3D Printed Orthotics: The Most Promising Projects of 2023

A traditional ankle-foot orthosis (AFO) starts from a plaster cast, then a technician heats and drapes a sheet of polypropylene over the positive mold. That process works, but it is labor heavy and the result depends on the technician's hands. A 3D scan of the limb produces a point cloud in seconds. Software trims the shell, sets wall thickness, and adds lattice where stiffness is needed. The printer then builds the part in one run.

The shift matters most for irregular anatomy. A diabetic foot with a Charcot deformity, a pediatric limb that changes shape every few months, a post-stroke patient with spasticity on one side: these are cases where a stock size never fits well. Printing lets the clinician change one variable, re-scan, and re-print without new tooling.

Cost structure is different too. There is no mold to amortize, so a one-off part is economically sensible. But the per-part cost does not fall as sharply as people expect at higher volumes, which is why printed and machined parts often coexist in the same product.

Lower-limb projects dominated the year. Several groups combined a handheld 3D scanner with FDM printing of PA 12 or TPU to produce AFOs and foot orthoses for diabetic ulcers and traumatic injury recovery. The common technical thread: a shell whose thickness varies by zone, thinner under the arch and thicker at the heel counter, which is hard to do with a draped sheet but trivial to program in a slicer.

Upper-limb projects focused on wrist-hand orthoses (WHO) for fracture stabilization and post-operative immobilization. These parts are mostly shell geometry with ventilation holes. Printing lets the designer place holes where the skin needs airflow and keep solid material at the load paths, something a perforated sheet cannot do.

Material trials were the third cluster. Groups tested PA 12, PA 11, TPU at different shore hardness, and PEEK for high-temperature sterilization cycles. PEEK printed parts can survive autoclave, which matters for reusable devices. The trade-off is printer cost and a narrower process window.

One pattern stands out across all three groups: the printed part is rarely the whole device. Straps, buckles, hinge pins, and adjustment mechanisms are usually machined or injection molded. That split is a practical decision, not a compromise.

PA 12 (nylon) is the workhorse. It has good fatigue resistance, takes repeated flex without cracking, and prints on SLS or MJF with consistent density. For an AFO that flexes thousands of times per day, fatigue life is the number that matters, not ultimate tensile strength.

TPU is chosen when compliance is the goal. A soft TPU shell can replace a rigid shell plus padding in some designs, which reduces part count. The downside is that TPU creeps under sustained load, so a strap that stays tight overnight will loosen over weeks.

PEEK and PEKK enter when the device must be sterilized or must carry load in a small cross-section. They print at 380–420 °C nozzle temperature and need a heated chamber. Not every service bureau runs them, and the cost per part is several times PA 12.

Carbon-fiber-filled nylon raises stiffness but lowers elongation at break. For a shell that must flex, that trade can cause brittle failure at the hinge line. We usually recommend unfilled PA 12 for flexing zones and CF-nylon only for rigid brackets.

• 1
  PA 12Best all-round fatigue life; SLS or MJF; good for AFO and WHO shells.
• 2
  TPUUse where compliance replaces padding; watch creep under long-term load.
• 3
  PEEK / PEKKAutoclave-capable and strong; high printer cost and tight process window.
• 4
  CF-nylonStiff but brittle; limit to rigid brackets, not flexing hinge zones.

Printing is a layer-by-layer process, so the Z axis is always the weak direction. A part loaded in tension across the layer lines will delaminate before the bulk material yields. If your orthotic has a thin strap eyelet or a clevis that pulls along the build direction, rotate the part or machine it instead.

Tolerance is the second boundary. A printed shell can hold roughly ±0.3 mm on a well-tuned SLS machine. A hinge pin bore that must press-fit a 4 mm shaft needs better than that. This is where CNC enters: bores, threads, and bearing seats are machined to ±0.005 mm, then bonded or fastened into the printed shell.

Surface finish matters for skin contact. As-printed SLS has a granular texture around Ra 8–12 μm. It can be dyed and sealed, but a machined and bead-blasted insert feels smoother. For a device worn 12 hours a day, that difference shows up in compliance.

Volume is the third boundary. Below a few hundred units, printing wins because there is no tooling. Above that, injection molding or CNC from a block can undercut the printed cost, especially for simple shells with uniform wall thickness.

Most serious orthotic projects end up hybrid. The printed shell carries the patient-specific shape. A machined aluminium or stainless hinge block carries the pivot. The two are joined with structural adhesive, heat-set inserts, or rivets. This lets the shell be reprinted when anatomy changes while the hardware is reused.

For the machined side, materials are usually 6061-T6 aluminium for weight, 316L stainless for corrosion resistance against sweat and cleaning agents, or Ti-6Al-4V when strength-to-weight matters most. PEEK is also machinable and survives autoclave, so it is a good match for reusable instruments.

Finishing is not cosmetic here. Anodizing adds wear resistance to aluminium hinge parts. Bead blasting removes tool marks that could trap bacteria. Laser marking puts a device ID and lot number on the part; minimum character height is 1.5 mm so it stays legible after handling.

Inspection closes the loop. A printed shell is checked by scanning against the original scan data. A machined insert is checked with calipers, pin gauges, and a CMM when the print calls for it. We inspect 100% of parts before shipment and can supply reports on request.