Experimental study on liquefaction resistance of partially saturated sands at s=70%-80% with emphasis on induced shear strain

Abstract

Liquefaction is one of the most important research topics in geotechnical engineering due to the severe damage it causes. In soil mechanics and geotechnical engineering, liquefaction is the behavior of fully saturated granular soils, mostly sands and silty sands, as a viscous liquid due to an increase in excess pore water pressure under a dynamic loading such as an earthquake. Consequently, the soil loses the interaction between the grains when the pore water pressure is equal to the vertical effective stress. Liquefaction causes loss of bearing capacity in structures, high settlements, and lateral spreading in the soil. A range of soil improvement methods have been developed so far to increase the liquefaction resistance of soils. However, most of the common methods are based on soil remediation before construction. After researchers realized that partially saturated soils have higher liquefaction resistance than fully saturated soils, soil improvement methods by reducing the degree of saturation of soils were started to investigate. In this context, the induced partial saturation (IPS) method, which can be applied to sites with existing structures, was developed. In this study, stress-controlled undrained cyclic simple shear tests were performed with the dynamic simple shear with confining pressure (DSS-C) device in the Istanbul Technical University Soil Mechanics and Geotechnical Engineering Soil Dynamics Laboratory. Experiments were conducted on fully saturated and partially saturated Sile (AFS 40-45) sand samples treated with IPS. Sodium percarbonate was used to apply the IPS method. Sodium percarbonate reacts with water to create air bubbles in the sample and causes the same volume of water to escape from the soil. After partially saturated and fully saturated samples were prepared using the wet pluviation method, the tests were processed in three stages: saturation, consolidation, and liquefaction. The final degree of saturation and relative density were determined from the values at the end of the consolidation stage. Fully and partially saturated samples were consolidated under vertical effective stress of 100 kPa, and experiments were conducted on different relative densities, degrees of saturation, and CSR values. The reason for choosing a vertical effective stress of 100 kPa is to develop a partial saturation resistance factor at 75% degree of saturation for the simplified procedure used to determine the liquefaction resistance. During the liquefaction stage, the sample was subjected to stress-controlled undrained cyclic loading in the sinusoidal waveform at 1 Hz. Within the scope of this study, cyclic loading was continued until the moment when the excess pore water pressure in the specimen was equal to the vertical effective stress, that is, when the excess pore water pressure ratio was equal to one (ru=1), and it is was accepted as the initial liquefaction. In total, 46 partially saturated and 10 fully saturated specimens were tested. Due to problems with the DSS-C testing device, reliable data could not be obtained from all experiments. Therefore, in this study, the results of the experiments on 25 partially saturated and four fully saturated samples were presented. Partially saturated specimens were obtained at an average relative density (Dr) of 40%, 50%, 60%, and an average degree of saturation (S) of 75%, and fully saturated specimens were obtained at an average relative density of 60%. One of the primary objectives of the thesis is to obtain the liquefaction resistance curves with certain cyclic stress ratios applied under S=75% degree of saturation and Dr=40%, 50%, and 60% relative density. With the tests performed, the effects of degree of saturation, relative density, and cyclic stress ratio on the liquefaction resistance of partially saturated sands were evaluated. Consequently, it is realized that the liquefaction resistance increases with the rise in the relative density or the drop in the degree of saturation. The cyclic stress ratio to reach liquefaction increased with increasing relative density or decreasing degree of saturation. It was noticed that if the cyclic stress ratio increased at constant relative density and degree of saturation, the excess pore water pressure generation accelerated; moreover, the number of cycles to liquefaction decreased. The effect of the relative density was also observed in the normalized liquefaction resistance curves. The partial saturation resistance factor (KPS) was developed for a 75% degree of saturation in order to evaluate the liquefaction potential of sand soil treated by the IPS to apply in the simplified procedure. In addition, in this study, the shear strain behavior of partially saturated sands under stress-controlled undrained cyclic loading by DSS-C device was investigated. It was observed that as the cyclic stress ratio increases, the shear strain corresponding to the number of cycles at the moment when ru=1 increases independently of the relative density. Since the liquefaction resistance of partially saturated specimens increased, high cyclic stress ratios were applied in order to reach liquefaction, unlike fully saturated specimens. Deviations in the initial reference points of some samples due to large shear strain were noticed. In this study, a partial saturation resistance factor (KPS) was developed for a 75% degree of saturation for use as a CRR curve in the simplified procedure; moreover, the shear strain behavior of partially saturated sand soils under high cyclic shear stresses was investigated.