Search

This paper presents two Bayesian hierarchical models—one utilizing the life-time data and other using the structural health monitoring (SHM) data, for degradation modeling and reliability assessment of rolling element bearings. The main advantage of the proposed life-time data based model is that, it accounts for the variability in failure times caused due to the difference in material properties, initial degradation, operating and environmental conditions by introducing Bayesian hierarchy in the model parameters. On the other hand, SHM data (such as vibration and strain) based model focuses on stochastic nature of bearing degradation, and models it using a two-phase Wiener process. In this model, the point of phase-transition is the time when the damage initiates. The detection of such a point is undertaken using Bayesian change point algorithms. For both the models, the model parameters and reliability are updated as more data becomes available. In this manner, the prior domain knowledge and life-time data or SHM data collected from the field can effectively be integrated to get updated reliability. Two case studies for rolling element bearings are presented to demonstrate the applicability to life-time as well as SHM data. [ABSTRACT FROM AUTHOR]

Degradation studies are often used to assess reliability of products subject to degradation-induced soft failures. Because of limited test resources, several test subjects may have to share a test rig and have their degradation measured by the same operator. The common environments experienced by subjects in the same group introduce significant interindividual correlations in their degradation, which is known as the block effect. In the present article, the Wiener process is used to model product degradation, and the group-specific random environments are captured using a stochastic time scale. Both semiparametric and parametric estimation procedures are developed for the model. Maximum likelihood estimations of the model parameters for both the semiparametric and parametric models are obtained with the help of the EM algorithm. Performance of the maximum likelihood estimators is validated through large sample asymptotics and small sample simulations. The proposed models are illustrated by an application to lumen maintenance data of blue light-emitting diodes. Supplementary materials for this article are available online. [ABSTRACT FROM AUTHOR]