Transfer Chute Blockage
Simulating a redesign with DEM
The transfer chute was blocking at between 1,500 and 1,800 tonnes per hour. The client asked us to review the design and redesign the chute so it could carry 2,500 t/h. We modelled the existing chute and the redesign with DEM simulation, using the same coal and belt speeds, to see where the blockage starts and whether the redesign clears it.
The four videos below compare the two designs from different angles. Everything shown is simulation, not site footage. No site, terminal or company is identified.
Overview: full chute, 18 seconds
The full transfer path: feed belt, head pulley, deflector, head chute, spoon and receiving belt. Current design on the left at 1,800 t/h; redesign on the right at 2,500 t/h. Particles are coloured by speed (blue is slow, red is fast). From about 7 seconds, coal piles up in the current design above the spoon. In the redesign, the stream passes through cleanly.
Spoon close-up
Close-up of the spoon. On the left, a slow-moving pile fills the head chute above the spoon opening and grows throughout the simulation. On the right, the wider spoon opening lets coal pass straight through.
Deflector impact zone
The impact zone at the head pulley. The redesigned deflector is longer and reaches lower, carrying the coal stream further down and lining it up with the spoon. Both deflector and spoon are ceramic-lined.
Belt loading and bed depth
Neither design loads off-centre (load centre within 3–4 cm of belt centre). The current design delivers only about 1,180 t/h to the receiving belt while 1,800 t/h is fed in. The redesign delivers the full 2,500 t/h.
What Changed in the Redesign
The spoon opening is wider: 937 to 1,344 mm (+43%), and taller: 1,165 to 1,487 mm.
The deflector reaches 900 mm lower than the original, starts 223 mm further back from the head pulley, and stands 2.59 m tall instead of 1.78 m.
Both the deflector and the spoon are ceramic-lined. Wall friction coefficient in the model: 0.2 for the ceramic surfaces, 0.4 for the steel head chute.
Belt speed is 4 m/s on both belts, unchanged between the two designs.
How the Simulation Was Set Up
The simulation uses the discrete element method (DEM). Each particle of coal is modelled individually, with forces, collisions and sliding against the chute walls and belt calculated at every time step.
Coal was sized to the site profile: 50–75 mm at 5%, 10–50 mm at 50%, under 10 mm at 45%. These are represented as 62, 40 and 25 mm particles in the same 5/50/45% split. Real fines are too numerous to model individually, so the simulation uses particles sized to the site profile rather than modelling every grain.
On site, the chute blocked at between 1,500 and 1,800 t/h. The model was calibrated so the current design blocks at about 1,800 t/h. In the simulation, blockage begins at about 7 seconds: coal held in the chute grows by about 170 kg/s, reaching about 3,180 kg by 20 seconds. About 1,180 t/h reaches the receiving belt.
The redesign, run at 2,500 t/h, holds a steady mass of about 1,360 kg in the chute. The redesign's maximum capacity is still being simulated at higher tonnages; the videos show it flowing freely at 2,500 t/h.
What Simulation Can and Can't Tell You
A DEM simulation shows the flow pattern: where material builds up, where it hits and wears the liner, how the load lands on the belt, and how a design change shifts those things.
It does not replace structural engineering, fabrication drawings or site sign-off. The simulation models the flow; structural capacity, fit and maintenance access still need to be checked separately.
The result depends on the inputs. If the material varies (for example wet and dry coal), the simulation needs to test more than one condition.
Questions About This Simulation
Full 3D Model, by Invitation
The complete interactive 3D model—every simulated particle, from any angle—is available to engineers on request.
Request accessRead More
How DEM simulation works, what it's good for, and where its limits are.
What flow analysis is, how simulation shows blockages and wear, and the data that makes it reliable.