Destroy the Myths of the Normal Curve
in Precision Machining
Myth: For stable processes, you should see a normal distribution.
For stable normal or "naturally varying" processes, yes! But precision machining does not exhibit a normal distribution, because the "voice of the process" indicates that it does not - and should not - randomly vary about the mean. In fact, if you witness this type of variation, your process is unstable and out of control! Precision machining exhibits a different stable distribution. The mean has no value to the control of the process whatsoever. CorrectSPC will show you why!
Myth: X bar-R charts are good for any distribution.
X bar- R charts are the worst charts for precision machining! The x data is a statistically insignificant amount of data for a diameter or a length, and the average of a sample set of insignificant data is of even less value. The range ends up being a value of the measurement error. As a result, using the average of insignificant data, and its related measurement error has no value in decision making for precision machining. Also, the calculations for the control limits are for the wrong distribution. Side-by-side plant floor case studies show that X bar-R charts yield the exact opposite of the correct process condition - signaling out of control when it is in control and vice versa! It simply does not work. CorrectSPC utilizes a revolutionary new SPC charting methodolgy that exhibits the GD&T definition of a circular or length characteristic. CorrectSPC will show you how!
Myth: For continuous improvement you must recalculate and compress your control limits.
First, the calculation for control limits of a normal distribution are wrong for precision machining! That is why when they are used, they compress the control limits to an unreasonable level.
Second, by understanding the true distribution of precision machining, you will see that compressing control limits increases variability by overcontrol!
Finally, the correct control limits for precision machining are constants - and never need recalculating!
There is a way to determine continuous improvement, and it has nothing to do with the X data control limits! CorrectSPC will show you how!