Vulcanisation Temperature and Time: The Cure Curve Behind Every Part

The cure curve is a rate, not a time

Silicone vulcanisation is a chemical reaction, and like most reactions it speeds up with temperature. What a data sheet calls the cure time is really a point on a curve: the time to reach a given state of cure at a given temperature.

The useful measure is t90, the time to reach 90 percent of the maximum torque on a moving die rheometer. It is a standard test, it is reproducible, and it gives a single number that can be compared between compounds and between batches.

On a log axis, t90 against temperature is close to a straight line, which means the relationship is exponential. That is convenient to remember and dangerous to extrapolate: a cure schedule developed at 160 C does not scale linearly to 200 C, and a part that cures in 15 seconds at 160 may scorch in 5 at 200.

Twenty degrees halves the cure

For a typical addition-cure silicone, t90 goes from 60 seconds at 120 C to 42 at 130, 30 at 140, 21 at 150, 15 at 160 and 10.5 at 170. Twenty degrees halves it: 60 at 120 becomes 30 at 140 and 15 at 160.

The rule is a useful planning tool, and it has limits. Below about 110 C the reaction is slow enough that the inhibitor dominates and the relationship flattens. Above about 190 C the cure is fast enough that the thermal mass of the part matters more than the oven setting, and a thick part will be overcured on the surface before the centre is cured.

That is why cure schedules are quoted with a thickness. A 2 mm section can be cured at 180 C for 10 seconds; a 10 mm section at the same temperature will have a fully cured skin and an undercured core, and no amount of extra time at that temperature will fix it without degrading the skin.

Table 1 — t90 against temperature, and the halving at every twenty degrees.

Properties follow the cure state

Tensile strength develops with cure state and it develops unevenly. At 50 percent of t90 the material has about 60 percent of its strength, at 70 percent it has 85, at 90 percent it has 97 and at 100 percent it has 100. Past 100 percent there is nothing to gain, and a slow loss begins.

The shape of that curve is why under-cure is the expensive side. A part cured to 70 percent looks and feels like a part cured to 100, passes a visual inspection, and fails a tensile test by 15 percent. The defect is invisible until the part is loaded.

Over-cure is less damaging but not harmless. Beyond about 110 percent of t90 the network continues to crosslink and the material stiffens, the elongation falls and the compression set starts to rise. At 130 percent the part is measurably harder than specification, which on a sealing part is a functional failure.

How to know where you are

A cure schedule is only as good as the measurement behind it. The three practical methods are a rheometer curve on the compound, a hardness check on the part, and a solvent swell test.

The rheometer is the reference: it gives t90 for the compound at the production temperature, and the schedule is set from it with a margin. The hardness check is the shop-floor method, and it catches gross under-cure but not a 10 percent shortfall. The swell test is the most sensitive, because an incomplete network absorbs more solvent, and it is the one to use when a part passes hardness and fails in service.

The measurement has to be on the part, not on a slab. A slab cured in a laboratory press at 5 mm thickness and a production part at 12 mm have completely different thermal histories, and the part is what the customer receives.

Thick sections and the thermal lag

Heat travels slowly through rubber. Silicone has a thermal diffusivity of about 1 x 10 to the minus 7 square metres per second, which means a 10 mm section takes minutes, not seconds, to reach the mould temperature at its centre.

Table 2 — Cure schedules by section thickness, and what happens if they are shortened.

So a thick part is cured from the outside in, and the surface sees far more heat history than the core. The usual solution is a lower temperature for longer, which lets the core catch up before the surface degrades. A 10 mm part might run at 130 C for 10 minutes where a 2 mm part runs at 170 C for 15 seconds.

The alternative is a two-stage cure: a short press cure to give the part enough green strength to be handled, followed by an oven post-cure that brings the whole section to temperature evenly. That is standard for thick mouldings and it is covered separately under post-curing.

Setting a cure schedule that repeats

A schedule that works once and drifts is worse than no schedule, because it produces intermittent scrap. Four things make it repeat.

1. Measure t90 on the actual compound at the actual production temperature, not from the data sheet.
2. Set the dwell at t90 plus 20 to 30 percent, which covers the thermal lag without overcuring.
3. Control the mould temperature with a controller, not with a dial, and log it.
4. Check hardness and swell on the first and last part of every run.
5. Recalibrate when the compound lot changes, because t90 varies by 10 to 15 percent between lots.

The lot-to-lot variation is the item most often missed. A compound from a different batch can have a t90 15 percent longer than the one the schedule was set on, and the result is a run of parts that are technically cured but 5 percent short of full strength.