Large-Format Additive

Large-format additive manufacturing (LFAM, and the well-known BAAM systems from Oak Ridge and Cincinnati) is FDM's logic scaled up: melt a thermoplastic, extrude it in beads, build a part layer by layer - but with a build volume measured in metres and a bead measured in millimetres to centimetres wide. The materials are workaday industrial plastics: ABS, PLA, PETG, polycarbonate and nylon, very often reinforced with chopped carbon or glass fibre.

That fibre matters. A short-fibre composite is far stiffer and more dimensionally stable than neat plastic, and crucially it shrinks and warps less as the big part cools - warping is the enemy at this scale, because a metres-long part that curls as it cools is scrap. The fibre also aligns along the bead as it extrudes, so the printed part is anisotropic: strong along the beads, weaker across the layer lines - a property you design around, orienting the print so the beads run with the main loads. Large additive is not about fine detail; it is about getting a lot of stiff material into an approximate shape, fast. Thermal management is the quiet discipline behind it: a metres-long part cools unevenly, and if a lower region contracts while a fresh upper bead is still hot, the whole part can warp or delaminate. So these machines run in heated enclosures, pace the layers to control cooling, and lean on the fibre reinforcement precisely because it holds dimension while the mass cools. It is a very different feel from a small plastic print - closer to managing a large casting than tending a desktop printer.

The defining feature of these machines is the extruder. A desktop printer pulls filament - plastic pre-formed into a precise 1.75 or 2.85 mm strand on a spool - which is convenient but expensive per kilogram and slow to feed. Large-format printers instead use a pellet (granule) extruder: raw plastic pellets, the same cheap feedstock injection moulders buy by the tonne, pour into a hopper, and a turning screw melts and pushes them out through a wide nozzle, exactly like a plastic-extrusion or injection-moulding screw.

Two things follow. First, cost: pellets are a fraction of the price of filament, so printing a big part is affordable in a way it never is on filament. Second, throughput: a pellet screw can push several to tens of kilograms per hour, versus tens of grams on a desktop machine - orders of magnitude faster. The extruder is often mounted on a 6-axis robot arm or a large gantry, giving reach and, on a robot, the freedom to print non-planar. The trade-offs are honest: wide beads mean a coarse surface and low fine detail, and pellet feeding and melt control are fussier than filament. You buy speed and cheap material and pay in resolution - which is exactly why the next step is often a milling cutter.

The single most valuable job for large additive right now is not the final part - it is the tool that makes the final part. Moulds, formwork and jigs are expensive, one-off, geometrically complex objects, which is precisely the territory where additive beats conventional making.

Concrete formwork is the clearest case. A bespoke curved concrete panel traditionally needs a hand-built timber-and-plywood mould that is costly, slow and thrown away after one pour. Printing that formwork from composite - often printed near-net then milled to a smooth face - produces a reusable, dimensionally exact mould for complex geometry at a fraction of the labour, and it is being used for precast facade panels and freeform concrete. The same logic serves GFRC moulds, thermoforming and vacuum-forming tools, composite lay-up moulds, and casting patterns. Because a printed mould can be a hollow, ribbed shell rather than a solid block, it uses far less material and cools evenly. The mental shift is important: large additive frequently earns its place one step back from the visible product, making the tooling that shapes concrete, glass-reinforced composite or thermoset - a quiet but genuinely transformative role. The economics are compelling wherever a shape is complex and needed only a few times: a printed mould can be produced in days instead of the weeks a hand-built pattern takes, revised by editing a model rather than rebuilding a jig, and stored digitally so it can be reprinted if it wears. For a facade of dozens of subtly different curved panels - the kind of geometry parametric design produces easily and traditional formwork prices out - printed moulds can be the difference between the design being buildable and being value-engineered away.

Because pellet printing is fast but coarse, and milling is precise but slow (and wasteful cutting from solid), the two combine beautifully. In a print-then-mill hybrid you additively build the part near-net-shape - close to final, deliberately a few millimetres oversize - then switch to a milling spindle and subtractively machine the surfaces that need accuracy and finish to tolerance.

This is the best of both families and directly recalls the subtractive/additive framing from Module 0. Additive gets a lot of material into roughly the right shape quickly and cheaply, wasting almost nothing; milling then delivers the flatness, smoothness and dimensional tolerance additive cannot reach on its own. On the most integrated machines a single 6-axis robot swaps end-effectors - print head, then spindle - and does both in one setup, so the part never loses its reference position. It is the standard workflow for printed moulds and tooling, where the working face must be milled smooth to leave a good surface on the cast part. The lesson generalises: additive and subtractive are not rivals; the strongest large-scale workflows sequence them, each doing what it is best at.

Be clear-eyed about the niche. Large additive wins for one-offs and very short runs of big, geometrically complex objects: bespoke moulds and formwork, tooling and jigs, patterns, exhibition and stage pieces, furniture-to-architectural sculpture, boat and automotive prototypes. Its advantages are no tooling cost, freeform geometry, fast material deposition, and the ability to print with recycled and reinforced plastics.

It loses to conventional processes the moment quantity or finish dominates. For thousands of identical parts, injection moulding is vastly faster and cheaper per unit. For a smooth cosmetic surface straight off the machine, additive cannot match a moulded or machined finish without post-work. For pure strength-to-cost in a simple shape, cut or moulded stock wins. And the printed part is anisotropic and only as good as its inter-layer bond, so load-bearing use needs testing and engineering sign-off. The honest summary: large additive is a tooling-and-one-off technology that most powerfully makes the moulds, formwork and prototypes for other processes - not a replacement for mass production. Match it to that job and it is transformative; expect it to replace the factory and it disappoints. There is a sustainability angle worth naming honestly too: many LFAM systems run recycled and reclaimed thermoplastic, and some parts can be shredded and reprinted at end of life, which is a real circular-economy advantage over thermoset composites that cannot be remelted. But that only holds for the thermoplastics; fibre-filled and cross-linked materials are far harder to recycle, so the green claim, like everywhere in this module, depends on the specific material choice rather than the process alone.