| dc.description.abstract | Liquefaction has been one of the major concerns in the field of geotechnical earthquake engineering. Various mitigation techniques such as vibroflotation, deep dynamic compaction, explosive compaction, grouting, deep soil mixing etc. have been employed over the last fifty years. These techniques either densify the in-situ soil or fill the voids with some external agent. Some of these techniques are costly, while others pose a threat to the adjacent structures and environment. Over the last couple of years, induced desaturation is emerging as a possible cost-effective and environment-friendly liquefaction mitigation technique. In this technique, the degree of saturation of the in situ saturated soil is reduced either by injecting air or by generating some kind of gas within the soil matrix.
The present study investigates the liquefaction resistance and cyclic response of the air-injected desaturated clean sandy soil. The large number of stress controlled undrained cyclic triaxial tests were conducted on samples with the degree of saturation in the range of 70 % to 99 %. Three relative densities of 30 %, 40 % and 60 % and initial effective confining pressure of 25 kPa, 50 kPa and 100 kPa were considered for the investigation. Cyclic shear stress ratio (CSR) was 0.175, 0.250, 0.300 and 0.400. It was found that the number of cycles required for initial liquefaction increased exponentially with the reduction in the degree of saturation. Cyclic strength of desaturated sand with relative density of 40 % with the degree of saturation of 80 % was found to be almost twice that of fully saturated sand. Depending upon the degree of saturation, relative density, confining pressure and CSR, five distinct failure modes were observed: 1) Hybrid cyclic liquefaction 2) Cyclic mobility-gradual/catastrophic 3) Cyclic softening-CE 4) Cyclic softening-E and 5) FAPSC-SS: Failure due to Accumulation of Plastic Strain on the Compression side as a result of gradual Strain Softening. Samples with a high degree of saturation underwent cyclic
mobility or hybrid cyclic liquefaction failure while those at a low degree of saturation underwent cyclic softening-CE or cyclic softening-E. Insight from the critical state soil mechanics framework revealed that nearly saturated samples with the initial state closer to critical state undergoes either catastrophic cyclic mobility or hybrid cyclic liquefaction whereas that farther from it undergoes either gradual cyclic mobility or FAPSC-SS. The phase transformation line was found to play an important role in understanding the cause of pore pressure reduction during the extension stage of loading in samples with a low degree of saturation. Longevity test revealed that at the end of seven days, there was negligible change in the degree of saturation of the triaxial specimen. It was found that Konstadinoun and Georgiannou (2014) pore pressure model, when modified for CSR and desaturation effect, could be used to predict pore pressure evolution in desaturated sand. Further, assessment of hypothetical desaturated soil domain by approximate method, revealed that desaturated soil with degree of saturation of around 70 % is adequate to prevent liquefaction under strong to very strong earthquakes having peak acceleration as high as 0.36 g.
The numerical investigation was performed employing open-source finite element software “OpenSees” developed by Pacific Earthquake Engineering Research (PEER) Center, University of California, Berkeley. Constitutive behavior of the soil was modelled using in-built pressure dependent multi yield material model “PressureDependMultiYield”. The triaxial specimen was modeled by eight-node linear isoparametric hexahedral element “BrickUP”. This element simulates the undrained response of coupled solid-fluid material based on Biot’s theory of poroelasticity. Further, an evaluation study was conducted to determine the constitutive parameters for the saturated and desaturated condition. Finite element model of 30 m thick soil domain was prepared and subjected to four earthquake motions from India, having peak acceleration between 0.10 g and 0.36 g. This soil domain was discretized into four node plane strain bilinear
isoparametric elements entitled as “quadUP”. This element captures coupled solid-fluid response when subjected to dynamic loading. The parametric study was conducted to understand the effect of permeability, degree of saturation, the thickness of the desaturated zone, on liquefaction resistance of the soil, measured in terms of pore pressure ratio. With an increase in the permeability for fully saturated condition, excess pore pressure was reduced, and dissipation of pore pressure was accelerated. Further, the higher the permeability, the higher was the acceleration amplification for fully saturated condition. With the reduction in the degree of saturation, reduction in the thickness of liquefied zone was observed. Moreover, for the degree of saturation of 81.4 %, the thickness of the non-liquefied zone was found to be equal to the thickness of the desaturated zone, irrespective of the input motion.
Keeping in view, experimental and numerical findings, the following points can be recommended: 1) Desaturation up to the degree of saturation of 80 % is adequate to double the cyclic strength of clean sand. 2) From approximate method it was found that desaturation up to degree of saturation of 70 % was adequate to prevent liquefaction against strong to very strong earthquakes having peak acceleration in the range of 0.10 g to 0.36 g, on the other hand numerical simulation revealed that degree of saturation of 81.4 % was sufficient to avoid liquefaction under same earthquakes. As numerical analysis is more rigorous than approximate method, it can be concluded that desaturation up to the degree of saturation of 80 % is adequate to prevent liquefaction. 3) Injected air gets entrapped into the voids and remains there for a long period of time. 4) The thickness of the desaturation zone, with the degree of saturation of 81.4 %, can be kept between 5 m to 15 m to prevent liquefaction under strong to very strong earthquakes having peak acceleration between 0.10 to 0.36 g. 5) Amplification factor at the ground surface for the degree of saturation of 81.4 % was found to be in the range of 0.32 to 1.76, under strong to very
strong earthquakes having peak acceleration between 0.10 to 0.36 g. This implies that desaturation of clean sand up to degree of saturation of 80 % is enough to achieve the two-fold goal: 1) to prevent liquefaction 2) to keep the acceleration amplification low. Further reduction in the degree of saturation may amplify motion significantly owing to presence of high matric suction, though this issue needs further investigation | en_US |
