An assessment of the relation between monitored settlements and the face stability of a slurry-TBM

Abstract

During the boring of a tunnel in soft soils with a slurry TBM, support pressure is used to achieve equilibrium at the face of the TBM. When this equilibrium is not reached, when the face support pressure is too low or too high, settlements will occur. In this research settlement and pore water pressure measurements are used to monitor the behavior of the soil and estimate the stability of the tunnel face. During boring of the tunnel, the exact stability of the face is not known. The TBM driver has to rely on the provided stratigraphy data, the advised face support pressures range provided by the geotechnical engineers and the experience of the tunnel boring team. Monitoring is not yet used to determine the face stability during construction. To do so, field data from a case study at RijnlandRoute is compared with analytical and numerical models. Sensitivity of the measurement equipment, and of both the analytical (DIN) and numerical model (Plaxis 3D) with respect to soil parameters, are considered. It has been found that the strength parameters of the layer in which the face is located have the highest influence on the minimum face support pressure. For the maximum face support pressure this is the volumetric weight of the entire soil profile above the face. For comparing the (soft soil) field data with numerical results, a Plaxis 3D model is built, and it is determined that HSsmall is a suitable constitutive model to capture the interaction between face stability, tunneling operations and soil behavior. It is shown that the tail void injection influences the settlements above and in front of the cutter head, but this influence is discarded and replaced by a wished in place lining, to simplify the numerical model. A scenario analysis on the sensitivity of soil parameters shows settlements do not vary significantly between the characteristic low and high values. In this analysis correlation of parameters is taken into account. The failure mechanism for the minimum face support pressure coincides with the active cave-in failure mechanism found in literature. For the maximum face support pressure the failure mechanism found in Plaxis 3D does not coincide with the expected hydraulic fracturing failure mode. The continuum representation of the soil in Plaxis 3D does not allow a hydraulic fracturing like failure mechanism to develop. Instead, a blow-out approximately 10 m. in front of the cutterhead occurs. This behaviour better resembles the failure mechanism of a EPB TBM. The field data gathered from the case study shows a thrust wave in front of the cutter head in both settlement and pore water pressure measurements. This thrust wave reaches up to 20 to 40 meters in front of the TBM. The heave induced by the thrust wave reduces the amount of settlements after the TBM passage. Excess pore pressures, induced by a high thrust wave, affect the face stability negatively. The excess pore pressure mainly depends on the advance rate. The higher the advance rate, the less time the pore pressures have to dissipate, leading to an increase in excess pore pressure. The accuracy of the settlements measurement devices is 0.8 mm., and of the spade cells 1.0 kPa. In general, the field data shows settlement curves corresponding to the Peck (1969) Gaussian curve (in lateral and longitudinal direction). Comparing the case study settlements with the numerically generated settlement curves show similar trends. The field data shows lower settlements than the numerical results. This can be due to the presence of excess pore pressures, the accuracy of the TBM data or its interpretation. A method to increase the accuracy, which is expected to result in a better fit with the numerical results, was found in a late phase of the research. This method takes into account the settlements which are induced by the thrust wave in front of the TBM. Comparing the numerical and analytically determined limit support pressures, it is found that the minimum face support pressure are similar in both methods. As similar failure mechanisms are found, numerical modelling seems a reliable way of determining the face stability. However, due to the limited range of applied support pressures available from the TBM data set, limit states could not be fully analyzed. To assure the reliability of a numerical model to determine the actual face stability based on surface settlements during the construction phase, additional research must be done. It is suggested to extend the methods used in this research with physical modelling. For maximum face support pressures, the analytical and numerical models do not coincide. Hydraulic fracturing cannot be modelled in Plaxis 3D. It is recommended not to use numerical modelling to determine the face stability based on settlements at face pressures higher than the face pressures at which an equilibrium is achieved.