Sand is not a liquid, even though it pours like one
Pour water out of a bottle and the flow rate depends on how full the bottle is — more water means more hydrostatic pressure at the neck, pushing it out faster, which is why the stream visibly slows as the bottle empties. Pour sand out of a funnel and something strange happens: the flow rate stays constant almost until the very last grains, regardless of whether the hopper is nearly full or nearly empty. This single observation, familiar to anyone who has watched an hourglass, is the signature of a granular material behaving fundamentally differently from a fluid.
The Janssen effect: force chains carry the weight sideways
In a liquid, pressure at depth grows linearly with the height of fluid above it. In a granular pile, weight is instead transmitted along chains of grain-to-grain contacts — force chains — which redirect a substantial fraction of the load sideways onto the container walls, where it is supported by friction. This is the Janssen effect: beyond a certain depth, adding more grains on top barely increases the pressure at the bottom, because the walls are already carrying most of the extra weight. Near the orifice of a hopper, this means the local pressure driving grains out becomes essentially independent of how much material sits above — hence the constant discharge rate.
The Beverloo law
Because the flow near the orifice is governed by a local free-fall-like acceleration rather than the pressure head far above, the mass discharge rate follows a Torricelli-like square-root law in the orifice diameter, empirically fit by W. A. Beverloo and colleagues in 1961:
Q = C · ρ · sqrt(g) · (D - k·d)^2.5 Q = mass discharge rate ρ = bulk density of the granular material g = gravitational acceleration D = orifice diameter d = grain diameter k ≈ 1-1.5, C ≈ 0.55-0.65 (empirical constants)
The exponent 2.5 rather than the naive 2 you might expect from a simple area-times-velocity argument (area ~ D², velocity ~ √D from free-fall) comes from a more careful treatment where the effective velocity at the orifice itself also scales with the square root of the effective size. The subtracted term k·d is the empty-annulus correction: grains at the very rim of the orifice cannot flow as freely as grains in the middle, so the orifice behaves as if it were slightly smaller than its geometric diameter. Crucially, Q depends on D, not on the height of the fill — the mathematical signature of the Janssen effect showing up directly in the discharge rate.
Arch formation and jamming
Narrow the orifice enough and something the Beverloo law does not capture takes over: the flow can stop completely, blocked by a self-supporting arch of grains spanning the opening. Each grain in the arch wedges against its neighbours, and the whole structure transmits the weight above it sideways into the hopper walls exactly like the Janssen mechanism at larger scale — a stable arch is essentially a tiny, load-bearing force-chain bridge. Because arch formation depends on the random local arrangement of grains at the moment the orifice narrows past a critical size, jamming is inherently probabilistic: an orifice a few grain diameters wide might flow freely for a while and then jam spontaneously, or jam almost immediately, with no two runs behaving identically. This is why real silos and hoppers are engineered with vibrators, air jets or orifices well above the empirically observed jamming threshold — usually taken as roughly 5 grain diameters for reliable flow of typical granular materials.
Why this matters beyond sand
Granular flow physics governs the handling of grain, gravel, pharmaceutical powders, plastic pellets and pretty much any particulate industrial material, and hopper jamming is a genuine, costly engineering problem — a blocked silo can halt an entire production line. The same force-chain and arching physics also appears, scaled up, in traffic jams at a bottleneck, in crowds squeezing through a doorway, and in the clogging of colloidal suspensions through microfluidic constrictions — all systems of discrete, interacting units funnelled through a narrow opening exhibit the same qualitative jamming transition.
Frequently asked questions
Why does an hourglass drain at a constant rate?
Because the weight of sand above the orifice is carried sideways to the container walls by force chains, not straight down onto the grains at the neck. The pressure at the orifice becomes independent of the height of sand above it once the pile is deep enough, so the discharge rate stays constant until the very end, unlike a liquid whose outflow speed depends on the hydrostatic head and slows as the container empties.
Why does the Beverloo law subtract k times the grain diameter from the orifice size?
Near the edges of the orifice, the flowing grains cannot pack and move as freely as they can in the middle, so there is effectively a dead, non-flowing annulus of about half a grain diameter around the rim. Subtracting kd, with k of order 1-1.5, corrects the geometric orifice size D down to the effective flowing width D - kd, which is what actually appears in the Torricelli-like square-root scaling.
Can a hopper jam even if the orifice is much wider than a single grain?
Yes. Arches can span orifices many grain diameters wide, especially with irregular, angular or cohesive particles. Jamming becomes overwhelmingly likely once the orifice drops below roughly 5 grain diameters, but it is a probabilistic event, not a hard geometric cutoff, so even a somewhat wider orifice can occasionally jam, and a somewhat narrower one can occasionally flow for a while before an arch finally forms.
Try it live
Everything above runs in your browser — open Granular Flow and change the parameters while it is running. Nothing is installed, nothing is uploaded, the whole model lives in one tab.
▶ Open Granular Flow simulation