Dissertation/Thesis Abstract

Pile Dynamics and Shearing Behavior of Granular Soils
by Kuei, Kevin Charles, Ph.D., University of California, Davis, 2020, 305; 27668372
Abstract (Summary)

The characterization of strength whether a resistance against penetration, or a resistance against liquefaction is important in geotechnical engineering. This dissertation studies two topics which broadly fall into these topical domains.

‘Forward’ and ‘inverse’ methods are routinely used in the analysis of piles for forecasting drivability or indirectly estimating pile capacity. A significant aspect of the latter problem is the adequate separation of ‘static’ and ‘dynamic’ components to total driving resistance in a process referred to as ‘signal matching’. The first part of this dissertation largely explores the inverse problem, and the realm of pile dynamics and inverse methods for estimating static capacity from direct measurements (i.e. dynamic load testing and signal matching). The work is predicated primarily on the direct measurements made available from the Reusable Test Pile, a site characterization tool developed for prototype pile design, and the Instrumented Becker Penetration Test (iBPT), a sister system developed for liquefaction assessment of gravelly soils at UC Davis. A novel or unique aspect of the system(s) is the multiple measurements or observation points made available from sensors embedded at the pile base and along the length of the shaft. In practice, usually only one such observation point at the pile head is available. Consequently, much of the presented work goes to developing the framework to incorporate these measurements for the analyses, verification, and evaluation of potential improvements to existing procedures. To this end, one-dimensional methods of wave equation analyses using simplified models of soil response for end-bearing and shaft resistance are adopted. The work incrementally builds from an element-level analysis at the pile base, to incorporating shaft resistance at a systems-level using any number of observation points via a multistage signal matching approach. Predictions from dynamic measurements are then compared with those obtained from direct static measurements, or as logically constrained based on driving mechanics. Aspects related to the pile driving process, model verification at an element- and systems-level, performance of alternative signal matching approaches, role or contribution of additional measurements, and details of non-uniqueness and uncertainty are discussed.

The use of sand-based methods for the evaluation of gravelly soils, such as in liquefaction analysis, is common in geotechnical practice. However, this assumption is largely necessitated due to limited guidance for evaluating particle size effects, case histories, and the inherent challenges of sampling and testing coarse-grained deposits. Consequently, the second part of this dissertation transitions to examining the soil element scale and how the shearing-response arises from its granular characteristics, in particular the role of particle size effects on granular behavior. Recognizing that most case histories of liquefiable gravelly soils are for very widely-graded materials, and that most sand-based procedures are for relatively clean sands of more uniform dimension, the study of gravelly soils is constrained to one of changing particle gradation. The Discrete Element Method is subsequently leveraged to control for particle-shape effects and to garner micro-mechanical insights and their relation to the macroscopic scale. Representative volume elements of differing polydispersity, and initial states are sheared under drained and undrained monotonic triaxial compression to understand fundamental differences in fabric evolution and mobilization of strength. The work builds incrementally from establishing the modeling framework, and investigating differences in fabric between granular assemblies of differing polydispersity and state, to a more application-based approach with revised stress-dilatancy relations accounting for dependence on gradation. Finally, the work concludes with an examination of the influence of ‘state’ definition on the interpretation of grading-dependent behavior.

Indexing (document details)
Advisor: DeJong, Jason T.
Commitee: Boulanger, Ross W., Martinez, Alejandro
School: University of California, Davis
Department: Civil and Environmental Engineering
School Location: United States -- California
Source: DAI-B 82/1(E), Dissertation Abstracts International
Subjects: Civil engineering
Keywords: Pile dynamics, Granular soils, Shearing behavior
Publication Number: 27668372
ISBN: 9798662476868
Copyright © 2020 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy