Impact Beds at the Load Point: Bars, Rollers and Belt Support

01Why the Load Point Punishes the Belt and Not the Chute

A conveyor's load point is the one place on the whole route where material arrives with energy it did not get from the belt. Everything downstream of the skirt boards is a transport problem. The first two metres are a collision problem, and the surface that takes the collision is almost always the belt, because the belt is the only thing in the path that is allowed to flex.

How hard that collision lands depends far more on drop height than on tonnage. The energy carried by a single lump rises with its mass, but it rises with the square of the speed picked up in free fall, and that speed climbs with the square root of the drop. Half a metre of extra chute can put more energy into one stone than doubling the throughput through a shorter drop. Two plants with the same belt width and the same capacity can therefore have load point lives that differ by a factor of five, which is why a transfer house destroys belts in months while the main incline runs for years.

Send us your drop height, lump size and belt speed for a bed layout review

The strike is only half of it. Between troughing idlers the belt is unsupported, and spacing is normally somewhere between one and one and a half metres. A lump that lands between two rolls pushes the belt down into that gap, bends the carcass over a short radius at the shoulders of the neighbouring rolls and pinches the cover between the stone and the roll shell. What comes out of that is a dent which never fully recovers, a ply separation that begins under the cover, and a section of industrial conveyor belt that looks sound from the walkway until it delaminates in a strip half a metre long.

Two failure modes arrive in the same event. Cutting from above, fatigue from below.

The useful question at a transfer is not how to build something strong enough to survive the impact, because the chute and the structure already do that. The job is to move that energy into a replaceable part, so the question becomes which component you are willing to destroy and how quickly it can be changed. Impact energy is not absorbed by strength. It is absorbed by deformation and by wear, and both of those have to happen inside a part that is small, inexpensive and reachable from the walkway. A sacrificial bar that loses ten millimetres in a year is doing exactly what it was bought to do. A belt cover that loses ten millimetres in a year is a shutdown, a splice and a crane.

That is why the most common answer at a cutting transfer is also the most expensive one, measured against the protection it actually delivers. When a belt starts getting cut, the first move is to reorder the same construction with a heavier cover, and the second is to upgrade to a cut and impact resistant grade. Both help, and both are limited in the same way, because cover compound resists abrasion and cutting without removing energy from the carcass. A twelve millimetre cover passes the same blow through to the plies as an eight millimetre cover does, so what it buys is time before the cover is breached rather than protection for what sits underneath. It also adds mass to accelerate at every start and raises the strain at the drive pulley. Any conveyor belt manufacturer will quote it without argument, because the request is easy to fill.

There is a worse version of the same mistake. Heavy cover conceals damage. A twelve millimetre cover can swallow an eight millimetre cut and look untouched while the breaker fabric below it is already torn. The next lump opens the wound, and what fails is not the cover but the splice or the carcass, usually on a shift when nobody is looking for it. On any belt where the cover is being asked to do structural work, inspection stops being a walk-past and becomes a measurement with a datum, a gauge and a schedule.

None of this makes cover grade irrelevant. Fine material at a modest drop, a clean chute and a steady feed will run for years on a well-chosen abrasion resistant conveyor belt, and that is the cheapest correct answer available. The distinction is about what the cover is being asked to do. If the cover is the wear surface, grade it properly and stop worrying about it. If the cover is being asked to absorb a stone that fell three metres, it is the wrong component in the wrong place, and no compound on the market changes that.

One transfer we were called to had been through three belt specifications in eighteen months, each thicker than the last, and the cutting had not moved. The chute discharged 2.4 m above the belt surface onto a 1,200 mm wide belt running at 2.8 m/s, and the gap between the last two idlers sat directly under the stream. A straight edge across the rolls showed 38 mm of sag with the belt loaded. The belt was not being cut by the stone so much as folded into a gap and pinched there, and a fourth belt would have failed the same way. The fix was 2.6 m of continuous supported surface.

The rest of this page stays on the bed itself, meaning the bars, the cushion underneath them, the frame that carries them and the geometry that decides whether the belt actually lands on them. Rollers get one section here only so the two options can be compared honestly; the mechanics of impact rolls, their disc stacks and the way they deflect are covered in how impact rollers protect conveyor belts, and that page is the right place to start if you have already decided on rolls.

02Anatomy of an Impact Bed: The Four Interfaces Under the Belt

An impact bed is a short, heavily built section of conveyor that replaces the troughing rolls under a transfer with a continuous sliding surface. It is not one product but a stack of four, and each interface in that stack does a different job. From the top down the order never changes: belt, bar, cushion, frame, structure. Confusing any two of those layers is how beds get specified badly.

The bar is the only part that touches the belt. It is a replaceable polymer or rubber block, usually one to two metres long, 100 to 200 mm wide and 50 to 100 mm thick, laid across the belt in a staggered pattern so no continuous joint runs the length of the bed. Both ends are chamfered and the fixing heads sit in counterbores with plugs over them, because anything standing proud of the surface will pick up a belt edge and start a tear.

Under the bars sits the cushion, a layer of rubber pads typically 25 to 75 mm thick and in the middle of the 55 to 70 Shore A range. It is the spring in the system. A point load spreads across four or five bars instead of one, and the surface deflects a few millimetres so the peak force is stretched over a longer time. Cushion that is too hard turns the bed into a rigid anvil. Cushion that is too soft bottoms out on the frame and takes a permanent set.

The frame holds the bars at the right height and the right angle, and it is the part that has to survive the horizontal drag of material that arrives with forward velocity and then has to be slowed to belt speed. Side rails, cross members and adjustable bar seats are normal; shimming under the rails to set the profile is normal too. What is not normal is a frame light enough to flex every time a lump lands, because a flexing frame works its bolts loose and eventually cracks at the welds.

The fourth interface is the structure below, and it is the one most often left out of the conversation. A bed does not absorb energy. It hands it down into the stringers and the chute steel.

Buyers often describe this assembly as an impact bar conveyor, which is a fair name for the section and a misleading name for the purchase. What you are buying is a loaded surface with a defined wear allowance, not a conveyor. The bars will be replaced three or four times in the life of the frame, and the frame will be replaced once in the life of the structure. Design the access around that ratio and the maintenance cost stays where it belongs.

The components around the bed matter as much as the bed. Skirt rubber hanging too low drags on the belt and adds load the bed was never sized for, while skirt rubber hanging too high lets the discharge stream escape sideways and wear the belt edges instead of the bar centres. Conveyor components that seal the chute and the load point belt support have to be chosen together, because they share the same two metres of belt and the same maintenance window.

03Three Answers at One Load Point, Compared Honestly

This page has to be explicit about its own boundary. The site already covers impact rollers, including what rolls do at a transfer, how the disc stack deflects and how the roll gap changes the pressure on the belt, and none of that is repeated here. What follows deals with the continuous support surface, meaning bars on a cushion in a frame, and uses rollers only as the thing a bar bed is compared against.

Comparing two options is not enough, because most plants arrive at this decision having already tried a third. The realistic choice at a cutting transfer is between a heavier cover on the same belt, a set of impact rolls in the same idler frames, and a bar bed that replaces the frames altogether. Put them side by side and the trade is not about strength at all. It is about how much energy each option removes from the carcass, how much of the wear lands on a part you can carry up the stairs, and what each one does to the rest of the conveyor.

The last row is the layout that most well-run plants end up with, and it is worth saying why. A bar bed is expensive in drag, because a belt sliding on polymer takes more force than a belt rolling on idlers. That penalty is worth paying for the two or three metres where the lumps land and is not worth paying for the next two hundred metres. Conveyor rollers are the better support everywhere the material has already matched belt speed, and the transition between the two support types is where the next problem usually starts.

Where bars lose is predictable. Sticky material that builds up between bars turns the surface into a rasp and wears the cover from below, which is a failure mode rolls do not have. A drive with no margin notices the extra drag at every start, especially on a loaded incline. Long beds on high-speed belts also run hotter than expected in still air. None of these are reasons to avoid a bed at a genuine impact zone, and all of them are reasons not to extend one further than the job requires.

Roll selection itself is a separate decision, and the geometry of trough angle and spacing either side of the bed belongs with the idler layout rather than here. If the section downstream of the transfer is being reviewed at the same time, the practical starting point is troughing idlers, trough angle, spacing and roller grade selection, because the bed has to match the profile those rolls impose. A bed set to 35 degrees feeding into a 45 degree trough is a belt training problem waiting for its first shift.

04Sizing the Bed: Drop, Lump Mass and Belt Speed

Sizing starts with a number almost nobody has on file, which is the mass of the largest lump rather than the average. A transfer handling screened material at a steady 900 tonnes per hour can be gentler than one handling 300 tonnes per hour of run-of-mine rock, because the second line delivers individual pieces that weigh more than a person. Ask for the top size and the shape, since a flat slab lands differently from a rounded cobble of the same mass.

Speed matters twice. It sets how much energy the lump carries, and it sets how far the material travels forward before it settles, which is the number that decides how long the bed has to be. A discharge stream that lands 800 mm downstream on a slow belt can land 1.6 m downstream on a fast one, and a bed sized for the first case will be half empty in the second. On a recent cement plant transfer running at 3.5 m/s, and on a port terminal running the same belt width, we watched the landing point move almost a metre between an empty belt and a fully loaded one, because material was riding on itself.

Length is a measurement, not a formula. The bed has to continue until the material has stopped moving relative to the belt, because everything before that point is still delivering energy sideways into the surface. In practice that means running the supported section at least a metre past the end of the skirt boards and then watching where the belt stops deflecting. Chalk a line across the belt, run it loaded, and see where the chalk survives. The place it survives is where the bed can stop.

Width is simpler but no less important, and it is one of the questions a conveyor belt supplier should settle before the drawing is issued. Bars are made to suit the belt width and, on a troughed bed, are built as a centre bar with two angled wing bars so that the surface follows the trough rather than sitting flat across it. A flat bed under a 35 degree trough forces the belt to flatten as it crosses the transfer and then re-trough immediately afterwards, which is a training problem and a fatigue problem at the same time.

05Bar Materials: What Wears, What Cuts and What Creeps

Nearly every bar sold today is some grade of ultra-high molecular weight polyethylene, and the arguments about which grade are usually arguments about the wrong variable. The polymer's job is to slide under a loaded belt with as little friction as it can manage while giving up material slowly and predictably. Friction and wear resistance both come from the same place, which is a very long molecular chain, and neither of them is visible in the colour of the block.

Two failure signatures matter more than any datasheet figure. The first is creep, where a bar under continuous load gradually flattens and the surface that was set 2 mm below the belt ends up 6 mm below it, so the belt stops touching the bed exactly where the load is heaviest, and the rubber conveyor belt goes back to spanning a gap. The second is embedment, where fine grit works into the polymer surface and turns a sliding face into a grinding face. A bar that has gone grey and gritty across its centre has stopped being a bearing surface, whatever its measured thickness says.

Rubber bars sit at the other end of the trade. They grip the material rather than letting it slide, which helps a discharge stream settle quickly and reduces the forward skid that causes spillage at the skirt line. They also absorb energy in their own thickness instead of passing it straight through. The trade-off is wear rate, because rubber sliding under a loaded belt with grit between the surfaces generates heat and tears.

Composite bars with ceramic or steel inserts exist for the transfers where nothing else survives, usually where the lump is both heavy and sharp and the drop cannot be reduced. They work, and they carry a specific risk that has to be designed out. Any insert that stands proud of the polymer, or any insert that wears slower than the polymer around it, becomes a hard edge sitting in the belt's path. A ceramic tile that survives while its surrounding polymer wears away will eventually score the cover along its whole length.

Material choice also decides how the bed behaves in the wet. Polymer keeps a low friction coefficient when the surface is wet, which is one reason a bed that feels acceptable in dry weather can become a slipping hazard for the belt during washdown. Rubber goes the other way and can become stickier as it wets, which helps material settle but adds drag. Neither behaviour shows up in a dry bench test, and both show up on the first rainy week after commissioning.

Where the bed and the belt are bought through the same conveyor belt distributor, the profile and cover questions are usually settled before either item is manufactured, because one technical contact owns both drawings. That is not the normal case. More often the belt comes from a conveyor belt factory that never sees the chute, the bars come from a local fabricator, and the drive belts arrive through a transmission belt manufacturer working to a completely different schedule. Add a plant that standardises its spare drive belts through a single V-belt manufacturer and you have four parties each holding a piece of the same two metres of belt, none of them holding all of it.

That split explains a pattern we see repeatedly. Belts ordered as wholesale conveyor belts on width, ply and cover grade alone arrive with a construction that is correct on paper and wrong for the transfer, because nobody wrote down the lump size or the drop. The bed is then designed around a belt that has already been made, which is the wrong order of operations.

06Setting the Bed: Profile, Clearance and the Transition

A bed that is built correctly and set incorrectly behaves like a bed that was never needed. The setting work is not complicated, but it has three dimensions that interact, and getting two of them right still leaves a belt that either floats above the surface or drags on it along the whole length.

Start with the profile. The top face of the bars has to reproduce the trough the belt is already running in, which means a centre section plus two wings set to the same angle as the adjacent idlers. On a 35 degree trough the wings are set to 35 degrees and the joints between centre and wing are chamfered so the belt does not bridge a step. What must not happen is a bed built to one angle feeding a conveyor built to another, because the belt will be forced to re-form twice in three metres.

Then clearance. At rest, with an empty belt, the bar face should sit just below the belt line, close enough that the belt touches as soon as material lands and not so close that it polishes the bars while running empty. That band is a couple of millimetres, set with shims under the bar seats. Too high and the belt spans the bed as if it were a gap between idlers, which defeats the whole purchase. Too low and the belt runs on the bars permanently, adding drag and heat on every metre.

The transition back to rolls deserves its own attention because it is where beds get blamed for problems they did not cause. If the last bar sits higher than the first roll, the belt steps down and the material tips forward, which shows up as spillage twenty metres downstream. If the last bar sits lower, the belt steps up over the steel conveyor roller and the splice takes a small impact on every revolution, which shows up as a splice failure six months later. The two surfaces should be set with the same straight edge and the same reference, and the roll frame should be shimmed to the bed rather than the bed to the frame.

Skirt sealing completes the picture. The rubber should touch the belt with light contact along its full length and should never be pulled tight enough to press the belt onto the bars. A skirt that is too tight adds the load of a second sliding surface, and the bed's drag calculation quietly becomes wrong. A skirt that is too loose lets the stream spread, and the bars wear at the edges instead of the centre, which is the wear pattern an experienced fitter reads first.

07Wear, Loosening and the Two Inspections That Catch Both

An impact bed fails in two ways and they need two different inspections. The slow way is wear, which announces itself over months and can be planned for. The fast way is loosening, which announces itself as a bang on a Sunday night and a torn belt on Monday morning. Plants that inspect only for wear tend to be surprised by the second one.

Wear is measured, not estimated. Pick four points along the bed, stamp a number beside each, and record the remaining thickness from the bar face down to the top of the fixing counterbore. That last dimension is the one that matters, because a bar with plenty of polymer left but a counterbore breaking through is finished. The practical rule most maintenance teams settle on is to schedule replacement when 20 mm of material remains above the fixing and to stop running when 10 mm remains, since below that the bolt head starts acting as a cutting tool rather than a fastener.

The wear pattern tells you whether the bed is set correctly, and it is worth reading before ordering a replacement set. Even wear across the full width means the stream lands where the chute was designed to put it. Heavy centre wear with untouched wings means the discharge is concentrated and the chute needs a deflector or a rock box rather than thicker bars. Heavy edge wear means the skirt is leaking or the belt is not running centred. Wear at the upstream end means the bed starts too late.

On a recycling line we looked at last year, the bars had been in for fourteen months and the maintenance team was proud of how little they had worn. The thickness readings were fine. What was not fine was the cushion, which had taken a permanent set of 9 mm under the centre bars and had stopped springing entirely, so the belt had been landing on what was effectively a rigid block. The bars had done their job of not wearing out. The bed had stopped doing its job of removing energy. We replaced six cushion pads and left the bars in place.

The second inspection takes ten minutes with a torque wrench and a marker pen. Check every fixing for preload, mark each head with a paint stripe so movement is visible from the walkway, and look along the frame for rust streaks that start at a weld. A frame that has cracked at the rail end will let the whole bed drop a few millimetres on one side, and the first symptom is usually a belt that starts training toward that side for no obvious reason. Bed geometry and belt tracking are more closely linked than most maintenance schedules admit.

Spares policy matters as much as inspection frequency. A bed under a heavy transfer consumes bars on a schedule you can predict after the first replacement, so the second set should be on site before the first is finished. Holding one full set plus a few cushion pads costs far less than losing a shift, and it lets the team change a damaged bar when they find it rather than when the next delivery lands.

08What to Send When You Ask for a Bed

Most impact bed enquiries arrive as a belt width and a photograph. That is enough to quote something, and not enough to quote the right thing, which is why the first drawing often comes back with a note asking for the same information a second time. The data that actually changes the design is short, and none of it requires an engineering department to produce. Our service team works from exactly this list when a site survey is requested.

Start with the drop, measured from the chute lip to the belt surface with the belt loaded rather than empty. Then the top size and bulk density, the belt width and speed, the trough angle either side, and the length of the skirt boards. Add the distance from the discharge point to the first idler frame if the chute is offset, because an offset stream wears a pattern the drawing should anticipate.

Two interface questions are worth settling before the order rather than after. The first is who supplies the skirt rubber and the clamp bars, because a bed supplier who assumes the chute contractor is doing it and a chute contractor who assumes the opposite will leave a gap that fills with material on the first shift. The second is access. If the bars have to come out through the chute, the chute needs a door; if they come out sideways, the walkway needs the clearance. A bed that cannot be serviced is a bed that will be run to destruction and then replaced as a complete assembly.

None of this takes long to collect. It is the difference between buying a shape that fits and buying a support surface that removes energy from the belt, which is what the load point needed in the first place and the cheaper answer once belt replacements are counted honestly. Our conveyor idler factory RFQ checklist covers the surrounding idler data if the same review is extending past the transfer.

Request a load point bed drawing for your transfer

09Frequently Asked Questions

How do I know whether my transfer needs bars or just better idlers?

Look at where the damage is rather than at how much there is. If the cover is cut in a line that sits between two idler frames, the belt is being folded into an unsupported gap and no roll change will fix it. If the damage sits directly over a roll and the belt is otherwise sound, the roll pressure is the problem and impact rolls are the cheaper answer. A transfer with a drop over two metres and sharp lumps usually needs continuous support, because the energy is high enough that the gap itself becomes the failure mechanism.

Will a sliding support surface overload the drive or make starting harder?

It adds drag, and the amount depends on bed length, belt weight and whether the surface is wet. Over two or three metres the extra torque is usually absorbed by the margin a conveyor already carries, but a loaded incline with a thin margin will notice it at every start. Give the drive supplier the bed length and the friction figure, then check the starting ramp. Beds that run far past the settling zone are the ones that cause trouble.

What wears out first, the bars or the rubber underneath them?

Usually the bars, and that is the intended order. They are the cheapest part in the stack and the easiest to reach. The exception is a bed that has been hit hard enough to take a compression set out of the cushion pads, which then stop springing and turn the whole assembly into a rigid block. That failure is invisible from a thickness reading, so press a bar down by hand during the walk-past and see whether it returns. If it stays down, the pads are finished even when the bars look good.

Why did the belt keep getting cut after we moved to a heavier cover grade?

Because a heavier cover adds material in the path of the blow without taking energy out of the carcass. The cover buys time before it is breached, and once it is breached the plies underneath see the same load they always did. The second effect is worse, since thick cover hides deep cuts and the next lump opens them at the splice. If the cutting pattern did not change after the upgrade, the energy is still arriving in the same place and the support under the belt is still the same gap it was before.

Can I retrofit a support frame into an existing chute without rebuilding it?

In most cases yes, provided the stringers are sound and there is room to hang a frame below the belt line. The work is usually done through a chute door or from the walkway with the belt held and the section isolated. What decides feasibility is the fixing centres and the height under the belt, not the age of the conveyor. Send a sketch of both and the frame can be built to suit rather than cut to fit.

How long does the supported section need to run past the skirt boards?

Far enough that the material has stopped moving relative to the belt, which is a measurement rather than a rule. A practical starting point is one metre past the end of the skirt, then chalk the belt and run it loaded to see where the mark survives. Everything upstream of that point is still delivering energy into the surface. On fast belts the settling distance grows quickly, so a layout copied from a slow line will come up short.