Scorching in Silicone: Cure That Starts Before the Mould

What scorching is, and what it is not

Silicone compound contains a curing system that starts to react at room temperature, slowly. Scorching is what happens when that reaction gets far enough ahead of the mould that the compound is partly crosslinked before it flows. It is not a cure problem in the mould; it is a storage and handling problem that shows up in the mould.

The industry measures it as scorch time, the time at a given temperature before the torque starts to rise, usually reported as t5 or ts2 from a moving die rheometer. A compound with a 40 minute scorch time at 120 C has roughly 40 minutes of safe flow once it is hot, and the clock starts the moment it is warm, not the moment it enters the press.

The confusion comes from the word cure. Curing is the reaction that makes the part; scorching is the same reaction happening where it should not. A scorched batch is not under-cured or over-cured, it is pre-cured, and the symptoms look like a flow problem because they are one.

Peroxide-cured compounds scorch more readily than platinum-cured ones, and the difference is not small. A peroxide system that has been sitting warm can be unusable in a day, while a platinum system in the same conditions will still run, though with a shorter safety margin.

The clock starts in the store room

Measured on a 50 Shore A peroxide-cured HCR compound, the scorch time was 46 minutes at 20 C, 34 at 25, 24.5 at 30, 17.4 at 35, 12.1 at 40, 8.3 at 45 and 5.6 at 50.

A store room at 30 C in summer is normal in most factories, and it cuts the safety margin by half compared with a 20 C store. At 40 C, which is what the top shelf of an uninsulated store reaches in July, the margin is a quarter of what it was in winter.

The curve is exponential because the reaction follows an Arrhenius relationship: the rate roughly doubles for every 8 to 10 degrees. That is why the difference between 20 and 30 C matters more than the difference between 10 and 20, and why air conditioning a compound store is a process investment, not a comfort one.

Time and temperature trade off against each other, which is the useful part. A compound that is good for 46 minutes at 20 C has consumed about the same fraction of its life after 34 minutes at 25 C, or after 12 minutes at 40 C. Multiplying time by the rate gives an equivalent age, and that is the number a store room should track.

Equivalent age is the number to track

Equivalent age is a way of adding up the time a batch has spent at different temperatures into one number, expressed as hours at a reference temperature. It is standard practice in food and pharmaceuticals and almost unknown in rubber, which is odd given how much it would help.

The calculation is simple enough to do on a spreadsheet: for each storage period, multiply the hours by a factor that doubles every 9 degrees above the reference. A week at 20 C is 168 hours. The same week at 35 C is 168 times 3.2, or about 540 equivalent hours, which is three times the ageing for the same calendar time.

Once a factory tracks equivalent age, the arguments stop. A batch that has 200 equivalent hours is fine; one that has 1,500 is suspect; one that has 3,000 should be tested before use. The numbers are specific to the compound, but the method is universal.

The practical version, for a factory that will not build a spreadsheet, is a rule: compound more than four weeks old gets a scorch test before it goes on the press, and compound that has been through a hot spell gets one regardless of age. That catches most of the problem for almost no cost.

Table 1 — Storage temperature against equivalent ageing, reference 20 C

Inhibitors, retarders and the shelf life claim

Most HCR compounds contain a scorch retarder, and the choice of retarder is what sets the shelf life on the datasheet. Retarders work by competing for the cure agent or by slowing the crosslink, and they are consumed over time, which is why shelf life is finite even in perfect storage.

A datasheet that claims six months shelf life is claiming six months at a stated temperature, usually 25 C or below, in the original packaging. Read the condition, not the number. A compound stored at 35 C does not have six months, it has about three.

The other thing to check is the packaging. Compound stored in an open bin picks up moisture and dust, and moisture affects peroxide cure more than platinum cure. Sealed, in the original liner, in a cool store, is the only condition in which the datasheet number means anything.

Retarder level is also a formulation variable, and more is not better. A compound with a long scorch time flows better but cures slower, which costs cycle time. The right answer is usually a moderate retarder plus a cold store, not a heavy retarder plus a warm one.

What scorching looks like on the floor

The first symptom is flow marks: a wavy, dull line on the surface where the compound stopped moving and then moved again. It looks like a weld line and it is often diagnosed as one, but a weld line from two flow fronts has a different shape and sits where the fronts meet, not where the flow hesitated.

The second is short fill that will not go away with more pressure. A scorched compound has a higher viscosity and a shorter flow life, so it fills the easy parts of the cavity and stops at the hard ones. Adding pressure raises the temperature and makes it worse, which is the diagnostic.

The third is a cure that looks uneven: hard at the gate, soft at the far end. That is the compound curing as it flows, so the material that arrived first has had the longest at temperature. On a well-behaved compound the cure is uniform across the part.

The fourth, and the one that gets misdiagnosed most often, is sticking. A pre-cured skin bonds to the mould differently from fresh compound, and the part tears on ejection. It looks like a release problem and it is a compound problem, which is why changing release agent does not fix it.

Table 2 — Scorch symptoms and what they are mistaken for

Storage, rotation and the FIFO rule

The fix is unglamorous. Store compound below 25 C, keep it sealed, rotate stock so the oldest is used first, and record the date and the store temperature. A factory that does those four things has almost no scorch problems; one that does none of them has a mystery that reappears every summer.

Cold storage has a trap: compound taken straight from a 10 C store into a warm room condenses moisture on the surface, and moisture causes blisters and porosity. The rule is to let sealed compound come to room temperature before opening it, which takes a few hours for a 20 kg bale.

The press side matters too. Compound sitting on a warm platen between shots is ageing at 40 to 50 C, which is the steep part of the curve. A covered tray, or a smaller batch brought to the press more often, costs nothing and buys back most of the margin.

And the store room thermometer should be read, not installed. A data logger that nobody looks at is decoration. A daily reading written on a sheet is a process control, and it is the one that would have caught the batch that caused last summer's rejects.

What to do with a suspect batch

Test it, do not guess. A scorch test takes about 20 minutes on a moving die rheometer, and comparing t5 against the incoming certificate tells you whether the batch has lost margin. If t5 has fallen by more than 30 percent, the batch is suspect; by more than 50 percent, it should not go into a cosmetic part.

A suspect batch can sometimes be used. Lower the mould temperature, shorten the flow path by using a smaller shot, or run it in a part where flow marks are hidden. Those are compromises, and they should be recorded as such, not treated as the new standard.

What cannot be done is to add retarder at the press. Retarder is a formulation component that has to be milled in evenly, and adding it by hand gives a compound with two cure behaviours in one bale. That is worse than the scorch.

The last option is to scrap it, and the calculation is usually simple: the cost of a bale against the cost of a shift of rejects plus a customer complaint. Most factories scrap too late rather than too early, because the bale is visible and the rejects are not.