Predicting the distribution of the volume of blasted rock mass in the surface prior to blasting itself is an extremely important and difficult task in surface mining, including the features of the interaction of various properties of rock mass and explosives materials, local geology, the initial distribution of explosive mass, time intervals between explosions, etc. Because the speed of the processes during the explosion is quite high, it is extremely difficult to control the explosion process in the real time. At the moment, the most effective approach is to simulate physics processes and determine all the initial parameters until the time of blasting. It is hard to overestimate the value of predicting the movement of blown up volume and includes saving on the further stages of mining operations, as well as minimising the cost of preparatory operations before an explosion.
Based on previous experience from the strength of materials and mechanics, as well as utilising statistics and an engineering approach, a numerical model was developed for simulating the volume distribution of blasted rock mass over the surface and caused by a series of explosions located in different space areas occurred at different time moments.
Using the strength of materials allowed to take into account the strength characteristics of destructible material of rock mass. Mechanics equations made it possible to establish the relationship between the geometric and energy characteristics of the discrete system, thereby determining the initial particle velocities after the destruction of the whole mass during a single explosion, as well as the trajectory of these particles.
The implemented model combines the simplicity of direct methods to depriving of time-consuming iterative algorithms with a time step and contains the necessary equations that satisfy the conservation of energy requirements to provide the natural results. Thus, we were able to implement an extremely fast approach for obtaining the physical modelling results of the blasted volume distribution. The high simulation speed allows to use this approach not only to predict the results of the explosion, but also opens up the possibility of using it in more complex algorithms where thousands or millions of numerical experiments are needed to perform reverse engineering tasks.
One of the priority areas for the further development of this model can be using of machine learning methods to tune empirical parameters for an additional improvement the accuracy, as well as using of multi-criteria optimisation to implement an automatic solution for searching the optimal source parameter values based on given results.