SINOCONVE conveyor belt manufacturer & supplier makes conveyor belt more efficient.

Steel Cord Rubber Conveyor Belt: Compound, Cover Grade and Adhesion

2026-09-29

Steel cord belts rarely fail because the wire was too weak. They fail because the rubber wrapped around the wire was wrong. A fabric carcass belt and a steel cord belt can carry the same tonnage on paper, yet the compound that has to grip a brass-plated cord, survive an 1,800 mm pulley wrap at 4.5 m/s and still release cleanly under a scraper is a different recipe, with different ingredients, different cure kinetics and a far narrower processing window.

That gap is where most purchase disputes start. A buyer compares two data sheets, sees DIN Y on both covers and 1,600 N/mm on both cords, and assumes the belts are equivalent. They are not. The cover grade tells you what the rubber can resist at the surface; it says nothing about whether the skim compound between the cords will still be holding at 40,000 hours. On a 1,200 t/h coal line we pulled a sample at 9 months and found the cord-to-rubber peel value had dropped from 15 N/mm to 6 N/mm while every cover test still passed spec. The belt looked healthy from the walkway and was two months from a splice blowout.

Send us your duty, cord size and cover grade — get a compound recommendation and a sample data sheet

The three variables that actually decide service life are the polymer system in the core compound, the adhesion package that bonds rubber to the brass coating, and the cover grade chosen against the real abrasion and temperature picture on site. Everything else on the specification sheet is downstream of those three. This article walks through how each one is specified, where they interact, and what a conveyor belt manufacturer can and cannot change once the belt is cured. If you are running heavy ore or long-centre overland systems, the same logic applies to an industrial conveyor belt selection and it is worth reading the compound section before the cord section.

One caution before we start. Compound design is not a checklist exercise. Two mills can use nominally the same polymer, the same cobalt salt and the same brass coating thickness and still produce peel strengths 40% apart, because mixing temperature, dump time and press cure state move the result more than the formulation does. Treat the numbers below as targets to be verified by test, not as guarantees.

01Why a Steel Cord Carcass Cannot Share a Fabric Belt Compound

The first thing to understand is that a steel cord belt is a composite held together by adhesion, not by fabric weave. In a fabric belt the warp and weft threads carry tension and the rubber mostly protects them. In a steel cord belt the tensile members are individual wires running at zero crimp, and they carry nothing at all unless the rubber grabs them along their whole embedded length. Load transfer is entirely interfacial. Lose the interface and the belt does not lose a little strength, it loses essentially all of it at that point.

That changes the compound priorities. A fabric skim compound is optimised for penetration into a weave and for holding threads apart at the correct gauge. A steel cord core compound, often called the skim or cushion compound, has to do something harder. It has to flow between cords that sit only a few millimetres apart, wet a smooth brass surface that offers almost no mechanical key, cure into a stiff but not brittle matrix that resists cord movement under repeated bending, and resist the ageing effects of heat and oxygen diffusing in from the cover side.

Stiffness matters more than people expect. Cords packed at 12 mm pitch with 4.5 mm of rubber between them will migrate if that rubber is soft, and cord migration shows up as a wavy belt edge, uneven tension and eventually centre cord breakout at the pulley. The rubber also has to be stiff enough to keep the belt from cupping between the pulley and the cord plane, which is what causes the classic longitudinal split above a cord after a jam.

Then there is the question of what the belt runs on. Steel cord belts typically run on large diameter pulleys with high wrap angles, which means the core compound sees far greater bending strain per pass than a fabric belt of the same strength class. Every one of those bending cycles creates interlaminar shear between the cords and the surrounding rubber. A compound with high hysteresis overheats under that shear, and heat accelerates the same ageing that destroys adhesion.

Steel cord rubber conveyor belt showing brass-coated cords embedded in the core compound

So the practical rule we give buyers is simple. Never accept a quotation that offers one compound for the whole belt. Steel cord construction always needs at least two distinct recipes, and on fire-resistant or high-temperature duty it needs three or four. Any conveyor belt supplier who cannot tell you the core compound and the cover compound separately has either not asked the mill or is not going to control the process. Ask for both recipes in writing, then ask for the peel test results that prove the pair works together.

There is also a commercial reason to separate them. Core compound is expensive. It carries the adhesion promoters, the cobalt salt and often a higher share of premium polymer, so the mill has a strong incentive to thin it. When you see a quotation that undercuts the market by 15% on a 1,800 mm wide ST 1600 belt, the first place to look is not the cord count but the skim gauge and the core compound formulation.

02Polymer Choice and the Cure Window in the Core Compound

Most steel cord core compounds start from natural rubber, usually blended with a smaller share of SBR or butadiene rubber to control cure rate and tack. Natural rubber brings the green strength the belt needs before cure, because the cords have to be held in position while the slab is built, and it brings the tear resistance that keeps a cord from tearing out of the matrix under impact. SBR improves ageing resistance and reduces cure reversion, but push its share too high and you lose tack and tear resistance. The blend ratio is the single most consequential choice in the recipe.

What the blend is really managing is the cure window. The compound has to be flat-cured, meaning it must reach full crosslink density without overcuring, because a steel cord belt slab is thick and the cords conduct no heat. Heat has to travel through rubber to reach the centre. A 20 mm thick belt on a platen press can take over 40 minutes to bring the mid-cord zone to cure temperature, and by that point the cover surface has been sitting hot for the whole interval. Get the vulcanisation system wrong and the surface overcures while the core undercures, which is the origin of more adhesion failures than any formulation error.

On site, overcured cover shows as a hard, glossy surface that no longer absorbs the impact of lump feed and starts to crack in a fine mosaic pattern. Undercured core shows as cords that can be peeled out of the belt after a splice with a hand knife, or as the acrid smell of uncured rubber when a cut section is opened. Both are press faults, not material faults, and both are visible within the first month.

Temperature resistance is the second driver of polymer choice. A standard core compound based on natural rubber is comfortable up to about 60 °C belt surface temperature in continuous duty. Above that, the natural rubber matrix softens, adhesion value falls and the belt starts to stretch under load faster than its cord elongation curve predicts. For hot clinker or sinter feed we move the core to an EPDM or chlorinated polymer base and rebuild the adhesion package around it, because the cobalt system that works with natural rubber behaves differently in a saturated polymer.

Compound element	Typical level in an ST core compound	What it does at the cord interface	What goes wrong if it is out of range
Natural rubber, plus SBR or BR blend	65–85 parts NR in the polymer base	Provides green strength so cords stay at pitch during building, and supplies tear resistance around each wire	Too much SBR drops tack and the cords shift during lay-up; too much NR and the cured matrix reverts on a hot vulcaniser
Cobalt salt adhesion promoter	0.8–1.5 phr as cobalt metal equivalent	Drives the sulphide bridge that links rubber to the zinc in the brass coating during cure	Below 0.8 phr the peel value never develops fully; above 2 phr the bond becomes brittle and fails under bending
Sulphur and accelerator package	Semi-EV system, 2.5–4.5 phr sulphur depending on thickness	Sets crosslink density and controls how long the cord zone has to reach temperature before scorch closes the window	A conventional high-sulphur system scorches early on thick slabs and leaves the mid-cord zone starved of free sulphur
Anti-reversion and anti-ageing agents	1–2 phr combined, plus 2–3 phr zinc oxide	Hold crosslink density through the long press cycle and slow oxygen attack from the cover side	Reversion shows as a soft, spongy cord zone in the middle of the belt cross-section and a slow loss of peel strength
Fillers, mainly carbon black and silica	35–55 phr, with silica limited to a few phr in the cord zone	Builds the stiffness that stops cord migration and controls heat build-up under repeated bending	Too little filler and cords migrate and the belt cups; too much and the compound cannot wet the brass surface

Two things are worth noting from that table. First, the adhesion promoter is a narrow window, not a case of more is better. Second, the filler load and the adhesion chemistry pull in opposite directions, which is exactly why the core compound has to be designed as a system and tested as a system.

03Brass Coating and the Adhesion Chemistry That Actually Holds the Cord

Steel cord does not bond to rubber by mechanical grip. The wire is drawn and then electroplated with a thin brass layer, typically 0.20 to 0.35 µm thick, with a copper to zinc ratio somewhere between 60:40 and 70:30. During cure, sulphur from the compound reacts at that brass surface and builds a copper sulphide film that is chemically continuous with both the metal and the crosslinked rubber. The bond is a chemical bridge a few nanometres deep, and everything about it depends on the copper to zinc ratio, the coating thickness and the amount of sulphur that reaches the interface in the right time window.

This is why the cord and the compound cannot be specified independently. Change the coating ratio and the cure kinetics of the compound have to change with it. A 67:33 coating that bonds beautifully to a semi-EV natural rubber system at 145 °C can bond poorly to a low-sulphur system, because not enough sulphur diffuses to the interface before the matrix gels. The matrix gelling is the key constraint. Once the rubber around the cord has crosslinked enough, sulphur mobility collapses, and any copper sulphide that has not formed by then will never form.

Coating thickness cuts the same way. Below about 0.15 µm the brass layer is too thin to sustain the reaction, and the bond is weak from day one. Above roughly 0.40 µm the coating becomes a ductile layer in its own right, and under repeated bending it deforms differently from the wire underneath. That mismatch is one of the mechanisms behind the corrosion-driven failures we see on wet coal lines. In one case a coastal coal terminal kept losing cord adhesion at roughly 18 months on an ST 1000 belt carrying washed coal at 8% surface moisture. The coating was 0.45 µm. Moving to 0.28 µm with the same compound almost doubled the time to first peel failure.

Water is the invisible variable in all of this. Even a perfectly formed bond degrades if moisture reaches the brass surface, because the sulphide film hydrolyses and the interface converts to a non-adherent oxide. That is why a skim compound has to be genuinely impermeable, and why a thin cover with an under-vulcanised interface is so dangerous. Water finds the path of least resistance and that path is usually the cord plane.

Interface variable	Practical target range	Mechanism at the rubber-brass boundary	Failure signature you will find in the field
Copper to zinc ratio in the plating	60:40 to 70:30, verified by cord supplier certificate	Sets how readily copper sulphide forms and how stable that film stays under load	Peel values scatter widely between cord lots, and the spliced zone fails before the body of the belt
Plating weight or thickness	0.20–0.35 µm, with a tolerance the mill must state	Too thin gives no reaction substrate; too thick creates a ductile layer that debonds under flexing	Wet duty belts lose adhesion at the bottom cover first, where water pools and flexing is highest
Free sulphur reaching the interface	Semi-EV system sized against belt thickness and press temperature	Sulphur must diffuse to the brass before the surrounding matrix gels and mobility collapses	Undercured core, rubber sticking to the cord on hand peel rather than shearing cleanly
Moisture at the interface	Cord stored dry, compound conditioned, no condensation before building	Water hydrolyses the sulphide film into a non-adherent oxide layer along the cord	Longitudinal rust streaks on cords inside a cut belt, with black oxide paste between the wires
Cure temperature and dwell at the cord plane	Plate temperature and time sized to reach the mid-cord zone, not just the surface	The reaction needs both the peak temperature and enough time before the matrix sets	A hard, glossy cover over a soft core, which is the most common press fault we audit

If you want one number to argue about with a mill, argue about the cord-to-rubber peel test. The standard method pulls individual cords out of a cured sample at a controlled rate and records the force per unit embedded length. For heavy duty ST belts we look for values that hold above 12 N/mm after ageing, and we want the failure surface to show rubber left on the wire rather than a clean brass finish. A clean wire means the bond gave way; rubber left behind means the rubber itself tore, which is the failure mode you want because it proves the interface held.

04Choosing Cover Grade Against Real Abrasion, Impact and Temperature

Cover grade is the part of the specification buyers read first and understand least. The letters and numbers come from different standards that do not map onto each other cleanly. DIN 22131 governs steel cord belts for general conveying and grades the cover by abrasion loss and tensile properties. ISO 15236 covers steel cord belts for general use and uses a similar but not identical grade structure. For the rubber itself, DIN 22102 defines the cover classes and ISO 4649 defines how abrasion loss is measured. RMA grades come from a different tradition altogether and are widely used in North American aggregate and mining work.

The confusion matters because a belt quoted as DIN Y and a belt quoted as RMA Grade II are not interchangeable in service, even though both are marketed as abrasion resistant. Abrasion loss is measured in cubic millimetres of rubber removed from a standard specimen under a defined load and abrasive sheet. Lower numbers are better. What the figure does not tell you is how the compound behaves under impact, under heat, or in the presence of oil, and those are usually the conditions that decide life on a real line.

Impact is where the abrasion number misleads most. A very hard, highly abrasion-resistant cover can be brittle, and a 300 mm lump of ore dropped from a transfer chute at 6 m/s will cut it and start a tear. On high-impact transfer points we often specify a slightly higher abrasion loss in exchange for much better cut and tear resistance, and we back it with a thicker cover and sometimes a separate cushion layer. This is the kind of trade the data sheet does not make for you.

Cover grade	Where it comes from and how it is defined	Roughly what it tolerates in service	Typical conveying duty
DIN Y, ISO 15236 grade Y	Abrasion resistant class measured to ISO 4649 method A, with a specified tensile strength and elongation at break	Good general abrasion resistance at moderate impact energy and belt surface temperatures up to around 60 °C	Iron ore, coal, gravel and crushed stone on long overland systems with controlled transfer heights
DIN W, ISO 15236 grade W	Wear class with a lower permitted abrasion loss than Y, usually with a higher hardness compound	Higher resistance to sliding abrasion, but less forgiving under sharp impact and on tight pulley geometry	Sharp crushed granite, sinter, coke and screened product where the abrasive is angular but the drop is small
Heat resistant grades per DIN 22131 and supplier data	Cover and core both rebuilt on a heat-stable polymer, tested for retention of tensile and elongation after oven ageing	Continuous material temperatures from about 80 °C to 120 °C, with short peaks above that depending on grade	Cement clinker, sinter feed, foundry sand and hot ash where the belt never fully cools between shifts
Fire resistant grades to AS 1332 or equivalent national standard	Cover and core compounds tested for flame propagation and, in some regimes, electrical resistance, on defined specimens	Limited flame spread after ignition, which is a safety property and not an abrasion property at all	Underground coal, enclosed transfer towers and any gallery where a frictional ignition could propagate
RMA Grade I and Grade II	North American classification of cover compounds by abrasion resistance, widely referenced in aggregate and hard rock work	Grade II covers general abrasive duty, Grade I covers severe abrasive duty with higher wear resistance	Quarry primary crushing, aggregate stockpiling and portable plant conveyors in North American operations

Notice that neither DIN nor RMA has a grade that describes oil resistance, chemical resistance or low temperature flexibility. Those are separate compound families. If your belt sees hydraulic oil drips under a crusher, or runs outdoors at minus 25 °C in winter, you are outside the graded cover system and need to specify the compound directly against the condition. We have written separately about oil resistant conveyor belt compounds and about abrasion resistant conveyor belt constructions, because those conditions need their own conversation. Heat is a third separate family, and a belt facing clinker at 110 °C is specified through the heat grade route rather than through an abrasion class.

Power transmission is a different branch of rubber engineering again, and a transmission belt manufacturer works to compound rules that do not transfer across to a conveying cover. Those belts run against grooved or toothed pulleys, so their limiting property is usually heat build-up at the flank or tooth, not sliding abrasion against bulk material. Mixing the two specification languages is a common cause of disappointment on both sides of a project. Keep the drive belts and the conveying belts in separate sections of the same enquiry, and let the supplier answer each on its own terms.

One more point on covers before we leave the subject. Cover thickness is not a substitute for cover quality. Doubling cover gauge on a poor compound buys maybe 30% more life and adds mass, weight on the pulleys and cost. Fixing the compound and the adhesion underneath typically buys a multiple.

05Four Ways Adhesion Fails, and What Each One Looks Like on Site

When a steel cord belt comes apart, the failure almost always reduces to one of four mechanisms. They look similar from the walkway and completely different on a cut section, which is why the diagnosis has to start from a physical sample rather than from a photograph of a torn belt. Getting the mechanism right decides whether you need a new belt, a different compound specification, or simply a repair to the press cycle.

Ageing and reversion of the core compound

The most common mechanism is slow chemical ageing. Heat and oxygen reaching the cord plane gradually reduce crosslink density and break down the interfacial sulphide layer. Peel strength falls from 15 N/mm towards 5 N/mm over months, and the belt keeps running because the remaining bond is still enough for normal tension. Then a jam, a stalled pulley or a chute blockage puts a short term overload into the cords, the weakened interface cannot redistribute it, and a long section of top cover separates from the carcass. On cut sections you find the cord zone soft and slightly greasy, with a reddish-brown oxide film on the wires.

We saw this pattern clearly on a limestone overland conveyor running 24 hours a day at 90 °C surface temperature. The belt was a heat grade on paper, but the core compound had been formulated on a conventional natural rubber system with only a modest anti-ageing package. Peel values at 14 months were 6.5 N/mm and falling. The replacement used the same cord and an EPDM based core at roughly 18% higher unit cost, and at 30 months the peel values were still above 11 N/mm.

Moisture ingress along the cord plane

The second mechanism is water. Once moisture reaches the brass surface it hydrolyses the sulphide film, converts it to a non-adherent oxide and the bond is gone permanently. Water enters through a cut, a damaged splice, a worn cover at the edge, or through a porous under-cured interface between cover and core. The signature is longitudinal rust staining and black oxide paste inside the belt, often extending far beyond the visible damage on the surface.

Wet duty is where this bites, and it explains why a rubber conveyor belt specified for a dry pit performs differently from one specified for a washed product line even when the cords are identical. Ask specifically how the mill controls moisture before building. Cord that has sat in a humid warehouse, or a slab built the day after a cold night, carries condensation at the interface and the press cannot cure it away.

Mechanical fatigue and cord migration

The third mechanism is mechanical rather than chemical. Small pulley diameters, high wrap angles and frequent starts create interlaminar shear in the cord zone. If the core compound is too soft, cords migrate laterally, the cord plane distorts and stress concentrates at the outer cords. The visible result is a belt that tracks poorly, then shows a wavy edge, then develops a longitudinal split directly above an outer cord. On a 1,400 mm wide ST 1250 belt at a transfer station, we measured 6 mm of lateral cord movement over 220 m of belt length on a compound that was 12% below specified hardness. Replacing the belt with the correct compound removed the tracking problem entirely within one week.

Impact damage belongs in this family too. A sharp lump striking the belt compresses the cover and the core, and if the compound has poor tear resistance the cord can be driven through the adjacent rubber, starting a rupture that propagates along the cord plane. That is why high impact transfer points need both a thicker cover and a compound with real tear strength, not just a low abrasion number.

Press and process faults that imitate material faults

The fourth mechanism is not a failure of the belt in service at all, but a failure of the press cycle that only becomes visible later. Insufficient platen pressure leaves porosity in the cord zone. A cool platen at one end of a long press leaves a gradient of cure state across the belt width. A press that opens too early leaves the mid-cord zone undercured while the surface reads correct on a hardness check. All of these show up as scattered, unpredictable peel values that no formulation change will fix.

This is the practical argument for a supplier audit before a large order, and we have covered the audit framework in our notes on vetting a wholesale conveyor belts programme. A mill with a calibrated press, a recorder on platen temperature and pressure, and a written cure cycle for each belt thickness will beat a mill with a better recipe and no process control every time. When you visit a potential conveyor belt factory, ask to see the last three press logs for a belt of your thickness and ask what the measured variation across the platen was.

Failure mechanism	Trigger or root cause	How to confirm it on a cut sample	What actually fixes it
Ageing and reversion	Sustained heat above the compound rating plus oxygen diffusion into the cord plane over months of running	Cord zone feels soft, slightly greasy, reddish oxide on wires, peel values below 8 N/mm while cover tests still pass	Move to a heat stable polymer base with a larger anti-ageing package, and control belt surface temperature at the source
Moisture ingress	Wet process material, cover damage, edge wear, or condensation on cords before the slab was built	Longitudinal rust streaks along the cords with black oxide paste, damage spreading well past the external cut	Impermeable skim compound, reinforced edge protection, dry cord storage, and prompt repair of any cover breach
Fatigue and cord migration	Small pulley diameter, high wrap angle, frequent starting, or a core compound below the specified hardness	Wavy edge, measurable lateral cord displacement, longitudinal split sitting directly above an outer cord line	Increase core stiffness, review pulley and roll geometry, and bring the belt back inside its rated tension window
Press and process faults	Low platen pressure, uneven platen temperature, or a cure cycle short for the belt thickness involved	Scattered peel readings along the belt length, porosity visible in the cord zone, hard glossy cover over a soft core	Calibrated press with logged temperature and pressure, a written cure cycle per thickness, and peel testing per production run

One habit worth building is to cut a sample from every belt on first delivery and keep it in a dark, dry store with the production date. When something fails four years later you have a baseline to compare against, and the comparison usually settles the argument about whether the belt was supplied wrong or worked outside its rating.

06Core Compound and Cover Compound Are Two Different Jobs

Buyers often assume the belt is one rubber with a pattern on top. In reality the core and the cover are designed against opposite requirements, and the fact that they are pressed together does not make them the same material. The core wants maximum adhesion to brass, high stiffness to hold cords at pitch, low heat build-up under flexing, and long-term resistance to ageing from the inside. The cover wants abrasion resistance, cut and tear resistance, low rolling resistance, and its own ageing performance against sunlight, ozone and material temperature from the outside.

Because those requirements pull differently, the two compounds are cured as a stack. The cover compound is laid over the core, they cure together in the same press cycle, and their cure systems have to be matched so that both reach their target state in the same cycle. If the cover cures much faster than the core, the interface between them becomes a weak plane. That weak plane is a common entry path for water and it is invisible from outside.

This matched-cure requirement is also why changing a cover grade is not a trivial substitution. Moving a belt from DIN Y to a fire resistant cover changes the cover compound family entirely, and the core compound usually has to change with it to keep adhesion and cure behaviour consistent. Any conveyor belt distributor who treats cover grade as a free option on an order form is creating a future failure. The two must be quoted as a pair, and the mill should be able to state cure compatibility in the data sheet.

Property	Core compound target	Cover compound target	Why the two are not the same	How it is checked
Primary job	Bond to brass coated cord and transfer load between wire and rubber	Protect the carcass from abrasion, impact, heat and the conveyed material	One is an interface material, the other is a sacrificial wearing surface	Cord pull-out test for the core, abrasion and tear tests for the cover
Hardness and stiffness	Relatively stiff for the class, to stop lateral cord movement under load	Softer in general, so it can absorb impact and conform to lumpy material	A stiff cover would crack at the impact point; a soft core would let cords migrate	Shore hardness on cured slabs, plus cord displacement measured on a running belt
Abrasion resistance	Not a design target, since the core never sees the material	Defined by the grade, such as DIN Y or W, with a stated abrasion loss figure	Optimising the core for abrasion would sacrifice the adhesion that matters there	ISO 4649 abrasion test on cover specimens, reported per production batch
Moisture and ageing protection	Needs to be as impermeable as the compound family allows, to keep water off the brass	Needs ozone and UV resistance on the exposed face and low water uptake at the edges	Different exposure conditions on each side of the same belt, so different additives	Water absorption and ageing tests, plus peel retention measured after oven ageing
Cure behaviour	System sized so the mid-cord zone reaches full cure before the press opens	System matched to the core so both reach target state in one cycle without scorch	A mismatch creates a weak plane between cover and core, which water will find	Rheometer curves for both compounds, plus press logs showing platen temperature

There is a related point about belt width and edge compound. The edge of a steel cord belt is where the outer cords sit closest to the surface, and it is where most damage starts. Many designs use a dedicated edge compound with higher tear resistance and sometimes a reinforced edge strip. If you are buying through a V-belt manufacturer who also supplies flat belting, ask whether the steel cord range uses a separate edge recipe or trims the belt from a wider slab without edge protection. The difference shows up in the second and third year of service, and by then the order is long finished.

07Where Steel Cord and Fabric Carcass Compounds Diverge

Fabric belts and steel cord belts often sit on the same specification sheet and get compared on the same columns, which is a mistake. The two carcass types place different demands on the rubber, and the compound differences are not cosmetic. Once you understand the divergence you can read a data sheet properly, and you can tell when a mill is offering a fabric belt recipe dressed up as a steel cord recipe.

The biggest difference is cord spacing. A fabric belt has a woven ply, so the rubber sits in a continuous sheet with threads embedded at close pitch. A steel cord belt has widely spaced, large diameter cords, sometimes 30 mm apart, with substantial volumes of rubber between them. That rubber is not filler. It is a structural element that has to be stiff enough to resist shear and heat build-up while remaining flexible enough to pass over pulleys, and the design problem is entirely different from a fabric skim.

Adhesion requirements diverge even more. Fabric plies need the rubber to penetrate the weave and anchor to the fibre surface, which is a mechanical and chemical process with the textile finish. Steel cord needs a chemical bridge to brass, as described earlier. The additive packages are different, the required sulphur availability at the interface is different, and the test methods are different. A compound optimised for fabric adhesion will not deliver strong brass adhesion, and vice versa.

Then comes stiffness and hysteresis. Steel cord belts are usually much longer, carrying higher tension over longer centres, and they flex far less often per unit length because pulley diameters are large. What they cannot tolerate is a compound that builds heat and retains it. A fabric belt on a short, fast conveyor with small pulleys is the opposite case, and low hysteresis matters more there. The same optimisation objective points in different directions depending on carcass type.

Design variable	Fabric carcass compound	Steel cord core compound	Practical consequence for the buyer
Cord spacing and rubber volume	Threads sit close together at a few millimetres pitch inside a continuous rubber sheet	Large wires at wide pitch, with substantial rubber volumes that act as structural material	Steel cord skim gauge and stiffness have to be engineered, not inherited from a fabric recipe
Adhesion mechanism	Rubber penetrates the weave and bonds to the treated fibre surface, partly mechanically	Chemical bridge from rubber to copper sulphide on the brass plated wire, with no mechanical key	Different additive packages; a fabric compound cannot be substituted into a cord belt
Typical pulley geometry	Smaller pulleys and higher flex frequency, so low hysteresis matters more per pass	Large diameter pulleys and long centres, so heat retention and slow ageing dominate	Specify the compound against the drive layout, not only against the material being carried
Tension distribution	Load is spread across many threads, so a single local defect has limited effect	Load concentrates in relatively few wires, so one lost bond concentrates stress nearby	Steel cord belts need better protection at edges, splices and transfer points
Repair philosophy	Local repairs and patch splices are common and usually adequate in the field	Splice and core integrity are critical, and field repairs are far more demanding	Budget for proper splice materials and training, not just for the belt itself

One related point about specialist constructions. A cleated incline belt built on a steel cord carcass, or a sidewall construction with corrugated flanges, adds a second cure operation for the profiles. That second cure heats the carcass again, and the core compound has to be stable enough to survive it without reverting. Not every mill can do this, and it is a good question to ask early if your project needs both high tension and steep incline capability.

The same logic applies when the duty is hot, wet or hazardous rather than merely steep. A heat resistant conveyor belt and a fire resistant conveyor belt both need the core compound rebuilt as well as the cover, and a transfer point handling sharp lumpy feed usually calls for an impact and cut resistant conveyor belt with a cushion construction rather than a harder cover. In each case the specialist property has to be carried through the full thickness, which is why these belts cost more than a standard grade and why they are worth it.

Steel cord rubber conveyor belt installed on a bulk material handling conveyor at a port terminal

For heavy duty applications we normally see steel cord specified in one of four places, and each one pushes the compound in a slightly different direction. Long overland lines in mining and quarrying put the emphasis on low elongation and consistent adhesion. Clinker and kiln feed duty in a cement plant puts the emphasis on heat stability in both the cover and the core. Ship unloaders and stackers in port bulk handling put the emphasis on impact and edge protection. High tonnage downhill lines put the emphasis on controlled elongation and brake behaviour, because a steel cord belt at 0.25% elongation on a 3 km centre returns a lot of energy when it stops.

08Curing, Testing and Batch Consistency on the Press

Everything written so far describes what a compound should be. Whether the belt you receive actually contains that compound depends on mixing control and on the press, and this is where the gap between mills is widest. A formulation is a document; a belt is a process output. Buyers who audit only the document get surprises.

Mixing comes first. Rubber compounds are mixed in internal mixers with a defined fill factor, rotor speed and dump temperature, usually in two stages with the accelerator added in the second stage at a temperature low enough to avoid scorch. Dump temperature variation of plus or minus 10 °C between batches changes the state of cure of the batch before it ever reaches the press, and that variation propagates all the way to the cord interface. A mill that logs dump temperature per batch has control. A mill that only logs it when something looks wrong does not.

Then the press. Steel cord belts are cured on platen presses that heat through the cover to reach the cords, and the cord plane is the last place to reach temperature. Pressure matters as much as heat, because the compound has to flow into the cord gaps and expel trapped air. Too little pressure and you get porosity, which is a void network along the cords that water will occupy within a season. Too much pressure early in the cycle and the compound is squeezed out of the cord zone before it has cured, leaving the cords under-covered.

After cure the belt is cooled under restraint, otherwise it shrinks unevenly and the cord tension distribution becomes non-uniform from edge to centre. A belt with uneven cord tension tracks badly and wears its edge. That is a press and cooling line problem, not a compound problem, and no amount of premium rubber will fix it.

Testing is the only way to confirm what was made. We ask for four groups of test data with every steel cord order, and we compare them against the pre-production sample rather than against a generic specification. Cord pull-out strength gives the interface number. Cover abrasion loss to ISO 4649 gives the wear number. Tensile strength and elongation at break on both cover and core give the mechanical picture. Oven ageing tests, typically 7 days at 70 °C or a defined equivalent, give the retention figures that predict long term behaviour.

Cured cross section of a steel cord rubber conveyor belt showing cover core and brass coated cords

Test or control point	Method or standard commonly referenced	Frequency we ask for	Acceptance logic and what it catches
Cord to rubber peel or pull-out	Mill standard method aligned with ISO 15236 or DIN 22131 requirements	Per production run, with a defined number of cords and a reported minimum	Catches adhesion promoter errors, moisture at the interface and undercured cord zones
Peel retention after ageing	Oven ageing, typically 7 days at 70 °C, followed by a repeat peel test	Per order, plus once per quarter for the standard production grade	Predicts long-term bonding; a large drop here means the belt will fail early even if new-belt peel looks fine
Cover abrasion loss	ISO 4649 method A, with a rotating drum and a defined abrasive sheet	Per batch of cover compound, reported as a maximum permitted loss	Confirms the declared cover grade and detects batch-to-batch shifts in filler or polymer
Full thickness tensile and elongation	Belt specific test on the completed belt, referenced to the declared strength class	Per production run on a cut sample from the belt itself	Verifies cord count, cord grade and that the cords were not damaged during building
Press log with temperature and pressure trace	Recorded platen temperature and hydraulic pressure for the whole cure cycle	Every belt, retained and traceable to the belt serial number	Catches uneven platen temperature, pressure loss and short cure cycles before they reach the customer
Cord supplier certificate and lot traceability	Cord manufacturer certificate covering plating ratio and plating weight per lot	Per cord lot, matched to the belt serial numbers it was used in	Catches plating variation that would otherwise appear as unexplained scatter in peel results

A word on how to read scatter. No mill produces identical peel values across a whole belt, and claiming to is a warning sign. What you want is a tight distribution with a minimum above the agreed floor. If the average is 14 N/mm and the lowest three cords are at 7 N/mm, the belt has weak zones and those zones will fail first. Ask for the minimum, not the average, and ask how many cords were tested.

The same principle applies to components around the belt. A pulley with a crowned face and a smooth, correctly profiled lagging protects the cord zone by keeping the belt tracking without edge stress. A properly selected troughing idler set with the right trough angle keeps the load centred and stops material from spilling onto the return run, which is what causes most edge damage. And a self-aligning idler at the critical change points reduces the lateral forces that make a marginal cord-to-rubber bond fail early. A belt with perfect adhesion can still be destroyed in a season by poor idlers, and a belt with adequate adhesion running on good components will often outlast its rating.

For buyers managing a full conveyor package rather than a single belt, it is worth reading how we structure conveyor components supply and how our quality assurance process handles batch records, because the compound discussion on this page only becomes real once the supplier can produce those records on request. If your application involves a second process step such as a vulcanised splice in a remote location, our field service notes cover what to prepare before the crew mobilises.

Finally, a note about the two words that get used interchangeably and should not be. A synchronous drive belt and a plain V-belt both deal with power transmission rubber, but the compound logic is different again, because those belts run on toothed or grooved pulleys and their failure modes are tooth shear and cord fatigue rather than abrasion. If your project includes drive belts alongside conveying belts, keep the specifications separate.

Request a compound and cover grade review for your steel cord belt

09Frequently Asked Questions

Can one compound serve both the core and the cover of these belts?

In practice, no. The core has to bond chemically to the brass coated wires and stay stiff enough to hold them at pitch, while the cover has to resist abrasion and impact on the outside. Mills that quote a single compound for the whole belt are usually cutting the adhesion package out of the design. Ask for two recipes and for the cure behaviour of the pair.

Why does cover grade alone not tell me how long a belt will last?

Because the cover grade describes the wearing surface only. It says nothing about the adhesion at the cord interface, the heat stability of the core or the press control behind the belt. We have pulled samples from belts that passed every cover test at nine months while the peel value at the cord plane had already halved.

What peel value should I put in a purchase specification?

For heavy duty ST class duty, look for a new-belt cord pull-out around 12 N/mm or higher and for that value to stay above roughly 8 N/mm after oven ageing. Also specify that the failure surface must show rubber adhering to the wire. A clean brass wire after the test means the interface gave way rather than the rubber.

How does moisture get into the cord plane, and how do I prevent it?

It enters through cover cuts, damaged edges, poor splices and porous undercured interfaces, and it destroys the bond chemically rather than washing it out. Prevention means an impermeable skim compound, good edge protection, dry cord storage before building, and repairing any cover breach quickly instead of waiting for the next shutdown.

Is a higher cover thickness a substitute for a better compound?

No. Adding cover gauge on a mediocre compound adds mass, pulley load and cost for perhaps 30% more life. Fixing the adhesion and the core compound typically buys a multiple of that, and on hot or wet duty the difference is larger still because the failure starts underneath the cover, not on its surface.

What should I ask for with the delivery documents?

Four things. The cord pull-out test result with the minimum value and the number of cords tested. The abrasion loss figure against the declared grade. Tensile and elongation for both cover and core. And the press log for your belt serial number, showing platen temperature and pressure through the cure cycle.