Silicone Post-Processing: Deflashing, Washing and Assembly

The sequence matters more than the method

Post-processing a silicone part is four operations in a fixed order: deflash, wash, dry, inspect. Changing the order creates problems that no amount of care in the individual steps can fix.

Deflash before washing, always. Flash that is washed first becomes soft and smears instead of cutting cleanly, and the smeared edge looks like a moulding defect. Wash before drying, because drying a part with cutting residue on it bakes the residue onto the surface. Inspect after drying, because a wet part hides surface defects and shows reflections that are not there.

The exception is a part that is assembled wet, such as a gasket that is fitted into a housing while still slightly tacky. Even then the deflash comes first, and the wash step is shortened rather than moved.

The order also sets who owns a defect. If flash is found after washing, it could have been a trimming miss or a torn edge created by the wash basket. If it is found before, the trimming station owns it. Clear sequence makes clear ownership, and clear ownership is what reduces the reject rate over months.

Deflashing: three routes and what each costs

Hand trimming with a knife is the default for small runs and for parts with complex edges. It is fast to set up, needs no tooling, and depends entirely on the operator. The flash line has to be cut, not pulled; pulling stretches the silicone and leaves a torn edge that will fail a visual standard even though the dimension is right.

Cryogenic deflashing tumbles the parts with a cryogenic medium so the thin flash becomes brittle and breaks off. It suits high volumes of small parts with simple flash lines and it removes labour from the step entirely. It does not suit parts with thin walls, because the wall can become brittle too, and it does not suit parts where the flash is thick.

Water jet deflashing uses a high pressure jet to cut the flash away. It handles thick flash and complex geometry and leaves a clean edge, but it needs a fixture to hold the part, and the fixture cost only makes sense above a certain volume.

The choice is usually volume against geometry. Under a few thousand pieces a year with mixed shapes, hand trimming wins. Above that with a stable shape, cryogenic. Where the flash is thick or the edge is cosmetic, water jet, regardless of volume, because the other two leave a witness line.

Table 1 — Deflashing routes against the work they suit

Washing is about extractables, not appearance

A silicone part can look perfectly clean and still fail an extractables test. The residue that matters is not visible dust; it is unreacted oligomer, mould release, and the low molecular weight fraction that migrates out of the cured rubber. Those come off in a wash, and the wash route determines how much stays.

Measured as extractables after cure on a 50 Shore A platinum compound, an unwashed part runs around 4,100 ppm. A solvent wipe brings it to 1,650. Hot water with detergent reaches 720. Ultrasonic with detergent reaches 380. A two-minute plasma treatment reaches 95.

The ranking is not surprising once you think about the mechanism. Wiping removes what is on the surface and pushes the rest around. Soaking in hot detergent swells the surface slightly and lets the mobile fraction diffuse out. Ultrasonics add cavitation, which agitates the boundary layer and reaches into the corners a cloth cannot. Plasma does something different: it oxidises the surface layer, which both cleans it and changes its chemistry.

For most industrial gaskets, ultrasonic with a mild detergent is the right answer. It gets extractables low enough for food and medical adjacent work at a cost that a job shop can carry. Plasma is for the cases where the surface also has to be bondable, because it does both jobs in one step.

Drying, and the water that stays behind

A washed silicone part holds water in a way that a moulded plastic part does not. The surface is hydrophobic, so water beads rather than sheets, and those beads sit in the corners and in any blind hole. A compressed air blow-off moves the beads but does not remove the film.

The practical fix is a warm air dry at 60 to 80 C for 15 to 30 minutes, which is below any temperature that would affect the cure but well above what is needed to drive off surface water. Tumbling during the dry helps, because it keeps the parts moving so water cannot pool.

Residual moisture matters for two reasons. It shows up as a haze on a glossy black part, which reads as a moulding defect. And it interferes with bonding and with printing, both of which are usually the next step after washing. A part that is bonded wet fails at the interface, and the failure is blamed on the adhesive.

If the part is going straight into a sealed bag, dry it harder. Moisture sealed into a bag with a silicone part will condense on the inside of the film in a cold warehouse and the customer will open a bag of parts that look wet.

Assembly is where clean parts get dirty again

A part that has passed the wash and the extractables check can fail at assembly. Handling with bare hands transfers skin oils and salts; a nitrile glove transfers plasticiser if it is a plasticised grade; a latex glove transfers sulphur, which is exactly what poisons a platinum cure if the part is being bonded to another silicone.

The glove rule for silicone assembly is nitrile, powder-free, and changed often. Not latex, for the sulphur. Not vinyl, for the plasticiser. And not the same pair all shift, because a glove that has handled a release agent will transfer it to every part it touches afterwards.

Assembly also changes the inspection standard. A gasket that is fitted into a housing has a visible compression set line, and that line is not a defect. The inspection sheet has to say so, or the next inspector will reject a good part for showing the mark of its own function.

Where the part is bonded, the bond is the weak point and the surface preparation is the whole game. Priming the substrate, not the silicone, is usually what decides whether the joint survives heat ageing. That is a separate subject, and it is the one that decides whether the assembly step is a cost or a warranty claim.

Where the hours actually go

On a typical silicone job, deflashing takes 38 percent of the post-processing hours, washing 22, visual inspection 18, assembly 14 and packing 8. The moulding cycle itself is often a smaller share of the total labour than any of the first three.

That distribution is why deflashing automation pays back faster than almost any other investment on a silicone line. Cutting the deflash time in half removes 19 percent of post-processing labour, and it removes the most variable 19 percent, because hand trimming is where operator-to-operator differences are largest.

It is also why tool design decisions made at the quotation stage matter so much. A tool that puts the flash line on a hidden face, or that splits the flash between two shut-offs so it breaks off cleanly, can cut the deflash step by more than any machine will. The cheapest post-processing is the flash that never had to be removed.

Table 2 — The defects that come from post-processing, not from moulding