Important: If you need to estimate very small defect rates, use Normal weighted instead of just Normal. This allows defect rates of just a few parts per million to be estimated well with only a few thousand simulation runs.

Tolerance Design is the investigation of how defect rates on the outputs can be controlled by controlling variability in the input factors.

In this case, the defect rate that is reported below all the factors is estimating the same quantity, the rate estimated for the overall simulation below the histograms (that is, if you clicked the Simulate button). Because each estimate of the rate is obtained in a different way, they might be a little different. If they are very different, you might need to use more simulation runs. In addition, check that the range of the factor scale is wide enough so that the integration covers the distribution well.

The mean and SD are updated when you change the factor distribution. This is one way to explore how to reduce defects as a function of one particular factor at a time. You can click and drag a handle point on the factor distribution, and watch the mean and SD change as you drag. However, changes are not updated across all factors until you click the Rerun button to do another set of simulation runs.

Assume we want a defect profile for factor X1, in the presence of random variation in X2 and X3. A series of n=N Runs simulation runs is done at each of k points in a grid of equally spaced values of X1. (k is generally set at 17.) At each grid point, suppose that there are m defects due to the specification limits. At that grid point, the defect rate is m/n. With normal weighted, these are done in a weighted fashion. These defect rates are connected and plotted as a continuous function of X1.