The work presented in this paper introduces a novel method of comminution; this is an impact crusher/grinder that pulverises lumps of up to 50mm size in fractions of a second. A crude device was proposed by Francis Clute (1966), but this needed better design and scientific understanding. Following an application for a new patent (Next Century Technologies et. al., 1997), a prototype machine was manufactured in 1998, and placed in a pilot plant. Since then, new patents have been applied for (Youds et. al). To date, two projects have been completed on this system, sponsored by the European Coal and Steel Community and the Department of Trade and Industry (DTI).
The energy required to reduce 50mm lumps to powder < 100 microns is in the range of 25-125 kWht-1 (Prasher, 1987). The material size reduction in these processes is achieved by impact and attrition of solid particles with a harder surface, usually made from treated and hardened steel. However, depending on the hardness of the particles appreciable wear on the impacted surface is possible. This means that machine components might have to be replaced regularly, and the product might be contaminated with metallic particles.
However, the interaction between the machine and processed material is not established. Therefore, the current work attempts for first time to achieve the mathematical/numerical modelling of this type of comminution. The main objectives of this work have been to identify the flow structure, the forces acting on solid particles, the main mechanism of breakage and possible scale-up rules for the machine. The proprietary FLUENT CFD code has been used, to predict the structures ofsingle-phase and two-phase flows, and to provide input for a particle breakage model.
In private communications, the following advantages have been claimed over conventional milling methods:
• High throughput rates of up to 5.5kgs-1 in one device, which is about six times the maximum throughput of a jet mill.
• The energy requirement is approximately 14kWht-1, compared with 20kWht-1 for conventional milling methods
• High size reduction ratios of up to 300:1 are achieved on a single pass, attributable to a rotor of speed up to 6000rpm
• The majority of product is less than 200 microns in diameter.
• The device is compact and has a limited number of moving parts. If the problems of rotor wear are overcome, it will need low maintenanceHowever, the main disadvantages to date are the high level of noise, and the potential for rotor wear. The fundamentals of operation, and the numerical model are discussed in the next sections.
The apparatus consists principally of a high-speed motor attached to a 610 mm diameter impeller . The motor is a 12-litre diesel engine with six cylinders in line, developing 222 kW and 105 Nm at 2000rpm. The motor drives the rotor with a belt driven twin-pulley assembly with a velocity ratio of 3:1. Attached to the impeller is a conical section and thereafter a straight duct, leading to the inlet of the apparatus. The length of the inlet duct can be changed from one to three metres, depending on the mass flow rate of the solid particles. The impellers tried so far have between eight and 12 vanes. To date, experimental results indicate that the harder is the material, the greater the number of vanes is required.
The measured velocity of the air at impeller speeds of 6000rpm is about 250ms-1. The inlet guide vane at the first point of impact with the particles is inclined to the left, whereas the outlet of the vane is normal to the impeller axis. Feed material is conveyed onto a scoop prior to the inlet duct section, and thereafter is accelerated into the machine. Thus, the airflow becomes anisotropic, because a swirl component of velocity occurs. The high-speed impeller transmits a weak free vortex upstream of the impeller zone. Based on preliminary work, observations from a high-speed video camera show that particles are accelerated in the inlet and conical section. However a number of localised vortices occur, as indicated by the backflow of particles in the near wall area of the conical section.
As particles are introduced to this vortex they change their path. Particles in the range of microns follow the airflow due to their low inertia. Observation of footage from a high-speed video camera indicates the following with respect to the larger particles. (1) They are likely to strike the wall of the inlet duct at some stage, (2) They are influenced by changes in the airflow, in that they tend to follow a spiral path towards the impeller vanes. After, the first impact with the impeller larger particles break,rebound backwards to the flow, and towards the conical section wall surface, which they strike before following a path along the wall. Eventually, after a number of impacts with the walls, they hit again with the impeller, where their size is small enough to allow them to pass through and leave the domain.
When the throughput rate is high, the big lumps, after collision with the impeller walls tend to move against the flow. It is then very likely that they hit with other particles coming from the opposite direction. This causes further size reduction, and a change in particle paths.
Downstream of the apparatus a cyclone collects the particles, where the larger particles are separated and collected at the bottom of the cyclone. The lighter particles exit the cyclone and are trapped in fabric bag filters. The filters are regularly pulsed with compressed air, and the filter cake is collected in metal bins. This way, the final size distribution includes a coarse and a fine product. A number of experiments have shown that the assembly reduces lumps of minerals up to 50mm original diameter to a mean product particle diameter of less than 200 microns. This has applied to recycled glass, glass frit, limestone, grit-stone, talc, coal and clay.
Owing to the complexity of the physical process, computational methods are needed to solve the flow field within the domain, and investigate turbomachinery characteristics for the shape of the impeller vanes.

