4.1 Hummingbird Characteristics
4.1.1 Flight Characteristics
4.1.2 Hovering Flight Characteristic
4.2 Technology: Hardware and Software Used in
188.8.131.52 How MSC/NASTRAN Works
184.108.40.206 Aeroelastic Analysis Using MSC/NASTRAN
220.127.116.11 Examples of Bulk Data Entries
4.3 MAV Technology and Science
4.0 Background Theory
This section is intended to provide the reader
with necessary background information on the hummingbird characteristics,
the hardware/software used in the project, and the research findings related
to MAVs. The first section will introduce the hummingbird and the special
characteristics it possesses. The next section will present a detailed
description of those characteristics and why they are important. The last
section will list software and hardware used in the research of the project
and also a detailed portion on MSC/NASTRAN. Finally, MAV research will
The hummingbird family has about 300 different species
and can be found on the continents of North, Central, and South America.
Figure 4.1 portrays the Phaethornis Yaruqui species. Hummingbirds are quite
small with a length of only 2 ¼ in to 8 ½ in, however, they
are not the smallest of all birds. They eat the nectar of flowers for survival
and can consume up to half their weight in sugar daily. The reason for
this enormous appetite is the hummingbird’s extraordinary flight capability.
No other bird can hover as long or as steady as the hummingbird, although,
a lot of energy is expended to stay aloft in this hovering attitude. The
disadvantage of hovering is the excessive energy required for its success.
The excessive energy requires the hummingbird to consume a lot of food.
The energy output of a hummingbird in hovering flight is ten times as much
as a man running nine miles an hour. Direct comparisons to a human being
show that a 170-pound man would have to consume about 130 pounds of bread
to keep up with a hummingbird’s energy-output [5, pp. 3-9].
As previously mentioned, the hummingbird can hover
more consistently and efficiently than any bird and is the only bird that
has the ability to hover with its body motionless. The hummingbird’s bone
structure allows it to move its wings quite differently from "ordinary"
birds. Figure 4.2 illustrates the bone composition of the wing of a hummingbird.
Ordinary birds articulate their wings at the shoulder,
wrist and elbow. During each wingbeat, their elbows and wrists will fold
and bend. Hummingbirds however, cannot articulate at the elbow and wrist,
but the wing is free to rotate in all directions at the shoulder joint.
Note in Figure 4.2, the upper set of bones is that of a hummingbird while
the lower set is that of a pelican. The juxtaposition of these two bone
structures exemplify the similarities and differences between ordinary
birds and hummingbirds. The hummingbird’s unusual arrangement of bone composition
results in notable flying characteristics. The exceptional hummingbird
is able to generate thrust from both the down beat and up beat of its wings.
In contrast, ordinary birds create the force needed to stay aloft only
by the down beat of their wings [5, pp. 204-6].
If a plot of wing-length to weight was made of all
natural flying creatures, one would get a graph as shown in Figure 4.3.
The figure displays the relationship between wing and total length for
the whole range of flying animals in logarithmic coordinates. The slope
is approximately 3, which means that the body-weight is proportional to
the cube of the wing length. Of course, it is expected that all points
will not lie on the average line. In fact, the hummingbird characteristics
lie close to the end of the bird region and are on the verge of the insect
region. Also, Figure 4.4 shows the relation between the wing beat rate
and the wing length. In this figure, we see that the insect or bird to
be modeled in the appropriate size range should have a wing beat close
to a hummingbird. [5, pp. 212-215]
4.1.2 Hovering Flight Characteristic
Figure 4.5 illustrates the figure "8" motif the wingtips
of the hummingbird trace in the air while hovering, as well as the wing
patterns at various positions. Notice the change in the pitch attitude
of the hummingbird as the speed of the bird changes from top speed to hovering.
The motion performed while hovering is illustrated in Figures 4.6 and 4.7.
As mentioned, the elbow and wrist do not move
relative to each other as seen in Figures 4.5 and 4.6. This motion is a
flight characteristic of the hummingbird. Also, notice the direction change
of the wing as it completes its cycle. It makes two 180 ° turns in
order to create the figure-8.
Technology: Hardware and Software Used in the Project
The following subsection lists the basic equipment
used in our research. It will describe the functions and reasons for using
the equipment. Also, a detailed discussion of the CT-Scanner and the MSC/NASTRAN
software will be presented. The first list includes the software used to
create the MSC/NASTRAN example. This list also contains material on how
to create the space-frame and the aeroelastic model of the hummingbird.
The second list presents the hardware used to film the hummingbird and
the computers on which the software was run.
The following list describes the software used
to create the space-frame and aerodynamic model of the hummingbird:
Adobe Photoshop: Scanned images were
corrected using Adobe Photoshop version 4.0 for better readability and
clarity of pictures and plots.
Adobe Premiere: Premiere was a graphics
tool utilized in creating individual pictures from a movie.
Apple Video: This software was used to
capture video footage from the VCR to the computer. It was the tool used
to create a computerized movie.
HpScan: Photographs and text in books were
scanned with the help of HpScan.
Kinetix 3-D Studio Max: This 3-D rendering
software enabled us to render the 3-D object and view it from any angle
Microsoft Word: Versions 95 and
97 were used to put together this report and other documents such as weekly
MSC/NASTRAN: Version 70 of this finite-elements
based modeling program will be used to perform aeroelastic analysis on
a hummingbird-wing model using the doublet-lattice method as well as a
Omnipage: Scanned text was converted from
bitmapped images onto ASCII format (text) with this optical character recognition
software. Some NASTRAN code was scanned and then converted to text for
comprehension of the aerodynamic segment of NASTRAN.
Omniview: The TIFF files were assembled
into a 3-Dimensional object with the aid of this program.
STL_UTIL: The Original 3-Dimensional object
was converted from the stereolithography to DXF format for easier manipulation
with this utility.
Visualization Toolkit: This program helped
us decimate the 3-Dimensional object to a smaller number of facits.
Windows 95, MacOS 8.0, Digital UNIX v 4.0:
These were the different operating systems used. MSC/NASTRAN was run on
the UNIX system. The MAC OS was the operating system running while the
film was transferred from the VCR to digital images. Windows 95 was the
base for scanning photographs and putting together the project and the
The following list describes the hardware used
to perform the filming and to use the software listed in the section above:
Computers: Macintosh, PCs, Digital DECs and
SGI machines were used.
Computer with Framegrabber: This computer
was connected to the VCR in order to transfer the necessary film or still
images from analog film format to digital format for pre-processing the
outline of the bird.
CT-Scanner: Computed Tomography is a completely
nondestructive technique for visualizing features in the interior of opaque
solid objects, and for obtaining digital information on their 3-D geometries
and properties. The scanner has not been used yet but hopefully will assist
us in determining a detailed bone composition of the hummingbird. See Figure
Digital Camera: A digital camera was used
to capture some images of the hardware used in the project such as the
hummingbird test setup.
High-Speed Video Camera, Tripod and Spotlights:
The High-Speed Video Camera belongs to the Department of Mechanical Engineering
and is under the supervision of Dr. Kenneth Ball. This camera has the ability
to record at the high rate of 2000 frames per second. A high-resolution
camera under the supervision of Dr. Clemens was used to conduct some filming.
One of the cameras was placed on the tripod and held still while the spotlights
illuminated the box and the bird.
Hummingbird Test Setup: This is where the
hummingbird was kept while being filmed. The box is made out of clear plastic
and it is therefore possible to film the bird through it. See Figures 5-1
The mirror assembly supported a mirror at 45 °
which enabled us to film the bird from the top. The hummingbird could then
be filmed from both the side and top views simultaneously. See Figure 2-7.
Ruby Throated Hummingbird: The hummingbird
is also included as hardware, since no experimental data would be available
Video Recorder: The VCR was connected to
the camera and registered the film onto a videotape using S-VHS recording.
MSC/NASTRAN is a finite element program developed
by the MacNeal-Schwendler Corporation. MSC/NASTRAN has very powerful analytical
capabilities. It is capable of analyzing stress, vibration, and heat transfer
characteristics of structures and mechanical components. [6, p. xiii] Finite
element modeling is a numerical method that is used for engineering analysis
and can be applied to all classes of field problems including structural
analysis, heat transfer, fluid flow, and electromagnetics. MSC/NASTRAN
will divide the structure into small elements (usually simple shapes like
triangles or rectangles), which form the model of the real structure. The
stiffness matrices for the individual elements are calculated. When combined,
these matrices form the stiffness matrix for the entire model. [7, p. 268]
The finite elements process in MSC/NASTRAN can be summarized by the following
The desired structure to be analyzed is input
into NASTRAN as a discretized mathematical model containing finite elements,
loads, constraints, and structural properties. Next, the computer running
MSC/NASTRAN will output any displacements, stresses, forces, mode shapes,
plots, or calculations. This data will be used to create graphs or tables
depending on the needs of the user. [6, p. 4]
How MSC/NASTRAN Works
MSC/NASTRAN needs an input file to begin its
run. The input file must contain the following in the order as presented
below [6, pp. 23-4]:
Aeroelastic Analysis Using MSC/NASTRAN
a. The Executive Control Statements
The primary function of this section
is to identify and determine the sort of analysis solution to be performed.
The CEND delimiter is always needed at the end of this section to
b. The Case Control Commands
c. The Bulk Data Entries
The purpose of this section is to specify and determine
the type of analysis output required as well as to control Bulk Data Input,
and select loads and boundary conditions.
This section is the last required entry and starts
with the BEGIN BULK delimiter. These entries include everything
about the model: size, geometry, coordinate systems, finite elements, loads,
boundary conditions, material properties, and element properties. This
section is ended with the ENDDATA delimiter.
An aerodynamic model beside the existent
structural model is needed to perform aeroelastic analyses in MSC/NASTRAN.
Since the structural gridpoints and the aerodynamic grid points will probably
not match, some splining techniques are used for both lines and surfaces.
These techniques are used to create the transformation matrix from structural
grid point deflections to aerodynamic grid point deflections. The aerodynamic
forces and moments at aerodynamic boxes are transferred to structural grid
points via the transpose of the transformation matrix. There are three
supersonic and one subsonic lifting surface aerodynamic theories. The subsonic
theory is called the Doublet-Lattice method, and is an extension of the
steady Vortex-Lattice method to unsteady flow. [8, p.1]
The Doublet-Lattice method will be used to perform
a number of analyses on the hummingbird-wing. All lifting surfaces are
presumed to lie almost parallel to the flow. Any interfering surface can
be analyzed, given that each is idealized as a set of trapezoidal planes.
[8, pp. 11,13]
Examples of Bulk Data Entries
Following are some basic bulk data entries
that will be used in both the space-frame and aerodynamic model of the
This command defines a geometric point in the
coordinate system. The directions of its displacement and its permanent
single-point constraints are entered. Each point is given an identification
number. [6, pp. 50-51]
CQUAD4 is a common element used to model structures.
The model is a quadrilateral flat plate connecting four grid points. [6,
The Membrane, transverse shear, bending, and coupling
properties of thin plate and shell elements such as CQUAD4 are entered
through the PSHELL entry. [n, 79-80]
This entry is the model for a basic beam that
can support tension and compression, bending, and shear in two perpendicular
planes. [6, p. 68]
CAERO1 is an aerodynamic entry and defines a trapezoidal
wing panel for the Doublet-Lattice or ZONA51 method. This is the basic
modeling tool for the aerodynamic model when using the Doublet-Lattice
method. [8, pp. 89,127]
This entry is used to identify associated interference
bodies in the subsonic case. This entry is required even if there are no
bodies defined. [8, pp. 89,127]
SET1 inputs the structural grid points to be splined
for the structural-aerodynamic splining. [8, pp. 104,128]
This card defines a surface spline. The spline
interconnects the aerodynamic boxes to the structural grid points selected
by the SET1 command.
[8, pp. 104,128]
Computed X-ray Tomography makes it possible
to examine the interior of an object without actually deforming or destroying
it. Two-dimensional images, called slices, illustrate the inside
of the object. These images are created by passing x-rays through the object
at many different angles. The difference in density of the object leads
to different x-ray absorption levels which create the contrast in the picture.
By making extremely thin, constant-thickness slices, one can produce a
detailed three-dimensional picture of the density variations of an object
by placing the slices on top of each other. On the next page, Figure 4.8
illustrates a CT-Scan picture composed of several different images. These
images were created depending on the preference of the user during the
MAV Technology and Science
Three main factors determine the resolution of the
CT image. These are the type and size of the x-ray source, the geometric
relations among the source, object and detectors, and the process used
to reconstruct the final image. The most modern instruments can have a
resolution of tens of m m (1 m
m = 10-6 m) per slice. The CT-Scanner at The University of Texas
at Austin has the capability to achieve such high resolutions. 
At the sizes envisioned for these devices,
normal aerodynamic rules no longer apply. Microflyers will have to operate
in an environment more common to small birds and large insects than that
of larger aircraft. The forces associated with air moving around the tiny
devices are more pronounced than with conventional aircraft in flight,
causing increased drag, reduced lift under the smaller wings at low speeds,
and decreased propeller efficiency. Such aircraft, weighing only 50 grams,
are more susceptible to wind gusts, updrafts, and rain.
Other challenges include developing tiny sensors,
engines, and power sources for such planes, as well as communications,
control and navigation systems for the tiny robot aircraft, which would
have to operate with little or no human input. Microflyers require an entirely
new approach to aircraft design and miniaturization. "We are not trying
to scale down conventional technology," says Robert Michelson of GTRI.
"We are scaling down new and unconventional technology ." Designing
the devices is not a simple task since making this aircraft requires considerably
more work than simply scaling down existing pilotless military drones or
making tinier replicas of radio-controlled model airplanes. "Nothing about
making micro air vehicles is going to be easy," said Dr. William R. Davis,
manager of the MAV program at the Massachusetts Institute of Technology's
Lincoln Laboratory. "With planes this small, all the rules change and everything
becomes challenging ." Building an aircraft smaller than 15cm is easy
enough, however, making them do something useful is the challenge.
The speed of the MAV is crucial so as to avoid
buffeting by gusts of wind that would make the craft useless. Aerodynamics
poses major problems for the MAV as well. As flying objects become smaller,
the viscosity of the air becomes increasingly important because for the
smallest insects, flying is more like swimming through honey. Microwings
are also susceptible to boundary layer separation. Small changes in the
angle of flight can result in extreme loss of lift, and in small aircraft
the boundary layer does not always reattach itself when the wing returns
to level flight.