Damage detection of steel girder railway bridges utilizing operational vibration response

In this paper, we develop a damage identification framework based on acceleration responses for railroad bridges. The methodology uses sensor-clustering-based time series analysis of bridge acceleration responses to the motion of the train. The results are expressed in terms of damage features, and damage to the bridge is investigated by observing the magnitude of these damage features. The investigation demonstrates the damage features by comparing the fit ratios of locations of interest so that damage can be identified and located and the relative severity of the damage assessed. The damage cases considered are stiffness loss, moment capacity reduction, and change in boundary conditions. In this study, a finite element analysis of a railway bridge model is used to verify our methodology. Our findings show that the proposed damage detection framework is very promising for continuously assessing the condition of railway bridges and thus will facilitate early detection of potential structural damage. This will be valuable for infrastructure owners seeking to develop more economical and effective maintenance strategies.