Presets
Playback
Parameters
Camera
Layers
Quality
Model check
The science
What is modelled
An inclined vibrating screen sorts crushed stone by size. Stone enters at the top of a wire-mesh deck, the deck shakes along a line tilted at the throw angle to its surface, and stones hop down the slope. Each time one lands it gets another chance to fall through an aperture; whatever reaches the end of the deck leaves as oversize. The coloured balls are sample parcels, each standing for an equal mass of feed, drawn about 2.5 times larger than their real size so they can be seen.
Equations
- In the frame of the deck (s down the slope, y normal to it) a stone feels
as = g sinθ + Aω² sin(ωt) cosαanday = −g cosθ + Aω² sin(ωt) sinα, where A is the amplitude, ω = 2πf, θ the slope and α the throw angle. It stays in contact whileay < 0, with normal forceN = −ayper unit mass and Coulomb friction μN (μ = 0.4) when it slides. - A landing has restitution e = 0.3 and a tangential impulse
μ(1+e)|vy|that removes sliding speed. The throw numberΓ = Aω² sinα / (g cosθ)says whether the deck can lift the stones at all: below 1 they only creep. - Passage: every landing (and every aperture pitch slid over) is a hit. A stone of size d smaller than the aperture a passes with the Gaudin probability
P = ((a − d)/(a + w))²for a wire of diameter w = 0.2a, times a bed-depth factorR = 1/(1 + (hbed/4a)²)that stands for stones that never reach the deck through a deep layer. The bed depth ishbed = Q / (ρb W v)with the measured mean speed v. - The feed follows a Rosin-Rammler law,
F(d) = 1 − exp(−(d/22 mm)1.5), truncated between 2 and 70 mm.
Model checks
- A single stone dropped on a still deck must bounce with flight times set by the restitution law.
- Vertical shaking must lift a stone off the deck exactly when Γ reaches 1, at the phase sin(ωt) = 1/Γ.
- The Gaudin formula is compared with a Monte-Carlo of random approach points over one aperture cell.
- Parcels fed must equal parcels finished plus parcels still on the deck, and the sampled feed must reproduce the Rosin-Rammler median.
Simplifications, stated plainly
- Stones are point parcels that do not collide with each other. Real beds stratify (fines work down through the coarse), jam apertures (pegging and blinding for stones close to the aperture size) and slow each other down. Here that crowding is one empirical factor R and is not calibrated on plant data.
- The deck is rigid and vibrates in a straight line; real screens flex, and circular or elliptical motions behave differently. The stones are spheres, so flat or elongated stones, which pass differently, are not represented.
- Friction, restitution, the wire diameter and the bed density are typical round numbers, not measurements of a specific plant. Absolute capacities (t/h per square metre) should not be read off the model; use it to compare settings against each other.
- The vibration of the drawn machine is magnified six times so it is visible; the physics uses the true amplitude. The throw angle, slope and aperture can be set beyond what one machine allows.
What to try
- Choose Too gentle: no throw. Γ falls below 1 and stones creep along the deck, so very few landings happen.
- Choose Overloaded and watch the bed depth and crowding factor cut the recovery of undersize.
- Raise the amplitude with the slope fixed: more throw gives more chances to pass but a shorter residence time. The best setting is a compromise.
Loading model...
Loading Blender model...