Abstract
Various models have been proposed to estimate the TBM penetration rate. Generally, the input parameters of these models can be divided into two categories: Machine parameters and geological engineering parameters. The engineering geological parameters will significantly influence the penetration rate if the machine operational parameters are kept within a reasonable near-optimal range. However, while some performance prediction models can be used for many common project settings, they have lower accuracy in certain applications. This study compared the observed penetration rate of a hard rock TBM with those predicted by MCSM, Norwegian University of Science and Technology (NTNU), Farrokh–Rostami, and Ramezanzadeh’s models in the Kerman water tunnel (KWT). Next, joint survey data are collected from the different zones of the KWT. For this purpose, a 3D discrete fracture network (DFN) model code was generated in Mathematica© version 13. The joint data’s orientation, persistence, and spacing were used to develop a 3D-DFN model for estimating the blockiness rate (BR) index. The BR index is the actual joint intensity in 2D (P21) and 3D (P32). In this study, the BR index is the newest rock mass parameter introduced and used to predict the penetration rate of TBM. This index can serve as a rock mass parameter that provides excellent and realistic results for predicting penetration rate (PR). The corresponding determination coefficient values of the PR with P32 and P21 are R2 = 0.96 and R2 = 0.98, respectively, and with CIA and UCS, are R2 = 0.42 and R2 = 0.49, respectively. Furthermore, using the DFN model showed its potential to be an accurate and reliable method for the overall estimation of the in-situ rock mass fragmentation, which highly controls the penetration rate of TBM.