Date of Award


Embargo Period


Degree Type


Degree Name

Doctor of Philosophy (PhD)


Civil and Environmental Engineering


Jacobo Bielak

Second Advisor

David O'Hallaron

Third Advisor

Amit Acharya

Fourth Advisor

Julio Lopez

Fifth Advisor

Enrique Bazan


The main objective of this thesis is to build a framework for performing earthquake simulations capable of including nonlinear soil behavior and the presence of the built environment in highly heterogeneous basins, and to study their influence on the final response of the ground in large urban areas exposed to high seismic hazard. To this end, we use a finite elements approach and extend the capabilities of Hercules, the parallel octree-based earthquake simulator developed by the Quake Group at Carnegie Mellon.

Nonlinear soil behavior is incorporated in ground motion simulations employing a rate-dependent plasticity approach to predict the nonlinear state of the material explicitly at every time step. The soil is modeled as a perfectly elastoplastic material. The presence of urban structures is modeled representing buildings as homogeneous blocks made up of the same type of hexahedral elements used in the mesh. These elements are generated automatically through a new set of application programming interfaces, which extend Hercules' meshing capabilities while preserving its core octree-based formulation.

Both implementations are tested under realistic earthquake conditions in heterogeneous geological structures. In the case of nonlinear ground motion modeling, results indicate that soil nonlinearities greatly modify the ground response, confirming previous observations for deamplification effects and spatial variability, and evidencing three-dimensional basin effects not fully observed before. In turn, the presence of building clusters causes multiple soil-structure interaction phenomena that change both the near ground-motion and the individual performance of the buildings themselves. This substantiates the argument that in soft-soil basins may it not be longer valid to ignore the presence of neighboring structures.

These new implementations represent important advances in computational seismology and help make a direct connection to subjects of interest in earthquake engineering. The analysis drawn from the applications presented here confirms aspects known from previous, though limited studies, and broadens our knowledge of the effects of nonlinear soil and the built environment on the ground motion due to earthquakes at a regional level not explored before.