3D Printed Elemento Helmet: How Lattice Energy Absorption Works
KASK moved a bicycle helmet shell and liner into additive manufacturing, and the design logic behind it is worth reading for anyone specifying structural parts. This page explains the mechanism, the material and print-orientation limits, and when a 3D printed Elemento helmet approach makes sense versus when machined or molded parts win. Written for design engineers and sourcing teams who need to judge a printed structure, not just admire it.

In this article
- 1
- 2
- 3
- 4
- 5
- 6
Why a 3D printed Elemento helmet absorbs impact differently
A conventional helmet liner is expanded polystyrene. It is molded into one density, one shape, one thickness. Impact energy goes in, the foam crushes, and the cell structure that was molded is the cell structure you get. There is no way to vary stiffness across the shell without gluing in a second foam piece.
The 3D printed Elemento helmet replaces that single-density block with a printed lattice. A lattice is a repeated cell geometry, and the whole point of printing it is that stiffness can be tuned cell by cell. Thicker struts near the crown, thinner struts at the temple, a denser band around the rim where the shell meets the head. None of that requires a new mold.
Impact energy is absorbed by progressive cell collapse. Under load, the struts bend, then buckle, then densify. Each stage has a different stiffness, and the plateau between buckling and densification is where the energy actually goes. A lattice with a long, flat plateau protects better than one that goes from stiff to solid in a few tenths of a millimeter.
That plateau is a design variable. FEA and drop tests on printed coupons tell you where it sits. The mechanism is not magic; it is the same crush behavior engineers already model in honeycomb and foam, just with more control over where the crush starts and how far it spreads.
Print orientation, strut diameter and the boundary conditions
Directional strength is the first boundary. A lattice printed with vertical struts is strong in compression along the print axis and weak in shear across it. Rotate the part on the build plate and the same geometry behaves differently. Any printed structure that takes a load in more than one direction needs cells designed for the directions that matter, not a single uniform pattern.
Strut diameter sets the minimum feature size, and that size is limited by the printer, not the CAD file. Below roughly 0.4–0.6 mm, most powder-bed polymer machines lose strut fidelity: the strut prints thin, partially fused, or not at all. Design the lattice around the process window you can actually hold, then confirm with a printed coupon.
Layer height and build angle also set surface roughness on the struts. A 0.1 mm layer gives a smoother strut and a more predictable buckling load than a 0.2 mm layer. Roughness matters here because it acts as a stress riser. Small notches on a strut wall start the crack earlier than the FEA model predicts.
Finally, resin or powder removal. Hollow lattices trap uncured resin or loose powder. If the cells are too small or the drain paths too few, the part carries extra mass and, worse, trapped material that can shift the balance. Design drain paths and vents as part of the lattice, not as an afterthought.
Material choice and where printed polymers stop
Sports helmets in this class typically use a photopolymer or a laser-sintered nylon with elastomeric behavior. The requirement is a material that yields rather than shatters: it must take repeated small impacts and one large one without fragmenting. Glass-filled grades add stiffness but cut elongation, and that trade is usually wrong for an energy-absorbing liner.
Temperature is the usual failure point. Many printed photopolymers soften well below the temperature a dark shell reaches in summer sun. A lattice that passes a room-temperature drop test can lose half its plateau stiffness at 50–60 °C. If the part sees heat, the material data sheet needs numbers at that temperature, not just at 23 °C.
Humidity and UV add a slower effect. Nylon absorbs moisture, which changes stiffness and dimension over months. UV degrades many polymers at the surface. Coatings help, but they add mass and a process step, and they can also seal a lattice that was designed to be open.
This is where the boundary sits: printed polymers are excellent for low-density, complex, energy-absorbing geometry, and poor for high-point loads, sharp threads, and metal-to-metal bearing surfaces. Those features belong in a machined or molded component that is bonded or fastened to the printed body.
What the 3D printed Elemento helmet teaches about hybrid design
Almost every production helmet is a hybrid. The printed lattice handles energy absorption. A molded or machined shell handles stiffness, puncture resistance, and the anchor points for straps and adjusters. The printed part is rarely asked to do the whole job.
That split is a useful template for other products: impact guards, drone frames, prosthetic sockets, robot end effectors, protective housings. Print the geometry that is expensive to mold and cheap to vary. Machine or mold the geometry that needs tight tolerances, threads, or a hard bearing surface.
At GreatLight we see this crossover constantly. A team prints a lattice prototype to validate the energy curve in a week, then asks for the mounting ring, the strap anchor, or the insert in aluminium or stainless so it survives real loads and real torque. The printed part proves the concept; the machined part carries the load.
Two practical checks before you commit. First, does the printed feature carry a point load or a torque? If yes, plan a metal insert from the start. Second, does the assembly need a tolerance tighter than about ±0.2 mm? Printed polymer usually will not hold it across a full production run, and the mating part should be machined to ±0.005 mm instead.
How to verify a printed energy-absorbing part
Test the coupon before the part. Print a lattice block with the same cell, strut diameter, material, and orientation as the production geometry, then run it in compression. The curve you get is the real plateau, not the modeled one. It costs a day and it saves a redesign.
Then test at temperature. Run the same coupon at the top of the service range and at the bottom. A plateau that shifts 30 percent between 23 °C and 55 °C tells you the part needs a different material or a metal load path. Do not extrapolate from room-temperature data.
Then test the assembly, not just the coupon. Bonding, fasteners, and the shell interface change how load enters the lattice. A coupon that crushes cleanly can still fail at the bond line if the joint is stiffer than the cell next to it.
For the metal side of the assembly, the routine is more familiar: raw material check, in-process monitoring, final inspection, and reports on request. GreatLight holds ±0.005 mm on machined features and Ra 0.8–1.6 μm as a standard machined finish, with finer finishes down to Ra 0.2–0.8 μm when a bearing surface or seal groove needs it.
Printed lattice versus machined or molded structure
Use this to decide which process owns which feature.
| Feature | 3D printed lattice | CNC machined metal | Injection molded |
|---|---|---|---|
| Complex internal geometry | Easy, no tooling | Limited by tool reach | Needs side actions |
| Energy absorption | Tunable per cell | Not a strength | Single density |
| Tolerance held | About ±0.2 mm | ±0.005 mm | ±0.05 mm on small parts |
| Threads and inserts | Poor, plan a metal insert | Cut directly | Molded or inserted |
| One-off cost | Low, no tooling | Low to medium | High tooling cost |
| Run of 10,000+ | Slow and costly | Competitive | Lowest unit cost |
| Heat resistance | Material dependent | High with the right alloy | Grade dependent |
Pick the process by the load, not by the trend
If the feature absorbs energy, varies in density, or exists in low volume, print it. If it carries a point load, holds a thread, or needs ±0.005 mm, machine it and join the two.
Common questions
Can a printed lattice replace foam in a production helmet?
For the energy-absorbing layer, yes, if the material holds its stiffness across the service temperature range and the lattice can be printed without trapped resin or powder.
The shell, strap anchors, and adjuster hardware usually stay molded or machined. Most production designs are hybrids.
What is the smallest strut we can print reliably?
It depends on the machine and material, but below roughly 0.4–0.6 mm most powder-bed polymer systems lose fidelity: struts print thin or partially fused.
Design the cell around a strut you can measure on a coupon, not the smallest number in the CAD file.
How does print orientation change the impact result?
A lattice is much stronger along the print axis than across it. Rotating the part on the build plate moves the weak direction, and it can move it right into the impact path.
Fix orientation first, then tune strut diameter. Changing both at once makes the test results hard to read.
Where do machined parts still fit in a printed assembly?
Threads, bearing bores, pivot pins, mounting rings, and any interface that sees torque or a point load. Printed polymer creeps under sustained load at those features.
Machining them in aluminium or stainless and bonding or fastening them into the printed body is usually cheaper than over-designing the print.
What tolerances should we expect from each process?
Printed polymer typically holds about ±0.2 mm on functional features. CNC machining holds ±0.005 mm (±0.0002 in). Small molded parts sit around ±0.05 mm.
Set the assembly tolerance from the tightest mating feature, then assign that feature to the process that can hold it.
Can you quote both the printed prototype and the machined hardware?
Yes. Send the model and we return a quotation with a free DFM analysis within 12 hours, and production can start within 24 hours. There is no minimum order quantity, from one prototype to 10,000+ part runs.
Uploads are secure and confidential, and an NDA is available on request.
Send the model and we will tell you which process fits
A quotation with free DFM analysis within 12 hours, and an engineer who will say when printing is the wrong answer.
12-hour quoteNo minimum order100% inspection