Highly ductile systems must be well characterized to facilitate performance-based design. Such characterization requires careful experimentation coupled with the development of appropriate simulation approaches. Many new bridge design and retrofit options that use advanced materials and new combinations of traditional materials are exhibiting very large ductilities. An example of a highly ductile system is that of precast segmental concrete bridge piers with unbonded vertical post-tensioning and localized use of highly ductile fiber-reinforced concrete. Precast segmental bridge support systems have been implemented in many non-seismic regions (Billington et al., 1999, Sritharan et al., 1999). Unbonded post-tensioned bridge pier systems have been receiving particular attention in seismic research across the world (e.g., Kwan and Billington, 2002a,b, Kurama et al., 1999, Ito et al., 1997, Mander and Chang, 1997, Priestley and MacRae, 1996, Priestley and Tao, 1993). The use of advanced ductile materials in combination with unbonded post-tensioning is new and has first been explored at Cornell.

One-fifth scale experiments on partial columns with unbonded post-tensioning and localized use of the ductile composites have proven the feasibility of this system (Yoon et al., 2002). As shown in Figures 1a and 1b, it was found that the highly ductile concrete located in hinge regions maintains its integrity beyond drifts of 20% while the unbonded post-tensioning does not yield and thus minimizes any residual displacement after reaching such high drifts. The load-carrying capacity of the system was also maintained at these large drifts. Residual drifts were on the order of 2-3%.

Experiments are now needed on full-height columns (roughly half-scale will be feasible with the proposed facility) due to the use of unbonded post-tensioning. Because the post-tensioning is unbonded, there is no localized yielding of the post-tensioned steel. Strain induced in the bridge columns is spread along the entire length of the post-tensioning. Therefore full bridge piers using the full length of strand are necessary for accurate investigation of such systems. Figure 1c illustrates the experimental concept. Full bridge pier experiments will require stroke capacities of 650-1000 mm. This facility will include a modular 1.8-m-high low reaction wall and a modular 7.2-m-high reaction wall. The experiments will be carried out on the upper surface of the low reaction wall and lateral load would be applied from the high portion of the strong wall. Equivalent gravity load would be applied using high capacity (force) actuators that would react off of the strong floor or off of a stiff frame supported at the top of the reaction walls. This support system will be designed to allow the actuators to move with the specimen as it undergoes large drift cycles.