The objective of most blasting operations in the mining industry is to destroy the in-situ structure of the rock mass, so that mechanical equipment can effectively and efficiently excavate the rock fragments. In achieving this objective, the engineer should be equally concerned with minimising the damage to the adjacent rock which will be left to form the load bearing structure.
However, one of biggest unknowns is just how much stress is actually required to initiate and then continue the fracturing in the rock, through the blasting process. If this process was better understood, selection of a suitable explosive, as well as the correct placement of the explosive within the rock mass could be better engineered. A further complication in diamond mining is that the diamond is contained within the kimberlite matrix. This product needs to be protected from any possible adverse blasting effects though at the time of this thesis research, this is still to be fully quantified.
It is therefore the objective of this research to attempt to identify the stress level at which kimberlite will begin to fail under explosive dynamic loading or simply, the near-field response of kimberlite to explosives.
In order to determine the physical parameters affecting the breakage of kimberlite in the near field, a field program as well as the appropriate instrumentation were devised to obtain this unique data. Boreholes were used to gain access to the area surrounding the explosive, so that the electrically resistant strain gauges could be placed at the desired positions to monitor the blast.
The introduction of a borehole into the rock created an interface through which the seismic signal would have to pass. Not only would the open borehole cause a free surface, at which spalling of the borehole wall could occur, but a free surface would cause the seismic wave, which at this point would be compressive, to reflect some of its energy as a tensile wave. To overcome this, a grout was manufactured to have the same acoustic impedance as the kimberlite rock, so that the artificially created void would be "transparent" to the seismic stress pulse.
To record the signal from the blast, electrical strain gauges were bonded directly to the rock and the borehole was only to be used as a means of placing the gauges at the required positions in space. The void created by the borehole was completely grouted with an acoustically similar material to the kimberlite once the gauges had been successfully bonded and the insertion tool removed.
Twelve test blasts were carried out altogether. These were done in three campaigns, with four tests being carried out in each campaign.
To investigate the dynamic breakage mechanism of kimberlite further, it was decided to use the Split Hopkinson Pressure Bar (SHPB) in the first instance and then move to the Plate Impact facilities, once a better understanding of its failure mechanism had been gained. This was to subject kimberlite specimens to different rates and magnitudes of stress and strain loading considered necessary to understand the dynamic response of kimberlite. In both cases numerical modelling was used to verify the results obtained.
Using a Split Hopkinson Pressure Bar (SHPB), a defined load could be set by the speed of the striker bar. Then by using a combination of high speed photography, electrical strain gauge recordings as well as the computer models, the kimberlite true dynamic failure mode was analysed.
As part of this research, work was commissioned at the Cavendish Laboratory, Cambridge University, England, to determine the Hugoniot Elastic Limit for kimberlite. This is the very first Hugoniot work to be carried out on kimberlite. The works suggests that the Hugoniot Elastic Limit for the tuffistic kimberlite breccia is in the region of 1 GPa.
The study concludes with a proposal for the behaviour for kimberlite breakage in the near field in response to explosive loading. It is based on the results from this research and published theories in shock physics and detonation physics.
Finally, an approach considered more appropriate to modelling rock breakage based on the outcome of this research is proposed. This approach combines detonation and geomechanical codes to fully describe the total blasting mechanics.