ISRM 2023 - PUBLICATIONS

T11: NATM versus TBMShengfeng HUANG, Misagh ESMAEILPOUR, Saadeldin MOSTAFA, Pooya DASTPAK, Rita SOUSAStevens Institute of Technology, United States of AmericaTBM penetration rate prediction using machine learning models and models’ generalizationMost existing models about penetration rate (PR) prediction have been developed and validated against data from one single project. This poses the question whether these models can perform well when faced with new data. We use two datasets of two tunnels built with the same construction method and in similar geological conditions. Different machine learning (ML) models are trained, validated, and tested with dataset from one tunnel and then generated to the other dataset. Additionally, the effect of several data processing techniques for splitting and scaling on the performance and generalization of the different models is tested. The results demonstrate that random forest (RF) and extreme gradient boosting (XGBoost) exhibit better performance than other models. Regarding generalization, CART and XGBooost model exhibit the best performance. The impact of splitting and scaling techniques on the generalization of the models becomes noticeable than on the performance of models. | Penetration rate, EPBM, Machine learning, Generalization, split and scale technique
T11: NATM versus TBMHoyoung JEONG (1), Yudhidya WICAKSANA (2), Sehun KIM (3), Seokwon JEON (3), Jeong-Gi UM (1)1: Pukyong National University, Republic of Korea (South Korea); 2: Bandung Institute of Technology, Indonesia; 3: Seoul National University, Republic of Korea (South Korea)Cutting performance evaluation of Actuated Disc Cutting by linear cutting testThis paper presents an investigation into Actuated Disc Cutting (ADC). The ADC is based on the undercutting mechanism and involves several cutting parameters. To enable an experimental analysis of ADC, a new lab-scaled rock-cutting system is introduced. This system facilitates the undercutting concept of rock cutting based on the traditional Linear Cutting Machine (LCM) system. The cutting parameters in ADC, such as eccentricity, linear velocity, and RPM are simulated by the introduced cutting system. The paper also introduces calculation methods for cutter forces and specific energy of ADC based on the system. A series of undercutting tests were performed on a designed hard rock specimen under different cutting conditions. The effects of cutting parameters on cutter forces and specific energy were discussed, and the optimal cutting conditions in ADC were analyzed. The proposed lab-scaled rock-cutting system and calculation methods can be utilized for future studies in this field. | Actuated disc cutting, undercutting, LCM, Cutting efficiency
T11: NATM versus TBMHan-Young JEONG (1), Mun-Gyu KIM (1), Yun-Joo NAM (1), Min-Gi CHO (1), Jung-Woo CHO (1), Sung-Bo SHIM (2)1: KITECH (Korea Institute of Industrial Technology), Republic of Korea (South Korea); 2: Kyungpook National University, Daegu, Republic of KoreaCompression test of TBM thrust jack for validation of bucking strength under inclined loading conditionThis paper studies a consecutive excavation technique to enhance the shield TBM advance rate. Helical type segments are considered for one solution for the continuous excavation of shield TBM. When using helical segments, the thrust jacks compress on the inclined side wall of the segments. The inclined contact can induce the side force and bending moment of the hydraulic cylinder, which can arouse oil leak or plastic deformation of rod and cylinder parts. The study examined the buckling reliability of thrust jack against to the compressive force on the inclined surface of helical segment. A numerical simulation was conducted for analyzing the structural stability. The reliability testing code was established and the testing specimen of jack and loading jigs were manufactured for the compression test of the jack. After the compression test was conducted, the testing results are summarized for validation. | Shield TBM, Thrust jack, Helical segment, Inclined surface, Compression test
T11: NATM versus TBMMun-Gyu KIM (1), Han-Young JEONG (1), Joo-Young OH (1), Jung-Woo CHO (1), Sang-Hwa YU (2), Ho-Young JEONG (3)1: KITECH (Korea Institute of Industrial Technology), Republic of Korea; 2: Hochang Engineering Co., Republic of Korea; 3: Pukyong University, Busan, Republic of KoreaPerformance validation of rock cutting-splitting method by scaled model tests and rock slope excavationA cutting-splitting method is developed as a mechanized excavation for rock slope and tunnelling. The technique has two procedures. The first step is the rock cutting as designed dimensions (i.e. depth, spacing, width). The second is the rock splitting process by inserting chisel into the cutting slot to split out a rock block. For propagation of tensile crack on the bottom of rock, a sufficient indenting force should be loaded on the cutting slot. An analytic solution was proposed to estimate the indentation force. Numerical analyses were conducted to analyze the tensile fracturing process during the splitting process. A series of scaled model tests were designed and conducted to evaluate the advantageous geometry for the method. Finally, in-situ excavation tests were carried out for a rock slope. The results were compared to the previous method (i.e., drilling-splitting method). It is concluded that the developing method can improve the cutting rate of hard rock. | non-vibrating, cutting-splitting, numerical analysis, scaled model test, rock slope excavation