SCEC Award Number 14028 View PDF
Proposal Category Collaborative Proposal (Integration and Theory)
Proposal Title Validating site response predictions in deterministic physics-å_based ground motion simulations
Investigator(s)
Name Organization
Dominic Assimaki Georgia Institute of Technology Ricardo Taborda University of Memphis
Other Participants Jian Shi, Georgia Institute of Technology; Graduate student assistant, University of Memphis
SCEC Priorities 6c, 6e, 6b SCEC Groups EEII, GMP, GMSV
Report Due Date 03/15/2015 Date Report Submitted N/A
Project Abstract
 In the first year of the project, we focused on the development of a new constitutive model that can predict realistic simple shear ground deformations over the entire strain range, from small strains to failure. Most geotechnical engineering site response models have been developed using laboratory experiments of cyclic material loading in the low to medium strain range (<1%), without imposed constraints on the shear strength of the material. As a result, when these experimental data are synthesized into stress-strain constitutive relations, the material response at higher strains (~5% to failure) --and therefore the predicted ground deformation-- is not reliable. On the other hand, elastic perfectly plastic (EPP) models have been developed with explicit consideration of capturing the material strength, but perform very poorly at lower strains, and overestimate the material hysteretic damping during cyclic loading. Our new hybrid constitute model bridges the gap between EPP models in solid mechanics and cyclic soil models in geotechnical engineering, and is capable of predicting both stiffness degradation and failure. In the second year of the project, we will use this model to 'correct' simulated ground motions at sites with known velocity profiles, and examine the extent to which predictions improve by explicitly accounting for site-specific response. Our long term plan is to extend this model to 3D.
Intellectual Merit To enable physics-based risk assessment of infrastructure systems, ground motion models should have capabilities to predict realistic ground deformations. In the first year of the project, we focused on the development of a new constitutive model that can predict realistic simple shear ground deformations over the entire strain range, from small strains to failure. lifelines and critical facilities. Thus, if ground motion models are to be used in physics-based risk assessment of these systems, they should have capabilities to predict nonlinear site effects and realistic ground deformations. In the first year of the project, we focused on the development of a new constitutive model that can predict realistic simple shear ground deformations over the entire strain range, from small strains to failure. Most geotechnical engineering site response models have been developed using laboratory experiments of cyclic material loading in the low to medium strain range (<1\% $), without imposed constraints on the shear strength of the material. As a result, when these experimental data are synthesized into stress-strain constitutive relations, the material response at higher strains (~5\% to failure) --and therefore the predicted ground deformation-- is not reliable. On the other hand, elastic perfectly plastic (EPP) models have been developed with explicit consideration of capturing the material strength, but perform very poorly at lower strains, and overestimate the material hysteretic damping during cyclic loading. Our new hybrid constitute model bridges the gap between EPP models in solid mechanics and cyclic soil models in geotechnical engineering, and is capable of predicting both stiffness degradation and failure. Our long term plan is to extend this model to 3D and implement it in SCEC ground motion simulations.
Broader Impacts Physics-based earthquake simulations are nowadays producing ground
motion time-series for engineering design applications. Of
particular significance to engineers, however, are simulations of near-field
motions and large magnitude events for which observations are scarce. These events are important because they can cause large ground deformations and ground failure (liquefaction, lateral spreading), namely effects that frequently control the risk of infrastructure systems vital to societies, like lifelines and critical facilities. Broader impacts of our model include a transformation in the way engineers, planners, stakeholders and the public design and evaluate urban resilience. Fully implemented in a 3D physics-based ground motion model, the nonlinear soil model will advance our understanding of the impact of earthquakes on infrastructure systems, and will improve emergency preparedness that can save lives and minimize economic losses from future earthquakes.
Exemplary Figure Figure 2