Abstract
The prediction of the advance rate of a Tunnel Boring Machine (TBM) in hard rock conditions is one of the most impor- tant concerns for estimating the time and costs of a tunnel project. In this paper, in the first step, a model based on Rock Engineering Systems (RES) is proposed to predict geotechnical risks (representing media characteristics) in rock TBM tunnelling. Fifteen main parameters that influence the geotechnical hazards were used in the modelling. In establishing an interaction matrix and also a parameter rating, the views of five experts were taken into account. The Vulnerability Index (VI) (geotechnical risk levels) for 2058 datasets out of 2168 sets of data from 53 geological zones in 11 km of the Zagros long tunnel was obtained. In the second step, based on the machine operating parameters such as torque, cutter head rotation per minute, cutter normal force and media characteristics (represented by VIs), which were used as input parameters and advance rate was used as an output parameter, while using 2058 datasets, linear and non-linear multiple regression analyses were carried out. 110 datasets (out of 2168 datasets), which were not used in the modelling, were ap- plied to evaluate the performance of regression models and other models in literature and the results were compared. The obtained results showed that the new linear model proposed with R2=0.83 and RMSE=0.12 has a better performance than the other models.
Highlights
In tunnel construction by Tunnel Boring Machine (TBM), the estimation of time and cost of a project are one of the most important parameters, which has always been a challenge between contractors and owners
The principles of the work carried out by Benardos and Kaliampakos, (2004) were adopted to define a model for predicting the level of geotechnical risks in rock TBM tunnelling. The merit of this modelling is that all of the important and obtainable parameters affecting geotechnical risks in rock TBM were taken into account
The performance of regression models were com(6) pared with the models from literature such as Khademi et al (2010), Hassanpour et al (2009, 2010), QTBM (Barton, 1999), NTNU (Bruland, 2012), Farmer and Glossop (1980), Cassinelli et al (1982), Innaurato et al (1991), and Graham (1977) (Table 14), using 110 randomly selected datasets from the Zagros long tunnel, which were not used in the modelling
Summary
In tunnel construction by TBM, the estimation of time and cost of a project are one of the most important parameters, which has always been a challenge between contractors and owners. Penetration and advance rates are key parameters in the performance estimation of a TBM, which results in estimating the time and cost of a project. Rock Mass Rating (RMR) parameters were employed for predicting TBM performance by Sapigni et al (2002); Bieniawski et al (2007); Khademi Hamidi et al (2010). Gong and Zhao (2007, 2009) and Gong et al (2007) introduced the index of boreability and rock brittleness to investigate the rock mass characteristics that affect the performance of TBM. The objective of the present piece of work is to introduce a new model to improve the accuracy of advance rate prediction in different geotechnical conditions in rock TBM tunnelling, considering both operational parameters of TBM (machine parameters) as well as media characteristics.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
