arXiv blog

The Mystery of Sand Flow Through an Hourglass

The sand flow through an hourglass is entirely different from a similar flow of liquid. But physicists say new measurements prove that conventional thinking about this problem is wrong.

kfc 05/19/2010

  • 7 Comments

Physicists have long known that the flow of sand through an hourglass is entirely different from the flow of liquid. In the case of a liquid, the rate of discharge depends on the pressure at the aperture which is determined by the height of the liquid above.

But things are different in granular media. In 1895, a German engineer called Janssen discovered that the pressure at the bottom of a container of granules does not depend on the height of the grains above. He reasoned that the grains form chains and bridges that transmit their weight to the side of the container where they are supported by friction. So the pressure increases asymptotically with depth up to a threshold.

In fact, if the height of the granules is twice that of the width of the container, the pressure at the bottom is effectively constant. It is this constant bottom pressure that Janssen and others have always assumed determines the mass flow out of the aperture in an hourglass.

Not so, say Maria Aguirre at the University of Buenos Aires in Argentina and a few friends who have carried out an experiment to test this idea.

Their idea is to take gravity out of the equation by placing the granules, or in this case small discs, on a conveyor belt which carries them through an aperture (see diagram above). This allows them to alter the exit velocity of the discs through the aperture by changing the speed of the belt.

The results throw up a surprise: they show that with the same bottom pressure it is possible to get different flow rates simply by changing the speed of the conveyor belt. That means it is the speed of the discs that determine the flow rate, not the bottom pressure.

That's a subtle but important result. Janssen's effect has been studied in detail for more than 50 years without anybody questioning the thinking behind it.

And it could have interesting practical applications. The new work suggests that it ought to be possible to increase the flow of particles through an aperture by increasing their velocity through it. That could be done with electric fields, for example, or with solid or liquid conveyor belts like the ones used by the Argentinian team.

That should be useful in the numerous industries where particles have to be mixed together in a repeatable way, such as the food, drug and construction industries.

But for those worried about soft boiled eggs, it shouldn't effect the flow of sand through an hourglass.

Ref: arxiv.org/abs/1005.2884: Pressure Independence Of Granular Flow Through An Aperture


TRSF: Read the Best New Science Fiction inspired by today’s emerging technologies.

Print

Close Comments

To comment, please sign in or register

Forgot my password

UncleAl

409 Comments

  • 635 Days Ago
  • 05/19/2010

Symmetry

Round disks as particle model are incomplete for both shape and dimension.  Let the circles become increasingly divergent ellipses.  Try bearings, M&Ms; prolate, oblate, three axis-divergent ellipsoids.  Flow varies with edge roughness versus particle dimension.  Particle size distribution matters (fines!).  Tribocharging between particles and with the walls matters.

Diffraction patterns vary with the perfection of the diffracting boundaries.  Rough-eged slits gives apodization.  Interface impedence is complicated.

Air is viscous at small scales - ask a dandelion seed.  Do hard-evacuated hour glasses have the same kinetics as otherwise identical air-filled or SF6-filled hour glasses?  The real world controlling variable might be voids between the particles affecting air cushioning and counterflow, hence sensitive to particle random packing density and very sensitive to clogging fines (dilatant fluid).  Don't forget humidity, for damp is sticky.

Reply

misc

1 Comment

  • 635 Days Ago
  • 05/19/2010

Umm...

Increasing the conveyor belt speed increases the "bottom pressure" directly. Therefore, it stands to reason that increasing the belt speed increases the flow rate. This does not contradict Janssen.

Reply

Torbjorn_Larsson_OM

175 Comments

  • 635 Days Ago
  • 05/19/2010

Umm..., indeed.

I'm constantly amazed by people's sheer willingness to comment on research without reading or even browsing. Don't you think a paper titled "Pressure Independence" would have controlled for pressure? Well, not only did they have to, they did. (Though, I'll admit, in a qualitative way only.)

"We emphasize that, in our experimental conditions,
the effective pressure exerted by the granular bed on the
downstream wall depends on the dynamical friction force
between the disks and the belt. So, at a constant veloc-
ity V , the bottom pressure continuously decreases as the
number of disks inside the system decreases during the
discharge. Thus, during the whole discharge, a constant
flow-rate Q is achieved while the pressure continuously
varies, which indicates that the granular flow rate is not
controlled by it."

This observation directly falsifies claims that "Increasing the conveyor belt speed increases the "bottom pressure" directly."

Reply

cobrasixtysix

14 Comments

  • 634 Days Ago
  • 05/20/2010

Interesting

Seems to me that as the belt moved faster and faster the pressure at the base would decrease more and more until the base material was at freefall, at which point I would think that more speed would have no further effect on flow increases, unless gravity or pressure were to be increased in some way.

Reply

Netizen

131 Comments

  • 634 Days Ago
  • 05/20/2010

Elongated grains of sand may get "stuck"

I haven't tested this theory but suspect that if the grains of sand are elongated in shape (but the same mass per grain) they will get stuck and not flow through the hourglass. I suspect therefore the shape of the grains is critical. Nobody questions this because the shape of grains of sand is fairly constant due to the mechanisms by which sand is formed.

Reply

herrera

1 Comment

  • 634 Days Ago
  • 05/20/2010

"affect", not "effect"

The word should be "affect" and not "effect" in the next to last paragraph.

Reply

thedailycritic

13 Comments

  • 633 Days Ago
  • 05/21/2010

hourglass

So what are the implications for 'the days of our lives'?

Reply

Bio

The Physics arXiv Blog produces daily coverage of the best new ideas from an online forum called the Physics arXiv on which scientists post early versions of their latest ideas. Contact me at KentuckyFC @ arxivblog.com

Follow The Physics arXiv Blog on Twitter

Subscribe to the arXiv blog RSS Feed

Advertisement
Advertisement

Facebook

Advertisement