Search

In the proposed modeling approach, when any of the essential key factors are not known initially, they can be approximated in various ways with a known impact on the accuracy of the final predictions. For example, on any program where reliability predictions are started at interim states of project completion, a-priori approximate estimates of the key factors are established for making preliminary predictions. Later these are refined for greater accuracy as subsequent program information of a more definitive nature becomes available. Specific steps to develop, validate and verify these new models are described.

New fundamental technique of reliability prediction for semiconductor diodes based on realistic mathematical models can be applied to component failure rate prediction including mechanical degradation, electrical degradation, environmental stress factors, and electrical load stress factors.

Mathematical models for predicting reliability of semiconductor diodes