Geomechanical characterisation of massive rock for deep TBM tunnelling

Abstract

A combined geological and rock mechanics approach to tunnel face behaviour prediction, based on improved understanding of brittle fracture processes during TBM excavation, was developed to complement empirical design and performance prediction for TBM tunnelling applications in novel geological conditions. A major challenge of this research is combining geological and engineering languages, methods, and objectives to construct a unified geomechanics characterisation system. The goal of this system is to describe the spalling sensitivity of hard, massive, highly stressed crystalline rock, often deformed by tectonic processes. Geological, lab strength testing and TBM machine data were used to quantify the impact of interrelated geological factors, such as mineralogy, grain size, fabric and the heterogeneity of all these factors at micro and macro scale, on spalling sensitivity and to combine these factors within a TBM advance framework. This was achieved by incorporating aspects of geology, tectonics, mineralogy, materials strength theory, fracture process theory and induced stresses. adequate prediction of rock mass behavior in response to tunnelling for each geological domain within the geometrical and mechanical tunnelling framework (Kaiser 2005). The need for quantifying geological descriptions for engineering geology applications such as open pit mine wall stability (Hoek 1999) and deep, hard rock, tunnel stability (Kaiser 2005) has been demonstrated but to effectively accomplish this, the rock behaviour and response must first be understood in order to define the values of importance for quantification (Kaiser 2005). In-situ behaviour can vastly differ from laboratory behaviour, depending heavily on textural properties (Diederichs et al. 2004), making understanding rock behaviour at the excavation boundary critical to properly quantifying geological characteristics. The development of the characterisation scheme followed the procedure outlined in Figure 1. The goal for this characterisation scheme was the development of a tool by which geological characteristics could be translated into indicators of susceptibility to spalling failure at the TBM cutter (small) and tunnel face (large) scales (Fig. 2). The ranges of rock types considered are massive rocks, with Rock Mass Rating (RMR) greater than 75 and Geological Strength Index (GSI) greater than 70. The characterisation scheme focuses on rock behaviour and response that leads to spalling-type yield, where sudden failure is induced through intact rock at the excavation boundary (Fig. 2). Existing methods, such as RMR (Bieniawski 1989), and Q (Barton et al. 1974), were developed to address conventional blocky ground issues of support and excavation in zones of raveling of the wall and face due to jointed rock masses. Several methods for TBM design, such as QTBM (Barton 2000) and the NTNU method (Bruland 1998), and TBM analysis methods developed by Buchi (1998) incorporate rock mass characteristics. Non spalling-type failure through massive rock, resulting from shear failure, does not occur by the same failure mechanism that leads to spalling-type failure. For this type of failure, conventional rock characterization, such as UCS, Cerchar Index (Dollinger et al. 1999) etc, have been shown to be sufficient for TBM design methods such as the CSM method (Bruland 1998, Rostami et al. 1996). Geological tools employed in this research include petrographical and textural rock description at the thin section, hand sample, and tunnel scale. Engineering tools include point load testing of drill core, rock mass classification (RMR and GSI), and TBM performance (penetration rate and thrust magnitude) data analysis. Materials science principles relating crystal deformation, and stiffness and strength properties (Illston et al. 1979, Nicolas & Poirier 1976) provided direction for interpreting the impact of tectonic deformation on rock strength. Figure 1. Characterisation scheme development begins with identification of the goal, rock mass and yield scope, determination of appropriate approach, and testing and calibration. Bruland 1998; Barton 2000; Dollinger et al. 1999; Rostami et al. 1996. 3 A GEOMECHANICAL CHARACTERISATION SCHEME FOR MASSIVE, HARD ROCK The geomechanical characterisation scheme was constructed to translate information available through geological description into information that relates directly to rock behaviour, focusing on spalling sensitivity as it impacts chipping and face instability. The flowchart in Figure 3 shows how geological sample characterisation factors are combined to obtain estimates of spalling sensitivity and fracture potential, which are used to make interpretations about the behaviour of the rock at the excavation boundary, either at the cutter or the tunnel face scale. A description of the development of the characterisation scheme summarized in Table 1 is presented in Villeneuve et al. (in prep). Figure 2.Schematic diagram illustrating the failure mode areas of focus. Top: schematic penetration rate vs gross thrust graph shows two separate processes during TBM cutter excavation: grinding at low thrust and penetration rate, versus chipping (a process akin to spalling) at high thrust and penetration rate. Bottom: tunnel cross sections demonstrate wall and face failure mechanisms; from top left: blocky ground resulting from discontinuities, squeezing due to shearing in low competence rock masses with respect to induced stress, spalling in the wall and/or face, depending on rock mass characteristics and induced stress geometry, and stress-fabric interaction inducing block formation and instability in the face. All designations for low, medium and high, denoting relative impact on fracturing and spalling behaviour, are related to the cutter-rock interaction and face instability realm. Low impact indicates characteristics that are unfavourable to spalling, and high impact indicates characteristics that promote spalling. This characterisation only suggests sensitivity to spalling and fracture potential, since the manifestation of spalling during TBM excavation also depends on the interaction of the tunnel, anisotropy and induced stress geometries specific to each tunnelling situation. 4 RELATIONSHIP BETWEEN THE CHARACTERISATION SCHEME AND GEOLOGICAL DESCRIPTION/HISTORY Testing and calibration of the geomechanical characterization is demonstrated by combining available geological, engineering and mechanical Figure 3. Characterisation schematic showing data collection, classification and combination to obtain fracture potential. Legend: FMM – mineralogy major; FMA – accessory minor; FM – mineralogy; FGP – grain size petrological; FGT – grain size tectonic; FGD – grain size distribution; FG – grain size and grain size distribution; FAF – fabric type; FAD – fabric scale; FA – anisotropy; FSS – spalling sensitivity; FSSA – spalling sensitivity with anisotropy; FFI – isotropic fracture potential; FFA – anisotropic fracture potential Table 1. Description of geological characterisation factors