A framework to generate fault-based behavior models for complex systems design

Abstract

Fault analysis has been identified as a crucial step during the design process. Any complex design problem requires careful consideration of fault modes, fault mechanisms, propagation of faults, etc. The verification and validation efforts in complex systems design can be improved by modeling faulty behavior. This can be done by using a library of pre-constructed faulty behavior models. Currently, a major limitation for modeling performance in complex systems design is that libraries only use nominal component behavior. Strictly using nominal behavior, as opposed to faulty behavior, leads to design uncertainty and poor verification and validation. In a reliability sense, the traditional method for dealing with uncertainty is to over-design the system. While this leads to a workable solution, it is not optimized in terms of design attributes and leads to wasted resources. This paper proposes an alternative method to capture faulty behavior by developing a framework to create component behavior models, with the goal of ultimately increasing design verification and validation during complex systems design. An example shows the implementation of the frictional wear fault mechanism for a gear component. While the results can be trivial for an individual component, the purpose of these models is to tackle complex systems design where the change in performance is measured at the system level.