Insect-inspired, tailless, hover-capable flapping-wing robots: Recent progress, challenges, and future directions

At rest, beetles fold and tuck their hindwings under the elytra. For flight, the hindwings are deployed through a series of unfolding configurations that are passively driven by flapping forces. The folds lock into place as the wing fully unfolds and thereafter operates as a flat membrane to generate the aerodynamic forces. We show that in the rhinoceros beetle (Allomyrina dichotoma), these origami-like folds serve a crucial shock-absorbing function during in-flight wing collisions. When the wing collides with an object, it collapses along the folds and springs back in place within a single stroke. Collisions are thus dampened, helping the beetle to promptly recover the flight. We implemented this mechanism on a beetle-inspired wing on a flapping-wing robot, thereby enabling it to fly safely after collisions.

https://science.sciencemag.org/content/sci/370/6521/1214.full.pdf

Mechanisms of collision recovery in flying beetles and flapping-wing robots.

This paper reports a fundamental investigation on the aerodynamics of two-dimensional flapping wings in tandem configuration in forward flight. Of particular interest are the effects of phase angle (φ) and center-to-center distance (L) between the front wing and the rear wing on the aerodynamic force generation at a Reynolds number of 5000. Both experimental and numerical methods were employed. A force sensor was used to measure the time-history aerodynamic forces experienced by the two wings and digital particle image velocimetry was utilized to obtain the corresponding flow structures. Both the front wing and the rear wing executed the same simple harmonic motions with φ ranging from −180° to 180° and four values of L, i.e., 1.5c, 2c, 3c, and 4c (c is the wing chord length). Results show that at fixed L = 2c, tandem wings perform better than the sum of two single wings that flap independently in terms of thrust for phase angle approximately from −90° to 90°. The maximum thrust on the rear wing occurs du...

Aerodynamics of two-dimensional flapping wings in tandem configuration

Development of a novel non-contact piezoelectric wind energy harvester excited by vortex-induced vibration

Energy harvesting for jet engine monitoring

Flying insects demonstrate extraordinary flight performance and have inspired the design of flapping wing micro air vehicles (FWMAVs). However, FWMAVs are not confined to undergoing the same wing kinematics as those observed on natural flyers. Rather than undergoing the ubiquitous normal hovering motion typically observed on flying insects, FWMAVs may instead opt to undergo the water treading motion which originates from aquatic propulsion. In this study, the aerodynamic performance of normal hovering and water treading motions are compared for 2D and 3D rigid flapping wings in hover. Numerical simulations are conducted at varying mid-stroke angles of attack (αM). The results show that for both 2D and 3D, water treading can achieve higher maximum mean lift coefficient compared to normal hovering. Additionally, water treading is more efficient than normal hovering at any target mean lift coefficient within the parameter range considered. Visualisation of the flow structures indicate that the performance augmentation of water treading motion can be attributed to three mechanisms. Firstly, compared to normal hovering, water treading motion delays the shedding of the leading-edge vortex. Secondly, water treading tends to yield more beneficial wing-wake interaction. Thirdly, normal hovering enters a high angle of attack (α), high drag phase near stroke reversal, which incurs high aerodynamic power. This high α phase is absent in water treading, resulting in higher efficiency. For 3D cases, the leading-edge vortex is more stable and hence the first and the second mechanisms become less significant. At αM=45°, water treading outperforms normal hovering in terms of hovering efficiency by up to 54% in 2D and 29% in 3D. Hence, the water treading motion is a promising alternative for FWMAV.

Aerodynamics of 2D and 3D flapping wings in water treading motion

Wind tunnel experiments have been conducted to investigate the effect of turbulence on the power output of a MEMS aeroelastic wind energy harvester of approximately 55mm x 55mm x 10mm in size The energy harvester consisted mainly of a cylinder (2mm diameter, 10mm length) attached to a MEMS platform equipped with piezoelectric The turbulence is artificially created via a turbulence generator comprising of a series of rectangular beams placed in the wind tunnel upstream of the energy harvester By varying the wind speed and width of the turbulence generator beams, it is shown that the time-averaged power output of the energy harvester increases significantly as wind speed and turbulence generator size increases In particular, replacing a 5mm-width turbulence generator with a 20mm-width turbulence generator can increase the power output by 1710% at wind speed of 6m/s Power output in the magnitude of tens of nanowatts is measured when the device is exposed to turbulent winds Results suggest that the turbulence capturing concept may be a promising means of harvesting wind energy for miniature Internet-of-Things devices, such as small wireless sensor nodes that are not connected to the power grid

Aeroelastic wind energy harvesting by piezoelectric MEMS device with turbulence capturing

An investigation into the efficiency of two-dimensional polygonal structures undergoing transverse vortex-inducedvibration (VIV) is presented. The study aims at enhancing the yield in high oscillation amplitude for effective energy harvesting, for elements with a mass ratio of ≈25, subjected to laminar cross-flow at Reynolds number of ≈150. The efficiency of an isotoxal-star element with octagonal geometry is numerically investigated and compared to canonical circular cylinders. Under static conditions, the isotoxal-star element yields higher magnitude of fluctuating transverse force compared to the circular cylinders. When undergoing VIV at zero structural damping, the two cylinders have almost identical maximum oscillation amplitude despite the difference in the vortex strength and transverse force. Further investigations using an isotoxal-star element with diamond geometry and a triangular element suggests that VIV amplitude at zero structural damping is predominately governed by the magnitude of transverse drag force of the cross-section. At a structural damping ratio of 0.10, VIV enhancement due to changes in cross-sectional geometry becomes more apparent. Results suggest that VIV amplitude under damped condition is dependent on both the magnitude of fluctuating transverse force and the magnitude of transverse drag force. The two forces play different roles with regards to VIV response: a higher magnitude of fluctuating transverse force from a stronger von Karman vortex trail fosters the initiation of VIV. Conversely, strong transverse drag originating from the strong vorticity generated during transverse motions reduces the VIV. As such, VIV enhancement is a consequence of streamwise velocity or crossflow, while conversely, its reduction is primarily attributed to stronger transverse velocities. Results therefore suggest that enhanced VIV response could be potentially obtained through tailoring of the cross-sectional area to yield increased fluctuating transverse force while minimising transverse drag, represented here by the triangular element. Yet the compromise with regards to maintaining an effective omnidirectional performance of the elements renders isotoxal-star structures with multiple vertices (hereby assumed octagonal) more suitable for practical energy harvesting applications.

Vortex-induced vibration of two-dimensional isotoxal-star elements for enhanced wind energy harvesting

This study proposes an alternative mass ratio definition for vortex-induced vibration (VIV) systems. The proposed definition, henceforth known as 'total mass ratio', differs from the conventionally used mass ratio definition (i.e. the ratio of the mass of the oscillating body to the mass of the displaced fluid) in that the added mass of the fluid is taken into account. Two-dimensional numerical simulations are conducted to compare the effect of mass ratio on the VIV response of two cylinders with different cross sections, namely, a fully-filled solid cylinder and a hollow cylinder with fluid-filled cavities; both are designed to generate identical flow characteristics. When plotted against the conventional mass ratio, VIV response amplitudes of the fully-filled and hollow cylinders are different despite the similarity in flow characteristics. Conversely, the trends of VIV response amplitude against total mass ratio for both cylinders overlap almost perfectly. Results therefore suggest that the total mass ratio definition yields a more complete representation of the mass property of the VIV system, especially when comparing cylinders of different cross-sectional geometries.

https://iopscience.iop.org/article/10.1088/2631-8695/abbac6/pdf

Mass ratio definition of vortex-induced vibration systems with added mass consideration

Flow behavior in the near-wall region, the variation of natural wavelength (a pair of counter rotating vortices) and its influence on the normalized torque in Taylor vortex flow (TVF), transition phase of Taylor vortex flow into wavy vortex and wavy vortex flow (WVF) have been investigated using Direct Numerical Simulation (DNS) in Taylor-Couette flow for radius ratio 0.5. Near wall region is defined by the viscous layer (VL) along the axial direction. VL is found to be thinner in the jet impingement and thicker towards the flow separation region. A distinct drop in the VL thickness with two peak values of VL across the drop is observed in the flow separation region, and the axial and radial spatial behavior of VL thickness in the drop region in TVF, transition of TVF into WVF and WVF provide a possibility of the origin of the non-axisymmetric disturbances and its propagation is in the flow separation region of outer wall (inflow region) which results in the formation of WVF. The study of flow structure reveals that in the transition of TVF into WVF, whole inflow region becomes unstable and this results in the appearance of periodic secondary flow in between two counter rotating vortices. This periodic secondary flow plays most dominating role in the appearance of strong azimuthal wave in the inflow region. Study finds that flow structure remains axisymmetric up to Re 425 beyond which WVF appears. Besides, the natural wavelength varies for TVF and WVF with the change in Reynolds number. The natural wavelength is found to be maximum in the transition phase of WVF with comparatively lower normalized torque. The VL thickness and normalized torque is strongly influenced by the natural wavelength, where decrease in natural wavelength results in the decrease in the thickness of VL and increase in normalized torque and vice versa.

Y. J. Lee

Papers

Aerodynamics of 2D and 3D flapping wings in water treading motion

Aeroelastic wind energy harvesting by piezoelectric MEMS device with turbulence capturing

Vortex-induced vibration of two-dimensional isotoxal-star elements for enhanced wind energy harvesting

Mass ratio definition of vortex-induced vibration systems with added mass consideration

On Taylor-Couette flow: Inferences of near-wall effect