This paper addressed the well-known methodology of the Fault Tree Analysis for modeling,rnprediction and failure mode mitigation and reliability improvement of a product. The analysis startsrnearly in the product development phase and continues as the design evolves. Detected failure modes thatrnare considered a threat to the product safety, or are found to be high contributors to the productrnunreliability are gradually addressed and mitigated as needed. Thus, reliability growth of a product isrnrealized in the product design phase while it is inexpensive and relatively simple to introduce necessaryrnimprovements.rnThe analyses are updated in the process of design reliability growth and the resultant reliabilityrnimprovement is continuously monitored. Information for failure mode probability necessary for the FTArnis obtained from the known and verified components failure rates found in software databases, obtainedrnfrom the component manufacturers’ life testing, or determined through a component reliability test. ThernFTA allows calculation of the product reliability. Mitigation of failure modes is accounted in the faultrntree, and the resultant improved reliability is recorded. The process lasts throughout the product designrnphase. Additional testing, when possible is then used for detection and possible mitigation of thosernfailure modes not identified during analysis.rnThe combined FTA/prediction methodology allows for a combination of hardware failure rate orrnfailure probability information with the probability of environmental and other operational effectsrnresulting into inexpensive product improvements and reduction in cost of long testing and productrnreadiness (time to the market) delays.
展开▼
