Abstract:

Reliability predictions are often used as the basis for important business decisions, including parts selection, environmental controls, qualification plans, warranties, maintenance plans, and spare requirements. Before a production can be trusted for use, especially in applications requiring high reliability and safety, it must be carefully assessed for validity. In other words, reliability models should only be employed once they are judged to be reliable themselves. It is thus imperative that they satisfy an established set of assessment criteria before being applied to electronic products. This paper establishes the minimum criteria needed to assess the value of a reliability model. The criteria emphasize the influence of environmental and operational stresses; variabilities in failure mechanisms, modes, and sites; design parameters; manufacturing processes and defects; and the statistical distribution of failures. If a model satisfies the criteria, it can be certified as robust enough to provide competent assessments. As an example, the criteria are used to assess various microcircuit reliability models.

Complete article is available to members of CALCE

© IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.