THESIS
2017
xix, 231 pages : illustrations ; 30 cm
Abstract
Anisotropic stiffness of soil is very important for predicting ground movements and the
serviceability limit state of geotechnical structures. Although stiffness anisotropy of various
sedimentary clays is recognized, anisotropic stiffness of unsaturated loess (an Aeolian deposit)
is rarely studied. As soil structure and suction play critical roles in the behaviour of unsaturated
loess, the effects of hydro-mechanical loading on anisotropic small-strain stiffness of
compacted and intact loess should be investigated.
The principal objective of this research is to examine evolving anisotropic stiffness of
unsaturated compacted and intact loess. A suction-controlled triaxial apparatus, equipped with
bender elements and hall-effects sensors, was used. To investigate the anisotropic s...[
Read more ]
Anisotropic stiffness of soil is very important for predicting ground movements and the
serviceability limit state of geotechnical structures. Although stiffness anisotropy of various
sedimentary clays is recognized, anisotropic stiffness of unsaturated loess (an Aeolian deposit)
is rarely studied. As soil structure and suction play critical roles in the behaviour of unsaturated
loess, the effects of hydro-mechanical loading on anisotropic small-strain stiffness of
compacted and intact loess should be investigated.
The principal objective of this research is to examine evolving anisotropic stiffness of
unsaturated compacted and intact loess. A suction-controlled triaxial apparatus, equipped with
bender elements and hall-effects sensors, was used. To investigate the anisotropic stiffness at
very small strains (below 0.001%), wetting-drying tests and isotropic compression tests were
carried out. Furthermore, constant suction and constant-? shear tests were conducted, to study
the small-strain (between 0.001% and 1%) stiffness.
It is found that, different from other saturated sedimentary clays, ?
ℎℎ (corresponding to the
horizontal shear plane) of unsaturated intact loess is less than ?
ℎ? (corresponding to the vertical
shear plane) at very small strains. This is mainly because horizontal layering is absence in the
intact loess whereas it is often found in other sedimentary clays. This stiffness anisotropy is
found evolving significantly with a change in suction. Similar trend was observed in the
compacted loess, implying that suction can induce stiffness anisotropy even at isotropic stress
state.
During shearing at small strains, intact loess exhibits stiffer response than the compacted loess
at deviatoric strain below 0·025%, mainly because more clay aggregates accumulate at the
inter-grain contacts in the intact loess. Contrary to the typical behaviour of compacted soils,
small-strain stiffness of the compacted loess increased with a decreasing suction. The increase
may be caused by significant wetting-induced collapse during the wetting process. In contrast,
the shear stiffness of the intact specimens remains unchanged, owing to slight swelling
observed during wetting. Comparing the stiffness degradation of loess to that obtained from
widely-used design chart proposed by Vucetic & Dobry (1991), the design chart is found
unconservative for the stiffness degradation of both compacted and intact loess at small strains.
Furthermore, a bubble model for unsaturated soil is developed within the framework of
kinematic hardening and bounding surface plasticity. To model effects of recent suction
history, an elliptical elastic bubble is defined inside a modified Cam-clay bounding surface.
The elastic bubble is considered as a function of suction, degree of saturation and plastic
volumetric strain. Translation of the elastic bubble is governed by suction, degree of saturation
and stress increments. Moreover, hardening modulus depends on not only stress and void ratio
but also suction, degree of saturation and distance between the elastic bubble and the bounding
surface. The proposed model is evaluated using suction-controlled constant−? shear tests on
completely decomposed tuff. It is evident that the new model is capable of well capturing
effects of recent suction history on non-linear stress-strain relation and shear modulus
degradation at small strains.
Post a Comment