GNGTS 2018 - 37° Convegno Nazionale

GNGTS 2018 S essione 3.2 691 Politecnico di Milano. A couple of tests were performed on a small-scale levee – representing the real site – experiencing different levels of water in the channel and various rainfalls. The survey layout, the materials and the geometry of the structure being known, we implemented a forward modelling approach to assess the 3D effects and the influence of buried electrodes to correct the measured data. Quantifying the effects related to 3D geometry and the survey configuration on 2D datasets is very important to correct the measured apparent resistivity values and to properly identify the variations of the embankment conditions (Sjödahl, 2002). Small-scale laboratory tests. The actual test site was scaled down for a scale of 1:12 (Fig. 1a) on a laboratory flume with transparent plexiglass walls and base size 80×200 cm . Two taps were put to allow water entrance and discharge and 6 sprinklers were placed on the top frame of the flume to simulate rainfalls. The embankment was built with homogenous sand with initial volumetric water content of 13%. Future tests are also planned to be performed with the soil collected from the real site. Two miniaturized cables equipped with 48 2 cm -long stainless- steel electrodes were designed, assembled and tested for laboratory purpose and connected to an IRIS Syscal Pro resistivity meter. Time-lapse ERT measurements were carried out with the Wenner array configuration (a = 3 cm ), with a horizontal resolution of 3 cm and a vertical one of 1.5 cm , and the electrodes were buried at a depth of 2 cm along the levee in order to avoid preferential paths for rainfall infiltration. In addition, a TDR was embedded in the structure to measure volumetric water content values during the tests and two GoPro cameras were set to take pictures every 10 s . Three sets of experiments were performed to investigate the ideas of this research. In one test, the seasonal filling and drawdown procedure typical for irrigation canals was reproduced (Fig. 1b). Levees of irrigation canals are subject to seasonal cycles of saturation and drying, Fig. 1 - a) Top: geometry of the monitoring site in San Giacomo delle Segnate. Bottom: The scaled down levee designed for laboratory studies. b) Top: an experiment simulating the filling period, showing 5.5cm of water in the canal for T29 time step. Bottom: an experiment when the water is discharged at T1.37 after the levee has experienced different water levels in the canal.