THESIS
2019
xvi, 156 pages : illustrations (some color) ; 30 cm
Abstract
Liquefaction refers to the dramatic loss of soil stiffness and strength during earthquake shaking.
Knowing the particle-scale information of granular particles and void space, collectively
termed as “fabric”, is crucial for obtaining a fundamental understanding of liquefaction in
granular soils. In this study, fabric evolution of granular soils in cyclic liquefaction is
investigated using 3D Discrete Element Method (DEM). A set of new fabric indices are
developed for quantifying particle-void distribution (?
?,?
?), which are found to have an excellent correlation with the post-liquefaction behavior, as well as irreversible changes in
fabrics developed in liquefaction processes, and finally proposed a fabric-based criterion for
jamming transition in flow deformation. The descripto...[
Read more ]
Liquefaction refers to the dramatic loss of soil stiffness and strength during earthquake shaking.
Knowing the particle-scale information of granular particles and void space, collectively
termed as “fabric”, is crucial for obtaining a fundamental understanding of liquefaction in
granular soils. In this study, fabric evolution of granular soils in cyclic liquefaction is
investigated using 3D Discrete Element Method (DEM). A set of new fabric indices are
developed for quantifying particle-void distribution (?
?,?
?), which are found to have an excellent correlation with the post-liquefaction behavior, as well as irreversible changes in
fabrics developed in liquefaction processes, and finally proposed a fabric-based criterion for
jamming transition in flow deformation. The descriptor ?
?, is found to be a good indicator for
the compressibility of the sample in the post-liquefaction stage. DEM simulations have been
conducted to identify critical factors that influence the granular soil’s resistance to multiple
liquefaction. It is observed that re-liquefaction resistance is greatly affected by the fabric
anisotropy of the sample by surpassing the effect of relative density. Within an unloading cycle
in the post-liquefaction stage, completely different soil fabric was formed if the liquefied soil
is consolidated at different unloading points, resulting in vastly different resistance to re-liquefaction.
In addition, medium dense granular assembly is subjected to different multi-directional loading
paths (Circular/Oval, Figure-8, Uni-directional) to explore the changes in soil micromechanical
structure. The trajectory of evolution in contact-based fabric tensor components is found to follow the shape of the loading path in the initial cycles of loading, whereas the trajectory of
evolution in new void-based fabric tensor components follow the variation in shear strain
corresponding to different directions. A possible new-criterion to decide soil liquefaction
triggering based on energy dissipation is discussed, which has more physical meaning
compared to the available criterion in case of multi-directional loading paths. Overall, the
discrete element modelling provides many insights that link microscopic fabric evolution to
the macroscopic cyclic soil behavior, which can be crucial information to develop micro-mechanically
based constitutive models.
Post a Comment