Search

Slope failure is a significant risk in both civil and mining operations. This failure phenomenon is more likely to occur during the high rainfall season, areas with a high probability of seismic activity and in cold countries due to freezing-thawing. Further, a poor understanding of hydrogeology and geotechnical factors can contribute to erroneous engineering designs. Several Limit Equilibrium Methods (LEMs) and numerical modelling tools have been developed over the years. However, the highlighted success of the Artificial Neural Networks (ANNs) in other disciplines/sectors has motivated researchers to implement ANNs to forecast the Factor Of Safety (FOS). This paper aims to develop ANNs to predict the value of the FOS for slopes formed by (i) uniform one soil/rock material and (ii) formed by two soil/rock materials. Each of these slopes contains three sub-models with 6, 7 and 8 input material parameters. Thousands of FOS values were generated for each sub-model using LEMs by randomly generating material input parameters. Over 80% of generated FOS values were used to train ANNs and the remaining 20% were used to for validation. The one-material models performed better than the two-material models overall. The first sub-model from the one-material models and the third sub-model from the two-material models exhibited the best performance compared to the other sub-models, achieving Mean Square Error (MSE) of 8.35E-04 and 5.10E-3, respectively. The third sub-model from the one-material models and the first sub-model from the two-material models have a MSE of 2.00E-3 and 9.80E-3, respectively. The second sub-models have shown the lowest performance compared to the other models. The minimal errors between LEMs and ANNs have led to the conclusion that ANN can be used as a tool for a quick and first-pass analysis by design engineers without undertaking rigours, complex, time-consuming and tedious computation of FOS using LEMs. An actual field-tested database can be usedto p

The presence of inherent discontinuities within a rock mass poses significant influence on its shear strength-deformation characteristics. Therefore, it is important to study the rock joint performance within the laboratory for the safe and economical design of underground structures (such as mine roadways) in jointed rock mass, stability analysis of jointed rock slopes and foundation design on a fractured rock mass. To study the rock joint mechanics and principles governing its shear behaviour, this research is focused on the behaviour of natural rough rock joint within the laboratory under Constant Normal Load (CNL) and Constant Normal Stiffness (CNS) boundary condition using servo-controlled direct shear apparatus at the University of Wollongong, Australia. It was observed from the experimental results that the CNS testing procedure truly simulates the shearing mechanism of an actual field unlike CNL, and the shear strength evaluated rock joint is underestimated under the CNL boundary condition. Also, the shear strength envelope under CNS exhibits non-linearity in contrast with the bilinear strength envelope under CNL boundary condition.

This comprehensive technical book on highwall mining covers theory and practice coupled with practical examples and design aspects. It contains eight extensive chapters elaborating broad-spectrum functionalities of highwall mining and its operational aspects, covering world scenario, economic potential, methods of coal extraction, design methodology including empirical web pillar design, numerical modelling for stress analysis, safety factor for web pillars, panel and barrier design, small-and large-scale numerical modelling, multiple seam interaction and design, coal web pillar strength, equivalent width concept, laboratory testing, new web pillar strength formula, effect of weak bands in coal seam, slope stability, safety and ground monitoring, hazards and regulatory requirements, case examples, norms and guidelines for practice. It also summarizes the results of research carried out by the CSIR Central Institute of Mining and Fuel Research (CSIR-CIMFR), India and the Commonwealth Scientific and Industrial Research Organisation (CSIRO), Australia on the subject.