Silicone Extrusion: Curing a Profile That Cannot Hold Its Own Shape

Extruded soft, cured in place

Silicone extrusion is unlike thermoplastic extrusion in one fundamental way. A thermoplastic profile is cooled after the die and becomes rigid as it cools, so the die and the calibration sleeve hold the shape. A silicone profile leaves the die as a soft, uncured rubber with almost no strength, and it has to be cured before it can be handled.

The curing happens in a tunnel immediately after the die. The profile passes through a heated chamber, usually with hot air at 200 to 400 degrees Celsius, and the peroxide or platinum cure system reacts. By the time it leaves the tunnel the profile is a rubber and can be pulled, cooled and coiled.

Because the profile is unsupported in the tunnel, it sags under its own weight. That is why silicone extrusions are usually round, or supported on a talc bed, or run vertically. A complex profile with thin walls is difficult to extrude in silicone for exactly this reason.

There are two routes to the uncured profile. Solid silicone rubber is compounded, strained and extruded on a screw; liquid silicone rubber is pumped and extruded through a gear pump and cured faster. The solid route is more common for profiles, the liquid route for tubing and thin sections.

The tunnel sets the line speed

The chart on the left plots line speed against tunnel temperature for a 6 mm wall profile. At 200 degrees the line runs at 3.5 metres per minute, at 250 it runs at 5.2, at 300 at 7.0, at 350 at 8.8 and at 400 at 10.2. The relationship is close to linear over that range, which makes the economics easy to model and the limits easy to miss.

The linearity stops at the top. Above about 400 degrees the surface cures faster than the core, the surface skins over, and the volatiles generated inside cannot escape. The result is a profile that looks cured and is porous, or one with a rough surface where the skin has ruptured.

Wall thickness is the other variable and it acts the same way as in any curing process: the time needed goes roughly with the square of the thickness. A 12 mm wall profile runs at about a quarter of the speed of a 6 mm one at the same tunnel temperature, which is why heavy sections are a different commercial proposition.

The practical control is a cure check, not a temperature setting. A tunnel temperature is a setpoint; whether the profile is cured is a measurement, and the measurement is usually a solvent swell, a hardness check or a simple stretch and recovery test on a sample from each run.

The defects and what they mean

The right hand panel gives the distribution: under-cure 28 percent, porosity 24, dimensional drift 18, surface roughness 14, salt or talc carry-over 10. Under-cure and porosity are both cure-related, which is why the tunnel dominates the defect list.

Under-cure is the one that reaches the customer. A profile that is under-cured looks correct, measures correctly and then fails in service: it takes a permanent set, it tears easily, it ages badly. The only defence is a cure test on every run, because the visual check will not find it.

Porosity is the defect that comes from curing too fast. The surface skins over before the interior has released its volatiles, so the gas forms bubbles inside the wall. On a translucent profile they are visible; on an opaque one they show up as a soft, spongy feel and a low density.

Dimensional drift at 18 percent is the consequence of the profile being unsupported. The extrudate swells after the die, sags in the tunnel, and the final dimension depends on the die, the draw-down and the sag. A change in line speed changes all three, so a speed change is a dimensional change.

Supporting a profile that cannot support itself

The talc bed is the traditional answer. The profile is extruded onto a bed of talcum powder, which supports it, prevents it sticking to itself and carries it through the tunnel. The cost is the powder, the dust and the cleaning, and the defect is carry-over.

Salt baths are the other traditional route, using a molten salt at 200 to 250 degrees that supports the profile and conducts heat far better than air. The cure is faster and more uniform, and the cleanup is worse. Salt carry-over at 10 percent is the price.

For simple round profiles the answer is often no support at all: run vertically downward, or run horizontally with a short unsupported span and accept a slight ovality. The design of the profile determines which route is available, which is why silicone profile design is a conversation with the extruder rather than a drawing sent for quotation.

Table 1 — Curing methods and what each one suits.

The compound decides the window

A silicone compound for extrusion is not the same as one for moulding. It needs a higher green strength so that the uncured profile holds its shape, a cure system that matches the tunnel, and a filler package that gives the right hardness and tear resistance after cure.

Green strength is the property that people forget. A compound with good cured properties but low green strength will sag in the tunnel and produce an oval profile, and no amount of tunnel tuning will fix it. The fix is in the compound, and it is usually a change of filler or of the polymer molecular weight distribution.

The cure system matters too. Peroxide cure is tolerant of the fast heat ramp in a tunnel and is the usual choice for solid silicone extrusion. Platinum addition cure is faster and cleaner but is sensitive to contamination, and a trace of sulphur or amine in the plant will stop it curing at all.

Which is why a silicone extrusion line and a silicone moulding shop in the same building need to be careful about what else is in the air. A platinum-cured extrusion running next to a peroxide-cured moulding operation is a recipe for intermittent cure failure that is very hard to diagnose.

Table 2 — Compound routes and the constraint each one imposes on the line.

Specifying an extruded silicone profile

A drawing that gives a shape and a hardness is not enough. The specification should give the compound type, the hardness with a tolerance, the cure system, the dimensional tolerances on the section, the cut length tolerance and the surface requirement.

The cure is the item that needs a test rather than a note. A specification that says fully cured is a wish; one that says a solvent swell of a stated value on a sample per batch is a control. The test is cheap and it is the only one that catches the defect that matters.

The dimensional tolerances have to reflect the process. A silicone profile cannot hold the same tolerance as a rigid PVC one, because it is soft, it sags and it stretches when measured. A tolerance that is achievable in PVC will produce permanent scrap in silicone.

Finally, the packing. Silicone profiles are usually coiled, and a coil that is stretched when it is wound will be shorter than specified and will have a set. The specification should say whether the length is measured under tension or relaxed, because the two numbers differ by several percent.