An injection mold is a precise metal block that splits into two halves. The cavity forms the outside of your part and the core forms the inside — and the gap between them, when the mold closes, is your part's shape, cut slightly oversize to allow for shrinkage. Knowing how they split explains why certain features drive tooling cost.
| Cavity (A-side) | Core (B-side) | |
|---|---|---|
| Forms | Outer surfaces — the faces you see | Inner surfaces — the hollow underside |
| Usually | The fixed half; the cosmetic show side | The moving half; holds the ejector pins |
| Marks | Kept clean for appearance | Carries the ejector-pin witness marks |
Molten plastic is injected into the closed gap, fills the shape, and cools. Then the mold opens and the part comes off. A quick example: on a simple cup, the outer wall and the base's outside are formed by the cavity; the inner wall, the inside of the base, and any internal ribs or bosses are formed by the core. Picturing which features sit on which half is how you anticipate draft, ejection, and the parting line.
As the plastic cools it shrinks and grips tightly onto the core — the inner form. Molds are deliberately designed so the part stays on the core (the moving B-side) when the mold opens, because that's where the ejector pins live. The pins then push the part off. That's the reason the core side usually carries the small round ejection marks, while the cavity (A-side) is reserved for the cosmetic surface. A buyer-useful point: if your show surface is on the face that would normally sit against the core, tell your toolmaker early — the mold can often be arranged so the cosmetic face is on the cavity side and the ejector marks land somewhere hidden. (It's also why those surfaces need draft to release.)
Where the cavity and core meet is the parting line — the seam around your part when the mold closes. It leaves a faint witness line, so it's placed on a neutral edge where a small seam won't hurt looks or function, and kept off sealing or sliding surfaces. Where it falls is a design decision made before the tool is cut.
A simple shape that pulls straight off the core is inexpensive to tool. Features that trap the part can't release from a plain two-part cavity and core, and each one needs extra tooling:
Each adds machining, mechanisms, and maintenance — so the more the geometry fights a straight pull, the higher the tooling cost. (See what drives mold cost.) A good DFM review suggests small changes that simplify the cavity and core without hurting the product.
The two halves are not interchangeable, and knowing which side is which explains a lot of mold behavior. The cavity (the female side that forms the outside/cosmetic surface) usually lives on the stationary half, bolted to the fixed platen where the sprue and the machine nozzle feed it. The core (the male side that forms the inside) usually lives on the moving half, which carries the ejector system. There is a deliberate reason for that split: the part shrinks onto the core as it cools, so you want the part to STAY on the moving half when the mold opens, where the ejector pins can push it off. If a part comes out on the cavity side instead, it is a fault to chase — usually an undercut or a finish that is gripping the wrong half.
The single fact that drives ejection design is that plastic shrinks toward its own center as it cools, so it clamps down onto any male feature inside it. A deep box shrinks tight around the core like a hand closing on a fist; a flat plate barely grips at all. That is why tall, deep, thin-walled parts need generous core draft and strong ejection, while shallow parts are forgiving. It is also why core cooling matters so much: a core that cannot shed heat stays hot, the plastic around it shrinks unevenly, and you get warp plus a part that fights ejection. The core is the hardest part of many molds to cool, which is why baffles, bubblers and beryllium-copper inserts show up there.
The cavity and core are the expensive, precise heart of the tool. They are cut from hardened or pre-hardened steel (or aluminum for lower volumes), roughed on a CNC mill, then finished — often by EDM (electrical discharge machining) for sharp internal corners and deep ribs a cutter cannot reach, then hand-polished or textured to the required finish. Most undercuts force a moving component into the core or cavity — a side-action for external features, a lifter or collapsible core for internal ones — unless a pass-through window can form them, and each of those is more steel to cut, more to go wrong, and more cycle time. This is the real reason a seemingly small geometry change (an internal snap, a side hole, a threaded boss) can move tooling cost and lead time: it changes what the cavity and core have to do mechanically, not just their shape.
A production mold rarely has one cavity and one core. To make parts faster, the same cavity/core pair is replicated — two, four, eight, sixteen "up" — fed by a balanced runner so every cavity fills at the same time, which is what keeps the parts identical. A family mold puts DIFFERENT cavities/cores in one tool to make a set of related parts in one shot, which saves tooling money but gives up some control because the parts compete for the same fill. How many cavities, and whether to run a family tool, is a volume-and-economics decision — our cavity count calculator works the throughput side of it.
You do not have to design the cavity and core — that is the toolmaker's job — but the choices that make them cheap are yours: put the parting line where it is natural, avoid internal undercuts you do not need, give inside walls enough draft to release from the core, and keep the wall uniform so the part shrinks evenly onto the core. A design-for-manufacture review catches the features that would force extra core/cavity mechanisms before any steel is cut, which is exactly where tooling money is saved.
What are the cavity and core? The two mold halves: the cavity (A-side) forms your part's outside, the core (B-side) forms its inside; the gap between them is your part's shape (cut slightly oversize for shrinkage).
Why does the part stay on the core? It shrinks and grips the core, and the mold is built so the part stays on the moving B-side where the ejector pins push it off.
What's the parting line? The seam where the two halves meet; it leaves a faint witness line, so it's placed on a neutral edge.
Why does geometry drive cost? Undercuts and trapped features need a slider, a lifter, or an unscrewing core beyond a simple cavity and core, which adds tooling cost.
How does this relate to draft/ejection? Both halves need draft to release, and the part is ejected off the core by pins on non-cosmetic faces.
JCH Design & Manufacturing can review your part to show how it will split into cavity and core, where draft, the parting line, and ejection fall, and which features drive tooling cost — with small changes to simplify it. Send your part; NDA first.
Book a 30-minute call