Timing Belts Conveyor Technology for Food Lines: Compliance and Grades

A food plant does not usually arrive at positive-drive conveying for elegant engineering reasons. It arrives there because of washdown. Once a line is hosed twice a shift with alkaline foam and a 70 °C rinse, friction-drive flat belts start to slip, tracking wanders, and the maintenance crew spends the small hours re-tensioning instead of running product. Teeth dropping into a matching pulley groove change that arithmetic. A toothed drive can run slack, and a slack belt tracks better across a long span than a tight one.

Almost none of that benefit survives a careless specification. Timing belts conveyor technology in food processing answers to a different set of constraints than a general industrial drive, and most of them are invisible on a datasheet. The polymer has to survive detergent chemistry and hot water without hydrolysis. The running surface has to be cleanable to a standard an auditor accepts on the first walk-through. Where the line passes a metal detector or an X-ray cabinet, the belt stops being a component and becomes a foreign-body risk that must be detectable in its own right. Add pitch accuracy that has to hold while the belt stretches, steams, cools and gets re-tensioned for years, and the specification problem turns serious.

What follows is the working order I use when a plant engineer sends over a drive for review, from duty profile through to the paperwork that supports it.

Send us your food-line drive details and we will size the belt with you

01Food-Processing Duty Profile: What These Belts Actually Face

A general industrial timing drive spends its life doing one thing at constant speed, in dry air, with a guard bolted over it. A food-line drive does not get that courtesy. It starts and stops with the filler, runs slow enough that product sits on the belt for minutes, and lives in a room where the relative humidity after sanitation sits above 85 % for two hours every night. The belt is not the only thing moving; the product is often wet, oily, sugary or acidic, and it leaves residue exactly where the teeth engage.

The consequences stack up in a specific order. Water and residue reach the tooth flanks and the pulley grooves, so friction between tooth and groove changes and the belt starts to ride differently. Then the drive loses the tension you set on Monday, because the carcass relaxes as moisture swaps in and out of the polymer. Then tracking drifts, and a drifting toothed belt does something a flat belt cannot do: it climbs the pulley flange, wears the tooth corners and starts to shear teeth off one at a time. I have seen an entire 32 mm pitch belt lose seven consecutive teeth across two weeks on a dairy line, purely because nobody re-checked tension after the first washdown cycle.

Temperature is the second axis, and it is worse than the number on the label suggests. A belt rated to 80 °C continuous does not see 80 °C. It sees a 92 °C caustic wash for twenty minutes, then a 6 °C chill tunnel, then ambient. That cycling is what kills polyurethane, and hydrolysis of ester-based TPU is the mechanism behind most of the cracking I find on the back side of belts that are only three years old.

Two of those five rows decide the grade. The rest decide the life. And note that none of them are solved by picking a bigger belt. Oversizing a food-line drive usually makes the problem worse, because a heavier belt needs tighter tension, and tighter tension loads the bearings and the pulley grooves that you were trying to protect.

02Compliance Framework: Washdown, Antimicrobial and Metal-Detectable

Compliance on a food-line belt is not one certificate. It is three separate arguments that buyers habitually collapse into a single request for "food grade paper", which is why so many approvals stall. As a conveyor belt manufacturer we see the same three questions arrive in the same order, and each one has its own evidence trail.

The first argument is food contact. In the European Union the framework is Regulation (EC) 1935/2004, with plastics covered by Regulation (EU) 10/2011 and its migration limits; in the United States the equivalent route for rubber articles intended for repeated use runs through 21 CFR 177.2600, and polyurethane resins sit under 21 CFR 177.1680. A declaration of compliance should name the regulation, the migration test conditions, the simulants used and the temperature and duration envelope. If a supplier sends a one-page letter that says "FDA approved" with no test conditions, you have a marketing document, not a compliance document, and your own auditor will say so.

The second argument is cleanability and microbiological performance. This is where washdown ratings, surface roughness and antimicrobial additives live. A common misunderstanding is that antibacterial additives replace cleaning. They do not, and ISO 22196 testing on a plastic surface tells you about the material, not about the crevices at your splice. For open hygienic design, the EHEDG guidelines and 3-A Sanitary Standards describe what a cleanable profile looks like, and the practical translation on a belt is this: no fabric edges exposed, no open tooth roots that trap product, and a splice that is flush on both faces.

The third argument is foreign-body control, and it is the one that gets plants into trouble. Under a HACCP plan, and under schemes such as FSSC 22000 or BRCGS, a belt that can shed material into product has to be either eliminated or made detectable. That is why detectable grades exist, and why their specification must be written in a way your quality team can verify at goods-in. Read the line audit practices we describe on our quality assurance page, because the same discipline applies to the belt you buy next quarter.

How detectable grades are actually built

Metal-detectable polyurethane usually carries a dispersed iron-based filler, and X-ray-detectable grades lean on denser fillers such as barium sulphate or fine stainless powder. Detection is a function of mass and geometry, not of the label. A 1.5 mm chip of a detectable belt may sit below the sensitivity of a ferrous aperture set at 2.0 mm, while a 4 mm fragment of the same belt triggers reliably. Ask for the fragment size the supplier validated at, then run your own test with cut fragments at the size your risk assessment assumes.

Colour is not decoration either. Blue belts dominate food plants because blue does not occur naturally in most product streams, so a fragment is visible to an operator during a visual check. A detectable filler plus blue compound plus a documented fragment-size test is the combination that closes the foreign-body argument. Anything short of that leaves the argument open.

If your team runs acidic product and wants a detectable belt as well, that combination is the one where compounders genuinely have to work, because some filler chemistries and some acid exposures do not coexist. Bring the product pH and the cleaning chemical list to the conversation instead of discovering the constraint after the first order.

You will notice that nothing in this section mentioned tension, pulley size or tooth count. Get the compliance layer right first, because changing grade later invalidates the mechanical sizing you did before it. That sequencing mistake is expensive, and it happens on roughly one in four projects that reach us with a frozen drive layout. If the food-contact route also runs through broader material questions, our chemical resistant conveyor belt page carries the compound discussion further than a food article reasonably can.

03Tooth Profiles for Food Lines: Which Pitch Fits Which Job

Pitch selection is where a food project diverges most sharply from a general drive design. On a machine tool you optimise for positional accuracy and speed. On a food line you optimise for cleanability first, then for the ability to open the belt for cleaning or replacement, and only then for torque capacity. That reordering eliminates a surprising number of otherwise sensible options.

The metric T-profile family, defined in DIN 7721 for T2.5, T5, T10 and T20, remains the default in European food machinery. Trapezoidal teeth are simple to mould, forgiving of small misalignment, and available in open-ended form so a maintenance team can fit a belt around a closed frame without dismantling the line. The AT variants use the same pitch with a reduced backlash geometry, which matters when you are indexing a product into a pocket and a few tenths of a millimetre of tooth play shows up as a jam. For higher torque in the same envelope, the curvilinear profiles grouped in ISO 13050, together with the widely used 5M, 8M and 14M pitches, carry more load per millimetre of width because the tooth flank is a curve rather than a straight line. The trade is alignment tolerance: curvilinear teeth need better pulley alignment and they trap product in the root more readily than a shallow trapezoidal form.

Width is the variable people get wrong. A narrow 16 mm belt running at high tension can carry the same torque as a 32 mm belt running loose, and the narrow belt will always be easier to clean because there is less surface and fewer edges. But narrow belts also track more sensitively and load pulley bearings harder. On a 1.4 m wide dough transfer I reviewed, the fix was not more width; it was two 32 mm belts on a common shaft driving separate belt paths, which kept the cleanable span short and removed the cross-belt drift entirely.

Minimum pulley teeth matter more than pitch

The single most common mechanical error in food timing drives is an undersized small pulley. Each tooth engagement bends the cord around the tooth root, so a small tooth count concentrates bending stress and shortens flex life. As a rule of thumb, do not go below 18 teeth on T5, 16 teeth on T10, and 24 teeth on 8M or AT10 in washdown service; the published minimums are usually optimistic because they assume dry, clean operation and a constant load. Where a compact layout forces a small pulley, expect to trade belt life for the space saving and record that decision so the next person knows why the belt is replaced every 18 months.

Back-bending is the other quiet killer. Many food conveyors use a small back-side idler to reverse the belt or to increase wrap angle, and every back-bend cycle fatigues the cord from the opposite direction. A belt that sees three back-bends per lap will not achieve the flex life of one that sees none, no matter what the catalogue says. Count the bends, add them to your duty cycle, and size the cord accordingly. If you are also maintaining friction drives on the same plant, our transmission belt manufacturer overview covers the flat and V-belt side of the same building, and the two selection processes should not be mixed.

One practical note on supply. A plant that standardises on two pitches across the site will hold fewer spares, train fitters faster and negotiate better, and the savings usually exceed whatever a third profile would have gained in mechanical efficiency. As a V-belt manufacturer we keep both sides in stock, but the belt selection logic should be settled by the plant, not by the catalogue. When a drive needs on-site attention rather than a different pitch, our service team can survey the layout before anything is ordered.

04Backing and Cover Choices: Hygiene, Release and Grip

On a toothed belt, the tooth side is fixed by the pulley. Everything you can still choose sits on the back. That is where hygiene, release behaviour, grip and cleanability are actually decided, and it is the part of the specification food plants under-specify most often. Two belts with identical pitch, width and cord can behave completely differently on the same line purely because of a coating a few tenths of a millimetre thick.

Bare polyurethane is the starting point, and for many transfer positions it is the right answer. It is tough, it tolerates mild chemistry, and a smooth finish releases dough and formed product well enough if the surface is not scratched. Where product is wet and sticky, a lightly textured or fine-ribbed surface helps because it lets liquid escape from under the product instead of building a film that lifts the item off the belt at the transfer point. The texture has to be shallow, though. Deep patterns trap protein and starch, and every trap point becomes a cleaning argument you lose at audit.

For anything that needs release without sticking, a PTFE-impregnated or silicone-coated surface is the common answer, and both carry food-contact routes when the compounder documents them under 21 CFR 177.1550 for the PTFE side. Silicone deserves respect and caution in equal measure. It handles heat well and releases beautifully, but it also transfers trace silicone to product and to downstream surfaces, which is a real problem if the next process is a seal, a print or a coating. I have watched a sauce line spend two weeks chasing adhesion failures at a lidding station because trace silicone had migrated from a belt two machines upstream and nobody had connected the two events.

Fabric-backed and grip-top variants look similar in a photograph and behave nothing alike. A woven grip top with a PVC or PU coating gives traction on inclines and handles cartons, trays and pouches, and its textured surface is easy to rinse. The cost is that woven constructions hold moisture in the interstices, so drying time after washdown stretches and the belt can harbour residue at the cut edges unless the edges are sealed properly. That is the argument for specifying edge condition explicitly, not leaving it to the cutting operation.

Where a line climbs, the choice between a grip-top timing belt and a different construction often comes down to whether the product is loose or in a container. Loose product on an incline needs a positive feature, which usually means cleats, sidewalls or a separate cleated belt, and we cover that design space on the sidewall conveyor belt page rather than repeating it here. Containers behave differently. They only need friction, and friction is easier to clean than a cleat because there is no crevice. When someone insists on cleats for boxed product moving up a 15 degree incline, I ask them what the cleats are for, and the answer is usually "the last belt had them", which is not a reason.

There is a fourth option that buyers rarely write down and always regret: an unspecified backing. If the purchase order says "food grade PU belt, 10 mm pitch, 40 mm wide" and nothing else, you will get whatever surface the factory had running that day. Put the finish, the hardness and the pattern depth on the drawing, and put the acceptance criterion in words a goods-in inspector can apply with a durometer and a visual check.

Release behaviour also sets how much chemical the line consumes. A surface that lets product slide clean at the transfer needs a shorter foam contact time than one that has to be soaked. I have measured a 25 % reduction in foam usage on a confectionery transfer simply by changing from a smooth bare TPU to a shallow ribbed back, and the belt paid for itself within a year on chemistry alone, before counting the extra production minutes recovered each night.

05Cord Selection and Wet-Environment Construction

Inside the belt, the cord carries the load and the polymer holds the shape. In a food plant the polymer gets all the attention and the cord gets almost none, which is backwards. A belt whose cover is perfectly compliant will still fail inside eighteen months if the cord loses strength in humid air, and the failure looks like a mysterious tension loss rather than a broken belt.

Steel cord gives the highest tensile capacity per millimetre of width and the lowest elongation, which is why it turns up on wide food conveyors where product positioning matters. It also brings two risks into a food plant. Bare steel cord corrodes, so the cord has to be brass-plated and fully encapsulated with no exposed ends, and any moisture path through the polymer at a splice becomes a corrosion site. Aramid cord is lighter, has no corrosion mechanism at all, and has excellent strength per unit mass; its weakness is that it is brittle in compression and it does not tolerate small pulleys. I measured a 25 % reduction in flex life on a 40 mm wide aramid belt after the design team replaced a 24-tooth pulley with a 16-tooth one to gain 60 mm of frame space, and the saving was paid back in belt changes within a year.

Polyester and glass cord occupy the middle ground and are usually the sensible default for light and medium food transfers. Polyester has a modest amount of stretch, which absorbs shock loads and takes some pressure off the bearing selection, but that stretch is also why tension has to be re-checked after the first week of operation and again after the first month. Glass cord sits closer to steel in stiffness and costs less, though it is unforgiving of sharp bends and needs protection at the tooth root.

Construction details finish the job. A fully moulded belt with no cut edges presents less risk of fibre shedding than a slit belt, and a truly endless moulded belt removes the splice as a hygiene feature entirely, at the cost of a longer lead time and a more invasive installation. Carcass fabric choice matters too. A hydrolysis-stabilised treatment on a polyester fabric will outlast an untreated one by a wide margin in a room that is wet for six hours a day.

Wide food belts with steel cord follow the same principles as heavy bulk conveyors, and the encapsulation and splice discipline described for steel cord conveyor belt construction applies here with tighter hygiene constraints on top. Where the food process also throws abrasive material into the mix, as happens with root vegetable washing and grain handling, cover wear becomes the leading failure mode instead of cord fatigue, and the compound guidance on our abrasion resistant conveyor belt page is worth reading alongside this one.

If your plant runs both toothed food drives and heavier rubber conveying in the same building, the material standards should still be consistent across both. Our industrial conveyor belt range and the rubber conveyor belt family follow the same compound documentation route, which makes a single audit pack possible instead of three. And if you ever need to know who actually builds the belt rather than who sells it, our own conveyor belt factory details are published for exactly that reason.

One last construction point that saves arguments later. Ask for the belt marking standard in writing before you order, and confirm what is moulded or printed on the back of the belt: compound code, pitch, batch and the food-contact reference. When two plants in the same group receive belts from different batches, the marking is what lets quality compare them without cutting a sample. Without it, a substitution is almost impossible to detect until a failure forces an investigation.

Spares policy completes the picture. Because food belts are often made to length, a replacement can take several weeks, and a plant that holds no spare runs a line at risk for a month. Record the exact belt length, width, pitch, tooth count, backing and splice type on the line card, together with the original supplier and the date of the last change. When you need the same belt again, that card is what stops a substitution, and for plants that buy across several sites the difference is visible in both price and consistency. If you buy in volume, our wholesale conveyor belts programme is built around that repeatability, and as a conveyor belt supplier we keep the full specification on file so the next order matches the last one. Plants that prefer to buy locally usually work through a conveyor belt distributor, and the same discipline applies: the card travels with the order.

06Cleaning, Tensioning and the Sanitation Window

Two maintenance tasks decide whether a food-line timing drive reaches its rated life, and neither of them is complicated. One is tension. The other is cleaning, and what makes cleaning difficult is that the drive is usually assembled into a frame that was never designed for the person who has to reach into it.

Start with tension. A toothed belt does not need the tightness a friction drive needs, which is precisely why crews over-tighten it. The correct target is enough tension to keep teeth fully engaged under peak torque with a small margin, and no more. Too little and the belt jumps teeth during an acceleration spike, wearing the tooth flanks a little on every jump. Too much and you load the pulley bearings, the shaft and the belt cord, and you get the characteristic whine that tells an experienced fitter the drive is unhappy before any failure appears. Set tension with a sonic meter where the span allows it, using the mass per unit length from the belt data sheet, and record the frequency reading in hertz on the maintenance sheet. A number on a sheet is repeatable. "Feels about right" is not.

Re-tensioning intervals should follow the environment rather than the calendar. A belt in a wet room relaxes faster, because moisture uptake changes the cord and fabric dimensions. Check at the first shift after installation, again after one week, then monthly for the first quarter, and after that align the checks with your planned maintenance window. On the dairy line I mentioned earlier, moving from quarterly to monthly checks in the first ninety days eliminated an entire year's worth of tooth failures without changing a single component.

Cleaning that respects both hygiene and the belt

Detergent chemistry has to be matched to the polymer, not to the dirt. A 2 % caustic at 70 °C is routine in many dairies and is tolerable for a hydrolysis-stabilised ether-based TPU. The same concentration at 90 °C, applied every night for two years, will shorten cover life markedly, and the first evidence is fine crazing on the back side that a torch reveals under raking light. Acidic descalers are the other common exposure, and alternating acid and alkaline programmes accelerate surface ageing even when each chemical is individually acceptable.

Pressure is the second variable. A 60 bar jet held close to a belt edges lifts compound at the tooth root and drives water into the splice, which is the mechanism behind most early splice failures I examine. Keep wash nozzles at 20 to 30 bar, keep the lance moving, and keep the temperature below the polymer's continuous rating. None of that is exotic advice, but it is ignored in plants where the sanitation crew is measured on the time the wash takes rather than on the condition of the equipment afterwards.

Drainage and drying matter as much as the wash itself. A belt that runs wet through the night grows biofilm in the tooth roots by week three, and no amount of antimicrobial additive compensates. Tilt guards, drill drain holes at the low points of the frame, and run the belt dry for ten minutes after sanitation before stopping for the night. The cost is ten minutes of energy. The saving is a cleaning problem you never create.

Hardware in the same enclosure deserves attention too. A worn conveyor pulley groove will ruin a new belt within weeks, and belts that drift usually trace back to guide or tracking hardware rather than to the belt itself, so a training idler or a worn bracket is worth checking before you blame the belt. Corroded conveyor brackets are equally common in wet rooms, and a bracket that has lost a millimetre of section will let the whole drive move under load.

07Tracking, Changeover and Drive Architecture on a Food Line

A food conveyor changes over more often than a production machine, and every changeover disturbs the drive. Belt speed changes, product weight changes, sometimes the direction of travel changes because a line runs both ways depending on the recipe. Each of those transitions loads the belt edge differently, and the accumulation of small asymmetric loads is what finally moves a belt off centre.

Crowned pulleys are the usual cure for tracking on flat belts and the usual cause of problems on toothed belts, because a tooth pattern does not ride a crown the way a flat band does. Set the drive and tail pulleys parallel within 0.2 mm across the face, use flanges only on the slack side, and give the belt a proper guide rather than relying on flange contact. Flange contact is where tooth corners disappear. I have measured 3 mm of flange wear on a 12 mm tall tooth in a plant that had fitted flanges on both sides of both pulleys "for safety"; removing one pair and re-aligning the frame restored tooth life to the expected 30 months.

Where product needs to move between belts, the transfer point is a design decision with hygiene consequences. A small gap between two toothed belts creates a knife-edge situation that traps product, and the nose bar that fixes it introduces another surface to clean. Many plants solve this by combining a toothed belt for the driven section with a low-friction slider bed for the transfer, and the components that make that work sit in the wider hardware range described under conveyor components.

Speed, torque and the servo habit

Food lines increasingly use servo drives, and servos are less forgiving of a loose belt than a fixed-speed motor. A quick index with a 0.15 s acceleration ramp puts the whole inertia of the belt and product into the tooth flanks, and a belt with 1.5 mm of play will jump teeth under that load long before it would fail on a constant-speed drive. When the machine builder specifies a servo, the belt specification should come with the acceleration profile in hand, not just the steady-state torque figure.

Do not forget what the belt is carrying and where it goes next. A toothed belt feeding a filling head is part of the filling machine, and machine builders often buy the belt with the rest of the hardware, which means the specification is inherited rather than chosen. Where a plant owns a mixed fleet of packaging and conveying equipment, the practical answer is a short approved-parts list, one backup belt per critical position, and a rule that no substitute goes on the line without a written comparison. We see that discipline most often on lines feeding food packaging equipment, where a changeover gap of one shift already costs more than the belt.

Cold stores and high-care areas add their own version of the same problem. A belt moved from a 4 °C chiller into a 25 °C packing hall changes dimensions slightly as it warms, and a drive tensioned in the cold is loose in the warm. Tension cold, verify warm, and accept the compromise deliberately rather than discovering it as a drift. Plants that move pallets and totes between these areas, including the ones described under logistics warehousing, are usually running several belt technologies side by side, and the same tensioning discipline applies to all of them.

Finally, plan the end of the belt's life before it arrives. A timing belt on a food line rarely fails without warning; it fails after a long period of small signals that nobody recorded. Tooth wear, tension drift, splice lift and edge fray all appear months before a stoppage. If the changeover team writes down what they see once a month, the replacement gets planned into a shutdown instead of landing on a Friday night, which is the difference between an hour of maintenance and a lost shift of production. Retired belts with a usable cover still serve useful purposes in secondary areas, and our note on recycling covers the options for handling belt material after it leaves the line.

08Grade Matrix and Ordering Checklist for Food-Line Belts

By this point the specification has more parts than a purchase order usually carries, and the failure mode is always the same. The buyer describes the application in prose, the supplier quotes a belt that fits the prose, and the gap between the two only becomes visible on the line. Writing the requirement as a checklist that maps each line condition to a specification item, an evidence item and an acceptance test removes most of that gap.

The columns matter more than the rows. For every requirement on the left there should be a specification clause, a document to request, a known substitution risk and something the goods-in inspector can actually check with a tool. If a row has no acceptance test, that requirement will be unverified, and unverified requirements are where recalls come from. Coverage across temperature, chemical exposure, abrasion and impact all interact; a heat-tolerant compound and an abrasion-resistant cover are not the same purchase, and the notes on our heat resistant conveyor belt and impact and cut resistant grades explain the difference in cover formulation.

Two rows deserve a warning. Oils and fats swell many polymers, so a belt that passes every hygiene test can still fail mechanically in a bakery or a snack line because the compound swelled in service; the compound choices on our oil resistant conveyor belt page address exactly that. And where product is washed with heavy water volumes, the drive end of the conveyor collects everything the belt sheds, which is why the condition of return conveyor rollers in the same frame belongs on the checklist too, even when they are not part of the belt order.

Hand that checklist to a supplier and the conversation changes character. Instead of a price and a lead time you get questions about the detector aperture, the detergent list and the smallest pulley in the drive, and those questions are the first sign that the belt will fit the line rather than the quotation.

Request a food-line belt specification review

09Frequently Asked Questions

Which tooth profile should we choose for a washdown product line?

On most European food machinery the default answer is a metric T profile, T10 for the main transfer positions and T5 for light indexing, because trapezoidal teeth rinse more cleanly than curved flanks and the belts are available open-ended for fitting around a closed frame. Choose an AT variant when the drive has to index product into a pocket with very little play, since the reduced backlash geometry removes the tooth-to-groove movement that shows up as a jam. Move to a curvilinear profile such as 8M when the same belt width has to carry noticeably more torque, and accept the extra pre-soak time that the curved tooth root demands. Whatever you choose, keep the small pulley at 16 teeth or more on T10 and 24 teeth or more on 8M if the room is wet every night.

Are antibacterial additives actually necessary, or is cleaning enough?

Cleaning is what removes soil, and no additive changes that. An antimicrobial compound in the cover helps against surface colonisation between cleaning cycles, and it is usually justified in high-care areas where residue never fully dries, but the activity falls off as the surface abrades, so a worn belt with an antimicrobial claim may offer very little. Ask for the ISO 22196 data. Read the organisms and the reduction figure, not the headline claim. Then set the additive cost against the cleaning cost in your own zone. Where the choice is between an antimicrobial compound and a validated procedure with proper drainage and a ten-minute dry run at the end of sanitation, the procedure wins.

How do we know a detectable grade will trigger our detector?

You do not know until you test it at your own aperture. Detection tracks mass, geometry and orientation, so a 2.0 mm ferrous aperture may ignore a 1.5 mm chip and still catch a 4 mm fragment reliably. Ask what filler is used, at what loading, and what the smallest validated fragment size is. Then cut fragments of that size from the delivered belt and pass them through the line at production speed. Keep the result with the belt record. A detectable grade with no validated fragment size is an intention, not a control.

Why do these belts fail at the splice even though the cord is intact?

In wet food rooms the splice fails long before the cord does. The usual cause is water driven into the joint. High-pressure lances held close to the belt lift compound at the seam and open a path for moisture; once water reaches the cord, corrosion or swelling follows and the joint lifts. Reduce nozzle pressure to 20 to 30 bar, keep the lance moving, and check the splice monthly for any gap over about half a millimetre or for exposed cord. Where the lead time allows it, a moulded endless belt removes the joint as a hygiene feature altogether and often pays back through fewer stoppages.

What tension setting stops a belt from jumping teeth after sanitation?

Set tension with a sonic meter on the longest free span, using the mass per unit length from the belt data sheet, and record the reading in hertz on the maintenance sheet. Frequency, not feel, is the repeatable measurement. Moisture uptake changes belt dimensions, so re-check on the first shift after fitting, again after a week, then monthly through the first quarter. After that, align the checks with planned maintenance. On wet lines that sequence removes the entire class of tooth failures that follows a wash cycle. If the belt still jumps teeth at the correct tension, look for an acceleration spike, an undersized small pulley or flange contact at the drive.

What documentation should we hold before approving a food-grade belt?

Four documents cover most audits, and they rarely arrive as a tidy set. First a declaration of compliance that names the food-contact regulation together with the migration or extraction test conditions. Then a compound data sheet, which has to give continuous and peak temperature, hardness and chemical compatibility against the detergents you actually use. Detectable grades need one more item, the detector validation statement, and that statement is worthless without the fragment size that was tested. The dimensional record for the order belongs with the line card, so pitch, length, width and tooth count are ready when the next replacement is specified. Where the belt is part of a wider quality system, the discipline described on our quality assurance page is the same discipline that keeps these four documents current rather than filed once and forgotten.