Advanced methods for rock discontinuities estimation in tunneling

The thesis is the summary of the three-years work done at the Department of Mechanics at the Politecnico di Torino, from January 2010 to December 2012, under the supervision of professor Luigi Garibaldi (Dipartimento di Ingegneria Meccanica e Aerospaziale) and professor Alberto Godio (Dipartimento di Ingegneria Ambientale, del Territorio e delle Infrastrutture). The subject of the thesis has been chosen in collaboration with Gd Test S.r.l., a company operating in the field of environmental engineering, based in Torino, relatively close to the Politecnico. During the three years, the company cooperated with me and the professors, in particular by giving an important help for the experimental tests. The topic is the prediction of lithological discontinuities ahead of a tunnel face, in a tunnel under construction, i.e. the identification of whatever type of zone producing a reflection of the seismic waves generated in a source point. Typical excitations are the blasting, the hammer or magneto-strictive sources. Moreover, the procedures can be improved to be applied in presence of the Tunnel Boring Machine, by exploiting the noise produced by its cutter head. The main objective of the work is duplex: to reduce the risk for people and machinery and to reduce the excavation cost by setting the excavation based on the rock configuration that has been predicted. The theme is a modification of the classical reflection seismology techniques, in order to be applied in tunnels. Indeed, the particular configuration of sources and receivers in a tunnel obliges to develop ad hoc methods for the analysis of the seismic data. The thesis proposes two methods for the identification of the discontinuities ahead of a tunnel face: the first comes directly from the concept of ellipsoids, fundamental in the reflection seismology, and the second is a backward method, able to provide an automatic prediction without extracting the time arrivals corresponding to the reflected waves. To run, this procedure needs only the coordinates of sources and receivers, together with the acquisition data parameters, and then it allows: i) to estimate an average value of the wave velocity; ii) to detect the discontinuities for each source; iii) to analyze and plot the number of superposing estimations for each node of the selected space domain. The final result can be interpreted as the probability to detect a discontinuity at a certain distance from the tunnel face. The thesis first describes the fundamental relationships in tunneling seismology, then it proposes a direct technique for the reflectors detection. Since it is demonstrated to be not too reliable in presence of low signal-to-noise ratio, the automatic method is presented. This procedure has been tested on synthetic and real data coming from the new Brennero tunnel under construction, which will be one of the longest tunnel in the world. The results indicate that the method runs very fast and it is reliable in the identification of lithological changes and discontinuities, up to more than one hundred meters ahead the tunnel face. Moreover, it can be used in every kind of tunnels, without any detailed knowledge of the soil and rocks around the tunnel, and with different excitations. The thesis deals also with another important issue during the tunnel construction: the monitoring of the rockburst phenomenon, a small seismic event caused by the rock alteration during the excavation. For this purpose, a specific procedure has been created, able to estimate the position of the source point.