Poor Cake Release: Causes and Cloth-Side Fixes
The cake does not drop cleanly at the discharge roller, carry-over comes back up the return run, and the scraper is doing work it was never meant to do. Before you order a new belt, find out which of three places the fault is coming from.
Poor cake release on a horizontal vacuum belt filter — the cake riding round the discharge roller instead of dropping — can start in the cloth, in the machine, or in the process: a hairy or filled cloth face is cloth-side; an oversized, scaled or worn discharge roller and a mis-set doctor blade are machine-side; a thin, wet cake is process-side. Scaling slurries such as the gypsum duties are where it bites hardest, because solids plate onto every surface they touch. The same symptom comes from all three places, which is why replacing the cloth on its own can leave the fault in place.
Does this match what you are seeing?
- Cake rides round the discharge roller instead of dropping, and comes back up the return run still on the cloth.
- Discharge is patchy — clean in places, a continuous skin in others.
- The doctor blade or scraper is doing all the work, and you can hear it.
- Solids are building up under the machine, in the drip pan, or in the cloth wash box.
- The cake looks wetter and more plastic than it used to at the end of the dry zone.
- Release got worse gradually rather than overnight, and filtration rate dropped with it.
If that is not quite it
Filtration rate fell first and the cake got wetter as a consequence Clogging / Blinding
Cake releases fine, but vacuum drops as the joint passes the box Joint / Seam Failure
Carry-over is worse on one side of the belt than the other Belt Tracking
What actually makes a cake let go
Discharge is a bending problem before it is a chemistry problem. The cloth leaves the vacuum zone flat and then turns through a tight radius at the discharge roller. The cake cannot follow that radius — it has almost no strength in bending compared with the fabric carrying it — so it fractures at the turn and falls away under its own weight. Everything that improves release works on one side of that balance: it either increases the separation the turn forces on the interface, or it reduces what is holding the cake to the cloth.
On the holding side there are three things at once. There is adhesion between the solids and the fabric surface. There is liquid: residual moisture between cake and cloth forms bridges that pull the two together, which is why a wetter cake sticks and a drier one lets go. And there is mechanical keying — particles sitting down inside the surface texture of the fabric, so the cake is not resting on the cloth but hooked into it.
That is the whole reason surface finish matters. A smooth monofilament face, a weave with long flat floats, a calendered surface: each of them reduces the texture available for keying, and each of them also closes the fabric down, which costs permeability. There is no version of this where you get a smoother cake-side face for free. The right answer is whichever side of that trade your circuit can afford to give up, and that is a decision to make against your filtration rate and your moisture target — not one to take off a web page.
Cake thickness and moisture belong in the same picture. A thicker cake is stiffer and heavier, so it fractures at the turn and drops; a thin cake has neither the weight nor the stiffness, and it smears round instead. A drier cake has shrunk and cracked away from the cloth, so there is less contact area and fewer liquid bridges left to break. This is why a dewatering fault presents as a release fault: nothing about the cloth has changed, but the cake now arrives at the roller wet and plastic.
And what carry-over does next
Cake that does not discharge does not disappear — it goes round the return run still attached, and it causes a second set of problems on the way. It is pressed back into the fabric at every return roller and again under tension over the vacuum box, which drives solids into the weave and blinds the cloth. It arrives at the cloth wash box as a solids load the box was never sized to remove, so wash water demand climbs and the nozzles plug — and a wash box that has stopped cleaning the cloth is the start of the next blinding cycle. On a rubber-belt machine, carried-over solids work in between the cloth and the transporter belt, where they pack into a ridge: uneven support over the vacuum box, a leaking edge seal, and a belt that will not run straight.
That is the loop worth breaking early: poor release feeds blinding, blinding makes the cake wetter, and a wetter cake releases worse. Carry-over is also product going back round the circuit instead of out of the plant — it matters most on duties where the wash stage is recovering something, as on phosphogypsum filtration. Where it shows up as drift rather than as carry-over, start at belt tracking.
Cloth, machine, process
Three columns, kept separate on purpose. Two of them are not ours to fix, and pretending otherwise would cost you a belt.
- Surface texture gives the solids something to key into: hairy or fibrous yarn surfaces present far more gripping area than smooth monofilament, which is why staple and multifilament constructions generally release less readily than monofilament in the same duty.
- The weave leaves an open, knuckled surface. Long floats in a satin or twill construction present a flatter face to the cake than a plain weave does, and a flatter face means less mechanical keying.
- The fabric surface has been abraded or fibrillated in service, so a cloth that released cleanly when new has grown a nap that holds the cake.
- Solids embedded in the surface pores from earlier blinding act as anchor points — the cake is then sticking to the previous cake, not to the cloth.
- The cloth is stiff or deformed in a way that stops it following the roller radius tightly, so the cake never gets the sharp bend that breaks it off.
- Discharge roller diameter is too large for the duty, or a nose bar has been replaced with a larger roll. The release mechanism is bending: the tighter the turn, the more the cake is forced to separate from a surface it cannot follow.
- The discharge roller is scaled, coated or worn out of round, so the effective radius is bigger and less even than the drawing says.
- Blade or scraper geometry is wrong — angle, contact pressure, or a blunt edge that burnishes the cake into the fabric instead of lifting it off. Too little contact leaves a skin; too much drives solids into the weave and wears the cloth.
- Belt speed is set faster than the dewatering zone can keep up with, so the cake reaches the roller wetter and more plastic than it should be.
- Vacuum is being lost before the discharge end — box seal, edge seal, a leaking joint — and the dry zone is not doing its job.
A new cloth will not fix these. It will look better for a while, and then the fault comes back.
- Cake is too thin. A thin cake has little weight and little stiffness of its own, so it smears round the roller instead of fracturing and falling; thin-cake, high-speed duties are the hard release cases.
- Residual moisture is high. A wet cake holds itself to the cloth through the liquid between them; a drier cake shrinks, cracks and loses contact area, which is most of why dewatering and release rise and fall together.
- Feed solids, particle size or crystal habit have drifted, changing how plastic the cake is at the discharge end.
- Wash water is landing too close to the discharge end and re-wetting a cake that had already been dried.
- Scaling duties — gypsum in particular — plate solids onto every surface they touch, including the cloth face and the discharge roller.
Correct these first, or the new belt inherits the fault.
Sticking is a live concern on scaling gypsum duties, where solids plate onto the cloth face and the discharge roller alike: phosphogypsum filtration (acid plant, fluoride-bearing) and FGD gypsum dewatering (wet-lime scrubber, chloride-bearing). They are separate duties with separate cloth decisions — see the two side by side.
Work through it in order
Every step says what to look at, what to measure and what to photograph. Do it in this order and you finish holding everything an engineering quote needs.
- 00
Stop and isolate first
- Look at
- Machine stopped, isolated and tagged out before anyone goes near the discharge end, a roll, or a blade.
- 01
Decide whether the cake is wet or stuck
- Look at
- Whether the cake arriving at the discharge roller is dry and cracked, or wet and plastic. These point at two different faults — a dewatering problem and an adhesion problem are not the same job.
- Measure
- Cake thickness at the end of the dry zone, left, centre and right.
- Photograph
- The cake surface just before the roller, across the full width.
- 02
Watch the discharge running, from the side
- Look at
- Where on the turn the cake actually lets go — immediately at the tangent point, part way round, or not at all.
- Photograph
- A side-on shot of the discharge roller with the machine running, if you can take one safely.
- 03
Measure the discharge roller
- Look at
- Scale, coating, flats, wear, and whether the surface is the same all the way across.
- Measure
- Roller diameter, and the diameter of any nose bar if one is fitted. Note what the machine drawing says it should be.
- Photograph
- The roller face and, separately, any build-up on it — before you clean it off.
- 04
Check the blade or scraper
- Look at
- Edge condition, contact across the width, whether it is polishing the cake rather than lifting it.
- Measure
- Blade angle and the gap or contact at left, centre and right.
- Photograph
- The blade edge close up, and the line it leaves on the cloth.
- 05
Look at the cloth face where carry-over sits
- Look at
- Whether the fabric surface is still smooth, or has gone fluffy, glazed, or filled with embedded solids.
- Measure
- Whether the affected area is the whole width or a band — a band points at feed or wash distribution, not at the cloth.
- Photograph
- Close-up of the cloth face, and of the back, in the worst area and in a clean area for comparison.
- 06
Follow the return run
- Look at
- Where the carried-over cake ends up: on the return rollers, on the rubber transporter, in the drip pan, or on the floor.
- Photograph
- The return run as a whole, plus any ridge of packed solids you find between the cloth and the carrier belt.
- 07
Open the cloth wash box
- Look at
- Blocked nozzles, spray pattern gaps, and how much solid the box is being asked to handle.
- Measure
- Wash water pressure and flow, as your plant normally reads them.
- Photograph
- The spray bar and the nozzle faces.
- 08
Check the dewatering side of the ledger
- Look at
- Vacuum level along the dry zone, box seal and edge seal condition, and the state of the seam as it passes the box.
- Measure
- Vacuum at the dry zone, against what the machine is supposed to hold.
- Photograph
- The seam, both faces — the same shot the engineering RFQ asks for.
- 09
Record the pattern
- Look at
- Gradual or sudden? Worse at one speed, one feed rate, one product grade? Better straight after a cloth wash?
- Photograph
- Nothing to photograph — write it down. It goes straight into the failure description on the RFQ.
- 10
Copy the nameplate
- Look at
- Machine brand, model and serial number.
- Photograph
- The nameplate, lit from an angle so the flash does not wash it out.
Not sure how to measure the belt itself? See the measuring guide.
What can be changed in the cloth
Four variables, and every one of them trades release against something else. That trade is the conversation; there is no construction that is simply better.
Yarn type on the cake face
Smooth monofilament on the face presents less gripping surface than a hairy staple or multifilament yarn. It is the single biggest cloth-side lever on release, and it trades against filtrate clarity and cake retention.
Weave construction
Satin and twill constructions give longer floats and a flatter cake-side face than a plain weave, so there is less for the cake to key into. The same change alters flow and particle retention, which is why it is a duty decision and not a preference.
Calendering and surface finish
Calendering flattens the yarn crowns and closes the surface, which usually helps release and usually costs permeability. It is a deliberate trade between a cake that leaves cleanly and a cloth that filters fast, and it has to be made against your circuit.
Built to your measured dimensions
Width and length taken from your old belt and your machine, so the cloth sits and turns the way the machine expects. A belt that is fighting the geometry does not release consistently either.
These cases are not solved by new cloth
A discharge roller that is oversized, scaled or worn out of round; a blunt or mis-set blade; vacuum lost before the dry zone; a cake made thin by a speed or feed change. Those are machine and process problems. A new belt will at most mask them for a while, and then the fault comes back with a belt you have already paid for. Fix the machine first, then change the cloth.
Tell us your machine and what you are seeing, and we will say honestly whether we can help.
Which of these we can actually supply
Our own confirmed manufacturing range is not published yet. Rather than print a number we cannot stand behind, the table below is a to-be-confirmed list. Send your case and we will come back with what the factory has confirmed in writing for your dimensions and your duty.
| Cloth-side option | What it addresses | Confirmed range |
|---|---|---|
| Monofilament face constructions available | Surface adhesion at the discharge roller | UNKNOWN — pending factory confirmation |
| Weave types available (plain / twill / satin) | How much surface the cake can key into | UNKNOWN — pending factory confirmation |
| Calendering and surface finishing options | Release versus permeability trade-off | UNKNOWN — pending factory confirmation |
| Permeability range per construction | Whether the release change costs you filtration rate | UNKNOWN — pending factory confirmation |
| Anti-stick or low-friction finishes | Release on sticky and scaling duties | UNKNOWN — pending factory confirmation |
| Maximum belt width and length | Whether your machine is inside our range at all | UNKNOWN — pending factory confirmation |
Process detail on how belts are built: factory and quality control. What the cloth family itself offers: HVBF filter cloth.
Replace it, or fix the filter first?
- The face has gone fluffy or fibrillated from abrasion and no longer presents a smooth surface.
- Solids are embedded in the surface pores and do not come out with a wash — the cake is sticking to the cake.
- Release is poor across the full width on a machine whose roller, blade and vacuum all check out.
- The construction is simply wrong for the duty — a hairy face on a sticky cake was never going to discharge cleanly.
- The discharge roller is scaled or coated. Clean it and look again before you order anything.
- The blade is blunt or mis-set. That is an afternoon, not a new belt.
- The cake is wet because vacuum is being lost. Find the leak; release will follow the dewatering.
- Cake is thin because belt speed or feed rate has been changed. Put the cake thickness back first.
- Carry-over is confined to a band across the width. That is a feed or wash distribution problem, not a cloth problem.
The quickest test on this list is cleaning the discharge roller and sharpening the blade. Do that before you price a belt. If the cloth body is sound and the machine cause is clear, do not change the belt — send us the photos and we will help you decide. We do not publish expected service life in months, because it depends entirely on your duty and your machine.
Decided it needs replacing? Next step is measuring the old belt →
The same facts, whether you ask us to diagnose or to quote
Tick these off as you collect them. Nothing here is stored or sent from this page — it is a scratch list for your own use.
Where cake release problems tend to come from
Phosphogypsum Filtration
WPPA acid-plant gypsum on horizontal vacuum belt filters, where wash recovery and release pull the cloth specification in opposite directions.
FGD Gypsum Dewatering
Wet-lime scrubber gypsum — chloride-bearing, a separate duty from phosphogypsum and not the same belt.
Horizontal Vacuum Belt Filter Cloth
Endless belts built to your machine dimensions, with the construction and finish options that bear on release.
Vacuum Belt Dewatering Filter Cloth
The cloth page for the dewatering duty — the last vacuum zone, where the cake has already been formed and washed and discharge is what is left to get right.
Common questions about cake release
Will a smoother cloth fix my cake release problem?
It can, if the cake is genuinely sticking to the fabric surface. It will not, if the cake is arriving wet because vacuum is down, or riding round because the discharge roller is too large or coated. And a smoother, more closed surface normally costs you permeability, so it is a trade to make against your circuit rather than an upgrade to order. Work through the checks first.
Why did release get worse when the cake got thinner?
A cake discharges because it cannot bend round the roller radius, so it fractures and falls under its own weight. A thin cake has little weight and little stiffness, so instead of breaking away it smears and stays on the cloth. If belt speed or feed rate changed shortly before the problem started, that is the first thing to look at.
Does poor cake release cause blinding, or does blinding cause poor release?
Both directions are real, which is why these two problems usually arrive together. A blinded cloth filters more slowly, so the cake reaches the discharge end wetter and sticks; the carry-over is then pressed back into the fabric on the return run and blinds it further. Deciding which came first is a matter of sequence — what fell first, filtration rate or discharge quality.
How long should a belt run before release gets this bad?
That depends on the duty — slurry, chemistry, scaling tendency, machine condition and how the belt is run. We do not publish a service-life figure. Send your operating conditions and the old belt details and we will discuss what is realistic for your case.
These answers are engineering judgement, not measured results, and no FAQ structured data is published from this page until they have been verified by our technical staff.
Measure your old belt and send it to us
Width, length, seam type, cake thickness and moisture, and the photos of the cloth face, the discharge roller and the blade from the checks above. That is enough for an engineering answer instead of a guess. No account, no registration.
Replacement cloth fabricated to your machine dimensions. Not an OEM spare part.