The design and continuous improvement of drilling and blasting activities in underground operations continues to mainly rely on documented rules of thumb as well as costly site specific trials. A comprehensive review of the literature has shown that there are few to no methodologies specifically developed for underground production blasting. The few that have been developed have applied empirical relations that do not adequately consider ring blasting geometries; have failed to appropriately respond to changes in basic design input parameters; and have proposed complex input and calibration requirements.
This thesis discusses the development, validation and application of a new fragmentation modelling framework specifically for underground production blasting conditions. A ''single ring" model is proposed as the basis of the overall framework and is the primary development of this thesis. This component is then extended into a stochastic model that allows the simulation of the impact of external operational factors on fragmentation outcomes such as blasthole deviation, dislocation and overall detonation performance.
The single ring component models the extent of both near field crushed zones and mid to far field fractured zones to predict the distribution of rock fragments expected to report to drawpoints. The output is based on the combination of two Rosin-Rammler based Sanctions which require the determination of three modelling parameters: the fines cut off point (fc), the expected mean fragment size (x50) and the "coarse" uniformity index (nc). Three original methodologies have been developed to determine these key parameters. These are; a model to predict the extent of crushing for a given explosive charge and rock combination; an empirical approach to estimate the post blast mean fragment size based on a more objective descriptor of in situ fracturing; and the implementation of a simple 3-D peak particle velocity (PPV) attenuation model that allows the determination of breakage and fragmentation uniformity characteristics for a given design layout and charging condition.
The applicability of the model in blast design is demonstrated with the introduction of a methodology to infer critical burden. This methodology was observed to be adequate for applications involving single ring blasts under confined or semi-confined conditions in hard competent rock masses. Further application of the single ring model is demonstrated through the analysis and evaluation of design parameters for a narrow inclined undercut, as part of the feasibility study of the Northparkes E26, Lift 2 block cave mine in New South Wales (NSW), Australia.
The extension of the single ring model into a stochastic modelling approach is explained and demonstrated through the simulation of fragmentation statistics obtained during the extraction of the first sub level undercut at the Ridgeway Gold mine in NSW, Australia. The proposed simulation methodology shows that operational statistics associated with the condition of SLC rings prior to firing, as well as detonation performance, may be modelled with the use of generalized beta functions. In the geotechnical conditions tested, the use of average in situ block size statistics, inferred from fracture spacing statistics, has been shown to be an appropriate input for blast fragmentation modelling and simulation purposes.
Whilst recognising the limitations of empirical and stochastic models, analysis and testing conducted in this research has demonstrated that the proposed framework provides a plausible methodology for modelling blast fragmentation in underground production blasting environments. The approach is aimed at the design evaluation stages in an operating mine and for studies at the conceptual, prefeasibility and feasibility stages of a project where different drill and blast scenarios and associated costs are still being assessed.
Four internationally refereed journal papers and one international conference paper have been published from the thesis work (see list of publications and Appendix A).