Sharp turning maneuvers with avian-inspired wing and tail morphing

The study addresses how avian-inspired morphing strategies can improve sharp turning maneuvers in fixed-wing drones, which typically need larger space to turn compared to multirotors. A turn consists of a roll phase (to establish bank angle) and a bank phase (sustained turning). Maximizing turn performance requires minimizing time to roll to the desired bank angle and maximizing lift during the bank to reduce turn radius. Conventional fixed-wing drones rely on ailerons for roll and elevator deflection to increase angle of attack and lift during banking. Birds, however, can perform sharper turns by asymmetric wing folding or wing twisting to induce roll and by symmetrically extending/sweeping wings and elevating/elongating tails to increase lift during banking. The research question is to quantify the relative impact of asymmetric wing folding versus asymmetric wing pitching (twisting) on rolling performance, and to quantify how symmetric wing and tail morphing affect lift and turning radius during banking, using a morphing, bird-inspired drone.

Prior work shows birds achieve agile turning via wing folding (area reduction causing lift asymmetry) and wing twisting (incidence changes causing lift asymmetry) and by symmetrically modifying wing and tail to increase lift in banked turns. Avian-inspired asymmetric wing folding and wing twisting have been implemented on drones and shown to enable roll control and steady bank angles, with effectiveness depending on angle-of-attack regimes and wing regions actuated. Earlier feathered drones demonstrated morphing benefits but indicated that asymmetric folding alone may require increased angle of attack to be effective, suggesting potential benefits from adding wing twisting. Despite these developments, a systematic, quantitative comparison of folding versus pitching for roll, and of symmetric wing and tail morphing for banking turn radius, remained largely unexplored.

Platform: Developed a feathered morphing drone (LisEagle), wingspan 1.52 m, mass 711 g, with wings capable of both folding (sweep) and pitching (uniform incidence change) either symmetrically or asymmetrically, and a morphing tail capable of fanning (area change) and elevator-like deflection (−30° to +15°). Wing folding allows one side extended and the other tucked; pitching allows ±10° incidence difference per side (20° total differential). Tail and wing symmetric extension increase lifting area and positive pitch moment. Instrumentation: Servos (KST X08S/X09S) actuate morphing surfaces. Propulsion via T-Motor F30-62 with 3.4×3″ propeller. Avionics include Pixhawk 4 autopilot, RTK GPS (10 Hz), pressure sensor for airspeed, IMU for attitude rates. Wind tunnel tests used an open-jet tunnel at 10 m/s with a Nano35 force/torque balance; propeller removed for measurements. Roll phase experiments: Wind tunnel measurement of roll moment coefficients versus angle of attack (−8° to 20°, 2° steps; cubic interpolation) for two configurations: (i) asymmetric wing folding (one wing extended, one tucked; no pitching), and (ii) asymmetric wing pitching (left −10°, right +10° with both wings tucked). Outdoor roll flight tests at ~12 m/s (60% throttle), initiating maneuvers from trimmed cruise; four trials per configuration, logging roll rates and bank angles over 1 s after actuation. Bank phase experiments: Wind tunnel measurement of lift and pitch moment for three configurations with tail elevator deflected upward by 15°: (i) wings and tail tucked; (ii) wings tucked, tail extended; (iii) wings extended and tail extended. Derived equilibrium angle of attack from zero pitch moment crossing and extracted corresponding lift coefficient. Developed a steady coordinated turn model relating turn radius to lift coefficient and bank angle (simplified with zero sideslip and steady conditions). Outdoor banking turns executed from cruise (~12 m/s), rolling into mean bank angles around 53° before configuring to one of the three morph states; five trials per configuration; turn radii estimated from trajectories; compared to model predictions.

• Asymmetric pitching vs folding for roll: In wind tunnel tests at cruise angles of attack (≈2–8°), asymmetric wing pitching produced substantially larger roll moment coefficients (≈0.079) than asymmetric wing folding (≈0.016). Pitching roll moment decreased with increasing angle of attack, whereas folding roll moment increased with angle of attack.
• Roll flight tests: From cruise, asymmetric wing pitching yielded much faster roll response and higher bank angles than asymmetric folding; the drone reached about 60° bank in 0.4 s with pitching versus only about 10° in 0.4 s with folding.
• Banking lift increase via symmetric morphing: With tail elevator +15°, the measured lift coefficients at pitch equilibrium were approximately: 0.52 (wings and tail tucked), 0.83 (wings tucked, tail extended), and 1.68 (wings and tail extended). Thus, symmetric wing and tail extension increased lift by a factor of about 3.2 relative to the fully tucked case.
• Turn radius predictions and measurements: The steady-turn model predicted that increasing CL from 0.52 to 1.68 reduces minimum turn radius from about 12 m to about 4.8 m (at similar conditions). Flight tests showed mean turn radii consistent with the trend: extended wing and tail achieved about 4.9 m mean radius, while the fully tucked configuration was around 12.1 m (discussion cites a 56.5% reduction from 12.1 m to 4.9 m). The tucked wing with extended tail condition yielded intermediate radii (reported mean ≈5.9 m in tests, with model range 6.9–8.9 m).
• Load factor and g-forces: The design withstands load factors up to 4. During turns, body g-forces remained below this limit (peaks: ≈1.46 g tucked vs ≈3.3 g extended wing and tail).
• End-to-end maneuver distance: For a representative 180° turn at 12 m/s and ~50° bank, combining fast roll via pitching with symmetric wing-and-tail extension reduced total space needed to about 9.9 m, compared to about 35 m when relying on asymmetric folding and keeping wing/tail tucked.

Results demonstrate that, at typical cruise angles of attack, asymmetric wing pitching is a more effective roll control strategy than asymmetric wing folding for rapidly achieving bank angle, thereby reducing the distance and time needed to initiate a turn. The aerodynamic trends explain this: pitching-induced roll authority diminishes at higher angles of attack due to lift saturation and possible reversal, whereas folding-induced lift asymmetry grows with angle of attack because it stems from area difference, suggesting folding may outperform pitching at high angles of attack. Adverse yaw was observed for both strategies; wind tunnel data suggest adverse yaw increases with pitching at higher angles of attack but remains nearly constant with folding, implying potential benefits of folding in high-AoA regimes to mitigate adverse yaw. During the bank phase, symmetric wing and tail morphing substantially raises lift, enabling higher bank angles and smaller turn radii, validated by both the steady-turn model and flight tests, despite real-world deviations from ideal steady, coordinated turns (variations in airspeed, sideslip, and altitude). Collectively, combining fast-roll authority from pitching with lift-enhancing symmetric wing and tail morphing yields large improvements in turn agility for fixed-wing drones in confined environments.

The study presents a bird-inspired morphing fixed-wing drone with wings capable of both folding and pitching and a morphing tail. Systematic wind tunnel and flight experiments show that asymmetric wing pitching provides superior roll control at cruise conditions compared to asymmetric folding, enabling faster achievement of large bank angles. Symmetric extension of wing and tail, coupled with tail elevation, significantly increases lift and reduces turn radius by more than a factor of two (e.g., from ~12.1 m to ~4.9 m), markedly improving agility in confined spaces. Future work should investigate roll control at higher angles of attack, where folding may outperform pitching, refine control to manage adverse yaw and maintain coordinated high-AoA turns, explore greater folding asymmetries and reductions in roll inertia, and implement precise feedback control for high-AoA flight to fully exploit morphing capabilities.

• Asymmetric folding tests used a limited area difference (~31%), which may understate its potential; greater asymmetry could yield higher roll moments.
• The pitching mechanism rotates the entire wing uniformly, potentially increasing tip angle of attack and inducing stall, which may reduce roll effectiveness versus birds’ spanwise twist.
• Adverse yaw occurs with both folding and pitching; effects may increase with pitching at higher angles of attack.
• Outdoor flight tests were not perfectly steady or coordinated: airspeed, bank angle, sideslip, and altitude varied due to pilot inputs and constant-throttle setup, causing discrepancies with the steady-turn model.
• Load factors during turns could not be precisely estimated (lack of resolved angles of attack and sideslip); body g-forces were used as conservative proxies.
• Wind tunnel tests excluded propeller slipstream effects, which may affect in-flight aerodynamics.