Will Adidas Launch Fully 3D Printed Sneakers?

• 1
  Why Fully 3D Printed Sneakers Keep Coming Back
• 2
  How Additive Actually Builds a Shoe Part
• 3
  Material Limits: Fatigue, Rebound, and Abrasion
• 4
  Where CNC Machining Still Beats Printing
• 5
  When Printing Makes Sense and When It Does Not
• 6
  A Checklist for Printed Shoe Components
• 7
  Printing vs Molding vs CNC Machining

Every few years an image of a one-piece shoe circulates and the same headline returns. Will Adidas launch fully 3D printed sneakers? The honest answer is that printing has already replaced some shoe parts, but not the whole shoe. The midsole is the usual target because it is a single flexible geometry with no assembly. The upper, the eyelets, and the outsole are harder problems.

From a manufacturing standpoint the question is not about brand intent. It is about which features survive a layer-by-layer build. A shoe has to bend millions of cycles, hold a foot, resist abrasion on concrete, and come in half sizes. Each of those requirements points at a different process.

We machine and print parts for aerospace, medical, and automotive customers, so we read this topic the same way we read any new design: what is the load path, what is the failure mode, and what does one unit cost at volume. That framing is more useful than speculating about a product launch.

• 1
  The midsole is the real storyIt is a single lattice part, which is exactly what additive does well.
• 2
  The upper is a textile problemWoven and knit uppers already beat printing on cost and breathability.
• 3
  Volume changes everythingAt 10,000+ units, molding usually beats printing on unit cost.

Powder bed fusion, the process behind most printed midsoles, spreads a thin layer of polymer powder and melts a cross-section with a laser or binder. The part grows in 60–120 μm layers. That layer height controls the smallest feature you can resolve and the surface roughness you get. A lattice with 0.4 mm struts is repeatable. A 0.15 mm strut is not.

The lattice itself is the design trick. A solid block of elastomer would be heavy and stiff. A lattice can be tuned cell by cell, so the heel compresses more than the forefoot. FEA drives the geometry, and the printer builds it without tooling. That is why printed midsoles appeared years before printed uppers.

Support structures are the hidden cost. Overhangs inside a lattice need support that must be cut or blasted out by hand. Every gram of support is a gram someone removes later. Designers reduce this by keeping lattice angles above roughly 40° from horizontal.

Post-processing decides the final feel. Printed elastomer parts are usually tumbled, dyed, or coated. Each step adds labor and variation, and variation in a shoe means a customer notices a difference between the left and right unit.

• 1
  Layer height60–120 μm typical; finer layers mean slower builds.
• 2
  Minimum strutAround 0.4 mm for a reliable lattice.
• 3
  Overhang ruleKeep unsupported angles above about 40°.
• 4
  Post-processingTumbling and coating add cost and unit-to-unit variation.

A midsole foam lives a brutal life. It sees millions of compression cycles, ozone, sweat, and road grit. Printed elastomers such as thermoplastic polyurethane handle this better than most people expect, but they still lose rebound over time. The printed lattice helps because cracks stop at a strut instead of running across a solid block.

Abrasion is the harder limit. The outsole contacts concrete and asphalt, and a printed polymer wears faster than a vulcanized rubber compound. That is why printed outsoles usually appear as a segment or a heel plate rather than the whole contact surface.

Heat matters too. A shoe left in a car in summer can see 60 °C or higher. Printed polymers soften and creep under sustained load at that temperature, so a lattice that feels right in the lab can flatten after a season. Molded foam compounds are formulated for exactly this.

None of this makes printing a bad choice. It sets the boundary. Printed parts do best where geometry matters more than abrasion resistance, and where volume is low enough that tooling cost cannot be amortized.

• 1
  FatigueA lattice slows crack growth; solid printed blocks do not.
• 2
  AbrasionPrinted polymer wears faster than vulcanized rubber.
• 3
  HeatAbove roughly 60 °C, creep and softening become real.

Shoe tooling is metal, and metal tooling is machined. Midsole molds, last masters, trim dies, and the aluminum plates inside a heat press all start as a block of 7075 or 6061 that gets cut. That work never goes to a printer because surface finish and dimensional repeatability matter more than shape freedom.

We hold ±0.005 mm on a 5-axis machine, which is the tolerance a mold cavity needs when two halves have to shut off cleanly. We run up to 4,000 mm of travel for large tooling plates. A printed mold would deform under injection pressure and would not survive the clamp force.

The same logic applies to prototype hardware around a shoe project. A wearable sensor housing, a test fixture that flexes a midsole 100,000 times, or a last with a specific heel curve are all machined parts. They need to be one piece of metal, not a stack of layers.

So the practical split is simple. Printing makes the soft, complex, low-volume part. Machining makes the hard, precise, load-bearing part that makes the soft part possible.

• 1
  Mold cavities±0.005 mm and a shut-off that has to seal.
• 2
  Test fixturesSteel and aluminum frames for flex cycling.
• 3
  Large platesUp to 4,000 mm for press and trim tooling.

Printing has no tooling cost. That is its whole advantage at low volume. One pair, ten pairs, or a thousand pairs can come off the same file with no mold to cut. For a limited release or a custom fit program, that math is hard to beat.

At high volume the math flips. Injection molding and compression molding spread the tool cost across tens of thousands of units, and the per-unit cost drops well below a printed part. A printed midsole also takes hours of machine time per unit, and machine time is the real constraint.

There is a middle zone where printing stays competitive: complex lattices that a mold cannot fill, short runs where a mold would take weeks, and designs that change between batches. If your geometry changes every season, printing avoids scrapping a mold each time.

For a buyer, the decision comes down to three numbers: annual volume, how often the design changes, and how much the geometry depends on a lattice. Get those three right and the process picks itself.

• 1
  Low volumePrinting wins: no tooling, no setup.
• 2
  High volumeMolding wins: tool cost amortizes, cycle time drops.
• 3
  Changing designPrinting avoids a new mold every revision.

If you are evaluating a printed component for footwear or any flexing consumer product, run these checks before you commit to a process. They catch most of the failures we see in review.

First, confirm the load path. A part that carries a person's weight in compression is very different from one that carries it in bending. Lattices handle compression well. Thin printed plates in bending delaminate between layers.

Second, set a minimum feature size and hold it. For polymer powder bed fusion, keep walls and struts at 0.4 mm or above. Below that, yield drops and you will reject parts after the build.

Third, plan the post-process before the design freeze. If the part needs dyeing, coating, or a bonded rubber outsole, those steps set the tolerance stack and the final cost.

Fourth, test to the real cycle count. A 10,000-cycle bench test tells you little about a shoe that sees a million steps. Match the test to the product.

• 1
  Load pathCompression favors lattices; bending favors solid sections.
• 2
  Feature floor0.4 mm minimum for polymer powder bed fusion.
• 3
  Post-processDecide dye, coat, and bonding before design freeze.
• 4
  Cycle testMatch bench cycles to the real service life.