Chute Design Guidelines

This page collects established rules of thumb for transfer chute design, to help you judge whether a chute you are modelling is reasonable before you spend calculation time on it. They are starting points for a design and a way of spotting one that is obviously wrong – not a substitute for the DEM calculation, which is what tells you how your material behaves in your chute.

The five objectives

The paper sets out what every chute has to achieve, and it is a useful checklist when you look at a model:

1. Guide material onto the receiving belt at belt speed, in the direction of travel.
2. Eliminate material spillage.
3. Enclose material dribbles.
4. Enclose the material from operating personnel.
5. Eliminate dust liberation.

Only the first is something a DEM calculation can show you directly – and it is the one most often got wrong. Watch the particle velocity where material meets the receiving belt: the closer it already is to belt speed and direction, the less the belt has to accelerate it, and the less wear and spillage follow.

Chute width and lump size

Chutes are sized for their largest lumps, not their tonnage:

• Minimum width of three times the maximum expected lump size. A 300 mm lump gives a chute around 900 mm to 1 m square.

The reasoning is that once a chute is wide enough for the biggest lumps it will comfortably pass the volumetric flow as well, so lump size is the governing dimension. This assumes oversize – rock slabs, tramp iron, timber – has already been removed upstream.

Chute angles

The angle of repose for dry run-of-mine gold ore is taken as about 38°, and against that:

• 50–55° is the optimum chute angle for that material. Anything less is avoided. The same holds for dry run-of-mine kimberlite.
• 45° is acceptable only where material is crushed, washed, sized and can be guaranteed dry – in diamond plants, the final recovery and sort house areas.
• 65° to 70°To handle cohesive, wet, or sticky materials (e.g., iron ore with high clay content below the water table), wall slope and valley angles must typically be a minimum of 65° to 70° to the horizontal to guarantee flow and prevent dead-material stagnation
• Over 70° is needed for material removed by belt cleaners.

That last figure is much steeper than seems reasonable at first, and the paper gives a memorable reason: belt-cleaned fines have effectively no angle of repose — they hang upside down. Any plate meant to carry scrapings away, such as the plate under a head pulley, has to be steep enough that wet fines cannot sit on it, because any build-up there leads quickly to blockage or spillage.

Note also that once large amounts of fines and water are present, the angle of repose becomes hard to predict at all. In a mud-rush the surcharge angle can be zero — an exceptional condition, and not one to design for.

A more conservative approach to wall steepness

The figures above come from experience with hard, dry run-of-mine gold ore and kimberlite. Treat 50–55° as the least a dry, free-flowing material will tolerate, not as a general default – for many materials it is too optimistic.

The deciding factor is usually not the coarse lumps but the fines, and whether they can get wet. Dry fines run freely; the same fines with a little moisture (rain on a stockpile, dust suppression sprays, wash water, the wet season) bind together, cling to the plates and to each other, and steadily build up. Once build-up starts, it narrows the flow path and roughens the surface, which encourages more build-up, until the chute blocks. A material's dry angle of repose says very little about how it behaves in that state.

This is why the iron ore industry commonly uses chute walls of around 70°: iron ore often carries a significant proportion of fines that clump together when moist. The same reasoning applies to any material with appreciable fines that may not stay dry – coal, bauxite, and ores containing clay among them.

Valley angles — the trap

For two plates at angle θ meeting at a right angle in plan, the valley runs at arctan(tan θ / √2). A few values worth knowing:

The remedies are to steepen the plates until the valley clears the minimum, or to design the valley out altogether – a vertical-sided chute has no valley to block.

This matters when you look at a model: a chute whose every plate measures acceptably can still block along its corners, and in the DEM calculation that shows up as material lodging in the corners rather than on the faces.

Drop height and change of direction

How much vertical height a transfer needs depends on how far the material has to be turned:

In practice the upper end runs a little higher than the published figure — up to about 5 × belt width is not unusual. A real 1800 mm belt 90° transfer checked against this page drops roughly 7.5 m, which is a little over 4 × belt width.

Small acute angles are called out as the most troublesome of all — worse than a square 90° turn. They force awkward geometry, leave shallow plate angles that material builds up on, and put the scraping area above open space where dribble is hard to catch.

Locating the chute: IP and WP

Transfer chute drawings locate the loading point with two marks, and the convention is worth knowing because it is how the geometry is dimensioned on a real drawing:

• IP — intersection point. Drop a vertical line from the centre of the feed belt and its head pulley. Where that vertical meets the receiving belt line is the IP.
• WP — work point. A point measured from the IP along the receiving belt, in the direction that belt travels. The WP marks where the feed chute lip sits above the receiving belt. Around 300 mm is typical, for both 90° and in-line transfers.

This is something the DEM calculation shows you plainly. Trace particles at the loading point and look at their velocity relative to the belt: if they are landing against belt travel, the lip is on the wrong side of the IP or too close to it.

How far forward of the pulley

The horizontal distance from the head pulley centre out to the chute is not a fixed dimension — it follows the discharge trajectory, so it depends on belt speed:

• A reasonable starting point is about two thirds of the belt width. On the 1800 mm belt drawing this page was checked against, that offset is 1250 mm.
• Faster belt — increase it. Material is thrown further forward.
• Slower belt — reduce it. At low speed material falls almost vertically off the front of the pulley, and a chute set too far forward simply misses the stream.

This is really a discharge trajectory question, and two points from the paper apply:

• Estimate the trajectory from the centre of area of the material profile, not from the belt line. Take the belt line and the discharged material will strike the chute hood higher than predicted.
• Belt speeds change over a plant's life as throughput requirements change, and the trajectory and impact zone change with them. A head chute that suited the original duty may not suit the current one.

The trajectory is also something you can simply watch: run the calculation and see where the stream actually goes. That is more reliable than any rule of thumb, because it accounts for your material, your load profile and your belt speed together.

Dead boxes and impact

A dead box lets material collect and then be struck by the incoming stream, so the wearing surface becomes the ore itself — a deflection plate whose hardness matches the feed exactly. It also kills momentum in a long or high chute.

The critical limitation is that a dead box is not self-cleaning. It relies on larger particles scouring it. With fine, wet or sticky material it simply blocks, which is why the paper reports them used for relatively dry gold ore but avoided for run-of-mine kimberlite with high clay and moisture content.

If you are modelling sticky material, see Simulating Wet/Sticky Material – a dead box is exactly the feature where cohesion will change your answer.

Feed chutes and loading skirts

The reference configuration here is the Langlaagte chute, long standard across the South African mining industry, originally developed to get fines to flow in run-of-mine ore. Its standard dimensions are all expressed relative to belt width.

• Skirt spacing of two thirds of the projected belt width, not two thirds of the actual belt width — the paper is explicit that the latter was common practice and is wrong. For a troughed belt these are materially different numbers, and using the flat width gives skirts that are too wide.
• The underside of the skirts should rise to a maximum of 50 mm above the belt in the direction of travel.
• Skirts should start a minimum distance in front of the tail pulley, clear of the transition zone where the belt has not yet reached its full trough.

On sealing: rubber seals are seals, not containment. If the skirts are set too high the seals flop over the belt edge, upset tracking and wear the top cover. Where de-dusting is not required, the paper's position is that unnecessary seals cost belt life and add drag — redesign the skirts rather than adding rubber.

Dust extraction

Where dust must be controlled, extraction is best applied after the transfer point onto the belt. Note the consequence for the drawing: the proportions of the chute and skirts are then largely set by the de-dusting requirement — the air volume and pick-up velocity — rather than by the material flow alone. On a dusty duty, the dust cover and skirt geometry may not be yours to choose freely.

Using these with the drawing tools

Several of the figures above map directly onto inputs in the parametric drawing tools: