2020
    Messages:
    6
    Likes Received:
    2
    Trophy Points:
    2
    Our company is moving in the TQM phases,1-5. We are currently at 2 and wanting to go to level 3. i am having difficulty understating something and battling to find clarity on what is required and why it is required.
    So in the MSA section the requirement is to use Control Limits and not specification limits on the MSA before SOP. To my knowledge Control limits are calculated by CL=m + or - 3*std deviation?
    What global is saying, is we must redo the study but use Historical Standard deviation from our SPC done with line side data collection.

    So when we redo the MSA using the same data from original MSA (using full range) instead if giving the Spec limits in minitab, we use the historical standard deviation option.

    1. What is the benefit to doing it this way?
    2. What is the difference between using Control Limits and historical Std Deviation
     
    kyleruppelt, Nov 27, 2020
    #5
  6. Miner

    Miner Moderator Staff Member

    Joined:
    Jul 30, 2015
    Messages:
    577
    Likes Received:
    493
    Trophy Points:
    62
    Location:
    Greater Milwaukee USA
    You have a lot going on in your question, some of which I think you may have misunderstood part of what you were asked to do.

    Please read my previous posts carefully. If I understand part of what you are asking, your initial studies used specification limits, which would be appropriate if your gage is used for inspecting product to tolerance and you are using % tolerance as the metric. You are now being asked to use a standard deviation derived from your ongoing SPC. This is appropriate when your gages are used for SPC, which they obviously must be since you have ongoing SPC data. You would use this to calculate % Process Variation (%PV). Note: You are not using control limits, but are using the actual process standard deviation derived from the SPC data. In Minitab, use your SPC data in a capability study to determine the process standard deviation.
     
    Miner, Nov 30, 2020
    #6