Search

The majority of steel structures in the U.S. follow a typical design practice. This practice consists of having relatively few frames of the building that resists lateral loads, and the remainder of the structure is designed to resist only gravity loads. The elements of this gravity system are connected using “simple shear” connections that are designed and assumed to behave as “perfect pins” that have no flexural strength or stiffness, thus generating no moment resistance at the connection. Nevertheless, past research has shown that the gravity system contributes to the lateral resistance of a building in non-negligible quantities. The main objective of this research is to perform preliminary analyses to support the design of a test specimen that will be used to evaluate the cyclic response of a multi-bay steel gravity framing test specimen subjected to lateral loads. This thesis explained and delineated the test program and concept that will be used in the large-scale system-level testing. Tests will be administered on variable-scale assemblies of two-by-three bays with dimensions of 16’x8’ and two half-stories of 6’. Three servo-controlled hydraulic actuators will be used to impose a displacement on the upper floor of the specimen. The first test specimen will evaluate the response with strong-axis decking metal deck (metal deck ribs oriented parallel to the load direction), and the presence and absence of seams at the column lines. Design and sizing of the specimen elements were done trying to have a continuation of previous research (Donahue, 2019) and based on the current practices in construction (Professional Advisory Committee). Several pushover analyses were done on SAP2000 (Computers and Structures, 2020), to a structural model of the test specimen, using several constructed moment-rotation spring models of hinges at the joints (to represent the simple shear connections). These moment-rotation spring models were developed and validated with the results of previous research (Donahue, 2019) and they accurately characterized the various detailing variations present in the previous studies. The results of these analyses showed the loads and displacements needed to “fail” the specimen, how each variation of the specimen could affect these loads and displacements, and the demands for designing critical elements like base beams, column support clevises, and the top truss diaphragm.

Chloride-induced corrosion of prestressed concrete structures is a complex phenomenon that has wide-ranging effects on structural reliability and potentially dangerous premature failures. Marine environments are ideal locations for corrosion to affect structures due to sea spray exposure and wetting and drying cycles providing plentiful access to chloride ions from the salt water. Many vital pieces of infrastructure such as wharves and seaports are located in corrosion-prone marine environments and many of these structures use prestressed concrete heavily in their construction. These structures also typically have the added complexities of submerged substructure elements that have highly variable and spatially dependent exposure conditions.Piles are substructure elements that play a particularly critical role in determining structural capacity of a wharves in the face of large hazard loads such as earthquakes, forceful tides and waves, or impact by ships in the event of an accident. These elements also undergo constant exposure to sea water (and thus chloride ions). The combination of these facts has made it crucial to fully understand the corrosion degradation characteristics of these structures so that capacity and reliability can be quantified in the best way possible. However, the variability in both the environmental and structural parameters that determine corrosion conditions makes their prediction and incorporation into structural analysis very difficult.The goal of this study was to provide accurate and realistic analysis of a prestressed concrete wharf pile subjected to corrosion resulting from severe chloride exposure conditions. This was accomplished in two major steps. First, an integrated probabilistic modeling framework that accurately simulated corrosion effects in prestressed concrete piles in a marine environment was established. This framework relied on models derived from empirical data or mathematical formulas representing the corrosion process. Next, a highly detailed finite element modeling framework using custom-built material models was established to build the pile structure and implement the corrosion deterioration characteristics at multiple time-steps before static pushover analyses were performed.Analysis results showed pile response under pristine conditions as being governed by bond “slip” failure between the concrete and tendons. As tendon corrosion progressed at subsequent time steps, various simulated failure methods caused a decrease in the overall composite strength of the pile in addition to a loss of ductility. This was demonstrated by the overall reduction in the 50th percentile force and displacement capacities, which fell 21.0% and 44.1%, respectively, at the final time step of 75 years. The spread of the peak force capacities is distributed over a generally wider range of displacements at 75 years than previous time steps, indicating a greater degree of uncertainty which must be taken into account. Also, multiple alternate failure criteria and force redistribution in deteriorated elements caused sudden failures that were important to account for. The impact of these failures can be noted by the reduction of the mean force and displacement capacities, which fell 34.3% and 66.7%, respectively, at 75 years. In addition, about 50% of simulations had reached a sudden failure due to tendon breaking or other material failure at 75 years.

In 1995, a framework for a reliability-based, performance-based seismic design procedure was proposed by Collins. The basic idea behind the procedure was to develop an equivalent single-degree-of-freedom (SDOF) model of the structure and then to use this model and uniform hazard spectra to predict maximum displacements and drifts in the multi-degree-of-freedom (MDOF) structure. Deterministic design-checking equations based on the equivalent SDOF system were developed, and design factors (analogous to load and resistance factors) were calibrated and used in design-checking equations to account for the uncertainty in the equivalent SDOF model, the seismic environment, and site soil conditions. The parameters of the SDOF model were obtained using information from two-dimensional static pushover analyses of the lateral force resisting systems of the structure. The objective of this research was to enhance or improve procedure by looking at alternative ways in which pushover analyses could be used. Specifically, this thesis looks at the following two topics: (1) The choice of the lateral force distribution is one of the most important aspects of such analyses since any single lateral force pattern may not capture all deformation characteristics. This is especially true when nonlinear behavior of the structure is considered. In the original procedure, only one force distribution was considered. This dissertation summarizes the results of a study to evaluate several possible force distributions. (2) The design procedure proposed by Collins was based on static pushover analyses of a two-dimensional model of a building. Such analyses are acceptable if buildings are regular and symmetric or if torsional effects can be otherwise accounted for. In this dissertation, the design procedure is extended to three-dimensional cases to explicitly account for the torsional behavior of asymmetric building structures.