Peak Particle Velocity for Blasting Rock Modeling

ABSTRACT Blasting mechanisms are a complex coupling of the rock mass properties and the explosive detonation performance as well as the blast design parameters. Both rock mass and explosive properties have parameters that are inherently not well defined. Thus, blast modeling is greatly challenging to predict rock fragmentation, blast vibration, and the broken-rock mass movement. For blast modeling, previous work has shown that near-field signature-hole blast vibration monitoring is an effective method to obtain critical blast model input. In a blast field, multiple charges create nonlinear additive strains at a given point in the rock and time, which further complicates blast modeling. This paper shows how the measured peak particle velocity ("PPV") from the signature-hole blast vibration serves as a key controlling parameter for modelling. Using PPV as the controlling parameter, pressures and strains can be approximated at a point of interest. Rock breakage is then related by simplified approximations to the pressures and strains. The modelling method described in the paper relates all blast design parameters to the PPV induced by multiple charges at a point in the rock. Therefore, using PPV as a key parameter allows blasting models to simulate all blast design parameters and significantly simplifies blast modeling. INTRODUCTION Predicting the results of rock blasting has been a long-standing challenge since its inception. However, the complex and dynamic mechanisms of rock blasting remain unclear (Fourney, 2015). The variables affecting the blast results are numerous, including rock mass properties, explosive properties, and blast design parameters. These variables interact in a nonlinear manner, making rock blast modeling more difficult than static or quasi-static rock mechanics modeling. If a blasting model is built purely based on the first principles of physics, it must involve many parameters related to rock properties and explosives. However, these parameters are often difficult to impossible to measure, which makes it challenging to obtain relevant blast modelling input. Moreover, most of these blast models cannot simulate full blasts and design parameters. Even a single modeling prediction requires a large amount of computing time and can only simulate a small number of blastholes. This forms a major impediment to the engineering applications of blast modelling.