1. What a structure must do (the short version)

• The frame must be light, stiff and hard to break: a drone’s airframe must cut the weight that sets its flight time, yet be stiff enough that vibration neither breaks it nor upsets control. A robot arm needs bigger motors the heavier it is.
• Among metals, stiffness per unit mass is nearly the same: the “elastic modulus divided by density” of aluminum, titanium and magnesium all come out at about 24 to 25. The difference comes from carbon-fiber composites (CFRP), but their stiffness changes strongly with direction Our calculation.
• Materials also affect heat and radio: papers have measured that carbon-fiber airframes shed motor heat poorly and block radio waves. Choosing only for “light and strong” causes trouble elsewhere.

Figure 2 divides a drone and a robot arm into parts. Each part needs different things.

Original SVG, detailed explanation and sources

Original SVG reference: Concept diagram (vector drawing). Our arrangement of example materials and requirements for each part of a drone and a robot arm. The materials are examples of options, not the make-up of any specific machine or product. The materials the UR5e data sheet lists are “aluminum, plastic, steel” [Ref. 15].

2. Why mass matters so much

Drones: weight acts on power to the power of 1.5

When a drone hovers, the propellers keep producing a thrust equal to the aircraft’s weight. From propeller theory (momentum theory), the power needed to hover is proportional to the 1.5th power of the aircraft’s mass. Bauersfeld and Scaramuzza give this relationship as an equation in their multicopter model (hover power proportional to (mg)^3/2) [Ref. 1] Sourced.

Let us check what the 1.5th power means with numbers. If total mass falls 5%, power falls to 0.95^1.5 = 0.926 times, about 7.4% less Our calculation (Fig. 3, left). With the same battery, flight time would in theory rise by about 8.0% Our calculation (1 / 0.926 = 1.080, assuming the battery mass and the aircraft’s efficiency do not change). For example, suppose a 1 kg drone has a 0.25 kg frame. Making the frame 20% lighter saves 0.05 kg, which is 5% of the total Our calculation (an illustrative assumption).

Robot arms: weight acts on torque

Robot arms also benefit from being light. With the arm held horizontal, the shoulder joint feels a turning force (torque) from the arm’s weight. If a 5 kg link has its center of mass 0.5 m from the shoulder, the torque is 5 x 9.81 x 0.5 = about 24.5 N·m Our calculation. If the link can be 4 kg, it is 19.6 N·m, about 20% less Our calculation (Fig. 3, right). If the shoulder drive can be smaller, the whole arm gets lighter still – a virtuous circle we can hope for (Inference).

3. Yardsticks for materials: specific stiffness and equal-mass beam stiffness

The yardstick that turns “light and stiff” into a number is specific stiffness: the elastic modulus E divided by the density rho (E/rho), a guide to stiffness per gram of material. Table 1 collects numbers from data sheets and papers.

The numbers in each row are typical values from manufacturers and papers Sourced. The E/rho column and the quasi-isotropic row are Our calculation (division, and the calculation in Chapter 4). The 90 deg strength is Hexcel’s IM7 value (64 MPa). The T6 yield and tensile values are for a 0.500 inch diameter specimen (Kaiser).

The table shows that aluminum (25.3), titanium (24.9) and magnesium (24.1) have almost the same E/rho Our calculation. So “titanium is light and strong, therefore stiff” does not hold; titanium’s strength is in its strength (yield) (we do not compare yield per density here). Pulled along the fibers, CFRP has an E/rho of about 104, a little over four times that of the metals Our calculation.

For the same mass, which bends least?

In a frame, materials are made into plates and bars. If you make a bar of the same weight, a lighter material can make it thicker. A thicker bar bends less (bending stiffness goes with the fourth power of thickness). Including this effect changes how materials compare. Take a beam with a square cross-section of side b and length L; the stiffness k for a force at its tip is k = 3EI/L³ with I = b⁴/12. Holding the mass m = rho b² L fixed gives k proportional to E/rho² Our calculation (derivation in “More detail” below).

More detail: why E/rho² appears

A cantilever (a bar fixed at one end) with a force F at its tip deflects by delta = FL³/(3EI). The stiffness is k = F/delta = 3EI/L³. For a square cross-section of side b, the second moment of area is I = b⁴/12, so k = Eb⁴/(4L³). Since the mass is m = rho b² L, b² = m/(rho L), and b⁴ = m²/(rho L)². Substituting gives k = E·m²/(4 rho² L⁵). With m and L fixed, k is proportional to E/rho².

This assumes the cross-section can be scaled up in a similar shape. Real thin-walled tubes are limited by buckling (thin walls wrinkle), by the minimum practical thickness and by joint strength. The expression is a guide for comparing materials, not a design conclusion (our commentary).

As Fig. 4 (right) shows, for equal mass, magnesium (13.9) is about 1.5 times aluminum (9.4) Our calculation, titanium (5.6) is about 0.6 times aluminum Our calculation, quasi-isotropic CFRP (25.4) is about 2.7 times Our calculation, and 0 degree unidirectional CFRP (66.5) is about 7 times Our calculation. Even with a smaller E, magnesium is lighter and so can be made thicker, which makes it bend less at equal mass. But the magnesium E used here is the apparent value from a NASA internship report and was not checked against another source (not confirmed).

4. Carbon-fiber composites: stiffness depends on direction

A carbon-fiber composite (CFRP) is fine fibers set in resin. The basics of making fibers and of molding methods are in the polymers explainer on CFRP and composite materials. Here we stick to what matters in robot and drone design: the dependence of stiffness on direction (anisotropy).

The Toray data sheet for T700S carbon fiber gives the fiber itself a tensile strength of 4,900 MPa, a tensile modulus of 230 GPa and a density of 1.80 g/cm³ [Ref. 7] Sourced. Set in epoxy at 60% fiber volume, the composite has a tensile strength of 2,860 MPa and a modulus of 134 GPa [Ref. 7] Sourced. Pulled across the fibers, however, the strength (90 degree tensile strength) is 81 MPa [Ref. 7] Sourced. Across the fibers the strength is about 1/35 Our calculation (81 / 2860).

As Fig. 5 shows, the modulus falls from 164 GPa along the fibers (0 degrees) to about 55.7 GPa, roughly one third Our calculation, with only a 15 degree turn. At 45 degrees it is 14.0 GPa and at 90 degrees 12 GPa, about 1/14 of the fiber-direction value Our calculation. Stacking layers at different angles (laminating) gives a “quasi-isotropic” laminate with similar stiffness in every in-plane direction, but its modulus is about 62.7 GPa, about 38% of the unidirectional 0 degree value (164 GPa) Our calculation. A statement like “carbon fiber is so many times stiffer than steel” means nothing until you check which direction and which lay-up the value is for.

5. Metals: aluminum, magnesium, titanium; fatigue and galvanic contact

Metal frame parts are made by machining, casting and sheet-metal work. Aluminum is commonly used for the motor mounts of drones and the joint housings of robots (commentary; that aluminum is among the UR5e’s materials is sourced [Ref. 15]). Magnesium bends less at equal mass (Fig. 4, right), and cast frames have been studied (next chapter). Titanium has high strength (yield 790 MPa, Table 1) and suits small parts that must be light and also strong (commentary).

Fatigue: “never yielding” is not enough

The airframes of drones and the joints of robots can suffer fatigue (failure under repeated loads even below the yield strength) from propeller vibration and repeated motion. The Kaiser data sheet gives the “fatigue endurance limit” of 6061-T6 (the stress it survives for 5 x 10⁸ cycles of reversed stress in an R.R. Moore test) as 97 MPa [Ref. 2] Sourced. That is about 35% of the 276 MPa yield strength Our calculation and about 31% of the 310 MPa tensile strength Our calculation. Designing as if “use up to the yield strength” is allowed would fall short under repeated loading (values vary with test method, surface condition and environment).

Galvanic contact: joining carbon fiber directly to aluminum

Bolting a CFRP plate to an aluminum part is a common design. A NASA report by Danford and Higgins (1983) measured that when 6061-T6 aluminum is coupled to graphite-epoxy (a carbon-fiber composite) in 3.5% salt water, corrosion proceeds like a battery [Ref. 14] Sourced. They also reported that coating the whole composite with paint and coating the aluminum with an epoxy system gave protection [Ref. 14] Sourced. The result is for salt water. For outdoor drones or robots in damp places, it suggests that contacts between carbon fiber and dissimilar metals need to be insulated (commentary).

6. Materials also steer heat and radio waves

Heat: a carbon-fiber frame sheds motor heat poorly

Drone motors turn part of the electricity into heat, which escapes through the frame. Milenin (AGH University, Krakow) built two 7-inch multicopters that differed only in the lower plate, CFRP or aluminum 6061-T6, and measured motor temperature over a 5-minute hover with an infrared camera [Ref. 8] Sourced. With no load both were similar at 37 to 44 °C, but with a 20 N payload the CFRP aircraft’s motors exceeded 90 °C while the aluminum aircraft’s stayed below 60 °C (about 30 °C cooler) [Ref. 8] Sourced. In the same author’s 2026 paper, a cast magnesium AZ31 frame gave 51.8 °C, aluminum 50.3 °C and CFRP 77.5 °C (2 kg payload) [Ref. 9] Sourced (Fig. 6).

The reason lies in how readily the materials conduct heat. The Kaiser data sheet gives the thermal conductivity of 6061-T6 as 167 W/(m·K) [Ref. 2], and the Toray data sheet gives T700S carbon fiber (the fiber itself) as 0.096 J/(cm·s·°C) = 9.6 W/(m·K) Our calculation [Ref. 7]. In the same units, that is about 1/17 of aluminum Our calculation. From these data we cannot tell the value for a resin-bonded composite or through the thickness of a plate (not confirmed). The paper concludes that the aluminum frame carried heat away and kept motor temperature low [Ref. 8].

Radio: carbon fiber hardly lets radio waves through

Carbon fiber conducts electricity, so it blocks radio waves. This becomes a problem when antennas for communication and positioning sit inside or near the airframe. Wahyudi et al. measured the loss of 5.3 GHz radio waves through two composites used for a surveillance UAV’s airframe (specimens of equal size and thickness) [Ref. 10] Sourced. Glass fiber (E-glass) gave -1.5 dB and carbon fiber (twill) -31.24 dB, a difference of about 940 times in power Our calculation (Fig. 7).

Merkli et al. also report that with a receive antenna on a carbon-fiber airframe, moving the antenna farther from the frame than 4 times (lambda/8), that is half a wavelength, reduced the difference from the reference by about 1 to 1.5 dB (700 to 3500 MHz; differences mostly under 3 dB; valid for that aircraft and setup only) [Ref. 11] Sourced. Keeping the antenna away from carbon fiber is the basic idea (commentary).

7. Plastics and 3D printing: build direction and heat resistance

Plastics (engineering plastics) are common for drone shells, robot covers and small parts. Injection molding suits making the same shape in large numbers cheaply [Ref. 12] (a statement in the review by Kantaros et al.). For small runs, prototypes and freedom of shape, 3D printing is used. The review by Kantaros et al. summarizes the use of carbon-fiber-filled PLA, PETG and nylon filaments for drone parts [Ref. 12] Sourced.

As a numerical example, take the data sheet of nylon 12 formed by laser sintering (SLS), EOS PA 2200. The tensile modulus is 1650 MPa (in X, Y and Z), the tensile strength is 48 MPa in X and Y and 42 MPa in Z, and the strain at break is 18% in X and Y and 4% in Z [Ref. 13] Sourced. Z is the build direction. Strength is only about 12% lower, but the elongation (toughness) falls to about 22% Our calculation (4/18). The deflection temperature (1.8 MPa) is 64 °C in X and 57 °C in Z [Ref. 13] Sourced (Fig. 9).

8. Limits, what we could not confirm, and a summary

What we could not confirm

• E and density of magnesium alloys: we used the apparent E from a NASA internship report and the density of pure magnesium. We could not confirm a primary manufacturer or standards source giving typical AZ31 values
• E/rho of steel: for steel used in gears and shafts we found no primary source giving the elastic modulus, so it is not in Table 1 (the Aubert & Duval steel sheet gives only density and strength)
• Thermal conductivity of CFRP plate (through thickness) and composites, and standard measured values of radio shielding: not confirmed. We have only fiber values and specimen values from specific papers
• Materials of specific models (DJI and others): we did not confirm the materials of any aircraft from an official specification. Only the materials listed in the UR5e data sheet were checked
• Newer data on galvanic contact between CFRP and metals: only the 1983 NASA report. Other environments and material pairs were not checked
• The latest edition of rules on drone structures: we checked the EASA special condition for Light-UAS (December 2020) but did not confirm whether it is the current issue
• ASTM standards: ASTM International’s site states that it prohibits AI use, so we did not open it. No ASTM standard number is given in the text of this article

Summary

• Drone hover power rises with the 1.5th power of mass; a 5% mass cut saves about 7.4% of power (ideal case) Our calculation
• The specific stiffness of metals is about the same (24 to 25). For beams of equal mass, light magnesium bends less Our calculation
• CFRP is outstandingly light and stiff only when aligned with the fibers. A 15 degree turn cuts the modulus to about one third Our calculation
• Papers have measured that CFRP frames shed motor heat poorly and block radio waves Sourced
• For 3D-printed nylon, elongation changes strongly with build direction, and the deflection temperature is not an upper service temperature Sourced

9. How to read the yardsticks and standards

The performance of airframe materials is measured with yardsticks for tension, bending, fatigue, heat and radio. Each depends on “what specimen, in which direction, at what temperature”. For design, aerospace practice is to use not a manufacturer’s typical value but a statistically derived design allowable (MMPDS for metals, CMH-17 for composites and so on) [Ref. 17, 18]. The table lists what we could confirm (checked 25 October 2026).

Numbers, titles and status of standards were checked on the official ISO pages (25 October 2026). The standard texts are paid, so we stay with the scope statements. The arrangement of the table is Our calculation. Because ASTM states that it prohibits AI use, we do not treat ASTM standard numbers in this article.

Cautions in reading (for materials engineers)

• Data-sheet values are not guaranteed design values: the Toray data sheet states that its values are lot averages for reference and not for specification [Ref. 7]. Aubert & Duval also say theirs are typical values, not guarantees [Ref. 3]
• A laboratory result is about one aircraft: Milenin’s temperature and vibration results come from one group’s particular designs. Do not generalize them to other aircraft or flying styles
• Standards are mostly voluntary: ISO standards generally become binding only when cited in law or contract (commentary). An EASA special condition applies within the operations and certification it covers
• Standards under revision or review can change: ISO 178 and ISO 9283, for example, are under revision or review as of the check date Not yet confirmed

10. Glossary

11. References (primary sources)

1. Bauersfeld L., Scaramuzza D. “Range, Endurance, and Optimal Speed Estimates for Multicopters”, IEEE Robotics and Automation Letters (2022; arXiv:2109.04741; hover-power equation (5) checked) – https://arxiv.org/abs/2109.04741
2. Kaiser Aluminum “Rod & Bar Alloy 6061 Technical Data” (data sheet KA-RBH-6061-8.10) – https://online.kaiseraluminum.com/depot/PublicProductInformation/Document/1025/Kaiser_Aluminum_6061_Rod_and_Bar.pdf
3. Aubert & Duval “Titanium Alloy TA6V ELI” (data sheet) – https://www.aubertduval.com/wp-content/uploads/2024/12/TA6V-ELI_GB.pdf
4. Aroh J. (NASA Kennedy Space Center) “Magnesium Alloys for Space Hardware Design” (internship final report, 2017; tensile-test tables checked) – https://ntrs.nasa.gov/citations/20170011677
5. NIST “X-Ray Mass Attenuation Coefficients: Table 1, elemental data (density)” (checked 25 October 2026) – https://physics.nist.gov/PhysRefData/XrayMassCoef/tab1.html
6. Hexcel “HexPly 8552 Epoxy matrix Product Data Sheet” (2023 edition; tables for UD carbon prepregs AS4 and IM7 checked) – https://hexcel.com/wp-content/uploads/2025/12/HexPly_8552_eu_DataSheet1.pdf
7. Toray Composite Materials America “TORAYCA T700S Standard Modulus Carbon Fiber Data Sheet” (Rev. 24 November 2025) – https://www.toraycma.com/wp-content/uploads/T700S-Data-Sheet.pdf
8. Milenin A. “Effects of FPV Drone Frame Materials on Thermal Conditions of Motors Under Extreme Payloads: Experimental and Numerical Analysis”, Processes 13, 3034 (2025), DOI 10.3390/pr13103034 (open access; checked at abstract level) – https://doi.org/10.3390/pr13103034
9. Milenin A. “Experimental Analysis of an AZ31 Magnesium Alloy Structural FPV Drone Frame: Comparison with Aluminum and Carbon Fiber”, Processes 14, 1361 (2026), DOI 10.3390/pr14091361 (open access; checked at abstract level) – https://doi.org/10.3390/pr14091361
10. Wahyudi et al. “Analysis of unmanned aerial vehicle airframe materials on circularly polarized antenna radiation characteristics”, Bulletin of Electrical Engineering and Informatics 14(4), 2649-2661 (2025), DOI 10.11591/eei.v14i4.8010 (open access; checked at abstract level) – https://doi.org/10.11591/eei.v14i4.8010
11. Merkli et al. “Characterizing UAV airframe effects on mobile-band RF and EMF measurements under controlled anechoic conditions”, International Journal of Microwave and Wireless Technologies 18(4), 632-646 (2026), DOI 10.1017/S175907872610350X (checked at abstract level) – https://doi.org/10.1017/S175907872610350X
12. Kantaros A., Drosos C., Papoutsidakis M. “Composite Filament Materials for 3D-Printed Drone Parts: Advancements in Mechanical Strength, Weight Optimization and Embedded Electronics”, Materials 18, 2465 (2025), DOI 10.3390/ma18112465 (review; open access; main text checked) – https://doi.org/10.3390/ma18112465
13. EOS “PA 2200 Material Data Sheet” (process Balance, 120 µm layers; as of 23 October 2026) – https://www.eos.info/polymer-solutions/polymer-materials/data-sheets/mds-pa-2200
14. Danford M. D., Higgins R. H. (NASA) “Galvanic coupling between D6AC steel, 6061-T6 aluminum, Inconel 718 and graphite-epoxy composite material: Corrosion occurrence and prevention” (1983; checked at abstract level) – https://ntrs.nasa.gov/citations/19840005195
15. Universal Robots “UR5e Specification” (data sheet, August 2023 edition) – https://www.universal-robots.com/media/1807465/ur5e_e-series_datasheets_web.pdf
16. EASA “Special Condition Light Unmanned Aircraft Systems – Medium Risk”, Doc. No. SC Light-UAS Medium Risk 01, Issue 01 (17 December 2020, Final) – https://www.easa.europa.eu/sites/default/files/dfu/special_condition_sc_light-uas_medium_risk_01.pdf
17. CMH-17 “Composite Materials Handbook-17” (description on the official site checked; 25 October 2026) – https://www.cmh17.org
18. MMPDS (Battelle) “Metallic Materials Properties Development and Standardization (MMPDS)” (description on the official site checked; the site announces the release of MMPDS-2026; 25 October 2026) – https://www.mmpds.org/
19. ISO “ISO 6892-1:2019 Metallic materials – Tensile testing – Part 1: Method of test at room temperature” (ISO standard page) – https://www.iso.org/standard/78322.html
20. ISO “ISO 527-1:2019 Plastics – Determination of tensile properties – Part 1: General principles” (ISO standard page) – https://www.iso.org/standard/75824.html
21. ISO “ISO 527-5:2021 Plastics – Determination of tensile properties – Part 5: Test conditions for unidirectional fibre-reinforced plastic composites” (ISO standard page) – https://www.iso.org/standard/80370.html
22. ISO “ISO 527-4:2021 Plastics – Determination of tensile properties – Part 4: Test conditions for isotropic and orthotropic fibre-reinforced plastic composites” (ISO standard page; shown as withdrawn, with ISO 527-4:2023 as the new version) – https://www.iso.org/standard/80369.html
23. ISO “ISO 14125:1998 Fibre-reinforced plastic composites – Determination of flexural properties” (ISO standard page) – https://www.iso.org/standard/23637.html
24. ISO “ISO 178:2019 Plastics – Determination of flexural properties” (ISO standard page) – https://www.iso.org/standard/70513.html
25. ISO “ISO 75-1:2020 Plastics – Determination of temperature of deflection under load – Part 1: General test method” (ISO standard page) – https://www.iso.org/standard/77576.html
26. ISO “ISO/ASTM 52900:2021 Additive manufacturing – General principles – Fundamentals and vocabulary” (ISO standard page) – https://www.iso.org/standard/74514.html
27. ISO “ISO 9283:1998 Manipulating industrial robots – Performance criteria and related test methods” (ISO standard page) – https://www.iso.org/standard/22244.html
28. ISO “ISO 10218-1:2025 Robotics – Safety requirements – Part 1: Industrial robots” (ISO standard page) – https://www.iso.org/standard/73933.html
29. Aubert & Duval “Steel 819B (36NiCrMo16)” (data sheet) – https://www.aubertduval.com/wp-content/uploads/2024/12/819B-ts_GB.pdf

12. Claim-to-source audit

Last updated: 25 October 2026 / Sources are limited to primary sources (manufacturer data sheets, peer-reviewed papers, NASA reports, official ISO and EASA pages). Numbers are typical data-sheet values or values from specific papers’ conditions, not guaranteed design values. Primary typical values for magnesium alloys, the elastic modulus of steel, thermal conductivity of composites, materials of specific models and the latest edition of rules could not be confirmed, so they are not stated. Figures 2 to 7 and 9 are concept diagrams or charts with our calculation made for explanation (Figures 6 and 9 use data-sheet and paper values); Figures 1 and 8 are frames for AI-generated images (no image placed yet).