In this paper, studies for structural condition assessment of the movable bridge using the Structural Health Monitoring (SHM) data collected from a laboratory based 4-span bridge-type steel structure will be presented. For this purpose, the authors expand on a time series analysis method using ARX (Auto-Regressive with eXogeneous input) models for damage detection with free response vibration data. A series of ARX models were developed for reference locations using adjacent location acceleration values as inputs. After creating the baseline ARX models from the undamaged structure, they were used with the data from one of the simulated damage scenarios. The error term of these damaged models, expressed as a proportion of the fit ratios was used as the damage sensitivity feature. Use of a normalized distance factor for the damage features was also shown to be highly effective at visually presenting the results. This ARX based time series analysis method was shown to be effective for damage detection, locating, and assessment of the relative damage severity for this relatively complex laboratory structure.
CITATION STYLE
Terrell, T., Gul, M., & Catbas, F. N. (2011). Structural health monitoring of a bridge model using ARX models. In Conference Proceedings of the Society for Experimental Mechanics Series (Vol. 4, pp. 357–364). Springer New York LLC. https://doi.org/10.1007/978-1-4419-9316-8_34
Mendeley helps you to discover research relevant for your work.