You are on page 1of 22

POWERED MANNED ORNITHOPTER

Project Report
ME597 Bio-Inspired Robotics

Raja Manish
School of Aeronautics and Astronautics
Purdue University

1
Acknowledgement

I thank Prof. Xinyan Deng for giving me the opportunity to pursue this course where I could
work in field of my interest. I also thank Prof. Justin Seipel for providing me wonderful
ideas. I think I am going to spend some of my good time working on this in coming days too.

R. Manish

2
Abstract

This paper presents the theoretical concept of developing a powered Ornithopter taking into
consideration ideas from various attempts to build such machine in the past. The wing design has been
made in such a way that it more or less imitates the natural flapping of the birds of suitable size and
pattern. The mechanism mentioned in this report is based on multiple bar linkages in order to
accomplish exact flapping behavior of an actual bird. The idea is theoretically simplified and then
computed and analyzed in the aerodynamic and other tools. Subsequent development and prototype
testing would verify the authenticity of the theoretical conclusions reached.

3
Contents Page

1. Introduction and Background 5

2. Proposed Ideas And Project Goals 6

3. Project Requirement 7

4. Design
4.1 Wings 8
4.2 Gear box 13
4.3 Frame 14

5. Testing and Analysis 15

6. Result And Discussion 18

7. References 19

8. Appendix 20
CAD drawing 21-22

4
1. Introduction and Background

Humans have always dreamed of flying like bird. The development of Ornithopters has been
around even before the invention of flying aircraft by wright brothers. Though the idea of flying like a
bird could never be achieved in sustainable way, there have been notable progresses in the field so far.
It has found that flapping flight has better propulsive efficiency than propeller based flight. The
invention of hang gliders, and aircraft enabled us to soar in air which is great accomplishment in
itself. But still, researchers are trying to understand and replicate the flapping flight of birds, which
has seen little achievements. This mystery need to be solved.

Several academic and non-academic organizations have involved in researches on powered or


unpowered, manned and unmanned flight using flapping wings for propulsion.

Fig. 1: The powered Ornithopter by DeLaurier [4] is one of the notable work in recent.

Since humans are naturally not evolved with stronger limb muscles as birds are, it becomes unlikely
for human to achieve similar performance without bridging the gap between the natural strength of the
body and the equivalent strength of the limbs required to propel a wing.

There has to be a system compact enough in size and weight which when hung to human body would
enable him to propel the wing flapping motion and hence generate lift as well as thrust.

5
2. Proposed Idea and Project goals

A bird is able to propel itself by following complicated movement of its wing having multiple degrees
of freedom (Fig. 2). Idea is to imitate the flapping motion of a large bird. Flapping includes pitch,
sweep as well as dihedral movement of the wing about the body.

Fig. 2: Axes of motion- sweep, pitch, and dihedral motion about x, y and z

In previous researches it was found that normal flapping leads to shock loading of the wing at the end
of each stroke which reduces effective thrust of the system. A simple yet effective method to
minimize such loading on body as well as wing has to be determined.

The project intends to build an independent framework containing power source, flapping wings and
control system that would be attached with a human so as to enable the complete human-machine
system in achieving a certain level of controlled flight.

Also, power would be derived mostly from artificial sources while human limb would perform control
operations.

6
3. Project requirement

 Light weight design


 Low wing loading shock
 Flexible wing movement instead of torsional type
 High controlled lift at low velocity
 Minimum take off distance
 Sustained flight for much time
 Minimum inertia effect on the wing
 Deliver best possible performance using minimum power input

Side view of created Design

7
4. Design

4.1 Wings

Since the forward speed is low in case of Ornithopters, the flapping wing has to be of high lift at low
speed and capable of providing sufficient forward thrust as well. From the studies of Liu ET AL [1] it
was found that wings with cross section of airfoil S1223(fig 3(a)) has high lift capability and has
similar characteristic when compared with birds like Seagull and Merganser. Flapping motion
requires periodic variation in angle of attack (AoA). As from the properties of the airfoil S1223 it was
found that it has large positive lift coefficient for wide range of AoA (fig 4). Hence this was chosen
for the wing section.

Cl
CD

CT

Fig. 3(a): S1223 airfoil compared to Seagull and Merganser airfoils. (Liu Et Al [1]); (b) Forces acting
on an aircraft, CT = Thrust, CD = Drag, W = weight, CL = Lift

T pmech
Here avg. thrust co-efficient, CT  , Avg. Power co-efficient, C p 
1 1
U 2 s U 3 s
2 2

CT
And Efficiency   , where T = avg. thrust, ρ = density of air, U = wind velocity, s = surface
CP
area, P = average mechanical power input

8
While drag decreases the efficiency of flight, it is essential for its stability. Total drag over the wing is
given by,

Dtotal = Dbody + Dtail + Dwing + Dviscous + Dinviscid

As speed of flight is low the inviscid drag can be neglected. The effective total can be calculated and
verified using computational tools.

S1223 has very large lift to drag ratio for wide AoA and hence is suitable for low speed operation
owing to its high camber (Fig 4).

Fig. 4: Performance curves of S1223 obtained from XFLGR/XFOIL for various Reynolds number and
Mach 0.05 (20 m/s)

This airfoil also has large value of negative moment coefficient. So the system has tendency to bring
the nose downwards. To counter the pitching moment of the system, a horizontal tail is to be
determined which would balance the effect of moment from a distance at aft. A standard symmetric
airfoil NACA 0012 was chosen for this purpose. It is thin an efficiently employed over wide variety
of aircraft from gliders to large airplanes.

9
Fig. 5: Performance curves of NACA 0012 obtained from XFLGR/XFOIL for various Reynolds
number and Mach 0.05 (20 m/s)

Wing Calculation

Consider maximum load on wings, W = 200-250 Kg.

Wing loading of large birds is about 20-25 Kg/m2

Following similar assumption, we have

Wing area = Avg. Total load/Wing loading = 225/25 ~ 9 m2

Birds with long narrow wing (aspect ratio > 4) such as Albatross have great soaring capability and has
less induced drag on the wing while short wide wings (aspect ratio <3) provide good maneuverability.
Since flapping flight requires change in inertia in each cycle, long narrow wings are not preferred for
wing design. Hence, a moderate aspect ratio of 4 was chosen.

If average wing span, b = 6m

Then average chord, C = Span/AR = 6/4 = 1.5m (From formula Aspect ratio = span x chord)

10
Now, general flapping freq. of wing is given by Lippisch Formula[3], f =

Where, W = 200-250 Kg; span b = 6 m; density of air ρ = 1.2 kg/m3

Hence, f = 1.2 to 1.4

In fluid mechanics, a dimensionless number known as Strouhal number is obtained to describe


oscillating mechanism [7]. Propulsive efficiency is high in birds over a narrow range of 0.2 < St < 0.4.

A f
Strouhal number, St = U ,

Where A = amplitude of oscillation = maximum oscillation angle covered x semi span,

f = frequency of oscillation,

U = wind velocity

 55 
    3
We have, A =  180  = 2.88 m, f = 1.4 s-1 and U = 20 m/s

We get, St = 0.20

Therefore frequency of flapping of 1.4 up to 2.8 is efficient for cruise.

The wing planform designed by DeLaurier was modeled as torsional beam (Fig. 6(a)). Though it was
able to perform quite well along its design theory, the partially rigid structure failed to sustain lift. The
design considered in this paper is based on multiple joint links allowing easy sweep motion along
with heaving and dihedral. Also, the wing has incremental dihedral angle so as to bring the CG near
the root as well as to reduce inertia effect (fig 6(b)). Tip vortices are less pronounced at low speed
flight, so it can be neglected in our case.

Fig. 6(a): DeLaurier Ornithopter wing

11
Kinematic Linkages with
Dihedral swept back wing

Fig. 6(b): Ornithopter Design created in this paper

Any airfoil beyond its limit of positive lift coefficient (CL) goes into stall because of flow separation
from boundary and thus loses its lift. Airfoil S1223 has positive CL from -4 to 14 degrees. Our
intended range of operation lies in much narrow bracket (2 to 6 degree, αup and αdown) inside the above
limit. Hence, we do expect any stall issue within our limit of use.

Following figure shows wing motion of typical bird with αdown and αup.

Fig. 6(c) Force diagram of a typical bird (α = Angle of Attack)[3]

12
4.2 Gear Box

The idea of the gear box is from Kestrel designed RC Ornithopter [8].

7575 N

0.165m 1m

0.051m
1250 N
386 Nm

9.65N
Fig. 7(a): Kestrel Ornithopter [8]; (b)
m mechanism involved

Our main purpose is to achieve the required angular wing movement using minimum power input.

Since frequency of flight = 1.4 to 2 beats/second ~ 90 to 120 beats/min

Considering equal load on each wing = 250Kg/2 = 125 Kg ~ 1250 N

Generally, moment of a bird wing is highest at a distance of one-third of semi span from root.
Assuming load to be concentrated at that point

Therefore required torque to move each wing = Load x Dist. = 1250 N x 1 m =- 1250 N m

This amount of torque is to be transferred from power source to each wing by converting rotary
motion to linear using off-centered worm wheel (fig.7 (b)). A standard IC engine has rotation speed of
not less than 3600 rpm and mostly near 5500 rpm to achieve most efficient torque-power curve.
Selecting an available rpm range of 3800-5200, we have Reduction ratio required = 3800/90 ~ 43
(chosen 3800 rpm)

This amount of speed reduction is possible either by using multiple spur gear combination or by using
single pair of worm-worm wheel. It is known that worm pair has very efficient performance where
high reduction is required. Also, it provides a reverse lock feature. So we selected a worm pair of ratio
1:40 for our purpose. Next, reduction in speed means torque multiplication. Therefore, torque required
at output shaft of engine to drive the wing is given as follows:

Torque to wheel = 2x 1250 Nm  0.051m = 772 Nm. TEngine = 772/40 = 19Nm @ 3800-5200 maximum
0 165m

Gears for each pair of wing element are arranged in out-of-phase manner (fig. 8). This out of phase
arrangement ensures that wing is always loaded and torque is continuously transferred from the
engine shaft to wing. By this way, sudden wing loading at the end of each stroke in reduced.

13
4.3 Frame

Frame is designed based on the weight criteria and the dimension of the internal drive system.

From aerodynamic point of view, frame need to be as compact as possible so as to minimize the
moment of inertia.

Material used is to be alloy metal at critical load bearing areas while carbon composite at rest of the
places.

Fig. 8 (b): 3D model of constructed Frame. Inner detail shows Drive mechanism. Note the out of
phase gears

Ball joint

rotational

1 DoF angular

TOP VIEW

Fig. 8(b): Wing planform showing kinematic link/joints

14
5. Testing and Analysis

Frame structure Analysis


Stress Analysis was performed over the frame keeping the frame base as fixed while upward load of
2500 N on the wing member and a high opposing torque moment of 4000Nm (our designed moment =
772 Nm) on the worm wheel. I was little skeptic about applying Kestrel Ornithopter mechanism but
these analysis predicted very good results (Fig. 8-9).

Stress is mostly concentrated at the worm wheel shaft which is obvious because of the torsional
loading on it.

Fig.9: Static loading- Stress concentration

15
Factor of safety was determined under full static load. Results obtained were good with minimum
FOS about 2.48 near the wheel shaft.

Fig.10: Factor of safety


Maximum deformation obtained is about 0.066mm at teeth of worm wheel which is acceptable.
Remember we applied 5 x Torque that we’ve got in calculation. So, actual FOS would be much more.

Fig.11: Maximum deformation under full load

16
Aerodynamics Analysis

Since Angle of attack varies during flapping, we consider a range of angles during our analysis

Fig.12: Static pressure variation at AoA = 0 deg. (front view, tool-XFLOW 3D CFD)
Note that the pressure difference between upper and lower surface of wing near the root is maximum
(about 210 Pa) which gradually decreases towards tip (about 44 Pa). Hence we are generating lift at
AoA = 0 near the wing root. Also, notice there is a sudden surge in pressure at center near the body.
This is because the body is not covered and is not aerodynamic during test.

Fig.13: Static Pressure variation at AoA = 8 deg. (front view, tool-XFLOW 3D CFD)
This time, pressure difference is maximum at the middle of each wing, i.e. about 900 Pa. Hence we
generate most of the lift at the middle of the each wing.

17
6. Result

The idea of kinematic link joints has been applied into the model. This concept need comprehensive
testing as this was not implemented practically in an Ornithopter. Some of the results obtained from
analysis are found to be in coherence with standard and theories. This is good at this stage of the
work. The wing geometry is near to the pattern of a large bird with an attempt to adopt similar
features. The aerodynamic analysis is being done on the model and the output obtained so far gave us
a good insight into a possible flight. While most of the concentration is on designing the flapping
wing, designs for the horizontal tail, landing gear as well as detailed dynamic analysis of wing is
being carried out. Also, current work is on the dynamics and control of the system.

I would say that the design idea presented here is preliminary in much sense and may have many
points for further refinement. Further work with possible prototype testing would enable us in
verifying each and every part of the system.

18
7. References:
[1] Liu, Tianshu ET. Al “Avian Wing Geometry and Kinetics”, AIAA Vol. 44, No. 5 May

2006

[2] Brown, R.H.J, “The flight of Birds”, Biol. Rev., 38, 1963

[3] Lippisch, A., “Man powered flight in 1929”, Journal of Royal Aeronautical Society,

vol. 64, July 1960

[4] DeLaurier, James, “An Ornithopter Wing design”, Canadian Aeronautics and Space

Journal, Vol. 40 No. 1, March 1994

[5] Selig, Michael S., “High Lift Low Reynolds Number Airfoil Design”, Journal of Aircraft,

Vol. 34, No. 1, Jan-Feb 1997

[6] Jackowski, Zachary John, “Design and Construction of an Autonomous Ornithopter”,

MIT, June 2009

[7] Strang, Karl Axel, “Efficient Flapping Flight of Pterosaurs”, Stanford University, Sept

2009

[8] Kestrel Ornithopter “www.randrmodelaircraft.com/Kestrel.htm”

19
8. Appendix

CAD drawings of the model:


i. Overview
ii. Inner detail

20
6 5 4 3 2 1

8299.25
2314.90 499.13 3573.51

874.53

714.60
423.86
D 1035.72 D

174.60 RIB (Symmetric)


152.3 Chord 1.215m
8.8 TYP 2 Nos. (1 each side)

RIB (Semi S1223)


Chord 1.215m
TYP 2 Nos. (1 each side)
Top View RIB (Profile S1223)
SCALE 1 : 40 Chord 1.5m
TYP 10 Nos. (5 each side) 1352.28

C C
841.15

7.2

Pilot to be attached
with harness (not shown)
Front View
SCALE 1 : 40
Right Hand Side View
SCALE 1 : 40

B B

ME 597 PROJECT

TITLE
Iso View
SCALE 1 : 50 Powered Ornithopter Preliminary design /
A Profile Views A
NOTE:
1. All dimensions are in mm. DRAWN SIZE DWG NO REV
2. Wing cover and right side wing ribs not shown for clarity. 12/14/2013
R. Manish
3. All Structural frame member made of composite metal/synthetics CHECKED A3 -
4. Landing Gear, Horizontal tail work in progress. SCALE
SHEET 1 OF 1
6 5 4 3 2 1
6 5 4 3 2 1

D D

4 bar Mechanism

1 DOF Joints

2 ST. horizontal Shaft


IC Engine
C C

Continued section
for tail attachment
Engine O/P shaft
25.0 driving 2 worms w/
Worm - Worm Wheel
Phase difference equal speed
Ratio 1:60
B 2 Nos. B

Right Hand Side View (Cutout/Ribs removed)


SCALE 1 : 10

ME 597 PROJECT

TITLE

Powered Ornithopter Preliminary design /


A Iso View Internal Detail A
NOTE:
SCALE 1 : 50
1. All dimensions are in mm. SIZE DWG NO REV
DRAWN
2. Wing cover and right side wing ribs not shown for clarity.
R. Manish 12/14/2013 A3
3. All Structural frame member made of composite metal
4. Landing Gear, Horizontal tail work in progress.
CHECKED
SCALE
-
SHEET 1 OF 1
6 5 4 3 2 1

You might also like