The Predictive error (PE) is the error associated with any shrinkage prediction for a chosen material, and is constant for that material.
It represents the confidence band for the shrinkage prediction and is calculated as the average error in predictions based on the error in the shrinkage model used.
The Predictive error (PE) is the error associated with any shrinkage prediction for a chosen material, and is constant for that material. It represents the confidence band for the shrinkage prediction and is calculated as the average error in predictions based on the error in the shrinkage model used.
The graph allows you to compare the tolerances that you've specified for the part dimensions with the errors associated with the predictions. This is done by converting the predictive error to a distance that allows direct comparison with the distance range given by the specified tolerances.
Ideally the predicted errors should lie within the desired part tolerances, for each critical dimension. If the predictive errors are well beyond the part tolerance, then it is clear that the desired tolerances cannot be met with the chosen material.