SLS 3D printing technology for custom orthotic insoles

• 1
  How SLS 3D printing technology turns a scan into an insole
• 2
  What geometry SLS 3D printing technology can and cannot hold
• 3
  Material behavior that decides the insole life
• 4
  From foot scan to finished insole: the steps that matter
• 5
  When SLS is the wrong route for an insole
• 6
  Step by step: setting up an SLS insole build
• 7
  SLS versus CNC versus molding for an insole program

Selective laser sintering works from a powder bed. A roller spreads a thin layer of polymer powder, usually 0.08 to 0.12 mm deep, across the build chamber. A CO2 laser, typically 30 to 70 W, traces the cross-section of the insole and melts the powder where the part should be. The platform drops by one layer, a new layer is spread, and the cycle repeats until the build is complete.

The key difference from fused deposition or stereolithography is what holds the part up. There is no support structure to cut away. The surrounding loose powder supports every overhang, so a medial arch that curls 40 mm above the base prints without any scaffold. That is why SLS 3D printing technology suits organic shapes like a footbed.

The powder bed also means parts come out inside a solid block of powder, often called the cake. You break the cake out, brush or bead-blast the part, and the surface is a matte grey. Unmelted powder that has not been through the laser can be sieved and reused, which is where the economics of the process comes from at higher volumes.

One consequence engineers should note early: the laser sinters powder, it does not fully melt it into a transparent solid. The result is a porous, slightly grainy structure. That porosity is fine for a footbed, but it changes how you seal, color, and clean the part later.

• 1
  Layer range0.08–0.12 mm typical for PA12 insoles.
• 2
  SupportLoose powder, not printed scaffolding.
• 3
  Post-processDepowder, bead blast, optional dye or seal.

The process has no draft angle limit and no tooling, so undercuts, internal channels, and variable wall thickness are all printable. A typical insole runs from 2 mm at the forefoot to 8 mm under the heel, with lattice or gyroid infill in the midfoot to tune stiffness. That kind of graded structure is impractical to mold and slow to machine.

Wall thickness is the first real constraint. Below about 0.8 mm, a single laser pass may not fully melt the layer, and the wall can warp or break during depowdering. Keep load-bearing walls at 1.2 mm or more. Thin cosmetic ribs can go to 0.8 mm if they carry no load.

Minimum feature size is another limit. Slots, holes, and text smaller than about 0.4 mm tend to close up or lose definition. If your insole needs drainage holes or registration marks, keep them at 0.8 mm or larger. Sharp internal corners should be filleted to at least 0.5 mm radius to avoid stress risers.

Long, flat, unsupported sections are where warpage shows up. A 250 mm flat plate printed in PA12 can curl 0.5 to 1.0 mm at the edges as the part cools. Adding a slight crown or a rib pattern removes most of that. For a footbed this is rarely a problem because the shape is already curved, but a flat insert plate is a different story.

• 1
  Load-bearing walls1.2 mm minimum, 2 mm preferred.
• 2
  Minimum hole0.8 mm to stay open after depowdering.
• 3
  Flat spansCrown or rib flat plates longer than 150 mm.

PA12 is the default powder for orthotic work. It has a flexural modulus around 1,600 MPa, good fatigue resistance, and it tolerates repeated load cycles without creeping much. A 90 kg user standing eight hours a day will not permanently flatten a 3 mm PA12 arch in normal use. PA11 is tougher and slightly more elastic, which some clinicians prefer for diabetic or high-activity users.

Glass-filled or carbon-filled PA12 raises stiffness sharply, often two to three times plain PA12. That is useful for a rigid shell in a post-surgical device. It is wrong for a cushioning layer, because it will not absorb impact and can feel harsh under a heel.

TPU powders exist for SLS and give a softer, rubbery part. They are harder to process, the powder ages faster, and the surface finish is tackier. For a shoe insert that needs real compliance, TPU is worth evaluating, but expect a narrower process window and more rejects.

Porosity matters for hygiene. Sintered nylon absorbs moisture and odor over time. Sealing the part with a medical-grade lacquer or a thin urethane coat closes the surface and makes cleaning practical. If the insole contacts broken skin, plan the coating step from the start rather than adding it after complaints.

• 1
  PA12Default. Balanced stiffness and fatigue life.
• 2
  PA11Tougher, more elastic. Good for high load.
• 3
  CF/GF PA122–3× stiffer. Rigid shells only.
• 4
  TPUSoft, but narrow process window.

A scan is not a CAD model. Most foot scans arrive as a point cloud or a mesh with 200,000 to 500,000 triangles and some noise. The first real step is cleanup: remove artifacts, close holes, and align the scan to a coordinate system so the heel and first metatarsal sit on known axes.

Then you offset the scan into a footbed. The foot is captured under load or unloaded, and those are different shapes. An unloaded scan needs a correction for the weight-bearing posture, or the insole will sit too high under the arch. This is a clinical decision, not a software default.

Next comes the shell. The top surface follows the corrected scan. The bottom surface follows the shoe last or the insole cavity. Between them you place the thickness map and any lattice. This is where SLS 3D printing technology earns its place: the two surfaces do not need to be parallel, and the internal structure can vary across the part.

Finally you orient the part in the build. A footbed laid flat builds faster but piles many layers on the heel, which is the highest-stress zone. Tilting the part 10 to 15 degrees spreads the layer lines across the arch and usually improves fatigue life at the cost of a slightly taller build.

• 1
  Clean the meshFix holes and noise before any offset.
• 2
  Correct postureLoaded and unloaded feet differ.
• 3
  Orient for stressTilt 10–15° to move layer lines.

If the part is a flat, 3 mm plate with two holes and you need 20,000 pieces a year, SLS is the slow and expensive answer. Injection molding or stamping will beat it on unit cost once tooling is amortized. The break-even for a simple insole shell is usually somewhere in the low thousands, but it depends on part size and mold cost.

If the insole must hold ±0.05 mm on a mating feature, such as a cleat that clips into a shoe chassis, sintered nylon will not hold it. The process shrinks during cooling and the shrinkage varies with position in the build. You can hold ±0.3 mm on a well-controlled build, not ±0.05 mm. Machine the mating insert or design a compliant snap instead.

If the surface must be optically smooth or food-safe without a coating, sintered nylon is a poor fit. The porosity traps residue. A machined or molded part with a closed surface solves that directly.

And if the material has to be a metal, SLS in the polymer sense is off the table. Metal sintering exists but is a different process with different cost and finish. For a metal orthotic component, CNC is the practical route.

• 1
  High volumeMolding wins past a few thousand units.
• 2
  Tight tolerance±0.05 mm is out of reach for SLS.
• 3
  Closed surfaceSintered nylon is porous by nature.