Large time dependent convergences of a tunnel excavated across a carboniferous rock mass are analyzed. These delayed convergences were simulated using an elasto-viscoplastic constitutive model that included a rock strength degradation process. The 3D numerical simulation of the tunnel excavation was performed accounting for both stage construction sequence and rate of excavation advancement. A comparison of the numerical results with the data measured on site allowed for a good calibration of the model parameters.
In this paper, a case study of a tunnel which has experienced large time-dependent displacements due to rock creep and rock damage is presented. This phenomenon has been studied by several authors, for example, Body et al. (2002); San drone et al. (2006); Strep and Gilda (2009); Pellet and Rosebud (2007) and Pellet et al. (2009). Despite these previous studies, there is still a lack of input, especially when comparing numerical results to measured field data. Therefore, additional investigations are still needed to validate the numerical simulations of the observed behaviour of rock masses. In this study, special attention is paid to properly calibrate parameters of the constitutive model.
The tunnel under study was excavated in the Alps for a high speed railway project, as reported by Rettighieri et al. (2008). It was excavated in a highly fractured carboniferous rock formation where large convergences of the tunnel wall, of about 1.5 meter, were observed (Figure 1). Due to the nature of the geological formations, it was foreseen that the principal difficulties in the tunnel excavation would be due to: High overburden stress in the highly disturbed and vectorized geological formation, Rocks prone to creep with poor mechanical properties (squeezing rocks), The tunnel was excavated in the following sequence: 5-meter long sections were excavated using the "drill and blast" method, Light temporary supports made of shotcrete, rock bolts, and steel sets were installed, The next section of excavation was undertaken. The average advancement rate was 1 meter per day. Therefore, the time to complete each cycle was approximately 5 days.
The mechanical properties of the rock mass were assessed based on in-situ investigations (Russ, 2009) and based on rock mass classifications. Geological Strength Index (GI) and Rock Mass Rating (RM) were used to constrain the Hook and Brown failure criterion parameters at the scale of the rock mass. Based on the ISM suggested method (ISM, 2007), the following set of values for the peak strength and for the residual strength were selected.
During the tunnel construction an extensive monitoring program was undertaken. Tunnel convergences of the tunnel wall were measured, and extensometer were used to measure displacements inside the rock mass (Rettighieri et al. 2008). The maximum convergence was more than 70 cm after 80 days. It has to be noted that the displacements increased with time even when the excavation was stopped.