Extreme Value Statistics for the Life-cycle Assessment of Masonry Arch Bridges

Arch bridges made of natural stone masonry nowadays rank among the oldest, structures still in service in the road and railroad infrastructure. Some of them with ages of more than 100 years, represent a historically very valuable building fabric. In Austria's railroad network alone there are around 1,000 arch bridges still in operation. In whole Europe the estimated number of masonry railroad bridges amounts to 70,000. Nowadays increasingly more attention is paid to the maintenance and rehabilitation of existing arch bridges than on their reconstruction, as a result of financial reasons as well as monument conservation aspects. Since those bridges usually have been projected for very different loads, assessments concerning the current bearing capacity and the future utilization are required. A recalculation with conventional calculation methods is often insufficient as results can considerably deviate from the actual load capacity due to various influencing factors. The challenge therefore is to find and verify more accurate methods to determine the actual condition of such arch bridges. A point of interest is the combination of finite element modeling strategies of masonry with inspection and monitoring strategies of existing structures. Inspections enable the detection of "damage", whereby damage is defined as changes in material as well as geometric properties of a structure. The applications of monitoring systems allow a continuous observation of a system over time. In case of long term observations, the output of this process is a periodically updated information regarding the ability of the structure to perform its tasks. Within the proposed contribution a generally valid methodology for the performance and lifetime assessment of masonry arch bridges based on nonlinear finite element analysis and monitoring data will be presented. Further two different approaches for the assessment of extreme values and probability of failure will be discussed. The two approaches will be applied to a railroad arch bridge located in Austria.