Mold and tooling: the sector where the bill comes due first

If we had to choose just one sector to start with in metal additive manufacturing, it would be this one. And not because of industrial romanticism: because of simple math.

In almost every application, the advantage of printing a part is that that part is better or cheaper. This isn't the case with molds. With molds, the advantage is that all the parts produced from that mold for the next few years are made faster. The savings don't happen once; they happen with every cycle, which is why it's the only sector where an expensive printed part can pay for itself in weeks.

The four families that work

Conformal Cooling Inserts. The channels follow the surface of the cavity instead of crossing the block in a straight line. This is the star application and has its own article, which considers whether or not it's worthwhile. What's important here: the entire mold isn't printed, only the insert — the hot zone insert, which is usually a small fraction of the whole.

Add-ons and slides in difficult areas. A deep rib, a fine pin that heats up and marks the part, an area with a weld line. Small parts with complex geometry that require electrical discharge machining and multiple clamping points.

Clamping and grippers. Here the advantage is not thermal but of mass: a robot gripper lightened with a lattice structure accelerates each handling cycle and reduces the load on the arm. And grippers are usually unique, free-form parts, made for a specific reference—which is the definition of a part that doesn't warrant machining tooling.

Nozzles, sprues, and hot runner elements. Internal geometry, thin walls, and a surface-to-volume ratio that machining doesn't handle well.

Why does the account appear first here?

Four conditions occur simultaneously, and it is rare for them to occur together:

The part is expensive anyway. An insert machined with electrical discharge machining (EDM) isn't cheap. The comparison isn't with a ten-euro part, it's with one costing several thousand.

The quantity is one. There is no series that amortizes tooling or programming. This is the case of the print or machine manual.

Geometry is impossible otherwise. A channel that follows a curved surface at a constant distance cannot be drilled. It is either additive, or it does not exist.

And the savings are recurring. Here's the difference with everything else. If the insert reduces cycle time, those savings are multiplied by all the parts produced in a year, and by all the years the mold lasts.

From there comes the only calculation that needs to be done, and it can be done on a napkin: cycle time saved in seconds × parts per year × machine hourly cost. If that exceeds the additional cost of the insert in less than a year, it's a done deal.

What alloy

In practice, maraging covers most plastic injection molding applications and is the best place to start. The complete catalog is available at materials.

The four design decisions that determine whether it lasts

1. How much wall do you leave between the runner and the cavity? This is the crucial decision. Too little and the wall will buckle under injection pressure or leave marks on the part; too much and the runner will stop cooling, meaning you've paid for a printed insert for nothing. The typical balance is around two to three times the runner diameter, measured from the cavity surface.

2. That the powder can escape. A conformal circuit is exactly the case of an internal channel manual with dead spots. If the powder remains inside, the insert won't cool, and there's no way to know until the mold malfunctions. Layout without blind low spots, outlets at both ends, and verification before assembly. The rules are in design rules for LPBF.

3. Which faces are machined, decided from the CAD. The cavity is always machined and polished. The bearing faces and the parting plane are also machined and polished. The rest are left rough. With its planned allowance, according to tolerances and finish.

4. The threads of the water connections, afterwards. Printed threads do not withstand tightening or seal. The hole is printed with the drill bit size and tapped during machining.

The cheapest way in: the hybrid insert

This is the practical aspect that is least talked about and the one that most lowers the barrier to entry.

It's not necessary to print the entire part. You start with a conventional machined base—a standard, inexpensive block of steel with no geometry—and only the part with the complex geometry is printed on top of it. The machine builds on this base, which acts as a platform.

This saves most of the manufacturing time, which is the main driver of cost, and reserves additive manufacturing for where it truly makes a difference. In many inserts, the area requiring conformal cooling represents twenty percent of the height.

It is also the gentlest route for a shop floor that wants to try: less powder per part, less machine time, less risk if it goes wrong.

What to do next, without skipping anything

The post-processing of a mold is not optional and its order matters:

1. Stress relief with the part still on the platform. If cut beforehand, it will warp — warp and stress relief.
2. Vampire cutting and support removal.
3. Maraging treatment, in the case of maraging, to achieve mold hardness.
4. Machining of the functional faces and cavity. After treatment, never before.
5. Polishing of the cavity, to the finish required by the plastic part.
6. Circuit leak test, under pressure. This is not a formality: it is a check that the powder has been expelled and that there is no through-porosity. An insert that leaks water inside a mold is an expensive problem.

When not

For the sake of honesty, and because it's what makes the rest useful:

• Large, simple molds. If straight cooling is working properly, there's nothing to gain.
• Plastic parts without thermal issues. If the cycle isn't limited by cooling, there are no savings, no matter how many conformal channels you add.
• When the bottleneck is elsewhere. If the cycle is dictated by extraction or handling, the mold isn't the problem.
• Very short production runs. If the mold is only going to make two thousand parts, there aren't enough cycles to recoup any investment.

How we approach it

If you have a mold with a heated zone that marks the cycle, the useful conversation doesn't start with the machine: it starts with how many seconds you think you can shave off and how many parts you make per year. With those two numbers and the geometry, the decision can be made in an afternoon.

The quote calculator gives the order of magnitude of the insert. And at the open days you can bring the part that's limiting your cycle and see it manufactured.

—

The complete account for conformal cooling is in conformal cooling. The dimensional rules are in design rules for LPBF. And what can be ordered in rough and what needs to be machined is in tolerances and finish.