PLA, PETG and TPU — choosing the right one

The short version

• PETG — most tough functional parts. Brackets, housings, mounts, guards, anything handled roughly or left outdoors. If you are not sure, this is usually the answer.
• PLA — parts that need to hold their shape accurately and are not under sustained load. Prototypes, fixtures, jigs, gauges, anything where crisp detail matters more than toughness.
• Flexible TPU — anything that has to bend, grip, damp or seal. Gaskets, soft jaws, grommets, bumpers, strain reliefs, non-marking pads.

You do not have to choose. Describe what the part does and where it lives, and the material gets chosen for you as part of the quote. That sentence is genuinely the most useful thing in an enquiry.

PETG, and why it is usually the default

PETG is the workhorse. It is tougher than PLA and far less brittle, which matters for the things that actually happen to parts — being dropped, over-tightened, knocked, levered against. Where PLA tends to snap, PETG tends to bend and survive. It also holds up considerably better outdoors and against sunlight, so anything that lives outside or in a vehicle should generally be PETG.

The trade is a slightly softer surface and marginally less crisp fine detail than PLA. For a functional part that is almost never the deciding factor.

PLA, and what it is actually for

PLA has a reputation as the beginner's material, which undersells it. It prints accurately with excellent detail and dimensional stability, which makes it genuinely the better choice for a part whose job is to be exactly the right shape — a gauge, a drill guide, an alignment fixture, a prototype being checked for fit. The planetary gear set was printed fine-layer for precisely this reason: on a gear tooth, how closely the printed flank follows the intended curve is what decides whether it meshes smoothly.

Its limits are real, though. PLA is comparatively brittle, so it fails suddenly rather than bending. It softens at modest temperatures — a part left in a car in an Australian summer is a genuine risk, not a theoretical one. And it degrades outdoors over time. For anything living in heat, sunlight or rough handling, PETG is the better call.

TPU, and when flexibility is the point

Flexible TPU is a different kind of material rather than a softer version of the others. It bends and returns, absorbs impact, grips rather than slides, and it is tough in a way rigid plastics are not — it tears far less readily than its softness suggests.

Use it where flexibility is the function: a gasket that has to compress, a jaw face that has to hold a part without marking it, a bumper that has to absorb a knock, a strain relief that has to flex repeatedly without cracking.

It is the wrong choice anywhere stiffness is the function. A TPU bracket is not a bracket. It also prints more slowly and cannot hold fine detail as sharply, so it costs more for the same size of part.

When the answer is none of these

This is the part of the page that matters most, and it is why the list is short.

Sustained heat. None of these three belongs near a heat source. Not next to a motor running hot, not in an oven, not near an exhaust or an element. This is the most common reason a printed part fails in service, and it is entirely predictable in advance.

Safety-critical parts, and heavy structural load. Worth separating this from ordinary load, because the two get confused. Parts carrying everyday and moderate loads are routine work here — brackets, mounts, housings, guides, gears and light drive components. The firm line is where failure is dangerous: anything holding a guard on, anything above a person, anything a compliance obligation attaches to, and anything carrying heavy sustained structural load. Those want metal and a different process, and you will be told so.

Food contact and drinking water. Printed parts have layer lines that hold bacteria and are not made to a food-safe standard here. Not offered, regardless of filament claims.

Watertight without help. Printed parts leak along the layers. A part that must hold liquid needs sealing afterwards, or designing around a bought seal or gasket.

Chemical exposure. Solvents, fuels and some cleaning agents attack these materials. Worth raising before anything is printed.

Fine threads and precision bearing faces. Better machined, or designed around a bought insert, bushing or heat-set thread. Sliding wear surfaces are a more interesting case: a printed wear part does wear faster than a metal one, but it also costs very little to reprint. Where a component is genuinely consumable — a guide, a wiper, a pad, a sacrificial face that protects something expensive — that trade often makes sense, and a printed part that is cheap to replace can beat a metal one that is not. Where the part has to hold its geometry for years under constant contact, it does not.

Higher-temperature or fibre-filled engineering filaments. Not printed here. The equipment in this workshop is not set up for them, and printing them badly is worse than not printing them at all. If your part genuinely needs one, you will be told and pointed elsewhere rather than sold something that will not hold up.

Layer height, and where the detail comes from

Material is only half of it. How finely a part is printed is chosen per job, from 0.1 mm where detail and surface matter to 0.3 mm where speed and bulk matter more. It is a real trade rather than a setting left on a default: a fine layer height follows a curve more closely and gives a better surface, and it takes considerably longer to print.

On some parts it decides whether the thing works at all. The planetary gear set was printed fine for exactly that reason — on a gear tooth, layer height is what determines how closely the printed flank follows the intended profile, and a coarse setting turns a smooth curve into a staircase that meshes roughly and wears quickly. On a bracket that nobody will look at twice, the same setting would be wasted money.

You do not need to specify it. Say what the part does and it gets chosen, along with the material and the print orientation, as part of the quote.

Why a short list is a feature

A supplier offering fifteen materials is not necessarily offering fifteen materials they print well. Three materials, used constantly, means the settings are known, the failure modes are familiar, and the advice on this page comes from parts that actually failed rather than from a datasheet.

It also means the honest answer arrives early. Being told at the enquiry that a part needs metal costs you a day. Finding out costs you a part, a machine, and a fortnight.

Frequently asked questions

Which material should I pick?

You do not need to. Describe what the part does, what it touches, where it lives and what it is exposed to, and the material gets chosen as part of the quote. If you have a preference, say so and you will be told whether it suits the job.

Can you print in a specific colour?

Within the range of filament stocked and available, usually yes. Exact colour matching to an existing part is not possible with off-the-shelf filament — expect close, not identical.

Is PETG waterproof?

Not on its own. Printed parts leak along the layer lines regardless of material. A part that has to hold liquid needs thicker walls and sealing afterwards, or should be designed around a bought seal.

Can you print a part that goes in a dishwasher or near a heater?

No. Sustained heat is the most common cause of a printed part failing in service, and none of these three materials belongs near it. You will be told this before anything is printed rather than afterwards.

Do you print higher-temperature or fibre-filled materials?

No. This workshop is set up for PLA, PETG and flexible TPU, and printing anything else badly would be worse than not offering it. Where a part genuinely needs another material, you will be pointed elsewhere.