Recent damage following earthquakes in Alaska has exposed a potential issue regarding external abutment shear keys. The keys, referred to as ���hider walls��� are meant to serve as fuses that remain intact for smaller earthquakes, but break way in larger earthquakes thus limiting the force that can be transmitted to the abutments. In that manner, the abutments are ���capacity protected elements��� whose force input is limited by the shear key capacity. For the concept to work, the following three criteria must be met: (1) The demand on the shear keys for the seismic event for which they are expected to remain intact must be reliably predicted. (2) The capacity of the shear keys must be reliably predicted; (3) The shear keys should fail in a manner that is easy to repair following on an earthquake. Through a series of experiments on existing and proposed shear key designs, development of analysis methods for the prediction of shear key strength, and parametric studies on abutment shear key demand, this research will provide engineers with the tools needed to achieve designs that satisfy the intended performance objective.

This research will investigate the seismic behavior of a variety of bridge structures containing high strength reinforcement and other unique details. The research will primarily utilize shake table testing to investigate the dynamic behavior of bridge systems. The test units will be subjected to sequences of historical earthquake ground motions, and response measured through a variety of instrumentation systems. Computational modelling will also be conducted affirm or revise prior design recommendations

This proposal addresses the impact of response spectra definitions on the design of multi-span bridges by conducting computational research to develop recommendations for hazard levels that are founded on non-linear bi-directional dynamic analysis. The research proposal also aims to develop simplifications to the direct displacement-based design approach for multi-span bridges that will facilitate implementation into bridge design practice. This will be accomplished by designing a series of bridge configurations to multiple hazard definitions. The resulting designs are then subjected to bi-directional non-linear dynamic analysis to assess the range of possible deformation demands that are then correlated to the assumptions made in selection of the seismic hazard for each design. By considering the instances that result in deformation demands exceeding the target for each hazard definitions across all rotation angles, it will be possible to select a seismic hazard definition that provides a known probability of exceedance for the design limit state. The data generated from the computational models will also facilitate simplifications to the direct displacement based design approach. The simplifications aim to reduce the effort to define the equivalent viscous damping, target displaced shape (and corresponding system displacement), and strength allocation to abutments vs. columns.

Alaska is subjected to the highest seismic hazards within the US. The state is also home to some of the most extreme climate in the country. Past studies have shown that low temperatures alter the behavior of reinforced concrete structures sufficiently that it must be considered in design to ensure the safety of the traveling public. Over the last several years, high strength steel reinforcement has become common in the market, with strengths as much as 50% higher than that typically deployed for seismic applications. It is well established that higher strength is not always desirable, especially if it results in a loss of energy dissipation capacity. Previous research has shown that high strength steel, even when meeting the requirements of appropriate ASTM designations, has reduced ability to deform compared to typical ���������������seismic steel������������������. Through the use of material and structure level tests at low temperatures, as well as computational modelling, the impact of low temperatures on high strength steel is explored in this research project. The concern stems from the well established impact of reduced fracture toughness at low temperatures, and the recent observation that the stress concentrations at the base of reinforcing bar ribs impacts the strain capacity of the bars (and hence ductility of the reinforced concrete member). This research aims to determine if low temperatures further impacts the performance of columns reinforced with high strength steel.

This proposal aims to develop a rapid seismic bridge assessment method that can be used for planning (via scenarios), and for post-earthquake assessment (inspection prioritization). Unlike existing methods which are largely probabilistic, and focused on high level assessment, the proposed methodology is sufficiently versatile that it can provide a range of information, spanning from deterministic bridge specific performance, to broader assessments of bridge vulnerability. The procedure relies upon the Direct Displacement-Based Design approach as the analysis engine, and has three components: (1) Bridge metadata; (2) Bridge limit state parameters; and (3) Seismic Hazard characterization. Given any 2 of the above, the third may be determined. While the approach will function with very course data (i.e. basic metadata such as span lengths, column diameter, and height; basic limit state parameters, such as limit state displacement; and course hazard definition (such as a 3 point code-based spectra), with more detailed information, the fidelity of the outcome increases significantly. The work described in this proposal will define limit state parameters for Alaska bridges and characterize the seismic hazard. The framework for the rapid assessment approach will also be developed and applied to a series of bridges. The final outcome will be detailed plans for development of a rapid assessment application which would be developed in a future phase of the research.

This proposal addresses the impact of bridge condition on behavior within the context of performance-based seismic design. Bridge engineers design structures assuming that their properties on day one remain constant throughout the life of the bridge. However, due to material degradation and lifetime transient loadings (including those from small earthquakes), the response of a structure to an extreme event such as an earthquake at some point in the future may be different from what the engineer calculates during the design phase. The research described in this proposal will assess the sensitivity of condition-dependent bridge response to inputs described below. There are five components to this research question. The interactions between these components are shown in Fig. 1 and described further to set the stage for the research that will be conducted in this multi-phase program.

Based upon a seismic analysis performed in 2013, the Anchorage Port Access bridge (Bridge number 0455) requires a seismic retrofit to safely accommodate the design earthquake. The bridge, shown in Figure 1, has several deficiencies including steel column-to-cap beam connections that were determined to perform poorly under seismic loading during the first phase of this study. The objective of this study is to develop retrofit recommendations through large scale testing, modelling, and analysis. During phase one of the work, a global computational model (using lumped plasticity frame elements) was developed to assess bridge deformation demands. As part of that work, hysteretic response of the components was represented through consideration of results of a local FEM model and experimental tests. Also included in that model was the impact of damping model choices, soil-structure interaction, and ground motion directionality. The local FEM model considered the details of the individual connections to assess stress concentrations, and was calibrated through the experimental tests. Lastly, the experiments that were conducted included half scale models of two representative bridge bents that were subjected to reversed cyclic loading. Details of each of these components of the first phase of work are described next.

Alaska is subjected to the highest seismic hazards within the US. As a consequence, their structural systems must be robust, redundant, and ductile such that interruptions to their service following an earthquake are minimal. Research over the last 20 years has led to improved design details that have been proven to work effectively by experimental, analytical, and field studies. However, much of Alaska is in a harsh environment with a comparatively short construction season. As a consequence, a premium is placed on rapid construction. Previous research has resulted in the development of an all-steel bridge system that can be easily constructed for temporary (or permanent) installations. That system has inspired the possibility of developing an alternative rapidly constructed bridge system for the more common bridge types deployed in Alaska ������������������ namely, Reinforced Concrete Filled Steel Tubes (RCFST) and Reinforced Concrete (RC) Column structures. These structural types are valued for their proven ductile performance. It is the goal of this research project to speed up the construction process, while retaining (or improving upon) their seismic behavior. To that end, this research aims to develop Accelerated Bridge Construction connections for RCFST and RC bridge systems that use ���������������external socket������������������ or ���������������external pocket������������������ connections. This is distinctly different from existing ���������������socket������������������ or ���������������pocket������������������ connections that are internal to the cap and can compromise seismic behavior. Lessons learned from the development of the steel bridge system (termed the ���������������Grouted Shear Stud (GSS) Connection) will be valuable as the connections described in this proposal are developed. The research will consist of large scale experimental seismic tests on candidate connections coupled with advanced (and simple) computational models for their analysis and design. The end product will be a suite of connection types that are supported by experimental and analytical evidence that will lead to a simple design approach for their deployment in practice.

Bridge girder cross-sections continue to become regional in nature, with many state DOTs adopting their own unique sections at either the state or regional level. Typically, girders are developed without consideration of a formal ���������������optimization������������������ of cross-section shape, or when any optimization was employed, the process of optimization and hence its outcome posed several limitations. For example, in many cases, the optimization focused on a single girder without considering any deck on it, whereas the lateral spacing of girders and thickness of the overhead deck are design variables which should be considered while optimizing the girder. Also, such optimization was often based on local search algorithms that do not guarantee global optimality, especially when the solution space is multi-dimensional and highly nonlinear. In most instances, optimization process only included ���������������quantifiable������������������ factors like material cost, volume or weight, labor cost, and formwork cost, etc. But solutions that are mathematically optimal with respect to the quantified factors are not necessarily and readily acceptable when considering non-quantifiable factors and preferences pertaining to practical and field implementation issues. Hence, it is important to extend the optimization procedure to enable outcomes from a formal optimization to be integrated with important subjective considerations. The objective of this research is to develop and apply contemporary meta-heuristic global search procedures for optimizing pre-tensioned decked bulb-tee girders for systematically identifying new optimized cross-section shapes. It is envisioned that, for a specific girder span length and a number of lateral girders, several maximally different alternative cross-sections with competitive structural and cost performance will be first identified; this will be repeated for different combinations of girder spans and numbers of girders to analyze and develop structural and cost performance characteristics variation with girder span length. Then in consultation with AKDOT and precast manufacturers, the alternative optimized cross-section shapes will be screened and fine-tuned based on practical considerations to identify a small set of ���������������best feasible������������������ cross-section shapes. The cross sections to be explored here will be compared for structural performance and material savings against existing ���������������optimized������������������ sections as well as various legacy sections used by Alaska DOT (existing decked-bulb-tee section), and sections employed elsewhere around the US (i.e., AASHTO girders, PCI Bulb-Tees). Preliminary exploratory analysis will be conducted to study the effects of optimized cross-section shapes on extending the girder spans and reducing the number of spans, and therefore the approximate (empirically estimated) net life-cycle cost savings of the bridge system considering the number of piers, foundations, abutments, etc. We expect the outcomes of this project will potentially benefit the AKDOT in improving the ability to span longer distances, reducing overall bridge construction cost, and using resources more efficiently.

This research consists of an analytical and experimental study aimed at understanding the non-linear behavior of bridge columns constructed from ASTM A706 Grade 80 reinforcing steel. Specifically, of particular interest are (1) Plastic hinge lengths (and spread of plasticity) as well as bond slip and development; (2) Reinforcing bar strain limit states such as onset of transverse reinforcement yield, onset of bar buckling, and tensile fracture; and (3) Hysteretic energy dissipation. To address these points we will develop a computational fiber-based model for analysis of columns, and conduct a series of sixteen large scale reversed cyclic tests on bridge columns. Direct comparisons with columns constructed from Grade 60 steel will be possible. This work follows upon a Phase 1 study at NC State for Caltrans that included 4 column tests. In addition to large scale tests, material level characterization will also be conducted as part of this work to assess the impacts of alloy composition on ductility.