FDM vs SLS 3D Printing

Most people’s mental model of 3D printing is FDM — a hot nozzle tracing out layers of molten plastic. It’s the process that put a printer on every maker’s desk, and for good reason. But it’s only one answer to the question of how you turn a CAD file into a physical object, and in a lot of cases it isn’t the best one.

Selective Laser Sintering takes a fundamentally different approach, and understanding why it’s different explains most of the practical trade-offs between the two.

How Each Process Actually Works

FDM (Fused Deposition Modelling) pulls a thermoplastic filament through a heated nozzle and deposits it along a toolpath. Each layer is a set of extruded roads that fuse to their neighbours and to the layer below. The part is built in air, so anything that overhangs beyond roughly 45° needs scaffolding underneath it that you break off afterwards.
SLS (Selective Laser Sintering) starts with a bed of fine polymer powder — typically nylon, with particles around 50 µm across. The whole bed is held just below the polymer’s melting point. A laser then scans the cross-section of the part, adding just enough energy to fuse those particles together. A recoater blade sweeps a fresh layer of powder across the top, the build platform drops one layer height, and it repeats.

That single architectural difference — building inside a supporting medium instead of in open air — cascades into almost everything else.

Geometry: The Biggest Gap

Because unsintered powder surrounds every part as it’s built, SLS needs no support structures at all. Overhangs, internal channels, captive assemblies, lattices, thin organic shells — all print without a single sacrificial structure to remove.

This matters more than it sounds. Supports on FDM aren’t just wasted material; they leave scarring on the surfaces they touch, they constrain how you orient a part, and they add real labour to every print. Designers working with FDM learn to think in terms of what the machine can hold up. With SLS, you mostly design the part you actually want.

The second geometric advantage is nesting. FDM builds on a plate, so you’re limited to what fits in a single layer of floor space. SLS fills a volume — parts can be stacked in three dimensions with a few millimetres of powder between them. A build chamber packed at 10–15% volumetric density will produce dozens of parts in a single cycle. For short-run production, this is the difference between a viable process and a hobby.

There’s one geometric catch: any enclosed cavity will trap loose powder. Hollow parts need escape holes, ideally two, sized generously enough to actually shake the powder out.

Mechanical Properties

FDM parts are anisotropic. Within a layer, the material is close to bulk plastic. Between layers, you’re relying on polymer chains diffusing across a weld line that formed in a fraction of a second against a cooler surface. Z-strength typically lands somewhere between 40% and 80% of XY strength depending on material, temperature, and how well the printer is tuned. If a load path runs perpendicular to the layers, the part fails at a layer boundary.

SLS parts are much closer to isotropic — commonly 85–95% of XY strength in Z. The whole bed sits near melt temperature, so each new layer fuses into a surface that’s still thermally soft rather than a cold one. Nylon 12 out of an SLS machine will typically land around 45–50 MPa tensile with elongation at break in the 15–20% range: genuinely tough, ductile parts that survive being dropped, flexed, and threaded.

For functional prototypes — clips, hinges, housings, brackets that see real load — this is usually the deciding factor.

Surface Finish and Detail

Neither process gives you an injection-moulded surface, but they fail differently.
FDM has visible layer lines and, depending on the nozzle, distinct road boundaries. Top surfaces can be quite smooth; curved vertical walls show stepping. It responds well to sanding, vapour smoothing (for ABS/ASA), and painting.

SLS produces a uniform, matte, slightly granular surface — a bit like fine sandstone — with no directional artefacts and no support scarring. Roughness is fairly consistent regardless of orientation, typically around 10 µm Ra. Downward-facing surfaces look the same as upward-facing ones, which is not something FDM can offer. It also takes dye extremely well; black-dyed nylon parts look and feel close to production components.
Minimum feature sizes are comparable in practice — under a millimetre for walls in both cases — though SLS holds fine detail better on complex organic geometry.

Materials

This one goes to FDM on breadth. Filament is available in an enormous range: PLA, PETG, ABS, ASA, nylon, TPU across a full spectrum of shore hardnesses, polycarbonate, PEEK and PEI on high-temperature machines, plus carbon- and glass-filled variants of most of them. Multi-material and multi-colour printing is well established. You can print something transparent, something conductive, something that dissolves.

SLS is dominated by one family: PA12 nylon, with PA11, TPU, PP, and filled nylons (glass bead, aluminium, carbon) making up most of the rest. Fewer options, but the workhorse material is genuinely excellent — the reason PA12 dominates is that it’s tough, chemically resistant, and dimensionally stable.

Powder also has a lifecycle that filament doesn’t. Unsintered powder is heat-aged through a build and can’t simply be reused indefinitely; it’s typically blended with virgin powder at some refresh ratio. Practical yield is high, but it’s not zero-waste in the way a spool is.

Time, Workflow, and Cost

Here FDM’s advantages are obvious and worth stating plainly.

An FDM printer is cheap, quick to start, and quick to stop. Load filament, slice, print, remove part. A small part is done in an hour. Machines start in the low hundreds of dollars.

SLS is a batch process with real overhead. The chamber has to come up to temperature before the build starts, and — this is the part people underestimate — it has to cool down slowly afterwards. Pull a nylon build out hot and the parts warp. Cool-down can take as long as the build itself. Then there’s depowdering and bead blasting to get the parts clean. You don’t run one SLS part in an afternoon; you fill a chamber and run it overnight.

And the machines are expensive. Industrial SLS systems run well into six figures. The benchtop generation that arrived over the last decade brought that down substantially, but you’re still typically looking at somewhere between fifteen and forty thousand US dollars before powder and post-processing gear.

So which should you use?

Reach for FDM when you need a part today, when you’re iterating quickly on form and fit, when the part is large, when you need a specific material property that only exists as a filament, or when the part is simple enough that supports aren’t a real cost.

Reach for SLS when the geometry is genuinely complex, when the part has to survive real mechanical loads in an unpredictable direction, when you need dozens or hundreds of something, or when you want parts that look and behave like production components rather than prototypes.

For most workshops the honest answer is that these are complementary rather than competing. FDM is where you figure out what to build. SLS is where you build the version someone else is going to use.

The gap that’s always bothered us is the price one. FDM went from twenty-thousand-dollar industrial machines to five-hundred-dollar desktop units in about a decade, and the design freedom that came with putting one on every bench was enormous. SLS hasn’t had that moment yet. That’s the problem we’re working on at Formfortis — a desktop SLS printer built around the same idea that made FDM ubiquitous: that the process matters more than the price tag, and there’s no fundamental reason the two have to stay coupled.