|
Our group is presently studying the solid/fluid and stick-slip/steady sliding transitions in granular matter. Granualr materials are a fascinating juxtaposition of the 3 normal states of matter - solid, liquid and gas. A granular material can behave as a solid when it is at rest, a liquid when it flows down an incline, and a gas when violently shaken. In fact, it is possible that all three states can be simultaneously present in a single sample of the material.
Granular materials are generally considered to consist of many (i.e. thousands or millions) single particles which are at least mesoscopic in size - small microscopic particles can often suffer thermal excitation and so destroy the solid phase; thus the particles must be at least big enough so that thermal fluctuations (kT), or fluctuations due to background noise or vibration, are negligible compared to the energy scale typically required to move the particles. Thus a powder can be considered as a granular material, though one must also be careful to avoid (or treat) cohesive or repulsive forces between grains, such as electrostatic forces or choesive forces due to excessive ambient humidity; this latter can be relevant even for macroscopic grains such as sand. At the other end of the scale, granular materials can be composed of rocks and boulders of size up to 10 metres or more. It is widely suggested that plantary and stellar accretion disks are best treated as a granular material.
The size of granular particles means that the structures formed are not in thermal equilibrium with the surroundings. Furthermore, the nature of the contact forces between grains, a non-linear interaction with threshold, means that proagation of stress in a granular material is highly disordered. Few connected grains form 'force chains' which carry most of the load to the underlying support while most grains carry little or no load. This is in marked contrast to solids and fluids where the internal pressure is normally homogenous. By applying an external force to the granulate, these force chains can shift and change so as to adapt to the applied force. Eventually, the granular material is brought to a barely stable state in which the smallest increase in external force will trigger a macroscopic rearrangement of the force chains and the granules.
In our experiments, we apply a steadily increasing shear force to a granular material which is confined to a circular channel of diameter 40cm, width 6cm and height <8cm. The channel is filled to varying heights with 2mm glass beads, and an annular plate is forced to rotate over the top of the channel. The plate is driven by a motor via a torsion spring which slowly builds up the shear force between the overhead plate and the
granular material until slip occurs.
 Previous results have shown clear differences in the statistical signature of the liquid and solid phases, a power-law distribution of event sizes, possible precursory activity before the occurrence of large events, and an almost one-to-one correspondence between granular shear stick-slip motion and the motion of magnetic domain walls in an external magnetic field.
We are presently adapting our experiment to add transparent walls, a torque sensor for measuring friction coefficients of the granulate, a motion sensor to measure dilation of the material etc. A high speed camera and long-exposure photography will help visually correlate mechnical observations of the system, in particular the extent of the shear band.
At present the experiment is being adapted to include a shaker to investigate the properties of the granular material when shaken at various intensities. We wish to measure the mean friction coefficient as a function of shear velocity and the depth of the granular layer at various shaking intensities.
We have recently started investigating the behaviour of a ratchet inserted in a shaken granular material. The test material, about 700g, is located in a glass beaker (12x9cm ca.) placed on a mechanical shaker, and an immersed asymmetric ratchet (which should play the part of a Maxwell's demon) is observed to see if a macroscopic rotary motion (perpendicular to the direction of energy input) is present. The control parameters are the excitation frequency ands and intensity, and also the properties of the ratchet (asymmetry, size, insertion depth etc). The xperiments reported were conducted in the absence of any macroscopic collection granular motion, e.g. convection or circulation.
Results show anything but a simple behaviour though some inferences can be made. When the imposed acceleration is below g (the acceleration due to gravity), the probe does not exhibit any significant motion whereas above this threshold, the probe begins to rotate in a direction dictated by its asymmetry - thus we can be sure there is no collective circular motion. More mportantly, the mean velocity of the probe vanishes as the lower band-pass frequency limit is increased above 60Hz, whereas it varies only weakly with the upper frequency cut-off. Importantly, a mechanic resonance with the granular material was found between 30-40Hz. Our intention is to identify the origin of this resonance through further testing and simulation.
|
