Silicone Compression Moulding: The Workhorse Process Explained

Why it is still the workhorse

Compression moulding is the oldest process in the rubber industry and it is still the most widely used for solid silicone rubber. A preformed blank of uncured compound is placed in an open cavity, the mould closes, heat and pressure cure the part, and it is demoulded. There is no runner, no sprue and no injection unit.

It survives because it is cheap to tool, tolerant of compound variation, and able to mould parts that no other process can: very large parts, very thick sections, parts with heavy inserts, and low volumes where the tooling cost of injection moulding cannot be justified. A compression tool for a simple part can cost a fifth of the equivalent injection tool.

It also survives because it is forgiving in a specific way that matters for silicone. Silicone compound is mixed, catalysed and often pigmented in-house, and its cure characteristics drift from batch to batch. A compression process absorbs that drift by adjusting cure time. An injection process with a fixed cycle has much less room to do so.

The four steps

1. Preforming. The uncured compound is cut or extruded into a blank of the right weight and shape. Too light and the cavity does not fill; too heavy and the excess has to escape as flash, which distorts the part and wastes material.
2. Loading. The blank is placed in the cavity by hand or by a robot. Placement matters: a blank that is off-centre fills one side first and traps air on the other.
3. Pressing and curing. The mould closes under pressure and is held at temperature for the cure time. The pressure has to be enough to close the mould completely but not so much that it squeezes the compound out of the cavity.
4. Demoulding and post-cure. The part is removed, often still hot, and then post-cured in an oven to drive off volatiles and complete the cure. The post-cure is what fixes compression set.

The step that is most often skipped is the post-cure, and it is the one that shows up six months later as a gasket that has taken a set and leaks. A compression moulded silicone part that is only press-cured retains decomposition by-products from the curing agent, and those by-products both smell and plasticise the rubber.

Cure time: temperature against thickness

Cure time is dominated by two variables and they pull in opposite directions. Raising the mould temperature speeds the chemistry: a 4 mm section cures in 12 minutes at 120 °C, 7 at 140, 4.2 at 160, 2.6 at 180 and 1.6 at 200. That is a factor of seven across a range that is entirely practical.

Thickness works the other way. Heat has to conduct into the part, and conduction time scales with the square of the distance, so doubling the section roughly quadruples the time to reach cure temperature at the core. Measured against a 2 mm section, a 4 mm section takes 2.2 times as long, a 6 mm section 4.5 times, and a 10 mm section 11 times.

Put the two together and the design consequence is clear. A thick section cannot be rescued by temperature, because raising the mould temperature overheats the surface while the core is still climbing. Above about 6 mm, the practical answer is to redesign the part, or to accept a long cycle, or to move to a process that heats the material more uniformly.

Table 1 — Cure time against section thickness at two mould temperatures.

Tooling, shrinkage and flash

A compression tool is simple: two plates, a cavity, often a common flash groove around the cavity edge, and heating by platens or by cartridge heaters. The simplicity is why the tool is cheap, and the flash groove is why the parts need deflashing.

Silicone shrinks 2 to 4 percent linearly, depending on the compound and the filler loading, and the shrinkage is not isotropic in a part with flow orientation. The tool has to be cut oversize by the shrinkage factor, and the factor has to be measured on a test moulding rather than taken from a data sheet, because two compounds from two suppliers at the same hardness will not shrink the same amount.

Flash is the other tooling consequence. In compression moulding the excess compound has to go somewhere, and it goes into the flash groove as a thin fin around the part. That fin has to be removed, by hand, by cryogenic deflashing or by tumbling. For a part with a cosmetic edge, deflashing is a real cost and it is worth designing the flash line onto a hidden face.

Table 2 — The tool features that decide whether a compression tool works.

The five defects

Table 3 — The five defects that account for most compression moulding scrap.

Backrind deserves a longer note because it is the defect that most often looks like a tool problem and is not. It happens when the mould opens while the part is still soft, and the rubber is torn at the flash line as the two halves separate. The cure is too short for the section, or the platen is cooler than the controller says, or the cavity is so overfilled that the excess is still flowing when the mould opens. All three are process, not tool.

When to move to transfer or injection

Compression moulding has three natural limits: cycle time on thick sections, tolerance on parts with inserts, and labour on high volumes. When any of those becomes the constraint, the alternatives are transfer moulding and liquid silicone rubber injection.

Table 4 — The three silicone moulding processes, and where the crossover sits.

The crossover is not only volume. LSR injection is chosen for parts that must have no flash, for parts with fine detail, and for medical parts where manual handling is a contamination risk. Transfer moulding is chosen for parts with inserts, because the closed cavity holds the insert in position while the compound flows around it, which compression moulding cannot do reliably.

What drives the cost

A compression moulded silicone part is priced from four things, and only one of them is the material.

• Material. Silicone compound is expensive, and the blank weight is not the part weight: the flash allowance is typically 10 to 20 percent, and it is scrap unless it can be re-milled into a low-spec compound.
• Cycle time. The press is the constraint, and a 20 minute cycle on a 10 mm section means a press produces three parts an hour regardless of how many cavities it has.
• Labour. Loading and demoulding are manual on most compression presses. A robot changes the economics but adds capital.
• Deflashing. Hand deflashing of a cosmetic part can cost more than the moulding. Cryogenic deflashing is faster and adds equipment cost.

The lever with the most effect is cavity count, because the press cost is fixed per cycle. A 16-cavity tool running a 4 minute cycle produces 240 parts an hour on one press, and the same part in a 4-cavity tool produces 60. That is why compression moulding is competitive at volumes that people assume belong to injection, as long as the part is small and the section is thin.