Crafted in 3D | Material Selector

Material Selector

Job Requirements

Set what the part has to do. The list narrows to materials that qualify.

Operating temperature

°C

Bears mechanical load at that temperature

Checked against the continuous-use rating, or the light-load deflection point where there is none. With the box clear, a material can qualify on a no-load rating far above the temperature where it bends under any load (unfilled PPS: rated 200 C, deflects at 90 C). Ticking the box adds the deflection test under load. It can only narrow the list.

Mechanical priority

Chemical exposure (select any)

Outdoor / UV exposure Flame rating required (V-0)

Max budget tier

$$$$$

Allowed forms

0 of 52 materials qualify

Select a material's name for its full property sheet. Click a column's i for what it means, or any flammability rating for its definition.

Operating temperature is matched to the continuous-use rating, or to the light-load deflection point where no maker publishes a rating (lower bound, conservative). Ticking load adds the deflection test at 1.80 MPa, or at 0.45 MPa where that is all the sheets give. A material with no published heat figure drops out when a temperature is entered. Cells are shaded by where each material ranks among those currently shown in that column, low to high, so you can scan each column’s spread at a glance. The shading fades when your shown options are close in value. Tensile and heat deflection are printed test bar figures from data sheets, and a range spans the makers we found. A row whose data sheets do not all say how the bars were made says so under Basis of the figures. Stiffness and flex figures are printed test bars too, except where that line says otherwise, as for a range that mixes annealed and as-printed bars. Glass transition is a reading of the polymer itself, mostly by DSC, not a bar test. Continuous use and flame are data-sheet ratings of the resin, not tests of a printed part. Impact cells name their unit; J/m cells are not ranked against kJ/m2 cells. Cost is the median regular list price per kilogram of the spools we priced on 3 October 2026. n/a means no published figure was found (the row’s Basis line says which). Best for lists typical uses of the polymer. Verify against the spool datasheet for anything design- or safety-critical. No printed part is food-safe as printed, and our parts are not certified for medical or other critical use unless we agree to it in writing.

What every column means, in plain terms, with the mechanics drawn out. Tap any row to expand it: the stats include a diagram of what is actually being measured and what counts as high or low. Tensile, heat deflection, impact and elongation are printed test bar figures from data sheets unless the cell or the row's Basis of the figures line says otherwise. UV, water uptake and chemical resistance come from published tests, ratings and tables, named by kind in that line. Stiffness and flex figures are printed test bars too, except where that line says otherwise, as for a range that mixes annealed and as-printed bars; glass transition is a reading of the polymer itself, mostly by DSC, not a bar test. Where nothing is published, a cell reads n/a (the row’s Basis line says which).

Identity and classification

The base polymer (for example PETG, nylon, or PEEK) decides the part's fundamental nature: roughly how strong, how tough, and how heat- and chemical-resistant it can be. The form (next row) then modifies that base. Think of the material as the engine and the form as the tuning. Everything else in this guide is a measurable consequence of these two choices.

Neat is the pure polymer with no fillers. It usually stretches further before it breaks than a filled grade of the same polymer. CF blends in short carbon fibers: the stiffest form, flatter and more stable in the print plane, but it breaks after less stretch and is slightly denser than the plain polymer. Treat it as neither an insulator nor ESD-safe unless the spool's sheet gives a resistance range. GF blends in short glass fibers: stiffer and flatter than the plain polymer. Whether a filled grade takes a notch better or worse, or costs more or less, than the unfilled one differs by polymer and maker: compare the Impact and Cost cells.

A quick bucket for what the material is for and how it is priced. Standard covers the everyday materials (PLA, PETG, PCTG, ABS, ASA) and the flexible TPU and TPE. Engineering materials are for functional parts with real mechanical, chemical, or thermal demands (PC, PET, the nylons, PP, PVDF, PEBA). High-Temp is the high-performance specialist class, from PPA up through PEEK. Most materials price instantly on the quote page. Tier tracks price as much as performance: the cost bands climb with it.

The band is the median regular list price per kilogram of the filament spools we priced for that row on 3 October 2026: under 30 USD is $, 30 to 60 is $$, 60 to 150 is $$$, 150 to 400 is $$$$, and 400 and over is $$$$$. Sale prices are left out. US prices are before sales tax; euro prices, used only where a row has no US price, are before VAT and converted at the European Central Bank rate of 2 October 2026 (1 EUR = 1.1225 USD). It is the price of the filament only, not the price of your part: an instant price counts the part's weight in that material and the hours it spends on the machine, and the hourly rate differs by material too. The step-down suggestion in each material's detail is worth taking when the part does not truly need the top grade.

Mechanical properties

Tensile strength is the most stress the material can take while being pulled apart, measured along the strong in-plane direction (1 MPa is about 145 psi). On the stress-strain diagram below it is the height of the peak: the highest force the part carries before it fails. Rough landmarks from the rows here, all printed test bars: PLA and PETG run 35 to 60 MPa, the nylons and PC about 50 to 80, PEEK about 100. A fiber-filled grade is stiffer, not always stronger: several carbon and glass rows sit no higher than their unfilled grade. Higher means it carries more pulling load before breaking. Read it together with elongation and impact, because a high number alone does not tell you whether it fails gracefully or shatters.

A printed wall is built layer on layer. This figure is the tensile strength of a bar printed upright and pulled across its layers, from the same data sheets as the Tensile column.

Modulus is stiffness, which is not the same as strength. On the stress-strain diagram it is the slope of the line at the start: a steep slope means the part barely moves under load and feels rigid, a shallow slope means it gives and flexes. (1000 MPa equals 1 GPa.) A part can be very stiff yet snap easily, or quite flexible yet nearly unbreakable, so stiffness has to be read next to strength and toughness. Fiber reinforcement raises modulus dramatically, which is exactly why CF and GF grades feel so rigid even though they are more brittle. Nylon cells give dry or as-printed, conditioned and wet figures, each with its condition, and the table ranks and shades by the dry or as-printed figure. Conditioned means a stated hold in humid air, hours or days (24 to 40 hours at 50 percent RH for the printed PA6 figures): a short hold, not months in humid air. Wet means soaked in water.

Where tensile pulls a part straight, flexural loading bends it, which is what most real brackets, arms, clips, and housings actually experience. Flexural strength is the most bending stress the outer surface survives before it yields or cracks. A tough or rubbery bar may bend without breaking; its data sheet then gives the stress at a stated bend, often 5 percent strain, and the cell shows that figure, so it is the stress at that bend and not a breaking strength. The diagram shows the setup: the part is supported at two ends and pushed in the middle. For a part loaded across its length, this number is often more relevant than the straight tensile figure. Nylon cells give dry or as-printed, conditioned and wet figures, each with its condition, and the table ranks and shades by the dry or as-printed figure. Conditioned means a stated hold in humid air, hours or days (24 to 40 hours at 50 percent RH for the printed PA6 figures): a short hold, not months in humid air. Wet means soaked in water.

Flexural modulus is stiffness measured in bending, and for printed parts it is usually the most useful stiffness number of all, because it directly predicts how far a beam or bracket deflects under a given load. A high flex modulus means the part barely sags; a low one means it visibly bows. Reinforcement raises it more than almost any other property, so a CF or GF grade is the first lever to pull when a part flexes too much under load. Nylon cells give dry or as-printed, conditioned and wet figures, each with its condition, and the table ranks and shades by the dry or as-printed figure. Conditioned means a stated hold in humid air, hours or days (24 to 40 hours at 50 percent RH for the printed PA6 figures): a short hold, not months in humid air. Wet means soaked in water.

Impact strength is the energy a notched test bar absorbs before it breaks: a pendulum strikes a bar with a V notch cut or printed into it, so the test shows how readily a crack or a sharp corner runs through the material. Each cell names its unit and test: kJ/m2 for the ISO tests (notched Charpy, ISO 179, or notched Izod, ISO 180) and J/m for the ASTM notched Izod (D256). The two units cannot be converted into each other, so the table ranks and shades the kJ/m2 cells only and lists J/m cells after them. Charpy and Izod bars are set up differently, so compare like with like. Most figures are printed bars from filament makers' data sheets; a cell says so when the bars were molded, extruded stock or another maker's filament. 'No break' means the bar bent without breaking. Nylon cells give dry and wet figures, which differ several times over (PA6: 2.3 to 6.9 kJ/m2 dry or after 24 hours at 50 percent RH, 28 wet).

Elongation at break is how far a test bar stretches, as a percentage of its length, before it breaks. High values mean the part bends and deforms before it fails; low values mean it breaks with little warning. The figures are printed bars pulled in the plane of the layers unless the cell says otherwise. A range spans the sheets we found; a cell that gives a conditioned nylon beside a dry one says which is which.

Thermal properties

Heat Deflection Temperature is found by heating a supported bar under a fixed load until it bends a set amount (see the diagram). The light-load version uses 0.45 MPa (66 psi) and is the optimistic heat number, appropriate for parts that are unloaded or only lightly stressed when hot, such as a cover or a duct. On any one sheet it is higher than the under-load figure, and the gap varies a lot: molded unfilled nylon 6 drops from 150 to 180 C at the light load to 55 to 65 C under the heavy one, and the printed nylon 6 row here shows the same pattern. In this table the two columns can come from different makers, so a few rows show the reverse. Do not use it for anything bearing real force at temperature.

The same heat-deflection test, but under a heavy 1.8 MPa (264 psi) load. This is the realistic hold-its-shape-under-working-stress number and the one to use for any load-bearing part that gets hot: a bracket on a car dashboard in the sun (in one Arizona study, dashboards of cars parked in summer sun averaged 69 C after an hour) or a part near an engine. It is always the lower of the two HDT values, so when in doubt, design to this one.

The glass transition is where the polymer's amorphous (disordered) regions change from a hard, glassy state to a soft, rubbery one. Whether Tg is the usable-temperature ceiling depends entirely on the polymer type, which the diagram shows. For amorphous plastics (PC, PEI/ULTEM, PSU, PPSU, PETG, ABS) there is nothing but amorphous material, so stiffness falls off near Tg and that sets the practical limit. For semi-crystalline plastics (PEEK, PPS, the nylons, PP, PVDF) the crystalline regions act like a skeleton that holds the part together well above Tg, all the way toward the much higher melting point, so the usable temperature sits far above Tg. This is why PEEK, with a Tg of only 143C, runs continuously at 250C, while ULTEM 1010, amorphous with a much higher Tg of 217C, tops out around 170C.

Continuous service temperature is a heat-aging limit: how hot the material can sit for thousands of hours and still keep most of its properties once it cools. It is not strength while hot: unfilled PEEK stock keeps about a tenth of its flexural strength at 260 C. It is the most life-relevant thermal number for anything that lives hot, as opposed to a brief peak it could survive once. As the modulus-versus-temperature diagram shows, semi-crystalline materials carry this number far above their Tg, while amorphous materials land at or below theirs. For a part that sits hot with no load, design to this figure with a margin. For a part that carries load while hot, use HDT under load, or HDT light where the load column is n/a. For unfilled semi-crystalline grades it is far lower.

Environmental and chemical

Each cell says what was published and how: a weathering test with its exposure (hours of lamp UV or xenon, or years outdoors) and what changed, or a maker's or producer's statement, marked as such. A cell with nothing published reads n/a. The Outdoor / UV box keeps a row when a weathering test kept at least 90 percent of its strength after 1,000 hours or more, or years outdoors; when the resin carries a UL 746C f1 outdoor rating; or when a maker or producer states UV or weather resistance and no source we found reports a larger loss or says it degrades. Color can change while strength holds: printed PEI 9085 bars kept their strength through 1,000 hours of lamp UV and turned darker and yellower. The authors of a 12-month outdoor test of printed ASA report that ABS, by comparison, yellowed, cracked at the surface and lost hardness, and one printer maker's data sheet lists long-term outdoor use as not suitable for PLA. ASA gives up strength slowly, with mild discoloration: in that test it lost up to 4 percent of its strength along the layers and up to 18 percent across them. That is why we suggest ASA for printed parts that sit in the sun. One filament maker advises paint or an anti-UV treatment for ABS parts used outdoors; we do not paint or coat parts.

Written as Resists and Weak against, item by item, from a filament maker's, resin producer's or stock-shape producer's table, with the strength, temperature, time or load the table gives. Where producers disagree, the item says who said what (one filament maker, another, the resin producer). Where no table exists for the grade, the cell says so, or shows the unfilled resin's table under a label that says it is not this grade. A material under Weak against may swell, soften, craze, crack under load or dissolve, and several cells show a chemical a part resists unloaded and fails under strain. The chemical boxes drop a row whose Weak against list names that family. For anything critical, check the exact chemical, strength and temperature against the producer's table.

UL94 is the standard small-flame test: a specimen is held in a set orientation, a flame is applied, and the tester records whether and how fast it self-extinguishes and whether it drips burning material. Ratings run from least to most flame-retardant, and the rating depends on part thickness. 'Not rated' or 'HB' effectively means it will burn, so treat any unrated material as flammable.

HBHorizontal Burn, the baseline. Burns slowly along a horizontal bar or stops before a mark. It will burn, just slowly. Most standard filaments (PLA, PETG, ABS, ASA, PP, nylon) are HB at best.

V-2Self-extinguishes within 30 seconds held vertically, but is allowed to drip flaming particles that could ignite material below.

V-1Self-extinguishes within 30 seconds and does not drip flaming particles.

V-0Self-extinguishes within 10 seconds and does not drip flaming particles. The V-0 ratings for PEEK, PEKK, ULTEM, PPS, PPSU and PVDF are base-resin ratings at a stated thickness (0.8 mm on the PEKK sheet, 1.5 mm on the PPS and PVDF sheets). A printed part is not itself UL listed. Standard polysulfone (PSU) is HB.

5VB / 5VAThe most demanding ratings, using a larger flame, for enclosures and structural fire-safety parts. 5VA is highest (no burn-through allowed); 5VB permits a hole. Few FDM materials reach these without special grades.

FSTBeyond UL94: aircraft and rail interiors must meet Flame, Smoke, Toxicity standards such as FAR 25.853. ULTEM 9085 resin meets the FAR 25.853 limits on its maker's data. A finished part still needs its own burn test. We do not certify parts for this use.

Water uptake as a percentage of weight, with the condition the source used: 24 hours in water, a stated number of days, saturation in water, or equilibrium in humid air at a stated relative humidity. The conditions give very different numbers for one material (molded PA6: 2.6 to 3.4 percent at 50 percent RH, 9 to 10 percent saturated in water), so compare like with like. Each cell says where its figure comes from. Water a polymer takes up swells and softens the part; the nylons take up the most (molded PA12 0.7 percent at 50 percent RH against 2.6 to 3.4 for PA6).

Selection guidance

The sweet-spot uses where this material's particular balance of properties makes it the correct choice. Start here when matching a part to a material: if your part looks like one of these, this row is a strong candidate.

The conditions where this material will disappoint or fail, usually paired with a pointer to the material that handles that case better. Reading this row saves more bad parts than any other, because most material mistakes are not picking a weak material, they are picking a capable material for the one job it happens to be bad at.

A material the row names as the step down, shown only where its cost band, from the same list prices, is lower than this row's, with its band beside it. It is compared with that material's row in the same form (CF or GF) where the selector has one. It is cheaper, not equal: check its figures against what the part has to do. Each row is priced from the spools found for it, which are not always one maker's line, so the gap can reflect the line priced as much as the polymer.