Soil Fabric and Anisotropy as Observed Using Bender Elements during Consolidation

E-Mail: nabeelshm@uoanbar.edu.iq

INTRODUCTION

For geotechnical design, the assumption that the soil material is homogeneous and under isotropic conditions has been widely used (Terzaghi 1943; Kaiser and Hewitt 1982; Terzaghi et al. 1996). However, several case histories related to the failure of foundations, slopes, and levees led geotechnical engineers to question the appli-
cability of the aforementioned assumption for soil materials (Bauer 1984; Petroski 1994; Stark and Eid 1998; Abdoun and Dobry 2002; Briaud et al. 2001; Briaud 2008; Tanaka et al. 2009). In particular, stress redistribution following changes in the internal or external conditions of a soil deposit, within or adjacent to given soil layers, may lead to failures (Lee and Rowe 1989; Sawangsuriya et al. 2007). In other words, the amount of soil anisotropy and the initial.

 METHODOLOGY

Two types of soil, kaolinite-rich clay and illite-rich clay, were used in this study to investigate the effects of soil type on the soil anisotropy. Kaolinite-rich soil (KaoWhite-S, Thiele Kaolin, Sandersville, Georgia) was used for this study. The liquid limit, plastic limit, and plastic index for the kaolinite-rich soil used in this study were 35%, 30%, and 5%, respectively. A total of 400 g kaolinite soil was mixed with deionized, deaired water to form slurries with initial water content of 50% and 100%. These slurries were poured into a 3.81-cm-diameter (triaxial specimens) or
6.35-cm-diameter (BP-CRS-BE specimens) slurry consolidometer, and a vertical effective stress of 137.8 kPa was applied (Zhao and Coffman 2016). The slurry was mixed to a higher water content (100%) to ensure that the soil was reconstituted instead of re- molded, based on the definition presented by Olson (1962). Illite-rich soil (Knight Hawk Coal, Percy, Illinois) was used for this study; the soil was sieved to pass a No. 200 sieve prior to making the slurry. Following a similar laboratory preparation method as was used for the kaolinite soil, 400 g of the illite soil was mixed with deionized, deaired water to form a slurry with an initial water content of 75%. Completion of primary consolidation of the slurry- consolidated soils was ensured prior to the soil specimens being trimmed. The time required to complete a 100% average degree of consolidation was determined by following the procedures presented by Casagrande (1936). An anisotropic slurry consolidometer was used to control the stress condition during the process of slurry consolidation for both
kaolinite-rich and illite-rich clay soils. Specifically, as a result of the slurry consolidation with a vertical applied load, there is a general tendency for the clay particle edges to align in the horizontal direction (Krizek et al. 1975). Although the slurry clay fabric may be initially flocculated prior to applying a load, the clay particles may slowly orient toward the horizontal direction due to the subjected stress during anisotropic consolidation (Krizek et al. 1975). Following preconsolidation of the slurried samples within the slurry consolidometer to a stress of 137.8 kPa, four kaolinite soil samples and two illite soil samples were used within the BP- CRS-BE device to collect the required HV and HH shear wave
measurement data. Three kaolinite soil samples and three illite soil samples were used within the triaxial device (preconsolidated to 207 kPa within a 3.81-cm-diameter slurry consolidometer instead of within a 6.35-cm-diameter slurry consolidometer). Table 1 sum- marizes the soil samples that were used for this study.

BP-CRS-BE Testing Methods

The device that was used to conduct the BP-CRS-BE tests, includ- ing one pair of horizontal BEs and one pair of vertical BEs, is shown in Fig. 1. Each pair of horizontal and vertical BEs was fabricated with a pair of polyoxymethylene slide bars. Using these bars, the vertically oriented bender elements were able to be horizontally inserted into the soil sample, and the horizontally oriented bender elements were able to be vertically inserted into the soil sample. This fabrication enabled horizontally propagated, horizon- tally polarized shear waves to be generated and received, thus measuring V s;HH . The slurry consolidometer–prepared samples were trimmed, extruded, weighed, and inserted into the 25.4-mm-tall stainless-steel
confining ring of the BP-CRS-BE device. After the combined sample and ring were placed into the device, the triaxial chamber was assembled around the triaxial insert. The triaxial chamber then was placed into the load frame, the chamber was partially filled with water, and the drainage lines that were connected to the bottom of the sample were purged of air by using water. The kaolinite sample, located within the triaxial insert inside the triaxial chamber then had back-pressure saturation applied at 206.7 kPa back- pressure for 30 min prior to beginning the consolidation phase of the test. This back-pressure was maintained during the consolidation phase of the test.

Through shear wave measurements within a BP-CRS-BE device of kaolinite and illite samples, conclusions were drawn with respect to effect of water content, effect of soil type, and inherent fabric anisotropy. The information which was gained from the w s ¼ 50% kaolinite sample, the w s ¼ 75% illite sample, and the w s ¼ 100% kaolinite sample is summarized here. Even when soils have the same constrained modulus, the differences in the measured shear wave velocity may be caused by the microfabric of clay particles. For instance, the shear wave velocity (and therefore shear modulus) of the w s ¼ 50% kaolinite samples was lower than that of the ws ¼ 100% kaolinite samples, which indicated that a lesser amount of preferred particle orientation dominated in the clay fabric within the w s ¼ 50% kaolinite samples compared with the w s ¼ 100% kaolinite sample. These conclusions also were validated using triaxial testing data from samples mixed to the same initial water contents. Two soils with a similar initial deformation behaved very differently during consolidation and shearing. Although the w s ¼ 50% kaolinite sample and w s ¼ 75% illite samples were prepared initially with similar amounts of axial strain in the slurry consolidometer, the soil samples behaved very differently during consolidation and shearing. The horizontally propagating, vertically polarized shear wave velocity measurement technique was used to gain in- sight into this difference. Although this difference was water- content related, it may be attributed more to the mineral structure of the illite mineral (2:1 sheet) compared with that of the kaolinite mineral (1:1 sheet). There was a limitation regarding the comparation of the kaolinite and illite soil samples that were not prepared at the same initial water content, although it did not affect the conclusions derived from this study. It is recommended that future study be conducted with different soil types at the same initial condition. The sheet arrangement of the two minerals also was shown to affect the amount of inherent soil anisotropy. Again, the data collected from the illite mineral (2:1 sheet) showed more hysteresis in the shear wave velocity measurements than the data collected from the kaolinite mineral (1:1 sheet). Moreover, the hysteresis was much more pronounced using the horizontally propagated, horizontally polarized waves.