Development and evaluation of Induced Partial Saturation (IPS), delivery method and its implementation in large laboratory specimens and in the field

Liquefaction of saturated loose sands during an earthquake is associated with the build-up of excess pore water pressure, leading to loss of shearing strength of the sand. Current liquefaction mitigation techniques used in practice are expensive, and cannot be implemented on sites with existing structures. An innovative, practical, and cost effective liquefaction mitigation technique called "Induced Partial Saturation" (IPS) was proposed by Yegian et al., (2007). The proposed technique involves injection of a chemical solution that generates minute gas bubbles within the pores of initially fully-saturated sand, thus increasing the compressibility of the pore water leading to reduction or even elimination of liquefaction potential.;The research presented in this dissertation was focused on: 1) development of an automated chemical solution preparation and delivery system, which will induce partial degree of saturation in liquefaction susceptible sands, 2) implementation of the system in large scale laboratory and field tests, and 3) verification that partial saturation reduces liquefaction potential.;The automated IPS delivery system that was designed and manufactured has four major components: 1) controlled-rate chemical powder (Sodium percarbonate) delivery, 2) solution preparation, 3) solution pumping at a constant pressure, and 4) solution injection. The various components were tested separately to evaluate their performances and limitations. The entire IPS delivery system was also evaluated in the laboratory and in the field to ensure its efficiency and ability to operate continuously during long periods of chemical solution injection into sand specimens or sand layers.;The IPS delivery system was implemented to treat two sand specimens prepared in the large laminar box of NEES Buffalo. The specimens were prepared and tested by shaking using the large shaker of the laboratory. Pore water pressure transducers, placed at various locations within the specimens, measured the excess pore water pressures during and after shaking of the specimens. Electric conductivity probes were used to assess the partial degree of saturation and its spatial distribution within the specimens. The results of the experiments demonstrated that the automated IPS delivery system was robust, efficient, and simple to operate reliably for long injection durations. The electric conductivity data confirmed that the IPS injection system did reduce the degree of saturation from 100% to about 90%, using a chemical solution with concentration of 1%. The shaking table tests confirmed that partial saturation created by the IPS delivery system prevented the excess pore water pressures to rise to levels that would cause liquefaction of the sand specimens. In the second test, the entire sand specimen was treated by IPS. During and after shaking, the excess pore pressures were small, and there were no sand boils created at the surface of the specimen. It is noted that in earlier tests of fully saturated sand specimen (no IPS treatment) the specimen shaken in the same manner as the IPS specimens, had liquefied during shaking with dramatic manifestations of sand boils on top of the specimen.;IPS was implemented in the field at the NEES UCSB, Wildlife Liquefaction Array (WLA), California. This site has liquefied during past earthquakes in the region. In this field research, IPS treatment was applied to a certain zone within a liquefiable silty sand, and then the T-Rex truck of NEES UT shook the treated as well as an untreated zone to evaluate the effectiveness of IPS treatment in reducing liquefaction potential. The excess pore water pressures that were generated and measured in both tests were small (pore pressure ratios smaller than 0.2). Comparisons of the excess pore pressures from both untreated and treated sites were inconclusive. There were a number of difficulties and challenges that contributed to the field test results being not very useful, including: small zone of treatment; medium dense silty sand, seismic energy transmitted by T-Rex to the silty sand was small; ground water used for chemical solution preparation and injection had high concentrations of ions that accelerated the chemical reaction; and the ambient temperature was over 100 degrees Fahrenheit, which also accelerated the chemical reaction, leaving little time for injection.;Finally, the concept of using a dynamic cone to assess the effect of IPS treatment at a site was explored. A simple cone with a pore pressure transducer incorporated at its tip was driven into fully and partially saturated sand specimens prepared in the laboratory. The tests clearly indicated that the penetration of the cone liquefied the loose fully saturated sand, but could not liquefy the partially saturated sand. The concept of this cone can be expanded to make it applicable under field conditions.