Additive Manufacturing Basics

In Module 0 you met the three families of making. Additive is the one where the machine adds material - almost always in thin, stacked layers - rather than cutting it away or reshaping it. Nothing is wasted into chips or off-cuts; the part simply accretes where the file says material should be.

The key mental move is that a printer cannot see your 3D model the way you do. It builds in two-and-a-half dimensions: it can move freely in X and Y across a flat plane, then step up a fixed amount in Z and do another flat plane. So before anything prints, the model has to be sliced - cut by a stack of evenly spaced horizontal planes into a sequence of 2D layers. Each layer is a flat outline (a perimeter) with some pattern filling it, and the machine draws that outline in material before rising to the next.

That is why a printer is happiest with flat-ish, gently changing geometry and struggles with steep overhangs: every layer needs something beneath it to land on. The slice is not a detail of the software - it is the deep logic of the whole family. Once you see an object as a stack of pancakes rather than a solid lump, additive fabrication stops being mysterious.

This layered logic also explains additive's signature look and its signature weakness in one stroke. The faint horizontal ridges you can feel on a printed part are the layers themselves, and the thinner you slice, the less you see them. The weakness is that the joins between layers are never quite as strong as the material within a layer - so a printed part has a grain, like timber, and behaves differently along it than across it. Almost everything else in this module is really a consequence of building in stacked layers, so it is worth fixing that single picture firmly in mind before any machine appears.

Between your model and the machine sits a piece of software called the slicer - the additive world's version of CAM. You feed it a mesh, usually an STL or 3MF file (a surface described as thousands of tiny triangles), tell it your machine and material, and it computes the whole print for you.

A slicer does far more than cut planes. It decides the layer height (how thick each slice is - typically 0.1 to 0.3 mm for desktop FDM), how many solid walls or perimeters trace each outline, how dense the internal infill lattice is, where supports are needed under overhangs, and how fast and hot to run. It then writes all of this as G-code - the same move-by-move instruction language you met in Module 1 - which the printer executes blindly. Popular slicers include Cura, PrusaSlicer, Bambu Studio and, for resin, Lychee or Chitubox.

The honest truth is that most print quality is won or lost here, not at the machine. A good model sliced badly prints badly; a modest printer with well-chosen settings can produce beautiful parts. Learning additive manufacturing is, to a large degree, learning to drive a slicer thoughtfully.

Additive is not one technology but a cluster of them, grouped by how each layer is formed. Three families cover almost everything you will meet.

Material extrusion melts a thermoplastic and squeezes it out of a moving nozzle, drawing each layer as a bead of molten plastic that cools and sticks. This is FDM (or FFF) - the cheap, ubiquitous desktop workhorse of the next lesson.

Vat photopolymer starts with a tank of liquid resin that hardens when hit by UV light. A light source - a laser in SLA, a whole-layer projector or LCD in DLP - cures one layer at a time against a build plate that lifts out of the vat. It gives the finest detail of any accessible process.

Powder bed fusion spreads a thin layer of fine powder and uses a laser or heat to fuse the grains where the part should be (SLS in nylon, MJF, or metal variants like DMLS). The surrounding loose powder holds the part up, so no supports are needed and complex, strong parts come out in one go.

There are others - material jetting, binder jetting, directed energy deposition - but if you understand these three, you understand the landscape. Lessons 4.2 and 4.3 take them one at a time.

Additive is not automatically better than cutting - it is a different tool with a different sweet spot. It wins clearly in four situations. Internal complexity: hollow parts, internal channels, lattices and shapes trapped inside other shapes are trivial to print and often impossible to mill, because no tool has to reach in. Organic, freeform geometry: a doubly-curved or branching form that would need five-axis machining prints without fuss. One-offs and low volumes: with no mould and no fixturing, the first part costs the same as the hundredth, so bespoke and prototype work is cheap. Consolidation: several machined-and-assembled components can sometimes be redesigned as a single printed part.

Subtractive still wins where additive struggles: large parts (printing is slow - a big FDM piece can take many hours), fine tolerances and smooth finishes straight off the machine, and the strength and material range of solid stock. Layered parts are also anisotropic - weaker along the layer-to-layer bond than across it - which matters for load-bearing pieces. A useful worked comparison: a small, boxy aluminium bracket might mill in twenty minutes on a CNC and come off flat and precise, while the same part printed in plastic could take three hours and still be weaker in one direction - but a branching, hollow, lattice-cored version of that bracket may be impossible to mill at all and print in the same three hours. The geometry, not habit, decides. The maker's skill, as ever, is choosing: reach for additive when geometry or one-off freedom is the prize, not because a printer is in the room.

Before you fall in love with additive, know what it does not give you for free. Resolution is bounded by the process: an FDM nozzle cannot draw a feature finer than its bead (around 0.4 mm), and the layer height leaves visible steps on gentle slopes - the classic ridged look of a cheap print. Resin resolves far finer, powder sits in between. If your model has crisp text or a razor edge, ask whether the process can even render it before you print.

Tolerance - how closely the finished dimension matches the drawing - is looser than machining on most desktop processes. A hole modelled at 5 mm may print at 4.7 or 5.2 mm depending on the machine, material shrinkage and how the layers stack, so parts that must mate precisely need test prints and often a deliberate clearance designed in. Finish off the machine is rarely presentation-ready: FDM shows layer lines, powder is grainy and matte, resin is smooth but needs the supports cleaned off. Nearly every serious print gets some post-processing - sanding, priming, painting, vapour-smoothing, dyeing - and that labour is part of the true cost. None of this makes additive worse; it makes it a real process with a real envelope. Designing inside that envelope, rather than being surprised by it, is what the rest of this module builds.