1. Introduction
The bench blast has been widely applied in many fields, such as mining engineering and hydraulic tunnel excavation. The effects of the bench blast would directly affect subsequent operations, such as loading, hauling and crushing. Otherwise, the bench blast would also cause some undesirable impacts, such as fly-rock, vibration and back-break, if the blast design is not optimized [
1]. To improve and optimize the bench blast design, a comprehensive, valid and convenient for application evaluation method of blast performance is therefore expected by mining engineers.
As efforts to find more scientific and the reliable methods to evaluate bench blast performance, various research studies were initiated. These works can be divided into three types from different viewpoints: (1) statistical analysis method; (2) harmful effects evaluation method; (3) numerical modeling method.
The statistical analysis method mainly employed the Artificial Neural Network (ANN) [
2,
3,
4,
5,
6,
7]. The ANN is an information processing system that is used for modeling complex relationships between input and output data. An ANN consists of an input layer, hidden layers and an output layer. Once the network has been trained by the previous blasting parameters and results, the results can be predicted by the network according to the new data inputs. However, in hidden layers, all the input parameters are calculated as whole, so this method can only show approximate relationships between parameters and results.
The numerical simulation method has been widely applied to predict the results of the bench blast, such as the Finite Element Method (FEM) [
8,
9,
10], Discrete Element Method (DEM) [
11,
12,
13] and Discontinuous Deformation Analysis (DDA) [
14,
15,
16]. The feedback from simulated results can provide a direction for design. However, the numerical model needs many parameters, such as geological and construction conditions, and some parameters are unavailable or difficult to be determined. The simulated results always have a significant difference with the results measured on site. Therefore, the numerical modeling method is only considered as a reference for bench blast design.
The harmful effects evaluation method mainly focuses on predicting the harmful effects induced by bench blasts [
17,
18,
19,
20]. For example, the famous Sadovsky equation [
21] was employed to predict the peak particle velocity (PPV) for a certain bench blast design, and some empirical formulas were obtained to evaluate the distance of the fragments flying away from the blast source [
22]. In this method, the single factor of bench blast performances was analyzed, instead of the comprehensive evaluation of the bench blast. Although they are still widely used due to their simplicity and relatively good reliability of results, these equations have one-sidedness because they only consider the harmful effects of blasting to evaluate the blasting effect.
To overcome the limitations of the abovementioned methods, a comprehensive evaluation method of bench blast in an open pit mine was proposed in this work. Eleven parameters were determined to evaluate the performance of bench blasts. Then, an evaluation model based on the linear programming and the fuzzy mathematics was proposed. Furthermore, a program code was developed on the platform of VC++. An application case of a bench blast was finally analyzed.
3. Program Development
In this study, a blasting effect evaluation system (BEES) was developed based on the aforementioned method. The detailed algorithm was designed, and the corresponding code was developed on the platform of VC++ (The version is 6.0, Microsoft). The program structure is shown in
Figure 1.
Boulder yield is calculated from rock fragmentation size distribution. The rock fragmentation size distribution is an important parameter for blasting effect evaluation, but the manual measurement is tedious and time-consuming. Today, software is available for use with any camera or photo type, or cameras with inbuilt software that automatically gives rock fragmentation size distribution both in table view and graphically. So, in this system, the image analysis technique is used for rock fragmentation size statistics. The boulder yield can be calculated easily by setting the standard value. The algorithm of fragment size statistics is shown in
Figure 2. The muck pile height can also be obtained by the method of image processing—the algorithm is shown in
Figure 3a, and its explanation on how muck pile height is determined is shown in
Figure 3b. The vibration can be measured by vibration measuring instrument. The distance meter can measure the distance of flying rock. The forward distance, under excavation, depth of side and back ditch and the flatness of terrace can be obtained by field observation and image processing. The blasting cost could be calculated by inputting the bench blasting engineering information.
In this system, the default data range and weight values are given by the expert evaluation method. That is to organize the experts in the field of Open-pit Bench Blasting Engineering, use professional knowledge and experience, and through intuitive induction, give a strong professional judgment and evaluation (see
Table 3 and
Table 4). However, the data ranges and weight values may be different for different construction environments and blast designs, it is not suitable for evaluation with the fixed data ranges and weight values. So, the data ranges and weight values should be changed with engineering conditions as it will be easier for deciders to obtain the advantages and disadvantages of blast design, and to better optimize the next blast design. In the system, the normal user and administrator models are provided. The administrator can adjust the data ranges and weight values according to the actual situation. The adjusted data are saved to the text in binary form and can be recalled next time.
After all the information input is completed, all the parameters are calculated and classified according to subclasses. The BEES provides dialogs for users to observe the evaluation of each parameter (see
Figure 4). According to all the parameter evaluation results, the blasting effect evaluation can be obtained—the algorithm is shown in
Figure 5—and the blasting effect evaluation results are displayed by the form of a graph (They can be seen in the next two figures,
Figure 6 and
Figure 7).
4. Application
To verify the evaluation and optimization method proposed in this paper, a practical bench blast engineering in a copper deposit was selected as an application case. The photo shot on site is shown in
Figure 8. This copper mine is mined by open-pit deep-hole step blasting. A No. 2 rock explosive is selected, and the structure of continuous charge is adopted. The total charge amount in the blasting area is 4392 kg. Detonators are used to detonate explosives. The latest Series IV advanced vibration and overpressure monitors are used for the blasting vibration measurement.
In general, there is visual error between photo measurement and site measurement; therefore, the correction coefficient is necessary. The camera is used to measure the image of the explosive pile particles. The visible area of the blasting pile is selected to shoot in an open direction. The area of each shot is about 40–30 m. The camera should be located in the front of the shooting area, and the elevation angle of the line of sight is within 35 degrees. Insert a pole, which should be located in the middle of the field of view, with an effective visible length of 1–2 m. The image scale coefficient is determined by the size of the post on the image and the actual size of the post. Then, the error of the image measurement method is corrected according to the measured particle size of the explosive pile sample, and the correction coefficient between the image measurement method and the geometric method is obtained. In this study, the coefficient is 0.8 and the rock fragment distribution is shown in
Figure 9.
With the different construction environments, the data ranges and weight values should be changed according to the construction condition. The BEES provides the default data ranges, and also provides the permission to modify the data. In this study, the data ranges are listed in
Table 5, and the weight values were not modified (see
Table 4).
The comprehensive blasting effect evaluation has two steps. First, in this study, a representative blasting operation was selected and 11 representative parameter values were obtained (see
Table 6). According to the values of the 11 parameters, the BEES evaluates blast effect from blasting safety, quality and cost, respectively, and shows the evaluation results by histogram—as shown in
Figure 6, the red column represents the evaluation result of the current index.
Second, the BEES shows the comprehensive evaluation result—the result is shown by histogram—as shown in
Figure 7, the red column represents the comprehensive evaluation result.
The blasting effect evaluation of this bench blast is between fair and good. The blasting cost evaluation is similar to the blasting effect evaluation. The blasting quality evaluation is between good and very good, but the blasting safety evaluation is poor. So, this blast design is not optimized. The next blast design should focus on burden and row spacing, etc., to control the blasting fly-rock and vibration.
5. Conclusions
Blasting effect evaluation is very necessary for blast design. In this paper, a method of blasting effect evaluation was proposed and an evaluation system was developed. The traditional evaluation methods are tedious and time-consuming. However, the developed evaluation system integrates several techniques, such as image analysis and parameter classification. All the selected parameters can be obtained easily to achieve rapid evaluation. The evaluation results are divided as subclass evaluation results and comprehensive evaluation results. The system has been used in practical bench blast engineering. The results are easier for deciders to obtain the advantages and disadvantages of blast design and provide the effective information for the optimization of the next blast design. The comprehensive evaluation system of blasting effect developed is not only convenient, accurate and objective, but also beneficial to enrich and perfect the fine management system of digital blasting management systems. The use of this software greatly reduces the burden of engineers and technicians and realizes the rapid statistics and evaluation of blasting effect.