Improved utility and application of probabilistic methods for reliable mechanical design

摘要

In a modern product development process such as in the automotive and aerospace sectors, extensive analytical and simulation approaches often are used to assess the ability of a design in fulfilling its requirements. Consideration of uncertainty in such situations is critical in ensuring a reliable design is produced. Probabilistic methods facilitate an improved understanding of design performance through characterization of uncertainty in the design parameters. The probabilistic methods developed over the past several decades have a range of capabilities and modes of application, for example, to predict reliability, for optimization, and to perform sensitivity studies, but have yet to be taken up routinely by industry due to a number of reasons. In this paper, issues that have typically inhibited their use or prevented a successful outcome are addressed through a systematic framework for improved utility and successful application of probabilistic designing for mechanical reliability.