UC Berkeley

Bridges often serve as key links in the local and national transportation networks. Any bridge closures can have severe costs not only in the form of repair or replacement, but also in the form of economic losses related to medium- and long-term interruption of businesses and disruption to the communities. In addition, continuous functionality of bridges is very important after any seismic event for emergency response and recovery purposes. Considering the importance of these structures, the associated structural design philosophy is shifting from collapse prevention to maintaining functionality in the aftermath of moderate to strong earthquakes (Resiliency). Moreover, the associated construction philosophy is being modernized with the utilization of accelerated bridge construction (ABC) techniques to reduce impacts of construction work on traffic, society, economy and on-site safety during construction. This dissertation presents two bridge systems targeting the aforementioned issues. A combined numerical and experimental research was undertaken to characterize the seismic performance of these bridge systems.

The first part of the study focuses on the structural system level response of highway bridges that incorporate a class of innovative connecting devices called the “V-connector”, which can be used to connect two components in a structural system, e.g., the column and the bridge deck, or the column and its foundation. This device, designed by ACII, Inc., results in an isolation surface at the connection plane via a connector rod placed in a V-shaped tube embedded into the concrete. Energy dissipation is provided by friction between a special washer located around the V-shaped tube and a top plate. Because of the period elongation due to the isolation layer and the limited amount of force transferred by the relatively flexible connector rod, bridge columns are protected from experiencing damage, leading to an improved seismic behavior. The V-connector system also facilitates the ABC by allowing on-site assembly of prefabricated structural parts including those of the V-connector.

A single-column, two-span highway bridge located in Northern California was used for the proof-of-concept of the proposed V-connector protective system. The V-connector was designed to result in an elastic bridge response based on nonlinear dynamic analyses of the bridge model with the V-connector. Accordingly, a 1/3-scale V-connector was fabricated based on a set of selected design parameters. A quasi-static cyclic test was first conducted to characterize the force-displacement relationship of the V-connector, followed by a hybrid simulation (HS) test in the longitudinal direction of the bridge to verify the intended linear elastic response of the bridge system. In the HS test, all bridge components were analytically modeled except for the V-connector, which was simulated as the experimental substructure in a specially designed and constructed test setup. Linear elastic bridge response was confirmed according to the HS results. The response of the bridge with the V-connector was compared against that of the as-built bridge without the V-connector, which experienced significant column damage. These results justified the effectiveness of this innovative device.

The second part of the study presents the HS test conducted on a 1/3-scale two-column bridge bent with self-centering columns (or broadly defined as “resilient columns” in this study) to reduce (or ultimately eliminate) any residual drifts. The comparison of the HS test with a previously conducted shaking table test on an identical bridge bent is one of the highlights of this study. The concept of resiliency was incorporated in the design of the bridge bent columns characterized by a well-balanced combination of self-centering, rocking and energy dissipating mechanisms. This combination is expected to lead to minimum damage and low levels of residual drifts. The ABC is achieved by utilizing precast columns and end members (cap beam and foundation) through an innovative socket connection. In order to conduct the HS test, a new hybrid simulation system (HSS) was developed, utilizing commonly available software and hardware components in most structural laboratories, namely, a computational platform using Matlab/Simulink (MathWorks, 2015), an interface hardware/software platform dSPACE (2017), and MTS controllers and data acquisition (DAQ) system for the utilized actuators and sensors. The operation of the HSS was verified using a trial run without the test specimen.

In the conducted HS test, the two-column bridge bent was simulated as the experimental substructure while modeling the horizontal and vertical inertia masses and the corresponding mass proportional damping in the computer. The same ground motions from the shaking table test, consisting of one horizontal component and the vertical component, were applied as input excitations to the equations of motion in the HS. Good matching was obtained between the shaking table and the HS test results, demonstrating the appropriateness of the defined governing equations of motion and the employed damping model, in addition to the reliability of the developed HSS with minimum simulation errors. The small residual drifts and the minimum level of structural damage at large peak drift levels demonstrated the superior seismic response of the innovative design of the bridge bent with self-centering columns. The reliability of the developed HS approach was the motivation of a follow-up HS study focusing on the transverse direction of the bridge where the entire two-span bridge deck and its abutments represented the computational substructure while the two-column bridge bent was the physical substructure. This investigation was effectively utilized to shed more light on the system level performance of the entire bridge system with innovative bridge bent design beyond what can be achieved via shaking table tests, which are usually limited by large-scale bridge system testing capacities.