A Framework for Early Assessment of Functional Failures to Aid in PHM Design

Abstract

Human artifacts have evolved into incredibly complex systems that rely on both hardware and software to function dependably with limited or no human operator control. As the monetary and human consequences for failure climb with technological progress, Prognostic and Health Management (PHM) systems are developed to help manage, mitigate, detect, and respond to failures. These PHM systems have become as integral and as important as any other subsystem in a complex machine. Methods have been developed to analyze and incorporate risk and reliability in the early stage decision making processes of complex system design. These methods, however, treat the development and capabilities of PHM in the conceptual stage of system design in a peripheral manner at best. The importance of PHM consideration early in system design is significant. The structure of PHM, one of the most complex of the subsystems, can begin to take shape. Design modifications that must be made in light of PHM limitations should be pushed to the earliest stage of design possible, where costs of changes are minimized. In this paper, several risk and reliability based design techniques are discussed in this context. In particular Function Failure Identification Propagation framework (FFIP) provides a systematic process to identify potential failure points and their resulting functional losses. FFIP is selected in this paper as fitting early stage PHM development very well. However, improvements are proposed to the existing FFIP process to better address PHM design needs during the conceptual design stage of a complex system. This paper presents an improved method, which is then applied to a liquid fueled rocket engine architecture. Future work, including using the information gathered here to start the conceptual design of the PHM system, is also discussed.