Presets
Playback
Parameters
Camera
Layers
Quality
Model check
The science
What is modelled
A shaking inclined deck sorts a mixed feed of platy mica flakes and rounded gangue grains (quartz and feldspar). The deck vibrates along a line tilted by the throw angle above its surface, like a vibratory conveyor. Flakes lie flat and slide with Coulomb friction; round grains roll with a small rolling resistance. Because the two kinds respond differently to the same shaking, flakes are carried up the deck into the concentrate bin and round grains roll down into the tails bin.
Equations
- Deck frame (s up the deck, z normal): the deck moves as
u(t) = A sin(ωt) (cos α, sin α), so a particle feels gravity(−g sin θ, −g cos θ)plus the inertial accelerationAω2 sin(ωt) (cos α, sin α). - Contact requires the normal force per mass
N/m = g cos θ − Aω2 sin α sin(ωt) > 0; otherwise the particle flies ballistically and lands inelastically (no bounce). - Flake in contact: sliding friction
μ Nagainst motion, stuck while|as| ≤ μ N/m. Round grain: rolling without slipping,a = (5/7)(as − cr N/m sgn v)for a solid sphere, at rest while|as| ≤ cr N/m. - Typical dimensionless intensity
Γ = Aω2/g; the drift speed of each kind is measured numerically (average over four start phases) and drawn against frequency, inclination and amplitude.
Simplifications
- Each particle is a point mass moving in the vertical plane of the deck; there are no particle-particle collisions, no rotation of flakes while airborne, no air drag, no bounce, and no sideways motion. Real feeds are layered, collide and cling by electrostatic forces.
- The friction coefficients are free sliders chosen in the plausible range for dry mineral flakes on steel (mica about 0.3 to 0.6); they are not measurements of a particular ore. The 5/7 rolling factor is the solid-sphere value; real grains are irregular.
- Particle sizes and friction values vary by a few percent around the slider value to show a spread. Real separators also use screens, air or water; this is only the dry friction and rolling step.
- Bouncing under strong shaking is chaotic: the drift curves are averages over four start phases and 1 s and still look jagged where flakes switch between stick, slide and flight. That jaggedness is real model behaviour, not noise added for effect.
- The vibration and the hop height are drawn exaggerated eight-fold, otherwise a 1.5 mm shake would be invisible. Everything else, including the 1 m deck, is to scale.
Loading model...
Loading Blender model...