| This dissertation developed new methodologies to extend existing degradation modeling capabilities to fielded systems, which may have many degradation failure mechanisms, and operate under diverse environmental and usage conditions. Degradation modeling is an important analytical tool to serve as a basis for reliability prediction. Reliability prediction can be made by observing and analyzing degradation patterns prior to the occurrence of failures. Prior research has produced many significant degradation modeling approaches. These were extended to consider a much broader class of reliability prediction problems.; Significant and important research has already been conducted for degradation modeling. This dissertation reviews three existing degradation models, namely (1) general path model, (2) degradation hazard function model, and (3) Markov model. All of these models have already been validated and have been demonstrated to be very useful for their intended applications, but they have respective limitations for more difficult problems. A common limitation is that none of them have been applied to fielded systems where stress-levels and failure thresholds are more complex. Failure thresholds are often assumed to be constant. Either no stress is explicitly considered or stresses are assumed to be stationary over time. These assumptions add restrictive limitations to the application of developed reliability prediction methodologies based on degradation modeling because in reality, many actual engineering problems, where degradation is observed and data is analyzed, have random, variable failure thresholds, multiple changing stresses, and multiple degradation measures. The objective of this research is to address these issues so the degradation modeling approach can be truly useful for a broader range of reliability prediction problems.; This dissertation developed extensions to the existing research so that degradation modeling can address new problem domains that previously did not meet the required assumptions and modeling constraints. These extensions will broaden the use of degradation modeling methods and can be used to make informed design decisions that will lead to more reliable and safer component and system designs. |
