Evidence

What has been measured, under what conditions

Every result Darwin publishes, organized by the question it answers. Each one states the exact material form tested, the conditions, where the result comes from, and what it does not yet establish.

Measured

A physical measurement of a material property under the stated conditions.

Demonstration

A functional demonstration under one set of conditions. Not a qualified performance envelope.

Engineering analysis

An engineering assessment or design expectation. Not a measured result.

Not yet evidenced

A plausible capability with no published evidence. Treated as a hypothesis until tested.

Mechanical & structural

Is it stiff and strong enough to carry load?

Material-level stiffness and strength — the starting point for any structural use, well before part- or structure-level testing.

In-plane shear modulus (G12)12.5–12.6 GPa

Shear modulus 12.5–12.6 GPa in stretch-oriented CNT film

TestedMeasured
Material form
Candidate CNT film, stretch-oriented (15% stretch), epoxy-coated for test. Two grades measured: HGR and XM10. A film-level specimen — not a finished part, laminate, or core.
Test conditions
Modified ISO 6721-2 torsional pendulum. G12 calculated on the initial sheet-thickness basis. Para- and meta-aramid papers measured alongside by the same method.
Why it matters
Shear stiffness governs how thin walls, skins, and sandwich cores carry load. Under the same method, stretch-oriented CNT film measured well above the aramid papers tested alongside it.
Provenance
Recent characterization of current candidate material, summarized from a technical study dated 21 September 2026. The underlying test presentation is not published here.
What it does not establish
Experimental characterization of candidate films — not design allowables, qualification, or certification data, and not the performance of any finished structure. Aluminum (~25 GPa) is a literature reference, not measured in this study. Density was not compared, so no specific-stiffness claim is made. Anisotropy and repeatability remain to be characterized.
Tensile strengthup to 3.5 GPa

CNT yarn tensile strength up to 3.5 GPa

DemonstratedMeasured
Material form
Continuous CNT yarn.
Test conditions
Continuous CNT yarn, material-level tensile measurement.
Why it matters
High tensile strength in a flexible, conductive yarn allows the same element to reinforce a structure and carry current or heat.
Provenance
Reported by Darwin. The test report, testing organization, and date are not published here.
What it does not establish
A material-level property for the yarn form, not the strength of a finished structural part or joint.
Study detail

Shear stiffness of stretch-oriented CNT film

Candidate CNT films were compared against conventional para- and meta-aramid papers using the same modified ISO 6721-2 torsional-pendulum method. Across the materials tested, stretch-oriented CNT film measured the highest shear modulus.

G12 shear modulus, initial sheet-thickness basisMeasured in this studyLiterature reference · not measured here
Stretch-oriented CNT film
12.5–12.6 GPa
Para-aramid paper
~7.5–7.6 GPa
Meta-aramid paper
~2.8–3.3 GPa
Aluminum
~25 GPa

Aluminum sits near ~25 GPa as a literature reference — not measured here. The strongest CNT result reached roughly half of that figure.

~4×

the tested meta-aramid paper, for the strongest stretch-oriented variants. The ratio is against meta-aramid, not para-aramid.

Orientation is an engineering variable
XM10, 15% stretch
~12.64 GPa
XM10, unstretched
~8.42 GPa

The same grade measured higher after stretching. Stretch orientation is a process variable that can be tuned — but it also introduces anisotropy that has not yet been characterized.

How it was measured
  1. 01 · SpecimenCandidate films epoxy-coated and wrapped around thin-wall polycarbonate tubes.
  2. 02 · FixtureMounted in a torsional pendulum with an aluminum flywheel as the rotating mass.
  3. 03 · MeasurementNatural frequency read with a magnet and pickup-coil arrangement.
  4. 04 · CalculationG12 derived from natural frequency and film or composited-layer thickness.
Two calculation bases

G12 was calculated on two thickness bases. They are reported separately and are not comparable with each other.

Initial sheet thickness
HGR, 15% stretch~12.50 GPa
XM10, 15% stretch~12.64 GPa
XM10, unstretched~8.42 GPa
Representative para-aramid~7.6–7.9 GPa
Meta-aramid paper~2.8–3.3 GPa
Measured composited thickness
HGR, 15% stretch~8.17 GPa
XM10, 15% stretch~12.86 GPa
XM10, unstretched~10.58 GPa
What this does not establish
  • Design allowables, qualification, or certification
  • Performance of a finished part, laminate, or sandwich core
  • Specific stiffness — density was not compared
  • Repeatability across lots, or longitudinal / transverse anisotropy
  • Environmental, temperature, or damage behavior
Validation still required
  • Repeatability and statistical characterization
  • Longitudinal and transverse anisotropy
  • Representative laminate and structure-level testing
  • Environmental conditioning and temperature performance
Electrical & EMI

How well does it shield and conduct?

Shielding effectiveness of thin conductive CNT layers, measured above 200 MHz.

Shielding effectiveness~40–50 dB

EMI shielding effectiveness ~40–50 dB

DemonstratedMeasured
Material form
Single conductive CNT layer at ~15 gsm areal density. The specific product form is not stated in the published summary.
Test conditions
Single conductive carbon layer at ~15 gsm areal density, measured above 200 MHz.
Why it matters
Shielding from a single very light layer is what makes it plausible to shield enclosures, radomes, and cable runs without the mass of metal foils or meshes.
Provenance
Reported by Darwin. The test report, testing organization, and date are not published here.
What it does not establish
Shielding effectiveness is a function of layup, areal conductivity, frequency band, and joint design. This value is representative of the tested configuration, not a universal specification, and does not establish shielding of a finished enclosure.
Shielding effectiveness~60 dB

Two-layer EMI shielding ~60 dB

DemonstratedMeasured
Material form
Two conductive CNT layers.
Test conditions
Two conductive carbon layers, measured above 200 MHz.
Why it matters
Shows shielding scales with layer count, so attenuation can be traded against mass and thickness for a given requirement.
Provenance
Reported by Darwin. The test report, testing organization, and date are not published here.
What it does not establish
Adding layers increases attenuation at the cost of mass and thickness. Configuration-specific; confirmed per build.
Heating & de-icing

Can it heat a surface and clear ice?

The material driven as a distributed resistive heater, on the bench and against icing conditions.

Areal heating flux~10 W/in²

Distributed resistive heating ~10 W/in²

DemonstratedMeasured
Material form
Conductive CNT macrostructure driven as a distributed resistive heater. The specific form is not stated in the published summary.
Test conditions
Carbon macrostructure driven as a distributed resistive heater. Power density is set by conductive-region geometry and drive voltage.
Why it matters
A thin conductive layer that heats evenly across its area is the basis for anti-icing and de-icing without discrete wire heaters.
Provenance
Reported by Darwin. The test report, testing organization, and date are not published here.
What it does not establish
Achievable flux depends on the surface, drive electronics, and thermal boundary conditions. Sized to the anti-icing / de-icing requirement per application.
Time to clear ice~20 min · zero-ice

De-icing to zero-ice in ~20 min under test conditions

DemonstratedDemonstration
Material form
CNT material used as an electrothermal heater layer.
Test conditions
De-icing demonstration at approximately −25 °C with air velocity of approximately 71 m/s. Zero-ice result achieved under these specific tested conditions.
Why it matters
Shows the heating function working against real icing conditions and airflow, not only on a bench.
Provenance
Reported by Darwin. The test report, testing organization, and date are not published here.
What it does not establish
A single demonstration under one set of environmental conditions. Not a qualified ice-protection performance claim and not generalizable to other icing envelopes without further test.
Thermal transport

Can it move heat where it is needed?

Thermal conductivity by form and direction. Both values are strongly directional.

Axial thermal conductivityup to 120 W/m·K

CNT yarn axial thermal conductivity up to 120 W/m·K

DemonstratedMeasured
Material form
Continuous CNT yarn, measured along the yarn axis.
Test conditions
Continuous CNT yarn, conductivity measured along the yarn axis.
Why it matters
A light, flexible path that moves heat along its length — useful for spreading or routing heat away from a source.
Provenance
Reported by Darwin. The test report, testing organization, and date are not published here.
What it does not establish
Highly anisotropic — the value is axial (along the yarn). Transverse conductivity is much lower.
In-plane thermal conductivity~20–40 W/m·K

Sheet thermal conductivity ~20–40 W/m·K

DemonstratedMeasured
Material form
CNT sheet, in-plane direction.
Test conditions
Carbon sheet form, in-plane thermal transport.
Why it matters
In-plane spreading helps even out hot spots across a surface, and supports uniform heating in heater layers.
Provenance
Reported by Darwin. The test report, testing organization, and date are not published here.
What it does not establish
In-plane value for the sheet form; through-thickness transport is lower and is handled by interface design.
Dimensional stability

Will it hold its shape across temperature?

Directional thermal expansion. Currently supported by engineering analysis, not measurement.

Coefficient of thermal expansionNear-zero / negative CTE

Near-zero, potentially negative thermal expansion

In developmentEngineering analysis
Material form
Directional carbon layups (pultruded and sheet forms).
Test conditions
Directional CTE tuned through carbon layup. Effective near-zero expansion is achievable in specific directions.
Why it matters
Precision structures — optical benches, antennas, instruments — lose accuracy as they expand and contract with temperature.
Provenance
Darwin engineering analysis. Not a published measurement.
What it does not establish
A directional, layup-dependent property. A true isotropic zero is not claimed; targeted to the directions that matter for a precision structure. Not yet demonstrated end-to-end for a given program.
Multifunctional

Can one layer do several jobs at once?

Each function above has been characterized on its own. Evidence that several functions hold together in one integrated layer has not been published yet.

Values describe the material under the stated conditions. They are not design allowables, qualification data, or a guarantee of performance in a finished part or system. Test reports are available to qualified partners where they can be shared.