Pillar strength in underground stone mines in the United States Esterhuizen, G.S.; Dolinar, D.R.; Ellenberger, J.L.
International journal of rock mechanics and mining sciences (Oxford, England : 1997),
2011, 2011-1-00, 20110101, Volume:
48, Issue:
1
Journal Article
Peer reviewed
Stone mines in the Eastern and Midwestern United States make use of the room-and-pillar method of mining to extract relatively flat-laying sedimentary formations. A survey of pillar performance was ...carried out to identify potential modes of instability. Pillars were found to have been successful in providing support to the overburden, but a small number of individual failed pillars were observed. Failure of the pillars was observed to be related to spalling of the hard brittle rocks, shearing along pre-existing angular discontinuities or progressive extrusion of soft infill materials on bedding planes. A method of estimating the pillar strength and selecting a safety factor for design was developed based on observations of stable and failed pillars, supplemented by numerical models. The developed pillar strength equation can be used to design stable stone mine pillars provided the rock conditions are similar to those included in the study.
The prediction of pillar stability (PS) in hard rock mines is a crucial task for which many techniques and methods have been proposed in the literature including machine learning classification. In ...order to make the best use of the large variety of statistical and machine learning classification methods available, it is necessary to assess their performance before selecting a classifier and suggesting improvement. The objective of this paper is to compare different classification techniques for PS detection in hard rock mines. The data of this study consist of six features, namely pillar width, pillar height, the ratio of pillar width to its height, uniaxial compressive strength of the rock, pillar strength, and pillar stress. A total of 251 pillar cases between 1972 and 2011 are analyzed. Six supervised learning algorithms, including linear discriminant analysis, multinomial logistic regression, multilayer perceptron neural networks, support vector machine (SVM), random forest (RF), and gradient boosting machine, are evaluated for their ability to learn for PS based on different input parameter combinations. In this study, the available data set is randomly split into two parts: training set (70 %) and test set (30 %). A repeated tenfold cross-validation procedure (ten repeats) is applied to determine the optimal parameter values during modeling, and an external testing set is employed to validate the prediction performance of models. Two performance measures, namely classification accuracy rate and Cohen’s kappa, are employed. The analysis of the accuracy together with kappa for the PS data set demonstrates that SVM and RF achieve comparable median classification accuracy rate and Cohen’s kappa values. All models are fitted by “R” programs with the libraries and functions described in this study.
The stability of crown pillar is critical during the transition from open pit to underground mining. Mining-induced fractures in the pillar form water seepage channels, which can cause potential ...water inrush hazards. A microseismicity-based method to establish seepage channel network and assess damage state of rock mass in the pillar is proposed. The formation processes of seepage channels and associated rock failure mechanism were analyzed. First, the spatiotemporal evolution of the microseismic (MS) events was presented, based on which the development process of the fractured zone was determined. Second, moment tensor inversion (MTI) was utilized to interpret the focal mechanism of the MS events. A 3D rose diagram was utilized to measure the fracture orientations and determine the main fracture surfaces, and a fracture network was subsequently established. Meanwhile, the distribution characteristics of the fracture radii and volumes were discussed. The results show that shear fractures were dominant in pillar and accounted for more than 90% of all MS events. The overall damage tensor of the pillar was subsequently assessed based on the MS-derived fractures, and the maximum damage direction was determined. Third, a fast chronological expansion method was proposed to iteratively build a connected network with a combination of MS event locations and the corresponding fracture orientations. The MS-derived connected network was used to estimate the distances of event-to-event seepage, from which the shortest seepage channel from each individual event to the network was determined in chronological order. Seepage channels between hydraulic recharge and discharge points were inferred. These results could be helpful for better characterization of seepage channel development and the implementation of pillar reinforcement.
This paper presents an integrated approach for field test and numerical modelling to investigate the relationship between gateroad stability and yield pillar size. The test site is located at ...Yuncheng city, Shanxi Province, China. Field tests indicated that when the yield pillar width was 17 m, the total convergence of the roof, yield pillar rib and virgin coal rib were 882, 587 and 352 mm, respectively, and severe roof sagging and yield rib spalling occurred during the panel retreat. A meticulously validated numerical model, incorporating a double-yield model for the gob materials and calibrated parameters, was developed to investigate the stress changes and yield zone distribution across the yield pillar with different sizes. The results of the simulation indicate that a yield pillar 17 m wide puts the gateroad in a high-stress environment; conversely, a yield pillar 8 m wide is subjected to a relatively low load and puts the gateroad in a good stress environment. Consequently, the rational yield pillar width was estimated at 8 m, and a support strategy was proposed. Field measurement data demonstrate that the newly designed pillar size and support pattern can efficiently ensure gateroad stability. The proposed numerical simulation procedure and calibrated method could be a viable alternative approach to yield pillar design. In addition, the design principle and support strategy for the yield pillar presented in this study can potentially be applied to other similar projects.
The influence of the strength of the pillar on the bearing characteristics of the backfill‐pillar system (BPS) based on the mechanism of backfill. Quantitative synergy relationships between backfill ...and pillar were defined and pointed out. There are four kinds of quantitative synergy relationships in the bearing process of the system. To verify the relationship, systems with different strengths of the pillar were selected to carry out numerical simulation experiments, and a mathematical calculation method to quantify synergy relationship in the BPS was designed. The results show that the system under the low strength of the pillar presents the overall increase in bearing characteristics, while the system under the high strength of the pillar presents “N”‐type bearing features. Four kinds of quantitative synergy relationships able to characterise the action mechanism between backfill and pillar in each stage of the bearing of the system. Through analysis and discussion, it was concluded that the essential reason for the difference in bearing characteristics is that the effect of displacement restraint, joint support, and passive compressive weaken with the increase of the strength of pillar. The effect of pillar strength on the internal synergy of BPS is well verified by numerical simulation test.
Two single-entry gateroad systems employing a yield pillar for bump control in a Chinese coal mine were introduced. The overburden depth of the longwall panels was approximately 390 m. When the ...width/height (W/H) ratio of the yield pillar was 2.67, coal bumps in the tailgate occurred in front of the longwall retreating face. However, in another panel, the coal bump was eliminated because the W/H ratio was reduced to 1.67. Under this condition, instrumentation results indicated that the roof-to-floor and rib-to-rib convergences reached 1,050 and 790 mm, respectively, during longwall retreat. The numerical model was used to back-analyze the two cases of yield pillar application in the hope to find the principle for yield pillar design. In order to improve the reliability of the numerical model, the strain-hardening gob and strain-softening pillar materials were meticulously calibrated, and the coal/rock interface strength was determined by laboratory direct shear tests. The results of the validated model indicate that if the W/H ratio of the yield pillar equals 1.67, the peak vertical stress in the panel rib (37.7 MPa) is much larger than that in the yield pillar (21.1 MPa); however, the peak vertical stress in the panel rib (30.87 MPa) is smaller than that in the yield pillar (36 MPa) when the W/H ratio of yield pillar is 2.67. These findings may be helpful to the design of yield pillars for bump control.
An investigation is made of the characteristic strata movement and mechanism underlying fault–pillar induced rock bursts (FPIRBs) in order to mitigate rock burst damage in fault areas. A mechanical ...analysis of the fault–pillar model is established and roof rotation criteria is obtained. A formula is derived for the average static stress in the pillar through theoretical analysis, physical simulations, and engineering practice. The results show that when a coalface approaches a fault area, two or more roof strata simultaneously fracture in the fault area, leading to an increase in the dynamic and static stresses in the pillar. The most important factors affecting FPIRB are the static stress in the pillar and the dynamic stress induced by fault slides. The roof block rotates more easily when the pillar width is smaller, the roof thickness larger, and the roof subsidence smaller. The average static stress in the pillar increases with decreasing pillar width and/or increasing roof fracture length. The stress is greater if there is a voussoir beam structure, in which case the stress is directly proportional to the squared length of the fractured roof, and inversely proportional to the squared width of the pillar just before rotation occurs. After rotation, it is directly proportional to the roof fracture length and inversely proportional to the pillar width. Based on the FPIRB mechanism and analysis of the mechanical model, six methods of FPIRB prevention are proposed. Also, we find that FPIRB occurrence can be effectively reduced by the use of de-stress blasting and large diameter drilling.
•Two or more roof strata fracture simultaneously in the fault area.•The main factors are the dynamic and static stresses in the pillar.•Smaller pillar widths facilitate easier roof rotation.•Larger roof thicknesses and smaller roof subsidence favor easier roof rotation.•The average stress rises as the pillar width drops and roof fracture length rises.
A reconfigurable, droplet‐directing surface is developed based on high‐aspect‐ratio shape‐memory polymer (SMP) pillars. The water droplet on the original or recovered SMP pillars can slide off the ...surface at a finite angle of inclination while being fully pinned on the deformed pillar array. This wettability contrast allows directed water shredding from the straight pillars to the deformed ones.
This paper describes the results of a back analysis of pillar failures at Troy Mine, Montana, and the use of this experience to make forward predictions on pillar stability in the nearby Montanore ...deposit which lies in a similar geomechanical setting. At Troy Mine, a progression of pillar failures in areas within the Middle Quartzite of the Revett formation led to the observed surface subsidence. The Troy Mine experience was used to understand the level of stresses and failure mechanism leading to the collapse of some pillars in the North Orebody to estimate pillar strength in quartzite beds within Troy’s mountainous terrain. The model elucidated that the dipping orebody geometry in relation to topography led to shear stresses in pillars at Troy Mine. Shear stresses resulted in significant loss of confinement in pillar cores (many theoretically in tension), even at width-to-height ratios that would be deemed stable under zero shear stress (flat seam under flat topography). A calibrated model was achieved, which allowed us to evaluate the impact that different pillar geometric characteristics (such as width, length, height, and shape) have on pillar performance under shear conditions for different depths and extraction ratios. Design charts were then generated to provide guidance on pillar geometry based on expected demand. Mine-wide models were developed to predict the level of vertical stress and horizontal shear stress for pillars in the different ore-bearing beds at Montanore. A sensitivity study was performed for various conditions, including extraction ratio, spatial location under the mountainous terrain, and local orebody geometry with the aim of performing a mine-wide evaluation of the factor of safety against shear. The results of the analyses performed in the present work show that the use of design methods that do not take the effect of shear stresses into account may result in under-designed pillars, while a false impression of rock mass strength could be derived from back analysis.
The support stress distribution of the printing sintering support pillars (PSSP), which became rounded due to the shrinkage that occurred in sintering formation, was different from that of the ...conventional cylindrical metal support pillars (CMSP). In this paper, ANSYS software was used to establish the mechanical simulation model of tempered vacuum glazing with PSSP. The influence of PSSP arrangement (spacing D, and three arrangement styles of square, regular triangle, and regular hexagon) on the mechanical properties of tempered vacuum glazing were analyzed, including the maximum stress σ1max on tempered glass (T-glass) surface, the maximum deformation ω1max of T-glass, and the maximum stress σ2max on the PSSP. The simulation results showed that when the PSSP were arranged in a square with D of 50 mm, σ1max, ω1max, and σ2max were 26.39 MPa, 7.82 μm, and 546.72 MPa, respectively, which met the requirements of the evaluation indexes of mechanical properties of tempered vacuum glazing, and had the least number of PSSP. The mechanical properties of tempered vacuum glazing with regular triangle arrangement of PSSP were better than those of square and regular hexagon arrangements, and the σ1max, ω1max and σ2max of which were 16.24 MPa, 7.25 μm and 489.66 MPa respectively, but with a too large number of PSSP. Optimization of the triangle layout angle and side length could reduce PSSP numbers. When the PSSP were arranged in an isosceles triangle with a 65° base angle and 50 mm bottom length, the mechanical properties of tempered vacuum glazing were the best. The triangular arrangement of PSSP had an optimization effect on the surface stress redistribution of tempered vacuum glazing, and the number of stress peaks was twice that of the square arrangement. Meanwhile, tempered vacuum glazing samples were made and the surface stress, thermal conductivity tests were conducted. The results showed that the measured values of peak and valley stress had an error of ±2.8 MPa compared to the simulated values. The simulation results were reliable and the surface stress redistribution of tempered vacuum glazing was present. The measured heat transfer coefficient of tempered vacuum glazing with PSSP was 0.7402 W m−2 K−1, slightly higher than that of tempered vacuum glazing with CMSP. The difference between them was not significant.
•Mechanical properties of tempered vacuum glazing were affected by pillar arrangement spacing and style.•Tempered vacuum glazing with regular triangle pillar arrangement had better mechanical properties than other arrangements.•The PSSP number could be reduced by optimizing the triangle layout angle and side length.•Triangular arrangement of PSSP had an optimization effect on the surface stress redistribution of tempered vacuum glazing.•The heat transfer coefficient of tempered vacuum glazing with PSSP was slightly higher than that with CMSP.
