UC San Diego

Continued advancements in high-speed computing and increased availability of earthquake strong motion data have been allowing for further progress in the area of soil-structure-interaction (SSI). Efforts in this dissertation are mainly concerned with three-dimensional (3D) computational analyses of pile foundations and bridge -foundation-ground systems. This includes Finite Element (FE) modeling of ground-pile foundation systems, documentation and assessment of recorded bridge strong motion data, and identification of dynamic bridge- foundation system characteristics. Currently, simplified approaches, such as p-y curves or the foundation stiffness matrix representation, are employed mainly when considering Soil-Structure-Interaction. However, there is much interest in more representative modeling techniques in order to improve our assessments of seismic pile foundation response. In an effort to address this challenge, 3D FE numerical investigations are conducted related to the response of piles and pile groups under lateral load. Distribution of loads and moments among the piles within the group is investigated. Effects of permeability and loading rate on lateral pile response are addressed for saturated relatively impervious cohesionless soil condition. Insights concerning the soil-pile interaction mechanisms are obtained based on the conducted analyses of the soil-pile foundation subsystems. Furthermore, numerical studies are conducted of long-span highway bridge-foundation systems under seismic loading conditions. Three-dimensional FE models of two existing bridges at Eureka California (the Samoa Channel Bridge and the Eureka Channel Bridge) are developed. Methodologies combining numerical modeling with insights gained from strong motion sensor records are investigated to capture the essential structure-foundation-ground system-response mechanisms. Focus is placed on the evaluation of dynamic properties and validation of the bridge FE models based on the recorded earthquake response. An optimization tool (SNOPT) is employed to evaluate the bridge foundation lateral stiffness. The studies show that computational modeling, along with analysis of the recorded ground-pile foundation data, provide an effective mechanism for understanding the entire structure-foundation-ground system response. The OpenSees platform and the user- interfaces OpenSeesPL, MSBridge, as well as SNOPT are employed in various sections of the study. In the domain of highly expensive and time consuming foundation design and/or retrofit, major beneficial outcomes can result from adoption of analysis tools which have been calibrated/ verified by actual recorded seismic performance data sets