Research in the Matuttis GroupGranular materials and friction
|
|
| Granular materials in motion may look like fluids, but as fluids go, they are pretty strange ones: Unlike the flow of Newtonian Fluids, where the upper surface of the fluid tries to stay horizontal, and the flow is determined by the viscosity and flow velocity, the flow of granular materials depends on the boundaries around the materials. | |
If the vessel is a narrowing hopper, all the grains will be in motion, like a fluid, which is called mass low. This is actually a problem in silo engineering, because in a silo with 50 tons of materials, all 50 tons are moving, and stopping the outflow will stop the 50 tons all at once - with considerable strain on the mechanic structure of the silo. |
|
| If grains flow through a hole in a horizontal bottom, there is only a narrow region where the particles drop downwards, and at the end, slopes towards the left and the right will be left over. In the case of a hopper, these dead zones of the flow mean that the hopper will never empty completely. |
|
| Granular
Dynamics and friction An important feature of granular dynamics is friction: Without friction, granular dynamics is not realistic. In the following, a heap is built up on a frictional floor, with the same coefficient of friction between the particles and the particles with the floor up to frame 400 with friction in the simulation. At frame number 400, the friction is switched off. The heap disintegrates and flows out like a fluid towards the sidewalls, forming a horizontal surface. (It is not a Newtonian fluid like water, because it can compensate normal stresses, but this "granular fluid" cannot form heaps.) |
|
| |
|
| In other words,
when we have stresses forming inside the heap, the
friction locks the particles onto each other and the
stress distribution (and the maintenance of the heap)
depends crucially on the friction between the
particles. Here is another heap formation with friction, but instead of the particles, the forces between the particles are drawn as blue lines, with the line thickness proportional to the forces. On the floor, the stresses are averaged over a representative volume and the largest stress component is drawn as red arrow: |
|
| |
|
| This is an example with clogging in a
hopper, where the particles are large enough and the
static friction is strong enough so that the
hopper cloggs. |
| Granular
dynamics, and modeling non-spherical grains |
||
| One of the most crucial
properties of dry granular materials is that they form
heaps: Poured out from a vessel, they automatically form a layered structure with a peak somewhere close to the middle. |
||
| Necessary for this is shape: The dynamics of granular particles is governed by a competition of rolling and sliding. Spheres have not need to slide, they can roll with minimum resistance, and with pure rolling, heaps can and will disintegrate to single layers. | |
|
| In our research, we model
granular particles as polytopes (polygons in two
dimensions, polyhedra in three dimensions) with
friction, to obtain granular dynamics which is as
realistic as possible: |
||
| Here is a rather curious movie of granular
dynamics. Actually, the experiment Alex and me have seen
performed decades ago in the experimental lecture. In a
vessel with a partition, particles are set up with the same
density. The system is vibrated so that the energy input by
the vibration is slightly less than the damping between the
particles. A small difference in the initial filling between
the left and the right half now leads to a higher energy
damping, in this case for the left half, so that all
particles finally accumulate on the left side. This can be
taken as an example of spontaneous symmetry breaking. Rather
more physically relevant, it as an argument that a granular
gas is not a "proper" homogeneous gas, but small
inhomogeneities will always lead to an ``inelastic collapse"
in the region around the largest initial particle
density. |
Last
edited: July 12
2026
Hans-Georg Matuttis