Damage Identification and Optimal Sensor Placement for Structures under Unknown Traffic-Induced Vibrations
This paper proposes a damage-identification and optimal sensor-placement approach for structures under unknown traffic-induced vibrations. Response reconstruction is performed for structures under traffic-induced vibrations to formulate the relationship between different sets of time-domain response...
| Main Authors: | , , |
|---|---|
| Format: | Journal Article |
| Published: |
American Society of Civil Engineers
2015
|
| Online Access: | http://purl.org/au-research/grants/arc/DE140101741 http://hdl.handle.net/20.500.11937/17356 |
| Summary: | This paper proposes a damage-identification and optimal sensor-placement approach for structures under unknown traffic-induced vibrations. Response reconstruction is performed for structures under traffic-induced vibrations to formulate the relationship between different sets of time-domain responses. Measured acceleration responses from a structure under traffic-induced ground motions are used for damage identification with a sensitivity-based iterative model updating method. Numerical simulations on a seven-story planar frame structure are conducted to verify the performance of the proposed approach. Damage identification is conducted based on the initial finite-element model of the structure and the measured responses from the damaged state under traffic-induced vibrations. Structural elemental stiffness factors are iteratively updated until the reconstructed responses match the measured ones as closely as possible. A two-phase optimal sensor-placement scheme is proposed for better response reconstruction and subsequent damage identification. Optimal sensor placement is investigated to identify the best locations for response reconstruction and sensitivity-based damage identification. With the use of optimal sensor locations, the introduced damages can be identified effectively and accurately with nearly exact damage severity estimation and very small false positives and false negatives under a 5% noise effect. |
|---|