Thermoplastics challenge


My challenge that brings me to the community has been attempting to statistically improve control over a wire finishing process. I've watched the pq systems two part webinar on control charts and capability and it had some insights that should be valuable. However I still have some questions due the complexity of the process.

So, I have a part measured for coating thickness over wire in four places, each measuring location will have a different mean by virtue of its physical location and orientation, the objective of course is to reduce the variation as there is just one specification.

Now I have 3 different substrate materials in 5 different gages that have a tolerance for diameter of their own that I'm measuring the coating thickness over with a caliper. In addition say I have X different coating materials and Y possible positions inside the coating equipment itself.

That's before I get into any of the other more complicated process parameters of the equipment itself and the thermodynamics involved. Let's ignore those for now, I can't address them here anyway.  

Currently I use an xbar and range chart that I can filter by date, part#, substrate material, and coating material. It has four data columns, one for each measurement location on any given piece.

In the past I have sampled several parts a few times a day, the webinar points out that this is actually not as good at capturing my variation as more frequently sampling just one part. Resulting in control limits that are tighter than they should be, but the process would still be out of control regardless.

My question is if I'm going to reevaluate the design of the control chart and the way I measure and sample, should I instead chart this as individuals and moving range if the end goal is to communicate just the one spec?

For example (chart includes some special causes and experimental results over one substrate type):


  • Steve DSteve D PQ Systems Employee
    Thanks for your question. Just a quick check for understanding. Would you say these columns would at least capture all the data you are examining?

    Substrate Material
    Coating Material
    Wire Gage
    Measured Location
    Coating thickness

    Also, is the one specification you have an Upper specification?


  • I have 4 columns for measurement data, the coating thickness, these are labeled upper left, upper right, lower left, lower right. The four combined represent one part, or one averaged point on the xbar chart.

    The other columns are identifiers.

    The specification I have is an upper and lower bound for the thickness taken over one side of the wire. Internally we use a +\- tolerance over both sides.

  • Steve DSteve D PQ Systems Employee
    Would you say the 4 measurements vary independently? For example, would it be likely that upper right would show out-of-control signals while the other 3 remain stable?
  • They vary independently for a few different reasons, some of those causes are what we address when we are trying to control them. Mostly orientation vs gravity, but also some thermodynamics and process parameters involved.

    Two out of the four show more out of control signals but their mean is closer to nominal. One out of the four has the lowest mean, and creates a problem when I have a data set that can show capability.

    That's basically my question though about possibly running this as independent and moving range charts. If there is only one spec, am I not capturing the full variation unless I'm treating them all as if they don't vary independently and run the chart blind to most of my variables? For diagnosing problems my chart has been useful, but I don't want to misrepresent the data as being more out of control than the process actually is.

  • Steve DSteve D PQ Systems Employee
    It sounds like you could look at the 4 metrics as different systems - monitoring each one with an Individuals/Moving range chart. This gives you a more granular look at the source for unwanted variation. This does not preclude having an additional X-bar chart that views the 4 measurements as a single system. Note that just having additional charts does not require that someone look at them all the the time. For example, the X-bar chart might be setup with alerting to notify you only when extreme conditions are detected. If you want to discuss this further, feel free to contact me via our application support channels.
Sign In or Register to comment.