ABSTRACT
The objective of this paper is to show the preliminary results of a research effort involving the integration of Structural Health Monitoring (SHM) data and structural reliability concepts. It has been demonstrated that the use of statistics of extremes to characterize the live load effects for larger timeframes, such as 75-year, can be a invaluable tool in the safety evaluation of monitored structures (Messervey 2009; Messervey et al. 2011).
Two strategies are applied to characterize the extreme value distribution (EVD) of the live load effects one based on the works cited above (Figure 1) and another based on the definition of an underlying distribution (Figure 2) and posterior simulation and fitting of the 50-year EVD (Figure 3). Gumbel distribution based transformation. https://s3-euw1-ap-pe-df-pch-content-public-p.s3.eu-west-1.amazonaws.com/9781315207681/cd556cd4-4dcf-4efe-8e29-56fc67b8bfbd/content/fig72_1.tif"/> PDF of the underlying distribution. https://s3-euw1-ap-pe-df-pch-content-public-p.s3.eu-west-1.amazonaws.com/9781315207681/cd556cd4-4dcf-4efe-8e29-56fc67b8bfbd/content/fig72_2.tif"/> Transformation and fitting of the 50-year EVD. https://s3-euw1-ap-pe-df-pch-content-public-p.s3.eu-west-1.amazonaws.com/9781315207681/cd556cd4-4dcf-4efe-8e29-56fc67b8bfbd/content/fig72_3.tif"/>
Using these two strategies to determine the live load effects and by applying sectional analysis to the monitored reinforced concrete girder the dead load effects and expected strain at failure of the section are obtained. The comparison of the results show that the assumption of a Gumbel distribution for the 1-day EVD is not the best fit for the collected data, however for the analyzed case, subtle changes in the reliability index were detected. The researchers will carry out further investigation regarding the use of smaller measurement times and the effects in these two strategies.
