Conveyor Components Manufacturers: Tensile Strength, Thickness and Carry Capacity Explained

We get asked to quote on component packages every week, and the same thing happens almost every time. A buyer forwards a spec sheet from another supplier and asks us to beat the price. The sheet looks complete. It has a tensile strength, a thickness, a capacity. Nothing is missing. What is missing is the duty data behind those three numbers, and without it the sheet cannot be compared to anything — not to ours, not to a competitor's, and not really to the machine the belt is going onto.

What follows is the checklist we use internally when a customer's engineer asks us to review somebody else's specification. It goes through the three numbers that carry most of the weight in any component quotation — tensile strength, thickness and carry capacity — and shows where each one genuinely constrains the design, how the tolerances should be read, how test reports are checked, and which answers from a supplier should make you stop and ask again. Before the end you will have six questions to ask, four documents to request, and a short list of red flags. None of it is theoretical. Every point traces back to a claim we had to verify, a roll we had to reject, or a failure we had to explain to a customer who had already paid for it.

Send Us the Spec Sheet — We Will Mark It Up Line by Line

01How a Component Spec Sheet Actually Governs a Conveyor Design

A specification is a sales document first and an engineering document second, and the two roles pull in opposite directions. The sales role wants every line filled with a big, clean number. The engineering role wants every line qualified with a test standard, a condition, and a tolerance. When you read a sheet, the lines that carry three attachments are the lines you can design with. The lines that carry a bare number are marketing until proven otherwise.

The three headline figures do three completely different jobs, and engineers who treat them as one performance score get burned. Tensile strength per unit width tells you whether the carcass survives the tension your drive will demand at the highest point in the profile, at the weakest cross-section, which is almost never the belt body — it is the splice. Thickness tells you how long the belt has before the carcass is exposed, and how it behaves as it goes around a pulley. Carry capacity is not a property of the belt at all. It is an output of the whole system, and the belt only appears in it as a width, a troughing geometry and a speed.

So the real question is never "which number is higher". It is "which number is closest to being exceeded on this conveyor". In a long overland haul the answer is tension, and the tightest point is a splice on the drive incline. In a short transfer conveyor under a coarse-ore chute the answer is impact and wear on the top cover. In a hot-clinker line the answer is adhesion retention, which quietly falls before tensile strength does. A competent conveyor components package is quoted against a duty description, not against a catalogue page.

Same Rated Strength, Very Different Splice

Here is a case that makes the point better than any table. On a 1,200 t/h coal line feeding a power station, the operating company replaced a five-ply EP160 belt with a four-ply EP200 belt to cut weight and reduce pulley loading. Both constructions are rated at 800 N/mm of belt width, so on the strength line of the specification nothing had changed. The splice did change. A four-ply belt has fewer steps and each step carries more load per unit of splice length, so the dynamic splice efficiency dropped from the high eighties into the mid seventies of belt strength. The new splices started pulling apart at around five months, always on the drive incline, always at the same step. The rated tensile number was never wrong. It had simply never been the governing number at the splice.

That is why we stopped accepting a single strength figure as sufficient. Every quotation we issue states the finished belt's minimum single-value tensile strength, the ply or cord construction behind it, and the splice efficiency we are prepared to guarantee on it. A conveyor belt manufacturer that gives you the strength number alone has given you one third of an answer. Ten lines of production — eight for fabric carcass, two for steel cord — is what sits behind our numbers, and every one of those lines produces belts whose construction can be traced from the roll label backward to the fabric batch.

02Tensile Strength: What the Rating Really Describes

Fabric and steel-cord carcass stock. The rated strength printed on the roll label describes the body of the belt — never the splice.

A tensile rating on a fabric belt is quoted in newtons per millimetre of belt width, and it is built by multiplying a per-ply figure by the number of plies. EP100, EP125, EP160, EP200, EP250, EP315 and EP400 are the common per-ply steps, where EP means the warp is polyester and the weft is nylon. A five-ply EP160 belt therefore carries a nominal 800 N/mm. Change the ply count and you change the rating; change the fabric from EP to a nylon-nylon carcass and you change how the same rating behaves in service, because the two constructions stretch differently under the same load. The same reading applies to any rubber conveyor belt, whatever the application sitting behind it.

Three things sit behind that number and they matter more than the number itself. The first is the yarn specification — the linear density of the warp, usually written in dtex, and the dip treatment that bonds the fabric to the rubber. Two suppliers can both sell "EP160" while using different yarns and different dips, and their belts will not last the same time. That gap is where most rubber conveyor belt manufacturers differ from one another without ever saying so. The second is the skim compound between plies, which determines adhesion, and adhesion is what actually transfers the load from one ply to the next around a splice. The third is the ply count itself in relation to pulley diameters, because a thick, stiff carcass on a small pulley bends more harshly and fatigues faster than a thinner one at the same rating.

For steel-cord belts the rating works differently. It is quoted directly in newtons per millimetre — ST630 through ST5400 — and the number describes the cord bundle rather than plies. Cord pitch, cord diameter, and the rubber that surrounds the cords carry the same weight as the headline figure, because cord-to-rubber adhesion is the failure mode that ends these belts. If you are comparing a steel cord conveyor belt quote, ask for cord diameter, pitch, and adhesion test results, not just the ST class.

Where the Rating Binds, and Where It Sleeps

Long-distance conveyance is the classic case where tensile strength governs everything. On a multi-kilometre overland conveyor the running tension is large enough that the design is essentially a conversation about strength, splice efficiency and take-up travel. A higher ST or EP class buys you either more capacity on the same structure or a longer single flight, and the safety factor you choose writes the whole quotation. Most mining specifications we quote against work to a safety factor somewhere between 8 and 10 on fabric belts with a screw take-up, and roughly 5.5 to 7 on steel-cord belts, depending on whether the tension is measured dynamically or statically.

Short and severe is the opposite case. On a 40-metre conveyor under a primary crusher, the total running tension may be a fifth of what the same belt would see on a 4 km haul, yet the belt is destroyed in nine months. Nothing about the tensile rating decides that outcome. What decides it is top cover thickness, compound abrasion resistance, and the spacing of impact idlers under the chute. Buyers who shop these two applications with the same checklist usually overpay on the long belt and under-specify the short one.

High temperature adds a third pattern. Heat attacks the rubber before it attacks the carcass, so a belt can pass its tensile test in the factory and still delaminate in a cement kiln discharge line within a season. The number that predicts that failure is aged adhesion retention — measured after a defined ageing cycle — not residual strength. When a specification for a hot line quotes only tensile strength, the document is describing the wrong failure mode. This is also why an abrasion or heat grade letter on a cover is not decoration; a abrasion resistant conveyor belt for a wet ore line and a heat-resistant belt for a clinker line are different products even when the carcass and the tensile rating are identical.

One more reading trap. Strength is quoted for the finished belt, and the finished belt is tested with coupons cut from a cured slab. The splice is a field or factory operation and its efficiency is expressed as a percentage of belt strength — values in the region of 60 to 90% for fabric belts with stepped splices, lower for a mechanical fastener, and a narrower band for steel cord depending on the splice design. If a supplier quotes a splice efficiency at or near 100%, that is not a strong claim, it is an unverified one. We ask for the dynamic splice test result and the splice drawing before we will sign a strength guarantee.

03Reading Tolerances: Nominal, Minimum and Average

Three different quantities get called "the value", and a buyer who does not name which one is not comparing specifications at all. Decide the population and the statistic in writing for every dimension that matters, before the order is signed. The table below shows how the same property is commonly reported, and what to write instead.

Look at how narrow the gap is between a conforming belt and a rejected one. A belt whose top cover averages 6.1 mm with a low point of 4.9 mm passes a "6 mm nominal" order and fails a "no point below 5.3 mm" order. Both belts are legal to sell; only one of them will survive a 1,400 mm chute drop for the design life. We have had exactly this argument twice, once with a port operator and once with a cement plant, and in both cases the specification said "6 mm" and nothing else.

How to Measure Without Being a Laboratory

Then decide what happens when a value misses. The realistic clauses are a re-test on a second coupon from the same roll with the same method, arbitration at an accredited third-party laboratory if the two results disagree, and an agreed remedy — credit, replacement or rejection at the supplier's cost — if the arbitration confirms the miss. Without that ladder, a miss always ends in emails. With it, a miss ends in a delivery note. We write the ladder into every serious order, and a conveyor belt supplier that resists it is telling you something about their process capability before you have spent anything. It is also the fastest way to tell conveyor components manufacturers apart before a purchase order exists.

04How to Verify a Test Report Without a Laboratory

Start with the standard. Tensile strength of a full-thickness belt is not one test, it is several, and the specimen geometry changes the result. A dumbbell specimen and a straight strip of the same belt do not fail at the same load, because edge effects and stress concentration differ. Abrasion resistance is worse. A rotating-drum abrasion result depends heavily on which method and which abrasive sheet were used, so a figure of 90 mm³ from one laboratory is not comparable with 90 mm³ from another unless both name the same method. When we compare a competing report to our own, the first thing we look at is whether both sides name the same standard edition. If they do not, we stop comparing numbers and start asking for a re-test.

The traceability line is where most audits end, and it is the line worth the most. A report that cannot be tied to a roll number is a statement about a belt, not about your belt. We once received a test report whose date of issue was nine days earlier than the cure date printed on the roll label. That is not a rounding error; it means the report belonged to another production run, and the entire submission had to be re-tested. The customer had already accepted the price, which made the conversation harder than it needed to be. Since then we have asked every incoming roll for a cure record alongside the test report, and we publish the same evidence for our own shipments through our quality assurance process.

Reports Travel Through the Chain, Not Just From the Mill

Where you buy matters as much as where the belt was made, because documents change hands. If you buy through a conveyor belt distributor who sources from a trading house, the report you receive may be a copy of a copy with pages missing, and nobody in the chain can answer a method question. Ask early who holds the original report and who will answer the re-test request. If the answer is "the factory in China, but we will forward it", budget two weeks for every technical exchange and build that into your schedule. A conveyor components supplier who cannot answer a method question is a pass-through on documentation, not a partner in it.

The same discipline applies to the drive side of the plant. A transmission belt manufacturer works to the same logic — cord adhesion, length tolerances and match-set claims all need evidence, and a V-belt set that is not matched behaves like a set of belts that are all slightly wrong. Buying components as a package from one accountable source means one test report structure, one traceability system and one conversation when something misses. It is also the reason we insist on visiting the lines that produce the belts we supply; a conveyor belt factory that will not let your engineer watch a tensile test in progress is not really offering verification, only documentation.

05Thickness and Cover Rubber: Two Dimensions, Two Different Jobs

Cover gauging on a finished roll. Top and bottom covers are recorded separately, at ten points along the length.

Total belt thickness is the number on the front of every catalogue, and it is the least informative figure in the document. What matters is the split. The top cover absorbs the material, and the bottom cover absorbs the pulleys and the return idlers. They wear at different rates for different reasons, they are consumed by different mechanisms, and only one of them can be traded away when a quotation needs to come down. A belt quoted as "10 mm" without a split could be 6 mm over 2 mm carcass over 2 mm, or 4 mm over 3 mm carcass over 3 mm. Those two belts have different wear lives and different bending behaviour, and the catalogue makes them look identical.

The top cover is a sacrificial layer, and its job is to be consumed slowly enough that the carcass never sees daylight. Wear rate is best expressed in millimetres per thousand operating hours for a given material, because that turns a specification into a maintenance plan. On a crushed copper ore line we surveyed, a 1,200 mm belt running 620 hours a month was losing roughly 1.4 mm of top cover per 1,000 hours under the loading zone and about 0.9 mm along the carrying run. With an 8 mm top cover and a practical wear limit of 1 mm of rubber remaining above the carcass, the usable allowance was 7 mm. That gives a little over 5,000 hours at the worst point, or about eight months, and it told the maintenance planner exactly when to book the replacement instead of discovering it during a shift.

The bottom cover is smaller and more easily under-specified. It takes the drive pulley traction, it runs over every return idler, and on a high-tension drive it is the surface that decides whether you can transmit the power at all. A belt with a generous top cover and a 1.5 mm bottom cover looks strong, and it will run until the return-side idlers polish through it and expose the fabric on the pulley side. On retrofits where the tension rating has been increased to lift more tonnage, the bottom cover is what we look at first, along with the pulley lagging grade. If you are buying an industrial conveyor belt by thickness alone, you are optimising the number nobody measures after installation.

Cover compounds are graded, and the grade is a better predictor of wear life than the millimetre figure. The common European scheme divides covers into four wear classes, running from a highly abrasion-resistant grade with an abrasion loss limit near 90 mm³ under ISO 4649 Method A, through two intermediate classes, to a light-duty grade in the region of 250 mm³. An 8 mm top cover in the light class can be consumed twice as fast as a 6 mm cover in the heavy class on the same conveyor. That is the trade you should be making, and it is invisible if the quotation only says "abrasion-resistant cover". Ask for the measured loss figure in mm³ and the standard it was obtained under.

Cover thickness also has a lower bound driven by the material, not by the tonnage. Sharp, angular rock with lumps above 300 mm needs depth to spread the impact energy, and a thin cover transfers that energy straight into the carcass, where it bruises the plies and starts delamination. That is the case for a thicker, softer top cover under a primary crusher, and for closer impact-idler spacing underneath it. A hot material is a third case again, where cover hardness changes with temperature and the cover has to tolerate thermal cycling without cracking. Where profiles and cleats are involved, the same carcass thoughts apply with an extra skin of complexity — a chevron conveyor belt or a cleated sidewall belt carries material on an incline, so its cover has to resist both abrasion and the peeling load at the foot of every cleat.

Finally, edge construction. Covers wrap the edges, and edge wear is what kills belts on poorly tracked conveyors. A 2 mm edge wrap on a wide belt in a transfer tower is not enough; the edges will fray, water will enter the carcass, and the fabric will delaminate from the outside in. Where you buy matters here too — a wholesale conveyor belts channel that ships from stock may not offer an edge-wrap option at all, while a mill that runs its own vulcanising presses will simply quote it. And on the drive side, remember that the belt is only half the transmission. A V-belt manufacturer dealing with the same wear problems will tell you the same thing about sheave alignment that we say about pulley alignment: geometry failures are diagnosed on the rubber, but they are not caused by the rubber.

06Carry Capacity: The Assumptions Behind the Tonnage

Carry capacity is the number most often quoted and least often explained, because it belongs to the conveyor rather than to the belt. A figure in tonnes per hour is the output of a cross-section, a speed and an assumption about how the material sits in the trough. Change any of those three and the figure moves, even though the belt has not changed at all. When a supplier prints a capacity against a component, the honest version of that claim always names the same four inputs before the number means anything. Material bulk density, surcharge angle of the load, troughing angle of the idler set and belt speed all have to be stated. If your quotation is missing those, the number it shows is a wish.

Troughing geometry does most of the work. A three-roll idler set at 35° carries substantially more material than the same width at 20°, because the load sits deeper in the trough and the effective cross-section grows by roughly a third. Move to 45° and the section grows again, which is why deep-trough conveying is popular on high-tonnage lines with free-flowing material. The catch is that deep troughs need the belt to be flexible enough to seat properly, and they put more strain on the edges. A belt that will not seat in a 45° trough has less capacity than its catalogue says, not more.

The material decides the rest. Surcharge angle falls with moisture and with angularity, and a sticky, wet ore can sit flat where a dry graded coal heaps up. Take a practical example from a port installation we were asked to review. The design was issued for 3,000 t/h on a 1,400 mm belt at 3.5 m/s, using a surcharge angle taken from a laboratory sample of dry coal. The coal that actually arrived at the berth ran about 9% moisture, and the measured surcharge angle was several degrees lower. Committed throughput settled closer to 2,250 t/h, and no component was ever at fault. The mistake lived entirely in a column of an assumption table that nobody checked against operating reality.

Sag Is the Limit Everybody Forgets

Even with the geometry right, the published capacity only holds while the belt is supported properly between idlers. Sag between idler sets is the practical constraint, and the usual design rule caps it around 2% of the idler spacing for normal service. Sag grows as the square of the spacing and falls with tension, so it is the interaction between belt tension, idler spacing and the weight of the load. Push a design to a wider idler spacing to save structure, and the same belt carries less material before the load starts behaving badly, edge contact degrades, and material spills at the loading point. This is why an idler and roller set change can cut the effective capacity of an unchanged belt.

07Six Questions and Four Documents: Screening a Supplier

Everything above compresses into a screening routine you can run in two emails and one call. The point is not to trap anyone. It is to find out, before money changes hands, whether the supplier understands the three numbers as design constraints or as lines in a brochure. Ask the six questions below in writing, request the four documents, and read the answers as a set.

1. Which figure on the sheet governs my conveyor, and what duty data did you use to decide? A serious supplier will ask about length, lift, tonnage, material and temperature before answering. One who answers in the same breath as the question has not modelled anything.
2. What is the minimum single-value tensile strength of the finished belt, and what splice efficiency do you guarantee on it? You want two numbers and a test standard, not a class name.
3. What is the carcass construction behind the rating? Fabric type, yarn linear density, ply count and per-ply strength for a fabric belt; cord diameter, cord pitch and cord count for steel cord. If this is treated as confidential and cannot even be described in outline, you cannot verify anything later.
4. What are the top and bottom cover thicknesses, the cover grade, and the measured abrasion loss in mm³? Four values, and they must be stated separately. Merged figures are the most common way a quotation hides a downgrade.
5. What do your aged test results show, and under which ageing cycle? Aged adhesion retention is the figure that predicts heat-related failure, and it is the figure most often missing entirely.
6. What are the capacity assumptions behind the tonnage in the offer? Bulk density, surcharge angle, troughing angle, belt speed and the sag allowance. Five inputs or the number is decorative.

Report against roll. Numbers are only useful when the document and the belt in front of you agree.

Then ask for paperwork. Four documents cover the whole audit, and each one answers a different question. Together they let you confirm what was sold, what was made, what was measured and who is accountable.

Red Flags, and What to Ask Instead

Where a supplier answers all six questions with numbers, names a standard for each, and volunteers the weak points of their own product, the commercial discussion gets shorter and cheaper. In our experience the offers that come back fastest with the loudest claims are the ones that cost the most to own. The same screening logic works across sectors — the duty data differs between mining and quarrying and a cement plant, but the questions do not. If you want the application-side counterpart to this document — what those duty conditions actually demand in each industry — we covered it in conveyor components for mining, which works through requirements and selection rather than document review.

08From Marked-Up Spec to Purchase Decision

What you do with the answers matters as much as getting them. Build one comparison sheet, one row per property, one column per supplier, and put the required statistic at the top of each row — minimum single value for strength, averaging rule plus single-point floor for thickness, measured value in mm³ for abrasion, five stated inputs for capacity. A row that cannot be filled in is a row that cannot be compared, and that is the whole purpose of the exercise. We use the same sheet when a customer asks us to bid against an incumbent, and roughly a third of the time it turns out that the two offers were never comparable in the first place. It is also the sheet we use when a buyer asks several conveyor components manufacturers to quote the same line.

Then decide who owns the splice, because on long and heavily loaded belts the splice is where the money goes. A factory-spliced and rolled belt arrives with controlled splice geometry and a documented test position. A belt shipped as plain rolls and spliced on site depends on your crew and their press. A conveyor component company that will not state which of the two it is selling is passing you a risk. Both are legitimate. What is not legitimate is a quotation that prices a strength guarantee without saying who makes the joint that carries it.

Finally, attach the marked-up specification to the purchase order as an annex, exactly as reviewed. Everything you resolved in the six questions and the four documents becomes contractual at that moment, and the argument about whether "6 mm" meant an average disappears. If your current supply route cannot survive that annex, this is the point at which to look at another one — and the moment to talk to us is before the order, not after the failure. The engineering teams behind reputable conveyor components manufacturers will happily mark up your specification and tell you which lines are thin, because it is cheaper for everyone than a claim.

Ask for Our Test Report Format and Tolerance Table

09Frequently Asked Questions

These are the questions that come up most often once buyers start reading specification sheets line by line instead of comparing headline numbers. The short answers are below; the reasoning behind each one is in the sections above. Where a question and our answer disagree with the sheet in front of you, the sheet is the thing to ask about. Whether you are screening ten conveyor components manufacturers or renewing a single order, the questions themselves do not change.

What does "minimum single value" mean on a belt specification?

It means the lowest reading that any accepted coupon from the roll may return. The nominal figure is the design intent, the average describes the typical result, and the minimum single value is the acceptance floor.

How do I compare two quotes when the numbers look identical?

Compare the method sections, not the results. Two offers of 800 N/mm can differ in ply count, yarn specification, dip treatment, skim compound and splice efficiency, and those differences decide service life. Ask for specimen geometry, orientation, specimen count and the lowest single value from each side. Most of the belt makers we compete against will supply them. Among conveyor components manufacturers, the ones who will not are worth avoiding.

Should I control total thickness or cover thickness in the order?

Control both, but write the cover figures first because they decide wear life. A practical wording uses an average of ten gauge points with a single-point minimum for each cover, measured on the finished belt rather than the cured slab, plus the carcass thickness that follows from the construction. If top and bottom covers are quoted as one number, that number is not a specification.

Can I verify a belt without a laboratory of my own?

Partly. You can gauge covers at several points, measure width and edge condition, weigh a sample, and read the traceability data against the roll label. What you can do without a lab is insist that the supplier's report is complete enough to be reproduced elsewhere, and that is a reading exercise rather than a testing one. Name an accredited third-party laboratory in the contract as the arbitrator before any dispute exists, so the escalation route is settled while everyone is still friendly.

Why do suppliers quote different tonnages for the same belt width and speed?

Because tonnage depends on assumptions the belt does not contain. Troughing angle, surcharge angle, bulk density, sag allowance and loading uniformity all move the cross-section, and two engineers can legitimately reach different figures from the same width. The difference is not dishonesty in most cases; it is an unstated input.

Which documents should be requested before a purchase order is issued?

Four, as a set. A test report tied to the delivered roll, a construction and material datasheet, a dimensional inspection record covering width, length and covers, and a process package with the management system certificate, cure records and the warranty exclusions in full. Suppliers of conveyor belting and components who answer all four requests promptly are usually the same ones whose belts pass the incoming inspection. That is the practical test of conveyor components manufacturers in any market.