Date of Award

Spring 1-1-2019

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

First Advisor

Shideh Dashti

Second Advisor

Abbie Liel

Third Advisor

Richard Regueiro

Fourth Advisor

Pedro Arduino

Fifth Advisor

Guido Camata

Abstract

A series of dynamic centrifuge experiments involving soil-foundation-mitigation-structure (SFMS) interaction on liquefiable soils are used to evaluate the ability of state-of-the-art numerical modeling in reliably predicting system performance (e.g., in terms of foundation settlement, rotation, roof accelerations, and column strains). This research comprises of two major components: (1) calibration and validation of two soil constitutive models with a series of fully drained and undrained, monotonic and cyclic, triaxial tests as well as a free-field centrifuge test on a layered, liquefiable soil deposit (without a building or a mitigation technique); and (2) numerical simulation of centrifuge tests that include multi-degree-of-freedom, shallow-founded, inelastic structures on mitigated and unmitigated liquefiable soil profiles, investigating the strengths and shortcomings of the available continuum numerical tools in capturing the key engineering demand parameters of interest.

Solid-fluid, fully-coupled, nonlinear, effective stress, 3D, finite element analyses are performed in parallel OpenSEES finite element platform with two different soil constitutive models. Overall, the numerical predictions agreed well with the experimental measurements in terms of acceleration, pore pressure, and settlement under structure in the presence of a relatively thin liquefiable layer or a mitigation strategy (when the extent of soil softening and volumetric strains due to sedimentation was reduced). However, these numerical simulations (and a continuum framework in general) could not capture volumetric strains due to sedimentation. This shortcoming became critical in thicker, unmitigated liquefiable layers and in the far-field, leading to a notable underestimation of total settlements. The cumulative foundation rotation or tilt was also generally underestimated numerically, regardless of the constitutive model. Improvements are needed to better capture numerically the accumulation of localized shear and volumetric strains below the foundation’s edges. The dynamic response of the soil-foundation-structure system and the required analysis time were highly sensitive to the choice of small-strain soil damping, soil constitutive model, and element type. They were also sensitive to the characteristics of the soil-foundation interface, but to a lesser extent. The results presented in this dissertation aim to guide future numerical modelers in simulating the seismic response of highly nonlinear SFMS systems and in better understanding the modeling sensitivities and uncertainties.

Share

COinS