Mining Conveyor Rollers Supplier: Roller Grade for Ore, Dust and Water

Roller selection on a mining site starts with a question that has nothing to do with price: what will kill this roller first? A 152 mm return roller under a 1,800 mm belt at a copper concentrator is asked to survive three separate attacks at the same time, and whichever attack wins the race decides your maintenance budget for the next five years. We walked a concentrator line in 2024 where the maintenance team had been swapping rollers on a ninety-day cycle and blaming the belt. The belt was innocent. The rollers had been bought as a general-purpose item, and the plant was moving wet, fine, abrasive concentrate through a transfer point that sprayed fines straight onto the return strand. Seized bearings had been dragging on the cover until it scuffed through to the carcass.

That plant eventually rebuilt its whole roller standard around three facts it had never written down. Dust chemistry, washdown water, and ore abrasiveness. Everything else — price, lead time, brand — came second.

Send us your ore type, belt width and washdown practice — we will return a roller grade recommendation

01Start With the Failure Modes, Not the Price List

Mining rollers fail in a small number of recognisable ways, and each way points back to a different specification field. Bearing seizure from dust ingress is the most common, and it rarely announces itself. The roller keeps turning, the shell surface temperature climbs, and by the time a fitter notices the grease has carbonised the raceway is already scrapped. Shell wear-through from abrasive ore is the second, and it behaves differently because it is gradual. A wall thickness that started at 3.2 mm can be down to 1.6 mm at the contact band after eighteen months of granite, and the shell then deforms oval under load. Shaft fatigue is the third, and it is the one that hurts most because it takes the belt with it when the roller collapses into the structure.

Ore does not negotiate. It grinds whatever surface you give it.

Corrosion sits behind all three of those. In a wet concentrator or a port terminal with salt air, the end cap and the shaft end are the first parts to lose their protective coating, and once the seal carrier face pits, the sealing lip no longer sits flat. That single pit turns a well-designed labyrinth into an open door. Sites that buy rollers purely on the delivered piece price almost always end up paying for the same roller three times over: once at purchase, once at the changeout, and once in production lost while a crew walks the line with a torch and a chain block.

The economics are easy to state and hard to remember under budget pressure. A roller that costs eighteen dollars instead of twelve, carrying a heavier wall and a proper labyrinth, may last three times as long in abrasive duty. Changeout labour on a high-level conveyor often exceeds the roller value within the first hour. Add the unplanned stop, and the arithmetic becomes absurd. This is why serious operators treat the roller standard as an engineered item rather than a consumable line in the stores catalogue. When we audit a plant's roller specification, we usually find the same gap — the belt side was specified properly, and the roller side was inherited from whatever the original equipment supplier shipped. That imbalance is expensive. A well-specified industrial conveyor belt running on badly matched rollers will still be damaged, because a seized roller acts like a stationary scraper against a moving belt.

Three threat families drive nearly every roller decision on a mine site. Abrasion from the ore itself, which sets shell material and wall thickness. Ingress from dust and fines, which sets the seal class and the bearing clearance. Corrosion from water and atmosphere, which sets coatings and shaft materials. Get all three right and the roller becomes boring, which is exactly what you want. Get one wrong and the whole string inherits the problem, because a single seized roller in a 40-roller panel raises the drag on its neighbours and accelerates their wear as well.

We have seen the reverse too. A quarry in the wet season replaced an entire panel with high-grade sealed rollers and then watched them fail in five months because the site had changed its washdown to a high-pressure jet that was driving water past a seal design meant only for fines. The rollers were good. The threat had moved. So the first job of any mining conveyor rollers supplier is not to sell you a catalogue number but to establish which threat dominates your circuit and where on the circuit it changes. Duty is rarely uniform along a line, and a supplier who quotes one roller type for the whole conveyor is telling you something about how they work.

A realistic roller programme therefore starts with a survey. Where does the ore arrive, how high is the drop, how wet is the product, what does the washdown actually do, and what is the ambient chloride load? We ask those questions before we quote, and on a recent granite installation those answers changed the recommendation from a 133 mm shell to a 159 mm shell with a thicker wall on the load side alone. The tonnage did not change. The drop height and the quartz content did. For a broader view of how circuits differ between hard rock and softer bulk materials, the site notes we keep on mining and quarrying operations are a useful starting point, and the roller and idler range itself is listed under conveyor rollers.

None of this replaces a site visit. It replaces guesswork.

Buyers who arrive with a clear threat statement get far better value than buyers who arrive with a part number. If you can tell a supplier the ore, the moisture, the drop height and the washdown regime, a competent conveyor belt manufacturer can map that to a roller grade in a single exchange. If you cannot, the supplier will default to a mid-range general-purpose roller, and mid-range is exactly the specification that survives nothing well. Our own engineering group spends a large part of each week on that mapping exercise, and the questions we ask are the same ones this article walks through section by section.

02Reading Ore Abrasiveness Before You Choose a Roller Grade

Abrasiveness is not the same thing as hardness, and confusing the two is the most expensive mistake we see in roller tenders. Quartz at Mohs 7 is not dangerous on its own if the particles are rounded and dry. The same quartz becomes a cutting tool once it is angular, fine and damp, because it then packs between the shell and the seal and abrades both at once. On a gold reef operation we measured 62% free quartz in the fines fraction with a Bond abrasion index of 0.68, and the roller contact bands showed 0.9 mm of radial wear in eleven months. On a lignite yard with a Bond index below 0.05, the same nominal roller geometry was still within 0.2 mm after three years.

Sites that want a defensible roller grade need three numbers. The Bond abrasion index or an equivalent laboratory abrasion figure, the particle size distribution of the fines fraction below 5 mm, and the moisture content at the point of loading. Any one of those alone will mislead you, and we have watched a plant over-specify to the point of buying heavy-wall stainless rollers for a dry, coarse, low-abrasion coal stream where a standard mild steel shell would have run a decade.

The shell is where abrasion is won or lost, but the failure mechanism is not simply thinning. As the wall wears, the section modulus falls, and the roller begins to bend between its bearings. That deflection raises the running torque, which raises the belt drag, which raises the power draw. Operators often notice the symptom on the drive panel long before anyone looks at the rollers. A 12% rise in the total conveyor power on a steady throughput is a reliable early warning that the roller string has degraded, and it is cheaper to act on that signal than to wait for a shell to perforate.

Material grade on its own tells you less than people assume. Cold-formed steel tube to a stated wall thickness performs differently from a seam-welded tube of the same nominal wall, particularly in fatigue. Specify the tube, the wall tolerance and the straightness limit, not just the steel grade. Wall thickness tolerances of plus or minus 10% are common in the market, and a batch that arrives at the bottom of that band will wear out measurably faster than a batch at nominal, even though the drawing says the same thing.

Note what that table is really doing. It is trading shell cost against changeout frequency, and the trade only pays if the downtime figure you apply is honest. Plants that value a stopped conveyor at the cost of the lost tonnage alone will always under-specify. Plants that include the labour, the crane, the permit, the risk assessment and the restart ramp will almost always justify the heavier wall.

One more variable sits on top of the ore itself, and it is the one buyers forget. Belt speed. A roller under a 5.5 m/s belt sees far more load cycles per year than the same roller under a 2.5 m/s belt, and bearing fatigue scales with cycles rather than with calendar time. We have measured L10 life estimates that halved when a plant raised a single line from 3.2 m/s to 4.6 m/s to meet a throughput target, with no change to the ore, the belt or the structure. If your tonnage plan is changing, your roller grade should be reviewed in the same meeting. The dimensional and grade options we stock are set out in the roller selection guide, and buyers consolidating several circuits into a single order normally do it through our wholesale conveyor belts programme so that the grades stay consistent across sites. Cement plants, which sit between quarry hardness and fine dust, are covered separately in our notes on cement plant handling.

Do not skip the lab work to save two weeks. A sieve analysis costs less than a single roller panel.

03Dust, Fines and Sealing Classes for High-Fines Circuits

Dust ingress is the highest-frequency failure cause in the roller population we service, and it has an unfair advantage: it works silently. A labyrinth seal does not fail in a single event. It fails over a series of thermal cycles, each one drawing a fraction of a milligram of fines past the lip, until the grease channels are packed and the bearing runs dry. By then the roller is a grenade with the pin half out. The next shift that loads the belt heavily has a seized shell cutting into the cover.

Three sealing families cover most mining duty, and the choice between them is driven by particle size rather than by dust concentration. A single lip running against a machined face works for coarse, dry dust above roughly 200 microns because the particles are too large to enter the gap. A two-stage labyrinth handles the 50 to 200 micron band, which is where most crushed ore fines sit. Below 50 microns, and especially when the material is damp or carries clay, you need a three-stage labyrinth with a grease-filled cavity, because at that size the particles behave like a fluid and will follow any continuous path into the bearing.

What surprised us most on a laterite nickel operation was the effect of humidity alone. There was no washdown, no spillage and the dust was classified as coarse, yet the failure rate was twice the site's own model. When we pulled sample rollers, the grease was emulsified. Overnight condensation inside the shell had been enough to turn the fines sitting in the seal into a paste, and that paste was migrating inward. The fix had nothing to do with the dust class. It was a change in the fill grade and a small change to the cavity geometry.

Sealed-for-life and re-greaseable rollers are not interchangeable, and the industry often treats them as though they are. A re-greaseable roller only justifies itself if the site actually greases it on schedule and if the grease port is engineered so that old grease exits rather than compressing into the seal. On a mine with a maintenance backlog, a re-greaseable roller that nobody greases performs worse than a properly sealed one, because the port itself becomes an ingress path. Ask the supplier which of the two they are offering and why, and be sceptical of an answer that depends on the site being perfect.

Fines also change how you inspect. Random spot checks miss seal degradation because it is a clustered failure, not a random one. It concentrates at the loading zone, at the skirt, and anywhere the belt path changes direction after a spill. Build your inspection route around those three locations and you will find problems two months earlier than a uniform walk of the line. Our field notes on dust proof idler sealing cover the seal stack drawings in more detail, and the interplay between bearing selection and seal choice is separated out in the article on bearing and seal selection. Where the impact is severe enough that the roller itself is being struck, the geometry question moves to the impact idler range, which uses a different rubber ring strategy.

Every roller batch should be traceable back to the seal architecture that was specified, and that traceability is worth more than a certificate in a folder. When a plant can tell you which seal class is running in which panel, a failure becomes a data point instead of an argument. That is the reason we insist that batches carry a marking that survives three years of ore dust, and it sits inside the same discipline that governs our quality assurance process. Buyers who standardise on a single conveyor belt supplier for both the belt and the roller string get an additional benefit here, because the belt cover grade and the roller seal class can be matched to the same dust sample rather than to two different assumptions. It also helps that the belt itself, whether a heavy rubber conveyor belt or a lighter food-grade construction, is specified against the same survey data.

04Water, Washdown and Corrosion Duty in Wet Mining Circuits

Water is the variable that most often turns a correct roller specification into a wrong one, because water changes quietly. A plant can install rollers into a scheme designed around dust control, and eighteen months later a new environmental permit requires the transfer points to be misted. Nobody reopens the roller specification. The seal that was sized for dry fines is now seeing a fine suspension, and a suspension is far better at penetrating a lip than dry dust ever was.

Chlorides matter as much as volume. A coastal port terminal we audited had modest washdown but an atmospheric salt load that pitted the end caps within one season. The fix was not a thicker coating on the shell, which was still in good condition. It was a change to the end cap material and the fastener grade, because the corrosion was concentrated where the assembly exposed dissimilar metals. Galvanic pairs do more damage on a roller than general surface rust, and they do it faster.

Three practical questions settle most water decisions. Is the water applied as a mist, a low-pressure hose or a high-pressure jet? Is the water process water with dissolved solids, or potable? And does the roller drain, or does it sit in a puddle after every washdown? A roller that stands in water has a different problem from a roller that is merely sprayed, because the wetted band is permanent and the seal sees continuous hydraulic head rather than a transient one.

One thing we say to every buyer considering stainless: it is not automatically the answer. Stainless resists corrosion and is often softer and more prone to galling than the plated part it replaces. In a highly abrasive duty with no water at all, specifying stainless can reduce wear life while raising cost. The rule we apply is simple. Match the material to the dominant threat, and if abrasion and corrosion are both severe, use a coated or hardened component rather than a bare stainless one.

Roller hardware is only part of a conveyor standard, and the parts that fail alongside rollers usually come from the same procurement decision. Brackets, frame fixings and pulley lagging all live in the same wet zone, and a roller that is well protected but bolted to a corroded bracket will still end up misaligned and eventually seized. Our conveyor brackets range is specified to be matched with the roller grade, and for the heavier shafts used in high-tonnage installations the steel roller construction carries the load better than a light formed shell. Terminals that combine salt air with heavy rainfall are a special case, and our notes on port bulk handling set out how we approach them.

The lesson from every wet site we have worked on is administrative rather than technical. Someone has to own the link between the cleaning regime and the roller specification. Plants that assign that ownership to one engineer have stable roller life. Plants that split it between operations, maintenance and procurement watch the specification drift for a decade.

Drive-side items follow the same logic, and many of the mining accounts we serve consolidate them with the roller order to keep one material standard across the plant. Belt drives, gearbox-side items and the sheave hardware are all part of that conversation, and buyers comparing power transmission grades typically start from our transmission belt manufacturer range before narrowing down to a specific section and length. The heavier wrapped constructions for crusher and pump drives sit in the V-belt manufacturer line, where the same grade-to-duty thinking applies.

05Matching Shell, Bearing and Shaft to Load and Speed

Load on a roller is not the weight of the ore sitting above it. It is the sum of the belt mass, the material mass, the idler spacing effect, and the dynamic amplification from the belt sag between supports. Buyers routinely understate that last term, and it is the reason a roller that appears to be loaded at 40% of its rated capacity fails at two years. We have instrumented enough lines to say this plainly: if the sag between rollers exceeds 2% of the spacing, the roller sees load spikes that no static calculation will show you.

Diameter is the cheapest lever you have. Moving from 133 mm to 159 mm increases the bearing size you can fit, cuts the number of load cycles per metre travelled, and reduces the contact pressure at the shell. It also changes the frame height, so the decision has to be made before steel is fabricated. Retrofitting a larger diameter into an existing structure is possible but rarely elegant, and we have seen plants spend more on adapter brackets than they would have spent on the correct rollers in the first place.

Bearing selection follows the load but is capped by the shell bore. The practical sequence is to size the roller diameter, then choose the largest bearing that fits with an adequate wall section, then verify the L10 life against the actual speed and load. A common mistake is to compute bearing life at the nominal belt speed and ignore the fact that a variable-speed drive may run the belt at 130% of nominal during recovery from a stoppage. Life is inversely proportional to roughly the third power of load, so a 30% load overshoot costs you a large fraction of the design life.

Shaft design is where cheap rollers hide their compromise. A shaft undersized for the bearing bore will deflect under the same load, and deflection is what starts the chain of seal damage. We specify shaft straightness and hardness as explicit line items because a soft shaft lets the bearing turn on the shaft instead of in the raceway, which produces a distinctive polishing pattern and a roller that runs true for a month and then develops a wobble. Look for that pattern during your first inspection round and it tells you exactly where the supplier cut the corner.

If you cannot inspect the plant yourself, ask for the numbers. A competent supplier can supply shaft diameter, material, straightness tolerance and hardness for any roller in the range, and a supplier who cannot is quoting from a photo rather than a drawing. This is the kind of detail that separates a genuine conveyor belt factory from a trading intermediary, and it is worth asking about before you place a trial order, not after. Buyers running high tonnage lines often combine the roller decision with a troughing geometry review, and the troughing idler page sets out how the three-roll arrangement affects the load split. Assemblies and frames that carry these rollers are listed under conveyor components, and the tracking benefits of getting the geometry right are covered in our note on self-aligning rollers and tracking.

Weight is a useful check when drawings are not available. Two rollers of identical external dimensions should be within a few percent of each other in mass, and a significantly lighter roller almost always means a thinner wall, a smaller bearing or a lighter shaft. Put a batch on a scale at goods-in. It is the cheapest engineering test you will ever run, and it catches substitution before the rollers reach the structure.

06Spacing, Trough Angle and Transition Zones

Spacing is the variable that most directly controls how much load each roller carries, and it is also the variable plants change most casually when they are chasing a cost reduction. Widening the spacing by 15% raises the load per roller by roughly the same proportion, increases belt sag, and pushes the belt into a deeper catenary between supports. The ore then sits in a moving trough rather than a stable one, spillage rises at the skirt, and the rollers under the loading zone take the extra material mass on top of the impact load they were already absorbing.

Trough angle is chosen for material behaviour, not for capacity alone. A 20 degree arrangement suits fine, free-flowing and often dry material because the belt stays flatter and the load profile is stable. A 35 degree arrangement is the workhorse for mixed ore and gives a useful capacity gain without making the belt train badly. A 45 degree arrangement maximises capacity but punishes the belt edge and demands good alignment, because the outer rolls now carry a large share of the load and any misalignment concentrates stress at the belt edge.

We watched a copper plant change a 35 degree string to 45 degree on a capacity upgrade, and the roller life in the outer positions dropped by a third within a year. The ore had not changed, the belt had not changed, and the tonnage had risen by 12%. The extra load at the wing rolls did the damage. The plant reverted to 35 degrees, accepted a slightly lower capacity and cut its roller consumption back to the earlier level. On that line, the throughput gain was not worth the consumable cost, and only the roller record made that visible.

Transition zones are where most of the unexplained failures live. Between the last troughing station and the pulley, the belt has to change from a troughed shape to a flat one, and if that change happens too quickly the belt edge is stretched and the roller at the transition point is loaded unevenly. The standard rule is to allow a transition length of at least one and a half times the belt width for fabric belts, and longer for thick, high-tension constructions. Shortening that distance to save steel is a false economy that shows up as edge damage and roller failure two years later. Our notes on training idlers explain how to control the belt through that region, and the return idler arrangement on the underside needs its own spacing review because the return strand carries a fraction of the load but almost all of the tracking risk. Pulleys, which set the geometry at the ends of the transition, are covered in the conveyor pulley range.

Also tighten the spacing at impact stations, and do it properly rather than approximately. Below a chute where lump ore lands, the load is transient and high, and the belt needs a support almost continuously for the first two or three metres. The rubber-ring construction used in the impact idler range absorbs that energy through deformation rather than transmitting it into the bearing. Skipping the impact station to save cost shifts the damage into the belt cover, and the impact and cut resistant belt constructions are designed to work with that station, not to replace it. When we see a plant replacing belts more often than rollers, the impact zone is usually the reason.

Spacing decisions should be documented as a drawing, not as a habit. The most useful single document a plant can hold is a marked-up general arrangement showing roller diameter, spacing and grade for each section of each conveyor. We have walked into plants where that document existed and plants where it did not, and the difference in roller consumption is stark. On the sites where it existed, the maintenance team could order exactly what was running and could see immediately when a contractor had substituted a different grade. Where it did not exist, every shutdown produced a mixed population of rollers with no way to trace which grade was performing well.

This is also the point where a buyer should test the reach of their supply chain. A conveyor belt distributor who keeps a genuine local stock position can supply the odd roller diameter or the short-run impact station without a full container order, and that flexibility matters more than unit price when a line is down. Ask what is held locally, in what diameters, and with what lead time. If the answer is that everything ships from a distant warehouse, then every emergency becomes a minimum order quantity.

Most mining roller problems are geometry problems wearing a consumable costume.

Buyers who want a second opinion on a proposed spacing before fabrication should send us the conveyor profile and the design tonnage. Reviewing it takes us less time than a failed panel takes to replace, and the arithmetic is far cheaper to change on paper. The buyer's checklist in our article on what buyers should know before ordering idlers is a reasonable template for the questions to ask, and the same discipline applies when the roller set is bought alongside the belt itself.

07Supplier Deliverables Beyond the Roller Itself

A roller order is only as good as the paperwork that travels with it, and mining buyers should treat the document set as part of the specification rather than as an administrative afterthought. When a failure happens two years into service, the difference between a warranty conversation and a shrug is usually a batch record that links the failed unit to a heat number, a wall thickness reading and a bearing lot. Suppliers who cannot produce that link are not being difficult. They never created it.

Ask for the material certificate first, and ask for it on the actual production lot rather than as a generic mill certificate held in the sales office. On tubes, the useful document is a dimensional inspection record showing wall thickness at several points along the length, because the tolerance band is where most substitution happens. On bearings, the useful document is a lot number and a manufacturer declaration, because a lot number is what makes a future failure traceable. We have seen orders where the bearing appeared to be a named brand on the packing list and a different product in the box, and only a lot number check would have caught it.

Then agree the sampling plan before production starts, not after the first complaint. A sampling plan turns a quality argument into an arithmetic decision, and it also protects the supplier, because a reasonable plan prevents a single outlier from triggering a full batch rejection. For rollers we normally propose a plan based on batch size with a defined accept and reject number, and we measure wall thickness, running torque, seal condition and marking on each sampled unit rather than only on the first one.

Packing and transport deserve their own line in the purchase order, and mining buyers underestimate this constantly. Rollers stacked directly on a container floor arrive with deformed shells and flattened seals, and the damage is easy to miss under a layer of dust on the first inspection. Specify end caps protected, rollers racked rather than loose, and a maximum stack height. If the shipment crosses a humid region, ask for moisture barrier packing, because condensation inside a container is enough to start corrosion on bare steel before the goods ever reach your yard. Our pre-shipment approach is described in the note on roller supply specifications, and the RFQ structure we recommend for idler and roller packages is set out in the idler RFQ checklist.

One last deliverable is rarely written down and always valuable. Ask the supplier for their recommended changeout interval per roller grade, expressed in operating hours rather than months. A supplier who can give you that number has a service history behind the recommendation. A supplier who cannot is guessing, and their guess will be based on the mildest duty they have ever seen. Where impact protection is a live question, the note on how impact rollers protect the belt shows how the interval changes once the transfer point is treated as part of the roller specification.

08Installation, Commissioning and Lifecycle Support

Even a correctly graded roller underperforms if it is installed badly, and most installation faults are invisible in the first month. The two that cause the most damage are a bracket that sits out of square and a roller that is forced into a frame with a hammer. Both end with the same result, which is a roller carrying a bending load it was never designed for, and both are simple to prevent with a straightedge and a bit of patience.

Alignment should be checked against the pulley centreline and the frame, not against the previous roller. We have found panels where the whole string had been progressively installed to a drifting reference, so each roller was individually acceptable and the line as a whole was not. Measure the diagonal of the panel frame before you start. If the diagonals differ by more than a few millimetres, fix the structure first, because no roller grade will compensate for a skewed frame.

Commissioning is the right moment to record baseline data. Take running torque readings on a sample of rollers, note ambient and bearing temperatures after two hours of steady running, and photograph the panel. Those three items give you a reference that makes every later decision easier, and they cost half an hour. A plant that installs 4,000 rollers and records nothing has thrown away the information that would have justified a longer interval or exposed a bad batch.

Then build the lifecycle plan around what the baseline shows rather than around a calendar. Rollers rarely fail uniformly, and the failure distribution tells you where to spend attention. If the first failures arrive at the loading zone, the transfer geometry is the problem and the roller grade is only the messenger. If they arrive at the tail, tracking and alignment need work. If they arrive uniformly across the line, the grade itself is too light for the duty. We have used that simple logic on scores of sites, and it holds even on circuits as different as iron ore and municipal recycling.

Spares strategy is the last piece, and it should be derived from the failure distribution rather than from a round number of cartons. Hold enough of the highest-wear positions to cover a single panel change plus one spare panel, and hold a smaller quantity of the grades used in lower-wear positions. Standardising on fewer diameters across the plant reduces the spares holding while making every part interchangeable, and that is usually a bigger saving over five years than the initial price negotiation. Our service team supports that planning directly and the scope is described on the service page. For lines that run through flexible sections, the garland idler arrangement is often easier to maintain because the idler set can be replaced without dismantling the frame, and the self-aligning idler variant reduces tracking interventions on long conveyors. Even in odd duties, like the sort of mixed-material streams we see in recycling plants, the same logic about baseline data and failure distribution applies.

Do not retire a failed roller without opening it. The grease condition inside a failed unit tells you more about the real duty than any specification sheet, and that evidence is destroyed the moment the roller goes into the scrap bin.

Talk to our engineers about a roller grade audit

09Frequently Asked Questions

How do I know which roller grade suits my ore before I order?

Start from three measurements rather than from a catalogue. Get the abrasiveness index or an equivalent lab figure, the particle size distribution of the fines below 5 mm, and the moisture content at the loading point. Those three numbers separate the threat into abrasion, ingress or corrosion, and each one points to a different shell wall, seal class or coating. If your plant has no lab data, a supplier with field experience can still make a defensible recommendation from the ore type, the drop height and the washdown practice, but you should treat that as a starting point to be confirmed after the first inspection round rather than as a final answer.

What wall thickness should I specify for abrasive ore?

For most crushed ore duty, a 3.2 mm wall is the minimum we would accept and 4.0 mm is the practical default once the material is heavy and angular. Granite and quartz-bearing ore justify 4.5 mm, and in those cases it is usually better to also move up a diameter so that the contact pressure at the shell falls at the same time. Wall thickness alone does not protect you, because a thin-walled large roller can perform worse than a thicker small one. Check the delivered wall with a gauge at goods-in, since the tolerance band is where the savings in a cheap quote are usually hidden.

Is a sealed-for-life roller better than a re-greaseable one?

It depends entirely on whether your site will follow the greasing schedule. Sealed-for-life units remove a maintenance task and eliminate the risk of a grease port that nobody uses becoming an ingress path, and they are the safer choice on plants with a backlog. A re-greaseable roller can outlast a sealed one if the grease quality is controlled, the interval is respected and the port is designed so that old grease can escape. Ask which type is being offered and why, and be suspicious of an answer that assumes everything on site runs perfectly.

How should washdown affect my roller choice?

Match the seal to the water first, then the material to the water chemistry. Mist and low-pressure hose washdown is usually handled with a coated shell and a stainless end cap, while a high-pressure jet needs a fully sealed assembly and nozzles repositioned so they do not play directly on the roller line. If the water is process water with dissolved solids or the site is coastal, add stainless shafts and isolating washers to break the galvanic path between dissimilar metals. Standing water is a separate problem, and it is solved by drainage before it is solved by any roller grade.

What documents should come with a roller shipment?

The set we insist on includes a tube dimensional record with wall thickness readings, a bearing lot declaration, the seal stack drawing, coating thickness and adhesion results, running torque and concentricity data, and a batch marking map. All of it should be tied to the production lot rather than to a generic certificate held in a sales office, because lot-level traceability is what makes a future failure analysable. The marking scheme matters as much as the certificates, since without it a mixed population in a panel cannot be traced back to the grade that performed well.

Can spacings be widened to reduce the number of rollers?

Only with a proper sag calculation, and rarely by much. Widening spacing raises the load on every remaining roller, deepens the catenary between supports and increases spillage at the skirt, and the loading zone still needs its own tight spacing regardless of what the rest of the line does. A few plants do save cost this way on the return strand, where the load is a fraction of the carry side, and that is a legitimate change. Doing it on the carry side of an abrasive line usually shifts the cost from the roller budget into the belt budget and the downtime budget.