Conveyor Belt Steel: Steel Cord, Steel Mesh and Metal Belts Explained

A search for conveyor belt steel drops you into one of the widest mismatches in industrial supply. Three separate product families hide behind those three words, and they share almost nothing. Different carcass, different drive, different splice, and a price per metre that can swing by a factor of six. Buyers who treat the phrase as a single product end up quoting a stainless woven mesh against an 1,800 m coal overland line, or paying cord-belt money for a 12 m cooling tunnel that only ever needed a chain-edge mesh.

The confusion is not the buyer's fault. Search results are dominated by catalogue pages, and every one of those pages belongs to a maker of exactly one family. A catalogue page will never tell you when its own product is the wrong answer. That is the document missing from the first page today, so this page is it. We split the term into the three families that actually reach a purchase order, then work through the operating window, the buying tests and the cost of getting it wrong for each one.

Short version, for anyone reading under time pressure. Long centre distances, high tonnage and a tight take-up tower belong to the cord family. Hot, open, hygienic or curved paths at modest load belong to metal. Impact-heavy transfer points sitting on an otherwise ordinary line belong to the reinforced fabric family. Everything below is the reasoning, the numbers and the field consequences behind those three sentences.

Tell us the centre distance, tonnage and material, and we will tell you which family fits

01Steel Is Not One Material, It Is Three Complete Systems

Start with what is doing the work inside the belt. In the cord family, the load path is a bed of parallel steel ropes embedded in rubber, and the rubber only protects them. In the metal family, the steel is the belt itself, formed into mesh or a continuous sheet. In the reinforced fabric family, the load path is woven synthetic fabric — and the steel arrives as a breaker layer whose job is to stop a tear, not to carry tension. Three answers to the same question, which is why the families cannot be swapped on a drawing board.

The second split is how the belt is driven. A cord belt or a fabric belt runs on a smooth pulley and is dragged by friction. A woven mesh belt is usually driven by sprockets engaging the edge chain or the mesh itself, so it cannot grip on a bare drum at all. A solid steel belt sits somewhere between the two, and its endless weld is the single part of the belt you cannot inspect from outside.

Put the families side by side and the boundaries become obvious. The table below is the one we use when a buyer sends us a line description and we have to say, in one reply, whether we make the right product for it or not.

Read the elongation column again, because it drives almost every complaint we hear. A belt that stretches 1.2 percent over a 900 m centre distance asks the take-up to absorb nearly 11 m of extra belt, and most towers were built for 3 m. No amount of good splicing fixes a take-up that is one third the size it needs to be.

Capacity behaves the same way. A cord belt at 1,600 mm width and 6 m/s will move well over 4,000 t/h of coal on a single flight, while a woven mesh of the same width is doing well to carry 40 kg per square metre of belt surface and would be scrap within weeks under that tonnage. Different physics, different budget line, different maintenance team.

02Family One, Steel Cord Rubber Belts, and the Window They Own

This is the family we build, so we will be blunt about where it does not belong. A cord belt is a rubber slab with a carefully spaced bed of galvanised steel ropes at its neutral axis. Strength is quoted in N/mm of width, and the grade name is a direct statement of that number. ST1000 means 1,000 N/mm, so a 1,200 mm wide ST1000 belt has a rated breaking strength of 1,200 kN. Our presses run grades from ST630 up through ST2500 in normal production, and higher grades on request.

Four numbers decide whether a line belongs to this family. Centre distance in metres, peak tonnage in tonnes per hour, maximum incline in degrees, and the take-up travel the structure allows. If the centre distance is beyond roughly 400 m at serious tonnage, or the incline is steep enough that the belt must carry the load rather than push it, cord wins almost every time. Under 200 m at moderate load, it is usually an expensive answer to a cheap question.

Cord construction codes, and why the code matters to a buyer

The rope inside the belt is a strand of strands, and the digits describe that geometry. The first number is the number of outer strands, the second is the number of wires in each strand. A 6x7 rope has six strands of seven wires, and an open 6x7 with a larger pitch is the classic choice where adhesion to rubber matters more than fatigue life. A 7x19 rope packs more, finer wires into the same diameter and resists bending fatigue far better, which is why it appears on small pulleys and reversing lines.

What a buyer needs from this layer is not the rope design itself but its consequence. Cord diameter sets the minimum pulley diameter. Cord pitch sets how narrow a splice step can be. Rope type sets how the belt behaves at a bend, and therefore how long it survives. Ask for the load-elongation figure at 10 percent of breaking strength; a supplier who cannot produce it is selling a grade name, not a belt.

Grade and structure together set the minimum pulley diameter, and this is where a purchase order goes wrong quietly. The rule of thumb we work to is 145 to 150 times the cord diameter on the main drive drum. An ST1600 belt built on 7.5 mm rope therefore needs a drive pulley of about 1,100 mm before any allowance for cover thickness and pulley lagging. Fit that belt to an existing 800 mm drum and the ropes take a reverse bend they were never rated for. Cord fatigue appears at the splice first, usually between eight and sixteen months in, and it looks like a small hairline crack on the top cover about 200 mm before the splice edge.

Elongation is the other half of the story. Steel cord stretches roughly 0.25 percent at working tension and only about 0.3 percent at break. On a 2,000 m centre distance, we have to plan for a take-up stroke near 5 m to cover elastic stretch plus the 1.5 percent of belt length that goes into the two splices during their first hot cure. On a 1,200 t/h coal line the operators had built a 4.5 m car and the towers were already cast. We sized that belt on 7x19 at a slightly lower operating tension and got the required travel down to 4.2 m, which is how the project survived without a structural change.

That last example is the whole argument for this family. The cord build does not just make the belt strong. It makes the line's take-up, pulley diameters and splice plan predictable, and predictability is what a 20 year mine plan is actually buying. If that is your situation, our own build of this product is on the ST grade range we press in house, and the detailed specification route is on the page linked there.

Where the family loses is simple and worth stating for anyone comparing quotes. Cord belts need a vulcanised splice, so they need a press, a crew, or a service visit at the mine. They need real structure at both ends. They do not like being run at 90 degrees round a transfer or down to a 400 mm terminal pulley. Below 200 m and 400 t/h, the extra capability is dead weight on the invoice, and a two-ply fabric belt built to DIN 22102 will do the same job for less money. We say that to buyers before they order, because a wrong family order is the only mistake in this business that costs the customer a full belt rather than a sample.

Two practical alternatives exist inside the same rubber-and-steel idea. If your issue is not distance but a steep angle, the answer may be a profiled surface rather than more strength, and that lives on our chevron profile range. If the incline is vertical and the space is small, the same carcass is built with raised edges on the sidewall belt page.

03Family Two, Metal Mesh and Solid Steel Belts, Judged on Heat and Hygiene

Metal belts are the one family where the belt is the machine. Nothing is embedded and nothing is bonded, so the whole product is visible in daylight, which sounds like an advantage until the day it cracks at a sprocket and you cannot tell how long it has been running that way. Two sub-types matter commercially, and they behave nothing alike.

A woven mesh is built from wire that is formed into spirals or rods and interlocked. Balanced weave is the general-purpose pattern, rod-reinforced weave adds straight rods across the width for open, hot or heavy duty work, and chain-edge mesh puts a roller chain either side so a sprocket can drive and track the belt without tensioning it. Wire is normally austenitic stainless, 304 for general service, 316L where chlorides or washdown chemicals are present, with 314 and nickel alloys appearing above 700 °C. Wire diameters we see range from 1.0 mm to 3.0 mm, and the open area of the finished mesh runs anywhere from 30 to 70 percent depending on the weave.

A solid steel belt is the opposite idea. A continuous sheet of carbon or stainless steel, 0.5 mm to 1.2 mm thick, joined into an endless loop by a longitudinal or lap weld and then ground flush. Nothing is open, nothing falls through, and the surface can be polished for direct food contact. That makes it the correct answer for chocolate cooling, dough conveying, glass handling and any process where a clean, flat, metal-only surface is the point.

Now the buying test that decides this family in one question. Is the product hot, wet, sticky, sharp-edged or hygienically regulated? If any of those is true and the distance is short, metal wins. If the product is a bulk solid arriving at 1,500 t/h from a crusher, metal has already lost, because the belt cannot carry the tonnage and the open weave will let half the material fall through to the floor underneath.

We see the trade from the other side too. A food plant running a 12 m bake tunnel does not need a reinforced rubber belt at all, but we still get the enquiry, because the buyer's internal template says conveyor. A woven 316 mesh running at 320 °C through a proving tunnel is the right answer there, and the honest reply is to say so and point them to a specialist. Our own rubber range is a straight fit for a bakery line that handles flour at ambient temperature and needs washdown, and for that job the mesh would be the expensive mistake.

Standards in this family are about contact and alloy rather than breaking strength. Food and pharma meshes are specified against FDA 21 CFR 177.2600 for the polymer parts and EU 1935/2004 for the assembly, with the stainless grade quoted to EN 10088 or AISI. Heat treatment and glass lines are specified by alloy and by expansion allowance rather than by a belt grade. Nobody publishes a breaking strength in N/mm, because nobody drives a mesh belt by friction. If a supplier quotes you an ST number for a mesh, they are quoting the wrong family.

Cost follows the opposite logic to rubber. A metal belt is dear per metre and cheap to live with, while a rubber belt is cheaper per metre and eats money in splicing and take-up structure. For a two-shift bakery on a 30 m cooler, the mesh capital is recovered in the first year through the absence of vulcanising visits. For a 900 m aggregate line, no metal belt survives long enough for that calculation to start.

Two adjacent products get confused with mesh often enough to mention. Spiral freezer belts and self-stacking high-density freezers use the same material in a different geometry, and the tracked components that keep them straight are covered on our training idler page for the rubber side of the same problem. Where a plant runs both families, the tracking hardware is usually what breaks first.

How to tell the two sub-types apart on a specification sheet

If the sheet gives a wire diameter, a pitch and an open area, you are buying mesh. If it gives a thickness, a width, a material grade and a weld type, you are buying solid sheet. There is no overlap and there is no belt that is both. A supplier offering a single product line that covers hot glass, wet chicken and coal ore is either misunderstanding your enquiry or hoping you will not check.

This is also the family with the shortest lead time and the least room for a custom carcass, which makes it attractive to a purchasing team under pressure. That is exactly when the wrong family gets ordered. For an overview of how our own metal-adjacent product line sits against an industrial conveyor belt range, the two catalogues can be read side by side, and the conveyor belt supplier page lists which families we actually carry.

04Family Three, Steel-Reinforced Fabric Belts, Built for Impact

The third family is the one buyers rarely know by name and most often buy by accident. The carcass is ordinary woven fabric, usually polyester warp with polyamide weft, quoted as EP125 through EP630 where the number is the breaking strength in N/mm per ply. What makes it a steel product is a breaker layer, meaning a fine galvanised steel mesh or a set of crossed steel cords, 1.0 to 2.0 mm thick, laid above the top ply and under the top cover.

The breaker does not carry tension. It is a crack arrestor. When a digger tooth or a piece of tramp steel punches the top cover, the breaker spreads the load sideways and stops the slit from running the length of the belt. On a quarry line, that single layer is the difference between a 300 mm gouge that you repair at the weekend and a 60 m split that stops the plant.

Carcass design itself is a set of compromises. More plies mean more strength and more stiffness in the trough. Two plies are easier to splice and track, and they are what most modern high-impact lines use because the strength comes from the fabric grade rather than ply count. A 2-ply EP400 belt at 1,000 mm width carries 800 N/mm, which is a serious rating for a belt that still bends round a 500 mm pulley.

Here is where the family gets over-sold. A breaker ply is not a strength upgrade. Buyers regularly ask for a steel-reinforced belt at the same width and hope to run it on a line that needed a cord belt, and the numbers do not allow it. Elongation of an EP carcass at working tension sits near 1 percent, against 0.25 percent for steel cord. On the 900 t/h, 780 m limestone line we surveyed two years ago, the existing EP800 belt was stretching 1.4 percent, so the take-up needed 11 m of travel. The tower had 3 m. Lengthening the tower would have cost more than the belt, and the operator eventually moved to a cord construction for that reason alone.

Impact is the other half. At a primary crusher discharge, a 1,200 mm drop with 300 mm lump size produces point loads that no carcass enjoys. That is why we build the belt with a thicker cover over the breaker, generally 8 mm top and 4 mm bottom, and why the plate at the feed point matters as much as the belt. Our standard wear compounds sit at a maximum abrasion loss of 120 mm³ under ISO 4649 for the X grade, and the same test on a Y grade allows 150 mm³. On a clinker line, going from an unqualified cover to a genuine DIN 22102 X compound took top cover life from a measured 3 months to 14 months on the same 1,100 mm wide belt, with the same material and the same drop height.

Splices are the weak point of fabric belts and should be planned, not improvised. A two-ply belt is normally spliced in two stepped stages with a hot cure, and the finished joint reaches roughly 80 to 90 percent of belt strength when the steps are cut correctly. Cutting a step 20 mm short loses more than the 10 percent you were budgeting for. On short drives we often see a mechanical fastener used instead, which is honest on a 40 m conveyor and catastrophic on a 600 m one, because the fasteners fatigue the fabric at every pass.

So where does this family belong? Lines between roughly 100 and 400 m, tonnage in the hundreds rather than thousands of tonnes per hour, short head stations, mobile and shiftable plant, recycled material with unpredictable metal contamination, and any site where a press visit for a cord splice is difficult to schedule. That is a big chunk of the world's conveyors, and it is why so much of our rubber conveyor belt output leaves the plant with a breaker layer fitted as standard.

Cover compound choice matters more than the steel here. A quarry feeding wet aggregate wants abrasion resistance, and the surrounding pages cover that route in detail, from abrasion resistant grades through to the impact and cut resistant build used at the drop point. Heat, oil and chemical loads each pull the compound in different directions, so a line that combines two of them needs an honest conversation rather than a single grade name.

We are a conveyor belt manufacturer first, which means we get asked to make the wrong family less often than a trading house would. The breaker ply is added at the layup stage in our own conveyor belt factory, so the position of the mesh relative to the top ply can be set to the drop height the customer describes, not to a stock build.

05Choosing by Operating Condition Rather Than by Product Name

Nobody stands at a plant and says "I need family three". They say the belt keeps splitting, or the tower has no room, or the product arrives at 400 degrees. So the workable way to choose is to start from the condition and let the condition pick the family. The matrix below is the one we run through on a call, and it is short on purpose, because more than four deciding conditions usually means the enquiry needs a site visit rather than a quote.

Read the fourth column before you read the third. Knowing what will fail is more useful at the quotation stage than knowing what will work, because in most procurement processes two of the three families can be made to work on paper. The one that fails physically is the one to rule out first.

One condition deserves special care. Genuine multi-family enquiries do exist, and they are usually cement plants and ports. A cement works may run a cord belt from the quarry, a fabric belt with a breaker through the clinker transfer, and a metal mesh through the packing hall. Three different products, three different budgets, one purchasing department. In that situation it pays to buy the three families from people who make the one you rely on most, and to be clear about which is which on the drawing, because a mixed order is where grade names get swapped.

Supply chain shape also decides who you buy from. A regional stockist will hold cut lengths of fabric belt for same-day delivery, which is worth more than a price difference on a plant that stops for a torn belt at 3 am. A conveyor belt distributor with local stock and a splicing crew is a service business, not a warehouse. If your volume is large enough to plan, wholesale conveyor belts bought directly from the maker on a rolling 12 month schedule remove one layer of margin, but they also remove the emergency truck at 3 am unless you agree that separately. Decide which of the two you are actually paying for.

The drive side of a plant has its own version of this confusion, and it is worth naming here because it shows up in the same orders. A crusher or a fan needs power transmitted by a wedge belt or a timing belt, and choosing that by price alone has the same result as choosing a belt by price alone. The width-to-depth ratio and the cord type inside a wedge belt decide how much torque it can pass before it slips and burns, and the pulley profile has to match the belt section exactly. When a bulk plant also needs drive belts, that part of the order is specified with a transmission belt manufacturer in the same way, and the sections themselves are listed on the V-belt manufacturer page. Two orders, two sets of numbers, one supplier if the supplier genuinely makes both.

06Six Things to Demand Before You Place the Order

Once the family is settled, the purchase becomes a document exercise. The belt you receive will be judged later by numbers that were agreed now, or not agreed at all. These are the six we would not sign an order without, and each one is a paper question rather than a factory visit.

Safety factor is the number most often left loose, and it differs between families for good reasons. Steel cord belts are conventionally sized at 6.7 to 1 against rated strength because the cords behave predictably, while textile belts are sized nearer 10 to 1 because the fabric creeps and the splice is less efficient. RMA practice and DIN 22102 point the same way. If a quotation offers you a cord belt at a 10 to 1 factor without explaining why, you are paying 40 percent extra material for nothing, and if it offers a fabric belt at 6.7 to 1 you should ask who is going to inspect the splice every quarter.

The certificate deserves a paragraph of its own. A real mill sheet carries numbers from the batch that produced your belt, including cord or ply adhesion in N/mm, cover abrasion in mm³ and elongation at the reference load. We have seen an imported cord order arrive with a certificate quoting a generic grade table rather than test data, and separated cords from the cover within a year on a line where the drive pulley was one size smaller than the belt needed. That enquiry would have been resolved before shipment by two figures on one page. Our own inspection route is described on the quality assurance page, and the paperwork a buyer should expect is set out item by item in the steel cord buyer checklist, which is the right companion to this page if you have already decided on that family.

Also fix the tolerance regime in writing. Belt width usually runs to plus or minus 1 percent, thickness to plus or minus 0.5 mm, and length at the specified tension rather than laid flat on the factory floor. That last point causes more arguments than any other single clause, because a 900 m belt measured unstretched can differ from the stretched length by several metres, and somebody has to pay for the difference. Say which tension you mean, in the order, in writing.

07What a Wrong-Family Order Costs in Practice

Mistakes in this business are rarely dramatic. They arrive as a slow explanation of why the maintenance budget is three times what it was last year. Below are the four patterns we see most, with the symptom that gives them away and the number that would have prevented them.

Two of those are worth a story. On a port line handling 1,500 t/h of fertiliser, the operators replaced a failing fabric belt with a cord belt of the same width and grade, and the take-up stopped working within a shift. Nothing was wrong with the belt. The tower had been sized around a belt that stretches, and the cord belt did the opposite of stretching, so the counterweight bottomed out and the drive lost grip. The plant needed a shorter tower stroke and a different tension setting, which took a week to arrange and a full day of production to prove.

The other story is a paper mill that specified a metal mesh to avoid a splicing contract. It was a good decision on paper and a poor one in service, because the washdown chemicals at that plant were chloride based and the mesh was 304 rather than 316L. Pitting appeared at the chain edge within eight months. Nothing about the belt was defective. The alloy simply did not match the environment, and the fix was an alloy upgrade rather than a stronger belt.

Both cases point at the same habit. Ask what the environment does to the material, not what load the material can take. Corrosion, heat, impact and abrasion each remove a family from consideration long before the strength calculation is interesting. Where a line has to cope with several of those at once, we handle the trade on the belt itself, which is why the heat resistant range and the fire resistant range sit next to the ordinary abrasion grades rather than replacing them.

08Matching the Family to the Industry You Are Actually In

Industry is a useful shortcut, because the same material handling problem tends to repeat across a whole sector. A quarry feeding a primary crusher faces the same three conditions anywhere in the world, and the family that survives there is the same too. Use the sector as a first filter, then check the four numbers from section 05 before you commit.

The sector view also explains why the first page of search results looks the way it does. Makers of metal mesh publish for bakeries and glass plants, makers of cord belts publish for mines and ports, and each catalogue is honest inside its own boundary. What none of them does is tell a cement plant that it needs two different families under the same roof. That is the gap this page exists to fill, and if your enquiry sits on the boundary between two families, the useful reply is a question about your worst transfer point rather than a longer quotation.

Sector pages go deeper than this table can, and each one carries the operating detail that belongs to it. Mining and quarrying lines with steep inclines and long hauls are covered on the mining and quarrying page. Hot clinker and dusty kiln feed sit with the cement plant applications. Shift tonnage through a jetty or stockyard is handled under port bulk material handling, and where the belt touches the product directly the rules change again, as set out for food and packaging lines. Two more sectors behave like nobody expects, namely recycling plants where contamination is the design case rather than the exception, and logistics and warehousing where the belt runs all day at low load and wears out on tracking instead of abrasion.

Pick the two sectors closest to your line, read both, and note where they disagree. The disagreement is usually the real engineering decision on your site.

Send us your line data and get a family recommendation, not a catalogue

09Frequently Asked Questions

How do I tell a cord belt from a fabric belt in the warehouse?

Cut a short piece and look at the edge. A cord belt shows a single row of round steel ropes running lengthwise at a stable pitch, typically 12 to 20 mm apart, with rubber above and below. A fabric belt shows layered woven plies, and a reinforced fabric belt also shows fine mesh above the top ply. If the sample bends easily in both directions and the cut face is soft all the way through, it is fabric. Ropes are stiff and can be felt individually with a thumbnail.

How far can a metal mesh belt run before it stops making sense?

Between 20 and 60 m for most industrial weaves, with special builds reaching about 100 m, and the load limit is the real constraint rather than the length. A mesh carrying 30 to 60 kg per square metre of belt area is at its practical ceiling, which is why a line moving hundreds of tonnes per hour cannot use it regardless of how short the span is. Add heat and hygiene to the case and the same belt becomes the obvious answer again.

Why does stretch matter more than strength on a long overland line?

Because the structure is already built by the time the belt is chosen. An EP carcass running near 1.4 percent elongation on a 780 m centre distance needs 11 m of take-up travel, while a cord build at 0.25 percent needs under 2 m for the same length. If the tower holds 3 m, the low stretch belt is the only belt that fits the tower, whatever the breaking strength comparison says. Length also multiplies the elongation, so the argument gets stronger as the line gets longer.

Can a steel-reinforced fabric belt do the work of a cord belt at a crusher discharge?

At the discharge, often yes, and that is exactly what the breaker layer is for. Along the haul, usually no. The steel mesh in a reinforced fabric build arrests a tear, it does not carry tension, so the belt still stretches 0.9 to 1.4 percent under load and still needs a generous take-up. Use it where impact and tramp metal are the problem and keep the cord build for distance, tonnage and incline.

Which standards should appear on the mill certificate?

For a cord belt, the relevant ones are ISO 15236-1, DIN 22131, AS 1332 and GB/T 9770, with cover abrasion quoted to ISO 4649. For a fabric carcass, DIN 22102 or the equivalent RMA grade is the reference, and a two-ply joint should be stated against an expected efficiency of 80 to 90 percent. What matters more than the list is that the numbers are batch figures from your belt rather than a generic grade table.

What happens if the drive pulley is smaller than the belt needs?

The ropes take a reverse bend beyond their rating and fatigue at the point where the belt leaves the drum, which on most installations is just ahead of a splice. The first sign is a hairline crack in the top cover roughly 200 mm before the splice edge, and it usually appears between eight and sixteen months after installation. The rule we work to is a drum of 145 to 150 times the cord diameter, so an ST1600 built on 7.5 mm rope needs about 1,100 mm, and an 800 mm drum is not a small compromise but a different belt.