This paper proposes method of uncertainty quantification of modal parameters and presents results acquired from monitoring of a 10-span highway bridge in Austria. Motivation for this research was to ensure robustness of damage detection based on long-term monitoring data. Therefore, the variability due to changing environmental and operating conditions was of primary concern. The proposed evaluation method uses clustering of modal identification results obtained at different conditions. Stochastic subspace identification (SSI) was used to identify modal parameters from ambient measurements. Besides the variability due to changing conditions, the accuracy of parameter extraction (within one data block) was also considered. The procedure contains 3 steps of data reduction: extraction of a set of modal parameters from one data block, statistical properties of this set, and statistical properties all sets from all data blocks. Compensation of temperature effects was performed by regression analysis. While the temperature influence on eigenfrequencies was significant, the mode shape and damping values showed lower but detectable sensitivity. Results of this uncertainty quantification are intended for the use in probabilistic damage detection.
|Publikationsstatus||Veröffentlicht - 2016|
|Veranstaltung||ISMA/USD2016 - |
Dauer: 19 Sept. 2016 → 21 Sept. 2016
|Zeitraum||19/09/16 → 21/09/16|
- Ehemaliges Research Field - Mobility Systems