Closed Form Solution for SHM-Based Bayesian Reliability Assessment

Abstract

Extensive structural health monitoring (SHM) of civil infrastructure provides massive amounts of data that must be promptly and effectively analyzed through appropriate models to infer the health state of structures. In general terms, the health state of a structure can be expressed as its structural reliability with respect to the most significant limit states. When the reliability is estimated based on SHM data, the problem is typically solved numerically with iterative approaches, which are computationally expensive and do not allow early warning and prompt response. This work presents a logically consistent approach for the Bayesian estimation of structural reliability based on SHM observations. Moreover, it illustrates how closed-form solutions can be obtained using linear models and Normal random variables. The proposed approach can effectively evaluate the sensitivity of structural reliability with respect to SHM observations; this can be used as a novel performance index for monitoring systems. Finally, this paper proposes an application of this approach to a real-life case study, the crack opening monitoring of the Settefonti highway viaduct in Italy.