Silicone Cord and Profile: Extruded Seals in Continuous Length

A cord is a seal with no tool

Extruded silicone cord is the cheapest seal there is, because there is no mould. Compound is pushed through a die, cured in a tunnel, and coiled. The die costs a few hundred euros and the part costs by the metre, which is why cord appears in every application where a gasket has to be made without a tooling budget: oven doors, enclosures, machine guards, glazing channels, lighting.

What a cord cannot do is hold a complex section to a tight tolerance. The die sets the section, but die swell, cure shrinkage and take-off tension all move it, so a cord quoted at 8 mm is typically plus or minus 0.2 mm and a complex profile is worse. For a groove that has to be filled exactly, that matters; for a groove that is designed to be overfilled, it does not.

The whole design approach changes with that. A moulded gasket is designed to a dimension. A cord seal is designed to a compression ratio, usually 20 to 30 percent of the cord diameter, so that the tolerance on the cord is absorbed by the squeeze rather than by the groove.

The faster line is also the better seal

Cure tunnel temperature against line speed: 8 m/min at 140 C, 14 at 160, 22 at 180 and 30 at 200. That is the expected relationship, hotter tunnel, faster cure, faster line. The interesting part is the second curve.

Compression set over the same temperatures: 28 percent at 140 C, 18 at 160, 12 at 180 and 9 at 200. Compression set is the permanent deformation left after the seal has been squeezed and released, and lower is better. So running the tunnel hotter gives a faster line and a seal that lasts longer, which is not the trade most people expect.

The reason is that compression set is dominated by how completely the crosslinks form, not by how fast. A cord cured at 140 C leaves the tunnel with a surface that is set and a core that is under-cured, and the core is what takes the permanent deformation. A cord cured at 200 C is crosslinked through the section in a fraction of the time, and the crosslink density is more uniform.

Where the temperature stops helping

The limit is not the cure, it is the compound. Above about 210 C a peroxide-cured silicone starts to scorch in the die, and a platinum-cured silicone can be inhibited by sulphur, tin or amine contamination that was harmless at lower temperatures. A tunnel at 220 C that works on Monday fails on Thursday when a new drum of compound arrives.

The second limit is the take-off. A cord cured hot and fast is soft at the die and has to be supported until it is set, so the first metre of the line has to be a smooth, tensioned run. Increase the temperature without improving the take-off and the cord stretches under its own weight, and the diameter goes out of tolerance while the cure is perfect.

The third is the die swell. Hotter melt swells more on exit, so the die has to be cut smaller to give the same finished diameter. A die cut for 180 C will produce an oversized cord at 200 C, and the correction is a new die rather than a new setting.

Table 1 — Extruded silicone cord settings and what each one controls

When a cord becomes a profile

A round cord seals against a flat surface. As soon as the groove is complex, or the seal has to clip onto something, or the section has to be asymmetric, the die becomes a profile and the problems change. A profile has thin sections and thick sections, and they cure at different rates.

The thin section is over-cured and the thick section is under-cured, on the same line at the same speed. The result is a profile that is brittle where it is thin and takes a set where it is thick. The fixes are to design the section with a more uniform wall, to cure at a lower temperature for longer, or to accept the compromise and post-cure the coil.

Profiles also need support. A hollow profile collapses under its own weight before it is cured, so it is extruded onto a mandrel, or with internal air pressure, or it is coextruded over a solid silicone core that holds the shape. A coextruded core is the most common answer for a sealing bulb, and it is why a bulb seal is stiff in one direction and soft in the other.

The joint is the part that fails

Cord is sold in coils and spliced into a loop, and the splice is where most cord seals leak. A butt splice with cyanoacrylate holds in tension and fails in compression. A vulcanised splice, made in a small heated mould with uncured compound at the joint, is the only one that behaves like the rest of the cord.

The splice also has to be positioned. On an oven door seal the splice goes at the hinge side, where the compression is lowest and where a small discontinuity is least likely to open. On a continuous gasket the splice goes where the joint will be clamped hardest.

Table 2 — Cord splices and where each one is honest about its limits

What to specify

A cord specification is short, and the short list is what keeps it honest.

• Diameter or section drawing, with the tolerance that the extrusion can actually hold.
• Hardness in Shore A, and whether the application needs a low-compression-set grade.
• Compression set with test conditions, since 9 and 28 percent are the same material at different cure temperatures.
• Service temperature, continuous and peak, because silicone at 200 C continuous is a different grade from 200 C peak.
• Splice type and position, if the cord is supplied as a loop.
• Whether the cord will be post-cured, which decides whether it outgasses in a sealed enclosure.

The compression set number is the one to insist on. Two suppliers quoting the same hardness and the same diameter can be 20 percentage points apart on set, and the difference is entirely in how they ran the tunnel.