Search Results

Severity-mitigating mechanisms (typically software-based) detect failures in a system and perform functions in order to reduce the severities of failures. Various approaches to FMEA analysis of severity-mitigating mechanisms exist within the industry. Three are compared and contrasted. Each method is compared against its ability to capture the three fundamental failures of a system that has severity-mitigating mechanisms: 1 a failure occurs and mitigating action is taken, 2 a failure occurs and mitigating action is not taken, 3 no failure occurs but mitigating action is taken. One method is advocated over the others because it: uses existing FMEA formatting; addresses all three cases; supports consistent linkage between FMEAs in a hierarchy of systems with any number of layers.

When analyzing the failure rate (or occurrence) of a system failure cause, the typical approach is to obtain an occurrence rating from the results of testing. However, in many cases, the occurrence of a system failure cause can be derived from a combination of occurrences of failure causes of the element (sub-system) failure mode coinciding with the system failure cause being assessed. This paper explores a few approaches for deriving occurrences from element FMEAs over a majority of cases before settling on a probabilistic approach that converts occurrences to worst-case failure rates to achieve the most fine-tuned combined occurrence rating. Finally, a “complex analysis” worksheet, where the logical combination of occurrences and failure rates is custom defined by the engineer, is introduced for handling special cases.