Work

Numerical Modelling of Cyclic Degradation of Natural Clay

Public Deposited

While infrequent, clay slope failures caused by earthquakes result in loss of life and substantial property damage. In strong seismicity areas, it is sometimes assumed that the stability of clay deposits can be evaluated through the residual undrained strength that is applicable at large deformations. A crucial factor missing in this conservative assumption is the understanding and quantification of the loss of natural clay shear strength during the seismic shaking. Consequently, the above assumption can neither address the fundamental issue of the mechanism of the slope failure initiation nor explicitly account for the specific parameters of individual earthquakes (e.g., amplitude and duration). The goal of this work is to study and quantify the strength degradation of natural clay subjected to cyclic loading. The assumption at the core of this thesis is that the strength loss of natural clay is related to the deterioration of its inherent structures (e.g., inter-particle bonds and fabric) caused by plastic deformation that develops during cyclic loading. Furthermore, such a strength reduction also results from the change of effective stress state as a result of the accumulation of excess pore pressure. A particular case studied in this work is the destructive landslides caused by the 1964 Alaska earthquake, one of the largest earthquakes in history. The strength loss of the Bootlegger Cove Formation (BCF) clay during cyclic loading has been recognized as a critical factor in the initiation of these landslides. The largest slide during this event was located at Turnagain Heights in Anchorage. Based on the in-situ tests at a site adjacent to the scarp of this slide, this work evaluates the BCF's in-situ sensitivity, an index of natural clay's susceptibility to structure degradation. Furthermore, based on the undrained strength of BCF clay interpreted from in-situ tests, a series of stability analysis are conducted, which back-calculates an upper bound on the strength reduction of BCF clay needed to initiate the slope failure. It is found that the computed strength reduction is compatible with the sensitivity of the BCF at the same elevation as the failure zone within the BCF. The back analysis can estimate the strength loss to initiate slope failure, but to quantify the strength reduction as a function of a specific seismic event, a general and more sophisticated method is needed. In this work, a bounding surface plasticity constitutive model is developed that accounts for the degradation of clay inherent structure as well as the change of effective stress state, i.e., the two major factors affecting material strength degradation. The model is developed in two steps. A basic model is proposed to represent the cyclic behavior of reconstituted clay, i.e., the intrinsic behavior. Compared with existing plasticity models for cyclic clay behavior, three major enhancements are proposed, including the mixed plastic flow rule, a new form of plastic modulus to uniformly reproduce cyclic softening and shakedown, and the adoption of a small strain elasticity model. The verification of the basic model with experimental observations shows that the aforementioned improvements ensure that the development of plastic deformation and excess pore pressure during cyclic loading is reasonably represented. To account for the effects of soil structure and its deterioration, the basic model is extended to include a new internal variable that represents the amount of soil structure and a destructuration law that quantifies the monotonic decrease of such a variable under irrecoverable deformation. The proposed plastic potential surface in the extended model is a function of fabric anisotropy and inter-particle bonds. The influences of these two aspects of soil structure on material plastic flow are explored based on the stress-dilatancy relation which is derived from the plastic potential surface. The appropriateness of this plastic potential surface to describe natural clay behavior is validated with experimental evidence. The validation of the extended model based on experimental data of seven natural clays shows that the proposed model is capable of reproducing the mechanical behavior of natural clay under monotonic and cyclic loading, and the strength degradation during cyclic loading can be reasonably quantified by this model.

Last modified
  • 02/20/2018
Creator
DOI
Subject
Keyword
Date created
Resource type
Rights statement

Relationships

Items