Additive Lattice or Gyroid Fabrication¶

Additive fabrication method — instantiates Functional Porosity Design

Prints an explicitly modelled void network — every strut, wall, channel, and orientation drawn by design — so the pore architecture is deterministic rather than left to a stochastic process.

Most ways of making a porous solid hand you a statistical result — a population of cells or interstices you steer but never place. Additive Lattice or Gyroid Fabrication is the exception: it builds the void network from an explicit geometric model, layer by layer, so the size and shape of each pore, the channels that connect them, and the direction the struts run are all drawn rather than discovered after the fact. The lattice may be a repeating unit cell (a strut truss, or a triply periodic minimal surface such as a gyroid) or a deliberately aperiodic field, but the defining move is the same: porosity becomes CAD geometry, and everything a designer can specify about a curve they can now specify about a pore. That determinism is what separates it from every foaming, packing, or phase-separation route in this archetype — and it is also what makes the build's fidelity to the file the whole ballgame.

Example¶

A surgical-implant team needs an acetabular cup — the socket half of a hip replacement — that bone will grow into and that flexes a little like bone rather than like solid titanium. Solid metal is far too stiff, so load bypasses the surrounding bone and it slowly resorbs (stress shielding). They model the cup's outer shell as a gyroid lattice in Ti-6Al-4V: pore size set to ≈600 µm so bone cells can migrate in and vascularize, wall thickness tuned so the porous shell's effective stiffness lands near cortical bone rather than near bulk titanium, and the gyroid's naturally interconnected channels guaranteeing no dead-end pockets where cells would starve. Because the network is one continuous minimal surface, there is no loose powder to work free in the body. They build it by laser powder-bed fusion, then check the printed pores against the model. The output is a part whose porosity is a specification, not a happy accident: ≈70% porous in the ingrowth shell, fully dense at the articulating rim, with the transition placed exactly where the surgeon wants it.

How it works¶

What distinguishes the method is that the pore network is authored, not processed:

Model the unit cell (or field). Pick a lattice family — strut truss, a TPMS such as gyroid or diamond, a stochastic Voronoi — and set its dimensions. This fixes pore size, wall/strut thickness, and open-vs-closed character directly.
Set orientation by design. Because struts are drawn, their alignment is a free variable: bias them along a load path or a flow direction to build in anisotropy on purpose.
Vary the cell across the part wherever a region needs different pores, since each cell is independently specified.
Build and reconcile. Print layer by layer, then compare the as-built network to the model — the method's fidelity, and its failure surface, both live here.

Tuning parameters¶

Unit-cell type — truss vs TPMS vs stochastic. Minimal surfaces give smooth, fully interconnected channels with gentle stress gradients; strut lattices are lighter and simpler but concentrate stress at their nodes.
Cell size and wall/strut thickness — the primary dials for pore size and effective stiffness; finer cells raise surface area and lower the printable floor.
Strut orientation / anisotropy — how strongly the network is aligned to a direction; buys directional stiffness or flow at the transverse direction's expense.
Relative density — the void fraction the cell targets; trades stiffness and strength against openness and lightness.
Build orientation on the plate — which way the part is grown; decides which struts are unsupported overhangs and therefore where as-built quality degrades.

When it helps, and when it misleads¶

Its strength is control no stochastic route can match: a designer can put a chosen pore size exactly where a function needs it, interconnect it on purpose, tilt its anisotropy toward a load or a flow — then reproduce the same architecture part after part. For lightweighting, tuned stiffness, and designed transport channels, nothing else is as direct.

The trap is treating the CAD file as the part. The as-built network is never quite the as-designed one: down-facing struts sag, thin walls thicken with adhered powder, and any feature below the machine's self-supporting overhang angle degrades or fails to form.^[1] Unremoved powder can clog nominally open channels, and the printed strut surface is rough enough to seed fatigue cracks. The discipline that keeps the method honest is to close the loop — measure the built pore network and load-test it rather than trusting the model — and to design within the process's real resolution and overhang limits, not the geometry's imagined ones.

How it implements the components¶

Additive fabrication fills the designed-geometry components — the ones you can only claim when pores are drawn rather than grown:

void_size_and_shape_distribution — the unit cell fixes pore size and shape directly, at near-zero dispersion when the build is clean.
connectivity_and_throat_topology — open-vs-closed and the channel/throat network are chosen (a gyroid is fully interconnected by construction).
orientation_and_anisotropy_profile — strut alignment is a design variable, so anisotropy is dialled in rather than inherited.

It does not set an overall void-fraction budget from process physics (that is Gas Foaming or Blowing), grade the architecture across the bulk (Graded-Density Manufacturing), or build high specific surface out of interstices (Particle Packing and Sintering Control); verification is a separate job — pore metrology belongs to Multi-Method Porometry, load-bearing proof to Mechanical Coupon and Fatigue Testing.

Instantiates: Functional Porosity Design — supplies the designed, deterministic pore architecture the archetype's other machinery then grades, measures, or maintains.
Sibling mechanisms: Gas Foaming or Blowing · Graded-Density Manufacturing · Particle Packing and Sintering Control · Sacrificial Templating and Leaching · Topology Optimization for Void Placement · Multi-Method Porometry · Mechanical Coupon and Fatigue Testing

References¶

[1] Powder-bed additive processes cannot form a down-facing surface below a minimum self-supporting overhang angle (commonly around 45° from horizontal, machine- and material-dependent) without sacrificial supports; struts and pore walls below that angle sag or trap powder. This is why build orientation is a first-class design decision, not a print-time afterthought. ↩