Date of Award

Fall 2000

Document Type


Degree Name

Doctor of Philosophy (PhD)


Civil & Environmental Engineering

Committee Director

Isao Ishibashi

Committee Member

Zia Razzaq

Committee Member

Ramamurthy Prabhakaran


Granular materials such as sand deposits have anisotropic characteristics due to the gravitational forces during depositional process and various stress histories. These anisotropic characteristics make evaluation of liquefaction resistance of soil very complex. Since obtaining undisturbed soil specimens for laboratory tests from discrete elements like sandy soils is very laborious, the soil specimens are reconstituted in laboratory with regards to the in-situ density of soil. Reconstituting the soil changes the soil fabric or anisotropic characteristics in comparison with field conditions. Many researchers have pointed out that liquefaction resistance is highly influenced by differences in the structure of the soil (or soil fabric) produced by different sample preparation techniques. Liquefaction resistance could vary with different soil fabrics even for the same soil with a specific density.

The objective of this research is to investigate the anisotropy of granular materials and to establish its relation to the liquefaction resistance. Anisotropy can be expressed by the directional mechanical properties of the soil. In order to investigate the effects of the anisotropy on the liquefaction resistance in a quantitative way, uniform medium dense sand specimens were prepared using three different techniques to create the different initial soil fabrics. Undrained cyclic triaxial tests were performed to determine the liquefaction resistance of each soil specimen. A relationship between liquefaction resistance and the number of cyclic loads was established for three types of specimens with different soil fabrics. The results clearly demonstrate that the preparation methods significantly affect the liquefaction behavior.

After preparation of the soil specimens, vertical and horizontal compression (P) wave velocity and vertical shear (S) wave velocity were measured, and the soil fabric effects on elastic properties of the soil were investigated under dry and saturated conditions. P-wave velocities in both directions showed small difference for all sample preparation techniques except for air pluviated (AP) specimens. AP specimens reflected cross anisotropy. However, S-wave velocities turned out to be very affected by soil-fabric. S-wave promised to distinguish nature of inherent anisotropy with regard to liquefaction resistance.

Anisotropic elastic constants of soil specimens were recovered quantitatively from elastic wave measurements and consolidation test data. Induced anisotropy due to different preparation techniques was identified by recovered anisotropic constants. The recovered elastic constants were found sufficient to express anisotropy of specimens. Accordingly, by using directional variation of elastic constants, three anisotropy indices were created. The anisotropy indices were then used to create pseudo cyclic stress ratio. Liquefaction cyclic stress ratios of the specimens were normalized with corresponding pseudo cyclic stress ratios. It was observed that the anisotropy effects on liquefaction resistance can be eliminated with those anisotropy indices. It was found that under cyclic loading, anisotropy indices obtained from experimental S-wave measurements were able to express anisotropy of granular materials and a relation to the liquefaction resistance.