Computational methods for simulation of extreme events on concrete structures

Necessary choices for an accurate simulation of the mechanical response of a reinforced concrete structure subjected to extreme loading are cataloged in this dissertation. Extreme loads are characterized by their rapid application to the structure, and are typically caused by blast or impact. The success of various constitutive models and discretization (finite element and meshfree) methods are evaluated by comparing simulation results with strictly controlled laboratory tests. Inadequacies stemming from the choice of analysis options are presented. Simulation results for system, member and specimen scales are compared to empirical data throughout the work. In particular, laboratory tests that are designed to quantify the dynamic strength increase of plain concrete are simulated. Several constitutive models are evaluated for their consistency to observations of concrete forced by extreme loads that cause fracture and fragmentation. Through these studies, a previously unidentified physical mechanism due to plastic response is proposed to explain the dynamic strength increase of plain concrete. Additionally, tests for static and dynamic behavior of the bond between steel and concrete are compared to experiments at laboratory and full scales. Nonlinear beam elements are embedded in the continuum to represent steel reinforcement. Pullout and large deformation capabilities are studied using a one-dimensional sliding interface. The novel mechanism proposed for the dynamic increase of fast compressive tests also applies to rapid bond pullout. Conservatism in the design process is improved by an understanding of the aforementioned dynamic processes. The lessons learned about computational models in this study may be applied to the economical prevention of rare, potentially catastrophic events that may occur during the service life of a structure. They serve as a basic guideline for developing efficient and accurate numerical models to simulate extreme event scenarios.