Density and Segregation in Moving Streams

This week I've been investigating particle segregation (or not) due to density differences and the effects of prior segregation (or lack of segregation) in conveyor belt streams.  Streams on conveyors are often segregated due to a number of factors like sifting segregation under vibration, the Brazil nut effect or other issues like the one you can observe in this short video.

One of the main causes of segregation is the bumping of the particle stream by the rollers under the conveyor - but we'll look at that in a future post.

In this post I've prepared five different experiments (using Algodoo) to look at density and segregation (or not) for particles of the same size, and differences of density and location in the conveyor belt-lots. Also note that all experiments assume no air current effect, which was the subject of my last post. Just select each experiment to see the Algodoo simulation videos

• Experiment 1 - Mixed belt-lot, particles with equal density
• Experiment 2 - Segregated belt-lot, particles with equal density
• Experiment 3- Mixed belt-lot, particles have density contrast
• Experiment 4- Segregated belt-lot, denser particles on top
• Experiment 5- Segregated belt-lot, denser particles below

If you like to work things out yourself perhaps watch the videos first and consider the questions included in the captions. Otherwise, I have included a short summary of each experiment below.  Also note that all experiments assume no air current effect, which was the subject of my last post.

1. Mixed belt-lot with particles of equal density - I'm seeing no segregation here and did not really expect any.  The belt-lot grouping is reflecting in the catch-bin pile.

2. Segregated belt-lot with particles of equal density. The belt-lot segregation is transferred to the catch-bin pile with the blue particles (on top) getting greater (angular) velocity as they leave wheel at the end of the belt.  Specifically, the larger the radius of a circular part, the greater the velocity imparted to the particles.

3. Mixed belt-lot with strong density contrast.  The denser gold particles have greater mass and as such are not as easily shifted by a lower density red particle, which tend to rebound from a collision with gold.  The result is that the higher density particles segregate to the centre of the pile and the lower density red particles roll (or in some cases are pushed) to the edges and upper surface of the catch-bin pile.

4. Segregated belt lot with density contrast - high density on top. The result of this experiment is very much the same as observed in experiment 2.  Really, the belt-lot segregation is the main cause of the catch-bin segregation, but if you look closely you will see that density does have an effect, as gold particles form a less movable pile due to the higher particle mass, and the red particles are forced to roll to the left.  See the comparative images above.

5. In the last experiment the configuration is reversed with the higher density layer on the bottom of the belt-lot. Here the belt-lot segregation is again the main control on the catch-bin pile but again density contrast is significant with the gold particles blocking nearly all mixing.

The main thing I've learned from these experiments is that the causes of segregation are complex and you need to observe closely the movement of particles in space and also understand the makeup of the components. Clearly density can have a significant influence on segregation but perhaps not so much as the original degree of mixing.

For the next posts I've prepared some simulations of conveyor streams similar to those in this post but with the additional complication of different particle sizes.