Abstract
The traditional failure modes and effects analysis (FMEA) is a conceptual design methodology for dealing with potential failures. FMEA uses the risk priority number (RPN), which is the product of three ranked factors to prioritize risks of different failure modes. The three factors are occurrence, severity, and detection. However, the RPN may not be able to provide consistent evaluation of risks for the following reasons: the RPN has a high degree of subjectivity, it is difficult to compare different RPNs, and possible failures may be overlooked in the traditional FMEA method. The objective of this research is to develop a new FMEA methodology that can overcome the aforementioned drawbacks. The expected cost is adopted to evaluate risks. This will not only reduce the subjectivity in RPNs, but also provide a consistent basis for risk analysis. In addition, the cause-effect chain structures are used in the new methodology. Such structures are constructed based upon failure scenarios, which can include all possible end effects (failures) given a root cause. Consequently, the results of the risk analysis will be more reliable and accurate. In the new methodology, the occurrence and severity ratings are replaced by expected costs. The detection rating is reflected in failure scenarios by the probabilities of either successful or unsuccessful detections of causes or effects. This treatment makes the new methodology more realistic. The new methodology also uses interval variables to accommodate uncertainties due to insufficient data. The new methodology is evaluated and applied to a hydrokinetic turbine system. This turbine is horizontal axis turbine, and it is under development at Missouri S&T