What cellular PVC actually is
Cellular PVC — foam PVC, sometimes called rigid cellular PVC — is PVC with gas dispersed through it. The gas comes from a chemical blowing agent that decomposes in the barrel, or from a physical blowing agent injected into the melt. The result is a material with a density of 0.45 – 0.75 g/cm³ instead of 1.4, which is why a foam PVC board is a third the weight of the solid board it replaces.
The reason it matters commercially is that solid PVC is heavy and expensive. A 20 mm solid PVC board weighs 28 kg/m² and costs accordingly. The same board in foam at 0.55 g/cm³ weighs 11 kg/m², can be nailed and screwed like timber, machines like hardwood, and takes a wood-grain surface. That combination is why foam PVC replaced timber in fascia, soffit, window surrounds and bathroom panelling.
The number that governs the whole process: density. Everything downstream — stiffness, screw holding, thermal insulation, surface hardness, cost — follows from it. And density is set by the balance between blowing-agent concentration and the line's ability to restrain the expansion, which means it is a process variable as much as a formulation one. The same compound run 5 % faster produces a lighter, softer board.
Three ways to make it
| Proceso | Density range | Skin | Surface | Used for |
|---|---|---|---|---|
| Celuka (internal mandrel) | 0.5 – 0.7 g/cm³ | Integral, 0.3 – 0.8 mm | Hard, glossy, smooth | Furniture, panelling, window surrounds |
| Free foam | 0.4 – 0.6 g/cm³ | Thin, 0.1 – 0.3 mm | Softer, visible cell texture | Where surface is laminated or painted |
| Co-extruded (foam core + solid skin) | 0.45 – 0.65 g/cm³ | Solid, 0.5 – 1.5 mm | Best — solid PVC surface | Exterior, high-end |
| Wood-plastic foam (PVC + wood flour) | 0.55 – 0.9 g/cm³ | Integral | Wood grain, embossed | Decking, cladding |
Table 1 — The four cellular PVC routes. Celuka is the volume process; co-extrusion is the quality one.
Celuka is the dominant process. The die has an internal mandrel or "torpedo" that cools the centre of the melt as it exits, so the surface freezes into a solid skin while the core stays hot and expands. The result is a board with a hard, closed surface and a uniform cellular core — the structure that makes it usable as a timber replacement.
Free foam lets the extrudate expand freely in air before it reaches the calibrator. It is cheaper and simpler, gives a lower density, but produces a thin skin and a surface that shows the cell structure. It is normally laminated or painted.
Co-extrusion puts a solid PVC skin over a foam core in a single die. It is the best surface and the best weathering, and it costs 20 – 35 % more because it needs a second extruder and a more complex die.
Formulation: what each part does
| Componente | Typical loading (phr) | Función | If it is wrong |
|---|---|---|---|
| PVC resin, K-value 57 – 60 | 100 | Matrix | Higher K = tougher but harder to process |
| Calcium-zinc stabiliser | 3 – 5 | Prevents dehydrochlorination | Too little: yellowing and burning |
| Chemical blowing agent (AC) | 0.4 – 1.2 | Generates gas | Too much: coarse cells, weak board |
| Blowing agent activator (ZnO) | 0.1 – 0.3 | Lowers decomposition temperature | Wrong level: gas released too early or too late |
| Acrylic processing aid | 2 – 6 | Melt strength and fusion | Too little: cells rupture, rough surface |
| Impact modifier (CPE or acrylic) | 4 – 8 | Toughness | Too little: brittle board, cracks at fixings |
| Calcium carbonate | 3 – 10 | Cost, cell nucleation | Too much: density up, impact down |
| Lubricants (internal + external) | 1 – 2.5 | Controls fusion and release | Imbalance: plate-out or burning |
| Titanium dioxide | 3 – 6 | Opacity and UV | Too little: yellowing outdoors |
| Cell nucleating agent | 0.2 – 0.6 | Uniform fine cells | Too little: coarse, irregular cells |
Table 2 — A working foam PVC formulation in parts per hundred resin. The blowing agent / activator pair is the balance that sets density.
The blowing agent and its activator are the two ingredients that decide whether the board is worth making. AC (azodicarbonamide) decomposes at about 200 °C on its own, which is above the PVC processing window. Adding zinc oxide drops the decomposition temperature to 170 – 185 °C, inside the window. Get the ratio right and the gas is released exactly as the melt leaves the die; get it wrong and either the gas comes out in the barrel (making a low-density, coarse, weak board) or it comes out too late (making a dense board with a rough, over-expanded surface).
Die and calibration
The die and the calibrator on a foam line are a single system. Unlike a solid profile, the die does not define the final dimension — it defines the melt stream, and the calibrator defines the board.
| Element | Design point | Effect if wrong |
|---|---|---|
| Die land length | 10 – 18 × gap | Too short: rough surface, uneven expansion |
| Die gap | 25 – 40 % of final thickness | Sets the expansion ratio |
| Mandrel (Celuka) | Cooled, 40 – 60 % of die width | Controls skin thickness and core density |
| Mandrel temperature | 20 – 60 °C below melt | Too cold: thick skin, dense core; too hot: no skin |
| Calibrator entry | Close to die, 30 – 100 mm | Too far: over-expansion before restraint |
| Calibrator taper | 3 – 8° entry | Too steep: drag marks; too shallow: unstable |
| Vacuum | 0.1 – 0.3 bar (lower than solid) | Too high: crushed cells at the surface |
| Cooling length | 6 – 12 m for a 20 mm board | PVC foam insulates, so cooling is slow |
Table 3 — Foam die and calibration. The vacuum is lower than for solid profile, which surprises people.
The vacuum point is worth explaining. On a solid profile, high vacuum pulls the melt hard against the calibrator and gives sharp detail. On a foam, high vacuum crushes the cells at the surface, producing a dense skin over a coarse core and a board that is heavier than specified with worse insulation. Foam lines run at 0.1 – 0.3 bar, just enough to control the surface, not enough to collapse it.
Cooling: the real bottleneck
PVC foam has a thermal conductivity of about 0.04 – 0.05 W/m·K, against 0.16 for solid PVC. It is an insulator — which is the point of the product, and also why foam lines are long.
| Board thickness | Cooling length at 400 kg/h | Line speed |
|---|---|---|
| 6 mm | 3 – 5 m | 1.8 – 2.5 m/min |
| 10 mm | 5 – 7 m | 1.2 – 1.8 m/min |
| 15 mm | 7 – 10 m | 0.9 – 1.3 m/min |
| 20 mm | 9 – 13 m | 0.7 – 1.0 m/min |
| 30 mm | 13 – 18 m | 0.5 – 0.7 m/min |
Table 4 — Cooling length for Celuka board. The thickness-squared relationship is brutal at 30 mm.
A board that is cut before its core is cool will bow, and the bow appears hours later — often after it has been packed. The practical rule is that the core must be below 60 °C at the saw, which for a 20 mm board means a tank long enough that the line speed drops to under 1 m/min. That is why foam lines look disproportionately long relative to their output.
Myth: "we can run faster if we cool harder." On foam you cannot, because the cooling is limited by the material's own conductivity, not by the water. Adding more spray does nothing once the surface is cold; the heat has to conduct out through the foam. The only real levers are a longer tank, a lower line speed, or a thinner board.
Surface: wood grain, lamination and painting
| Acabado | Método | Durabilidad | Coste |
|---|---|---|---|
| Smooth white | Polished calibrator | Bien | Baseline |
| Embossed wood grain | Textured calibrator or post-embossing roller | Good — the grain is in the surface | +5 – 10 % |
| Co-extruded cap, wood grain | Solid PVC cap + embossing | Best — weathering is in the cap | +20 – 35 % |
| Film lamination | PVC or PET film, adhesive | Good indoors; can delaminate outdoors | +15 – 30 % |
| Painted | Spray or roller | Fair — needs re-coating | +10 – 20 % |
| Foil wrapped (profile wrapping) | PVC/PET foil, hot melt | Good indoors | +20 – 40 % |
Table 5 — Surface options for foam board. Co-extrusion is the only finish that is genuinely exterior-grade.
Embossing works on foam because the surface skin is still thermoplastic when it leaves the calibrator. A textured roller or a moving belt presses the grain in at 90 – 110 °C, and the skin retains it. Emboss depth is 0.3 – 1.0 mm; deeper than that and the roller breaks through the skin into the cellular core, which shows as a rough patch.
Defects and causes
| Defecto | Apariencia | Causa | Fix |
|---|---|---|---|
| Density too high | Board heavier than specified | Blowing agent too low, or melt too cold | Increase AC; raise melt temperature 5 °C |
| Density too low | Board light and soft | Too much gas, or over-expansion before the calibrator | Reduce AC; move the calibrator closer |
| Coarse cells | Visible large bubbles in the core | Poor nucleation, or gas released too early | Add nucleating agent; check activator level |
| Thick skin / dense surface | Board heavy, core fine | Mandrel too cold | Raise mandrel temperature |
| No skin | Rough, open surface | Mandrel too hot, or vacuum too low | Cool the mandrel; check vacuum |
| Surface ripples | Wavy lines across the board | Unstable expansion at the die exit | Stabilise melt temperature; adjust land length |
| Bowing after cutting | Board curves hours later | Core still hot at the saw | Longer tank; slower line |
| Yellow / brown streaks | Discoloured lines | Local overheating or a die dead spot | Strip and clean; check for hang-up |
| Cell collapse at edges | Dense, glossy edges | Vacuum too high at the calibrator entry | Reduce vacuum; soften the entry taper |
| Poor screw holding | Fixings pull out | Density too low at the surface, or skin too thin | Raise density; thicken the skin via mandrel temperature |
Table 6 — Foam board defects. Six of the ten are density or skin-thickness problems, which are the same problem seen from different angles.
Properties and where foam PVC is used
| Propiedad | 0.55 g/cm³ board | Solid PVC |
|---|---|---|
| Densidad | 0.55 g/cm³ | 1.40 g/cm³ |
| Flexural strength | 18 – 25 MPa | 70 – 80 MPa |
| Flexural modulus | 0.7 – 1.1 GPa | 2.8 – 3.2 GPa |
| Impact (Charpy, notched) | 3 – 6 kJ/m² | 4 – 8 kJ/m² |
| Thermal conductivity | 0.04 – 0.05 W/m·K | 0.16 W/m·K |
| Screw pull-out, 20 mm board | 600 – 1,200 N | 2,500 – 3,500 N |
| Water absorption, 24 h | < 0.5 % | < 0.1 % |
| Service temperature | −20 to 60 °C | −20 to 65 °C |
Table 7 — Foam versus solid PVC. Stiffness drops by two-thirds but so does weight, and insulation improves fourfold.
The stiffness loss is the design constraint. A foam board is roughly one-third the modulus of solid PVC, so a shelf or a fascia that worked in 10 mm solid needs 16 – 20 mm in foam. That is fine for cladding, where stiffness is not the governing requirement, and it is why foam is not used for structural profiles.
Screw pull-out is the second constraint. A foam board holds a screw at 600 – 1,200 N, which is adequate for cladding fixings but not for anything load-bearing. Fixings should be into the skin, not the core, and should be at least 20 mm from an edge.
Frequently asked questions
What is the difference between Celuka and free foam?
Celuka uses a cooled internal mandrel so the surface freezes into a solid skin while the core expands — giving a hard, smooth surface. Free foam lets the extrudate expand in air before calibration, giving a lower density but a thin skin and a visible cell texture. Celuka is the process for anything that will be seen or handled.
Can foam PVC be used outdoors?
Only with a co-extruded solid cap or a laminated/painted finish. A plain Celuka board outdoors will chalk and yellow, and its surface will pick up dirt in the open cells. Co-extruded board with a 0.5 – 1.5 mm solid cap is genuinely exterior-grade and is used for fascia and window surrounds.
Why is my board heavier than the specification?
Either the blowing agent is low, the melt is too cold so the gas does not expand fully, or the mandrel is too cold and the skin is too thick. Check the mandrel temperature first — it is the most common cause and the easiest to correct.
Can foam PVC be welded?
Yes, with hot-air or hot-wedge welding using a PVC rod. The weld is not as strong as on solid PVC because the core is cellular, so welded joints in load-bearing applications need a mechanical backing. Solvent welding works on the skin but not through the core.
What is the maximum thickness?
Practically 30 – 40 mm. Beyond that the cooling length becomes uneconomic and the core density becomes hard to control. Thicker sections are usually made by lamination rather than extrusion.
Is foam PVC food safe?
Standard foam PVC is not a food-contact material. Food-contact grades exist but are uncommon; the blowing agent residues and the stabiliser system are the issues. For food display, use PET or polypropylene foam instead.
How much does a foam board die cost?
A single-cavity Celuka die for a 20 mm board runs €12,000 – €28,000 including the mandrel and the calibrator, in China. A co-extrusion version with a cap layer is €25,000 – €50,000. Add 3 – 5 trials and 6 – 10 weeks of lead time.
Why does my board bow after it is cut?
Because the core was still hot when the saw cut it. PVC foam conducts heat slowly, so the surface can be cold and rigid while the core is at 90 °C. As the core cools it contracts and the board curves. The fix is a longer cooling tank or a slower line — not a different formulation.
Making or specifying cellular PVC board?
Send the target density, thickness and finish. We will tell you which process fits, what line speed is realistic, and what the die and calibration will cost.
ES 
EN 
FR 
DE 
PT 
AR 
RU 
JA 
KO 
IT 
NL 
TR 
PL 
ID 