Metal 3D Printing Applications From Formnext 2024

• 1
  Key takeaways
• 2
  What Formnext 2024 Actually Showed
• 3
  Multi-Material Nozzles, Copper Cores and Thin-Wall Lattices
• 4
  Large-Format Deposition and Binder Jetting for Production Runs
• 5
  Medical Implants, Heat Exchangers and Hybrid Machining
• 6
  How We Take a Printed Part to a Finished Part
• 7
  Which Process Fits Which Part

Formnext 2024 was less about new printer brands and more about what the machines can now produce. The metal 3D printing applications with real traction on the floor were the ones solving a specific manufacturing headache: a mold that runs too hot, a wear part that fails every six weeks, a hydraulic manifold with nine cross-drilled holes that keep leaking. Those are problems engineers recognize, and they are the reason a shop starts looking at additive at all.

Three hardware shifts made those applications practical. First, multi-material deposition heads now switch between two metal powders inside a single build, which means a tool steel surface on a copper core without a brazed joint. Second, large-format deposition cells grew their working envelope past 1 m, so an automotive housing no longer has to be split into four printed segments and welded. Third, laser systems and scan strategies got cleaner, so as-built density on DMLS parts is now consistently above 99.5% when the parameter set is qualified for that alloy.

The counterweight is that none of this removes the need for subtractive work. Nearly every part we saw at the show was a hybrid: printed to near-net shape, then machined on the interfaces. Anyone evaluating these metal 3D printing applications should budget the CNC operation into the part cost from day one, not treat it as a surprise at the end.

The InssTek-style multi-material nozzle was the clearest example of a process doing something machining cannot. A directed energy deposition head feeds two powders into the melt pool and grades the composition across the part. In tooling, that means a copper core for heat transfer with a tool steel skin on the wear face, printed in one continuous operation. The old route was two parts, a braze joint, and a joint that eventually cracks under thermal cycling.

The limitation is resolution. Directed energy deposition leaves a rough surface, typically Ra 8–12 μm before finishing, and the minimum feature you can hold is closer to 1 mm than to 0.1 mm. This is a process for large molds, die inserts and repair work on existing tooling, not for a 4 mm impeller. If your part has internal channels under 1.5 mm or wall sections under 0.8 mm, DMLS is the better route.

Copper cores printed with conformal cooling channels were everywhere at the show, and the math is straightforward. A channel that follows the cavity at a constant 6–8 mm from the surface pulls heat out evenly, so the part cools without the hot spots that cause warpage. Cycle time on a high-run tool drops, and scrap rate drops with it. The trade is that copper alloys for DMLS are still a narrow list, and thermal conductivity after the build depends heavily on the heat treat and on how well the powder bed was qualified.

Thin-wall lattices and topology-optimized brackets were the third cluster. These are parts where the design intent is stiffness per gram: a satellite bracket, a robot end-effector link, a heat sink fin array. Lattice struts down to 0.2–0.3 mm are routine in DMLS. The catch is that a lattice is hard to inspect and hard to clean. Trapped powder inside a closed lattice is a real risk, so add drain holes and specify the escape path on the drawing.

• 1
  Choose DED multi-material whenThe part is large, needs a wear-resistant skin, and has features above roughly 1 mm.
• 2
  Choose DMLS whenYou need fine internal channels, thin walls or lattice struts below 0.5 mm.
• 3
  Budget a finishing passAs-built surfaces run Ra 8–12 μm; sealing faces and bores need machining.

Large-format deposition targets the parts that used to be castings. A housing printed in one piece removes the pattern cost, the draft angle and the four-week tooling lead time. For a low-volume industrial machine or an EV prototype, that is a real advantage. The trade-offs are surface finish, which is rough, and the need for a machining allowance on every mating face. Add 1.5–2.5 mm of stock on faces you intend to mill.

The parts that fit best are boxy, generously sized and low in feature density: motor housings, pump bodies, structural frames, manifolds without fine internal geometry. If the part is the size of a fist with twenty tight tolerances, deposition is the wrong tool. Send it to a 5-axis mill instead. We run 16 simultaneous 5-axis centers with travels up to 4,000 × 400 × 150 mm and a Ø400 mm rotary table, which covers most of the housing work that would otherwise be printed.

Binder jetting took a different path at the show: it is now a production process, not just a prototyping one. Because the laser is replaced by a binder and a sintering furnace, parts come off the line in batches, nested in a build box, with no support structures to cut away. That makes it competitive on small, complex parts at volumes in the hundreds to low thousands. Sintering shrinkage runs about 15–20% depending on alloy, so every dimension on the drawing has to be scaled, and the furnace cycle sets the final tolerance.

The practical limit on binder jetting is tolerance and density. Expect around ±0.3% of dimension after sintering, which is fine for a bracket and not fine for a bearing seat. Porosity is higher than DMLS, so parts that see fatigue loading or pressure containment should be evaluated carefully. For a non-critical housing or a flow component, binder jetting at volume is often cheaper per part than machining from bar.

Porous titanium implants remain the strongest medical case for metal additive. A Ti-6Al-4V lattice with 60–80% porosity mimics trabecular bone, and bone grows into it. No subtractive process produces an interconnected porous structure inside a solid shell. The regulatory side is the real cost: ISO 13485 documentation, process validation, and a locked parameter set. If you are developing an implant, plan the quality system before the geometry.

Compact heat exchangers and cold plates were the other high-value application. A printed cold plate with gyroid or pin-fin internal geometry can pull far more heat out of a package than a drilled plate, because the surface area per unit volume is much higher. The design rule is to keep channels above roughly 0.5 mm and to give the powder a way out. A sealed gyroid core with no escape path will trap powder and fail cleaning.

The last application is not a printed part at all. It is the hybrid cell: a deposition head or a DMLS build followed by in-machine milling, so the part is finished without re-fixturing. This is where a printed part becomes a usable part. Critical bores, seal faces and bearing seats need machined tolerance and a real surface finish, and no printing process delivers that as-built.

At our end, the printed blank arrives and the CNC work starts. We hold ±0.005 mm (±0.0002 in) on critical features and Ra 0.8–1.6 μm on sealing surfaces, with finer finishes down to Ra 0.2–0.8 μm when the drawing calls for it. Aluminium 6061, 7075 and 2024, stainless 316L and 17-4PH, Ti-6Al-4V and Inconel all machine differently after a build, and the heat treat condition matters more than the alloy name on the certificate.