Re: Tom Ryan's reply to the SPC theory point

[GrantBlair@aol.com]

In a message dated 9/7/2001 1:18:29 AM Eastern Daylight Time, 
tpr42345@aol.com writes:

<< If we are going to substitute an X-chart for a c-chart, which indeed might
 be a good idea if the Poisson assumption is not met, we are immediately
 confronted with the problem of how to determine the control limits.  This
 perhaps unanticipated obstacle cannot be avoided simply by using
 3-sigma limits.  Although the Poisson model might not be appropriate,
 we can't act as if we have a symmetric distribution. >>

Tom is on-target...there is no fundamental basis for substituting any other 
chart for a c-chart. Nor is there a practical basis, for the following 
reasons:
-Forget about LCL....The only acceptable  TARGET for a C-chart is at or near 
zero. This is the reason
C=0 sampling is the only attribute scheme accepted in the automotive industry.
-Processes which produce any other common cause level are inherently flawed 
and must be redesigned. They are the equivalent of a  box which always 
produces "red beads".
-Control charting with C charts lends itself to continuous improvement. It's 
the easiest chart to calculate and to use on the floor, and once the 2 
previous conditions are met, easiest to tell when something's gone wrong.
I can give many practical examples of this, but think the following one used 
by a team applying Deming's philosophy will suffice:
Our plant was approached by a Japanese customer with an interest in obtaining 
plastic chip to be used for molding electrical components. They had a problem 
with carbonized polymer, which changed the product from an insulator to a 
conductor. Even small amounts of carbonization (black specks) could cause 
arcing at high voltage levels.
Our sister plant had "solved" the problem with 100% visual inspection and 
speck removal of batches fed over a backlighted board
We built a minature version of the board and visually inspected a 
representative sample from each batch. Results were plotted on a C-chart.
Next, we attacked every special cause observed, then related it to common 
cause conditions.
Special cause conditions observed when extrusion nozzles were changed led to 
a quick-change procedure, followed by further investigation of temperature 
uniformity.
Raw material-driven upsets led to monitoring of all lines regardless of 
product type, when it was learned 
black specks were an early signal of other potential product problems.
After a few months work, we were able to meet our goal of a C-chart with an 
UCL of 1. The most interesting lesson learned from this was that our sister 
plant never "got it". They assumed the customer could be lured away with a 
price break, and were shocked when the customer turned them down. They never 
understood that 100%
inspection built poor quality into their process.

GrantBlair@aol.com
Ninety Six SC