https://www.cdc.gov/niosh/mining/UserFiles/works/pdfs/psius.pdf

Pillar strength in underground stone mines in the United States

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.

Comparative performance of six supervised learning methods for the development of models of hard rock pillar stability prediction

Predicting pillar stability is a vital task in hard rock mines as pillar instability can cause large-scale collapse hazards. However, it is challenging because the pillar stability is affected by many factors. With the accumulation of pillar stability cases, machine learning (ML) has shown great potential to predict pillar stability. This study aims to predict hard rock pillar stability using gradient boosting decision tree (GBDT), extreme gradient boosting (XGBoost), and light gradient boosting machine (LightGBM) algorithms. First, 236 cases with five indicators were collected from seven hard rock mines. Afterwards, the hyperparameters of each model were tuned using a five-fold cross validation (CV) approach. Based on the optimal hyperparameters configuration, prediction models were constructed using training set (70% of the data). Finally, the test set (30% of the data) was adopted to evaluate the performance of each model. The precision, recall, and F1 indexes were utilized to analyze prediction results of each level, and the accuracy and their macro average values were used to assess the overall prediction performance. Based on the sensitivity analysis of indicators, the relative importance of each indicator was obtained. In addition, the safety factor approach and other ML algorithms were adopted as comparisons. The results showed that GBDT, XGBoost, and LightGBM algorithms achieved a better comprehensive performance, and their prediction accuracies were 0.8310, 0.8310, and 0.8169, respectively. The average pillar stress and ratio of pillar width to pillar height had the most important influences on prediction results. The proposed methodology can provide a reliable reference for pillar design and stability risk management.

https://www.mdpi.com/2227-7390/8/5/765/pdf

Predicting Hard Rock Pillar Stability Using GBDT, XGBoost, and LightGBM Algorithms

Investigation on support pattern of a coal mine roadway within soft rocks — a case study

To understand the fracture mechanism in all kinds of rock engineering, it is important to investigate the fracture evolution behavior of pre-fissured rock In this research, we conducted uniaxial compression experiments to evaluate the influence of ligament angle on the strength, deformability, and fracture coalescence behavior of rectangular prismatic specimens (80 × 160 × 30 mm) of brittle sandstone containing two non-coplanar fissures The experimental results show that the peak strength of sandstone containing two non-coplanar fissures depends on the ligament angle, but the elastic modulus is not closely related to the ligament angle With the increase of ligament angle, the peak strength decreased at a ligament angle of 60°, before increasing up to our maximum ligament angle of 120° Crack initiation, propagation, and coalescence were all observed and characterized from the inner and outer tips of pre-existing non-coplanar fissures using photographic monitoring Based on the results, the sequence of crack evolution in sandstone containing two non-coplanar fissures was analyzed in detail In order to fully understand the crack evolution mechanism of brittle sandstone, numerical simulations using PFC2D were performed for specimens containing two non-coplanar fissures under uniaxial compression The results are in good agreement with the experimental results By analyzing the stress field, the crack evolution mechanism in brittle sandstone containing two non-coplanar fissures under uniaxial compression is revealed These experimental and numerical results are expected to improve the understanding of the unstable fracture mechanism of fissured rock engineering structures

An Experimental and Numerical Study on Cracking Behavior of Brittle Sandstone Containing Two Non-coplanar Fissures Under Uniaxial Compression

/pdf/pillar-and-roof-span-design-guidelines-for-underground-stone-35w1rkq6tn.pdf

Pillar and roof span design guidelines for underground stone mines

Practical Estimation of Rock Properties for Modeling Bedded Coal Mine Strata Using the Coal Mine Roof Rating

Observations of pillar conditions in limestone mines showed that the presence of weak bands in the limestone can result in pillar damage at stresses that are lower than one would otherwise expect. The objective of this National Institute for Occupational Safety and Health study was to investigate the mechanism of failure caused by the presence of weak bands by using a series of numerical models based on realistic physical properties. The effect of the weak bands on pillar strength was also investigated by developing a series of pillar models with different width-to-height ratios and determining their strength by simulating a servo-controlled loading condition. The material properties of the bands and the limestone were varied and their effects on pillar strength were determined. Model results demonstrated failure development similar to the pillars observed in the field. The weak bands were seen to develop tensile stresses in the limestone as they extrude under increasing stress. The limestone fails in tension at a much lower stress than its expected uniaxial compressive strength. The results showed the load bearing capacity of the pillars can be reduced significantly by the presence of multiple thin weak bands. The degree of strength reduction is largely dependent on the compressive strength, the frictional resistance and the thickness of the weak bands. Single weak bands do not affect pillars as severely as multiple bands. However, the model results show that a single thick weak band, which comprises more than about 2 percent of the pillar height, can cause a significant reduction in pillar strength. The weakening effect of the bands become less severe as the pillar width-to-height ratio is increased. Field observation of spalling and failure associated with weak bands in limestone pillars show that the failure mechanism in the models resembles the observed failure. The model results provide insight into the important factors affecting pillar strength in the presence of weak banded materials.

/pdf/effects-of-weak-bands-on-pillar-stability-in-stone-mines-enuq50m9p2.pdf

Effects Of Weak Bands On Pillar Stability In Stone Mines: Field Observations And Numerical Model Assessment

A survey of pillar conditions was carried out at 21 operating limestone mines that use the room-and-pillar method. The surveyed pillars were all located in rock that was classified as “Good” to “Very Good” using the RMR rock mass classification system. The average pillar width was 14.5m and the width to height ratio varied between 0.53 and 3.52. Pillar conditions were rated using a visual rating system that accounts for the effect of geological structures as well as stress related fracturing.

John L. Ellenberger

Papers

Pillar strength in underground stone mines in the United States

Pillar and roof span design guidelines for underground stone mines

Practical Estimation of Rock Properties for Modeling Bedded Coal Mine Strata Using the Coal Mine Roof Rating

Effects Of Weak Bands On Pillar Stability In Stone Mines: Field Observations And Numerical Model Assessment

Pillar Stability Issues Based On A Survey Of Pillar Performance In Underground Limestone Mines