How to choose 3D printer filament: the properties that actually matter

"Which 3D printer filament is best?" is the question every beginner asks, and it is the wrong one. The filament that makes a beautiful desk figurine is a poor choice for a drone frame. The material that survives a hot car will annoy you on a quick draft print. The right question is: what does this part need to survive?

Most guides answer by listing 3D printer filament types and uses. This one starts from the decisions. We will walk through the handful of properties that decide whether a printed part works, in plain language first and measured numbers second. You do not need an engineering degree. You just need to know what to look for.

Where we quote exact figures for our own materials, they are typical values from our published technical data sheets. For material classes we do not yet publish data on, we say so and describe typical class behaviour honestly.

Before you buy: can your printer print it?

Start from the part, but check the printer before you buy the spool. Five things decide whether a filament will even work on your machine:

• Filament diameter. Usually 1.75 mm, sometimes 2.85 mm. Buy the one your printer takes.
• Maximum nozzle temperature. PLA and PETG sit comfortably below 250 ¬∞C. Some engineering materials need considerably more.
• Maximum bed temperature. Matters for ABS, ASA, nylon and other higher-temperature materials.
• Enclosure. PLA and PETG print happily on open printers. ABS and ASA want an enclosed, draught-free space.
• Nozzle material. Standard brass is fine for PLA, PETG and TPU. Fibre-filled filaments are abrasive and need a wear-resistant nozzle.

One more thing for ABS and ASA: an enclosure controls warping, but it is not the whole story. Print these materials with sensible ventilation in the room.

A material can be right for the part and still wrong for your printer. Check both before spending money.

A quick decision guide

If you only came for the answer, start here. The rest of the guide explains every line.

• Your first spool ever: PLA.
• Visual model, prototype or fine detail: PLA.
• General functional part: PETG.
• Holding weight for months: PETG or an engineering polymer. Never choose by tensile strength alone.
• Must not bend at all: PLA, the stiffest of the common materials.
• Should bend and bounce: TPU, the flexible one.
• Will get dropped or crashed: tough grades of ABS, ASA or nylon first, PETG second, PLA last.
• Lives somewhere warm: PETG. Somewhere genuinely hot, like a car in summer sun: ASA or ABS.
• Pulled apart across its layer lines: mind print orientation and Z-strength. PETG keeps roughly 94 % of its strength across layers per our TDS.
• Needs gears, threads or parts that fit together: a tightly controlled filament diameter plus a calibrated printer.
• Lives outdoors in the sun: ASA.
• Lives near oils, grease or coolant: PETG, and check the chemical resistance table for your exact chemicals.

How to read a filament spec sheet (no engineering degree required)

Most filament marketing runs on adjectives: strong, tough, easy, engineering-grade. A technical data sheet (TDS) replaces adjectives with numbers, and that is the whole trick to choosing well. A line like "tensile strength 48 MPa, ISO 527" tells you what was measured and by which standard method.

One honest qualification. A shared standard makes comparison meaningful, but only when the details behind it are similar too: how the test specimens were printed, their orientation, their conditioning and the test conditions. The standard gives two sheets a common language. The details tell you whether the comparison is fair. If a filament page offers no numbers and no methods at all, that silence is also information.

Two more things to know before the numbers make sense.

First, a printed part is like a stack of glued layers. It is strong along the layers and weaker across the glue lines. Engineers call this anisotropy, and sheets show it as X-Y values (along the layers) and Z values (across them). The gap between those two numbers often says more than either number alone.

Second, TDS figures are typical values, not promises. Your printer, your settings and the shape of your part all shift the result. Honest sheets say this openly.

The nine properties that decide your part

1. Strength: the load it holds before breaking (tensile strength)

Strength, in everyday terms, is how hard you can pull on the material before it breaks. On a data sheet it is called tensile strength and measured in megapascals (MPa): load per unit of area. A higher number holds more.

Picture a printed wall hook holding a bag of tools. Our PLA-PRO prints at a typical 56 ±3 MPa along the layers and our PETG at 48 ±2 MPa, and both hold serious weight when printed in the right orientation.

So PLA wins? Not necessarily. A tensile test lasts minutes. Your hook hangs there for months, and plastics slowly and permanently deform under constant load, an effect called creep that grows with temperature. PLA is particularly sensitive to it, so for sustained loads PLA is the risky choice. PETG or an engineering polymer usually holds up better, and long-term load and temperature still matter for any plastic. The lesson is simple: never choose a material from tensile strength alone.

2. Stiffness: how much it bends (Young's modulus)

Stiffness is how much the material resists bending and stretching. A printed ruler should be stiff. A phone case should not be. The data sheet number is Young's modulus, again in MPa: higher means more rigid, lower means more flexible, and neither is better until you know the part.

For a straightedge, a tool jig or a camera mount that must not flex, PLA leads the common materials at a typical 2 850 ±100 MPa, ahead of PETG at 2 200 ±60 MPa (both per our TDS). For a clip or a bumper that should flex, TPU is the answer: a rubber-like filament class around Shore 95A hardness, roughly as firm as a hard skateboard wheel. The classic beginner mistake is printing a snap-fit clip in PLA because it happened to be loaded. PLA stretches only 6 ±2 % before it breaks, so it cracks instead of clicking.

These two properties get mixed up constantly, so here is the difference. Stiffness is how much it bends. Strength is how much load it takes before breaking. A material can be stiff but easy to snap, like a biscuit, or soft but hard to tear, like a rubber band.

3. Toughness: surviving drops and crashes (impact strength)

Toughness is the material's ability to absorb a hit without shattering. It is a different property from strength, and for many parts the more important one. Sheets report it as Charpy impact strength in kJ/m²: a standard test bar is struck by a pendulum, and the number is the energy it soaks up while failing. A related clue is elongation at break: how far the material stretches before snapping. More stretch usually means more forgiveness.

The drone frame is the honest example. On paper PLA looks strong, and it is, but with a typical notched Charpy of 3.0 ±0.5 kJ/m² and 6 % elongation it is brittle: in a crash it cracks and shards. Our PETG absorbs 5.0 ±0.6 kJ/m² and stretches 9 ±2 %, so it bends and scuffs where PLA snaps. Many ABS, ASA and nylon grades offer substantially higher impact resistance still, with big differences between formulations. If your part will be dropped, stepped on or flown into a tree, read the impact row first.

A useful detail: sheets often list Charpy twice, unnotched and notched. Unnotched (a typical 16 ±3 kJ/m² for our PLA-PRO, 20 ±3 for PETG) describes a smooth bar. Notched describes a bar with a deliberate weak spot, because sharp corners, screw holes and layer lines all concentrate stress the same way. Real parts have those, so design against the notched number, and give printed corners generous rounding. It is free toughness.

4. Heat resistance: when plastic goes soft (Tg and HDT)

Every common filament is a thermoplastic: warm it enough and it softens. Two numbers tell you when. The glass transition temperature (Tg) is where the material starts to soften and lose stiffness. It is not melting. The part does not turn into a puddle, it just stops being reliably rigid. The heat deflection temperature (HDT) asks the more practical question: at what temperature does a loaded test bar visibly start to bend? For a loaded part, HDT is the number to design against.

The phone mount on a car dashboard is the classic failure. A car parked in summer sun easily passes 60 °C inside. PLA, with a typical Tg of 60 °C and an HDT of 55 °C at 0.45 MPa per our TDS, slumps into modern art. Our PETG holds a Tg of 80 °C and an HDT of 70 °C at 0.45 MPa, enough for a warm interior but not for a loaded mount in direct sun on the windscreen. For that job the class-typical answer is ASA or ABS, whose Tg sits around 100 °C. Match the number to the hottest hour your part will ever see, not to the average day.

5. Layer adhesion: why prints split along their lines (Z-strength)

Remember the stack of glued layers? Layer adhesion is the quality of that glue, and it is where printed parts usually fail. Like a brick wall, a print rarely breaks through the bricks. It breaks along the mortar. The measurable version is the Z tensile strength compared with X-Y: how much of the material's strength survives the layering.

The ratio differs between materials more than most people expect. Per our TDS values, PLA-PRO keeps 36 of its 56 MPa across layers, roughly 64 %. PETG keeps 45 of 48 MPa, roughly 94 %, one of the best layer-adhesion profiles among common filaments. If your part is loaded across its layers, say a hook printed flat so the bend pulls its layers apart, that ratio decides whether it works or splits along a line.

The first fix is not a different filament. It is orientation: turn the part so the main load runs along the layers wherever the shape allows. When the shape will not cooperate, the Z number decides.

6. Diameter tolerance: the spec almost everyone skips

Filament is sold by diameter, usually 1.75 mm, and every sheet lists a tolerance: how far a spool may wander from that size. It reads like fine print. It is not.

Here is why. Your slicer calculates how much plastic to push through the nozzle from the filament's cross-section. When the diameter is off, the flow is off, and the error is amplified: because flow depends on the diameter squared, a 1 % diameter error becomes roughly a 2 % flow error. The percentage doubles. Along one long print, diameter drift shows up as uneven walls, blobby seams and parts that measure slightly wrong.

Filament tolerances on the market commonly run about ±0.02-0.05 mm depending on the manufacturer. Infilus targets ±0.02 mm and measures the diameter in-line on every spool as it is extruded, not on samples afterwards. We track roundness on the same laser system, because an oval filament changes flow the way a diameter error does. Our TDS lists roundness above 99.5 %, meaning the cross-section stays within half a percent of circular.

Be clear about what this buys you. Tight diameter control does not control dimensional accuracy on its own. Flow calibration, cooling, shrinkage, printer mechanics and slicer settings all matter too. What it does is remove one important source of extrusion variability before the spool ever reaches your printer, and that matters most for gears, threads, thin walls, mating parts and long prints.

7. Moisture: the silent print killer (hygroscopic filament)

Most filaments are hygroscopic: they slowly absorb water from the air, the way salt clumps in a humid kitchen. That absorbed water boils inside the hot nozzle, and the symptoms are unmistakable: popping sounds, stringy wisps between features, rough surfaces, tiny bubbles and layers that bond weakly. If a spool that printed beautifully a month ago suddenly prints badly, suspect moisture before you touch the settings.

Data sheets report moisture absorption under a defined test (ISO 62, 24 hours). Read it as a comparative property between materials, not as a stopwatch for how long an open spool survives beside your printer. Real uptake depends on temperature, humidity and the exact formulation. Our typical 24-hour values: PETG 0.18 %, PLA-PRO 0.30 %. Nylon, class-typically, absorbs several times more and demands much stricter discipline.

The practical rules are simple. Store opened spools sealed with fresh desiccant, and keep highly hygroscopic materials like nylon out of room air as much as possible. Dry a spool when it has been exposed too long, following the sheet for that exact material: ours specify 45 °C for 4 h for PLA-PRO after more than a week in humid air, and 65 °C for 4 h for PETG after 48 hours. Hotter is not better. Too much drying heat deforms both the filament and the spool.

8. Sun, weather and chemicals

Outdoor and workshop parts usually fail from their environment, not from load. Two properties cover it.

UV resistance decides outdoor life. A garden camera bracket printed in PLA or PETG will yellow, grow brittle and weaken over seasons of direct sun. That is class behaviour, not a brand defect. Both manage short and medium-term duty outdoors. For parts that must still behave after several summers, ASA is the material engineered for the job, holding its colour and its mechanics, with ABS a step behind.

Chemical resistance decides life on the workbench. Per our TDS, PETG copes well with oils, greases, alcohols and weak acids, which is why it suits parts living near lubricants and coolant. PLA tolerates oils and alcohols only briefly. Neither survives ketones such as acetone, aromatic solvents or strong acids and alkalis. Never trust the words "chemical resistant" on their own. Check the resistance table for your material and your chemical, before printing rather than after.

9. Ease of printing: the honest property

No ISO test measures how much of your evening a filament costs, but for a beginner this may be the most important property on the list.

PLA is the forgiving end, and the starting conditions on our sheets show it: 200-220 °C at the nozzle, a 50-60 °C bed, the cooling fan high, no enclosure. PETG asks a little more: 230-250 °C, a 70-85 °C bed, the fan held back, plus a bit of retraction tuning to control its one habit, stringing. Still no enclosure.

On speed, be sceptical of any universal number. How fast you can print is a printer property as much as a material property: what limits it is volumetric flow, the cubic millimetres per second your hotend can actually melt. A modern high-flow machine and an older printer are different worlds at the same nominal speed. Start from the profile recommended for your printer and material, then tune.

ABS and ASA are the demanding end. They shrink more strongly as they cool, so they want an enclosed printer, stable air and sensible room ventilation. Nylon adds strict dryness discipline on top. Be honest about your printer and your patience. A material you cannot print reliably is the wrong material for you, whatever its numbers say, and there is no shame in printing PLA for a year first.

PLA vs PETG: the everyday decision

For most printers, most days, the real choice is between these two: PLA, a plant-derived polylactic acid (the base polymer behind NatureWorks Ingeo and most of the world's PLA), and PETG, a glycol-modified cousin of the polyester in drinks bottles. Here is the measured comparison, typical values per our published TDS:

There is no overall winner in that table, and that is the point. PLA is the tool for rigid, detailed and easy prints, and plenty of them are functional: jigs, cases, brackets that live indoors at room temperature. PETG is the tool for tougher parts that face heat, moisture and everyday handling. Choose by the part's job. When you are ready to print functional parts, our PETG guide covers the printing side step by step.

FAQ: choosing filament in practice

Choose by the part, not the polymer

Different 3D print filaments are good at different things, and honest data sheets show you exactly what. That is all choosing filament for 3D printing comes down to. Read the numbers. Look at the direction of the load, not just its size. Think about the hottest day, the fall from the workbench, the screw hole in the corner and the month beside an open window. Then choose the easiest material that does the job.

Explore the Infilus range