The Elsmar Cove Forum The CORRECT steps to implement an SPC chart
 Forum User Name Keep Me Logged In Password
 Register Photo Albums Blogs FAQ Social Groups Calendar Search Today's Posts Mark Forums Read

# The CORRECT steps to implement an SPC chart

Posted 17th November 2010 at 10:46 AM by bobdoering
Updated 22nd November 2010 at 04:27 PM by bobdoering

You might have had some SPC training in the past, but it probably skipped over the correct steps to implement charting on the shop floor. Check it out, here they are:

1. Develop the total variance equation (Yes! This is the very first step!)

2. Determine which variance factors are adjustable, which are noise, and which can be set as a constant.

3. Minimize the variation of each of the participating variables - get the process in a steady state and capable. This includes eliminating your "special causes".

4. Accurately determine the correct distribution of each of the remaining variances

5. Determine which of the remaining variance factors you are going to chart

6. Pick the correct chart to evaluate that variance factor (variable)

Did anyone share that with you in your SPC training? If not...you missed out on the important stuff...

1.  Quote: In Reply to Parent Post by calin.furdui Can anyone help me with some excel file that is calculating Cpu, Cpk, Cp, PPM, and staff like this and is also doing a bell curve with the histogram for things like flatness or perpendicularity ? It is important to note that to calculate capability on unilateral tolerances with a physical limitation (such as '0' for flatness), the bell curve does not apply. It is the wrong distribution to model that variation. You will generally need to use a beta or weibull distribution to describe those variations. Often, people will use those correct distributions, then transform them to calculate capability. Best tool, use Distribution Analyzer at variation.com Posted 2nd March 2012 at 10:52 AM by bobdoering
2.  This is regarding knowing the distribution before SPC. Doesn't central limit theory apply for sufficiently large sample size data? Posted 3rd March 2012 at 08:12 AM by alias710
3.  Quote: In Reply to Parent Post by alias710 This is regarding knowing the distribution before SPC. Doesn't central limit theory apply for sufficiently large sample size data? CLT has been used to support the Shewhart charting methodology. Its usage has been debated, and is generally felt to be a weak or inadequate statistical basis for that charting methodology. Whether you accept that, or Chebychev's inequality to support Shewhart's chart, you need to recognize that it only holds that if your process data is independent and random, then Shewhart's charts can be used to determine if the processes has a special cause. That leaves two very important reasons to understand the distribution: 1) To see if the process distribution meets the statistical assumptions for Shewhart charts, and 2) to determine if the process is capable and stable. Shewhart's charts and their calculation hold no connection to capability, so the correct model distribution is needed for that evaluation. Therefore, as an overall SPC implementation process, it is necessary. Posted 5th March 2012 at 01:24 PM by bobdoering
4.  Thanks for the clarification Bob! Posted 15th March 2012 at 08:18 AM by alias710
5.  Need one more help. How useful is SPC for Press parts and assembly of parts? If not what else is recommended? Thanks... Posted 15th March 2012 at 08:31 AM by alias710
6.  Quote: In Reply to Parent Post by alias710 Need one more help. How useful is SPC for Press parts and assembly of parts? If not what else is recommended? Thanks... I really depends on the process that generates the output characteristic - especially if it is variable characteristic. I hate to answer so generically, but the question does not pose enough specifics: kind of variable, sources of variation that affect it (total variance equation), etc. Posted 26th March 2012 at 02:27 PM by bobdoering

Do you find this blog post helpful and informational?

The time now is 09:37 AM. All times are GMT -4.