"Rock tests are performed before the start of every mining or civil engineering project as part of a detailed feasibility study. The feasibility study is costly and it comprises drilling, sample collection, sample handling and laboratory testing. Numerical modeling techniques, such as Particle Flow Code (PFC), can be used to provide reliable estimates of rock strength values. The numerical models for unconfined compressive strength (UCS), direct tension, and Brazilian tests were developed in PFC, and validated using data from literature. A particle size range of 3-5 mm with Dmax/Dmin = 1.67 gave the best results. The numerical errors were in the range of 6-22% for UCS, 21-80% for direct tension, and 5- 10% for Brazilian tests. About 1,800 confined compression tests were also performed in PFC to obtain formation material properties. However, the PFC algorithm takes a very long computational time to complete the process, and thus, there is a need for more efficient and faster methods. In this research, the author uses artificial intelligence methods including, Artificial Neural Network, Mamdani Fuzzy Logic, and Hybrid neural Fuzzy Inference System (HyFIS) to solve this problem. These methods, along with the Multiple Linear Regression method, were used for the predictive analysis. Based on R2 and RMSE statistics for the testing phase, HyFIS is the best predictive model. This study is the first attempt to develop self-learning artificial intelligent models for predicting formation material properties. In addition, this research study investigates the shovel excavation process using the discrete element technique in PFC to examine the shovel digging phase. The shovel excavation simulator provides a tool for optimizing strategies for maximizing its performance that provides a major breakthrough in the shovel excavation frontier"--Abstract, page iii.
Alagha, Lana Z.
Awuah-Offei, Kwame, 1975-
Ph. D. in Mining Engineering
Missouri University of Science and Technology Department of Mining and Nuclear Engineering
Ma'aden Phosphate Company
Saudi Mining Polytechnic (SMP)
Itasca Consulting Group
Missouri University of Science and Technology
xv, 258 pages
© 2018 Muhammad Waqas, All rights reserved.
Dissertation - Open Access
Electronic OCLC #
Waqas, Muhammad, "Discrete element and artificial intelligence modeling of rock properties and formation failure in advance of shovel excavation" (2018). Doctoral Dissertations. 2731.