Computer Forum OPTIMUM BLAST DESIGN

The computer blast model SABREX (Scientific Approach to Breaking Rock with Explosives) is a powerful new tool that can give miners the optimum bang for the buck. To highlight the role of all relevant blasting parameters on blast results, Sabrex combines computer graphics with the analytic approach of describing the interactions of explosives with rocks. The micro-computer version of the model can rapidly simulate blasts at almost any mine site, (The program has been used at such open-pit operations as Fording Coal, Quintette Coal, Crows Nest Resources, Teck Corp. and the Iron Ore Co. of Canada and in underground mines operated by Inco Ltd., Rio Algom and Denison Mines.) However, no blasting model can claim to simulate all blasting situations accurately and yield unequivocal results every time.

The SABREX model consists of three essential elements:

* detonation properties of explosives;

* strength and elastic properties of rock; and * blast geometry.

The treatment of detonation properties is unique to this model because it can accommodate the less-than- ideal behavior of commercial explosives. Unlike molecular and other military explosives, commercial explosives are multi-component systems. The components are often dictated by the needs of convenient packaging and economy. As a result, these explosives can sustain “ideal” chemical reaction only in very large diameters. In practice, the same explosive composition is packaged or used in a range of diameters.

In addition to the diameter of the borehole, non-ideal detonation is strongly controlled by the extent to which the explosive fills the hole (called decoupling), the mode of initiating the blast and the confinement provided by the surrounding rock. The extent of this non-ideality is also reflected in the energy partitioning (between shock and gas). In the reduced scale of practical charges, the reaction rates of commercial explosives are influenced by the surroundings. The loss of energy through lateral expansion results in lower detonation velocities and concurrent reduction in pressure and energy.

The complete SABREX architecture, as shown on this page, consists of several interacting modules. The external supporting modules, such as information from high-speed films, photogrammetric studies of muck piles and vibration analysis that are used to confirm and refine SABREX predictions, are also shown. In the absence of experimental data to run the non-ideal detonation code (cpex), there is provision for utilizing the output from an ideal-reaction core as an initial step to prediction.

The rock properties used as inputs to SABREX are:

* density

* Poisson’s ratio

* Young’s modulus

* unconfined static compressive strength

* unconfined static dynamic strength

* dynamic tensile strength

* the shock attenuation co-efficient.

The latter two are obtained experimentally under shock-loading conditions. The inputs on blast geometry are usually the following:

* borehole diameter

* blasting pattern

* borehole inclination

* bench height

* depth of subgrade drilling

* collar distance

* initiation pattern and delay times.

Information on unit costs or explosives, accessories and drilling complete the required input file.

The SABREX predictions consist of the following:

* fragment size distribution

* muck pile profile

* grade level fragmentation

* collar block fragmentation

* flyrock control

* backcracking, including damage below subgrade

* cost-per-unit volume of rock broken.

The size distribution of blasted rock can be predicted with either the Crack or the KUZ-RAM module.

The Crack module is based on postulated crack patterns around each borehole as a function of the equilibrium explosion pressure in the expanded borehole during the shock phase, as well as the strength and attenuation properties of target rock. Tensile breakage is the failure mode in this model. The size distribution analysis is carried out in two dimensions on any specified plane perpendicular to borehole axis by the Monte Carlo sampling method. The space bounded by the cracks and sampled in orthogonal directions by this technique essentially represent the fragment size in that space.

The KUZ-RAM model provides an alternative method of estimating fragment size distribution. It is based on comminution principles and assumes the available explosive energy to have a functional relationship with fragment size distribution.

The Heave module can yield either a relative measure of rock movement with respect to the reference blast or the complete profile of the muck. The choice is dictated by the desired precision in the blast simulation exercise. The detailed profile is generated by calculating the velocity and time of initial movement of blocks within the burden and tracking their path until they come to rest. The initial velocity is calculated from a knowledge of the explosion gas expansion curve. Because of the enormous complexity of multi-block motion accompanied by collision, the simulation is based on the simplification that the motion of the first row limits the motion of subsequently fired rows, and on the assumption of a swell condition and angle of repose for broken rock.

The Rupture module yields the extent of cracking of rock, called the rupture envelope, behind the blast hole as well as below subgrade. This is obtained by calculating the limiting distance from the borehole for failure of rock by tension, compression or shear from the hole pressurized at the equilibrium pressure. The rupture envelope as well as the grade break- out angle are generated from the postulated crater dimensions in this region. As mentioned previously, a variety of blast-monitoring techniques can be used to verify SABREX predictions and, in the process, increase the reliability of its predictions.

A typical rupture envelope simulation for an open pit is shown on this page. In this case, BT and BC represent the burden at toe and collar respectively. The backbreak profile is represented by CB, BB and SB at three key regions of the bench. The case shown consisted of only one continuous charge; different explosives for the toe and collar regions, as well as deck charges, can also be accommodated by the model. These results have important applications in berm and slope designs.

A graphics display of size distribution helps in rapid selection of the optimum explosive-blast geometry configurations. Photogrammetric survey of the muck pile can be used to refine these predictions and make them more sensitive to changes in some key blasting parameters.

Verification of the model predictions against surveyed muck profiles, although limited in number, has confirmed the essential soundness of the approach. This is particularly significant in view of the complex nature of the problem and the simplifying assumptions made in treating motion of broken rock mass.

Cost figures in a SABREX cost analysis are weighed against specific performance figures — fragmentation, heave, for example — to arrive at the most desirable explosive-blast geometry combination.

Considerable additional research is required to characterize rock masses in terms of response to shock loading, especially those with pronounced inhomogeneities such as bedding planes, faults and joints. The same applies to describing a more comprehensive rock failure criterion under combined dynamic loads of compression, shear and tension.

The exact role of stemming and its quantification in the modelling process would require further investigation. Our knowledge of crack propagation and branching in inhomogeneous rock mass remains largely qualitative. There is also an urgent need to develop a more accurate method of assessing fragment size distribution in very large muck piles than the photogrammetric techniques now employed. Despite these gaps in our knowledge, we can already foresee interactive systems operating on the bench or underground, consisting of drill monitors, borehole logging equipment, and explosives mix and pump trucks. The integrated system, with the help of real-time blast simulations, can deliver customized explosive products not just for each region of the pit, but for each borehole. Bibhu Mohanty is senior research scientist at CIL Inc.’s Explosives TechnicalCentre in McMasterville, Que. Paul Tidman is senior scientist at the ICI Explosives Group Technical Centre, also at McMasterville. Gordon Jorgenson is manager, blasting physics, with CIL Inc. in North York, Ont.