Thermal conductivity tells how readily heat moves through a material, and for ordinary FFF/FDM filaments it is usually low compared with metals. In controlled tests of near-solid printed specimens, common unfilled polymers such as PLA, PETG, and ABS often fall around 0.17–0.23 W/m·K, but the value for a finished print can change markedly with infill, internal voids, print direction, temperature, and formulation. A 2025 same-method study measured 100% infill PLA at 0.182 W/m·K, PETG at 0.198 W/m·K, and ABS at 0.172 W/m·K near room temperature.[a]
| Material | Test Temperature | Thermal Conductivity | Test Context |
|---|---|---|---|
| PLA | 22.12 °C | 0.182 ± 0.006 W/m·K | 100% infill, ASTM E1530 guarded heat flow |
| PETG | 22.36 °C | 0.198 ± 0.006 W/m·K | 100% infill, ASTM E1530 guarded heat flow |
| ABS | 22.03 °C | 0.172 ± 0.006 W/m·K | 100% infill, ASTM E1530 guarded heat flow |
These are measured values for specific printed specimens, not universal constants for every PLA, PETG, or ABS spool. The value printed on a datasheet is only comparable when the material state, test method, direction, temperature, and specimen structure are also known.[a]
What Thermal Conductivity Measures
Thermal conductivity is commonly written as k or λ and expressed in watts per meter-kelvin, W/m·K. A higher value means heat can pass through the material more readily for the same temperature gradient and geometry. A lower value means the material offers more resistance to conductive heat flow.
The property describes heat transfer through the material. It does not tell how hot a part can become before it softens, creeps, loses stiffness, or deforms under load. Those behaviors are described by other thermal properties.
| Property | What It Describes | Typical Unit |
|---|---|---|
| Thermal conductivity | How readily heat is conducted through the material | W/m·K |
| Glass transition temperature (Tg) | Temperature region where an amorphous polymer phase becomes much more mobile and softer | °C |
| Heat deflection temperature (HDT) | Temperature at which a test specimen reaches a defined deformation under a specified load | °C |
| Specific heat capacity | Energy needed to raise the temperature of a unit mass | J/kg·K |
| Thermal diffusivity | How quickly a temperature disturbance spreads through a material, combining conductivity, density, and heat capacity | m²/s |
High heat resistance does not mean high thermal conductivity. PEEK and PEI can retain useful properties at temperatures where PLA would soften, yet printed specimens of these engineering polymers still show thermal conductivity values in the low fractions of a watt per meter-kelvin rather than metal-like heat transfer.[f]
How Thermal Conductivity Is Measured in Printed Polymers
There is no single universal test setup behind every thermal-conductivity number. ASTM E1530 uses a guarded heat flow meter technique for determining thermal resistance under steady-state conditions and can be used to determine conductivity for suitable homogeneous opaque solid specimens. The active ASTM E1530-25 edition covers specimens less than 25 mm thick and emphasizes control of contact resistance between the specimen and test surfaces.[b]
ISO 22007-2:2022 specifies the transient plane heat source, often called the Hot Disk method, for measuring thermal conductivity and thermal diffusivity of plastics. It can be configured for isotropic materials as well as materials with uniaxial anisotropy, which is relevant when a printed structure conducts differently along and across deposited roads.[c]
Other research uses light-flash or laser-flash methods, hot-box assemblies, or instrument-specific guarded heat flow systems. The result can still be useful, but a value from one method should not be treated as directly interchangeable with a value from another method unless the specimen and test conditions are sufficiently comparable.
A resin datasheet, an injection-molded plaque, a filament strand, and an FFF part are different specimen states. For printed components, voids, bead interfaces, orientation, crystallinity, additives, and actual density can shift the measured result. A number reported for bulk resin should not be relabeled as the conductivity of a printed part.
Why the Printed Structure Can Matter More Than the Polymer Name
FFF does not create a perfectly continuous block of polymer. The part is assembled from deposited roads, with boundaries between adjacent roads and between layers. Even when a slicer is set to 100% infill, microscopic gaps can remain. Air has a much lower thermal conductivity than ordinary thermoplastics, so voids interrupt conductive paths and lower the effective conductivity of the printed structure.[d]
Infill Density Changes the Effective Conductivity
The 2025 PLA/PETG/ABS study provides a controlled example. PLA specimens printed with grid infill measured 0.129 W/m·K at 40% infill, 0.140 W/m·K at 60%, and 0.155 W/m·K at 80% near room temperature. The 100% infill PLA specimen measured 0.182 W/m·K. The material label stayed the same while the internal structure changed.[a]
| PLA Structure | Measured Temperature | Thermal Conductivity |
|---|---|---|
| 40% grid | 21.60 °C | 0.129 ± 0.004 W/m·K |
| 60% grid | 21.73 °C | 0.140 ± 0.006 W/m·K |
| 80% grid | 21.27 °C | 0.155 ± 0.005 W/m·K |
| 40% honeycomb | 21.59 °C | 0.131 ± 0.005 W/m·K |
| 60% honeycomb | 21.83 °C | 0.149 ± 0.005 W/m·K |
| 80% honeycomb | 21.96 °C | 0.164 ± 0.006 W/m·K |
| 100% infill | 22.12 °C | 0.182 ± 0.006 W/m·K |
This comparison also shows why a material-only ranking can mislead. In the same experiment, changing PLA from 40% grid to 100% infill changed conductivity by about 41%, while the difference between 100% PLA and 100% PETG near room temperature was under 9%. For some designs, structure can therefore move the result more than switching between two ordinary unfilled filament families.[a]
Infill Pattern Changes the Heat Path
Infill percentage tells how much material the slicer intends to place inside the part; it does not describe how continuous the heat path will be. In the 2025 study, honeycomb specimens generally conducted more heat than grid specimens at the same nominal infill level. A separate PETG study that held density at 25% found a much wider geometry effect across twelve patterns: concentric PETG measured 0.057 W/m·K, while several patterns produced higher values because their internal paths connected the hot and cold boundaries differently.[e]
The PETG result should not be used as a universal value for “25% PETG.” It belongs to that specimen geometry and hot-box test. Its value is in showing that topology matters. A sparse part designed for insulation should minimize continuous polymer bridges in the heat-flow direction, while a heat-spreading part benefits from more continuous material paths.
Flow, Porosity, Walls, and Solid Skins Also Matter
Actual density depends on extrusion as well as the infill setting. A study of material-extrusion PLA varied the extrusion multiplier from 0.97 to 1.05. As measured porosity fell from 5.82% to 0.05%, thermal conductivity rose from 0.157 to 0.188 W/m·K. Above an extrusion multiplier of 1.03, conductivity stopped improving even though excess material began to hurt dimensional accuracy.[d]
Wall count and top/bottom thickness can alter the same heat path. Perimeters and solid skins are usually denser than sparse infill, so two parts with the same nominal infill percentage can have different effective conductivity when one has thicker shells. Layer height, bead width, flow calibration, and under-extrusion can also change how much polymer-to-polymer contact exists. There is no reliable rule that simply making the layers thinner will always raise or lower conductivity; the resulting interfaces and void geometry matter.
Print Direction Makes Thermal Conductivity Directional
A printed polymer should not automatically be treated as isotropic. Heat moving along a continuous extrusion road can encounter a different path than heat moving across layer boundaries. This directional behavior is called thermal anisotropy. It can come from the printed architecture itself and, in filled filaments, from the orientation of fibers, flakes, or nanotubes during extrusion.
The practical distinction is between in-plane conduction, where heat travels mainly along the deposited layers, and through-plane conduction, where heat must cross layer interfaces. The preferred direction is not guaranteed to be the same for every material because filler orientation can reverse or amplify the behavior.
Research on a 20 wt.% multiwalled-carbon-nanotube/PLA composite illustrates how strong orientation effects can become in a deliberately conductive formulation. A vertically aligned printed structure reached about 0.575 W/m·K through-plane at 35 °C, compared with about 0.218 W/m·K for a horizontally aligned structure made from the same composite.[h] Those values are not representative of ordinary PLA; they show why the heat-flow direction must be stated when a printed composite is anisotropic.
Temperature Can Shift the Measured Value
Thermal conductivity is normally reported at a measurement temperature or over a temperature range. In the 2025 guarded-heat-flow tests, 100% infill PLA rose from 0.182 W/m·K near 22 °C to 0.202 W/m·K near 50 °C. PETG rose from 0.198 to 0.211 W/m·K over a similar range, and ABS rose from 0.172 to 0.184 W/m·K.[a]
The change was not perfectly linear. PETG measured 0.207 W/m·K near 60 °C and ABS measured 0.177 W/m·K near 60 °C, both below their values at about 50 °C. PLA was not taken to the 60 °C point because the specimens showed visible deformation near that temperature.[a] That last detail also demonstrates the separation between conductivity and heat resistance: a specimen can conduct heat in a measurable way while its shape stability is already becoming unsuitable.
Measured Values for PLA, ABS, TPU, PEEK, and PEI
A 2023 study used a DTC-25 guarded heat flow meter on FDM specimens printed at 100% fabrication density. Because the materials were tested within one experimental program, the values are useful for comparison inside that dataset.[f]
| Printed Material | Measured Thermal Conductivity | Interpretation |
|---|---|---|
| PLA 3080 | 0.22 ± 0.06 W/m·K | Low-conductivity unfilled thermoplastic |
| ABS | 0.22 ± 0.06 W/m·K | Similar to PLA within the reported uncertainty |
| TPU | 0.26 ± 0.05 W/m·K | Slightly higher in this test set |
| PEEK | 0.25 ± 0.05 W/m·K | High temperature capability does not produce metal-like conductivity |
| ULTEM 1010 (PEI) | 0.20 ± 0.06 W/m·K | Low thermal conductivity despite high-temperature service capability |
| ULTEM 9085 (PEI) | 0.20 ± 0.06 W/m·K | Close to ULTEM 1010 in this test |
The uncertainty ranges overlap for several materials. These measurements should not be converted into a universal ranking of all commercial PLA, ABS, TPU, PEEK, or PEI formulations.[f]
Carbon-Fiber and Metal-Filled Filaments Need Their Own Data
Fillers can create new heat-transfer paths through a polymer matrix, but the result depends on filler type, loading, particle or fiber shape, orientation, dispersion, and remaining porosity. A “CF” label therefore does not specify one conductivity value.
A 2026 Hot Disk study of MEX materials measured PETG at 0.23 ± 0.01 W/m·K and PETG-CF at 0.37 ± 0.01 W/m·K. In the same program, ABS measured 0.19 ± 0.01 W/m·K and an ABS glass-fiber formulation measured 0.22 ± 0.01 W/m·K. The different size of those changes shows why the base polymer and filler package must be evaluated together rather than assuming that any fiber-filled filament will behave the same way.[g]
Carbon fillers can also make conductivity more directional because fibers, nanotubes, or flakes may align during extrusion. For thermal design, a datasheet that reports one scalar value without stating direction can be incomplete when the formulation is strongly anisotropic.
Metal-filled filament is not thermally equivalent to solid metal. In the 2023 DTC-25 study, an as-printed aluminum-filled filament containing a high fraction of metal measured 0.40 ± 0.05 W/m·K.[f] That is higher than the unfilled polymers in the same experiment, but the polymer-rich printed structure still behaves very differently from a machined or cast metal heat sink.
Thermally Conductive Filaments Are a Separate Composite Class
Specialized thermal-management filaments can use graphite, graphene, carbon nanotubes, boron nitride, or other conductive fillers to move beyond the range of ordinary unfilled thermoplastics. The filler network is the main reason these formulations can behave differently. Some fillers also conduct electricity, while others such as boron-nitride systems can be designed to remain electrically insulating. “Thermally conductive” and “electrically conductive” therefore describe separate properties.
Research composites can reach values well above standard PLA or PETG, but a laboratory formulation should not be assigned to a retail filament that only shares the same base polymer. Product-specific TDS data, measurement direction, filler loading, and test method are needed before using a specialty filament in a thermal model.
From Material Conductivity to the Thermal Resistance of a Part
Thermal conductivity is a material property. A finished component also has dimensions, contact areas, interfaces, cavities, and exposed surfaces. For simple one-dimensional conduction through a uniform slab, the material value can be converted into thermal resistance with:
Thermal resistance: Rth = L / (k × A)
Conductive heat flow: Q = k × A × ΔT / L
L is the heat-flow thickness in meters, k is thermal conductivity in W/m·K, A is area in m², and ΔT is the temperature difference in kelvins or degrees Celsius.
For example, consider an ideal solid wall 5 mm thick with a heat-transfer area of 0.01 m² and an assumed conductivity of 0.20 W/m·K. Its one-dimensional conductive thermal resistance is 2.5 K/W. With a 20 °C temperature difference across the wall, the ideal conductive heat flow is 8 W.
This calculation is a screening estimate, not a complete model of an FFF enclosure. Sparse infill changes the effective k; contact resistance appears where parts touch; external surfaces exchange heat by convection and radiation; and complex geometry creates multidimensional heat paths. A thin solid plate, a thick low-infill block, and a hollow enclosure made from the same filament can therefore reach very different temperatures.
Designing for Insulation or Heat Spreading
The preferred material and slicer setup depend on whether the part should resist heat flow or spread heat more evenly. Thermal conductivity should be treated together with mechanical load, service temperature, dimensional accuracy, and the geometry of the heat path.
| Thermal Goal | Material Direction | Print-Structure Direction | Main Trade-Off |
|---|---|---|---|
| Reduce conductive heat transfer | Low-conductivity polymer or insulating composite | Lower effective density, longer heat path, fewer continuous polymer bridges | Lower infill can reduce stiffness and impact resistance |
| Spread heat through a housing | Thermally conductive composite with measured data | High density, continuous walls, orientation aligned with desired heat flow | Filled materials may be abrasive, brittle, or harder to print |
| Isolate a temperature sensor | Low-k polymer | Small contact area and narrow conductive path | Mount still needs enough mechanical stability |
| Transfer heat to a temperature sensor | Higher-k material or composite | Dense contact region and short conduction path | Interface quality can dominate a small sensor mount |
| High-temperature structural part | Choose heat resistance first, then check conductivity | Use the density and orientation required by the load | PEEK, PEI, PC, and similar materials have demanding print requirements |
For insulation, lowering nominal infill is only one option. The geometry should also avoid direct polymer bridges between hot and cold faces. For heat spreading, the opposite approach is usually more useful: dense shells, controlled extrusion, minimal voiding, and a filament whose conductivity has been measured in the direction that matters.
What a Useful Thermal-Conductivity Datasheet Should State
A conductivity number without specimen context has limited engineering value. Before comparing two filament products, check whether the accompanying data identifies the conditions that produced the number.
- Value and unit: normally W/m·K.
- Measurement temperature: conductivity can change with temperature.
- Test method or standard: for example ASTM E1530 or ISO 22007-2.
- Specimen state: raw resin, molded plaque, filament, or printed part.
- Printed density or infill: nominal 100% and actual density are not always the same.
- Heat-flow direction: especially for layered or fiber-filled materials.
- Conditioning and post-treatment: drying, annealing, sintering, or heat treatment can alter the structure.
- Formulation: pigments, plasticizers, fibers, minerals, and other additives can move the result away from a generic polymer value.
These details also explain why published values for the “same” polymer may disagree. For example, the studies cited here report PLA values that differ because they use different formulations, print structures, temperatures, and instruments. Averaging those numbers into one supposedly precise PLA constant would remove the information needed to interpret them.
Frequently Asked Questions
Which common filament has the highest thermal conductivity?
There is no universal winner across all brands and print conditions. In the 2025 same-method comparison at 100% infill near room temperature, PETG measured 0.198 W/m·K, PLA 0.182 W/m·K, and ABS 0.172 W/m·K, so PETG was highest within that test set.[a] The differences are small enough that formulation and printed structure can change the ordering in another dataset.
Does 100% infill give the highest thermal conductivity?
It generally creates a more continuous polymer path than sparse infill. In the 2025 PLA, PETG, and ABS tests, the 100% specimens produced the highest conductivity among the tested 40%, 60%, 80%, and 100% densities.[a] A slicer setting of 100% does not guarantee zero porosity, so flow quality and bead fusion still matter.
Is PEEK much more thermally conductive than PLA?
Not necessarily. One 2023 guarded-heat-flow study measured PEEK at 0.25 ± 0.05 W/m·K and PLA 3080 at 0.22 ± 0.06 W/m·K.[f] PEEK’s major thermal advantage is its ability to remain useful at much higher service temperatures, not metal-like heat conduction.
Does carbon fiber always increase filament thermal conductivity?
Carbon fiber can raise conductivity, but the amount depends on fiber loading, orientation, polymer matrix, dispersion, and porosity. A 2026 test measured PETG-CF above unfilled PETG, but that result belongs to those formulations and cannot be assigned to every PETG-CF spool.[g]
Can PLA or PETG be used as a replacement for a metal heat sink?
Ordinary PLA and PETG are poor substitutes where the design depends on metal-class heat spreading. The printed values cited here are only a few tenths of a watt per meter-kelvin. Thermally conductive composites can improve heat transfer, but they need product-specific measurements and may still require a different geometry than a conventional metal heat sink.
Is the lowest-conductivity filament always the best thermal insulator?
No. Lower material conductivity helps, but the complete part also depends on wall thickness, infill geometry, air cavities, contact area, convection, radiation, and mechanical requirements. A low-k material printed as a dense thin bridge can transfer more heat than a slightly higher-k material used in a thicker, low-density structure.
Sources
- Materials — Experimental Investigation of Thermal Conductivity of Selected 3D-Printed Materials — Supports the PLA, PETG, and ABS conductivity values, temperature dependence, infill-density results, infill-pattern comparison, and ASTM E1530 test conditions. (Peer-reviewed research article.)
- ASTM International — ASTM E1530-25, Guarded Heat Flow Meter Technique — Defines the active guarded heat flow method, specimen scope, and thermal-resistance measurement basis. (Official standards organization.)
- International Organization for Standardization — ISO 22007-2:2022 — Defines the transient plane heat source method for thermal conductivity and diffusivity of plastics, including anisotropic testing conditions. (Official international standard.)
- Progress in Additive Manufacturing — Effect of Filament-Extrusion Rate on 3D-Printed PLA — Supports the measured relationship between extrusion multiplier, porosity, and thermal conductivity in printed PLA. (Peer-reviewed research article.)
- Polymers — Influence of 3D Microstructure Pattern and Infill Density on PET-G — Supports the effect of infill topology on PETG thermal performance and the 25% concentric specimen result. (Peer-reviewed research article.)
- Materials — Experimental Measurement of Thermal Conductivity of FDM Materials with a DTC-25 Meter — Supports the same-program measurements for PLA, ABS, TPU, PEEK, PEI, and aluminum-filled filament. (Peer-reviewed research article.)
- Materials — Validation of Filament Materials for Injection Moulding 3D-Printed Inserts — Supports 2026 Hot Disk measurements comparing unfilled and fiber-filled MEX materials. (Peer-reviewed research article.)
- ACS Applied Nano Materials — Thermally Conductive 3D-Printed Carbon-Nanotube-Filled Polymer Nanocomposites — Supports the example of direction-dependent conductivity in an aligned CNT/PLA composite. (Peer-reviewed research article.)