Ask any supplier how strong the silicone-to-plastic bond is, and you will almost always hear the same two words: "strong adhesion." Ask for a number, and the conversation gets quiet. There is a reason for that — and once you understand it, you will never accept those two words as a specification again.
Bond strength sounds like a single value, the way tensile strength or hardness is. It is not. A silicone-to-plastic bond is the product of at least five variables, and a number that omits them is meaningless.
| Variable | Why it changes the number |
|---|---|
| Substrate | Silicone bonds to PC, ABS, PP, PA66, PBT, POM and metal very differently. PP and POM have inherently low surface energy and are the hardest to bond without a primer. |
| Silicone grade | A self-bonding LSR cross-links chemically to the substrate; a standard grade does not, and needs a primer or a mechanical lock. |
| Bond route | Mechanical interlocking, primer, or self-bonding LSR produce bonds of fundamentally different character and strength. |
| Joint geometry | Contact area, undercuts, grooves and gate placement decide whether the joint fails in peel or shear — and at what load. |
| Test method | Pull-off (Newtons), 90° or 180° peel (N/mm), and lap shear measure different things and return different numbers. |
A supplier who quotes one bond strength figure without naming the substrate, the grade, the bond route, the geometry and the test method is hiding the very conditions that determine the number. That is why "strong adhesion" is so common — it promises everything and specifies nothing.
Before you can discuss a number, you have to know which of two fundamentally different bonds you are buying. They fail differently, and they are verified differently.
Undercuts, holes, grooves or knurled features physically capture the silicone, so the joint holds under peel and shear load regardless of surface chemistry. It is the most process-independent route — which is why CYES makes it the default wherever geometry allows.
Risk profile: fails only if the geometry is wrong or the insert shifts. Predictable, and verifiable by geometry review before molding.
A primer or a self-bonding LSR grade makes the silicone cross-link at the molecular level to the substrate — a water-tight, peel-resistant bond for seal-critical joints. It is the only option for a flush, uninterrupted surface.
Risk profile: depends on a clean, uncontaminated insert, a controlled cure and correct primer coverage. A single dirty insert can produce a silent bond failure.
This is why CYES specifies the bond route per part in DFM, against your drawing and service conditions — and why we never cite a generic capability we have not demonstrated. See the full mechanism in silicone overmolding & bonding.
Suppose a supplier finally gives you a figure. Before you compare it to another supplier's figure, check which test produced it — because the three common methods are not interchangeable.
| Test | What it measures | Unit | Best for |
|---|---|---|---|
| Pull-off | Force to separate the silicone from the substrate normal to the surface | Newtons (N) | Small bonded areas, buttons, sealing faces |
| Peel (90° / 180°) | Force per unit width to peel the silicone back from the substrate | N/mm | Edges, flaps, joints loaded in peel |
| Lap shear | Force per unit area to slide the two materials past each other | MPa (or N/mm²) | Structural joints loaded in shear |
A pull-off force in Newtons and a peel strength in N/mm are measuring different failure modes — one is a point load, the other a line load. Comparing them across suppliers is as meaningless as comparing a mold tolerance to a finished-part tolerance. The number is only useful when the test method and the specimen are named.
Instead of a headline number, CYES runs a real verification loop on the actual bonded part — and reports the measured result per program. This is the difference between a bond that is claimed and a bond that is demonstrated.
Bonded samples are pulled and peeled on a CLD8 5 kN universal testing machine (tensile, elongation, peel and tear) to confirm the joint meets the specification — not just that it looks bonded.
Bonds are tested across the service temperature range and after thermal cycling and aging, so the number you get reflects the field, not a fresh-out-of-the-mold coupon.
The exact grade, primer or self-bonding LSR selection, and the applicable pull / peel / aging test are confirmed per part during DFM — reported honestly, never as a generic claim.
We run these tests on the named equipment we publish — the CLD8 5 kN universal testing machine listed in our testing & measuring equipment — so the verification itself is auditable. The measured result is documented and released with the PPAP where a program calls for it. See manufacturing engineering for the full verification context.
Print these. They will expose more in five minutes than an hour of supplier talk.
Ask for the exact substrate and silicone grade. "Silicone on plastic" is not a material system — PC, ABS, PP, PA66, PBT and POM all bond differently.
Mechanical interlocking, primer, or self-bonding LSR? If they cannot name the route, they have not engineered the joint.
Pull-off, peel or lap shear — and the measured result with its unit (N, N/mm or MPa). A method without a number is a promise, not a result.
Was the value measured on a real molded part or on a laboratory coupon? A coupon number can be flattering and still fail on your actual geometry.
This checklist is the same discipline behind our verified manufacturer policy — every number has a scope, and a number without a scope is how rework and field failures hide.
Tell us the plastic or metal you are overmolding onto — we'll confirm the bond route, the test method, and the measured result for your exact part, honestly documented.
Get a Bond-Verified Quote See Overmolding & Bonding