Computational Simulation for Mineral Processing

The use of computational simulation to predict the performance of existing equipment or for the design of improved mineral processing equipment will be referred to as Computational Mineral Processing (CMP) in this paper. The emphasis will be on solving the governing equations for liquid solid flows in complex geometric domains to obtain, typically, the three-dimensional distributions of velocity and pressure for the liquid phase and the velocity and concentration for the solid phase. The underlying technology, computational fluid dynamics (CFD), is very well-developed in other industries (aircraft, car and chemical) and forms a vital part of product design and analysis. The subsequent design of improved mineral processing equipment may also involve finite element structural analysis (FESA). Computational fluid dynamics (CFD) and finite element structural analysis (FESA) are part of the generic discipline of Computational Engineering and Science (CES). Ciment and Scherlis (1993) define CES as `The systematic application of computing systems and computational solution techniques to mathematical models formulated to describe and simulate phenomena of scientific and engineering interest'. Increasing computer power (Kaufmann and Smarr, 1993), is a dominant factor determining the rapid growth of industrial utilisation of CES. This increase is happening at both the supercomputer and the workstation level. Indeed, today's typical workstation, a Hewlett-Packard 735, is as powerful as a CDC 7600, the leading supercomputer in 1970. Future computer architectures are expected to become increasingly parallel (Gartner Group, 1990) allowing a sustained performance of a teraflop (a million megaflops) to be achieved well before the Sydney Olympics in the year 2000. It is recognised that the modem development of most disciplines in the physical sciences rests on the three complementary strategies of experimentation, analysis and computational simulation. As the 21st century is approached the volume of computational simulation, both fundamental and applied, is growing dramatically and the role of experimentation is diminishing. In the area of equipment design, CES provides the following advantages over experimental testing and measurement: 1. lead time in design and development is significantly reduced; 2. CES can simulate conditions not reproducible in experimental tests; 3. CES provides detailed and comprehensive information; and 4. CES is more cost-effective and time-efficient. The ultimate benefit for industry is greater productivity and hence greater profitability. The broader issues in relation to CFD are discussed by Fletcher (1993b). In the five- to ten-year time frame, growth in computer power will make it practical to combine CFD and FESA with optimisation procedures to produce a design process that is almost fully computerised. It is expected that fully automatic