The configuration chosen in Module 02 defines the external shape of the vehicle. This module defines how that shape is kept intact under the loads it will experience in flight, on landing, during handling, and across the full range of operating temperatures and conditions. Structural design is not about making things strong — it is about making them strong enough, in the right places, for the right loads, without carrying more weight than necessary. In an unmanned system where every gram of structure is a gram not available for payload or fuel, that distinction is not academic.
Structural requirements and load cases
Before any material is selected or any cross-section is sized, the load cases must be defined. A load case is a specific combination of forces and moments that the structure must withstand, drawn from the flight envelope, the ground handling requirements, and the worst-case operational scenarios defined in the CONOPS.
The primary load cases for a fixed-wing UAV airframe are: maximum positive and negative g loading in symmetric pull-up and push-over maneuvers; asymmetric rolling pull-out (the hardest combined bending and torsion case for the wing); landing impact, including hard landing with a gust-induced attitude; catapult or hand-launch acceleration loads at the attachment points; and ground handling loads including the vehicle being carried incorrectly and surviving without damage.
Load paths: how forces travel through a structure
The most important concept in structural design is the load path — the route by which an applied force travels through the structure to the ground or reaction point. A well-designed structure has clear, short, direct load paths. Forces travel through the stiffest available route, so placing material along the natural load path maximizes its contribution to structural efficiency. Material placed off the load path adds weight without adding strength.
In a fixed-wing UAV, the primary load path runs from the wing spar (which carries the bending moment generated by the distributed aerodynamic lift) through the spar carry-through in the fuselage, to the opposite wing spar. The fuselage is in bending between the wing attachment and the tail, reacting the pitching moment from the horizontal stabilizer. The motor mount carries the thrust and torque reaction from the propulsion system, which must be transferred into the fuselage without creating local stress concentrations that initiate fatigue cracks.
Understanding the load path before selecting materials ensures that high-strength, high-stiffness material is placed where it does the most work — along the primary load-carrying members — while lighter, lower-cost material fills non-structural volumes.
Materials: the options and their trade-offs
Unmanned systems draw from a wider materials palette than conventional aviation because the production volumes, certification requirements, and operating life expectations are different. The most commonly used structural materials in the systems described in this series are:
Aluminum alloys (6061-T6, 7075-T6): The workhorse of small UAS structural components — motor mounts, frames, centering rings, brackets, and custom fittings. 6061-T6 offers good strength (yield 276 MPa), excellent machinability, and corrosion resistance. 7075-T6 offers higher strength (yield 503 MPa) at a modest weight penalty, preferred for highly stressed fittings and components where fatigue life is critical. Both machine cleanly, are readily available, and can be bonded or fastened without special processes. The fuel tanks for the Jet-Wing were machined from 7075 — a material that tolerates the thermal cycling of jet fuel while providing the required burst pressure margin.
Fiberglass (E-glass / S-glass): The standard hull material for USV and larger UAV fuselages. E-glass fabric in epoxy resin provides good stiffness and strength at low cost, with straightforward hand layup processing. S-glass offers approximately 40% higher tensile strength than E-glass at modest additional cost, used where weight-to-strength ratio is more critical. Fiberglass composites absorb moisture over time, which degrades the matrix-dominated properties — particularly interlaminar shear strength — in humid operating environments. The SeaDoo USV hulls used fiberglass/Kevlar hybrid construction specifically because the Kevlar provides impact resistance and crack arrest against the hydrodynamic loads of high-speed surface operation.
Carbon fiber reinforced polymer (CFRP): The highest specific stiffness and specific strength of any material in common UAS use. Unidirectional carbon fiber in epoxy resin has a specific modulus approximately five times that of aluminum at roughly 40% of the weight. The trade-offs are brittleness — CFRP fails suddenly at ultimate load with minimal plastic deformation, making inspection of impact damage critical — and higher cost of both material and processing. CFRP is used for spar caps, spars, and primary structure on any program where weight is the dominant design driver. The Group-3 UAV wings used carbon fiber spar tubes and carbon fiber composite wing skins to achieve the 5-meter wingspan at a weight that permitted catapult launch.
Kevlar (aramid fiber): Exceptionally high tensile strength and impact resistance, but low compressive strength and poor compatibility with epoxy matrices compared to carbon fiber or glass. Kevlar is almost never used as a primary structural material — it is used for impact resistance layers in hybrid laminates (where it arrests crack propagation from an impact), for pressure vessel overwraps, and for applications where the load is purely tensile and impact resistance is more important than stiffness. The AUV/UUV pressure hull used a Kevlar overwrap over a machined structural liner for exactly this reason: the Kevlar arrests any crack that initiates in the liner before it can propagate to leakage.
Composite laminate theory
A composite laminate is a stack of fiber-reinforced plies, each with its fibers oriented in a specified direction, bonded together by a matrix material (typically epoxy resin). The stiffness and strength of the laminate in any given direction depends on the orientation, thickness, and stacking sequence of those plies. This is what makes composites uniquely powerful — the designer specifies not just the material but its directionality, placing stiffness and strength where the load path requires it.
The key orientation conventions: 0° fibers run parallel to the primary load direction and carry axial loads efficiently. 90° fibers run perpendicular to the primary load and resist transverse loads and Poisson contraction. ±45° fibers resist shear loads and contribute to torsional stiffness. A quasi-isotropic laminate (equal proportions of 0°, 90°, +45°, and -45°) behaves approximately like an isotropic material in-plane, useful when the load direction is uncertain or varies. A spar cap designed to carry pure bending will be predominantly 0° fiber because the load is axial. A tube designed to carry torsion will be predominantly ±45° fiber because the load is shear.
Sandwich construction
Many UAS structural panels — fuselage skins, wing skins, control surfaces — use sandwich construction: thin, stiff face sheets bonded to a lightweight core material. The sandwich dramatically increases the panel's bending stiffness with minimal weight increase, because bending stiffness scales with the cube of the panel thickness. Doubling the core thickness approximately doubles the weight of the core (which was already small) while increasing bending stiffness by a factor of four.
Common core materials are Rohacell foam (polymethacrylimide), Nomex honeycomb, and balsa wood. Rohacell is preferred for precision structural applications — it maintains consistent density and compressive strength, resists moisture, and bonds reliably to CFRP face sheets. Nomex honeycomb offers the highest strength-to-weight of any core material but requires specialized bagging and cure processes and is difficult to use for complex curved geometries. Balsa is used for simple, low-cost applications where weight is less critical than cost.
Finite element analysis in practice
FEA predicts how a structure deforms and where stresses concentrate under applied loads, before any physical part exists. For UAS structural design, the workflow is: build a simplified geometry model in CAD, apply material properties and laminate definitions, apply the design loads at the appropriate attachment points, run the analysis, and examine the stress and displacement results.
The key outputs to examine are: maximum principal stress (to identify potential failure locations in brittle materials like CFRP), von Mises stress (for ductile materials like aluminum), and deformation to verify that the structural deflection under load does not compromise aerodynamic performance or create clearance issues. Safety factors are applied by ensuring that the maximum predicted stress does not exceed the allowable stress — typically the material ultimate strength divided by the factor of safety.
Joint design: bonded and mechanical
Joints are the most common failure initiation point in composite structures, because load transfer between dissimilar materials or between structural members creates stress concentrations that the bulk material never experiences. The two primary joint types are bonded joints and mechanical fasteners.
Bonded joints transfer load through the adhesive layer in shear. They are efficient, add minimal weight, and distribute load over a large area — but they are permanent, difficult to inspect for internal bond quality, and susceptible to peel forces at the joint edges. Bonded joints should never be designed to carry peel loads; the adherend geometry should be designed to keep the adhesive in shear throughout the load range.
Mechanical fasteners (bolts, rivets) are inspectable, removable, and reliable for transferring large concentrated loads. In composite structures, they require careful design to avoid bearing failure (the composite crushes around the fastener hole) and bypass failure (the laminate fails in tension across the reduced cross-section at the hole). Minimum edge distances and spacing ratios, consistent with the relevant design standard, must be maintained.
The motor mount is the most demanding joint in most UAS designs — it must carry the full thrust force, the torque reaction from the motor, and the gyroscopic moments from the spinning propeller, all at a point where the geometry is constrained by the airframe envelope. On the Group-2 3D-printed UAV, a custom aluminum fuselage frame was bonded into the FDM Carbon-Nylon structure specifically because the motor mount loads exceeded what could be reliably transferred through the printed matrix.
Mold-making: plug, mold, and part
For composite structures with curved surfaces — fuselages, wing skins, hull bodies — the manufacturing process begins with a plug (also called a master): a physical model of the desired part surface, machined or hand-shaped to the precise geometry required. A mold is produced from the plug by laying composite material or pouring casting resin over its surface. Parts are then produced from the mold.
The plug-to-mold-to-part workflow is the same whether the structure is a UAV fuselage or the custom top deck and conformal fuel tanks on the SeaDoo USV conversions. In both cases, the geometry of the finished part is embedded in the mold, so every part produced from that mold is dimensionally identical to the plug. This is the key manufacturing advantage of composite tooling for small-volume production: once the mold is made, the geometric accuracy of subsequent parts is essentially free.
Mold release agent is applied to the mold surface before each part layup to prevent the part from bonding to the mold. For the USV fuel tank molds, a wax-based release system was used — the gelcoat layer of the finished part provides the watertight outer surface, and the underlying fiberglass/Kevlar laminate provides the structural load-carrying capacity.
Additive manufacturing for structural components
3D printing has fundamentally changed what is possible in rapid-prototype UAS development, but its structural limitations are frequently underestimated. The two processes relevant to structural UAS components are FDM (Fused Deposition Modeling) and SLS (Selective Laser Sintering).
FDM deposits material layer by layer, creating parts with anisotropic properties — stronger in the plane of deposition than through the layers. Layer delamination under tensile loading perpendicular to the print plane is the dominant failure mode. For structural FDM parts, print orientation must be chosen so that the primary load is in-plane, and high-performance materials — Carbon-Nylon, PA12-CF — should be specified rather than standard PLA or ABS. The Group-2 all-additive UAV used FDM Carbon-Nylon throughout its fuselage and wing structure: the layer orientation was optimized in each component for the primary load direction, and the design included an aluminum reinforcement frame at the motor mount specifically because the through-layer tensile loads at that location exceeded what the FDM material could reliably carry.
SLS fuses powder material with a laser, producing parts with nearly isotropic properties and no layer delamination weakness. SLS PA-12 Nylon has consistent mechanical properties in all directions and excellent surface finish without post-processing. It is the preferred process for structural UAS components that cannot be oriented favorably for FDM: the nozzle bodies, pressure components, and complex geometry parts of the APCP motor programs used SLS PA-12 for exactly this reason — the geometry could not be simplified to align the print plane with the primary loads.
3D printing vs. CNC machining: benefits and trade-offs
The question of whether to 3D print or CNC machine a structural component is not a matter of one technology being better than the other — it is a function of geometry, load, volume, tolerance, and schedule. Understanding the trade-space allows the designer to make the right choice for each component rather than defaulting to a single process for the entire system.
Where additive manufacturing wins: The most compelling advantage of 3D printing is geometric freedom. CNC machining is constrained by toolpath access — internal voids, undercuts, and enclosed channels are difficult or impossible to machine without multipart assemblies and subsequent joining operations. A 3D printer is indifferent to internal geometry. This enables topology-optimized structures that carry loads only where needed, internal cooling channels in thermally loaded components, and the consolidation of multi-part machined assemblies into a single printed part. The reduction in part count directly reduces assembly time, eliminates joint interfaces as potential failure points, and removes the tolerance stack-up that accumulates across mating machined surfaces.
For low-volume production — the typical scenario in UAS development — additive manufacturing eliminates tooling cost entirely. A design change that requires machining a new fixture or modifying a jig has a hard cost in both money and time. The same change in a printed part costs nothing beyond the time to modify the CAD file and run the printer. This is the single most significant advantage of additive manufacturing for rapid development programs, and it explains why the Group-2 all-additive UAV could be designed, manufactured, and test-flown in 65 days — design iterations that would have required new tooling in a conventional manufacturing workflow were trivially implemented overnight.
Where CNC machining still wins: Dimensional accuracy is the primary domain where machining is superior and likely to remain so. CNC machining holds tolerances of ±0.01–0.05 mm routinely. FDM holds ±0.3–0.5 mm. SLS holds ±0.1–0.2 mm. For bearing bores, sealing surfaces, threaded features requiring high pull-out strength, and mating interfaces where assembly fit is critical, CNC machining is the correct process. Attempting to print these features introduces variability that creates assembly problems and potential reliability issues in service.
High-cycle fatigue applications favor machined aluminum or steel over printed polymer — not because printed materials inherently have poor fatigue properties, but because the fatigue behavior of printed parts is less predictable and less well-characterized than that of wrought aluminum alloys with decades of documented behavior. For components that will experience tens of thousands of load cycles in service — flight control linkages, landing gear attachment points, hinge brackets — machined metal is the conservative and defensible choice.
Production volume also matters. For quantities above approximately 50–100 parts, the per-part cost of 3D printing exceeds that of CNC machining or injection molding due to the slow throughput of additive processes. A printed part that costs 40 minutes of printer time to produce can typically be machined in 3–5 minutes once the fixture is set. The crossover point depends on the geometry, but for any program that expects to produce more than a handful of each component, the economics of printing deteriorate rapidly with volume.
The hybrid approach — printed structural geometry with machined metal inserts at load concentration points — is often the most practical solution for UAS airframes. It captures the geometric freedom and tooling cost advantages of additive manufacturing for the bulk of the structure, while using machined metal at the interfaces where dimensional accuracy and material predictability are non-negotiable. This is not a compromise; it is the correct engineering use of each technology within its domain of advantage.