GNGTS 2014 - Atti del 33° Convegno Nazionale

42 GNGTS 2014 S essione 3.1 (Lancellotta et al. , 1991) led to the identification of four main joint sets (dip direction/dip): K1 (110/75), K2 (0/80), K3 (150/15) and K4 (50/75). These discontinuities tend to isolate the north- western frontal portion of the cliff which increases its instability as a function of the rock joints along the fractures and its foot edge. A volume of about 12,000 m 3 has been estimated forming the instable sector. Particularly along a K4 discontinuity there is a clear evidence of movement: a decimetric step dislocates the lawn and the confining walls of the panoramic square. Geophysical investigation. Focusing on the gravitative processes that affect the studied rock mass, a good knowledge of the fracturing state is an important requirement for the stability analysis, with particular reference to orientation, spacing, opening and persistence, filling and hydraulic conditions of the discontinuities. With this aim, we extended a previous cross-hole seismic tomography (Colombero and Comina, 2014) realized in the yard in front of the sanctuary, in order to obtain a tomographic image between three available inclinometric boreholes (Fig. 1b). To perform the tests a Borehole Impacter Source by Geotomographie GmbH was used as in-hole source in the S2 borehole, with three different locations till a depth of about 6 m (after this depth an obstruction of the hole casing inhibited deeper measurements) while a sledge hammer, impinging both vertically and horizontally on a steel rod, was employed as surface source in different locations along the lines connecting the three holes. A prototype borehole string equipped with 8 three-component geophones (10 Hz) at 1 m spacing, stiffly connected by a PVC bar that permits to control geophone orientation, was progressively lowered (with a 2-geophone superposition each subsequent positioning) in the S1 and S5 soundings, till the maximum available depth (respectively of 27 and 21 m). On the lawn surface 4 three-component geophones (2 Hz) were placed along the lines connecting the three boreholes, with a reciprocal spacing of 2.5 m (along S1-S2) and 4 m (along S2-S5). First break manual picking was performed on the acquired seismic traces, on the pertaining components, to allow for both P- and S-wave velocities imaging. Data were inverted to obtain a tomographic image between the investigated volume with the use of GeoTomCG software, which performs three-dimensional tomographic analysis with source and receiver positions in any configuration within a 3-D grid. The software allows for curved-ray calculations which have been observed to be more accurate in case of strong velocity contrasts. Curved ray tracing is performed with a revised form of ray bending, derived from the Um and Thurber (1987) Fig. 1 – Building of a 3-D seismic velocity model for microseismic event localization. a) The cliff of Madonna del Sasso (NW Italy) with the location of the four triaxial geophones of the monitoring network. b) The point cloud obtained from a previous laser-scanner survey with the results of the cross-hole seismic tomography between the inclinometric boreholes S1, S2 and S5. c) Rendering of the 3-D velocity model for P waves obtained from DSM, cross-hole tomography results, geological and geomechanical observations and measurements.