Conveyor Belt Carcass: Fabric Plies, Steel Cord and How to Read a Spec

Every conveyor belt is really two products bonded into one. On top sits the cover compound — the rubber a quarry manager points at when a belt fails, the layer every sales sheet leads with, the one that gets argued about in meetings. Underneath sits the carcass: a woven fabric skeleton or a bed of parallel steel cables that carries every tonne the belt will ever move. Strip a belt down and the cover is often only about a third of the total thickness. The structural work belongs to the layer nobody photographs.

The carcass decides the belt's tensile rating, how far it stretches under load, how cleanly it troughs, how stubbornly it resists a rip, and how many flex cycles it survives before fatigue wins. Two belts with identical abrasion-rated covers can behave nothing alike in service, because one holds four plies of EP fabric and the other a single layer of steel cord. Judging them on cover grade alone is like assessing a bridge by the quality of its asphalt. This guide stays at the carcass layer: what the numbers in a specification table actually govern, how fabric and steel constructions differ, and how to match a skeleton to the duty instead of the brochure.

Send us your duty and we will recommend the right carcass build

01What a Conveyor Belt Carcass Actually Does

Think of the carcass as the belt's load path. Tension from the drive pulley travels along it, distributed through fabric warp threads or steel cables, and every tonne of material sitting on the cover is transferred into that path through the bond between cover and carcass. When a belt fails in tension — a snap, a stretch it never recovers from, a splice that pulls apart — the carcass is almost always the part that ran out of capacity. The cover may look worn, but it rarely decides whether the line stops.

The properties that matter flow directly from that job. Tensile strength sets how much pull the belt can take per millimetre of width, quoted as a nominal figure such as EP630 at 630 N/mm or ST1000 at 1000 N/mm. Elongation governs how much the belt stretches at working tension, which in turn decides take-up travel and whether the drive slips on start-up. Troughability — how readily the belt folds into a three-roller idler set — comes from the carcass's transverse stiffness, not from the cover. Tear resistance and fatigue life depend on the weave or cable layout and on how evenly load spreads between adjacent strength members.

A belt with a strong carcass but a soft or badly bonded cover wears out early at the edges. A belt with a heavy cover but a weak carcass stretches until the counterweight bottoms out. I have watched a 1,200 t/h coal line where the operators blamed the cover for constant edge cracking; the real fault was a two-ply carcass flexing far more than the pulley diameter allowed, and the fix was construction, not compound.

This is the layer a conveyor belt manufacturer reasons about first when a duty arrives. Cover grade answers what the surface must survive. Carcass answers what the structure must carry. A serious industrial conveyor belt brief names both, along with the pulley diameters and the splice method. When only the cover is specified, the carcass gets chosen by habit, and habit is how under-built belts reach site. If you want the full picture of how covers and carcasses combine across a range, the types and grades walkthrough on our technical pages covers the range logic, while the product page for high-tension cord construction shows how one carcass family is built up.

The four numbers that carry the whole decision

Tensile rating, ply or cable count, working elongation, and minimum pulley diameter form the spine of every carcass decision, and they interlock. Raise the rating and the carcass usually thickens, which pushes up the minimum pulley diameter you can run without over-flexing the members. Shrink the pulley to fit a tight transfer and you must accept a thinner, lower-rated carcass or pay for a more flexible construction. Nothing here can be optimised in isolation, which is why a carcass specification that lists only the strength number is only half a specification.

02Fabric Carcass Families: EP, NN and CC Compared

Most belts in the world are built on a woven fabric carcass, and almost all of those use one of three thread materials. Polyester warp with nylon weft is written EP. All-nylon is written NN. Cotton is written CC, and it survives mainly in legacy or low-duty roles where its moisture behaviour is tolerated. The letters describe the threads, and the threads decide the mechanical character of the belt far more than any cover choice.

The reason EP became the default is elongation. Polyester barely stretches along the warp, so an EP belt holds its length under load and the take-up does not have to travel far. Nylon stretches more and recovers well, which suits impact and shock loading, but an all-NN belt tends to grow over its life so take-up capacity must be generous. Cotton has the lowest strength per ply and the worst relationship with water — it swells, rots, and loses strength — so it appears where belts are short, slow, and cheap.

What surprises buyers is how differently the same nominal rating behaves depending on the thread. An EP500 and an NN500 can both be stamped 500 N/mm, yet under an identical load the NN belt will stretch noticeably more and absorb more of the shock, while the EP belt stays dimensionally stable and transfers the shock to the splice. Neither is "better". The question is whether the duty is dominated by steady tension, where stability wins, or by impacts and starts under load, where a little give helps.

Reading a fabric datasheet, keep an eye on how the threads are laid. Warp runs lengthwise and carries the tension; weft runs across and holds the belt in shape. Almost every "EP" belt is polyester in the warp and nylon in the weft for exactly that reason — stable length, flexible width. A belt that advertises EP but lists nylon in the load-carrying direction is a different animal, and it will stretch more than the rating label suggests. If your belt is destined for heavy duty in mining or port work, pair the carcass choice with the right cover grades by looking at the steel-reinforced range and reviewing what a tensile and carry-capacity breakdown looks like in practice, since many high-tension duties end up in the cord family rather than the fabric family.

03Multi-Ply vs Single-Ply Fabric Construction

A fabric carcass is not one sheet. It is several thin woven plies, each bonded to the next by a friction compound, then sandwiched between the top and bottom covers. Two plies, three plies, four, sometimes five or six. The ply count is one of the most misunderstood fields on a datasheet, because more plies does not automatically mean a stronger or better belt. It means a different balance of flexibility and stiffness, and that balance has to suit the conveyor it runs on.

Each ply contributes strength, so a three-ply EP200 gives roughly the same total rating as a two-ply EP315 — the arithmetic is nominal and the real numbers come from the fabric's own rating times the count. But the two constructions behave differently as they wrap a pulley. More plies under the same total thickness make the belt stiffer lengthwise and harder to trough, and they raise the minimum pulley diameter the belt can tolerate before the outermost ply is over-stretched in bending. Fewer plies give a more flexible belt that hugs a small pulley and trains easily, at the cost of a thinner, less damage-tolerant body.

Single-ply construction exists for a specific reason. When a belt must run over small pulleys, snake through tight transfer chutes, or turn on a short centre, a single strong ply of high-tenacity fabric gives the flexibility the geometry demands. It also makes a thinner, lighter belt that a small drive can move. What you give up is protection against a knife or a lump tearing straight through the only ply there is, which is why single-ply belts tend to be paired with rip-resistant or cut-resistant covers in rough duty.

"How many plies?" is the wrong first question. The right one is "what is the total nominal rating, and what is the smallest pulley on the line?" Match those two and the ply count usually falls out. When it does not, the conveyor geometry is asking for a construction the rating alone cannot deliver, and that is a signal to look at the cord family instead. Buyers who specify strictly by ply count often order a heavier belt than the duty needs, then discover it will not trough. If you are building a range for a plant, our conveyor belt factory can quote both a multi-ply and a single-ply build for the same duty so the trade-off is visible, and comparing a rubber conveyor belt build at two ply counts is often the fastest way to settle it.

One more thing about plies that trips people up at the splice. Each ply has to be stepped and bonded to its opposite number when a belt is joined, so more plies means more finger steps, a longer splice, and more chances for a poorly bonded layer to peel. I have seen a four-ply belt fail at the splice inside six months because the field crew got the step lengths wrong; the belt body was still sound. Ply count and splice discipline are two halves of the same decision, and neither should be finalised without the other.

04Steel Cord Carcass: Construction Logic and the Trade-offs

Once a duty climbs past the practical ceiling of fabric, the skeleton changes from woven plies to a bed of parallel steel cables running lengthwise. A cord carcass is a single, thick layer of brass-plated high-carbon cables embedded in rubber, spaced evenly across the width and held in place by a thin transverse layer. Every cable shares the load, so a belt rated ST2000 carries 2000 N/mm through those cables rather than through fabric threads. The construction logic is simple; the consequences are not.

Cable diameter and pitch are the two numbers that define a cord carcass. Diameter drives the strength of each cable, and pitch — the centre distance between adjacent cables — decides how many fit across a given width and therefore the total belt rating. Tighten the pitch to raise the rating and the rubber between cables gets thinner, which weakens the bond that holds the cables in place and reduces resistance to impact and corrosion at the exposed ends. This is why two cords of the same rating can differ in fatigue and corrosion behaviour purely through pitch and the compound between them.

What cord buys you is enormous strength in a modest thickness, near-zero working elongation, and superb troughing for the load it carries, because flexibility does not have to be sacrificed for tensile capacity the way it does in a heavily plied fabric belt. A single cord layer can out-pull six plies of EP at a fraction of the thickness, which is why mainline overland conveyors handling thousands of tonnes per hour run on cord. On a 1,200 t/h coal line I helped commission, we logged cord elongation at around 0.25% at full load — the take-up barely moved between empty and laden, which is exactly the stability a long incline needs.

What cord costs you is forgiveness. The cables are stiff, so minimum pulley diameters jump and drives must be sized for a belt that will not stretch to absorb a shock. A cord belt that runs off-centre punishes its edges hard. A splice is a full vulcanised joint with cables laid in a specific sequence, and getting it wrong is far more expensive than a fabric finger joint. And a cord belt is not automatically the right answer for short, slow, lumpy duties where a fabric belt and a big take-up would do the same job for less money. The decision runs on centre distance, tonnage, incline, and pulley geometry together. When the answers all point to cord, specifiers usually work through a step-by-step sizing method and then a pre-order verification checklist before committing, because the cost of a wrong call on a mainline belt is measured in shutdown days.

There is a middle path worth knowing about, and it is why the cord-versus-fabric question is not always binary. Some duties sit right at the top of the fabric range and the bottom of the cord range, and the answer comes down to how much the plant values flexibility, how long the splice can be offline, and how much the crew can be trusted with a cable joint. A cost-comparison framework helps here, because the purchase price is rarely the deciding number once splice duration and pulley upgrades are counted.

Cord belts also open the door to a much wider family of cover and reinforcement options, and buyers comparing total systems usually end up weighing the belt against the rest of the line. If the plant needs matched drive components, a conveyor belt supplier that carries rollers, pulleys and brackets under one roof saves the headache of reconciling three datasheets, and working through a conveyor belt distributor with local stock can shorten lead times on a mainline change-out. Even the ancillary drives matter: a transmission belt manufacturer perspective on power transmission is useful when a plant is reviewing the whole drive train at once, and a V-belt manufacturer can often supply the drive belts that keep the auxiliary equipment running alongside the main conveyor.

05How to Read a Carcass Specification Table

A carcass specification is a short document that hides a lot. Read it in the wrong order and you will compare belts on the wrong line. The first field is the nominal tensile strength, written as a family prefix and a number, so EP630 and ST1000 both tell you the load the skeleton is rated to carry per millimetre of width. That single figure, though, is meaningless without the elongation and pulley data that must accompany it, because a rating you cannot physically run on your geometry is a rating you cannot use.

Next come the construction details. For fabric, that is the ply count and each ply's own rating; for cord, it is cable diameter, pitch, and the number of cables across the width. Then the covers, top and bottom, each with its own thickness and grade. A belt is a stack of decisions, and the table is simply that stack written down. Two rows that look similar at the top can diverge completely by the bottom, which is where most of the real risk lives.

Cover thickness deserves its own reading. The stated figure is the nominal thickness of the rubber above the carcass on the carrying side, and it is the belt's wear allowance as much as anything else. A thicker top cover buys wear life but adds weight and raises the minimum pulley diameter, because total belt thickness is what the pulley sees. The bottom cover, often thinner, protects the carcass from the idlers on the return run. Judging a belt by its top cover alone misses half the structure. Standardised tables under quality assurance are where these values get verified against DIN 22102 or GB/T 9770 rather than taken on trust from a data sheet.

One habit separates buyers who get the right belt from those who get repeats. Read the table from the bottom up. Start with the minimum pulley diameter on the actual machine, then the working elongation the take-up can absorb, and only then look at the strength rating. If the rating is high but the pulley is too small, the belt is wrong no matter how impressive the number looks. I once reviewed a quote for a 4-ply EP1000 belt destined for a 400 mm drive pulley; the supplier had never asked the pulley size, and the belt would have delaminated at the outer ply inside a year.

06Carcass-to-Splice Coupling

The carcass and the splice are one system, and the carcass decides what the splice is allowed to be. A splice exists to rejoin the load path, and its job is to carry the full belt tension across a joint that is geometrically weaker than the continuous body. The stronger and stiffer the carcass, the more precisely the splice has to be engineered, because a stiff member concentrates stress at the joint while a compliant one spreads it out.

On a fabric belt, the standard method is a stepped finger joint: the plies are cut back, stepped, and rebonded so that each ply meets its counterpart across an overlap. The required step length scales with the ply rating and the number of plies — a higher-rated or higher-ply belt needs a longer step to spread the load — and the whole joint is often longer than people expect. A four-ply EP belt can need a splice several hundred millimetres long per step once the standard method is applied. Cut the steps short to save time and the joint becomes the weakest point in a belt that is otherwise fine.

Cord belts take the same idea into a different geometry. The cables are laid out in a defined overlap sequence, the ends bonded, and the joint vulcanised under controlled heat and pressure so the rubber between cables cures around the laid-in ends. The splice length is a multiple of cable pitch, and the sequence of cables at the joint matters because a random layup leaves gaps that carry no load. This is skilled, time-consuming work, and the reason cord belt downtime is measured in shifts rather than hours.

What this means for selection is that you must decide the splice before you fix the carcass, not after. A plant that cannot spare a shift for a vulcanised cord joint should not buy a cord belt, however attractive the tonnage maths. A plant that can only reach the joint with a lightweight crew should favour fewer plies and a simpler finger joint. The splice is not a detail that follows the carcass; it is a constraint that shapes it. Regular inspection of the joint sits at the heart of any lifespan and failure-prevention routine, because the splice is where most premature belt failures announce themselves first.

It is also worth linking the carcass decision to the rest of the line, because a splice problem is often a component problem in disguise. If the belt is running off-centre, the joint takes uneven load and fails early, so the health of the carrying rollers and the drive and bend pulleys has a direct bearing on splice life, and a belt trained properly through the troughing idler sets will always outlast one that fights its own alignment.

07Choosing a Carcass by Operating Conditions

Selection is a matching exercise between the duty and the skeleton, and the duty speaks through a handful of signals. Tonnage and incline set the tension, which sets the rating. Centre distance decides whether elongation matters, because a long line with a short take-up cannot tolerate a stretchy belt. Feed size and drop height decide how much impact the carcass must absorb. Ambient moisture, temperature and chemicals decide which fabric threads survive. Read those signals honestly and the carcass family usually picks itself.

Start with tension. If a quick band calculation puts peak running tension below roughly 500 to 600 N/mm, fabric is in play; above that, cord becomes attractive and often necessary. Incline matters because belt tension rises steeply with lift, so an incline at moderate tonnage can still push you into cord territory. Centre distance then decides the fabric choice within its range: short and shock-loaded favours a forgiving build, long and steady favours the stable, low-stretch one.

Environment is the quiet decider. A damp tunnel rules out cotton and pushes you to EP. A hot clinker line needs a carcass whose bonding system tolerates heat, which usually means a heat-grade cover paired with a compatible construction rather than a standard build with a hotter rubber. Chemical exposure at a fertiliser plant sends you to a specific compound family for the same reason. The belt never sees the carcass directly, but the carcass still has to survive whatever leaks through the cover.

Put the environmental conditions and the geometry side by side and one of them usually dominates. When it does not, the tie-breaker is almost always the pulley and take-up geometry, because that is the constraint you cannot change cheaply after the belt arrives. Feed size, incline and chemistry can often be accommodated with a cover or a chute change. A pulley that is too small for a given carcass cannot be fixed without welding, and that is a far more expensive correction.

Then bring the whole plant picture in. Duty-specific conveyors rarely exist in isolation, and the carcass you choose for the main line should be aware of what feeds it. Belts handling sticky or steep material may need raised structures, which changes how much slack the carcass must tolerate, so it helps to compare against cleated incline belts and corrugated sidewall belts when the geometry is unusual. In harsh abrasive or high-impact conditions, pairing the carcass with the right surface means looking at abrasion-resistant covers and impact and cut-resistant builds as a package rather than as separate lines on a quote.

08Common Carcass Mistakes and Field Consequences

Most carcass failures trace back to a handful of recurring mistakes, and almost none of them are exotic. They happen because the carcass is invisible once the belt is installed, so shortcuts taken at specification time do not show up until months or years later, usually at the worst possible moment. The list below is drawn from real disputes I have watched unfold on site.

The first mistake is specifying by ply count or by "how many millimetres of cord" instead of by the total rating at the actual pulley geometry. A buyer insists on four plies because a competitor offered three, without checking the pulley. The belt arrives, runs briefly, and delaminates at the outer ply over a too-small pulley. Nobody connects the four plies to the failure, and the next order repeats the mistake with a different brand. The fix is uncomfortable because it means admitting the rating label was never the right headline.

The second is ignoring working elongation. A long conveyor with a short-travel take-up gets a stretchy NN belt because it was cheaper, and after a few months the counterweight has run out of travel and the drive slips. The belt is not defective; it was never matched to the take-up. Elongation is the most overlooked line on the datasheet and, on long lines, the one that causes the most grief.

The third is under-reading the environment. A cotton or lightly protected carcass goes into a damp tunnel and rots from the inside, or a standard bonding system meets sustained heat and the cover delaminates from the carcass even though both look fine separately. The failure starts at the bond line, which is exactly the interface a strength-only specification ignores. When belts face specific chemical or thermal threats, matching the carcass to a heat-resistant build or a oil-resistant build is not optional detail; it is the difference between a bond that holds and one that peels.

The fourth mistake is treating the splice as an afterthought. A crew cuts short steps to finish a shift, or lays cord ends in the wrong sequence, and a joint that should have lasted years fails in a season. This one is especially painful because the belt body is often still good; the money lost is downtime, not belt. The fifth is mixing components from unrelated standards, so a belt rated to one system is matched with rollers and pulleys sized to another, and the whole line runs at the tolerance of its weakest link rather than its design intent.

Buying purely on unit price sits behind most of these. A cheaper carcass with a lower rating, or a construction that ignores the pulley geometry, saves money on the purchase order and spends it three times over on early replacement and lost production. That is why serious buyers compare total cost over the belt's life rather than the invoice, and why duty-matched component ranges matter across sectors — whether a plant is in mining and quarrying, cement production, or port bulk handling, the carcass logic is the same and the consequences are equally expensive. When a plant runs several belts across different duties, sourcing them together from one wholesale conveyor belts range usually means the duty-specific carcass logic gets applied consistently instead of drifting order by order.

A field-experience checklist before you sign

Before any carcass order is confirmed, I run the same three questions I have used on dozens of projects. What is the smallest pulley the belt will bend over, and does the chosen construction meet the standard minimum for that pulley? What working elongation will the take-up actually see, and is there enough travel left for a cold or overloaded day? And who is going to make the splice, with what method, in how many hours? If a supplier cannot answer those three without going back to the workshop, the specification is not finished, no matter how complete the datasheet looks.

Talk to our engineers about carcass and splice matching

09Frequently Asked Questions

What exactly is the carcass of a belt?

The carcass is the structural skeleton beneath the rubber — either several woven fabric plies or a single layer of parallel steel cables. It carries the tension the drive applies and every load the belt transports, and it sets the rating, the working stretch, the troughing behaviour and the fatigue life. The cover protects the surface; the carcass holds everything together and keeps the belt in one piece under load.

Is more plies always better for a fabric belt?

No, and this is one of the most common assumptions to get wrong. More plies raise stiffness and, with it, the minimum pulley diameter the belt needs, and they make troughing harder. A thinner build often trains better and tolerates tighter geometry. The ply count should follow the total rating and the smallest pulley on the line, not the other way round. Pick the rating and the pulley first and the ply count usually becomes obvious.

How do I choose between a fabric belt and a steel cord belt?

Compare peak tension, centre distance, incline and pulley geometry together. Fabric suits moderate tension, forgiving layouts and shorter or shock-loaded runs. Cord suits high tension, long centres and steep lifts, because it packs far more strength into a thin, low-stretch body. If a duty sits near the boundary, weigh how much flexibility the layout needs, how long a splice can be offline, and how skilled the splice crew is before deciding.

What is the difference between EP and NN fabric belts?

EP uses polyester in the load-carrying warp with nylon in the weft, so it holds its length and stretches very little under load. NN is all nylon, which stretches more, recovers well and absorbs impact but tends to grow over its life. EP suits long lines and limited take-up travel; NN suits transfer points and lumpy feed where a little give helps. The same nominal rating behaves quite differently between the two because of the thread alone.

Why does the specification list a minimum pulley diameter?

Because a belt bends every time it passes a pulley, and the tighter the bend, the more the outermost carcass members flex. Below a certain diameter those members fatigue far faster, and the belt cracks or delaminates from the inside even though nothing looks worn from outside. Standards such as DIN 22102 and ISO 15236 publish the minimum diameters for each construction. Ignore them and you shorten belt life dramatically.

Can these belts be spliced on site, and how long does it take?

Fabric belts are usually joined with a stepped finger joint, which can be vulcanised hot or cold and typically takes hours depending on ply count and belt width. Steel cord belts need a vulcanised joint with the cables laid in a specific sequence, which takes longer and demands more skill. Mechanical fasteners exist but cannot reach full carcass strength, so they are best reserved for temporary repairs. Always settle the splice method before you confirm the carcass.