Seismic response and prevention of liquefaction failure of sands partially saturated through introduction of gas bubbles

Liquefaction of sands is an earthquake-induced phenomenon that can cause devastating damages to our built environment. Liquefaction mostly occurs in loose saturated sands when subjected to repeated or seismic loading, due to excess pressure generation of incompressible pore water. This research reports on developing a new liquefaction mitigating technique referred to as Induced Partial Saturation (IPS) which will be a cost-effective and practical solution for new as well as existing structures. The liquefaction mitigation measure that was explored improves earthquake resistance of loose saturated sands by introducing some amount of gas/air in the voids of the sand. Preliminary research performed by Eseller (2004) and Yegian et al. (2007) demonstrated that IPS can be a potential mitigation measure against liquefaction. This research further evaluates in depth the seismic response and liquefaction benefit of sands mitigated by IPS as well as it investigates the sustainability of IPS in sands under various conditions in nature through an integrated experimental and analytical research program.;The short- and long-term sustainability of entrapped gas/air bubbles was tested in large scale experimental setups under hydrostatic, and upward, downward, and lateral hydraulic gradients as well as under horizontal excitation. A cyclic simple shear liquefaction box (CSSLB) was designed and manufactured which can induce uniform shear strains in a sand specimen through the use of a shaking table. The CSSLB is versatile and accommodates the insertion of various types of transducers needed in this research. A new laboratory IPS technique was developed to prepare partially saturated sand specimens with uniformly distributed gas bubbles and at controllable degrees of saturations. The new technique involved mixing dry sand with a dental product "Efferdent" (main ingredient being an oxygen source: Sodium Perborate) and raining the mixture in water leading to generation of oxygen gas bubbles in the voids. Then the uniformity of the soil characteristics (relative density and degree of saturation) of sand specimens prepared by IPS was evaluated using a multiple P and S wave measurement facility that was developed for use in large soil specimens. The P-S wave measurement set-up was used to investigate the potential use of P-wave measurements as a means for estimating partial degree of saturation in sands. After evaluating the soil characteristics, cyclic simple shear strain tests were performed on fully and partially saturated sand specimens prepared by IPS. The effect of important parameters including relative density, degree of saturation, induced shear strain levels, initial effective stresses, and applied number of cycles, on excess pore water pressure ratio (ru) generated in sand specimens was investigated through a series of cyclic simple shear strain tests. Finally, an empirical model (RuPSS, excess pore water pressure ratio (ru) in Partially Saturated Sands) was developed to predict excess pore water pressure ratios in partially saturated sands. The model was based on the experimental data obtained from the cyclic simple shear strain tests.;The results from this research show that gas/air bubbles entrapped in sand specimens by IPS remain in the voids even under various flow and ground shaking conditions encountered in nature. It was confirmed by the P and S wave measurements that the new IPS laboratory technique is capable of achieving uniform partially saturated sand specimens. Also P wave measurements performed in partially saturated sand specimens at several degrees of saturation demonstrated that contrary to published information, P wave velocity can not provide estimate of degree of saturation (S) for S< 90%. According to cyclic simple shear strain tests, IPS prevents sands from liquefying by limiting maximum r u (rumax) to less than 1. Also, IPS leads to increased number of cycles required to reach rumax, thus resulting in smaller values of excess pore pressures during small to moderate earthquakes. Also, excess pore water pressure ratio is reduced in specimens with lower degrees of saturation, higher relative densities and under lower strain amplitudes. The developed empirical model RuPSS predicts excess pore water pressure ratio in partially saturated sands during earthquakes by incorporating the effects of degree of saturation (S), relative density (Dr), initial effective stress (sigmav'), the induced shear strain amplitude (gamma) and earthquake magnitude (M). The developed model can be used in field applications of IPS, by providing predictions of either ru for a design level S, or S for a limiting value of ru.