Why lattices attract attention
A heat sink is a surface-area problem inside a volume constraint. Triply periodic minimal surfaces (gyroid, Schwarz-D and friends) pack a great deal of continuous surface into a small envelope, with fluid paths that avoid the dead zones you get behind a straight fin.
What L-PBF actually imposes
The design freedom is real, but it is bounded. Three constraints matter more than the rest:
- Minimum wall thickness. Below roughly three to four times the layer height, walls become unreliable: partial fusion, surface porosity and dimensional drift.
- Overhang angle. Unsupported surfaces below about 45° from horizontal degrade. In a closed lattice you cannot remove supports afterwards, so the geometry has to be self-supporting by construction.
- Powder evacuation. Every internal channel must have a route out for unfused powder. A sealed cell full of trapped powder is added mass, not added performance.
Surface roughness cuts both ways
An as-built internal surface is rough, which increases wetted area and turbulence, and that is helpful for heat transfer. It also increases pressure drop. Whether that trade is favourable depends entirely on the pumping budget, and it is worth measuring on a printed coupon rather than assuming a smooth-wall correlation holds.
A workable method
Model the thermal performance, then run the geometry back through the three manufacturability constraints above and re-model whatever had to change. Print a coupon. Measure pressure drop and thermal resistance on the real part. The gap between the first simulation and the measured coupon is the useful information.
