What Is Overmolding: A Second Shot Over a Shell
Home
Yiwu, Zhejiang, China
Mon – Fri 9:00 AM – 6:00 PM · Sat 9:00 AM – 4:00 PM
[email protected]

What Is Overmolding: A Second Shot Over a Hard Shell

Alan Xiao, Founder of Sourcing Pioneer
Alan Xiao
· Updated November 6, 2026 · 9 min read

Overmolding moulds a second material over a part that already exists. The hard shell gives the shape and the strength. The soft layer gives the grip, the seal or the colour. Whether the two stay together is a question of heat, not glue.

That last point is where most quality arguments start.

A buyer sees a soft grip lifting at one corner after six months. The factory says the material was wrong. The buyer says the process was wrong. Both may be right, because the bond is decided inside the tool in about a second.

This page covers the three ways to make a two-material part. It also covers the one temperature that decides the bond, and what to write on a drawing before tooling is cut.

Warm beige studio still life of a hard plastic shell with a soft grip overmolded on one edge
Hard core, soft edge. The join line is where the warranty claim will come from.

The Short Answer

Overmolding is a two-step moulding process. A rigid part, called the substrate or insert, is made first.

That part is then placed into a second tool. A second material is injected over it and takes the shape of the remaining cavity.

The result is one part with two materials. The soft material usually holds the part, seals it, or makes it look better.

The substrate can be plastic, or metal. Insert moulding around a machined metal boss is common, because it gives the finished part a thread that can be screwed and unscrewed without stripping.

The soft layer is usually a rubber-like plastic, often called a TPE. That stands for thermoplastic elastomer, which simply means a plastic that bends like rubber and can still be melted.

Three Routes to a Two-Material Part

Not every two-material part is made the same way, and the price gap between the routes is large.

RouteHow the part is madeToolingWhere it fits
Insert overmoldingSubstrate is moulded first, then loaded by hand or robot into a second toolTwo standard toolsLow to medium volume, or a metal insert
Two-shot mouldingBoth materials in one machine, the tool rotates or slides between shotsOne complex tool, a two-barrel machineHigh volume where labour per part matters
Co-injectionTwo materials injected into the same cavity at once, usually skin and coreOne tool, special nozzleOne material hiding inside another
Assembly insteadTwo separate parts joined by adhesive, a clip or ultrasonic weldingTwo simple toolsLarge parts, or pairs that will not bond

The first route is the one most Chinese factories will quote by default, because it uses ordinary tools. A worker loads the substrate, the press closes, and the soft layer goes on.

That manual step is also its weakness. Cycle time grows, and part-to-part consistency depends on the loader placing the substrate in the same spot every time.

Two-shot moulding removes the loader. The tool itself moves the part from one cavity to the next, so the second shot lands on a surface that is still hot.

That detail matters more than the labour saving, and the next section explains why.

The fourth route is the one nobody quotes and everybody should consider. If two materials will not bond, gluing or clipping them is cheaper than a tool that cannot hold them.

The Bond Is Decided by One Temperature

Here is the fact that explains most overmolding failures, and almost no supplier quote mentions it.

For the two layers to weld, the surface of the substrate has to reach its own melting point. Not soften. Melt.

A published study on polyamide overmolding measured this directly. The grade used melts between 218 and 223 degrees.

The injected material arrived at 260 to 280 degrees. That sounds hot enough. But the hot melt meets a cold wall, and the wall pulls heat out of it fast.

The authors put it plainly: the surface of the substrate must reach its melting temperature to get a weld.

When it does, the two layers share one continuous mass. When it does not, they are only touching, and the bond is a surface effect rather than a weld.

You can see the difference under a microscope. The paper shows two zones from the same test series. One failed with a clean edge, which means the substrate never melted. The other failed with a torn, rough surface, which means the substrate did melt and the two layers were joined.

This is why a rotating two-shot tool beats a hand-loaded insert tool on bond strength. The second shot lands on a substrate that has had no time to cool. Far less heat is needed to bring the surface back to melting.

Diagram comparing a welded joint with a torn surface against an unmetled joint with a clean edge
A torn break means the layers melted together. A clean break means they only touched.

Where the Weld Stops and the Glue Begins

The same study found something a buyer can actually check on a sample.

The bond was not the same across the whole contact area. Welding happened at the corners of the rib. The middle of the plate only had adhesion.

That is a strange result until you think about where the melt front arrives. It reaches the edges of the cavity first, while the centre is still filling. Then packing pressure pushes the part against the cold wall.

So the strongest bond sits where the material ran first, and the weakest sits where it stopped.

For a buyer, this changes what a sample proves. Pulling a soft grip off a sample tells you the average, and the average hides both the good corner and the weak centre.

The practical test is to cut the part, not pull it. A cross-section through the joint shows how much of the contact area actually melted. Two suppliers can both pass a pull test with very different amounts of real weld.

There is an industrial test for the pull side, and it is worth naming in the order. Peeling an overmoulded rib is covered by DIN EN 6033, and the study used it because the results were repeatable.

The Settings That Went the Wrong Way

This is the part of the study that should make a buyer suspicious of confident process advice.

The team varied four settings and expected the usual answers. Three of the four moved the bond in the opposite direction to what they expected.

Raising the melt temperature helped, as expected. Hotter material carries more heat to the interface, so more of the substrate surface reaches melting.

Raising holding pressure made the bond worse. So did holding it longer.

Raising the mould temperature also made the bond worse, and this one has a chemical cause. The warmer the tool, the more residual monomer can migrate to the surface, and leftover monomer in the joint weakens it.

The study measured that monomer content and it tracked the bond strength. Plates with 1.34 percent left-over monomer held best. At 1.82 percent the joints were much weaker.

The lesson is not that pressure is bad. It is that the bond is a heat problem first, and pressure only helps once the interface has actually melted.

A supplier who promises a fix by turning up the packing pressure is guessing. The lever that works is the interface temperature.

What Makes a Good Material Pair

Material choice is a pair question, not a single-material question, and this trips up buyers who change one side.

One rule of thumb is that the substrate should melt at a higher temperature than the layer going over it. Otherwise the second shot can soften the part it is meant to grip.

Beyond that, chemical likeness decides whether the two can weld at all. Two grades of the same plastic weld easily. A soft plastic over a very different hard plastic often cannot, and needs a mechanical grip instead.

Note what the study’s own setup was. Both sides were the same polyamide, and even then the weld only formed at the corners.

That is the honest benchmark for a dissimilar pair. If a same-material joint is only partly welded, do not accept a promise that an unlike pair will hold by chemistry alone.

When chemistry will not do it, the design has to. That means holes, slots or an undercut that the soft layer flows into and locks behind. The soft material then holds by shape, not by adhesion, and a shape survives a bad day in the oven.

Diagram of a mechanical undercut and through hole locking a soft overmold layer against a hard substrate
A through hole or an undercut holds by shape. It keeps working when the chemistry does not.

Seven Lines for the Drawing

“Soft touch grip” is not a specification. These seven lines make it one, and each closes a question the factory would otherwise answer by habit.

  1. Substrate material and grade. Name the polymer and the grade, not just “ABS”. Two grades of the same polymer can differ in how they take a second shot.
  2. Overmold material and hardness. Name the family and a hardness figure on the Shore A scale, which is the soft-side scale used for rubber-like plastics.
  3. Bond method. Say whether the layers must weld by chemistry, lock by shape, or both. Then say where the mechanical lock sits.
  4. Soft layer thickness. Give a target and a minimum. A layer that is too thin will not fill the lock and will tear at the edge.
  5. Tooling route. State insert or two-shot, and state who pays for the loader and the second tool.
  6. Bond test. Name the peel test and the acceptance figure. Ask for a cross-section as well, because a peel figure alone can be met by a part that is only welded at the corners.
  7. Colour, gloss and texture. Give a colour reference and a texture grade, because the soft layer is usually the part the customer touches first.

Line six is the one worth arguing about before the tool is cut. A peel figure is a number, and a number can be reached by two very different parts.

Line three decides the risk. A design that relies on chemistry alone has no fallback if the pair turns out not to weld. A design with a lock still holds if the weld is weak.

One TPE compounder’s own design guide puts the workable wall band at 0.71 to 3.81 mm. It also says the soft layer should not change thickness by more than four to one. It also caps the flow length at 150 times the thickness. Those are the numbers to put behind line four.

Where the part has to take a thread, the substrate may be metal. The rules for that are the same as any other insert. The tooling logic is set out in the guide to design for manufacturing and tooling.

Where It Fails in the Field

These are the faults that show up after shipping, and what each one usually means.

What the customer seesUsual causeWhat to check
Soft layer lifts at one edgeThe interface never reached meltingCross-section the joint, do not just pull it
Holds at the corners, loose in the middleMelt front reached the edges firstGate position, and whether packing helped or hurt
Soft layer tears off in useNo mechanical lock, chemistry doing all the workLook for a hole, slot or undercut in the drawing
Substrate warps or sinksSecond shot ran hotter than the substrate can takeThe melting points of both materials
Colour drifts between batchesRegrind share changed, or the soft layer was overbakedAsk for the regrind percentage in writing
Flash at the join lineTool shut-off has wornTool age and the number of shots already run

Two rows matter more than the rest. The first and the third.

A part that lifts at an edge may still pass a pull test, because the corners carry the load. And a part with no lock has nothing left once the weld ages.

Field failures in soft-touch parts also reach the electrical side. A grip that lifts on a powered product exposes a joint, which is why a test report for the finished part belongs with the drawing. The testing sequence for a finished product is set out in the guide to the test report a finished part needs.

What the Tool Costs, and Why

Tooling is where the three routes diverge in price, so it is worth knowing what the money buys.

A steel tool is built to run for well over a million parts, and the tool itself can cost hundreds of thousands of dollars. An aluminium tool is far cheaper, but it wears and it will not hold tight dimensions at volume.

A two-shot tool adds a rotating or sliding core and needs a machine with two barrels. That is why the quote is higher than two separate tools, even though it removes the loading step.

For a first run, the cheaper answer is usually an insert tool and a hand-loaded substrate. The cost of an injection mould in China sets the base figure, and the two-shot premium is added to it.

Volume is what decides it. Below a few tens of thousands of parts, the loader is cheaper than the rotating core. Above that, the labour per part starts to beat the tooling difference, and the bond is better too.

One detail from the process itself explains why cycle time is so short. The cavity is filled to about 95 to 98 percent before the machine switches from speed to pressure control. The shot itself often takes well under a second. Everything about the bond is decided in that window.

How to Compare Two Quotes

Four questions separate a quote you can hold a factory to from one you cannot.

First, does it name the tooling route and say who loads the substrate? A quote that leaves this open is not comparable with one that does not.

Second, does it give the hardness of the soft layer? Without a Shore A figure, two suppliers can both say “soft touch” and deliver very different parts.

Third, does it offer a bond test, or only a visual check? A visual check will not catch a joint that is welded at the corners and glued in the middle.

Fourth, is trial and validation in the price? First-shot samples are the cheapest part of a project and the most expensive thing to skip.

After those four, price and lead time belong in the comparison. A soft grip that lifts in the first year costs more than the gap between two quotes.

If the soft layer has to be a specific feel, the material family is the next decision. The trade-offs between the soft plastics are set out in the guide to soft material choices. The same logic applies to a grip as to a toy.

For a part that has to survive outdoors, the shell around the grip matters as much as the grip. Powder coating and anodizing are the two usual answers on the hard side. The difference between them is set out in the powder coating explainer.

THE WELD LINE

Ask for a Cross-Section, Not a Pull Test

Send me the drawing, both materials and where the part will be used. I will tell you whether the pair can weld, where a lock is needed, and which tooling route fits your volume. That costs nothing before the tool is cut.

Send the Drawing
Is overmolding the same as insert molding?

Insert moulding is one form of it. The insert is the part you place into the tool, and it can be plastic or metal. Overmoulding is the wider term for moulding a second material over an existing part.

Does overmolding need adhesive?

Not if the two materials weld or lock together. Adhesive appears when the pair will not bond and the design has no mechanical grip, and it adds a manual step that is hard to inspect.

Why does my soft grip peel at the edge?

Usually the substrate surface never reached its melting point, so the two layers only touched. A cold tool, a cold substrate or a thin second shot all cause it.

Can any plastic be overmolded?

The substrate can be most rigid plastics, or metal. The soft layer is usually a thermoplastic elastomer. The question is never one material, it is whether the pair welds or needs a mechanical lock.

Is two-shot molding worth it?

Above a few tens of thousands of parts, usually yes. It removes the manual loading step and the second shot lands on a warmer substrate, which gives a better bond.

How thick should the soft layer be?

Thick enough to fill any lock and to resist tearing at the edge. Ask the moulder for a minimum, then check it on a cross-section rather than on the outer surface.

Questions & Comments

We read every one and reply within 24 hours

A clean break at the joint means the substrate never melted. Tell me both materials and where the part will be used, and I will say whether the pair can weld at all.

Subscribe
Notify of
guest
0 Comments
Most Voted
Newest Oldest
0
Would love your thoughts, please comment.x
()
x