GNGTS 2018 - 37° Convegno Nazionale

GNGTS 2018 S essione 3.2 669 The most recent literature suggest the use of geophysics as a tool for identifying the liquefied zones after earthquakes (Abu Zeid et al. , 2012; Apostolopoulos et al. , 2013; Giocoli et al. , 2014). Applied geophysics techniques were also used to quantify the phenomenon, evaluating the differences in electrical resistivity before and after the detonation of explosive charges in a liquefiable unit (the so called Blast Tests , e.g. Amoroso et al. , 2017). The hypothesis, which we want to test in this work, is the following: during the generation of the excess pore pressures, grains of sand “vibrate” and they tend to separate from each other at each cycle of the seismic input. If so, we should observe be two important variations in at least two geophysical parameters, without producing sensitive settlements or displacements: electrical resistivity and v s velocity. In the light of this hypothesis and to test it, a simple laboratory experiment has been conducted, trying to validate the abovementioned statement, in particular from the point of view of electrical resistivity. Jinguuji et al. , (2003), were the first who try to follow the process in the laboratory, studying the variation in density pre and post liquefaction using electrical resistivity and cyclic dynamic loads. On the contrary, the proposed experiment consisted of micro-ERT (Electrical Resistivity Tomography) time-lapse measurements carried on a sandbox subjected to an impulsive “seismic” input, simulating a real ERT field acquisition. A 70x45x40 cm box is filled with fine sand (HOSTUN Sand) (Fig. 1 right panel). The sandbox is equipped with four tubes at each angle to saturate the system gradually from the bottom. A monitoring well is installed externally to the tank to control the water level and its oscillations. The sandbox is positioned above a wood plate where a sledgehammer pendulum of 6 kg is used to hit the system generating the seismic input (Fig. 1 left panel). On the surface of the sand, 24 inox steel electrodes are connected to Syscal Pro georesistivimeter for ERT measurement. Resistivity data were acquired for all the experimental steps with a Schlumberger configuration of 121 quadrupoles measured every about 130 s. The experiment was divided in four main steps: 1. Saturation step : every ERT 0.5 l of water was poured in the model. This step ends when the water table reaches a depth of approximatively 8.7 cm below ground surface; 2. Stationary step : ERT time-lapse measurements for back-ground analysis; 3. Dynamic step : the pendulum hits 7 times the system at the beginning of each ERT. During this step, the water level increase was monitored through the external pipe. Fig. 1 - Experimental setup: a sledgehammer is used as seismic source. Each hit is followed by an ERT measurement to evaluate the effect of the seismic input.