Download as pdf or txt
Download as pdf or txt
You are on page 1of 13

Wing kinematics and aerodynamic forces in

miniature insect Encarsia formosa in forward


flight
Cite as: Phys. Fluids 33, 021905 (2021); https://doi.org/10.1063/5.0039911
Submitted: 09 December 2020 . Accepted: 10 January 2021 . Published Online: 25 February 2021

Xin Cheng (程欣 ), and Mao Sun (孙茂 )

ARTICLES YOU MAY BE INTERESTED IN

Effects of wing-to-body mass ratio on insect flapping flights


Physics of Fluids 33, 021902 (2021); https://doi.org/10.1063/5.0034806

Aerodynamic interaction of bristled wing pairs in fling


Physics of Fluids 33, 031901 (2021); https://doi.org/10.1063/5.0036018

Why coronavirus survives longer on impermeable than porous surfaces


Physics of Fluids 33, 021701 (2021); https://doi.org/10.1063/5.0037924

Phys. Fluids 33, 021905 (2021); https://doi.org/10.1063/5.0039911 33, 021905

© 2021 Author(s).
Physics of Fluids ARTICLE scitation.org/journal/phf

Wing kinematics and aerodynamic forces


in miniature insect Encarsia formosa
in forward flight
Cite as: Phys. Fluids 33, 021905 (2021); doi: 10.1063/5.0039911
Submitted: 9 December 2020 . Accepted: 10 January 2021 .
Published Online: 25 February 2021

Xin Cheng (程欣), and Mao Sun (孙茂)a)

AFFILIATIONS
Institute of Fluid Mechanics, Behang University, Beijing 100191, China

a)
Author to whom correspondence should be addressed: m.sun@buaa.edu.cn

ABSTRACT
Miniature insects fly at very low Reynolds numbers, and the effect of air viscosity is large. Previous studies in this area are on hover flight.
Here, we study the forward flight, by measuring the wing kinematics and analyzing the flows of a typical miniature insect (Encarsia formosa,
wing length of about 0.5 mm). In the beginning of the upstroke, the wings quickly accelerate backward at a very large angle of attack and
smash on the air (“impulsive rowing”), generating a large thrust; in the rest of the upstroke, the wings come together and move upward,
slicing through the air and generating a small negative vertical force and negative thrust. In the beginning of the downstroke, the wings fling
open and produce a leading-edge vortex (LEV) on each wing; in the rest of the downstroke, the wings move downward and forward with the
LEV staying attached, generating a large vertical force and some negative thrust. The large thrust produced by the “impulsive rowing” over-
comes the body drag and the negative thrust produced by the wings in the other parts of the flapping cycle; the vertical forces, produced by
the “flinging” and by the downward/forward motion of the wings carrying the LEVs created at the fling, provide the weight supporting force.
That is, the tiny insect overcomes the strong viscous effect by fast smashing the wings on the air, by fast flinging open the wings, and by using
the LEVs created at the fling.
Published under license by AIP Publishing. https://doi.org/10.1063/5.0039911

I. INTRODUCTION aerodynamic forces is the delayed-stall mechanism.10–21 In the


Miniaturization is one of the principal directions of evolution delayed-stall mechanism, a separated leading-edge vortex (LEV)
in insects.1 As a result, many insects are very small. The average with large strength attaches to and moves with the wing in the
insect-wing length (R) is 5 mm. Near half of the existing winged- entire upstroke and downstroke, and the LEV produces a low pres-
insect species are R between 0.5 mm and 4 mm;2 they are sure or suction force on the upper surface of the wing, resulting in
referred to as miniature insects. The wing length of medium and the lift and drag. The mechanism can also be explained using the
large insects is 5 mm–50 mm, e.g., those of fruit-flies, honey bees, vortex dynamic theory: The LEV, together with the root vortex, the
and hawkmoths.2–4 starting vortex, and the tip vortex, forms a vortex loop. As the wing
Medium and large insects in normal hovering stroke their wings moves, the attached LEV facilitates a linear growth of the area of
back and forth in an approximately horizontal plane [Fig. 1(a)]; the the vortex loop with time, producing a large time rate of change of
back and forth strokes are referred to as the upstroke and down- the first moment of vorticity (or fluid impulse), hence a large aero-
stroke, respectively.3 It has been shown that in an upstroke and dynamic force.22–25 A strong LEV attached to the wing is essential
downstroke, the wings operate at high angle of attack (35 )5–9 to the delayed-stall mechanism.
and that a lift and drag are produced, with the drag being a little The Reynolds number (Re) of the wing in the flapping mode
smaller than the lift. The lift supports the insect weight, and the shown in Fig. 1(a) is defined as
drag in the upstroke and that in the downstroke cancel out each cU 2UfR2 ðr2 =RÞ
Re ¼ ¼ ; (1)
other. The major aerodynamic mechanism responsible for the v vðR=cÞ

Phys. Fluids 33, 021905 (2021); doi: 10.1063/5.0039911 33, 021905-1


Published under license by AIP Publishing
Physics of Fluids ARTICLE scitation.org/journal/phf

forces are produced by fast downward acceleration of the wing, at


nearly 90 angle of attack, similar to the case of a plate rapidly started
at its normal direction.26 These two approximately upward pointing
forces provide the weight-supporting vertical force. The horizontal
components of these two forces have opposite directions and cancel
out each other in a flapping period, and the wingbeat-cycle mean hori-
zontal force is zero, as required for hovering flight.
As far as we know, all the previous studies on the flapping
motion and aerodynamic forces of miniature insects are within the
FIG. 1. (a) Wing-tip trajectory (projected onto the symmetrical plane) and the flap- context of hovering flight.3,26,30–37 In forward flight of miniature
ping motion of a wing section of a relatively large insect. (b) Those of a tiny wasp,
insects, how the flapping-motion pattern and aerodynamic-force
Encarsia formosa.
generation would change is unknown. In the present work, we study
the forward flight of a typical miniature insect, the tiny wasp
where c is the mean chord length of the wing, U is the mean wing Encarsia formosa, the hovering flight of which has been studied pre-
speed at the radius of gyration of the wing (r2), and  is the kinematic viously by many researchers.3,26,34 First, the forward-flight wing
viscosity of the air (here U ¼ 2Ufr2, where U is the stroke amplitude kinematics is measured by using three high-speed cameras
and f is the wingbeat frequency). Re represents the effect of air viscos- equipped with micro-lenses and extension tubes. Then, the flows
around and aerodynamic forces on the insect are obtained by
ity: lower Re means the wing moves in a more viscous flow. As insect
numerically solving the Navier–Stokes equations using a well-
size (R) decreases, U, r2/R, and R/c do not change greatly and f
tested flow solver. Finally, the flight and force production mecha-
increases approximately according to R−0.67 (see Ref. 2). Then, Re is
nisms are analyzed, and how the miniature insect overcomes the
approximately proportional to R1.33. Thus, for small insects (R very
large viscous effect is revealed.
small), Re will be very low, or the viscous effect will be very large. For
the smallest insects such as the tiny wasp Encarsia formosa (wing
length is about 0.5 mm), Re is 10 (see Refs. 3 and 26). At such low II. MODELS AND METHODS
Re, if the flapping mode of the medium and large insects [Fig. 1(a)] A. Experimental method and insects
is used, little lift can be generated, while the drag is very large.27,28 The method of measuring the forward-flight wing kinematics is
This is because with such a large viscous effect, during the forma- similar to that used previously by the authors for hovering flight,26
tion of the LEV, the vorticity is highly diffused and the resulting except that the flight chamber in the present study is “longer” than
LEV is very weak.29 Therefore, very small insects must use a “new” that used in hover flight: in Ref. 26, the flight chamber is a cubic of the
flapping mode. size of 40  40  40 mm3, while in the present forward-flight study, it
Recent studies have shown that the flapping mode and force pro- is a cuboid with the size of 80  40  40 mm3 [Fig. 2(a)]. Three
duction mechanisms of very small insects are very different from those orthogonally aligned synchronized high-speed cameras
of their larger counterparts:26,30 the tiny wasp Encarsia formosa and (FASTCAM Mini UX100, Photron, Inc., San Diego, CA, USA;
the thrip Frankliniella occidentalis have a very deep U-shaped upstroke 10 000 fps, shutter speed 20 ms, resolution 896  488 pixels)
[Fig. 1(b)], and at the start of the downstroke, the wing performs the equipped with 60 mm micro-Nikkon lens and 12 mm extension
“fling” motion, which was previously discovered by Weis-Fogh.3 The tubes are used to film the flight of the insects [Fig. 2(a)].
deep U-shaped upstroke produces large transient drag that points Quantitative kinematics are extracted from the filmed data by an
almost upward, and the “fling” motion, at its later stage, also produces interactive graphic user interface developed using MATLAB
transient drag pointing almost upward. These two transient drag (MATLAB V. 7.1, Math Work, Inc., Natrik, MA, USA). The

FIG. 2. (a) A sketch of the experimental system. (b) Shapes of the model body. (c) Shape of the model wing. (d) Portions of computational grids.

Phys. Fluids 33, 021905 (2021); doi: 10.1063/5.0039911 33, 021905-2


Published under license by AIP Publishing
Physics of Fluids ARTICLE scitation.org/journal/phf

method has been described in detail in Ref. 26, in which kinematic revolving with an acceleration phase, a constant-speed phase at Re
data of hovering flight of the same species were measured. ¼ 500 and a deceleration phase;43 and flapping44 at Re ¼ 180), and
Insects used in the experiment, small wasp Encarsia formosa comparison has also been made with theoretical (Stokes and Oseen
Gahan (the same species as that used in the hovering flight study), solutions) and measured results for a sphere at very low Re (Re
were acquired from the Laboratory of Biological Invasion of Institute ¼ 5–20). The method has been described in detail in Refs. 26 and 39,
of Plant Protection, Chinese Academy of Agricultural Sciences; they and only an outline of the method is given here. The method is based
were descendents of wild-caught E. formosa. The wasps were housed on the idea of artificial compressibility, first developed by Rogers
in small chambers in groups of three to five individuals and fed with et al.45 for a single grid and then extended by Rogers and Pulliam46 to
20% sugar solution. The experiment was conducted at room tempera- overset grids. In the method for a single grid, the time derivatives of
ture 25  C–27  C and relative humidity 50%–60%. the momentum equations are differenced using a second-order, three-
More details of the experimental setup, the method used to point backward difference formula. The continuity equation requires
extract the 3D body and wing kinematics from the filmed data, and that the divergence of the velocity field is zero. In order to solve the
the method used to measure the morphological parameters (wing time discretized momentum equations for a divergence-free velocity at
length, wing shape, wing area, etc.) can be found in Ref. 26. a new time level, a pseudo-time level is introduced into the momen-
tum equations and a pseudo-time derivative of pressure divided by an
B. Flow governing equations, computational model, artificial compressibility constant is introduced into the continuity
and numerical procedure equation. The resulting system of equations are iterated in pseudo-
time until the pseudo-time derivative of pressure approaches zero;
The governing equations for the flows around the insect are the
thus, the divergence of the velocity at the new time level approaches
incompressible, three-dimensional Navier–Stokes equations. The
zero. The derivatives of the viscous fluxes in the momentum equation
dimensionless form of equations is as follows:
are approximated using second-order central differences. For the
r  u ¼ 0; (2) derivatives of convective fluxes, upwind differencing based on the
  flux-difference splitting technique is used. A third-order upwind
@u 1
þ u  ru ¼ rp þ r2 u; (3) differencing is used at the interior points, and a second-order upwind
@s Re differencing is used at points next to boundaries. For the case of overset
where u is the non-dimensional fluid velocity, p is the non- grids, the solution method for a single grid is applied to each of the
dimensional fluid pressure, s is the non-dimensional time, and Re is wing grid and the background grid, and data are interpolated from one
the Reynolds number; in the non-dimensionalization, c, U, and c/U grid to another at the inter-grid boundary points using tri-linear inter-
are taken as reference length, velocity, and time, respectively (defini- polation. Details of this algorithm can be found in Refs. 45 and 46.
tions of c and U have been given in Sec. I). For the boundary conditions, at the far-field inflow boundary,
The body of the insect is modeled as follows: The shape of the the velocity components are specified according to the constant inflow
body is obtained from the video images (side and top views of velocity while pressure is extrapolated from the interior; at the far-
the body). The cross-section of the body is taken as an ellipse, and the field outflow boundary, pressure is set equal to the inflow pressure
body model is given by fitting ellipses to the body shape in the video and the velocity is extrapolated from the interior. On the wing sur-
images. The side top views of the model body are shown in Fig. 2(b). faces, impermeable wall and no-slip boundary conditions are
The wings are modeled as flat plates with rounded leading and trailing applied, and the pressure on the boundary is obtained through the
edges, and the thickness of the plate is 0.03c. The planform of the normal component of the momentum equation. The insects move
model wing [Fig. 2(c)] is obtained from the measured wing shape. forward at some constant flight speed, and this value is given at the
There were relative movements between the wings and between inflow boundary.
the body and the wings, and therefore the equations were solved over The grid size used here is the same as that used for hovering flight
moving overset grids. In the overset grid system [Fig. 2(d)], there was in Ref. 26: the wing grid had dimensions 61  91  65 in the normal
a body-fitted curvilinear grid for the body and for each of the left and direction, around the wing, and in the spanwise direction, respectively
the right wings, and there was a background Cartesian grid, which (first layer grid thickness was 0.001c); the body grid had dimensions
extends to the far-field boundary of the domain. The background 99  77  53 along the body, in the azimuthal direction, and in the
Cartesian grid was generated algebraically. The wing grid is of O–H normal direction, respectively; the background grid had dimensions
type and that of the body is O–O type, and they were generated by a 121  121  121 in the x, y, and z directions, respectively. Detailed
Poisson solver. For the wing grids, dynamically deforming grids are grid resolution tests were conducted to ensure that the flow calcula-
used to treat the time-various bending deformation of the wing near tions are grid independent.26
the end of an upstroke or downstroke. A procedure of combining the
method of the modified trans-finite interpolation and the method of
III. RESULTS AND DISCUSSION
solving the Poisson equation was used to generate the dynamically
deforming grids; a more detailed description of the procedure can be A. Wing-flapping motion at forward flight
found in Du and Sun.38 Constant-speed forward flights of five wasps were captured in
The CFD solver used to solve the flow equations is an in-house our experiment. These insects are named, respectively, as W1, W2,
code developed more than a decade ago13 and has been tested in sev- W3, W4, and W5. The movies in Fig. 3 (Multimedia view) give the
eral works of our group,26,33,39,40 using experimental data of various video sequences for some of the wasps. As an example, the sequences
wing motions (translating at Re ¼ 100;41 rotating at Re ¼ 1000;42 of the flight of one of the wasps (W1) are shown in Fig. 3.

Phys. Fluids 33, 021905 (2021); doi: 10.1063/5.0039911 33, 021905-3


Published under license by AIP Publishing
Physics of Fluids ARTICLE scitation.org/journal/phf

The method for describing the wing motion as a function of time [Fig. 4(a)]. As seen in Fig. 4(a), b + v is the angle between the long-
is the same as that used by Cheng and Sun26 and Lyu et al.30 for minia- axis of the body and the stroke plane.
ture insects: For a filmed wasp, data of five flapping cycles were proc- The position (and motion) of a wing relative to the stroke plane
essed. In each of the five cycles, when the wings are at the end of the can be determined by three Euler angles and by the maximum bending
downstroke and at the end of the upstroke, the wing-tip points are displacement of the wing (see Ref. 26 for the detailed description of
projected onto the plane of symmetry of the insect. A linear regression wing bending). The Euler angles [Fig. 4(b)] are the positional angle of
line of the projections is then determined. The plane that passes the wing, / (in the stroke plane), the pitch angle w, and the deviation
wing base and at the same time is parallel to the above line is defined angle h; and the maximum bending displacement is denoted by dm.
as the stroke plane. The stroke plane is shown in Fig. 4. In the figure, Figure 5 gives the measured Euler angles and the spanwise bending in
(X, Y, Z) is a reference frame with the origin at the wing base: the one wingbeat cycle in a forward flying wasp (W1); the corresponding
X-axis in the horizontal direction and pointing forward, the Z-axis in data of other wasps are given in the Appendix (see Fig. 14). Flight
the vertical direction and pointing upward, and the Y-axis in the speed Vf, stroke plane angle b, and body angle v are also obtained
horizontal direction and pointing sideward (Fig. 4); (x, y, z) is another from the measurement. From the measured data, the stroke ampli-
reference frame also with the origin at the wing base, the x–y plane tude Ф (defined as Ф ¼ /max − /min), the wingbeat frequency f,
coinciding with and the z-axis normal to the stroke plane, and the the advance ratio Vf* (defined as Vf* ¼ Vf/2ФfR), and the Reynolds
y-axis pointing sideward, coinciding with the Y-axis [Fig. 4(b)]. Both number of wing Re (defined in Sec. I) are determined.
frames move with the body at constant speed. The stroke plane is Morphological data required for carrying out the flow computation
approximately horizontal when the wasp is at hovering flight.26 At were measured in the experiment; they are the wing length R, the
forward flight, it tilted forward by an angle b [see experimental data area of one wing S, the distance between the two wing roots lr,
in a later part of the present section (Sec. III A)]; b is referred to as and the radius of the second moment of wing area or radius of gyra-
the stroke plane angle [Fig. 4(a)]. The angle between the long-axis tion r2. These kinematical and morphological data are listed in
of the body and the horizontal is referred to as the body angle, v Table I.

FIG. 3. Video sequences of W1 at forward flight; three camera views are shown. Times noted are microseconds from the instant when the upstroke starts. Complete video
sequences can be seen in the movies. Movie 1: Flight of W1. The left, middle, and right parts of the movie show the flight captured by the top-view camera and two side-view
cameras, respectively. Playback speed is 15fps, 0.15% of the actual speed of the movie. Movie 2: Flight of W2. The left, middle, and right parts of the movie show the flight
captured by the top-view camera and two side-view cameras, respectively. Playback speed is 15fps, 0.15% of the actual speed of the movie. Movie 3: Flight of W3. The left,
middle, and right parts of the movie show the flight captured by the top-view camera and two side-view cameras, respectively. Playback speed is 15fps, 0.15% of the actual
speed of the movie. Multimedia views: https://doi.org/10.1063/5.0039911.1; https://doi.org/10.1063/5.0039911.2; https://doi.org/10.1063/5.0039911.3

FIG. 4. (a) Reference frames. (b) Euler


angles defining the wing kinematics.

Phys. Fluids 33, 021905 (2021); doi: 10.1063/5.0039911 33, 021905-4


Published under license by AIP Publishing
Physics of Fluids ARTICLE scitation.org/journal/phf

FIG. 5. (a) Measured Euler angles, /


and h. (b) Measured Euler angle, w, and
maximum bending displacement (dm/R).
Data are for wasp W1 (mean 6 s.d.; n
¼ 5 wingbeats); T, stroke period; R, wing
length.

TABLE I. Flight parameters for the five wasps.

ID Vf (m/s) Vf* f (Hz) Ф (deg) b (deg) v (deg) R (mm) S (mm2) lr (mm) r2 (mm) Re

W1 0.33 0.32 354 140.0 35.9 42 0.59 0.13 0.22 0.38 9.5
W2 0.27 0.25 351 146.8 30.52 45 0.61 0.14 0.23 0.39 10.5
W3 0.32 0.32 349 139.2 33.0 50.5 0.59 0.13 0.22 0.38 9.3
W4 0.25 0.26 328 137.3 22.7 58 0.60 0.13 0.22 0.38 8.9
W5 0.31 0.30 337 136.5 34.3 48 0.64 0.15 0.24 0.41 10.4

Using the data given in Fig. 5 and Table I, the stroke pattern of (t/T  0.25–0.5), the wings move close to each other at the back of the
the wing motion of W1 is plotted in Fig. 6; those of the other four insect and at the same time move upward (see the left and right panels
wasps are given in the Appendix (see Fig. 15). Figure 6 also shows the of Fig. 6); this phase is called the closing/up-moving phase. In the third
orientation of the wing and the velocity of the wing relative to the air phase (t/T  0.5–0.65), the wings fast “fling open”3,26 (see the right
at various times of the stroke cycle; here, the wing is represented by panel of Fig. 6). In the fourth or the last phase (t/T  0.65–1), the
the wing section at the radius of gyration of the wing, and the relative wing sweeps downward and forward at large angle of attack (see the
velocity is the vector summation of the velocity due to wing-flapping left panel of Fig. 6); this phase is called the downward/forward
(/_ and h)
_ and the velocity due to the insect's forward flight (V_ f ) (see sweeping phase. A flapping cycle of insects commonly comprises two
Fig. 7). half-strokes, the upstroke and the downstroke. Here, the impulsive
From Fig. 6, it is seen that a flapping cycle can be divided into rowing phase and the closing/upward-moving phase are in the
four phases. In the first phase (t/T  0–0.25; t denotes time, and T is upstroke, and the fling phase and the downward/forward sweeping
the flapping cycle period, T ¼ 1/f), the wings move backward at very phase are in the downstroke.
large velocity at large angle of attack (see the left panel of Fig. 6); this Comparing the forward flight case in Fig. 6 (left panel) and the
phase is called the “impulsive rowing” phase. In the second phase hovering flight case in Fig. 1(b), it is seen that motion of the wing

FIG. 6. Left) Diagram showing the wing-


motion pattern of W1; the dashed curve
indicates the wing-tip trajectory relative to
the body (projected onto the X–Z plane);
black lines indicate the orientation of the
wing at ten time instances, with dots mark-
ing the leading edge; the blue arrows rep-
resent the relative velocity of the wing
(projected on to the X–Z plane). (Right)
Back view of the orientation of the wing at
nine time instances; the blue arrows rep-
resent the relative velocity of the wing
(projected on to the Y–Z plane).

Phys. Fluids 33, 021905 (2021); doi: 10.1063/5.0039911 33, 021905-5


Published under license by AIP Publishing
Physics of Fluids ARTICLE scitation.org/journal/phf

FIG. 7. Components of the relative velocity of the wing.

relative to the body is generally similar in the two cases. The major dif-
ference between the two cases is that in the forward flight case, the
body of the insect tilts forward (the body angle is 45 ), while in the
hovering flight case, the body is approximately vertical (the body angle
is 90 ).

B. The aerodynamic forces


The flows around and aerodynamic forces on the insect are com-
puted by solving the incompressible Navier–Stokes equations, as
described in Sec. II. In the present study, the vertical component
FIG. 8. The time histories of the computed vertical (a) and horizontal (b) forces of
(Z-component) of the total aerodynamic force of a wing or the body is the wings (c) and the body (d) in one flapping cycle for W1.
referred to as the vertical force, and the horizontal component
(X-component) is referred to as the horizontal force or thrust; the
Y-component of the force is the side force (note that the side force
of the left wing cancels out that of the right wing). The vertical force of the thrust that propel the insect's forward motion is produced by
and the horizontal force (thrust) on the two wings are denoted by the impulsive rowing of the wings (t/T  0–0.25) (the impulsive
Vw and Hw, respectively, and those on the body are denoted by Vb rowing also provides some vertical forces). It can also be seen that
and Hb, respectively. Figure 8 gives the time histories of the com- in addition to the negative horizontal force on the body (body
puted aerodynamic forces of the wings and the body in one flapping drag), in some relatively long periods (t/T  0.25–0.55 and t/T
cycle for W1; results for the other four individuals are given in the  0.65–1), the wings also produce a relatively large negative hori-
Appendix (see Fig. 16). zontal force (negative thrust). These negative horizontal forces
The following is observed from Fig. 8. In the impulsive rowing are mainly overcome by the large thrust in the rowing phase (t/T
phase (t/T  0–0.25), in which the wings move fast backward at high  0–0.25). That is, the insect paddles itself through the air like
angle of attack, a very large thrust peak [Fig. 8(b)] and a relatively large people rowing a boat.
vertical force peak [Fig. 8(a)] are produced. In the closing/upward- These can also be seen clearly from the plot of the total aerody-
moving phase (t/T  0.25–0.5), in which the wings are close to each namic force on the wings (Fig. 9).
other and move upward, a negative vertical force [Fig. 8(a)] and a neg-
ative horizontal force [Fig. 8(b)] are produced. In the fling phase (t/T
 0.5–0.65), a relatively large positive vertical force [Fig. 8(a)] and C. How the aerodynamic forces are produced
some thrust [Fig. 8(b)] are produced. Finally in the downward/forward As seen in Sec. III B, large aerodynamic force peaks are produced
sweeping phase (t/T  0.65–1), a relatively large positive vertical force during the impulsive rowing phase, the fling phase, and the down-
[Fig. 8(a)] and negative thrust [Fig. 8(b)] are produced. As for the ward/forward sweeping phase [Figs. 8(a) and 8(b); t/T  0–0.25; t/T
body, a small vertical force [Fig. 8(c)] and a relative large negative hori-  0.55–1; Fig. 9]. These force peaks provide the weight supporting the
zontal force [Fig. 8(d)] are produced. vertical force and the thrust that overcomes the negative horizontal
From the above results, it is seen that the majority of the vertical forces of the wings and the body drag.
force that supports the insect weight is produced by the fling (t/T Let us define the lift and the drag of the wing as the components
 0.5–0.65) and the downward/forward sweeping of the wings (t/T of the total aerodynamic force that are perpendicular and parallel to
 0.65–1) (the fling also provides some thrust) and that the majority the relative velocity of the wing at the radius of gyration of the wing

Phys. Fluids 33, 021905 (2021); doi: 10.1063/5.0039911 33, 021905-6


Published under license by AIP Publishing
Physics of Fluids ARTICLE scitation.org/journal/phf

at the side edges [Fig. 11(a)], similar to that of a plate fast started in its
normal direction. From vortex dynamics theory,47 generation of large
opposite sign vorticity at different locations of the wing will result in
large time rate of change of the first moment of vorticity, giving the
large aerodynamic drag on the wing. The large drag can also be
explained in another simpler way: the wing starts from a relatively
small speed and fast accelerates at very large angle of attack, like a plate
smashes on the fluid, producing a large fluid impulse [Fig. 11(b)]; this
will result in a large positive pressure on its lower surface and a large
suction pressure on its upper surface [Fig. 11(c)], giving in the large
drag.
Next, we look at the fling phase (t/T  0.55–0.65); the flow
field data are plotted in Fig. 12. In the fling phase, the wings fast
FIG. 9. The total aerodynamic force on the wings at various times (green arrows) rotate about their trailing edges, starting from zero rotational veloc-
in one stroke cycle (left panel, side view; right panel, back view). ity, also generating large and opposite-sign vorticity at different
locations [Fig. 12(a)]. The wings also smash on the fluid and pro-
duce a large fluid impulse [Fig. 12(b)], giving a large positive pres-
(r2), respectively. The vertical and horizontal forces of a wing come sure on its lower surface and a large suction pressure on its upper
from its lift and drag. So, let us look at the lift and drag of the wings surface [Fig. 12(c)]. The force production mechanism is similar to
and analyze how they are generated. Figure 10 shows the lift (L) and that of the impulsive rowing motion. Note that in this phase, a
drag (D) of the wings of W1 in the flapping cycle. leading-edge vortex (LEV) is produced by each of the wings [see
As seen from Fig. 10, very large drag and much smaller lift are the panel in Fig. 12(a) at t/T ¼ 0.64].
produced by the wing. That is, the large thrust and large vertical force Finally, we examine the downward/forward sweeping phase (t/T
in the rowing phase, fling phase, and downward/forward sweeping  0.65–1); the flow field data are given in Fig. 13. In this phase (which
phase are mainly contributed by the wing drag. This can be seen more is subsequent to the fling phase), the LEV produced in the fling phase
clearly from Fig. 9: the total aerodynamic force vector is almost in the attaches to and moves with wing, or the LEV is recaptured by the wing
opposite direction of the wing velocity vector. [Figs. 13(a) and 13(b)]. This is the well-known delayed stall mecha-
In order to explain how the large wing drag is produced, we nism of aerodynamic force production. That is, the large aerodynamic
examine the flow field data of the wings. First, we consider the impul- force in the downward/forward sweeping phase is due to the LEV or
sive rowing phase (t/T  0–0.25). In this phase, the wings fast acceler- delayed-stall mechanism. Another way to explain the force production
ate backward at very high angle of attack; i.e., the wings fast smash on in this phase is as follows. When the wing having a vortex (the LEV)
the air. Figure 11 gives the flow field data in this phase. It is seen that on its upper surface, a suction pressure will be produced by the vortex
when the wings move, starting from zero velocity, vorticity of opposite [see the panels in Fig. 13(c) at t/T ¼ 0.70, 0.75, and 0.80], giving the
sign is created continuously at the leading and trailing edges and also wing drag.
As aforementioned, the fling phase and the downward/
forward sweeping phase are the downstroke of the very small wasp
(the fling is at the beginning of the downstroke). For the previously
studied larger insects (no fling at the beginning of a downstroke),
the wing needs to move (azimuthally rotate) about 30 –40 before
the LEV is formed; when the Reynolds number is very low, i.e., the
viscous effect is very large, the vortex formed after such a long
time is rather diffused and cannot be very effective.27–29 Here, to
overcome the large viscous effect, the small wasps “cleverly” use the
fling motion at the beginning of the downstroke. This is a very
interesting point.
We, thus, see that smashing on the air at large acceleration by
the wings is one method used by the tiny wasps to produce a large
aerodynamic force in a very viscous flow: in the impulsive rowing,
very large translational acceleration is used; in the fling, very large
rotational acceleration is used. Another method is using the fast
fling to produce a strong leading-edge vortex and keep the vortex
on the wing in the rest of the downstroke (as aforementioned, in a
normal downstroke used by large insects, when the flow is very
viscous, a strong leading-edge vortex cannot be produced). Using
large acceleration to produce a large aerodynamic force in very vis-
FIG. 10. The time histories of the lift L (a) and drag D (b) of the wings of W1 in the cous flow has also been found in the case of hovering flight of the
flapping cycle. tiny wasps in our previous study.26 But at there, the method was

Phys. Fluids 33, 021905 (2021); doi: 10.1063/5.0039911 33, 021905-7


Published under license by AIP Publishing
Physics of Fluids ARTICLE scitation.org/journal/phf

FIG. 11. Vorticity fields, flow fields, and


pressure distribution on the wing surface
in the impulsive rowing phase (t/T
 0–0.25). (a) Isovorticity surface plots of
the wing; vorticity is non-dimensionalized
by U/c, and the magnitude of the non-
dimensional vorticity is 2. (b) Non-
dimensional flow velocity vector plots at
section 0.7R (R is the wing length) at the
corresponding times. (c) Pressure distribu-
tions on the upper surface of the wing
(first row) and the lower surface of the
wing (second row) at the corresponding
times; the pressure is non-
dimensionalized as (p–p1)/0.5qU2, where
p is the pressure, p1 the reference pres-
sure, and q is the air density.

used in vertical-force production. Here, we found that the method Having explained the aerodynamic-force production mecha-
is used in thrust production. Using fast started rotation about nisms of the tiny wasps in forward flight, let us consider the aerody-
the trailing edge to produce a strong leading-edge vortex in a very namic power required to produce the aerodynamic forces. From
viscous flow and keep the leading-edge vortex in the rest of the the computed aerodynamic force and the wing motion data, the
downstroke for large aerodynamic force production is a mechanism power used by the wing to generate the forces can be calculated,
observed for the first time in insect flight. and the mean power and mean vertical force (average over one

FIG. 12. Vorticity fields, flow fields, and


pressure distribution on the wing surface
in the fling phase (t/T  0.55–0.65). (a)
Isovorticity surface plots of the wing; vor-
ticity is non-dimensionalized by U/c, and
the magnitude of the non-dimensional vor-
ticity is 2. (b) Non-dimensional flow veloc-
ity vector plots at section 0.7R (R is the
wing length) at the corresponding times.
(c) Pressure distributions on the upper
surface of the wing (first row) and the
lower surface of the wing (second row) at
the corresponding times.

Phys. Fluids 33, 021905 (2021); doi: 10.1063/5.0039911 33, 021905-8


Published under license by AIP Publishing
Physics of Fluids ARTICLE scitation.org/journal/phf

FIG. 13. Vorticity fields and pressure distribution on the wing surface in the fling phase (t/T  0.55–0.65) and the subsequent downward/forward-sweeping phase (t/T
 0.65–1). (a) Non-dimensional isovorticity surface plots (top view) at t/T ¼ 0.65, 0.70, 0.75, and 0.8; vorticity is non-dimensionalized by U/cm, where cm is the mean chord
length of the wing. (b) Non-dimensional spanwise vorticity contours at the corresponding times. (c) Non-dimensional pressure distributions on the upper surface of the wing
(first row) and the lower surface of the wing (second row) at the corresponding times.

flapping cycle) can also be calculated. Dividing the mean power by motion pattern and aerodynamic force generation of a very small
the mean vertical force gives the specific power, i.e., power used for wasp, Encarsia formosa, in forward flight. Our findings are the
per Newton weight lifted, denoted by P*. For the tiny wasps in for- following.
ward flight, P* ranges from 2.18 W/N to 3.05 W/N; the average P* The flapping cycle consists of four phases: impulsive rowing (the
for the five wasps is 2.56 W/N. For comparison, we also computed wings fast accelerate backward at very large angle of attack); clos-
the values of P* for the wasps in hover flight, studied by the authors ing/upward-moving (two wings come together at the back of the
previously.26 For the eight wasps in hover flight,26 P* ranges from insect and move upward); fling (the wings fling apart like opening a
2.08 W/N to 2.52 W/N; the average P* for the eight wasps in hover book); and downward/forward sweeping. The first and second
flight is 2.36 W/N. We see that at medium speed forward flight, the phases are in the upstroke, and the third and fourth phases are in
power requirement is slightly larger than that in hovering flight, the downstroke. In the impulsive rowing phase (the first half of the
9.5% larger. upstroke), the wing smashes in the air and generates strong drag
In order to achieve the large aerodynamic forces, as aforemen- that points forward and slightly upward, providing a large thrust
tioned, the tiny wasps adopt four different phases within one flapping peak and a small vertical force peak. In the closing/upward-moving
cycle (impulsive rowing, closing/upward-moving, fling, and down- phase (the second half of the upstroke), the wing slices through the
ward/forward sweeping). It is of great interest to know whether these air and generates much smaller drag, giving a small negative verti-
four stages are unique to Encarsia formosa in forward flight or they are cal force and negative thrust. In the fling phase (the start of the
a strategy that is widely used by other miniature insects in a relatively downstroke), large drag is also produced by fast pushing the air,
low-Re regime in forward flight. This will be investigated in our future providing some vertical forces and thrust; at the same time, a strong
experiments. leading edge vortex (LEV) is created on each wing. In the down-
ward/forward sweeping phase (the second, longer part of the down-
stroke), the wing carries the LEV on its upper surface and produces
IV. CONCLUDING REMARKS large drag that points upward and backward, giving a large vertical
The wings of miniature insects operate at very low Reynolds force and negative thrust.
number (Re in the order of 10), and the viscous effect is large. If their The vertical force produced in the impulsive rowing, the fling,
wings flap like those of larger insects (Re in the order of 100–1000), and the downward/forward sweeping provides the weight supporting
aerodynamic forces required for weight supporting and propulsion force; the large thrust produced by the impulsive rowing and the fling
cannot be produced. A new flapping mode must be used to overcome overcomes the body drag and the negative thrust produced by the
the large viscous effect. In the present paper, we studied the wing wings in the rest of the flapping cycle.

Phys. Fluids 33, 021905 (2021); doi: 10.1063/5.0039911 33, 021905-9


Published under license by AIP Publishing
Physics of Fluids ARTICLE scitation.org/journal/phf

The tiny insects overcome the strong viscous effect by fast ACKNOWLEDGMENTS
smashing the wings on the air in the first half of the upstroke, by
fast flinging the wings at the beginning of the downstroke, and by This research was supported by grants from the National
carrying the LEV created at the fling to move in the rest of the Natural Science Foundation of China (Grant Nos. 11832004 and
downstroke. 11721202).

APPENDIX: DATA FOR OTHER FOUR WASPS

FIG. 14. Measured wing kinematics for


W2 (a), W3 (b), W4 (c), and W5 (d)
(mean 6 s.d.). For more details, see the
legend of Fig. 5.

Phys. Fluids 33, 021905 (2021); doi: 10.1063/5.0039911 33, 021905-10


Published under license by AIP Publishing
Physics of Fluids ARTICLE scitation.org/journal/phf

DATA AVAILABILITY
The data that support the findings of this study are available
from the corresponding author upon request.

REFERENCES
1
A. Polilov, At the Size Limit—Effects of Miniaturization in Insects (Springer
International Publishing, Cham, Germany, 2016).
2
R. Dudley, The Biomechanics of Insect Flight: Form, Function, Evolution
(Princeton University Press, 2002).
3
T. Weis-Fogh, “Quick estimates of flight fitness in hovering animals, including
novel mechanisms for lift production,” J. Exp. Biol. 59, 169–230 (1973).
4
C. P. Ellington, “The aerodynamics of hovering insect flight. II. Morphological
parameters,” Philos. Trans. R. Soc. B: Biol. Sci. 305, 17–40 (1984).
5
C. P. Ellington, “The aerodynamics of hovering insect flight. III. Kinematics,”
Philos. Trans. R. Soc. B: Biol. Sci. 305, 41–78 (1984).
6
S. N. Fry, R. Sayaman, and M. H. Dickinson, “The aerodynamics of hovering
flight in Drosophila,” J. Exp. Biol. 208, 2303 (2005).
7
Y. Liu and M. Sun, “Wing kinematics measurement and aerodynamics of hov-
ering droneflies,” J. Exp. Biol. 211, 2014 (2008).
8
S. M. Walker, A. L. R. Thomas, and G. K. Taylor, “Deformable wing kinemat-
ics in free-flying hoverflies,” J. R. Soc. Interface 7, 131 (2010).
FIG. 15. The wing-motion patterns of W2, W3, W4, and W5.
9
C. Hefler, R. Noda, H. H. Qiu, and W. Shyy, “Aerodynamic performance of a
free-flying dragonfly-A span-resolved investigation,” Phys. Fluids 32, 041903
(2020).
10
C. P. Ellington, C. van den Berg, A. P. Willmott, and A. L. R. Thomas,
“Leading-edge vortices in insect flight,” Nature 384, 626–630 (1996).
11
H. Liu, C. Ellington, K. Kawachi, C. V. D. Berg, and A. P. Willmott, “A compu-
tational fluid dynamic study of hawkmoth hovering,” J. Exp. Biol. 201, 461–477
(1998).
12
R. J. Bomphrey, R. B. Srygley, G. K. Taylor, and A. L. R. Thomas, “Visualizing
the flow around insect wings,” Phys. Fluids 14, S4 (2002).
13
M. Sun and J. Tang, “Unsteady aerodynamic force generation by a model fruit
fly wing in flapping motion,” J. Exp. Biol. 205, 55–70 (2002).
14
S. P. Sane, “The aerodynamics of insect flight,” J. Exp. Biol. 206, 4191–4208
(2003).
15
Y. L. Yu, B. G. Tong, and H. Y. Ma, “An analytical approach to theoretical
modeling of highly unsteady viscous flow excited by wing flapping in small
insects,” Acta Mech. Sin. 19, 508–516 (2003).
16
Z. J. Wang, J. M. Birch, and M. H. Dickinson, “Unsteady forces and flows in
low Reynolds number hovering flight: Two-dimensional computations vs
robotic wing experiments,” J. Exp. Biol. 207, 449 (2004).
17
J. Lee, H. Choi, and H.-Y. Kim, “A scaling law for the lift of hovering insects,”
J. Fluid Mech. 782, 479–490 (2015).
18
G. Liu, H. Dong, and C. Li, “Vortex dynamics and new lift enhancement mech-
anism of wing-body interaction in insect forward flight,” J. Fluid Mech. 795,
634–651 (2016).
19
K. B. Lua, Y. J. Lee, T. T. Lim, and K. S. Yeo, “Aerodynamic effects of elevating
motion on hovering rigid hawkmothlike wings,” AIAA J. 54, 2247–2264
(2016).
20
D. D. Chin and D. Lentink, “Flapping wing aerodynamics: From insects to ver-
tebrates,” J. Exp. Biol. 219, 920–932 (2016).
21
X. G. Meng, “Ceiling effects on the aerodynamics of a flapping wing at hovering
condition,” Phys. Fluids 31, 051905 (2019).
22
J. D. Eldredge and A. R. Jones, “Leading-edge vortices: Mechanics and model-
ing,” Annu. Rev. Fluid Mech. 51, 75–104 (2019).
23
M. Sun and J. H. Wu, “Large aerodynamic forces on a sweeping wing at low
Reynolds number,” Acta Mech. Sin. 20, 24–31 (2004).
24
X. X. Wang and Z. N. Wu, “Stroke-averaged lift forces due to vortex rings and
their mutual interactions for a flapping flight model,” J. Fluid Mech. 654,
453–472 (2010).
25
J. Yao and K. S. Yeo, “Forward flight and sideslip manoeuvre of a model hawk-
moth,” J. Fluid Mech. 896, A22 (2020).
26
X. Cheng and M. Sun, “Very small insects use novel wing flapping and drag
FIG. 16. The time histories of the computed vertical (a) and horizontal (b) forces of principle to generate the weight-supporting vertical force,” J. Fluid Mech. 855,
the wings (c) and the body(d) in one flapping cycle for W2, W3, W4, and W5. 646–670 (2018).

Phys. Fluids 33, 021905 (2021); doi: 10.1063/5.0039911 33, 021905-11


Published under license by AIP Publishing
Physics of Fluids ARTICLE scitation.org/journal/phf

27
L. A. Miller and C. S. Peskin, “When vortices stick: An aerodynamic transition 37
S. H. Lee, M. Lahooti, and D. Kim, “Aerodynamic characteristics of unsteady
in tiny insect flight,” J. Exp. Biol. 207, 3073–3088 (2004). gap flow in a bristled wing,” Phys. Fluids 30, 071901 (2018).
28
J. H. Wu and M. Sun, “Unsteady aerodynamic forces of a flapping wing,” 38
G. Du and M. Sun, “Effects of unsteady deformation of flapping wing on its
J. Exp. Biol. 207, 1137–1150 (2004). aerodynamic forces,” Appl. Math. Mech. 29, 731–743 (2008).
29
Y. Z. Lyu, H. J. Zhu, and M. Sun, “Aerodynamic forces and vortical structures 39
M. Sun and X. Yu, “Aerodynamic force generation in hovering flight in a tiny
of a flapping wing at very low Reynolds numbers,” Phys. Fluids 31, 041901 insect,” AIAA J. 44, 1532–1540 (2006).
(2019). 40
X. G. Meng and M. Sun, “Aerodynamic effects of wing corrugation at gliding
30
Y. Z. Lyu, H. J. Zhu, and M. Sun, “Flapping-mode changes and aerodynamic flight at low Reynolds numbers,” Phys. Fluids 25, 071905 (2013).
mechanisms in miniature insects,” Phys. Rev. E 99, 012419 (2019). 41
K. Taira and T. Colonius, “Three-dimensional flows around low-aspect-ratio
31
G. R. Spedding and T. Maxworthy, “The generation of circulation and lift in a flat-plate wings at low Reynolds numbers,” J. Fluid Mech. 623, 187–207 (2009).
rigid two-dimensional fling,” J. Fluid Mech. 165, 247–272 (1986). 42
J. R. Usherwood and C. P. Ellington, “The aerodynamics of revolving wings II.
32
S. Sunada, H. Takashima, T. Hattori, K. Yasuda, and K. Kawachi, “Fluid- Propeller force coefficients from mayfly to quail,” J. Exp. Biol. 205, 1565–1576 (2002).
dynamic characteristics of a bristled wing,” J. Exp. Biol. 205, 2737–2744 (2002). 43
F. Manar and A. R. Jones, “The effect of tip clearance on low Reynolds number
33
M. Sun and X. Yu, “Flows around two airfoils performing fling and subsequent rotating wings,” AIAA Paper No. 2014-1452, 2014.
translation and translation and subsequent clap,” Acta Mech. Sin. 19, 103–117 44
J.-S. Han, J. W. Chang, and J.-H. Han, “The advance ratio effect on the lift aug-
(2003). mentations of an insect-like flapping wing in forward flight,” J. Fluid Mech.
34
L. A. Miller and C. S. Peskin, “A computational fluid dynamics of ‘clap and 808, 485–510 (2016).
fling’ in the smallest insects,” J. Exp. Biol. 208, 195–212 (2005). 45
S. E. Rogers, D. Kwak, and C. Kiris, “Steady and unsteady solutions of the
35
N. Arora, A. Gupta, S. Sanghi, H. Aono, and W. Shyy, “Lift-drag and flow incompressible Navier-Stokes equations,” AIAA J. 29, 603–610 (1991).
structures associated with the ‘clap and fling’ motion,” Phys. Fluids 26, 071906 46
S. Rogers and T. Pulliam, “Accuracy enhancements for overset grids using a
(2014). defect correction approach,” in 32nd Aerospace Sciences Meeting and Exhibit
36
A. Santhanakrishnan, A. K. Robinson, S. Jones, A. A. Low, S. Gadi, T. L. (American Institute of Aeronautics and Astronautics, Reston, Virigina, 1994).
Hedrick, and L. A. Miller, “Clap and fling mechanism with interacting porous 47
J. C. Wu, “Theory for aerodynamic force and moment in viscous flows,” AIAA
wings in tiny insect flight,” J. Exp. Biol. 217, 3898–3909 (2014). J. 19, 432–441 (1981).

Phys. Fluids 33, 021905 (2021); doi: 10.1063/5.0039911 33, 021905-12


Published under license by AIP Publishing

You might also like