ELASTIUM Launches 3D Printed Sneakers Made From Foam Particles

• 1
  How 3D printed sneakers bond loose foam beads
• 2
  What geometry a 3D printed sneakers midsole can hold
• 3
  Which elastomers work, and which do not
• 4
  When the foam part needs CNC machining instead
• 5
  Practical rules before you commit a file
• 6
  Printed foam vs CNC machined parts on the same shoe

Foam-particle printing, often called particle-bed printing, does not melt a filament onto a plate. The printer spreads a thin layer of elastomeric beads across a build bed, then a printhead deposits a binder or a solvent that makes the beads stick where the part should be. The bed drops, a fresh layer spreads, and the cycle repeats.

The bond here is point contact between beads, not a solid melt pool. Each bead keeps most of its own shape, so the finished part is roughly 60-80 percent bead and 20-40 percent void, depending on bead size distribution and how hard the roller compacts each layer. That void fraction is what gives the part its spring.

After printing, the green body is cured. Cure temperature and time set the final stiffness more than the printer settings do. A part cured at the low end of the window stays soft and rebounds slowly. Push the temperature up and the beads fuse deeper, so the part gets stiffer and loses some of its cushioning.

This is why two midsoles printed from the same file can feel different. The file controls geometry. The cure schedule controls feel. Anyone specifying a printed foam part should lock both, or the second run will not match the first.

Particle-bed printing handles lattice and gyroid structures that would be impossible to mold. Internal ribs, graded density, and undercuts all come free, because there is no tool pulling out of the part. If you want a midsole that is stiff at the heel and soft at the forefoot, you change the cell size in the file.

The catch is resolution. A bead is typically 0.3-1.0 mm across, so any feature smaller than roughly three beads will not form cleanly. Thin walls under 1.5 mm tend to break during depowdering. Sharp internal corners trap loose beads that are hard to clear.

Dimensional tolerance is the bigger limit. Printed foam parts usually land in the ±0.3 to ±0.5 mm range on outer dimensions, and worse on thin sections, because the beads shift during spreading and shrink a little on cure. That is fine for a midsole surface. It is not fine for a mating bore or a locating shoulder.

Surfaces come out matte and slightly pebbled, roughly Ra 6-12 μm. If the part needs a smooth cosmetic face or a printed graphic that reads cleanly, it will need a skin, a coat, or a machined insert.

Most printed foam midsoles use thermoplastic polyurethane beads, EVA-based beads, or a PEBA-type elastomer. TPU is the common choice because it bonds well with solvent binders and holds rebound over many compression cycles. EVA is cheaper and softer but loses height faster under repeated load.

PEBA beads give the best energy return per gram, which is why they show up in performance running shoes. They also cost more and need tighter process control, because the same bead batch can print differently if humidity in the shop moves.

Not every elastomer prints. Beads that are too tacky clog the roller and bridge across the bed. Beads with a wide size spread pack unevenly and leave density gradients that show up as soft spots. A tight particle size distribution matters more than the chemistry label.

Silicone and natural rubber beads are generally not used in this process. They either cure too fast in the bed or will not hold a binder bond. If a design calls for those materials, molding or casting is the better route.

Printed foam is good at cushioning and bad at interfaces. The moment a part needs a threaded insert, a press-fit bearing seat, a flat mounting face, or a bore that must hold ±0.05 mm, printing alone will not get there. That is where we come in.

One common route is a hybrid: print the compliant body, then machine the hard interface features into it or bond a machined insert into it. We run 127 high-precision CNC machines at our Dongguan and Singapore plants, including 16 simultaneous 5-axis machining centers, so a printed midsole can be trimmed, faced, or pocketed to a datum.

We hold ±0.005 mm (±0.0002 in) on machined metal and plastic features, with surface finishes from Ra 0.2-0.8 μm on fine work up to Ra 1.6-3.2 μm as-machined. If the shoe needs an aluminum heel clip, a titanium eyelet plate, or a mold for the foam itself, that geometry is a machining job.

Material choice follows the load. Aluminum 6061-T6 for clips and plates, 7075 for high-stress brackets, 17-4PH stainless for wear surfaces, POM or PEEK for low-friction inserts. We quote from one prototype to 10,000+ part runs, with no minimum order quantity.

Keep the smallest feature at three beads or larger. If your bead is 0.5 mm, do not draw 1 mm ribs. They will not survive handling, let alone a wear test.

Design a flat datum pad on any face that will later be machined. A printed surface is not a reliable datum, so give the machinist a clean plane to pick up. Without it, the first operation becomes guesswork.

Plan for shrinkage. Print a test bar first and measure it after the full cure, not straight off the bed. Cure shrinkage of 0.5-2 percent is normal, and it is not linear across a graded-density part.

Keep wall thickness as uniform as you can. Thick sections cure slower in the middle than at the skin, which builds internal stress and can bow a long midsole. Split thick regions with a lattice instead of solid fill.

Decide early whether the part is cosmetic, structural, or both. A part that is both will almost always need a secondary operation, and that operation should be in the plan before the file is frozen.