Filter Cloth Blinding: Causes and Cloth-Side Fixes
Flow falls week after week, the cake comes off wetter, and the wash no longer brings the cloth back. Before you order a new belt, find out whether the medium is really the restriction.
Blinding is the progressive loss of open area in a filter medium: solids lodge in and bridge across the openings of the fabric, and dissolved species precipitate as scale inside it. On a horizontal vacuum belt filter it shows up as falling filtrate rate, wetter cake, and wash water that channels through the cake instead of displacing the mother liquor out of it. The medium is only one of the three places it can start — a dead wash header and a leaking vacuum box produce very similar numbers in the control room.
Does this match what you are seeing?
- Filtrate rate has fallen steadily over weeks, and it did not come back after the last belt wash.
- Cake comes off wetter than it used to, at the same belt speed and the same feed.
- The same wash rate recovers less than it did — wash water is finding a channel instead of displacing liquor through the cake.
- Held up to a light, the cloth no longer shows clear light points across parts of its width.
- The vacuum gauge sits as high as ever, or higher, while throughput is down.
- The cloth face feels stiff, boardy or gritty, or carries a hard deposit you can scratch at.
If that is not quite it
Throughput fell in a step, and the vacuum reading fell with it Joint / Seam Failure
The cloth wanders off-line and the seal leaks at the edge Belt Tracking
Flow is normal but the cake will not let go at the discharge roller Poor Cake Release
Two different things are called blinding
They look the same on the flow meter and they are treated completely differently. Knowing which one you have is most of the diagnosis.
Particles bridging the open area
Solids close to the size of the fabric openings do not pass and do not sit cleanly on the surface either. They wedge in the interstices between yarns and bridge across them, and once a bridge has formed, finer material builds on top of it. Solids held on the face are a surface layer that a wash can lift off; solids held inside the structure are depth blinding, and the wash never reaches them.
Scale growing inside the fabric
Where the liquor is close to saturation, a change in temperature or concentration across the machine drops solids out of solution — calcium sulphate is the classic case, and carbonate and hydroxide scales behave the same way. The crystals grow in place, keyed into the weave rather than resting on it. This is why a plant can double its wash pressure and see nothing change: pressure removes deposits that sit on a surface, and scale does not.
A medium that is closing up passes less air, so the vacuum level holds or climbs while filtrate flow falls. A leak does the opposite — air flow rises and the vacuum level drops. If throughput fell and the gauge fell with it, you are more likely looking at a seam or joint failure or a failed seal than at a blinded cloth.
Why washing gets worse before flow does
On a wash duty the wash liquid is supposed to move through the cake as a front, displacing the mother liquor ahead of it. That only works while the resistance is even. As parts of the medium close up, or as the cake cracks over them, the wash finds the lowest-resistance route and runs through it, leaving most of the cake untouched. Wash efficiency therefore falls at an unchanged wash rate — often the first sign, and on duties where the wash recovers product rather than just rinsing the cake, the expensive one.
The wash chemistry is duty-specific and the two gypsum duties do not share it: phosphogypsum filtration washes soluble P2O5 out of an acid-plant cake, while FGD gypsum dewatering washes chloride out of a scrubber cake. The two duties side by side →
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.
- The open structure of the medium sits in the same size band as the bulk of the fines, so particles bridge the openings instead of passing through or being held on the surface.
- Multifilament and staple yarns present far more internal surface for solids to lodge against than monofilament does. That is a selection trade-off against filtrate clarity, not a defect.
- The face of the fabric lets cake key into it rather than release from it, so every pass leaves a little more behind than the wash takes off.
- The fabric is already blinded through its depth, or its structure is damaged — no wash circuit recovers that.
- Wash header pressure low at the header, or falling along its length, so part of the width is barely being washed at all.
- Nozzles blocked, worn, mis-aimed, or spaced so the spray fans no longer overlap — worn orifices open the fan out and soften the impact long before they stop spraying.
- Belt wash running on the wrong face, running short of the full width, or shut off during production to save water.
- The vacuum system not holding: a leaking box gasket, a worn wear-belt seal or a flooded filtrate receiver all cut the driving force and look like a blinded cloth from the control room.
- Cake too thin, because belt speed or feed rate is off. A cake filters through itself; without one, the slurry works directly on the medium.
A new cloth will not fix these.
- Feed particle size distribution has drifted finer than the medium was selected for — a grinding or classification change upstream lands on the filter.
- Flocculant under- or over-dosed. Too little and the fines stay free to bridge the openings; too much and the polymer itself gels on the face.
- Scale-forming chemistry. Calcium sulphate, carbonate and hydroxide scales precipitate when temperature or concentration shifts across the machine, and scale grows inside the weave rather than resting on it — which is exactly why spray washing does not touch it.
- Hard wash water, or filtrate recycled as wash water, putting fines and scale back into the cloth about as fast as the wash takes them out.
Correct these first, or the new belt closes up the same way.
Blinding bites hardest where the duty depends on cake washing and where the liquor carries a scaling species. On the machines this site covers, that description fits gypsum-type duty on horizontal vacuum belt filters; the same mechanisms apply to recessed-plate and vertical pressure filters, where the medium is a pressure filter cloth rather than an endless belt.
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 — and, more usefully, knowing whether the cloth washes back.
- 00
Stop and isolate first
- Look at
- Machine stopped, isolated and tagged out before anyone goes near the cloth, a wash nozzle or the vacuum box.
- 01
Get the trend, not today’s number
- Look at
- Filtrate rate, cake moisture and vacuum level over weeks, not over one shift. Blinding comes on gradually; a leak or a pump fault arrives as a step change.
- Measure
- Filtrate rate and cake moisture now, against the same readings when the belt was new.
- Photograph
- Nothing to photograph — take the trend off the historian or the shift log and write it down.
- 02
Read the vacuum against the air flow
- Look at
- Which way the vacuum reading has moved while throughput fell. This one observation separates two opposite faults.
- Measure
- Vacuum level at the box, and the pump duty if it is instrumented.
- Photograph
- The gauge, with the reading legible.
- 03
Look at the cloth across the full width
- Look at
- Is the cloth closing uniformly, or in stripes, bands and patches? Stripes running in the direction of travel usually point at the wash — one nozzle washes its lane and the next one does not. Uniform closing points at the medium, the feed, or scale.
- Measure
- Where the closed areas sit relative to the nozzle spacing. If they line up, you have your answer.
- Photograph
- The belt face across the full width, taken square on, plus a close-up of a closed area and of a clean area on the same belt.
- 04
Check the wash header and every nozzle
- Look at
- Blocked and worn orifices, wrong standoff, fans that no longer overlap, and a header fed from one end that is starved at the far end.
- Measure
- Wash water pressure at the far end of the header as well as at the feed end, and flow if it is metered.
- Photograph
- The header running, from a safe position, so the spray pattern across the width is visible.
- 05
Check cake thickness and feed distribution
- Look at
- Whether there is a cake at all through the first zone, and whether it is the same across the width.
- Measure
- Cake thickness at left, centre and right at the end of the drainage zone.
- Photograph
- The cake surface across the full width.
- 06
Run it on water alone
- Look at
- With the slurry off and the belt washed, how freely clean water passes through the cloth compared with what it did when new.
- Measure
- Nothing precise is needed here — the relative rate is the answer.
- Photograph
- Nothing. This is a feel-and-see test.
- 07
Cut a coupon and try to wash it back
- Look at
- Take a small piece from scrap or from the belt at change-out. Rinse in clean water, then — strictly under your plant’s own chemical handling procedure — treat it with the descaling agent your site already uses. Permeability coming back with water alone means surface fines; coming back only with the chemical means scale; not coming back at all means the structure is blocked or damaged.
- Measure
- Hold the coupon to a light before and after. A sound fabric shows discrete light points.
- Photograph
- The coupon against a light source, before and after. This is the single most useful photograph you can send us.
- 08
Ask what changed upstream
- Look at
- Grind or classification changes, recirculating load, flocculant dose, a new wash water source, a temperature drop across the machine, filtrate being recycled as wash water.
- Photograph
- Nothing to photograph — write it down. It goes straight into the failure description on the RFQ.
- 09
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
Construction chosen against your particle size distribution
The opening structure is selected so the bulk of your solids either sit on the surface or pass, rather than landing in the size band that bridges across the openings and stays there.
Filament type and surface finish
A monofilament surface gives solids less to key into than multifilament or staple yarn. Calendering and other face treatments change how the cake sits and how much of it the wash takes off.
Washability treated as a selection criterion
A medium can filter well and wash badly. Where the duty depends on cake washing, that gets weighted when the cloth is specified instead of being discovered on the machine.
Built to your measured dimensions
Width, length, seam and edge taken from your old belt and your machine, so the cloth is not being forced to fit and then blamed for the consequences.
These cases are not solved by new cloth
A wash header that is not washing, worn or blocked nozzles, a vacuum system that cannot hold vacuum, and a feed whose particle size distribution has moved are machine and process problems. A new belt gives you a clean start and then closes up the same way, on a cloth you have already paid for. Fix the wash and the vacuum first, confirm what changed upstream, 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 duty and your dimensions.
| Cloth-side option | What it addresses | Confirmed range |
|---|---|---|
| Weave and open structures available | Bridging behaviour against your particle size distribution | UNKNOWN — pending factory confirmation |
| Polymers available | Chemical duty, temperature, and resistance to the scaling species | UNKNOWN — pending factory confirmation |
| Filament types (mono / multi / staple) | How much internal surface solids have to lodge in | UNKNOWN — pending factory confirmation |
| Surface finish and calendering options | Cake release and how much washes back off | UNKNOWN — pending factory confirmation |
| Air permeability grades | How open the medium starts out, before any duty | UNKNOWN — pending factory confirmation |
| Maximum fabricated width | Whether your machine is inside our range at all | UNKNOWN — pending factory confirmation |
Process detail on how belts are built: factory and quality control. The cloth family itself: horizontal vacuum belt filter cloth.
Does it wash back, or is it finished?
This is the whole decision, and step 07 answers it. Everything else is working out which fault put you here.
- Permeability does not come back after a clean water rinse and after a chemical clean — the structure itself is blocked.
- Hard scale that cannot be removed without damaging the yarns.
- Blinding together with loss of fabric strength, or with yarn damage visible under a glass.
- Blinding that returns quickly on a machine that has checked out clean, with the wash circuit working. A like-for-like replacement will do exactly the same thing.
- Permeability comes back once the wash header is unblocked or repaired.
- Permeability comes back with a chemical clean the plant can run on a set schedule — then the live question is the scaling chemistry, not the cloth.
- Throughput recovers once a vacuum leak is found. The cloth was never the restriction.
- The closed areas line up with dead nozzles. Fix the wash, run it, then reassess.
- Blinding started when something changed upstream, and clears when that change is reversed.
If the coupon washes back and the wash circuit is the fault, 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 how long a medium stays open depends entirely on your slurry, your wash 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 blinding bites hardest
Phosphogypsum Filtration
Acid-plant gypsum on horizontal vacuum belt filters, where the wash stage recovers soluble P2O5 — so wash channelling costs product yield, not just cake moisture.
FGD Gypsum Dewatering
Wet-lime scrubber gypsum: a chloride-bearing duty with its own wash requirement, and a separate belt from the acid-plant one.
Horizontal Vacuum Belt Filter Cloth
The cloth family this duty runs on, and the construction choices that decide how fast a medium closes up.
Vacuum Belt Dewatering Filter Cloth
The cloth page for the dewatering duty — the last vacuum zone, where a medium that has closed up reads as residual moisture in the cake rather than as a filtrate rate.
Questions about blinding
Is blinding a cloth problem or a process problem?
Usually both at once. The medium decides how easily solids bridge its openings and how much of what lodges there washes back out; the process decides what arrives at the cloth and what precipitates inside it. That is why a cloth that ran well for a long time can start blinding quickly with nothing changed on the filter — the feed changed. Work through the checks above before you conclude it is the fabric.
How do I tell blinding from a vacuum leak? Both cut my throughput.
Watch the vacuum reading, not the filtrate rate. A medium that is closing up passes less air, so the vacuum level tends to hold or climb while filtrate flow falls. A leak — an opening seam, a failed edge seal, a bad box gasket — does the opposite: air flow rises and the vacuum level drops. If your gauge fell, look at the seam and the seals first.
Can a blinded belt be cleaned and put back on?
Sometimes, and the coupon test in step 07 is what tells you. If permeability comes back with clean water, you had surface fines and a wash problem. If it comes back only after a chemical clean, you have scale, and the question becomes whether you can run that clean on a set schedule without damaging the fabric. If it comes back with neither, the fabric is finished and no cleaning schedule will change that.
Will a more open cloth solve it?
It may raise the flow and it may cost you filtrate clarity, cake release, or both — and if the fines are bridging because they sit in the same size band as the openings, a slightly more open weave can blind just as fast. That trade-off is decided against your particle size distribution and your wash configuration during the engineering review, not from a web page.
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, the photos from the checks above, and the answer to one question: does it wash back? 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.