Shear modulus, G₀, at very small strains (0.001% or less) is one of important parameters for predicting dynamic response and behaviour of soils for many engineering problems such as design and analysis of earth retaining structures for earthquakes and static loads. With the advancement of use of bender elements in soil testing, anisotropic G₀ can now be determined by measuring velocity of shear wave propagating in different planes of a soil specimen. It leads to the evaluation of anisotropy of G₀. While most studies of the anisotropy have been conducted on saturated soils, experimental studies of anisotropy of unsaturated soils have rarely been carried out. _0(ij)...[ Read more ]

Shear modulus, G₀, at very small strains (0.001% or less) is one of important parameters for predicting dynamic response and behaviour of soils for many engineering problems such as design and analysis of earth retaining structures for earthquakes and static loads. With the advancement of use of bender elements in soil testing, anisotropic G₀ can now be determined by measuring velocity of shear wave propagating in different planes of a soil specimen. It leads to the evaluation of anisotropy of G₀. While most studies of the anisotropy have been conducted on saturated soils, experimental studies of anisotropy of unsaturated soils have rarely been carried out.

This study aims to characterize anisotropic shear moduli (G_0(ij)), of intact and recompacted specimens of a completely decomposed tuff (CDT) under both saturated and unsaturated conditions. A CKC triaxial test apparatus was modified to measure shear wave velocity and hence shear moduli at different suctions in three polarization planes at different stress states using bender elements. Three series of triaxial tests were conducted on intact and recompacted specimens under both saturated and unsaturated conditions. The test programme included isotropic compression tests at constant suction, drying and wetting tests at constant net mean stress and shearing tests at constant stress ratio and suction.

Experimental results indicate that degree of anisotropy, expressed by G_0(hh)/G_0(hv)), of intact and recompacted specimens is 1.45 and 1.05 on average respectively. In isotropic compression tests conducted on intact and recompacted specimens, for a given suction, the degree of anisotropy appears to be independent of net mean stress ranging from 110 kPa to 500 kPa. In drying and wetting tests conducted on recompacted specimens, for a given net mean stress, change of the degree of anisotropy is less than 2% as matric suction varies within a range of 0 kPa to 250 kPa. In shearing tests, for given values of both stress ratio and suction, variation of the degree of anisotropy is less than 4% as deviatoric stress increases from 110 kPa to 500 kPa. It appears that net mean stress and matric suction do not induce anisotropic shear modulus but an increase in stress ratio results in a reduction of degree of anisotropy. These findings are consistent with predictions using empirical equations proposed in this study. During a drying and wetting cycle at constant net mean stress, there is a marked hysteresis loop of shear wave velocity and hence shear modulus measured during a cycle of drying and wetting for recompacted specimens.