Seismic base isolation is increasingly applied for the earthquake-resistant design of bridge structures. Combining large horizontal flexibility and high energy dissipation capacity, seismic isolation allows to considerably reduce seismic forces while controlling seismic displacements. Consequently, it greatly enhances the seismic performance of the structure and, most of the time, reduces long term and short term costs, allows preserving the bridge functionality, etc.
However, seismic responses of isolated bridges and the performance of the seismic isolation systems (SISs) are highly dependent on the seismic ground movement properties, especially their frequency content and spectral properties. In this context, optimal isolation characteristics for bridges located in such regions and performance of simple analysis methods should vary. SISs were primarily developed for high seismicity areas (HSAs) while in moderate seismicity areas (MSAs), seismic isolation is still used on a relatively rare basis. Consequently, more adapted SISs and rapid analysis methods for MSAs and particularly those in cold regions are of high interest and their availability should encourage a wider application of the technology.
This research subscribes to this context. Its main specific objectives are to: 1) identify the range of optimal characteristics of SISs for HSAs and MSAs; 2) revisit the limits of application of the simple single-mode spectral analysis method for moderate and high seismicity areas and supplement the method with useful predictions for the base-isolated bridge design; and 3) carry out analytical and numerical formulation/investigation of the behaviour of an SIS combining laminated natural rubber bearings and U shaped metallic dampers, a potentially well fitted solution for SIS applications in moderate and high seismicity cold regions.
Based on the global seismic activity map, the classification of seismic zones is carried out from the point of view of seismic isolation, through the frequency content and spectral parameters. Common characteristics of seismic records for each seismicity class are identified and their differences highlighted in relation to seismic isolation efficiency. Results confirm that earthquakes in MSAs are characterized by a concentration of energy at high frequencies and the extension of the vibration period leads to a more rapid decrease in the seismic spectral acceleration, compared to HSAs. On the basis of an extensive parametric analysis where SIS properties are varied, the optimal ranges of the main two SIS hysteretic features, namely the characteristic strength (Qd) and the post-elastic stiffness (Kd), are specifically determined and clearly distinguished between HSAs and MSAs. Results show that SISs for bridges in HSAs require high energy dissipation capacities, while the devices with low to moderate damping capacities are more appropriate for bridges in MSAs. More specifically, Vancouver and Montreal sites, representative of HSAs and MSAs, respectively, are considered. With the actual seismic design data, the optimal values of the characteristic strength to weight, Qd/W, are found in the range of [0.08 to 0.12] for the Vancouver site and in the range of [0.015 to 0.045] for the Montreal site. Optimal post-elastic stiffness is found to correlate strongly to optimal characteristic strength and empirical relations are proposed for each site.
In addition, the application limits and performances of the single-mode spectral analysis method (SMSA) for the seismic analysis of isolated bridges in North America are reviewed and completed. To do this, the seismic demands of isolated bridges, generated within the framework of the parametric study, as predicted by the SMSA, are compared with the results of nonlinear time history analyses, including inside and outside the applicability limits specified in current codes. In this way, the advantages of these limits and their effects on the accuracy of the SMSA are investigated and clarified. The results obtained indicate that the limits on the equivalent viscous damping and on the restoring force specified in the current codes are the most effective conditions, while the conditions related to the limits on the effective period and the post-elastic period can be ignored. To complete the SMSA, a predictive relationship of the residual displacement as a function of the restoration system and the expected margins of error of the method are proposed for the isolated bridges located in the two Canadian sites of Vancouver and Montreal.
Compared to the range of optimal characteristics of SIS for each seismic zone, natural rubber bearings (NRBs) exhibit insufficient energy dissipation capacity. Therefore, the methods to fill this gap have been reviewed and studied. Lead-rubber bearings are expensive and not environmentally friendly, while high damping rubber bearings (HDRB) suffer from major flaws in meeting optimal designs in cold regions due to their inherent "scragging" phenomenon combined with their high sensitivity to cold temperatures. In such a context, adding metallic dampers appears to be a more effective solution for improving the isolation performance of NRBs. The addition of U-shaped dampers (UDs) is identified as one of the most promising techniques but the UD behaviour remains poorly understood and documented. Such a behavior is therefore studied analytically and numerically and the available experimental studies are used to validate the results obtained. Predictive equations for the main hysteretic characteristics of UD and NRB-UDs are proposed. The results show that the UD provides high and stable EDC in all directions without significantly affecting the flexibility of the original bearing. In addition, the NRB-UDs seismic performance is easily controlled by changing the UDs’ crosssection and the number in NRB-UDs integrated systems, making it an effective solution that can be widely applied in the refurbishment of conventional isolators, especially NRBs, and can also meet a wide range of design requirements. Additional numerical and experimental studies are however necessary to extend the results to other steel grades, to establish failure criteria, and to evaluate the residual displacements resulting from these SISs.
| Date | 22 Jul 2021 |
|---|
| Original language | French |
|---|
| Awarding Institution | - École de technologie supérieure
|
|---|
| Supervisor | Lotfi Guizani (Supervisor) |
|---|