• Geomechanics and Engineering
• /
• v.30 no.2
• /
• pp.107-121
• /
• 2022

Coal pillar assessment is of broad importance to underground engineering structure, as the pillar failure can lead to enormous disasters. Because of the highly non-linear correlation between the pillar failure and its influential attributes, conventional forecasting techniques cannot generate accurate outcomes. To approximate the complex behavior of coal pillar, this paper elucidates a new idea to forecast the underground coal pillar stability using combined unsupervised-supervised learning. In order to build a database of the study, a total of 90 patterns of pillar cases were collected from authentic engineering structures. A state-of-the art feature depletion method, t-distribution symmetric neighbor embedding (t-SNE) has been employed to reduce significance of actual data features. Consequently, an unsupervised machine learning technique K-mean clustering was followed to reassign the t-SNE dimensionality reduced data in order to compute the relative class of coal pillar cases. Following that, the reassign dataset was divided into two parts: 70 percent for training dataset and 30 percent for testing dataset, respectively. The accuracy of the predicted data was then examined using support vector classifier (SVC) model performance measures such as precision, recall, and f₁-score. As a result, the proposed model can be employed for properly predicting the pillar failure class in a variety of underground rock engineering projects.

• Proceedings of the Korean Society for Rock Mechanics Conference
• /
• 2008.10a
• /
• pp.123-134
• /
• 2008

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.9 no.1
• /
• pp.91-97
• /
• 2007

Loads on the pillar and ground deformation in 2-arch tunnel, which is excavated in the rock mass with regular discontinuities, depending on the dips of discontinuities and the construction sequences were experimentally studied. Large scale model tests in the biaxial test facility were performed. Tested model (width 3.3m, height 3.0m, and length 0.45m) for 2-arch tunnel in 1/10 scale were built with various dips. Test results show that discontinuities have greate affects on the behavior of the 2-arch tunnel, especially on the pillar loads and ground deformation.

• Geomechanics and Engineering
• /
• v.10 no.2
• /
• pp.175-188
• /
• 2016

The construction of a new cavern modifies the state of stresses and displacements in a zone around the existing cavern. For multiple caverns, the size of this influence zone depends on the ground type, the in situ stress, the cavern span and shape, the width of the pillar separating the caverns, and the excavation sequence. Performances of underground twin caverns can be unsatisfactory as a result of either instability (collapse) or excessive displacements. These two distinct failures should be prevented in design. This study simulated the ultimate and serviceability performances of underground twin rock caverns of various sizes and shapes. The global factor of safety is used as the criterion for determining the ultimate limit state and the calculated maximum displacement around the cavern opening is adopted as the serviceability limit state criterion. Based on the results of a series of numerical simulations, simple regression models were developed for estimating the global factor of safety and the maximum displacement, respectively. It was proposed that a proper pillar width can be determined based on the threshold influence factor value. In addition, design charts with regard to the selection of the pillar width for underground twin rock caverns under similar ground conditions were also developed.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.15 no.6
• /
• pp.599-612
• /
• 2013

This study aims to investigate an economical efficiency of two excavation methods with respect to the room-and-pillar method for the underground space and conventional excavation method, i.e. 2-arch tunnelling method. For feasibility study, an excavation cost for both room-and-pillar method and 2-arch tunnelling method was estimated when the same space in operation was required. It was assumed that properties of reinforcements and rock were adopted from literatures. However, an excavation shape of the room-and-pillar method was assumed not to be the rectangular shape which is a general type in the room-and-pillar method but to be an arch shape in order to compare with the conventional excavation method (2-arch tunnelling) and to achieve the maximum bearing capacity of the structure during excavation. Consequently, the wider space in use or required and the better condition of rock we assumed, the more economical advantage we have in the room-and-pillar method than the 2-arch tunnelling method.

• Journal of the Korean GEO-environmental Society
• /
• v.12 no.11
• /
• pp.39-49
• /
• 2011

In this paper it was focused on the characteristics of displacement and stress due to the construction of diagonal twin tunnel. In this research, the characteristics have been analyzed with the presumption that the tunnel's diameter (D) is 13m and the ground was formed by weathered rock. In analysis, the width of pillar is 2.0D, 2.5D, 3.0D, the height of soil cover is 3D, 4D, 5D, and the installation angle of diagonal twin tunnel has been changed into $15^{\circ}$, $30^{\circ}$, $45^{\circ}$. And the program used in this analysis is FLAC which is widely used in solution problems of ground engineering in order to gain and analyze occurring shotcrete and rockbolt stress and nearby ground displacement according to pillar width, the height of soil cover and installation angle of diagonal twin tunnel. As a result, in the weathered rock grounds, when the width of pillar is more than 2.0D, the height of soil cover is over 5.0D, and the installation angle of diagonal twin tunnel is lower, they lessen effects on the ground and favorable on the stability of tunnel.

• Proceedings of the Korean Society for Rock Mechanics Conference
• /
• 2004.10a
• /
• pp.171-180
• /
• 2004

Underground mining causes surface subsidence long after the mining operation had been ceased. Surface subsidence can be in the form of saucer-shaped depression or collapsed chimneys or sinkholes. Sinkhole formations are predominant over shallow-depth room and pillar mines having weak overburden strata. In this study, occurrences of sinkholes due to mining activity are assessed based on local geological conditions and mining parameters using fuzzy reasoning techniques. All input and output parameters are represented with linguistic hedges. Numerous fuzzy rules are developed to relate sinkhole occurrences with input parameters using fuzzy relational matrix. Based on the combined fuzzy rules, possibility of sinkhole occurrences can be ascertained once the geological and mining parameters of any area are known.

• Journal of the Korean Geotechnical Society
• /
• v.25 no.5
• /
• pp.65-73
• /
• 2009

Large scale model tests and numerical analyses are performed to investigate the stress distribution of pillar due to surface loading nearby two-arch tunnel which is constructed in the regularly jointed rocks. It is observed that the influence of discontinuities on the stress distribution in the discontinuous rock mass and the underground stresses induced by surface loading are greater than those of linear elastic theory. Especially, lines of equal stresses are developed to the direction of inclination according to the inclined grade. In cases of discontinuities imbedded in parallel with or vertical to the ground, the pressure bulbs are formed symmetrically, however, the inclined ones result in stress distribution in parallel with and vertical to the planes of discontinuities. Results indicated that stress distribution is seriously affected by the angle of discontinuity. When stresses propagating to the pillar need to be estimated, relative location of surface loading, grade of discontinuous plane, and location of two-arch tunnel should be carefully considered.

• Tunnel and Underground Space
• /
• v.23 no.2
• /
• pp.130-140
• /
• 2013

In this study, scaled model tests were performed to investigate the stability of twin tunnels with small clearance, where the pillar widths were 0.5D and 0.25D, respectively. The tunnels were supposed to be constructed in anisotropic weathered rocks with $30^{\circ}$ inclined bedding planes, and the model tests were conducted under the condition of lateral pressure ratio, 1. Six types of test models which had respectively different pillar widths and support conditions were experimented, where crack initiating pressures, maximum pressures, failure modes of pillar and deformation behaviors around tunnels were investigated. The models with wider pillar were cracked under higher pressure than the models with shallower pillar. The models with lining support were cracked under higher pressure and showed less tunnel convergence than the unsupported models. The models with both lining and pillar reinforcement were proved to be most stable among the tested models. In particular, as the model of 0.25D pillar width with only lining support showed shear failure of pillar according to the existing bedding planes, so both lining and pillar reinforcement were thought to be indispensable in that case of tunnel.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.13 no.2
• /
• pp.115-131
• /
• 2011

Recently, considering the aspects of disaster prevention and environmental damage, the construction of a twin tunnel is increasing. When constructing a twin tunnel, the stresses are concentrated at the pillar so that stability of the tunnel is decreased. Since the previous studies on the behavior of a twin tunnel pillar are mainly restricted to the estimation of the tunnel behavior and the analysis of surface settlement, there is a limit to a quantitative stability estimation of the pillar. Therefore, it was quantitatively investigated how the pillar width of a twin tunnel affects its stability. To ensure this end, global tunnel safety factors obtained numerically using shear strength reduction technique, local safety factors of a pillar using the equation that Matsuda et al. suggested, and strength/stress ratios of the pillar were estimated and their results were analyzed for two sections with different rock covers. For a reasonable design of a twin tunnel pillar, it was turned out that strength/stress ratio, the local pillar safety factor, and global tunnel safety factor should be used interrelatedly rather than independently.