Best in Bull
Why the heel is not a lattice
An open honeycomb is the stiffest thing you can build in compression and the weakest thing you can build for holding a heel in place. Here is the arithmetic that made us close it up.
Every printed insert you have seen photographed is an open lattice, all the way through, heel to toe. It photographs beautifully. We built one, measured it, and then closed the heel.
Here is why.
A honeycomb is a bundle of tubes
Load a vertical-walled honeycomb straight down its axis and it is the stiffest structure you can make out of a given amount of material. That sounds like an argument for using it. It is the opposite.
A heel cup does not work by resisting downward load. It works by resisting the heel splaying sideways. A honeycomb cell is a tube, and tubes splay. Push a tube over and it hinges at its base. So an open-lattice heel cup gives you the one property you did not need and none of the property you did.
The wall that crushes on the second step
The other half is fatigue. A thin wall in a lattice does not fail by breaking, it fails by buckling — and once a TPU wall has creased, it creases in the same place forever.
Plate buckling for a wall of thickness t braced by its neighbours at spacing b:
σ_crit ≈ 4π²E / (12(1−ν²)) · (t/b)²Taking E for TPU 95A at 20 MPa — an estimate, it moves with print orientation, layer bonding and moisture — and a peak heel-strike pressure around 600 kPa:
| zone | wall | solid fraction | σ_crit | σ in the wall | margin |
|---|---|---|---|---|---|
| 0.86 mm lattice heel | 0.86 mm | 20 % | 2.7 MPa | 3.0 MPa | buckles |
| 2.40 mm heel | 2.40 mm | 48 % | 19.3 MPa | 1.25 MPa | 15× |
| 0.86 mm forefoot | 0.86 mm | 25 % | 4.5 MPa | 2.4 MPa | 1.9× |
The first row is the interesting one. A 0.86 mm wall — two extrusions at a 0.4 mm nozzle, which is the thinnest honest wall on that machine — goes over at heel strike. For about a week that reads as *cushioning*. After that it is a crease, and then it is a crack.
The forefoot row is the reason the forefoot stays open. Push-off never reaches the same peak, the margin holds, and that is exactly where the honeycomb is earning its keep: flex, moisture, airflow.
What we did instead
The heel is solid. The cup crest is solid. The honeycomb is a field of recessed pockets in the top surface rather than holes through the part, and the wall thickness is graded along the foot so there is no seam anywhere for the two behaviours to delaminate at.
Measured off the shipped mesh, US Men's 11:
| overall | 290.0 × 98.0 × 14.17 mm |
| heel floor → cup crest | 5.6 → 13.9 mm (8.1 mm cup) |
| arch, medial / centre / lateral | 10.8 / 7.8 / 4.6 mm |
| forefoot section | 3.4–3.9 mm |
| pocket pitch | 6.4–8.6 mm |
| land between pockets | 1.6 mm |
| volume · mass in TPU 95A | 104.9 cm³ · ≈127 g |
The part we got wrong first
We shipped a revision with a heel cup taken from the wrong file — a ridge with its crest 11–15 mm inboard, behind a flat flange. It is a real orthotic pattern. It is not ours.
The cup we actually wear peaks at the perimeter: 12.0 mm on the lateral side and 12.7 mm on the medial against a 3.1 mm floor, with the rise starting about 16 mm from centre. We had measured that off our own insert and then designed past it. The current cup is that geometry, scaled.
That is the whole method, really. Build it, measure the thing you actually built, and believe the measurement over the intention.
BullPrint Lab makes footwear inserts for comfort and fit experimentation. Nothing here is a medical device and nothing here is medical advice — no claim is made about treating, preventing or diagnosing anything.