3D printing makes it possible to complete the partial replacement of an ankle prosthesis

A total ankle replacement swaps the whole joint. A partial replacement keeps the healthy side of the joint and only rebuilds the worn or collapsed portion, usually the talar dome or a section of the tibial plafond. That smaller scope is what makes the case interesting from a manufacturing standpoint: the implant has to sit on a surface that no two patients share.

The bone-facing side is not a flat plane. It follows the resected surface, and the resection itself follows the patient's remaining cartilage, subchondral bone and cyst cavities. On a machined-only route you would approximate that geometry with a few planar cuts and accept gaps. Gaps mean stress shielding and loosening. This is where 3D printing makes it possible to keep the interface close to the patient's own anatomy rather than to a catalog size.

The bearing side is the opposite problem. It is a curved metal or ceramic surface that articulates against a polyethylene insert, and it needs a defined radius, a defined surface finish and a defined clearance. Those are tolerance-driven features. Additive alone will not hold them.

• 1
  Bone interfacePatient-matched, porous or textured, driven by imaging data.
• 2
  Bearing surfaceSpherical or toroidal, tolerance-driven, machined.
• 3
  Fixation featuresScrew holes, keels, pegs. Machined or printed depending on load path.

Laser powder bed fusion on Ti-6Al-4V is the common route for the bone-facing body. A 30–60 μm layer height and a 60–100 μm laser spot give walls thin enough for trabecular-like lattice and stiff enough for a load-bearing talar component. You can vary strut diameter across the part to match the stiffness gradient between the subchondral plate and the cancellous bone behind it.

Printed lattice does more than save weight. Bone grows into open pores in the 300–800 μm range, which gives secondary fixation without cement. That is a geometry you cannot cut with an end mill. This is another place where 3D printing makes it possible to do something the subtractive route simply cannot reach.

What the printer will not give you is a bearing surface. As-built titanium lands around Ra 8–15 μm with partially sintered particles on downward-facing surfaces. It also drifts. Expect ±0.1 mm on a well-controlled machine, and more on thin unsupported sections. So the printed blank is a near-net shape, not a finished implant.

The printed body arrives with a machining allowance of 0.3–0.5 mm on every surface that has to be accurate. First operation is usually a fixture built to the printed geometry itself, so the datum is the patient-matched surface rather than a random face. From there, a 5-axis cycle machines the articulating radius, the insert pocket and the fixation holes in one setup.

Simultaneous 5-axis matters here because the bearing surface is a compound curve around a tilted axis. Repositioning the part mid-cycle adds stack-up error that eats into a ±0.005 mm tolerance budget fast. The same setup also drills the screw holes normal to the curved surface, which avoids the oval entry you get from a 3-axis approach.

Surface finish on the articulating side is typically held at Ra 0.2–0.8 μm. That range is what keeps polyethylene wear predictable. Below that you gain little and spend a lot of cycle time; above it, wear debris climbs. The bone-facing lattice is left as-built and only cleaned, never machined, because cutting it would close the pores.

After machining, parts go through ultrasonic cleaning, passivation for titanium, and 100% dimensional inspection. Reports come with the shipment when the drawing calls for them.

• 1
  DatumsFixture locates on the patient-matched surface, not the print plate.
• 2
  Allowance0.3–0.5 mm on bearing and pocket surfaces.
• 3
  Finish targetRa 0.2–0.8 μm on articulating surfaces.
• 4
  Do not machinePorous lattice. Cutting closes the pores.

Ti-6Al-4V ELI is the default for the bone-facing body. It prints well, it passivates, and its elastic modulus sits closer to bone than CoCr does, which reduces stress shielding at the interface. Standard Ti-6Al-4V (TC4) is acceptable for non-implant tooling and trial components but ELI grade is the one to specify for anything that stays in the body.

CoCr alloys come in when the articulating surface is metal-on-polyethylene and wear resistance is the priority. CoCr prints less cleanly than titanium and is harder to machine, so the printed body is usually kept minimal and the bearing geometry is ground and polished. Hardness helps here; the same property makes tool life short.

PEEK and carbon fibre reinforced PEEK are used for trial components, drill guides and non-load-bearing spacers. They machine fast and can be printed for form checks. They are not a substitute for a metal bearing surface in a load-bearing ankle component.

Stainless 316L and 17-4PH appear in instrument trays, trial handles and drill guides rather than in the implant itself. 17-4PH gives you the hardness for a guide that has to survive repeated sterilization.

For an implantable component the drawing is only half the paperwork. The build file, the powder lot, the machine parameters and the heat treatment record all belong in the device history record. A printed part cannot be re-verified after the fact the way a machined part can, so traceability is captured forward.

Inspection on a patient-matched part is a mix of CMM work on the machined features, optical or CT scanning on the lattice, and a fit check against the anatomical model when the customer supplies one. CT is the only practical way to see inside a porous structure without cutting it up.

GreatLight holds ISO 9001:2015, IATF 16949:2016, ISO 13485:2016 and ISO 27001:2022. The ISO 13485 scope covers the medical device work; ISO 27001 covers how patient imaging data is stored and transferred. Uploads are handled under NDA on request.