Verification of effectiveness and design procedure of gravel drains for liquefaction remediation

Abstract

The current design practice of using gravel drains as a liquefaction countermeasure involves the selection of the drain spacing and drain diameter to keep the peak excess pore water pressure ratio low. As it has mostly been verified through small-scale 1 g shaking tests, its validity for the field-scale prototype has yet to be well investigated. In this study, a series of centrifuge tests was conducted to gain insight into the stress-dependent behavior of loose sand deposits with a level surface improved by gravel drains. Meanwhile, the current design procedure was validated with experimental data. The results revealed that the effects of gravel drains in suppressing the excess pore pressures depend significantly on the depth of the drains. The current design procedure has failed to elucidate the depth-dependent behavior of sand deposits. One of the important features of the mechanical properties of the soil used for the design of gravel drains was revealed through laboratory tests in which the coefficient of volumetric compressibility, mv, was found to be highly dependent on the stress level, while mv was assumed to be fixed in the design procedure. It was also found that the water flow regime in gravel drains can be a turbulent flow. The Reynolds number in drains increases from the bottom to the top, and the permeability coefficient decreases accordingly, resulting in more significant well resistance than expected based on the current design procedure. In the present study, when stress level-dependent mv and Reynolds number-dependent kw were used as input soil parameters, the axisymmetric diffusion equation, with consideration given to the well resistance, satisfactorily predicted the excess pore pressures in sand with gravel drains.