TEL:0086-18803691565
Most terrazzo tile defects are not machine faults. They are process faults that appear hours or days after the tile leaves the press, which is exactly why they cost so much: by the time you see the crack, the white bloom or the colour band, the whole batch is already stacked on pallets and counted as finished goods.
This guide works backwards from the symptom. For each defect it covers what the problem looks like, what usually causes it, what to correct on the spot, and what to change in the process so it does not come back. It applies to cement-based terrazzo tiles made on a press, on either a fully automatic or a semi-automatic line.
Machine adjustments are easy to make and hard to reverse, so resist the urge to start there. Read the tile first. The table below maps the visible symptom to the cause that most often produces it, and to the one measurement that will confirm or rule it out.
| What you see | Most likely cause | Check this first |
|---|---|---|
| Fine map cracking across the face | Mix too wet, or the face dried before curing began | Water content — use the squeeze test below |
| One straight crack running from a corner | Demoulded before the tile had green strength | Time left in the mould against tile thickness |
| White powder bloom on the surface | Soluble salts migrating to the surface with water | Cement alkali level and where the tiles are stored |
| Colour drifts from pallet to pallet | Pigment dosed per mixer load instead of per batch, or poorly dispersed | Pigment scale accuracy and the mixing method |
| Face layer lifting off the base layer | Too long a gap between the face and the base going into the mould | The delay between the two filling steps |
| Chipped corners on most tiles | Worn or burred mould edge, or early demoulding | Mould edge condition under a light |
| Pinholes scattered across the face | Air trapped in a stiff mix, or vibration too short | Vibration time and the sand-to-fines balance |
| Tiles bow after a few days | Top and bottom faces curing at different rates | Whether tiles are drying uncovered in direct sun |
| Stones sitting proud or popping out | Dusty aggregate, or chips too coarse for the face thickness | Whether the aggregate is washed before use |
Two different failures produce two different crack patterns, and they are worth telling apart.
Map cracking looks like a fine net over the whole face, with no preferred direction. It is a shrinkage problem. The face loses moisture faster than the body behind it, so the surface tries to contract while the core holds its size, and the surface tears. The usual trigger is a mix that was made wetter than it should be — often because a batch looked dry and someone added water to make it flow.
A single straight crack, usually starting at a corner or the middle of one edge, is a strength problem at demoulding. The tile was lifted while it was still too weak to carry its own weight. Thin tiles and large formats are the most exposed, and cold weather makes it worse because strength develops more slowly.
What to do: neither crack can be repaired invisibly on a through-body tile. Sort those tiles out of the saleable stack immediately rather than hoping the grinder will hide them — grinding opens a crack wider instead of closing it.
What to change: stop judging mix consistency by eye. Use the squeeze test: take a handful of mix and close your fist. It should hold the shape of your fingers and leave the palm damp, not wet, with no water running between your fingers. If water runs, the batch is too wet and it should not go to the press. For the demoulding side, add time in the mould rather than subtracting it, and add more in cold weather.
Efflorescence is a salt deposit. Soluble salts present in the cement, the sand or the aggregate dissolve in the mixing water, travel to the surface as the tile dries, and crystallise there as a white or greyish powder. It is not a surface stain, so washing the face does nothing lasting — the salt is still inside the tile and will come back the next time it gets wet.
Four conditions make it much more likely: a high-alkali cement, aggregate that was never washed, a mix with more water than the cement can bind, and storage that lets the tile take on moisture from the ground or the air.
What to do: dry-brush the bloom off to assess how deep it goes, then stop the moisture path. Tiles sitting on a damp floor or stacked uncovered through a rainy season will keep blooming no matter what you apply to the face.
What to change: keep a single low-alkali cement source and do not switch suppliers mid-order without a trial batch; wash aggregate and let it drain before use; keep stored tiles on pallets, off the ground, under cover. Efflorescence that appears in the first week is usually a water problem. Efflorescence that appears after months in storage is a storage problem.
Colour is the defect customers reject most often and the one producers notice latest, because a single pallet looks fine on its own. The comparison only becomes obvious when two pallets sit side by side on site.
There are three independent causes and they are easy to confuse. The first is pigment dosing: if pigment is scooped per mixer load rather than weighed per batch, the variation between two consecutive batches can already be visible. Iron-oxide pigments are usually dosed in the range of two to five percent of the cement weight depending on the shade, and at those levels a small weighing error is a visible shade change.
The second is dispersion. Pigment tipped dry into the mixer in one lump will not distribute evenly no matter how long the mixer runs. Pre-mix the pigment into water to form a slurry and add the slurry to the batch.
The third cause surprises people: the cement itself. Two deliveries of cement from different clinker, or cement that has aged in a humid store, shift the base grey of the tile before any pigment is added. A shade that was correct last month can drift with no change at all in your own process.
What to change: weigh pigment on a scale that reads to the gram, not by volume; add it as a slurry; keep one cement source; and keep a master sample board stored indoors. Compare every batch against the board under the same light. Daylight from a window changes through the day, so compare under a fixed lamp or you will chase shadows.
Delamination shows as a face layer that lifts, blisters or peels away from the base layer, sometimes only after the tile has been ground. It is almost always a timing failure rather than a materials failure.
A pressed terrazzo tile is built in two stages: the decorative face mix goes in first, then the structural base mix, then the press closes. The two layers bond because they are compacted together while both are still fresh. If the face mix has already skinned over — its surface dried and closed — before the base arrives, there is nothing left for the base to grip, and the press simply compacts two separate slabs into one mould.
The second contributor is a face mix that is too dry or too lean. A face that will not compact into a continuous skin cannot bond to anything.
What to change: shorten the interval between the two filling steps, and keep the face mix covered between mixing and use so it does not dry at the surface. If your line runs slowly at shift start, do not fill a large batch of face mix and let it stand — fill what you will press within the next few minutes. A delaminated tile cannot be salvaged, so this is one defect where prevention is the whole answer.
Chipped corners are usually reported as a mould problem and are usually a handling problem. The three real causes, in order of how often they occur: demoulding too early, a worn or burred mould edge, and pallets that are not flat.
Run your fingertip along the inside edge of a mould that has been in service for a while. A burr you can feel will take a chip out of every tile that passes through it. Check the pallets the same way: a pallet with a high spot makes the tile rock on the stack, and a rocking green tile loses its corners before it loses anything else.
What to change: inspect mould edges on a fixed schedule rather than when a customer complains; replace or dress burred moulds; keep pallets flat and reject warped ones; and do not stack green tiles higher than the bottom layer can support. Stack height is a common shortcut that shows up later as edge damage on the bottom three layers.
Pinholes are small craters on the face, roughly the size of a pinhead, often clustered rather than evenly spread. They are trapped air. A stiff mix holds air pockets that vibration should have released, and a grading with too few fines leaves gaps that cannot close under pressure.
Two mistakes make it worse. The first is adding water to make the mix flow instead of correcting the grading — that trades pinholes for cracking. The second is over-vibrating to compensate, which drives the heavier aggregate down and the paste up, producing segregation and a face that is rich in cement at the top and starved of stone.
What to change: fix the sand-to-fines balance so the mix compacts without extra water, and set vibration time by observation of the tile as it densifies rather than by a number someone used on a different machine. If the stone is sinking, you are vibrating past the point of use.
A tile that is flat at demoulding and bowed a week later is telling you that its two faces cured at different rates. The face that dries first shrinks first, and the tile curls towards it. Uncovered drying in direct sun is the classic trigger: the top face can be several degrees hotter and much drier than the bottom.
Pallet condition contributes as well. A pallet that sags in the middle under a stack of tiles will transfer that curve into every tile above it, and the effect accumulates with stack height.
What to change: cure under cover, with both faces exposed to similar humidity, and keep curing stacks out of direct sun and away from draughts. Check pallets for flatness, not just for damage. If you must store finished tiles for months before shipping, store them flat, fully supported, and not leaning against a wall.
Pop-out is a chip of aggregate that separates from the face and leaves a crater. It has two causes and both are about bond rather than about the press. Aggregate that was never washed carries a coating of dust, and cement paste bonds to the dust instead of to the stone, so the bond fails at the first stress. Aggregate that is too coarse relative to the face thickness leaves too little paste around each stone to hold it.
There is a cosmetic version of the same problem: stones that end up sitting proud of the face after grinding, which reads as an uneven surface rather than a hole.
What to change: wash aggregate and let it drain, do not use it dry and dusty; size the coarse fraction to the face thickness you are actually pressing; and pre-wet aggregate so it does not pull water out of the paste at the interface.
A large share of "bad batch" complaints turn out to be an input that changed without anyone noticing. These are the four checks worth doing before you adjust the machine.
Most of the defects in this list are decided in the first few days, when the tile is weakest and most easily influenced. A tile that survives demoulding is not finished — it is at the most vulnerable point in its life.
The order that works is consistent rather than complicated: leave the tile in the mould long enough to hold its shape, move it to a covered curing area where it is not exposed to sun or draught, keep it flat and fully supported, and do not grind or polish it until it has developed enough strength to survive the operation. Grinding a tile too early produces edge chipping and a face that tears rather than cuts.
Storage deserves the same attention as production. Tiles stacked on damp ground, uncovered, or leaning at an angle will develop efflorescence, warping and edge damage that had nothing to do with how they were made. Every one of those defects will be blamed on the machine.
Defects are cheap to fix at the press and expensive to fix at the port. This routine takes a few minutes per shift and catches most of the list above before a batch is completed.
| When | Check | Why it matters |
|---|---|---|
| Start of shift | Mould edges, pallet flatness, mixer cleanliness | Chips and warping start here, not at the press |
| Every batch | Squeeze test on the mix, pigment weight on the scale | The two errors that cause most cracking and colour drift |
| Every batch | Colour against the master sample board under fixed light | Catches drift while the batch can still be blended |
| Once a day | Break one tile from the day's output and look at the fracture | Reveals delamination and weak bond that the surface hides |
| Once a week | Silt test on the sand, moisture check on cement storage | Catches an input that changed without anyone noticing |
| Once a week | Walk the curing and storage area after rain | Efflorescence and warping are created in storage, not production |
If you are still specifying a mix, the material side is covered in more detail in our notes on terrazzo tile raw materials and on tile composition. If your defects appear on a line that is already running, the machine settings themselves are worth reviewing against the production sequence described in our process guide.
For producers moving from hand-cast work to a pressed line, the practical constraints of layout, curing space and line configuration are set out in our guide to starting a terrazzo tile factory.
If a defect on your line does not match anything in this list, describe the symptom and send a photograph of the fracture surface — the way a tile breaks usually identifies the cause faster than anything written down.
