Abstract

Predicting the penetration rate plays a key role in tunnel projects using a tunnel-boring machine (TBM). Developing accurate prediction models can improve project management, and save budget and time in tunnel projects. In this research, the Gelas water-tunnel project data were used to obtain new statistical models for predicting the TBM penetration rate per revolution (PRev) utilizing the toughness index (T_i), modulus ratio (E/UCS) and joint parameters (J_P). The relationships between various geomechanical properties and rock classification systems, including uniaxial compressive strength, Brazilian tensile strength, Young's modulus, joint parameter, toughness index, rock quality designation, rock mass rating, geological strength index, rock mass quality and rock mass index, were analysed and considered on the TBM performance in sedimentary, igneous and metamorphic rocks. The statistical analysis clearly showed that T_i revealed a significant correlation with the actual PRev (R² = 0.75). In addition, the PRev was computed using T_i, and J_P showed good agreement with the coefficient of determination (R²): i.e. 0.79. The results indicated that the T_i decreased by increasing the modulus ratio, so the PRev increased. This model can be used easily as it provides a straightforward predictive model using a multi-parameter model.