Silicone conducts heat poorly. On a thick part, the skin can be crosslinked and sealed while the core has not yet reached cure temperature. Most thick-section defects are, at bottom, that time lag.
Thin parts rarely give trouble: the temperature spread is small, skin and core cure almost together, and the window is wide. Thick sections compress it — skin over-cured while the core is still under-cured.
The One-Line Mechanism: Heat Moves Inward, and So Does Cure
In compression molding, heat enters from the mold surface, so the outer layer always heats faster than the core. The skin reaches cure temperature first, skins over first, and finishes first. The core is one step behind at every step. The greater the wall thickness, the larger that lag.
The remedy in the literature targets exactly this. Lei, Shi, Luo and co-authors, in “Preparation of Thick Functional Products of Silicone Rubber”, Silicone Material (有机硅材料), 2012, 26(3): 164–169, report that a staged temperature ramp, extended cure time, and pressure-holding cooldown produce dense, defect-free thick sections, with reported process windows of 15 °C staging increments, 10–20 min ramp intervals, average heating rate 0.75–1 °C/min, and demolding temperature selected by section thickness from room temperature up to 80 °C.
Not one of those numbers says “push the temperature higher.” They all do the same thing: wait. Split the rise into steps, hold at each one, and the core keeps pace with the skin.
Four Failure Modes Unique to Thick Sections
Skin seals over, trapping gas inside
The outer layer crosslinks into a dense shell, closing the escape path for core gas and volatiles. That is as far as the mechanism goes here — the bubble types and their countermeasures live in Silicone Molding Defects: A Field Guide to Bubbles, Shorts, Sticking and Under-Cure.
Uneven crosslink density through the section
On one part, the skin may be near correct cure or slightly over while the core is still under-cured. Consequences:
- Hardness varies through the section — different readings at different points on the same part.
- Rebound and compression set become directional — performance depends on where the sample is taken, not only on the compound.
- Post-cure starts from two different reaction states, widening the gap further.
Ramping post-cure too fast → internal cracking and blistering
When a thick part moves into post-cure, a fast ramp heats the skin quickly while the core is still cold; residual low-molecular species and trapped gas expand with nowhere to go. **The four problems post-curing cannot fix are a separate chain — we leave it there.**
Uneven cooling → warpage and internal stress
Cooling also runs outside-in: the skin sets first, the core shrinks later, and the two restrain each other, building internal stress; asymmetric geometry turns that into warpage. That is why the literature places pressure-holding cooldown alongside staged heating — the cooling leg needs as much control as the heating leg.
Wall-Thickness Bands and What Typically Goes Wrong
| Wall thickness band | Typical difficulty | Direction of control |
|---|---|---|
| Thin (roughly ≤3 mm) | Little difficulty; wide process window | Conventional parameters; watch short shots and flash |
| Medium (roughly 3–8 mm) | Internal temperature spread appears; occasional skin-seal porosity | Staged ramp plus well-timed venting |
| Thick (roughly 8–15 mm) | Skin/core cure asynchrony becomes the main contradiction; core venting is hard | Staged ramp, extended cure time, pressure-holding cooldown |
| Very thick (roughly >15 mm) | Heat-transfer time dominates cycle time; cooling stress and warpage risk significant | Process levers approach their limit — consider design-side material reduction or a process change |
What the Process Side Can Do
Staged ramping and repeated venting
Splitting the rise into steps with dwell time between them is the most direct way to let the core catch the skin; venting between steps gives gas a way out.
The same 2012 study also reports the opposite result: as venting amount increases, large parts tend toward core cracking (the reported preferred window is 0.20%–0.27%). The shop-floor instinct to “vent a few more times” can split the core outright on a thick section — here the ceiling matters more than the floor.
Staged ramping into post-cure
Post-cure cannot be a single jump either. The same staged ramp plus dwell logic applies there, for the same reason: bring skin and core into the target band together.
The formulation itself also shifts that window. Lyu, Feng, Wen and co-authors report, in “Preparation of vinyl-containing perfluoropolyether-b-polysiloxane and its properties in silicone rubber”, Acta Materiae Compositae Sinica, DOI: 10.13801/j.cnki.fhclxb.20240604.004, that scorch time (t10) and optimum cure time (t90) both decrease as the block copolymer loading rises. Move the formulation and the previous window no longer holds — and thick sections are especially sensitive, because their window was narrow already.
What the Design Side Can Do: Shrinking the Lag Itself
The process side does one thing: it makes the core catch the skin. The design side does another: it shrinks the lag itself. The second usually has more room to work, and it rarely gets written up on its own.
Uniform wall and gradual transitions
Abrupt thickness change is the worst geometry: the core of the thick zone and the skin of the thin zone sit at the same mold temperature yet need completely different cure times. Converting the step into a gradual transition (ramp or radius) keeps the heat path continuous and reduces the skin/core lag.
Coring out and material reduction
Adding weight-relief pockets or cored cavities inside a thick section turns “solid thick” into “thin wall plus ribs.” Wall thickness drops, heat-transfer time drops, and the synchrony problem eases. Of the available changes, this one usually gives the most improvement for the least cost.
The cost: tooling gets more complex, and coring changes overall stiffness — recheck assembly and service loads.
Where inserts sit changes heat flow
Metal inserts conduct heat far faster than silicone, so compound around them heats and cures early, while the far side can become a shadow zone that lags. The closer an insert sits to the center of a thick region, the more it disrupts the temperature field.
Five DFM recommendations
| What to change | Why it works | Cost / caution |
|---|---|---|
| Convert abrupt thickness steps into gradual transitions | Continuous heat path; smaller skin/core cure lag | May alter cosmetic parting lines; recheck assembly clearances |
| Core out thick regions / add weight-relief pockets | Effective wall drops; heat-transfer time falls sharply | More complex tooling; reduced stiffness needs assessment |
| Distribute inserts symmetrically, away from thick-zone centers | Prevents metal inserts from disrupting the temperature field and creating shadow zones | May be constrained by functional layout |
| Thinner wall wherever stiffness allows | Wall thickness is the most direct difficulty variable | Stiffness and service life must be revalidated |
| Rib a large flat area instead of thickening it | Trades structural stiffness for wall thickness, avoiding solid thick zones | Ribs may cause sink marks — evaluate together |
When a Thick Section Should Not Be Compression Molded
Change the design first, then talk parameters. In four cases, however precisely you tune the cure curve, it returns less than changing wall thickness or the process route.
- Wall thickness beyond the economical range for compression molding → evaluate liquid silicone injection instead. The two routes are compared in LSR vs HCR Silicone: What’s the Difference and Which Should You Choose?.
- Very low volume plus very thick section → tooling and process-validation cost is out of proportion to output. Do the arithmetic before cutting steel.
- Thick parts needing very tight dimensional consistency → compression molding has limited tolerance capability, and thick sections make shrinkage harder to hold. Change process or relax tolerances.
- Geometry inherently hostile to heat transfer (large solid areas, extreme thick-to-thin ratio) → fix the design first rather than pinning hope on process parameters.
We do not offer simulation-based prediction of the curing process (CAE thermal-field modeling, finite-element simulation). Thick-section windows come from practices reported in the literature, confirmed round by round against actual tooling — we make no simulation commitment.
What to Send With Your RFQ
For thick sections, quote accuracy depends almost entirely on these items:
- Maximum and minimum wall thickness (sets the heat-transfer time scale)
- Whether metal inserts are present (location and material)
- Annual volume (determines whether tooling pays off)
- Tolerance requirements on critical dimensions (align tolerance capability early)
See our sampling and trial-molding workflow, or Contact us with your drawing.
Thick sections are hard for one reason, and it is a chain: wall thickness → thermal gradient → skin/core cure asynchrony. The process side can make the core catch up. The design side can shorten the distance it has to cover. The same change costs one revision on the drawing and ten rounds of trial molding. Whether a thick part can be made: look at wall thickness first, the compound second. Whether the cost comes down afterwards is decided by annual volume. Send the drawing, wall-thickness callouts, and annual volume: within 24 hours you get the call — whether this wall thickness cures through in one shot, whether the drawing needs a change, and which process route fits your quantity.
Shenzhen Lixinyuan Technology Co., Ltd. | Custom silicone OEM/ODM | MOQ 50 pcs | Sampling in 5–7 days | Contact us
References
Process parameters cited here are typical conditions reported in the public literature below, not measurements from our own production; actual runs are validated per your drawing and compound.
- 雷卫华, 石耀刚, 罗世凯, 陈立新, 曹君, 周安伟. 硅橡胶厚功能制品的制备工艺研究[J]. 有机硅材料, 2012, 26(3): 164–169.
- 吕宁宁, 冯裕智, 文敬滨, 温萍, 路明霄, 唐旭东. 含乙烯基全氟聚醚-b-聚硅氧烷制备及其在硅橡胶中的性能[J]. 复合材料学报. DOI
Frequently Asked Questions
Why are thick silicone sections harder than thin ones?+
Silicone conducts heat poorly: by the time the outer skin has cured, the core may not have reached curing temperature, leaving a cured shell over an under-cured centre. Uneven contraction during cooling then adds distortion and internal stress. The thicker the section, the harder cure synchrony becomes.
Any design guidance for thick sections?+
Keep wall thickness as uniform as possible, ramp transitions instead of stepping them, avoid solid masses by coring or using inserts, and never place critical mating surfaces at the thickest point. Validate all of this at sampling stage — far cheaper than changing tooling after production starts.
When is compression moulding unsuitable for thick parts?+
Four cases: extreme wall-thickness variation that cannot be redesigned, strict core density requirements, tolerances beyond compression moulding capability, and thick sections that must also be highly transparent (the interior tends to haze). For these we recommend redesigning or changing process — and we say so up front rather than accepting and reworking later.