Statistical performance sensitivity-a valuable measure for manufacturing oriented CAD

The authors delineate the random and deterministic components of the manufacturing and design process. The model introduced shows the differences between statistical sensitivity and classic sensitivity. The statistical sensitivities are an extension of the classic sensitivities, applied to the general statistical outcome of the manufacturing process. The calculation procedure involves the efficient estimation of all the statistical sensitivities by using one Monte Carlo simulation. The calculation of these sensitivities and their application to a Salin and Key active filter are discussed.<<ETX>>