Buy article PDF
Instant access to
the full article PDF
for the next 48 hrs
US$ 35
|
Structural Engineering and Mechanics Volume 74, Number 2, April25 2020, pages 157-173 DOI: http://dx.doi.org/10.12989/sem.2020.74.2.157 |
|
|
|
Finite element model updating of long-span cable-stayed bridge by Kriging surrogate model |
||
Jing Zhang, Francis T.K. Au and Dong Yang
|
||
| Abstract | ||
| In the finite element modelling of long-span cable-stayed bridges, there are a lot of uncertainties brought about by the complex structural configuration, material behaviour, boundary conditions, structural connections, etc. In order to reduce the discrepancies between the theoretical finite element model and the actual static and dynamic behaviour, updating is indispensable after establishment of the finite element model to provide a reliable baseline version for further analysis. Traditional sensitivity-based updating methods cannot support updating based on static and dynamic measurement data at the same time. The finite element model is required in every optimization iteration which limits the efficiency greatly. A convenient but accurate Kriging surrogate model for updating of the finite element model of cable-stayed bridge is proposed. First, a simple cable-stayed bridge is used to verify the method and the updating results of Kriging model are compared with those using the response surface model. Results show that Kriging model has higher accuracy than the response surface model. Then the method is utilized to update the model of a long-span cable-stayed bridge in Hong Kong. The natural frequencies are extracted using various methods from the ambient data collected by the Wind and Structural Health Monitoring System installed on the bridge. The maximum deflection records at two specific locations in the load test form the updating objective function. Finally, the fatigue lives of the structure at two cross sections are calculated with the finite element models before and after updating considering the mean stress effect. Results are compared with those calculated from the strain gauge data for verification. | ||
| Key Words | ||
| cable-stayed bridge; fatigue life; health monitoring; mean stress effect; model updating; surrogate model | ||
| Address | ||
| Jing Zhang and Dong Yang: Department of Civil Engineering, Hefei University of Technology, Hefei, Anhui Province, China Jing Zhang, Francis T.K. Au and Dong Yang: Department of Civil Engineering, The University of Hong Kong, Pokfulam Road, Hong Kong, China | ||