This question already has an answer here:

I have this bowl of sodium polyacrylate (Insta-Snow) that has been drying out for a few weeks. At this point it consists of solid pieces ranging in size from 3 or 4 mm across to much smaller ones that form a fine powder. I can stir them around and mix them but if I hold the bowl at a slight angle and shake it for few seconds they sort themselves out with the smallest particles at the higher part of the bowl and the largest ones at the bottom. The behaviour reminds me of seeing sand grains sorted on a beach by the action of waves.

How does this self sorting happen?


If I put the particles in a cylindrical cup and shake them they sort themselves the opposite way with the largest particles on top but in the bowl it's the finest particles on top. This seems to be a case of the reverse Brazil nut effect. Clearly interactions between the particles and the container are part of the explanation.

Before ShakingAfter Shaking

I'm not sure what tags apply - feel free to edit.


marked as duplicate by John Rennie, Martin, user36790, Qmechanic Apr 22 '16 at 12:58

This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question.

  • 5
    $\begingroup$ This seems a bit strange because this question is about why you should find the larger objects at the top. $\endgroup$ – ACuriousMind Apr 18 '16 at 1:08
  • $\begingroup$ I tried putting them in a small cup and shaking them. Then I did get the larger ones on top and the powder on the bottom but it took a lot more shaking. $\endgroup$ – M. Enns Apr 18 '16 at 1:25
  • $\begingroup$ The sideways forces are responsible for the sorting. $\endgroup$ – Peter Diehr Apr 18 '16 at 1:25
  • $\begingroup$ You have produced an "average" density gradient with the average density greatest at the bottom. If you are a small grain near the top what are you going to do? Fall through the gaps left by the big grains. $\endgroup$ – Farcher Apr 18 '16 at 4:32

The observation of granular convection as discussed here depends on the force of gravity.

What you are seeing is the settling of the small granules according to the force that the sides of the bowl are imposing. Try hitting in one direction only.

When shaken, the particles move in vibration-induced convection flow; individual particles move up through the middle, across the surface, and down the sides. If a large particle is involved, it will be moved up to the top by convection flow. Once at the top, the large particle will stay there because the convection currents are too narrow to sweep it down along the wall.

  • $\begingroup$ Thanks for the lead on granular convection. It seems what I'm observing is the reverse Brazil nut effect which occurs in conical containers. $\endgroup$ – M. Enns Apr 18 '16 at 14:37
  • $\begingroup$ It all has to do with the way the forces balance, and the "fluid" flow established. $\endgroup$ – anna v Apr 18 '16 at 15:14

Not the answer you're looking for? Browse other questions tagged or ask your own question.