A mechanical method for predicting TBM penetration rates

• Published in: Arabian journal of geosciences 2020-05, Vol.13 (9), Article 335
• Main Authors: Wang, Ruirui, Guo, Xu, Li, Jianbin, Wang, Jian, Jing, Liujie, Liu, Zhengyu, Xu, Xinji
• Format: Article
• Language: English
• Subjects: Boring machines; Control methods; Correlation coefficient; Correlation coefficients; Design parameters; Dredging; Drilling & boring machinery; Earth and Environmental Science; Earth science; Earth Sciences; Excavation; Mathematical models; Mechanical analysis; Model accuracy; Original Paper; Parameters; Penetration; Predictions; Rivers; Rocks; Tensile strength; Torque; Tunnel construction; Tunnels; Water supply
• ISSN: 1866-7511; 1866-7538
• DOI: 10.1007/s12517-020-05305-x

In the tunnel boring machine (TBM) excavation process, the accurate prediction of the TBM performance, especially the penetration rate, is of great significance to the in time planning, cost control, and safety judgment. In this paper, we propose a prediction method and its corresponding model of penetration rate based on a mechanical analysis. The variables of the method included the design parameters (tunnel radius and cutter arrangement), the operating parameters (the thrust, torque and revolution per minute), and the tensile strength. The rock breakage depth of a single cutter was calculated using the force balance. Then, the rock breakage volume per revolution was obtained by summing up the rock breakage of each cutter based on the cutter arrangement. Finally, based on the premise of knowing the rock breakage volume per revolution, the numerical solution of the penetration rate was acquired. On the basis of the field data from the 4th Section of the Water Supply Project from Songhua River, the accuracy of the model is verified by 59 samples, with a correlation coefficient of 0.64 between prediction and actual results and a mean absolute percentage error of 14.6%. Furthermore, a basic form of the PR prediction equation suitable for different projects was proposed using the single variable control method. For CREC188 TBM in particular, the undetermined coefficient values of the equation were determined using 7-fold cross-validation and the calculation equation was verified with another 50 samples with a correlation coefficient of 0.78 and a mean absolute percentage error of 10.9%.