Date of Award

Spring 1-1-2016

Document Type


Degree Name

Doctor of Philosophy (PhD)

First Advisor

Shideh Dashti

Second Advisor

John S. McCartney

Third Advisor

Dobroslav Znidarcic

Fourth Advisor

Tad Pfeffer

Fifth Advisor

Scott Brandenberg


The majority of stiff-unyielding underground structures have performed well in recent earthquakes. The few cases of failure are mainly related to construction in poor ground conditions (e.g., soft fill, liquefiable soils, sloping ground) and inadequate or no seismic design. Currently there is disagreement among engineering professionals and design methodologies on seismic forces and deformations experienced by these structures, sometimes resulting in inaccurate design that maybe un-conservative or over-conservative. The current procedures used to evaluate seismic loading and deformations of these structures rely on over-simplified analytical methods or advance numerical tools that have not been validated against physical model tests. Previous analytical and numerical studies have identified key factors such as structural flexibility, base fixity, and wavelength as being important in the seismic response of these structures. However the commonly used simplified methods don’t take into account these important factors.

Further, there is a lack of well-documented case studies and experimental research systematically evaluating the seismic response of stiff-unyielding buried structures that are restrained against excessive deformations at their base and roof. Therefore, a centrifuge experimental study of the seismic response of these structures was performed by varying the key parameters and monitoring seismic lateral earth pressures, racking displacements, bending moments, and soil-structure interaction. The results of these experiments provide a better understanding of the underlying mechanisms of soil-structure interaction near these structures and their overall performance. The results also serve as a database for other researchers to calibrate their numerical tools.

A series of sixteen centrifuge model experiments representing 11 to 12 m-high reinforced concrete underground reservoir structures were conducted. The structure stiffness, backfill soil type and slope, embedment, container type (rigid versus flexible boundaries), fixity conditions, and ground motion characteristics were varied to evaluate their influence and relative importance on structural performance. The structures were buried in dry medium-dense sand at 60% relative density and compacted, site-specific silty sand with different backfill slopes. A suit of earthquake ground motions and sinusoidal motions with different frequencies were applied to the buried models. The application of sinusoidal motions in particular allowed for a comprehensive study of the influence of loading frequency in relation to the fundamental frequency of the site and structure.


Advisor: Craig Davis