Tunnel construction in squeezing conditions is very demanding. Difficulties are met in making reliable predictions at the design stage. During excavation squeezing is not easily anticipated, even when driving into a specific geological formation. Squeezing conditions may vary over short distances due to rock heterogeneity and changes in the mechanical and hydraulic properties of the rock mass. Indeed, the selection of the most appropriate excavation-construction method to be adopted (i.e. mechanized tunnelling versus conventional tunnelling) is highly problematic and uncertain.
Squeezing behaviour stands for large time-dependent convergence during tunnel excavation. It takes place when a particular combination of induced stresses and material properties pushes some zones around the tunnel beyond the limiting deviatoric stress at which time-dependent deformations start. It is closely related to the excavation and support techniques which are adopted. Deformations may terminate during construction or continue over a long period of time.
The consequences of squeezing consist of large tunnel closures, considerable deformations of the tunnel face, high pressures on the support, or on the shield of the Tunnel Boring Machine (TBM) in case of mechanized tunnelling, and eventually, in extreme conditions, local instabilities and collapses. Due to the fixed geometry and the limited flexibility of the TBM, allowable space to accommodate ground deformations is restricted, and the ground can slowly lock the machine. On the contrary, in conventional tunnelling a considerably larger profile can be excavated initially in order to allow for large deformations to take place. The obvious consequence is that in deep tunnels, whenever severe squeezing conditions are anticipated, conventional tunnelling appears to be preferred over mechanized tunnelling.
In engineering practice, the difficulties to deal with squeezing conditions are connected to: (1) the evaluation of the time-dependent characteristics of the rock mass by means of laboratory or in-situ tests, (2) the use of an appropriate constitutive model, and (3) the choice of a suitable excavation and support system.
I started studying the squeezing behaviour of tunnels during my MSc thesis in Civil Engineering at the Politecnico di Torino in 2003, under the supervision of Prof. Giovanni Barla, who made me interested in Rock Mechanics and Rock Engineering. The thesis, which belongs to a wider research program on the excavation of tunnels in difficult conditions, started from the observation that tunnel construction in squeezing conditions is very demanding due to the difficulty in making reliable predictions at the design stage. This problem was analysed from both a numerical and experimental point of view.
One of the main problems of design practice is the use of an appropriate constitutive model that can take into account correctly the time dependent behaviour of the rock mass. Previous studies carried out by the Rock Mechanics and Rock Engineering (RMRE) group showed that the rheological models which are currently used in the design practice, like the CVISC model, are not able to reproduce correctly all the features of time dependency involved in tunnel excavation. For this reason the more advanced viscoplastic model of Lemaitre was chosen as the subject of the thesis. The model was derived analytically, implemented into the finite difference code FLAC2D and then calibrated on the results of laboratory tests on clay shales samples from the Raticosa tunnel (Firenze-Bologna), which experienced very important squeezing problems during construction. Even if the Lemaitre's model showed to match very well both the results of laboratory tests and the monitoring data of the tunnel, it revealed some important limits that cannot be neglected (Bonini et al., 2009).
The experimental part of the thesis was dedicate to the set-up of the High Pressure Triaxial Apparatus (HPTA; Barla et al, 2010), that was especially developed for the DIPLAB Laboratory of DISTR to allow the study of time dependent behaviour of weak rocks at high confining pressure. Preliminary tests performed on limestone samples permitted to calibrate and modify the apparatus, and showed its very high potential.
This thesis took me about one year, but the results were very positive, and the worthy of publication (dignit? di stampa) was awarded. This, in conjunction with the passion for the Geotechnical Research and the nice atmosphere in the RMRE group encouraged me to continue my studies with a PhD course in Geotechnical Engineering in 2005.
Because the research started during the degree thesis was not concluded and some problems remained to be addressed, in accordance with Prof. G. Barla, the same main subject was chosen for the PhD thesis. Also in this case numerical and experimental works were carried out.
The availability of rock samples obtained from the Saint Martin La Porte access adit of Torino-Lyon Base Tunnel, which experienced very important squeezing problems during excavation, and the peculiar characteristics of the material has determined the choice of coal as the rock material of interest for the experimental study. An experimental program was carried out by using the HPTA apparatus, with attention to strength, deformability and time dependent behaviour. The results obtained permitted to evidence some important characteristics of time independent and time dependent behaviour of coal.
Regarding the numerical research, the evident limits of the Lemaitre's model and the lack of a suitable constitutive model in literature, lead to the formulation of a new viscoplastic constitutive law, called SHELVIP (Stress Hardening ELastic VIscous Plastic) (Debernardi and Barla, 2009). The goal was to propose a rather simple constitutive law that can describes all the most important aspects of time dependency involved in tunnel excavation and can be used with confidence both for research and design practice. The SHELVIP model couples the elastoplastic and time dependent behaviour by using a plastic yield surface, as frequently adopted in tunnel design analysis, and the definition of a state of overstress referred to a viscoplastic yield surface. A stress hardening rule was introduced. The model was developed in all its detailed aspects and implemented into the finite difference code FLAC2D. The SHELVIP model was calibrated by using the laboratory tests performed on coal specimens. It is shown to fit very satisfactorily the experimental results of creep and stress relaxation triaxial tests. Numerical analysis of the Saint Martin La Porte tunnel showed that the SHELVIP model can be used with confidence in order to reproduce by numerical analysis the behaviour of tunnels under severe squeezing conditions.
The results of the PhD thesis, concluded in May 2008, was very positive and very appreciated, but some question remained to be addressed and further developments of the SHELVIP model were needed. The creep failure and the interaction between creep and strength, which are very important for tunnel excavation, need to be introduced into the SHELVIP model. For this reason a fellowship with Politecnico di Torino, co-ordinated by Prof. G. Barla started in April 2008. The final goal is the use of the newly developed SHELVIP model for the study of the mechanized excavation of tunnel in squeezing conditions.
During this year the SHELVIP model was implemented in the code FLAC3D, in order to perform three-dimensional analyses of tunnels, and some modifications in the mathematical formulation of the SHELVIP model were introduced, but not already implemented in the code, to allow the creep rupture to be reproduced.
Regarding the research carried out during these years on viscoplastic behaviour of weak rocks, it is important to observe that the study of the mechanized excavation of tunnels in squeezing conditions is possible only if an adequate constitutive model is used. In fact, only the time dependent characteristics of the rock mass make the excavation of tunnels by means of Tunnel Boring Machine (TBM) still an open problem.