Condition assessment and monitoring of deteriorating concrete structures with Bayesian methods

One of the major factors affecting structural performance in time is deterioration of its components due to environmental conditions. Because large uncertainties are associated with the process of structural deterioration, probabilistic inference methods are more suitable for strength assessment and prediction. This contribution proposes a new framework of condition monitoring and remaining strength prediction for deteriorating concrete structures. A Bayesian dynamic linear model is used to describe the dynamics of the condition monitoring parameters. This model incorporates a certain deterioration model and will be used to describe the process of structural performance deterioration. When monitoring information becomes available, the evolution trend could be predicted and updated within the Bayesian framework. Meanwhile, the Cumulative Bayes Factors (CBF) are calculated to detect abnormalities in the parameter sequences. The advantage of using cumulative Bayes factors is that it can track changes of the sequence structure (evolution trend) even if outlier data at certain time points exist. Hence, it can avoid false alarms and can be used in consecutive monitoring, giving a timely warning when an out-of-range deterioration/accident occurs and providing information for optimal maintenance strategies