Overmolding Design Guide: Bonding and Substrates

Overmolding bonds a second material onto a molded substrate in the same tool, and the joint holds or fails in the first four seconds of the second shot. Design the substrate for mechanical interlock first, choose a bondable pair second, and use adhesion promoters only as a backup — not as the primary design strategy.

Key takeaways

  • Mechanical interlock carries the bond when the two materials are incompatible; chemical bonding only supplements it.
  • Substrates of PP, PE and acetal bond to almost nothing without surface treatment — pick a substrate you can bond before you pick the overmold material.
  • Shore A 40 to 70 TPE grades bond a wide range of engineering plastics at melt temperatures below 230 °C, so they rarely distort the substrate.
  • Overmold wall thickness of 0.8 to 2.5 mm, applied over a substrate edge rather than over a flat face, gives the most bond per unit of cost.
  • Bond quality is measured in peel or pull force per unit of bond line, not in appearance, and it must be measured after thermal cycling.

What overmolding actually does

Overmolding is a two-material sequence inside one tool. The first shot forms a rigid substrate; the tool indexes or transfers, and the second shot injects a thermoplastic elastomer over it. The elastomer cools against the substrate and forms the joint.

Three things happen at that interface, and they are not equally important. The elastomer shrinks onto the substrate and grips it — that is the mechanical component. The polymer chains of the two materials entangle or chemically bond across the interface — that is the chemical component. And the geometry raises or lowers the load a given bond can carry before it peels — that is the design component.

Most failed overmolding projects have optimised the second item and ignored the third. A compatible material pair can still peel if the bond line has no undercut to sit in.

Chemical bonding: which pairs actually stick

Chemical bonding requires chain entanglement across the interface, which requires the two melts to be mutually soluble or closely related. That is why the same family sticks to itself and why completely different families do not.

The practical rule is bond-ability by polymer family. TPE-S and TPE-E bond readily to PP, PP co-polymer and PE, but only when the grade is compounded for those substrates. TPU bonds well to ABS, PC, PC/ABS, PA and PBT. TPE-V bonds to PA, PBT, ABS and PC/ABS. Silicone bonds to almost nothing except silicone and some hard-coated metals.

The problem families are pure polypropylene, pure polyethylene, acetal (POM) and PTFE. Their surfaces have low surface energy, so the elastomer cannot wet or entangle with them at all. An overmold on a pure PP substrate will look perfect on day one and peel in service.

Overmold material Bonds well to Needs primer Typical Shore A Melt temperature
TPE-S PP co-polymer, PE, polar-tagged PP compounds Pure PP, virgin PE 30–75 190–215 °C
TPE-V PA6, PA66, PBT, ABS, PC/ABS Glass-filled PA above 30% fill 50–90 210–235 °C
TPU ABS, PC, PC/ABS, PBT, PA, TPU Acetal, pure PE, PTFE 70–95 (Shore A/D) 200–225 °C
Silicone (LSR) Silicone, primed metal inserts Almost all thermoplastics 20–70 Around 180 °C

Note the melt temperature column. A TPE that needs to run at 235 °C to develop adhesion will locally soften a 1.5 mm PC substrate. Substrate distortion is the second most common overmolding defect after poor bond, and it is caused by choosing the material for adhesion alone.

Selecting the substrate for the bond, not for the part

Substrate selection is usually driven by stiffness, cost and the working environment, then the overmold material is chosen afterwards. Reversing that order is the single most reliable way to get a bonded assembly that passes first article and fails in the field.

If a soft-touch grip is the requirement and the substrate is currently pure PP because it is cheap, the options are narrow: switch the substrate to a PP co-polymer or a TPE-compatible PP grade, add a primer step, or redesign the joint as a pure mechanical interlock. All three are workable, and the material change is usually the cheapest, because the tooling is not yet cut.

If an elastomer grip on an acetal slider, the chemical route is closed. Plan for an undercut, a through-hole, or a dovetail, and accept that the grip will be a clipped-on part rather than a bonded one — which is often the more serviceable design anyway.

If the assembly is a structural bracket in glass-filled PA66 at 30% or more glass, the bond line is not the design driver. Glass fibres do not bond, and the fibre ends exposed at the substrate surface reduce the effective bond area. Increase the interlock depth by 0.5 mm and drop the bond expectation accordingly.

Mechanical interlock: the part that survives

Regardless of how well the materials bond, the joint should fail in the elastomer before it peels off the substrate. That condition is met by geometry, not by chemistry.

The reliable features are simple and cheap to machine into the substrate cavity. A dovetail or trapezoidal groove along the bond line resists peel far better than a straight edge. Through-holes that the second shot fills create discrete pins of elastomer that carry load in shear. A textured or bead-blasted substrate surface raises the effective bond area without changing the tool. And a substrate edge that steps down by 0.3 to 0.5 mm lets the elastomer wrap around the edge and lock in place.

The one feature to avoid is a flat bond face pulling in pure tension. An elastomer on a flat face has one failure mode and it is the fast one. Even a 15-degree undercut turns that into a peel-dominant load, which the same material pair survives at several times the force.

Wall thickness and the geometry that holds

Overmold thickness between 0.8 and 2.5 mm covers most cosmetic and grip applications. Below 0.8 mm the melt front freezes before it reaches the far edge of the flow path, and short shots concentrate at the bond line. Above 2.5 mm the elastomer sinks and the differential shrinkage against the substrate increases, which is where the visible peel starts.

Flow length matters more than thickness. A rough working figure is that the flow length achievable at a given wall thickness is about 100 to 150 times that thickness for a TPE at 210 °C into a 60 °C tool. A 1.2 mm overmold therefore fills about 120 to 180 mm from the gate — past that, add a second gate rather than thicken the wall.

Gate placement deserves its own decision. Gating directly onto the bond line shears the interface and is a common cause of peel at the gate. Gate into a free area and let the melt flow into the bond region from the side.

Surface preparation and primers

Primer is a process, not a material. It adds a solvent flash step, a cure window, an operator skill requirement and a failure mode that only shows up after ageing, so it should be the last design choice for a bonded interface and never a safe fallback.

When chemistry forces it, the sequence is: clean the substrate, confirm the surface energy has been raised by the treatment (flame, plasma or corona, all of which raise the surface energy of the substrate), apply primer within the specified open time, and bond before the primer loses activity. Each of those steps needs a documented check, not a visual one.

The reliable substitute is a deliberate mechanical interlock designed into the substrate. It costs nothing at the tool, it does not age, and it survives the rework loop that a primer wet process never does.

Testing the bond before the tool is cut

Bond strength is not visible. A bonded assembly that looks right can carry a fifth of the load of an equivalent one, so acceptance has to be measured.

The two useful tests are peel and pull, both expressed per unit of bond length or bond area so results are comparable across parts. Peel loads a strip at 90 degrees and exposes the interface weakness directly; pull loads the elastomer normal to the substrate and measures a blend of adhesion and interlock.

Both should be run on samples from the T1 shot, then again after ageing: thermal cycling between −20 and 85 °C for 100 cycles, plus 168 hours at 85 °C and 85% relative humidity for parts going into humid or outdoor service. Interface failure after ageing, with cohesive failure of the elastomer at T1, is the classic sign that the bond was marginal all along.

Our answer: we would rather see a peel test on T1 samples than a material datasheet claim. RCH Plastic reference: mold making in-house with molding run through vetted partners means the T1 samples, the process parameters and the reported bond figures come from the same run, so a marginal bond shows up at the same time as a marginal dimension.

Mistakes that account for most failures

Choosing the elastomer before the substrate. The substrate sets the bond ceiling, and it is the harder of the two decisions to reverse once the tool exists.

Relying on chemical bonding with no interlock. It works in the lab and peels in a dropped product, because the chemical component is the one that degrades with ageing and contamination.

Overmolding across a sharp substrate edge. The edge concentrates stress and the elastomer tears there. A 0.3 mm radius on the substrate edge removes the concentration for no tooling cost.

Skipping the thermal ageing test. Room-temperature peel results are optimistic by a wide margin for anything that leaves the factory.

Related guides in this series

Send the substrate material, the overmold material and the joint cross-section, and our engineers will say in three working days whether the bond is achievable by chemistry alone or needs an interlock designed in. Request a free DFM review.