New Prediction Models for Mean Particle Size in Rock Blast Fragmentation

The paper refers the reader to a blast data base developed in a previous study. The data base consists of blast design parameters, explosive parameters, modulus of elasticity and in situ block size. A hierarchical cluster analysis was used to separate the blast data into two different groups of similarity based on the intact rock stiffness. The group memberships were confirmed by the discriminant analysis. A part of this blast data was used to train a single-hidden layer back propagation neural network model to predict mean particle size resulting from blast fragmentation for each of the obtained similarity groups. The mean particle size was considered to be a function of seven independent parameters. An extensive analysis was performed to estimate the optimum value for the number of units for the hidden layer for each of the obtained similarity groups. The blast data that were not used for training were used to validate the trained neural network models. For the same two similarity groups, multivariate regression models were also developed to predict mean particle size. Capability of the developed neural network models as well as multivariate regression models was determined by comparing predictions with measured mean particle size values and predictions based on one of the most applied fragmentation prediction models appearing in the blasting literature. Prediction capability of the trained neural network models as well as multivariate regression models was found to be strong and better than the existing most applied fragmentation prediction model. Diversity of the blasts data used is one of the most important aspects of the developed models.     

Modelling the Impact of Rockmass and Blast Design Variation on Blast Fragmentation

Most blast fragmentation models assume the rock mass properties. explosive properties and blast design variables to be constants and uniformly distributed within a blast. However, in reality all these input variables vary within a blast resulting in variation in the resulting fragmentation size distribution. A stochastic modelling approach is introduced in this paper to quantify this variation. This technique takes the input variables as statistical distributions rather than constants and through several thousand iterations, generates a statistical representation of the expected fragmentation resulting from a poduction blast. A case study of three production blasts from a large open pit mine are presented and the modelled fragmentation 'envelope' shows good agreement with the fragmentation 'envelope' estimated from Split image analysis. The various blast-related parameters influence different parts of the fragmentation distribution, e.g., rock strength and explosive velocity of detonation have most impact on the fines. The technique is used to identify the parameters that have the greatest influence on various size fractions. Such an analysis will be useful to direct resources to efficiently minimise the variation.     

The BLAST-CODE model - A Computer-Aided Bench Blast Design and Simulation System

Blast design is a critical factor dominating fragmentation and cost of actual bench blasts. However, due to the varying nature of rock properties and geology as well as free surface conditions, reliable theoretic formulae are still unavailable at present and in most cases blast design is carried out by personal experience. As an effort to find a more scientific and reliable tool for blast design, a computer-aided bench blast design and simulation system, the BLAST-CODE model, is developed for Shuichang surface mine, Mining Industry Company of the Capital Iron and Steel Corporation Beijing. The BLAST-CODE model consists of a database representing geological and topographical conditions of the mine and the modules Frag + and Disp + for blast design and prediction of resultant fragmentation and displacement of rock mass. The two modules are established in accordance with cratering theory qualitatively and modified quantitatively by regression of the data collected from 85 bench blasting practices conducted in 3 mines of the Shuichang surface mine. Blasting parameters are selected based upon quantitative and comprehensive evaluation on the effect of the factors such as rock properties, geology, free surface conditions and detonation characteristics of the explosive products in use. In order to ensure practicality and reliability of the system, the BLAST-CODE model allows automatic adjustment to the selected parameters such as burden B and spacing S as well as explosive charge amount Q of any blasthole under irregular topographic and/or varying blastability conditions of the rock mass to be blasted. Simulation of the BLAST-CODE model includes prediction of fragmentation and displacement that are demonstrated in terms of swell factor, characteristic rock size x c and size distribution coefficient n by Rossin-Ramler's equation, and 3-dimentional muck pile profile. The BLAST-CODE model also permits interactive parameter selection based on comparison of the predicted fragmentation and displacement as well as the cost for drilling, explosives, and accessories until the most effective option can be selected.     

Predicting rock size distribution in mine blasting using various novel soft computing models based on meta-heuristics and machine learning algorithms

Blasting is well-known as an effective method for fragmenting or moving rock in open-pit mines. To evaluate the quality of blasting, the size of rock distribution is used as a critical criterion in blasting operations. A high percentage of oversized rocks generated by blasting operations can lead to economic and environmental damage. Therefore, this study proposed four novel intelligent models to predict the size of rock distribution in mine blasting in order to optimize blasting parameters, as well as the efficiency of blasting operation in open mines. Accordingly, a nature-inspired algorithm (i.e., firefly algorithm – FFA) and different machine learning algorithms (i.e., gradient boosting machine (GBM), support vector machine (SVM), Gaussian process (GP), and artificial neural network (ANN)) were combined for this aim, abbreviated as FFA-GBM, FFA-SVM, FFA-GP, and FFA-ANN, respectively. Subsequently, predicted results from the abovementioned models were compared with each other using three statistical indicators (e.g., mean absolute error, root-mean-squared error, and correlation coefficient) and color intensity method. For developing and simulating the size of rock in blasting operations, 136 blasting events with their images were collected and analyzed by the Split-Desktop software. In which, 111 events were randomly selected for the development and optimization of the models. Subsequently, the remaining 25 blasting events were applied to confirm the accuracy of the proposed models. Herein, blast design parameters were regarded as input variables to predict the size of rock in blasting operations. Finally, the obtained results revealed that the FFA is a robust optimization algorithm for estimating rock fragmentation in bench blasting. Among the models developed in this study, FFA-GBM provided the highest accuracy in predicting the size of fragmented rocks. The other techniques (i.e., FFA-SVM, FFA-GP, and FFA-ANN) yielded lower computational stability and efficiency. Hence, the FFA-GBM model can be used as a powerful and precise soft computing tool that can be applied to practical engineering cases aiming to improve the quality of blasting and rock fragmentation.     

Fragmentation prediction using improved engineering formulae

In the last decade, fragmentation prediction has been attempted by many researchers in the field of blasting. Kuznetsov developed an equation for the estimation of average fragment size, x 50 , based on explosive energy and powder factors. Cunningham introduced a uniformity index n as a function of drilling accuracy, blast geometry and a rock factor A associated with a “blastability index”, which can be calculated from the jointing, density and hardness of the blasted rock mass. Knowing the mean size and the uniformity index, a Rosin-Rammler distribution equation can then be derived for calculating the fragment size distribution in a blasted muckpile. Analysis of existing data has revealed serious discrepancies between actual and calculated uniformity indices. The current integrated approach combines the Kuznetsov or similar equation and a comminution concept like the Bond Index equation to enable the estimation of both the 50% and 80% passing sizes ( k 50 and k 80 ). By substituting these two passing sizes into the Rosin-Rammler equation, the characteristic size x c and the uniformity index n can be obtained to allow the calculation of various fragment sizes in a given blast. The effectiveness of this new fragmentation prediction approach has been tested using sieved data from small-scale bench blasts, available in the literature. This paper will cover all tested results and a discussion on the discrepancy between measurement and prediction due to possible energy loss during blasting.     

A Fragmentation Modelling Framework for Underground Ring Blasting Applications

Empirical approaches for predicting fragmentation from blasting continue to play a significant role in the mining industry in spite of a number of inherent limitations associated with such methods. These methods can be successfully applied provided the users understand or recognize their limitations. Arguably, the most successful empirical based fragmentation models have been those applicable to surface blasting (e.g., Kuz-Ram/Kuznetsov based models). With widespread adoption of fragmentation assessment technologies in underground operations, an opportunity has arisen to extend and further develop these type approaches to underground production blasting.

This paper discusses the development of a new fragmentation modelling framework for underground ring blasting applications. The approach is based on the back-analysis of geotechnical, blasting and fragmentation data gathered at the Ridgeway sub level caving (SLC) operation in conjunction with experiences from a number of surface blasting operations.

The basis of the model are, relating a peak particle velocity (PPV) breakage threshold to a breakage uniformity index; modelling of the coarse end of the size distribution with the Rosin-Rammler distribution; and modelling the generation of fines with a newly developed approach that allows the prediction of the volume of crushing around blastholes.

Preliminary validations of the proposed model have shown encouraging results. Further testing and validation of the proposed model framework continues and the approach is currently being incorporated into an underground blast design and analysis software to facilitate its application.     

A determination method of rainfall type based on rainfall-induced slope instability

Explosives are still the cheapest source of breaking rock in the mining or tunnelling operation and can be applied in varying geological conditions. It generates various troubles such as ground vibration, air overpressure, and fly rocks. It is well known that the maximum charge per delay (MCPD) has to be optimum for safe blasting and can be achieved through trial blasts, which is a complicated and costly process. Therefore, it is required to reduce the number of trial blasts. In this study, a total of 18 blasts were conducted in an underground coal mine and were simulated using similar ground conditions using Ansys software. The Peak particle velocity values obtained in the mines and through the models were compared. The error in PPV found between the actual and predicted by simulation is less than 15%. It can help us design the MCPD in rock excavation operations, visualise damages using simulation in Ansys software, and economical compared to field trials.

     

Development of Rock Engineering Systems-Based Models for Flyrock Risk Analysis and Prediction of Flyrock Distance in Surface Blasting

The environmental effects of blasting must be controlled in order to comply with regulatory limits. Because of safety concerns and risk of damage to infrastructures, equipment, and property, and also having a good fragmentation, flyrock control is crucial in blasting operations. If measures to decrease flyrock are taken, then the flyrock distance would be limited, and, in return, the risk of damage can be reduced or eliminated. This paper deals with modeling the level of risk associated with flyrock and, also, flyrock distance prediction based on the rock engineering systems (RES) methodology. In the proposed models, 13 effective parameters on flyrock due to blasting are considered as inputs, and the flyrock distance and associated level of risks as outputs. In selecting input data, the simplicity of measuring input data was taken into account as well. The data for 47 blasts, carried out at the Sungun copper mine, western Iran, were used to predict the level of risk and flyrock distance corresponding to each blast. The obtained results showed that, for the 47 blasts carried out at the Sungun copper mine, the level of estimated risks are mostly in accordance with the measured flyrock distances. Furthermore, a comparison was made between the results of the flyrock distance predictive RES-based model, the multivariate regression analysis model (MVRM), and, also, the dimensional analysis model. For the RES-based model, R ² and root mean square error (RMSE) are equal to 0.86 and 10.01, respectively, whereas for the MVRM and dimensional analysis, R ² and RMSE are equal to (0.84 and 12.20) and (0.76 and 13.75), respectively. These achievements confirm the better performance of the RES-based model over the other proposed models.     

Prediction of engineering properties of rocks from microscopic data

The purpose of this study is to develop the empirical equations for the prediction of the physical and mechanical properties of limestone and marble from microscopic data including their mineralogical and petrographical properties and to test the validity of model equations by using multivariate statistical methods. This study was performed on 15 different rocks, composed of six limestone and nine marble samples. Stepwise multiple regression analysis was applied to predict the engineering properties of both the marble and limestone rock samples considering petrographical properties as inputs. In order to determine the overall significance of the empirical equations for prediction of the physical and mechanical properties of marble and limestone samples, the F test was also performed. As a result of this study, it is found that the empirical equations developed in this study are statistically significant.     

Discrimination of Mine Seismic Events and Blasts Using the Fisher Classifier,Naive Bayesian Classifier and Logistic Regression

Seismic events and blasts generate seismic waveforms that have different characteristics. The challenge to confidently differentiate these two signatures is complex and requires the integration of physical and statistical techniques. In this paper, the different characteristics of blasts and seismic events were investigated by comparing probability density distributions of different parameters. Five typical parameters of blasts and events and the probability density functions of blast time, as well as probability density functions of origin time difference for neighbouring blasts were extracted as discriminant indicators. The Fisher classifier, naive Bayesian classifier and logistic regression were used to establish discriminators. Databases from three Australian and Canadian mines were established for training, calibrating and testing the discriminant models. The classification performances and discriminant precision of the three statistical techniques were discussed and compared. The proposed discriminators have explicit and simple functions which can be easily used by workers in mines or researchers. Back-test, applied results, cross-validated results and analysis of receiver operating characteristic curves in different mines have shown that the discriminator for one of the mines has a reasonably good discriminating performance.     

排土场散体岩石粒度分布与剪切强度的分形特征

应用分形理论研究了矿山排土场散体岩石粒度分布的分维规律,建立了分维数与排土场散体物料剪切强度参数的定量关系式。研究表明,排土场岩石块度分布具有良好的分形结构,分维数值大小随着排土场高度的增加而增加,但不超过3。当采样尺度范围一定时,分维数越大,散体中细颗粒含量越多,平均粒径也越小。分维数与散体岩石的剪切强度参数摩擦角?呈负指数关系。分维数值可用于排土场粒度资料的统计分析与剪切力学强度参数的预测。     

Characterization of underground rock fragmentation

Summary This paper focuses on the methodology and techniques developed to characterize the rock fragments produced by blasting in an underground environment. This work formed part of an integrated approach to the optimization of blasting design at a Canadian mine. Details are given of the photographic and image analysis techniques adopted, together with data from a program of full scale, study blasts in the mine. Features of the observed fragmentation are reviewed which related to controlled variation in the blast designs, together with other factors which were observed both to influence fragmentation characteristics and to interact with loading equipment productivity.     

MK模型的拟合方程和预测方程在土-水特征曲线中的应用及参数敏感性分析

土-水特征曲线可以预测非饱和土的各种性质（如：非饱和渗透系数、剪应力和热学性能等）。但测量土-水特征曲线耗时久且花费昂贵。为了解决这一问题,目前,很多研究都致力于从基本的岩土工程性质预测土-水特征曲线。基于此,以MK（Modified Kovács）模型的2种形式（拟合方程和预测方程）为土-水特征曲线模型,以Matlab编程语言中的cftool为拟合工具,以西宁黄土、粉砂土、红黏土和冰碛土4种细粒土为研究对象,对比拟合方程和预测方程描述细粒土土-水特征曲线的效果和差异,分析MK模型中黏附饱和度 1解 的变化规律,提出了基于MK模型的饱和度进行参数敏感性分析的计算公式。结果表明：拟合曲线和预测曲线在描述4类典型细粒土土-水特征曲线时均具有较好的效果,但拟合曲线整体上优于预测曲线;土壤质地和黏粒含量影响 值;饱和度对拟合参数 的敏感性较大,对拟合参数 的敏感性较小。     

岩体水力学基础(三)──岩体渗流场与应力场耦合的集中参数模型及连续介质模型

本文在分析了岩体系统的结构性和水力学特征后，提出了岩体渗流场与应力场耦合数学模型的机理分析、混合分析及系统辨识建模方法。运用系统辨识方法建立了岩体渗流场与应力场耦合的集中参数模型，并应用于解决实际问题；运用机理分析和混合分析方法建立了岩体渗流场与应力场双场耦合及与温度场三场耦合的连续介质分布参数模型。     

Multivariate statistical analysis approach for prediction of blast-induced ground vibration

Excavation of coal, overburden, and mineral deposits by blasting is dominant over the globe to date, although there are certain undesirable effects of blasting which need to be controlled. Blast-induced vibration is one of the major concerns for blast designers as it may lead to structural damage. The empirical method for prediction of blast-induced vibration has been adopted by many researchers in the form of predictor equations. Predictor equations are site specific and indirectly related to physicomechanical and geological properties of rock mass as blast-induced ground vibration is a function of various controllable and uncontrollable parameters. Rock parameters for blasting face and propagation media for blast vibration waves are uncontrollable parameters, whereas blast design parameters like hole diameter, hole depth, column length of explosive charge, total number of blast holes, burden, spacing, explosive charge per delay, total explosive charge in a blasting round, and initiation system are controllable parameters. Optimization of blast design parameters is based on site condition and availability of equipment. Most of the smaller mines have predesigned blasting parameters except explosive charge per delay, total explosive charge, and distance of blast face from surface structures. However, larger opencast mines have variations in blast design parameters for different benches based on strata condition: Multivariate predictor equation is necessary in such case. This paper deals with a case study to establish multivariate predictor equation for Moher and Moher Amlohri Extension opencast mine of India. The multivariate statistical regression approach to establish linear and logarithmic scale relation between variables to predict peak particle velocity (PPV) has been used for this purpose. Blast design has been proposed based on established multivariate regression equation to optimize blast design parameters keeping PPV within legislative limits.     

Prediction of Flyrock Distance in Open Pit Blasting Using Surface Response Analysis

Flyrock is a rock thrown to greater distance than desired and is a dangerous and unwanted phenomenon in surface mines, particularly, when blasting is proceeding close to human occupation and dwellings. The prediction of flyrock distance is critical in defining the statutory danger zone of blasting and has evaded blasters for quite some time. Control of flyrock with its distance prediction involves identification of key variables and understanding their influence. Theoretical models though provide a good understanding of the phenomenon, the confidence that can be assigned to such models is still very less. This study presents novel method to identify, merge and consolidate independent variables into a simplified equation for flyrock distance prediction without compromising on the actual field applications. Field investigations were carried out in several mines and relevant data were generated relating to flyrock. The key parameters, namely, explosive, blast design and rock mass nature were characterized and analysed. An empirical model involving the key contributors for flyrock generation and distance prediction were assimilated and a new equation was developed based on actual data collected by employing surface response analysis. The developed model was found to be statistically significant and validated. Sensitivity analysis was conducted to ascertain the role of independent factors on flyrock distance.     

A Method to Estimate In Situ Block Size Distribution

This paper presents a new technique for estimating the in situ block size distribution in a jointed rock mass. The technique is based on Monte Carlo simulations using more realistic fracture geometry as its input compared to other block size estimation methods described in the literature. This geometry represents fractures as either polygons or triangulated surfaces and therefore models persistence and truncation of fractures accurately. Persistence has been shown to be critically important for the accurate prediction of block size and shape. We show that for rock masses with relatively small discontinuities, the results of our predictions differ markedly from previous methods which over-predict fragmentation.     

Improvements in Blast Fragmentation Models Using Digital Image Processing

One of the fundamental requirements for being able to optimise blasting is the ability to predict fragmentation. An accurate blast fragmentation model allows a mine to adjust the fragmentation size for different downstream processes (mill processing versus leach, for instance), and to make real time adjustments in blasting parameters to account for changes in rock mass characteristics (hardness, fracture density, fracture orientation, etc). A number of blast fragmentation models have been developed in the past 40 years such as the Kuz-Ram model [1]. Fragmentation models have a limited usefulness at the present time because: 1. The input parameters are not the most useful for the engineer to determine and data for these parameters are not available throughout the rock mass. 2. Even if the input parameters are known, the models still do not consistently predict the correct fragmentation. This is because the models capture some but not all of the important rock and blast phenomena. 3. The models do not allow for 'tuning' at a specific mine site. This paper describes studies that are being conducted to improve blast fragmentation models. The Split image processing software is used for these studies [2, 3].     

Prediction of ground vibration due to quarry blasting based on gene expression programming: a new model for peak particle velocity prediction

Blasting is a widely used technique for rock fragmentation in opencast mines and tunneling projects. Ground vibration is one of the most environmental effects produced by blasting operation. Therefore, the proper prediction of blast-induced ground vibrations is essential to identify safety area of blasting. This paper presents a predictive model based on gene expression programming (GEP) for estimating ground vibration produced by blasting operations conducted in a granite quarry, Malaysia. To achieve this aim, a total number of 102 blasting operations were investigated and relevant blasting parameters were measured. Furthermore, the most influential parameters on ground vibration, i.e., burden-to-spacing ratio, hole depth, stemming, powder factor, maximum charge per delay, and the distance from the blast face were considered and utilized to construct the GEP model. In order to show the capability of GEP model in estimating ground vibration, nonlinear multiple regression (NLMR) technique was also performed using the same datasets. The results demonstrated that the proposed model is able to predict blast-induced ground vibration more accurately than other developed technique. Coefficient of determination values of 0.914 and 0.874 for training and testing datasets of GEP model, respectively show superiority of this model in predicting ground vibration, while these values were obtained as 0.829 and 0.790 for NLMR model.