Cast silicone moulds
Boxed moulds with the split line chosen rather than simply accepted. Once the box was built properly, the patch stayed on this route.
Cast mouldsProject record · Electrical assembly · Cast route, multi-cavity
An electrical patch that would not lie flat.
This thin cast silicone patch needed to sit flat on a housing, but it would not. The part and material were right; the mould was the problem. Rebuilding the box fixed the fit, and moving to a multi-cavity tool made it worth running.
The brief
A thin patch that had to settle into place on its own.
An electrical assembly business came to us with a thin silicone patch. It sits between a housing and a mounting face to seal and insulate the assembly. The patch had a simple outline, with no undercuts or fine detail to complicate the mould. The key requirement was that it lay flat on the housing before clamping. That meant the operator could use both hands for the fixings, rather than holding a patch that curled out of position.
They had been buying the parts from an existing supplier and had a drawer full of rejects. The parts were the right size, hardness and colour. They simply were not flat. The corners lifted, always the corners, and never by quite the same amount on two parts in a row.
They arrived expecting a conversation about materials: softer grade, harder grade or a different filler. We had that conversation anyway. But we did not think the material was where the fault lived. When a curl changes from part to part, it is not usually a property of the material. It is usually a property of the thing the material was cured in.
The diagnosis
A flexible mould inside a flexible box produces a curled part.
The original mould was a single silicone block, poured and then trimmed, resting loose on a bench while it cured. That is a perfectly reasonable way to make many things. For a thin, flat mould, though, it is not the right approach.
A thin cavity in a soft mould has almost no stiffness across its width. When you put warm silicone into it, the block relaxes. Lift it to move it, and the block sags. Whatever shape the mould face is holding when the part gels is the shape the part keeps. In a thin section, there is not enough material in the part to resist that shape afterwards. The corners are worst because that is where the block is least supported.
We did not just assert it; we proved it. We cast the same material in the original mould and in a trial mould backed by a machined rigid plate, using the same batch and curing both side by side. The plate-backed parts came out flat and stayed flat. The bench-cured parts curled, while those we deliberately moved during curing curled more.
So the fix turned out to be a tooling job rather than a materials job, which was welcome news for the customer's budget.
Why a rigid base kills warp in a thin cast part.
Illustrative schematic. Real tools include additional detail (runners, vents, ejector positions).
What we changed
Rebuild the box, then repeat it twelve times.
We first proved the single-cavity fix, then multiplied it.
Put the mould in a rigid frame
Instead of pouring silicone into a soft box, we cast it into a machined frame. The rigid frame carries the load, while the silicone forms only the face.
Reference the cavity to a flat base
We took the cavity floor from a ground plate, so the surface your patch sets against is genuinely flat rather than only approximately flat.
Cure it where it was poured
We fill the tool, close it and leave it to cure where it was poured. Nothing is carried across the workshop with liquid silicone inside, because that is where half the curl came from.
Take the parting line off the sealing face
We moved the split to the patch edge. This keeps any flash where it can be dressed, rather than on the sealing face that needs to seal against the housing.
Scale up to twelve cavities
Once one cavity was making a flat part every time, we repeated that cavity across a single frame. This means a whole set can be filled and cured under the same conditions.
Check flatness on a plate, not by eye
First-off patches are checked against a surface plate, not held up and judged by eye. A patch that looks flat in your hand is not proof that it is flat.
The stages
A typical run, drawn out on a timeline.
Drag the timeline left and right, or use the arrow keys. Each stage explains what happens and roughly how long it takes.
Drag · tap a stage for detail
Total working time on a job of this shape: about 4 weeks (26 working days). Lead times shift with drawing changes, sign-off pace and material availability.
The commercial bit
The numbers, out in the open.
Rather than quote a figure that would only be right for one job, here is a worked example, based on a realistic project of this shape. Your quote follows the same shape, with your part, your volume and your material.
Worked example
An electrical OEM needing 2,000 patches a year, dead-flat critical.
| Machined aluminium base plate | £420 one-off |
|---|---|
| Silicone mould on the base | £340 one-off |
| Tooling amortised over 6,000 parts (3 yrs) | ≈ £0.13 per patch |
| Cast, cure, trim | ≈ £0.82 per patch |
| Landed cost per patch | ≈ £0.95 |
Assumes a 60 mm × 90 mm × 2.5 mm patch in 60 Shore A. Flatness spec ≤ 0.3 mm across face.
Where the job ended up.
For two years, we treated it as a material fault. No one had told us that the mould was causing the corners to lift.
Manufacturing engineer, electrical assembly manufacturer
Behind the work
The pages behind this job.
Two services carried this part, while a third is worth reading if your own part is still changing shape.
Multi-cavity tooling
Repeating a proven cavity lets the whole set fill and cure together. This can lower your cost per part, but any change affects every cavity.
Multi-cavity tooling
CAD and design for manufacture
This is where we agree the split line and fill arrangement before anything is cut. It is the cheapest stage at which to fix a flatness problem.
CAD and DFMMore records
Other work of this kind.
A feature that could not hold air
This aerospace part had finer detail than anything we had previously held in a soft tool.
Aerospace componentFlash on a sealing lip
We positioned the parting line on the one face of the ear pads that needed to feel like nothing.
Silicone ear padsWhen a master stops the cure
The modelling clay pattern that left the mould face tacky, and the test that identified it.
Jewellery masterNext step
Is your thin part curling?
Curl on a thin cast part is nearly always caused by what happened during cure, rather than by the material itself. The useful questions are: is the mould properly supported, is the cavity floor genuinely flat, and was the mould picked up before the part had gelled? Two of those three checks are free to fix.
Please send a handful of the bad parts rather than just one. A single curled patch shows that it curled; six show whether the curl is consistent. Consistency points to geometry, while inconsistency points to handling. If you have the existing mould, please send that too, as it usually answers the question by itself.
If the part hasn't been made yet, bring us the geometry. We'll advise where the split line and fill point should go before anyone builds a box.
At a glance
The idea in a picture.
Silicone hardness in everyday terms
Swipe to see the whole diagram
Ready to start?
Tell us what you need to make, and we'll help you work out how to mould it.
Send us a drawing, photograph or sample, and we'll come back to you with a mould strategy, material recommendation and price.
Real people, a real workshop
You deal with the people who make it.
No call centre and no middlemen. When you contact TCI Mouldings, you speak to the team who design, build and check your mould.
- A real Warwickshire workshopUnit 15, Waterloo Park, Bidford-on-Avon, Alcester B50 4JG.
- Real people on the phoneCall the team on 01789 224228, Monday to Friday.
- The same team on every jobDesign, production and aftercare come from the people you see here.
- Part of TCI GroupFive specialist brands, one engineering team. Visit TCI Group.
Thomas LaneManaging Director
George PrestonDesign & Project Manager
Kelly BallingerHead of Partnerships
Jim AndertonProduction Manager
TheodoreEmployee Welfare