Most small plastic parts get their own mold. Launch a family of five related parts, and you’ve bought five complete tools, each with its own plate stack, ejector system, waterlines, locating ring, and sprue bushing.

None of that hardware is specific to your part. Set it beside the base from another customer’s tool, and you couldn’t tell them apart. A modular mold base lets you buy it once: one hardened frame stays mounted in the press, and each part in the family gets its own tool-steel insert, the core and cavity that carry your geometry.

Where the Money in a Small Mold Actually Goes

The cavity and core — the steel your geometry actually lives in — usually account for about 60% of the quote on a small single-cavity tool, a bit more or less depending on how much detail the part carries. The other 40% or so covers everything else: the plate stack, the ejector pins and return system, the cooling circuits, the sprue bushing and locating ring, and the machining that squares it all up.

Forty percent doesn’t sound like much until you multiply it. Five parts, five dedicated tools, five ejector plate stacks doing the same job. On a five-part family, that’s two whole tools’ worth of spend on generic hardware that never touches a customer.

The base doesn’t know or care what part it’s molding, which is exactly why you don’t always have to duplicate it.

One Frame Stays in the Press

The mechanics are simpler than the name suggests. A master unit die—the DME MUD system most shops build to—is a hardened frame with its own ejector plates, waterlines, and clamp slots. You mount it in the press once. Each part in your family lives in an insert unit, cavity and core, plus its own runner and gating, that slides into the frame’s pocket and locks in.

Pulling a full tool out of a press and hanging the next one generally costs an hour of press time before you get a first shot, and that’s assuming the day goes well. With inserts, the frame never leaves the press, so there’s no crane to book, nothing to re-square, and nothing to re-plumb beyond a couple of quick-disconnects. An operator lifts one insert out, sets the next one into the pocket, and the swap usually takes under five minutes. People tend not to believe that part until they’ve watched it.

We build the inserts in tool steel — P20 for most work, H13 or S7 where the resin is abrasive, or the geometry takes a beating. Aluminum inserts are cheaper up front. They also wear out, and a wear-out on an insert you planned to run for a decade is a second tooling bill you didn’t budget.

Five Parts, Three Years, Three Ways to Pay for It

Take a five-part family in glass-filled nylon at about 3,000 units per part per year — 9,000 shots per part over three years, 45,000 total. Call a dedicated single-cavity steel tool for one of those parts one unit of tooling spend (real dollars land in the tens of thousands and vary enough by part that we quote them rather than blog them). Stack the three routes side by side:

  • Dedicated steel, one tool per part: 5.0 units. Each of the five arrives with its own base, ejector system, and set of waterlines, and the shop then has to set up, maintain, and shelve each one.
  • Aluminum tools: about 0.5 each, so 2.5 to start. Glass fill and other abrasive resins spoil the arithmetic, since they begin wearing aluminum cavities somewhere in the 5,000 to 10,000 shot range. Over 3 years, budget at least one rebuild per part at roughly 0.35 per rebuild, which puts the total near 4.25 units — and that figure says nothing about the weeks a tool sits out of the press waiting to be rebuilt.
  • One modular mold base plus five steel inserts: the base at about 0.4, inserts at about 0.6 each. Total 3.4 units.

The shared-base route costs about a third less than dedicated steel and gives you steel wear life. A P20 insert runs 300,000 to 500,000 shots before it needs attention; H13 goes past a million. At 3,000 a year, that’s tooling that outlives the product it was cut for.

Most product families, at launch, sit in the band where sharing a base is the better buy. That band has an edge on either side. Low volumes rarely support even an insert—at a few hundred units a year, validating with a 3D-printed part or a bridge tool is usually cheaper than cutting one.

The upper edge depends on how much of the press a single part occupies. Once a part runs around 20,000 to 30,000 units a year, a dedicated multi-cavity tool will generally earn back its cost in cycle time.

Where Shared-Base Tooling Breaks

The pocket sets the envelope and shut height in the frame, and that ceiling is firmer than most people assume. A part that grows past the insert footprint doesn’t get a bigger insert; it gets its own tool. Space is short for other things as well — water has to be routed somewhere, a slide has to seat somewhere, and a unit built to come out of the press in five minutes doesn’t leave much room for either, so side actions and complicated cooling layouts run into the same limit.

Serialization is the more common day-to-day problem, because only one part in the family can run at a time. If two parts spike in the same month, both wait on the same frame, and one ships late. The alternative, a family mold, avoids the queue by molding every part on every shot whether you need them or not, and fill balance in those tools is usually poor.

Resin is the other practical limit. That five-minute swap only holds up if the barrel doesn’t have to be purged and brought to a different temperature, so it’s worth grouping the family by resin, or at least sequencing the changeovers with resin in mind.

Who Owns the Inserts When You Leave

We build to a standard DME frame size, and the inserts belong to you, cut to dimensions that any other shop running that same frame can take.

If the program ever moves, the inserts go along with you, and the handover looks like any other mold transfer — the Build-To files and process sheets come along too.

Not every shop sets it up that way. Some build on proprietary bases, and an insert cut for a proprietary frame won’t fit anything else, so the tooling is yours in name but stays where it is, and your spot in that shop’s changeover queue stays with it. That’s worth asking about before steel gets cut, because the same question asked later comes with a much bigger bill attached.

When a Part Has Earned Its Own Tool

Press time is the number to watch. Once a single insert is taking up a third or more of the hours the frame has available, or the forecast has one part heading past roughly 20,000 a year, it has outgrown sharing. It is ready to graduate to a dedicated multi-cavity tool.

The gate location that filled cleanly, the cooling layout that held dimension, the shrink the resin really gave you, the window the press settled into once the job was stable — by the time a part is busy enough to move, the insert has handed you all of it, over thousands of production shots. None of that was available the first time around, when the design had to lean on a mold-flow estimate and some educated guessing.

So the dedicated tool is the easier of the two designs, and that is the second way the insert phase pays for itself. Drawn from a real record, it tends to need fewer revision loops and reaches first article sooner than the original did.

What does a 3D-printed prototype tell you about any of that? Fit and feel. The insert is the first tooling that tells you how the part molds.

Buy the base once. Buy the cavities as your product line earns them. Move a part off the frame when it’s busy enough to deserve its own tool, and design that tool from what the insert already taught you. Roughly 40 cents of every tooling dollar on a product family goes to duplicating steel that never touches your part — the modular mold base is how you stop paying for it five times.

If you have a part list and a volume forecast in hand for a product family, send the Build-To files and the forecast through our quote page. We’ll compare shared-base against dedicated for your own parts and tell you where each one lands on the crossover, including any that would be better off skipping the frame altogether.