This action might not be possible to undo. Are you sure you want to continue?
AIRCRAFT ANALYSIS SOFTWARE
BY
Benjamin Sweeten
Submitted to the graduate degree program in Aerospace Engineering
and the Graduate Faculty of the University of Kansas
in partial fulfillment of the requirements for the degree of
Master‟s of Science
_____________________________
Dr. Shahriar Keshmiri, Chairperson
Committee members* _____________________________*
Dr. Ray Taghavi
_____________________________*
Dr. Saeed Farokhi
_____________________________*
Dr. Richard Hale
Date Defended: April 27, 2010
ii
The Thesis Committee for Benjamin Sweeten certifies
that this is the approved Version of the following thesis:
CFD ANALYSIS OF UAVs USING VORSTAB, FLUENT, AND ADVANCED
AIRCRAFT ANALYSIS SOFTWARE
Committee:
_____________________________
Dr. Shahriar Keshmiri, Chairperson
_____________________________
Dr. Ray Taghavi
_____________________________
Dr. Saeed Farokhi
_____________________________
Dr. Richard Hale
Date approved:____________________
iii
Acknowledgements
I would like to thank NASA and the National Science Foundation for the
contributions to the University of Kansas Aerospace Engineering Department.
Without their support and funding, this research would have not taken place. Also,
thank you to Dr. Shahriar Keshmiri for his leadership and guidance during my
graduate and undergraduate career. Dr. Keshmiri helped in every aspect of my
research, from selecting classes, funding, and research topics. Thank you to Dr.
Richard Hale for his guidance on the Meridian UAV. He has high expectations of his
students, and is willing to put out the same amount of effort if not more. Thank you
to my professors that I have had throughout my education. They have given the
knowledge needed for my career path. Thank you to Dr. Lan, for allowing me to use
his program and giving me guidance during the learning process of this program.
Thank you to Andy Pritchard, the Aerospace Engineering Department‟s airplane and
power plant mechanic. Andy worked with me on many projects and research. He
offered his guidance and experience to everyone, and he was a major influence on my
graduate career. Thank you to all my fellow colleagues and students for their help.
In particular Dave Royer, Jonathan Tom, and Bill Donovan for answering and helping
with all questions that were asked of them. Without their hard work the Meridian
would have never flew. I would like to thank my family for offering their support
and anything I needed during my college career. Lastly, I would like to thank my
fiancée Jessica Donigan for her love and support. She dealt with me in a very
difficult time and has always been there for me. Thank you to everyone that I may
iv
have neglected to mention. There have been many people that gave me guidance and
support during my graduate and undergraduate career.
v
Abstract
The University of Kansas has long been involved in the research and
development of uninhabited aerial vehicles, UAVs. Currently a 1,100 lb UAV has
been designed, built, and flown from the University. A major problem with the
current design of these UAVs is that very little effort was put into the aerodynamics.
The stability and control derivatives are critical for the flight of the vehicle, and many
methods can be used to estimate them prior to flight testing. The topic of this
research is using high fidelity computer software, VORSTAB and FLUENT, to
determine the flying qualities of three different UAVs. These UAVs are the 1/3 scale
YAK54, the MantaHawk, and the Meridian. The results found from the high fidelity
computation fluid dynamics programs were then compared to the values found from
the Advance Aircraft Analysis, AAA, software. AAA is not considered to be as
accurate as CFD, but is a very useful tool for design. Flight test data was also used to
help determine how well each program estimated the stability and control derivative
or flying qualities.
The YAK54 results from both programs were very close to each other and
also to the flight test results. The results from the other two UAVs varied largely, due
to the complexity of the aircraft design. VORSTAB had a very difficult time
handling the complex body of the Meridian. Its results showed the aircraft was
unstable in several different modes, when this is known to not be the case after
several flight tests.
vi
From these results it was determined that VORSTAB, while a high fidelity
program, has difficulty handling aircraft with complex geometry. If the aircraft is a
traditional style aircraft with noncomplex geometry VORSTAB will return highly
accurate results that are better than AAA. The benefit of AAA is that a model can be
created rather quickly and the results will typically be within an acceptable error
range. A VORSTAB model can be very time consuming to make, and this can
outweigh the improved results. It is rather simple to determine if the VORSTAB
results are valid or not, and the input file can be easily improved to increase the
accuracy of the results. It is always a smart idea to use both software programs to
check the results with one another.
FLUENT was used to determine the possible downwash issue over the
Meridian fuselage. This software is a widely accepted program that is known to
produce very accurate results. The major problem is that it is very time consuming to
make a model and requires someone with a large amount of knowledge about the
software to do so. FLUENT results showed a possibility for a large boundary layer
near the tail and flow separation at high angles of attack. These results are all
discussed throughout the report in detail.
vii
Table of Contents
Page #
Acknowledgements ...................................................................................................... iii
Abstract ..........................................................................................................................v
List of Figures .............................................................................................................. ix
List of Tables ............................................................................................................. xiii
List of Symbols ............................................................................................................xv
1 Introduction .........................................................................................................1
2 Literature Review................................................................................................3
2.1 Wavelike Characteristics of Low Reynolds Number Aerodynamics ........... 3
2.2 A Generic Stability and Control Methodology For Novel Aircraft
Conceptual Design ..................................................................................................... 4
2.3 Theoretical Aerodynamics in Today‟s Real World, Opportunities and
Challenges ................................................................................................................. 5
2.4 The Lockheed SR71 Blackbird – A Senior Capstone ReEngineering
Experience ................................................................................................................. 6
3 Stability and Control Derivatives........................................................................8
3.1 Longitudinal Motion ..................................................................................... 8
3.2 LateralDirectional Motion ......................................................................... 14
3.3 Perturbed State ............................................................................................ 20
4 AAA ..................................................................................................................22
5 VORSTAB ........................................................................................................23
5.1 Creating a Model......................................................................................... 24
6 FLUENT ...........................................................................................................26
7 YAK54.............................................................................................................29
7.1 AAA Modeling of the YAK54 .................................................................. 32
7.2 VORSTAB Modeling of the YAK54 ........................................................ 35
7.3 Method Comparison YAK54..................................................................... 51
7.4 Linearized Model of the YAK54 ............................................................... 55
8 MantaHawk .......................................................................................................56
8.1 AAA Modeling of the MantaHawk ............................................................ 57
8.2 VORSTAB Modeling of the MantaHawk .................................................. 60
8.3 Method Comparison MantaHawk ............................................................... 73
8.4 Linearized Model of the MantaHawk ......................................................... 77
9 Meridian UAV ..................................................................................................78
9.1 AAA Modeling of the Meridian ................................................................. 80
9.2 VORSTAB Modeling of the Meridian ....................................................... 83
9.3 Linearized Model of the Meridian ............................................................ 135
9.4 FLUENT Modeling of the Meridian ......................................................... 135
9.5 FLUENT Model Generation ..................................................................... 141
9.6 Method Comparison Meridian .................................................................. 158
10 Conclusions and Recommendations ...............................................................165
11 References .......................................................................................................169
viii
Table of Contents Continued
Appendix A ................................................................................................................171
Appendix B ................................................................................................................172
ix
List of Figures
Figure 1: EarthFixed and BodyFixed Axes System ................................................... 9
Figure 2: YAK54 ....................................................................................................... 29
Figure 3: Unigraphics CAD Model of YAK54 ......................................................... 31
Figure 4: YAK54 VORSTAB Lift Curve Slope ....................................................... 36
Figure 5: YAK54 VORSTAB Drag Curve................................................................ 36
Figure 6: YAK54 VORSTAB Pitching Moment Coefficient Curve ......................... 37
Figure 7: YAK54 VORSTAB Longitudinal Derivatives .......................................... 37
Figure 8: YAK54 VORSTAB LateralDirectional Derivatives due to Sideslip ....... 39
Figure 9: YAK54 VORSTAB LateralDirectional Derivatives due to Roll Rate ..... 39
Figure 10: YAK54 VORSTAB LateralDirectional Derivatives due to Yaw Rate .. 40
Figure 11: YAK54 VORSTAB Sideforce Coefficient due to Aileron Deflection .... 42
Figure 12: YAK54 VORSTAB Rolling Moment Coefficient due to Aileron
Deflection .................................................................................................................... 42
Figure 13: YAK54 VORSTAB Yawing Moment Coefficient due to Aileron
Deflection .................................................................................................................... 43
Figure 14: YAK54 VORSTAB Lift Coefficient due to Elevator Deflection ............ 45
Figure 15: YAK54 VORSTAB Drag Coefficient due to Elevator Deflection .......... 45
Figure 16: YAK54 VORSTAB Pitching Moment Coefficient due to Elevator
Deflection .................................................................................................................... 46
Figure 17: YAK54 VORSTAB Sideforce Coefficient due to Rudder Deflection .... 48
Figure 18: YAK54 VORSTAB Rolling Moment Coefficient due to Rudder
Deflection .................................................................................................................... 48
Figure 19: YAK54 VORSTAB Yawing Moment Coefficient due to Rudder
Deflection .................................................................................................................... 49
Figure 20: MantaHawk ............................................................................................... 57
Figure 21: MantaHawk VORSTAB Lift Coefficient .................................................. 61
Figure 22: MantaHawk VORSTAB Drag Coefficient................................................ 61
Figure 23: MantaHawk VORSTAB Pitching Moment Coefficient ............................ 62
Figure 24: MantaHawk VORSTAB Lift Coefficient due to Pitch Rate ..................... 62
Figure 25: MantaHawk VORSTAB Pitching Moment Coefficient due to Pitch
Rate ............................................................................................................................. 63
Figure 26: MantaHawk VORSTAB LateralDirectional Derivatives due to
Sideslip ........................................................................................................................ 64
Figure 27: MantaHawk VORSTAB LateralDirectional Derivatives due to Roll
Rate ............................................................................................................................. 65
Figure 28: MantaHawk VORSTAB LateralDirectional Derivatives due to Yaw
Rate ............................................................................................................................. 65
Figure 29: MantaHawk VORSTAB Lift Coefficient due to Symmetrical Deflection 68
Figure 30: MantaHawk VORSTAB Drag Coefficient due to Symmetrical
Deflection .................................................................................................................... 68
x
List of Figures Continued
Figure 31: MantaHawk VORSTAB Pitching Moment Coefficient due to Symmetrical
Deflection .................................................................................................................... 69
Figure 32: MantaHawk VORSTAB Sideforce Coefficient due to Asymmetrical
Deflection .................................................................................................................... 71
Figure 33: MantaHawk VORSTAB Rolling Moment Coefficient due to
Asymmetrical Deflection ............................................................................................ 71
Figure 34: MantaHawk VORSTAB Yawing Moment Coefficient due to
Asymmetrical Deflection ............................................................................................ 72
Figure 35: Unigraphics CAD Model of the Meridian UAV ....................................... 79
Figure 36: Meridian Empennage Models 1 and 2 ....................................................... 86
Figure 37: Meridian VORSTAB Model 2 Lift Coefficient due to Angle of Attack... 88
Figure 38: Meridian VORSTAB Model 2 Drag Coefficient due to Angle of Attack. 88
Figure 39: Meridian VORSTAB Model 2 Pitching Moment Coefficient due to Angle
of Attack...................................................................................................................... 89
Figure 40: Meridian VORSTAB Model 2 Longitudinal Derivatives due to Pitch
Rate ............................................................................................................................. 89
Figure 41: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to
Sideslip ........................................................................................................................ 91
Figure 42: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to Roll
Rate ............................................................................................................................. 91
Figure 43: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to Yaw
Rate ............................................................................................................................. 92
Figure 44: Meridian VORSTAB Model 2 Sideforce Coefficient due to Aileron
Deflection .................................................................................................................... 95
Figure 45: Meridian VORSTAB Model 2 Rolling Moment Coefficient due to Aileron
Deflection .................................................................................................................... 95
Figure 46: Meridian VORSTAB Model 2 Yawing Moment Coefficient due to Aileron
Deflection .................................................................................................................... 96
Figure 47: Meridian VORSTAB Model 2 Lift Coefficient due to Elevator Input ..... 98
Figure 48: Meridian VORSTAB Model 2 Drag Coefficient due to Elevator Input ... 98
Figure 49: Meridian VORSTAB Model 2 Pitching Moment Coefficient due to
Elevator Input.............................................................................................................. 99
Figure 50: Meridian VORSTAB Model 2 Sideforce Coefficient due to Rudder
Input .......................................................................................................................... 101
Figure 51: Meridian VORSTAB Model 2 Rolling Moment Coefficient due to Rudder
Input .......................................................................................................................... 101
Figure 52: Meridian VORSTAB Model 2 Yawing Moment Coefficient due to Rudder
Input .......................................................................................................................... 102
Figure 53: Meridian VORSTAB Model 3 Lift Coefficient due to Angle of Attack. 103
Figure 54: Meridian VORSTAB Model 3 Drag Coefficient due to Angle of
Attack ........................................................................................................................ 104
xi
List of Figures Continued
Figure 55: Meridian VORSTAB Model 3 Pitching Moment Coefficient due to Angle
of Attack.................................................................................................................... 104
Figure 56: Meridian VORSTAB Model 3 Longitudinal Coefficients due to Pitch
Rate ........................................................................................................................... 105
Figure 57: Meridian VORSTAB Model 3 LateralDirectional Derivatives due to
Sideslip ...................................................................................................................... 106
Figure 58: Meridian VORSTAB Model 3 LateralDirectional Derivatives due to Roll
Rate ........................................................................................................................... 106
Figure 59: Meridian VORSTAB Model 3 LateralDirectional Derivatives due to Yaw
Rate ........................................................................................................................... 107
Figure 60: Meridian VORSTAB Model 3 Sideforce Coefficient due to Aileron
Deflection .................................................................................................................. 109
Figure 61: Meridian VORSTAB Model 3 Rolling Moment Coefficient due to Aileron
Deflection .................................................................................................................. 109
Figure 62: Meridian VORSTAB Model 3 Yawing Moment Coefficient due to Aileron
Deflection .................................................................................................................. 110
Figure 63: Meridian VORSTAB Model 3 Lift Coefficient due to Elevator Input ... 112
Figure 64: Meridian VORSTAB Model 3 Drag Coefficient at due to Elevator
Input .......................................................................................................................... 112
Figure 65: Meridian VORSTAB Model 3 Pitching Moment Coefficient due to
Elevator Input............................................................................................................ 113
Figure 66: Meridian VORSTAB Model 3 Sideforce Coefficient due to Rudder
Input .......................................................................................................................... 115
Figure 67: Meridian VORSTAB Model 3 Rolling Moment Coefficient due to Rudder
Input .......................................................................................................................... 115
Figure 68: Meridian VORSTAB Model 3 Yawing Moment Coefficient due to Rudder
Input .......................................................................................................................... 116
Figure 69: Meridian VORSTAB Model 4 Lift Coefficient due to Angle of Attack. 117
Figure 70: Meridian VORSTAB Model 4 Drag Coefficient due to Angle of
Attack ........................................................................................................................ 118
Figure 71: Meridian VORSTAB Model 4 Pitching Moment Coefficient due to Angle
of Attack.................................................................................................................... 118
Figure 72: Meridian VORSTAB Model 4 Longitudinal Derivatives due to Pitch
Rate ........................................................................................................................... 119
Figure 73: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to
Sideslip ...................................................................................................................... 120
Figure 74: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to Roll
Rate ........................................................................................................................... 121
Figure 75: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to
Yawing Moment ....................................................................................................... 121
Figure 76: Meridian VORSTAB Model 4 Sideforce Coefficient due to Aileron
Deflection .................................................................................................................. 124
xii
List of Figures Continued
Figure 77: Meridian VORSTAB Model 4 Rolling Moment Coefficient due to Aileron
Deflection .................................................................................................................. 124
Figure 78: Meridian VORSTAB Model 4 Yawing Moment Coefficient due to Aileron
Deflection .................................................................................................................. 125
Figure 79: Meridian VORSTAB Model 4 Lift Coefficient due to Elevator Input ... 127
Figure 80: Meridian VORSTAB Model 4 Drag Coefficient due to Elevator Input . 127
Figure 81: Meridian VORSTAB Model 4 Pitching Moment Coefficient due to
Elevator Input............................................................................................................ 128
Figure 82: Meridian VORSTAB Model 4 Sideforce Coefficient due to Rudder
Input .......................................................................................................................... 130
Figure 83: Meridian VORSTAB Model 4 Rolling Moment Coefficient due to Rudder
Input .......................................................................................................................... 130
Figure 84: Meridian VORSTAB Model 4 Yawing Moment Coefficient due to Rudder
Input .......................................................................................................................... 131
Figure 85: Meridian Fuselage Axial CrossSection for FLUENT ............................ 137
Figure 86: Meridian Fuselage with Farfield Divisions ............................................. 138
Figure 87: Meridian Farfield Meshes ....................................................................... 140
Figure 88: Meridian FLUENT Velocity Magnitude Profile for All Angle of
Attacks ...................................................................................................................... 145
Figure 89: Meridian FLUENT Velocity Angle Profile for All Angles of Attack .... 148
Figure 90: Velocity Profile in Boundary Layer with Favorable and Adverse Pressure
Gradient..................................................................................................................... 150
Figure 91: Velocity Profile with Flow Separation at Point S (dashed line u = 0) .... 151
Figure 92: Meridian FLUENT Total Pressure Profile for All Angles of Attack ...... 152
Figure 93: Meridian FLUENT Stream Function Profile at All Angles of Attack .... 154
Figure 94: Meridian FLUENT Vorticity Magnitude Profile at All Angles of
Attack ........................................................................................................................ 157
Figure 95: Circulation around an Airfoil Producing Lift .......................................... 158
xiii
List of Tables
Table 1: YAK54 Lifting Surface Dimensions ........................................................... 30
Table 2: YAK54 Flight Conditions ........................................................................... 31
Table 3: YAK54 AAA Moment of Inertia and Trimmed Values .............................. 32
Table 4: YAK54 AAA Longitudinal Derivatives...................................................... 33
Table 5: YAK54 AAA LateralDirectional Derivatives ............................................ 34
Table 6: YAK54 AAA Stability Requirements ......................................................... 35
Table 7: YAK54 VORSTAB Longitudinal Coefficient and Pitch Rate (1/rad) ........ 35
Table 8: YAK54 VORSTAB LateralDirectional Derivatives (1/rad) ...................... 38
Table 9: YAK54 VORSTAB LateralDirectional Due to Aileron Deflection
(1/rad) .......................................................................................................................... 41
Table 10: YAK54 VORSTAB Longitudinal Due to Elevator Deflection (1/rad) ..... 44
Table 11: YAK54 VORSTAB LateralDirectional Due to Rudder Deflection
(1/rad) .......................................................................................................................... 47
Table 12: YAK54 VORSTAB Stability Requirements ............................................. 50
Table 13: YAK54 AAA and VORSTAB Stability and Control Comparison ........... 52
Table 14: YAK54 Stability and Control Derivatives Typical Ranges ...................... 54
Table 15: MantaHawk AAA Longitudinal Derivatives .............................................. 58
Table 16: MantaHawk AAA LateralDirectional Derivatives .................................... 59
Table 17: MantaHawk AAA Stability Requirements ................................................. 59
Table 18: MantaHawk VORSTAB Longitudinal Derivatives (1/rad) ........................ 60
Table 19: MantaHawk VORSTAB LateralDirectional Derivatives (1/rad) .............. 64
Table 20: MantaHawk VORSTAB Longitudinal Derivatives due to Symmetrical
Deflection .................................................................................................................... 67
Table 21: MantaHawk VORSTAB LateralDirectional Derivatives due to
Asymmetrical Deflections .......................................................................................... 70
Table 22: MantaHawk VORSTAB Stability Requirements ....................................... 73
Table 23: MantaHawk AAA and VORSTAB Comparison ........................................ 74
Table 24: MantaHawk Stability and Control Derivatives Typical Ranges ................. 76
Table 25: Meridian Characteristics ............................................................................. 79
Table 26: Meridian AAA Longitudinal Derivatives ................................................... 81
Table 27: Meridian AAA LateralDirectional Derivatives ......................................... 82
Table 28: Meridian AAA Stability Requirements ...................................................... 83
Table 29: Meridian VORSTAB Model 1 Longitudinal Derivatives (1/rad) ............... 84
Table 30: Meridian VORSTAB Model 1 LateralDirectional Derivatives (1/rad) ..... 84
Table 31: Fuselage Diameter for Meridian Models 1 and 2 ....................................... 86
Table 32: Meridian VORSTAB Model 2 Longitudinal Control Derivatives and
Affected by Pitch Rate ................................................................................................ 87
Table 33: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to
Sideslip and Roll Rate ................................................................................................. 90
Table 34: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to Yaw
Rate ............................................................................................................................. 92
xiv
List of Tables Continued
Table 35: Meridian VORSTAB Model 2 Lateral Directional Derivatives due to
Aileron Deflection ...................................................................................................... 94
Table 36: Meridian VORSTAB Model 2 Longitudinal Derivatives Affected by
Symmetrical Tail Deflections ..................................................................................... 97
Table 37: Model 2 Lateral Directional Control Derivatives Affected by Asymmetrical
Tail Deflection .......................................................................................................... 100
Table 38: Meridian VORSTAB Model 3 Longitudinal Derivatives at due to Pitch
Rate ........................................................................................................................... 103
Table 39: Meridian VORSTAB Model 3 LateralDirectional Derivatives .............. 105
Table 40: Meridian VORSTAB Model 3 LateralDirectional Derivative due to
Aileron Deflection .................................................................................................... 108
Table 41: Meridian VORSTAB Model 3 Longitudinal Derivatives due to
Symmetrical Tail Deflections ................................................................................... 111
Table 42: Meridian Model 3 LateralDirectional Derivatives Affected by
Asymmetrical Tail Deflections ................................................................................. 114
Table 43: Meridian VORSTAB Model 4 Longitudinal Derivatives ........................ 117
Table 44: Meridian VORSTAB Model 4 LateralDirectional Control Derivatives . 120
Table 45: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to
Aileron Deflection .................................................................................................... 123
Table 46: Meridian VORSTAB Model 4 Longitudinal Derivatives due to
Symmetrical Tail Deflection ..................................................................................... 126
Table 47: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to
Asymmetrical Deflection .......................................................................................... 129
Table 48: Meridian VORSTAB Stability and Control Derivatives Dr. Roskam‟s
Typical Ranges.......................................................................................................... 133
Table 49: Meridian AAA and VORSTAB Comparison ........................................... 134
Table 50: Meridian Farfield Edges ........................................................................... 139
Table 51: Meridian Edge Meshes ............................................................................. 139
Table 52: Meridian Fuselage CrossSection FLUENT Lift and Drag ...................... 143
Table 53: Meridian Fuselage Lift and Drag Method Comparison ............................ 159
Table 54: Meridian AAA and VORSTAB Comparison ........................................... 161
Table 55: Meridian VORSTAB and AAA Stability and Control Derivatives Typical
Ranges ....................................................................................................................... 163
Table 56: YAK54 Longitudinal Modes ................................................................... 173
Table 57: YAK54 LateralDirectional Modes ......................................................... 173
Table 58: MantaHawk Longitudinal Modes ............................................................. 175
Table 59: MantaHawk LateralDirectional Modes ................................................... 176
Table 60: Meridian Model 2 Longitudinal Modes .................................................... 177
Table 61: Meridian Model 2 LateralDirectional Modes .......................................... 178
Table 62: Meridian Model 3 Longitudinal Modes .................................................... 179
Table 63: Meridian Model 3 LateralDirectional Modes .......................................... 180
xv
List of Symbols
Symbol Description Units
Normal
a ..................................................... Speed of Sound ..................................................... ft/sec
A or AR ........................................... Aspect Ratio ......................................................... 
b ......................................................... Wing Span .............................................................. ft
c̃ ............................................... Mean Geometric Chord .................................................... ft
c
r
....................................................... Root Chord .............................................................. ft
c
t
........................................................ Tip Chord ............................................................... ft
C
D
................................................. Drag Coefficient ...................................................... 
C
Do
....................... Variation of Airplane Drag with Angle of Attack .......................... 1/rad
C
Doe
................... Variation of Airplane Drag with Elevator Deflection ........................ 1/rad
C
Du
............................... Variation of Airplane Drag with Speed .................................. 1/rad
C
Dq
........................... Variation of Airplane Drag with Pitch Rate ............................... 1/rad
C
l
.......................................... Rolling Moment Coefficient ............................................. 
C
l
............. Variation of Airplane Rolling Moment with Angle of Sideslip ................ 1/rad
C
loa
........... Variation of Airplane Rolling Moment with Aileron Deflection ............... 1/rad
C
lor
........... Variation of Airplane Rolling Moment with Rudder Deflection ............... 1/rad
C
lp
..... Variation of Airplane Rolling Moment with Rate of Change of Roll Rate ....... 1/rad
C
lr
..... Variation of Airplane Rolling Moment with Rate of Change of Yaw Rate ....... 1/rad
C
L
................................................... Lift Coefficient ....................................................... 
C
L_ih
............ Variation of Airplane Lift with Differential Stabilizer Angle ................. 1/rad
C
Lo
........................ Variation of Airplane Lift with Angle of Attack ........................... 1/rad
C
Loe
.................... Variation of Airplane Lift with Elevator Deflection ......................... 1/rad
C
Lq
............................ Variation of Airplane Lift with Pitch Rate ................................ 1/rad
C
Lu
................... Variation of Airplane Lift with Dimensionless Speed ....................... 1/rad
C
m
....................................... Pitching Moment Coefficient ............................................ 
C
m_ac
.............. Pitching Moment Coefficient about Aerodynamic Center ...................... 
C
mo
............. Variation of Airplane Pitching Moment with Angle of Attack ................ 1/rad
C
moe
......... Variation of Airplane Pitching Moment with Elevator Deflection .............. 1/rad
C
mq
................. Variation of Airplane Pitching Moment with Pitch Rate ..................... 1/rad
C
mT
................... Variation of Airplane Pitching Moment due to Thrust ......................... 
C
mTo
. Variation of Airplane Pitching Moment due to Thrust with Angle of Attack .... 1/rad
C
n
......................................... Yawing Moment Coefficient ............................................ 
C
n
............. Variation of Airplane Yawing Moment with Angle of Sideslip ................ 1/rad
C
noa
........... Variation of Airplane Yawing Moment with Aileron Deflection ............... 1/rad
C
nor
........... Variation of Airplane Yawing Moment with Rudder Deflection ............... 1/rad
C
np
.... Variation of Airplane Yawing Moment with Rate of Change of Roll Rate ....... 1/rad
C
nr
.... Variation of Airplane Yawing Moment with Rate of Change of Yaw Rate ....... 1/rad
C
nT
....... Variation of Airplane Yawing Moment due to Thrust with Sideslip ............ 1/rad
C
Tx
........................... Variation of Airplane Thrust in X Direction................................. 
C
Txu
............. Variation of Airplane Thrust in Xaxis with respect to speed ................... 
C
y
............................................... Sideforce Coefficient .................................................. 
xvi
Normal Continued
C
y
................... Variation of Airplane Sideforce with Angle of Sideslip ...................... 1/rad
C
yoa
................. Variation of Airplane Sideforce with Aileron Deflection..................... 1/rad
C
yor
................. Variation of Airplane Sideforce with Rudder Deflection ..................... 1/rad
C
yp
.......... Variation of Airplane Sideforce with Rate of Change of Roll Rate ............. 1/rad
C
yr
.......... Variation of Airplane Sideforce with Rate of Change of Yaw Rate ............. 1/rad
dc/do.......................................... Downwash Gradient ................................................... 
do/d ........................................... Sidewash Gradient .................................................... 
D ............................................................. Drag ................................................................. lbs
e ............................................ Oswald‟s Efficiency Factor ............................................. 
F
y
........................................................ Sideforce .............................................................. lbs
h ............................................................ Height .................................................................. ft
i
h
....................................... Horizontal Tail Incidence Angle........................................... deg
I
xx
, I
yy
, I
zz
.................... Airplane Moments of Inertia about XYZ ............................ slugsft
2
I
xz
................................ Airplane Products of Inertia about XYZ ............................. slugsft2
L .............................................................. Lift .................................................................. lbs
L, boundary layer ................. Flow Distance Along Body.................................................. ft
N ................................................... Yawing Moment .................................................... ftlbs
N.D. .............................................. Nondimensional...................................................... 
M .................................................. Pitching Moment ................................................... ftlbs
M
∞
......................................... Free Stream Mach Number .............................................. 
p .......................................................... Roll Rate ....................................................... rad/sec
P
s
..................................................... Static Pressure ..................................................... lbs/ft
2
P
T
.................................................... Total Pressure ..................................................... lbs/ft
2
q ......................................................... Pitch Rate ...................................................... rad/sec
q =0.5pV
2
............................. Aircraft Dynamic Pressure ........................................... lbs/ft
2
r........................................................... Yaw Rate ...................................................... rad/sec
u .............................................. xcomponent of Velocity ............................................. ft/sec
U ........................................... Axial Free Stream Velocity ........................................... ft/sec
v .............................................. ycomponent of Velocity ............................................. ft/sec
V .......................................................... Velocity........................................................... ft/sec
V ........................................... Volume Coefficient of Tail .............................................. 
w ............................................. zcomponent of Velocity ............................................. ft/sec
S .............................................................. Area................................................................... ft
2
Re ................................................ Reynolds Number ..................................................... 
x .......................................... Axial Station Along Fuselage ................................................ ft
x
ac
...................Distance from Aerodynamic Center and Reference Point .......................... ft
x
cg
.................... Distance from Center of Gravity and Reference Point ............................. ft
x
vs
................. Axial Distance Between Vertical Tail a.c and Airplane c.g. ........................ ft
2d ................................................. Two Dimensional ..................................................... 
2ddp ............................... Two Dimensional Double Precision ....................................... 
3d ................................................ Three Dimensional .................................................... 
3ddp .............................. Three Dimensional Double Precision ...................................... 
xvii
Greek
o .................................................... Angle of Attack........................................................ deg
o_dot .............................. Rate of Change of Angle of Attack ................................ deg/sec
2
 ........................................................... Sideslip .............................................................. deg
õ ..................................... Average Boundary Layer Thickness....................................... 
o* ............................................ Displacement Thickness ................................................ 
o
e,a,el,r
..................................... Control Surface Deflection ............................................... deg
A .........................................................Change in ........................................................... 
n ............................................. Dynamic Pressure Ratio ................................................ 
I ........................................................ Circulation ...................................................... m
2
/sec
ì ........................................................ Taper Ratio .......................................................... 
v ................................................. Kinematic Viscosity ............................................... m
2
/sec
t .............................................................. 3.14 ................................................................ 
p ........................................................... Density .......................................................... lbs/ft
3
t .................................... Angle of Attack Effectiveness Factor ...................................... 
u ................................................ Momentum Thickness.................................................. 
e ................................................... Angular Velocity ................................................ rad/sec
v ................................................... Stream Function ...................................................... 
Acronyms
AAA .................................. Advanced Aircraft Analysis
CAD ..................................... Computer Aided Design
CFD ................................ Computational Fluid Dynamics
CReSIS .................... Center for Remote Sensing of Ice Sheets
KUAE ................. University of Kansas Aerospace Engineering
NSF .................................. National Science Foundation
PC ............................................. Personal Computer
TAS .............................................. True Airspeed
UAV .................................. Uninhabited Aerial Vehicle
Subscripts
a ......................................................... Aileron
B ........................................ Body Stability Axis System
e ......................................................... Elevator
el ......................................................... Elevon
f ........................................................ Fuselage
h................................................... Horizontal Tail
o....................................... Coefficient at Zero Deflection
r .......................................................... Rudder
v..................................................... Vertical Tail
w .......................................................... Wing
wf ................................................ Wing Fuselages
1............................................. Steady State Condition
1
1 Introduction
The shape and design of an aircraft can dramatically influence how the aircraft
handles and is controlled. The stability and control derivatives are essential for flight
simulation and handling qualities. There are several equations that can be used to
estimate some of the derivatives, but not all of them. These equations are just
estimations and can be magnitudes off. Therefore, better methods have to be used
before investing millions of dollars on an aircraft.
Wind tunnel tests, are a method that results in derivatives that are highly
accurate. The problems with wind tunnel tests are that it is very expensive and can be
very time consuming. Also, the wind tunnel models are scaled down to fit in the
tunnel, and this can have a dramatic change on the results, since the results do not
always scale up as easily. Air will flow over a smaller body differently than a larger
body, due to the changes in Reynolds number and other flow characteristics. Using
an experienced wind tunnel expert and a highly accurate tunnel can minimize these
problems, but will be very expensive. Over the past couple of decades computer
simulation has become much more prevalent. Computational Fluid Dynamic
software is much more accurate than it once was and is becoming more user friendly,
but it still requires an expert to create a 3D full aircraft CFD model. The mesh
generation for a model can be difficult and requires a great deal of experience. This
software is expensive to purchase, but can be used over and over again. Also, many
different test cases can be run to determine flying qualities in various situations.
There are also several different programs that are readily available that can produce
2
high fidelity results, and some of these programs can be purchased at a reasonable
price.
Three different computer programs were used to determine the stability and
control derivatives on three different UAVs. Two of the aircrafts were being
designed and built to fly while the other one was already a production aircraft. The
1/3 scaled YAK54 model was a production aircraft purchased by the University of
Kansas, and the Meridian and MantaHawk were designed at the University of Kansas.
The Meridian is a 1,100 lb aircraft that was designed to fly in the Polar Regions.
Advanced Aircraft Analysis (AAA) and VORSTAB were used on all three aircraft
and FLUENT was also used on the Meridian. FLUENT is a very high fidelity CFD
program, but requires a large amount of experience and time. High level CFD
programs can be very expensive and time consuming when performing aerodynamic
analysis. This is why engineers prefer to use engineering level programs, such as
AAA, to generate the derivatives quickly. The main goal of this research was to use
high fidelity CFD programs to test the validity of these engineering level programs.
The stability and control derivatives found from each software program were
compared to each other and conclusions about the software were drawn.
3
2 Literature Review
It is a wise idea to examine current and past research going on in the field of
study. This gives the researcher a chance to see what is currently going on, or has
previously been examined in the past. It can also give the researcher ideas on topics
and experiments to conduct. The research that was conducted in this paper is
aerodynamic analysis, using high fidelity CFD programs, to determine the stability
and control derivatives of UAVs with low Reynolds numbers. Therefore, the
literature review topics consisted of stability and control analysis software, low
Reynolds number aerodynamics, and the CFD software that was used in this research.
A brief summary of each article will be given and then the conclusion drawn from
these papers.
2.1 Wavelike Characteristics of Low Reynolds Number Aerodynamics
Lifting surfaces will demonstrate several uncharacteristic flow patterns when
flown at a low Reynolds number. These patterns include a drag increase greater than
the rate of increasing lift, acoustic disturbances, and variances in drag across the span
of the lifting surface. This has a great effect on the design of micro or small UAVs
since they have rather small Reynolds numbers. Spanwise flow can usually be
ignored at higher velocities, but due to the low Reynolds number the flow can travel
in the spanwise direction. Using velocity potential theory, boundary layer theory, and
sinusoidal wave theory, the drag variation can be modeled as a sinusoidal wave along
the span. These results were then comparedt to the research conducted by Guglielmo
4
and Selig, where the drag magnitude was observed to be happening in a wave form
along the span. The goal of this research was not to exactly match the Guglielmo and
Selig data, but to demonstrate that the drag magnitude and flow can be modeled using
sinusoidal wave theory. This goal was successfully accomplished even though it did
not match the trend observed by Guglielmo and Selig. All of this can be found in
detail in Ref [1].
This research shows how the low Reynolds number can affect the flow around
the aircraft‟s lifting surfaces. At low speeds the drag magnitude can vary along the
span of the lifting surface, and in turn this can dramatically affect the other stability
and control of the aircraft. If the flow is traveling at different speeds and in different
directions (spanwise) the aircraft will not react how it typically would at higher
speeds. The control surfaces would not have the same impact when the flow is
varying.
2.2 A Generic Stability and Control Methodology For Novel Aircraft
Conceptual Design
Stability and control is the most serious requirement for flight safety, and yet
there is not a standard or reliable method for determining stability and control in the
design phase. There are several methods used and many are considered acceptable
within the industry. A major weakness, of most methods, is the design and sizing of
the control effectors. Currently very simple methods are used for the sizing, and are
done so in the cruise, landing, and takeoff conditions of the flight envelope. This
5
research shows that sizing should be done so in the grey areas of the flight envelope,
where nonlinear aerodynamics prevail.
A method for generating stability and control was designed over a four year
period and is called AeroMesh. This method is capable of handling both
conventional and unconventional and symmetric or asymmetric flight vehicles.
Design constraints and various flight conditions are first implemented into the
program. An input file is then created for a CFD program called VORSTAB. This
CFD software will estimate the stability and control derivatives as well as determine
the size, position, and hinge lines of the control effectors. A 6 degreeoffreedom
model is then used to determine stability and control in the trimmed and untrimmed
condition. This 6DOF model uses control power to determines the stability and
control derivatives in the trimmed and untrimmed conditions. All information was
taken from Ref [2].
This research shows just how important a high fidelity CFD program can be
do the design of the control effectors and their sizing. The design criteria are at the
extremes of the flight envelope, so the control effectors are designed at the point of
nonlinear aerodynamics. Using VORSTAB can help eliminate the use of simple
methods that are low fidelity.
2.3 Theoretical Aerodynamics in Today’s Real World, Opportunities and
Challenges
CFD has revolutionized the aerodynamic industry, but it still faces many
challenges in predicting and controlling various flows. These flows include UAV
6
low Reynolds number, high angle of attack, boundary layer transition, three
dimensional separation, and others. Since this is the case it is wise to combine
theoretical, computational, and experimental approaches when analyzing the flow.
These problems that a typical CFD program, has with the flow, is discussed and ways
to analytically solve these problems are given. Multiple approaches are applied to the
flow to find solutions. Using all three of the solution methods allows the users to see
the short comings of each method. It is a very important and critical skill set to know
and understand how to set up a problem up from the beginning, and then make
approximations using mathematical and physicsbased models. This principle should
then be applied to a modern computational method. All information was taken from
Ref [3].
It can be seen that not only a CFD program should be used during the design
process, but also other methods. The research in this report covers both analytical
methods, AAA, and high fidelity methods, VORSTAB and FLUENT. Understanding
how to set up problems is very important due to the high complexity of modern CFD
programs. A small error in the input can dramatically influence the results. It is also
very important for the user to be able to interpret the results, and this skill set comes
from understanding the theoretical methods.
2.4 The Lockheed SR71 Blackbird – A Senior Capstone ReEngineering
Experience
At the University of Texas at Arlington, the senior aerospace class re
engineered the Lockheed SR71 Blackbird in a two part design course. Currently in
7
the aerospace industry, it is very rare for a company to start a design from scratch, but
rather add to or modify previous research.
There were no changes made to the SR71 model, but rather the aircraft was
reanalyzed. A CAD model of the aircraft was created, and from there the stability
and control derivatives were determined in a two different ways. VORSTAB was the
primary method for obtaining the derivatives, but also the Dr. Roskam method was
used. The Roskam method is outlined in an eight book series about aircraft design.
This second method is the same as using the AAA software. Both methods
derivatives were then compared to the actual SR71 data. The VORSTAB results
were off by an order of magnitude, but followed the correct trends with the exception
of the yawing moment coefficient due to sideslip and yawing moment coefficient due
to roll rate. All other aspects of the design process were completed ranging from
aircraft systems to flight performance. All information was taken from Ref [4].
The reengineering of the SR71 by the senior design class at the University of
Texas at Arlington is very similar to the topic of this thesis. Several methods of
analysis were used to determine the stability and control derivatives, and the results
were then compared to one another.
8
3 Stability and Control Derivatives
The stability and control derivatives come from the aerodynamic forces and
moments acting on upon the aircraft components. These components are defined as
the wings, tails, fuselage, and any other surface on the aircraft. The flow around an
entire aircraft is too complex to allow formulas to determine the derivatives. Wind
tunnel tests or high fidelity computational fluid dynamics should be used to estimate
the control derivatives with a high level of accuracy. To define and understand the
stability and control derivatives one must have a basic understanding of aerodynamic
principles; it will be assumed that the reader has this basic knowledge. The aircraft
forces and moments are broken into two distinct directional motions, longitudinal
motion and lateraldirectional motion. Coupling between longitudinal and lateral
directional dynamics is assumed to be zero for stability and control derivative
estimation.
3.1 Longitudinal Motion
Longitudinal motion is due to the forces and moments of the aircraft in the xz
plane in terms of the stability axes system, Figure 1 from Ref [5]. These forces and
moments include drag force, lift force, and pitching moment. The aircraft orientation
also affects these three, as well as the deflection of control surfaces. These forces and
moment will be discussed in subsections of this chapter.
9
Figure 1: EarthFixed and BodyFixed Axes System (From Ref [5])
3.1.1 Drag Coefficient, C
D
Drag is the force that acts in the opposite direction of motion. There are two
types of drag that contribute to the entire aircraft drag. Parasite drag is drag due to
the shape of the aircraft when there is zero lift produced, and induced drag is the drag
produced due to the production of lift. The following are influences on the drag:
airplane wetted area, skin friction, angle of attack, control surface deflection, speed,
and dynamic pressure. Eq [1] and Eq [2] are two equations for determining drag, Ref
[5] and Ref [6] respectively.
 Eq [1]
 Eq [2]
10
3.1.2 Change in Airplane Drag due to Change in Angle of Attack, C
Do
This coefficient represents variation of the drag coefficient with angle of
attack. Typically, drag increases as the angle of attack increases, or in other terms,
the drag changes as the angle of attack moves away from the steady state condition.
It also increases as the Mach number increases. This derivative, C
Do
, can be
estimated by differentiating Eq [2] which then produces Eq [3], Ref [6].
 Eq [3]
3.1.3 Change in Airplane Drag due to Change in Elevator Deflection, C
Doe
Deflection of the elevator produces an incremental change in aircraft drag,
described by the derivative C
Doe
. This derivative is determined at an angle of attack
of zero. It is acceptable to neglect the change in drag due to these control surface
deflections in low speeds, but high fidelity computer simulations will still calculate
this derivative. However, when trim drag is important, this increase in drag cannot be
neglected. The elevator deflection might be used to trim the aircraft, and therefore it
is important for the overall drag.
3.1.4 Lift Coefficient, C
L
The lift is defined as the force acting on a surface that is perpendicular to
oncoming flow in the upward direction. This means that it is also perpendicular to
the drag force vector. Lift is what keeps the aircraft in the air. The aircraft wings,
fuselage, and tails all can produce lift. The fuselage does not produce much lift, and
is usually found as a wingfuselage combination. The horizontal tail is affected by
11
the fuselage boundary layer, propeller slip stream, or jet exhaust. Therefore, when
determining the horizontal tail lift, the downwash effect has to be accounted for. The
following equations can be used to determine the overall lift of the aircraft, calculated
for zero angle of attack, Ref [5].
 Eq [4]
 Eq [5]
The wingfuselage combination and horizontal tail lift coefficients can be broken into
components. Methods for estimating these lift coefficient values can be found in Ref
[7].
3.1.5 Change in Airplane Lift due to Change in Angle of Attack, C
Lo
Similar to the drag, the lift is affected by the angle of attack of the aircraft. As
the angle of attack increases, the aircraft lift will increase until the stall point. This
allows the aircraft to climb. Once a certain angle of attack is reached, the lift will
begin to decrease because the aircraft has reached a stall point. The angle of attack of
the stall point is dependent upon the geometry of the aircraft, the flow, and the
altitude. This is why an aircraft has a maximum angle of attack. As the subsonic
speed increases, the C
Lo
increases, but around transonic region it begins to decrease.
The angle of attack also has an effect on every lifting surface. At certain angles of
attack, the downwash or flow off the wings could blanket out the tail and render them
ineffective. Eq [6] can be used to determine the lift coefficient due to angle of attack,
Ref [5]. Eq [6] is found by partial differentiation of the firstorder Taylor series of
the lift Eq [7], Ref [5].
12
 Eq [6]
 Eq [7]
3.1.6 Change in Airplane Lift due to Change in Elevator Deflection, C
Loe
The deflection of the elevator will change the camber of the horizontal tail
airfoils and therefore change the lift of those airfoils. Highly cambered airfoils
usually have higher lift. Therefore, depending on the camber direction of the
horizontal tail and the direction of deflection, the lift of the horizontal tail will
increase or decrease. The effect of the elevator deflection on the total aircraft lift
coefficient can be found in Eq [8], Ref [5]. This is also found by partial
differentiation of the firstorder Taylor series of lift Eq [7], Ref [5].
 Eq [8]
3.1.7 Pitching Moment Coefficient, C
m
This is defined as the aerodynamic force that creates a moment that causes the
aircraft to pitch, or rotate upwards and downwards. Lifting forces create this resultant
force that causes the aircraft to pitch. The point that the aircraft rotates about is
typically defined as the center of gravity. Center of pressure is defined as the point at
which the pitching moment coefficient is equal to zero. Aerodynamic center is the
point about which the pitching moment coefficient does not vary with angle of attack.
As the angle of attack changes the center of pressure location will change. These two
points create the pitching moment coefficient; the center of pressure location changes
cause the change in the pitching moment. Most aircraft are inherently stable as long
13
as the center of gravity is ahead of the aerodynamic center. Elevator deflection and
angle of attack can dramatically change the pitching moment. The following
equations define the pitching moment coefficient and estimation of this value, Ref
[5].
 Eq [9]

Eq [10]
3.1.8 Change in Airplane Pitching Moment due to Change in Angle of Attack, C
mo
As stated previously, the center of pressure can move forward and aft as the
angle of attack changes. This results in a changing moment arm and an increasing or
decreasing pitching moment. The angle of attack also changes the lift on the aircraft,
and therefore changes the aerodynamic force that creates the pitching moment. This
derivative, C
mo
. is called the static longitudinal stability derivative which should be
negative for an inherently longitudinally stable aircraft. For example, if the aircraft
that is statically stable is pitched upward it naturally returns to steady state and
pitches down or viceversa if pitched downward. If it was not stable, the aircraft
would want to continue pitching upward and could flip over. The horizontal tail has a
large affect on this since it is used to pitch the aircraft. Horizontal tail incidence
angle can dramatically affect this derivative due to the lift it creates on the tail. Eq
[11] can be used to estimate this derivative, Ref [5].
14

Eq [11]
3.1.9 Change in Pitching Moment due to Change in Elevator Deflection, C
moe
This derivative is referred to as the longitudinal control power derivative and
is typically negative. The effectiveness of the elevator is basically due to the volume
coefficient of the horizontal tail, Eq [12], and the angle of attack effectiveness of the
elevator, t, Ref [5]. The larger the size of the elevator is, the more effect it has on the
pitching moment. For example, a fully moving horizontal tail has just as much effect
as the incidence of the horizontal tail. Eq [13] is used to estimate the derivative, Ref
[5].
 Eq [12]
 Eq [13]
3.2 LateralDirectional Motion
The rolling motion is referred to as the lateral motion, and the yawing motion
is referred to as the directional motion. These two motions are results of control
surface deflections and sideforces, where sideslip plays a large role in lateral
directional motion. This is the angle of directional rotation from the aircraft
centerline to the direction of the wind. The sideslip angle can be thought of as the
directional angle of attack. The forces and moments that are defined in the lateral
15
directional motion are sideforce, yawing moment, and rolling moment. Similar to
longitudinal control, there are several variables that affect these forces and moments.
3.2.1 Rolling Moment, C
l
The rolling moment is the aircraft‟s rotation about the xaxis in the stability
coordinate system. Several different things can cause and influence the rolling
moment, and those are sideslip, angle of attack, the moment reference center (usually
center of gravity), deflection of control surfaces, and airspeed. The control surfaces
that affect the rolling moment are the aileron (lateral control surface) and rudder
(directional control surface). Elevator deflection influence can usually be ignored
since the deflections are symmetrical and theoretically cancel each other out. Eq [14]
is the dimensional form of the rolling moment and Eq [15] shows the first order
Taylor series form of the rolling moment, Ref [5].
 Eq [14]
 Eq [15]
3.2.2 Change in Airplane Rolling Moment due to Change in Sideslip, C
l
This derivative is often referred to as the airplane dihedral effect. The reason
for this is because the airplane dihedral angle can have a huge influence on the rolling
moment especially when at a sideslip. If the aircraft is at a sideslip and has a dihedral
angle on the wings, one of the wings will be hit with more air than the other. This
will cause a higher lift on that wing and in turn cause the airplane to roll. Rolling
moment derivative due to sideslip can be estimated by summing the dihedral effect of
16
the individual components of the aircraft, Eq [16], Ref [5]. There are many factors
that play into the individual components‟ dihedral effect. For example, the wings‟
location on the fuselage can affect the direction that the aircraft will want to roll. The
vertical tail will also see a higher sideforce when the aircraft is at a sideslip. For a
detailed explanation and ways to estimate the individual components of the dihedral
effect refer to Ref [5].
 Eq [16]
3.2.3 Change in Airplane Rolling Moment due to Change in Aileron Deflection, C
loa
A positive aileron deflection is defined as the right aileron up and the left
aileron down. This produces a rolling moment by decreasing the lift on the right
wing due to the negative camber of the aileron, and increasing the lift on the left wing
due to the positive camber of the aileron. These aileron deflections will also produce
a yawing moment, and this is why most ailerons are deflected differentially. This
differential deflection will help minimize the yawing moment that is produced. Flow
separation can also occur with large aileron deflections, and this can reduce the
effectiveness of the ailerons.
3.2.4 Change in Airplane Rolling Moment due to Change in Rudder Deflection, C
lor
The purpose of the rudder is to produce a yawing moment, but due to the
typical location of the vertical tail and rudder a rolling moment is produced. With the
rudder deflected, the free stream air will encounter the rudder and produce a
sideforce. The resultant sideforce is typically located above the center of gravity and
17
will produce this rolling moment. This derivative is usually positive, but at high
angles of attack it can switch signs because of the vertical tail moment arm location
changes. Eq [17] shows how to estimate this derivative, Ref [5].
 Eq [17]
3.2.5 Sideforce Coefficient, C
y
This is the aerodynamic force that causes the aircraft to yaw and can cause a
rolling moment if above or below the center of gravity. With zero angle of attack,
sideslip, and control surface deflection, the sideforce should equal zero for a
symmetrical aircraft. The sideforce is a result of sideslip, angle of attack, control
surface deflection, and symmetry of aircraft. For an unsymmetrical aircraft, a
sideforce could be produced from the side that has a larger amount of surface area
being hit by the free stream air. Eq [18] calculates the dimensional sideforce and Eq
[19] shows the first order Taylor series, Ref [5].
 Eq [18]
 Eq [19]
3.2.6 Change in Airplane Sideforce due to Change in Sideslip, C
y
Similar to the effect sideslip has on rolling moment, this derivative can be
broken down into individual components. The wings‟ contribution depends on the
dihedral angle. A larger dihedral will produce a large sideforce, because there is
more surface area for the sideslip free stream to contact; however, the wings‟
contribution is generally negligible. Fuselage contribution depends on its shape and
18
size. Large fuselages have more contact surface area to produce larger sideforces.
The vertical tail can produce a large sideforce due to the large moment arm from the
center of gravity and the size of the tail. This derivative can be estimated using
methods found in Ref [7].
3.2.7 Change in Airplane Sideforce due to Change in Aileron Deflection, C
yoa
This contribution to the sideforce is very small and more often than not
negligible. If these rolling moment controls are close to a vertical surface the
sideforce cannot be neglected. This happens by the increase in lift on one side and a
decrease on the opposite side. These changes in lift are actually changes in pressure
which, if close to a vertical surface, can produce a sideforce. Wind tunnel tests have
to be completed to measure this in a reliable fashion.
3.2.8 Change in Airplane Sideforce due to Change in Rudder Deflection, C
yor
The rudder has a large influence on the sideforce. The purpose of a rudder
deflection is to create a sideforce that will produce a yawing moment. Depending on
the location it will also produce a small rolling moment. This sideforce depends on
the size of the vertical tail in relation to the wings. The lift curve slope of the vertical
tail also plays into the influence of sideforce. The sideforce contribution of the
rudder can be determined using Eq [20], Ref [5].
 Eq [20]
19
3.2.9 Yawing Moment, C
n
The aircraft yawing moment is the rotation about the zaxis in the stability
coordinate system. For a symmetrical aircraft the yawing moment is equal to zero for
zero angle of attack, sideslip, and control surface deflections. The same things that
influence the rolling moment influence the yawing moment. Those influences are
angle of attack, sideslip, speed, control surface deflections, and location of the
moment reference center. Eq [21] is the dimensional form of the yawing moment and
Eq [22] is the first order Taylor series, Ref [5].
 Eq [21]
 Eq [22]
3.2.10 Change in Yawing Moment due to Change in Sideslip, C
n
This derivative is referred to as the static directional stability and plays a large
role in Dutch roll and spiral dynamics. The derivative can be estimated by summing
the components of the aircraft, Eq [23], Ref [5]. The wings‟ influence can be
neglected since the flow is usually in line with the airfoil. The fuselage, on the other
hand, can play a large role, but it depends on the shape and the amount of projected
side area forward or aft of the center of gravity. Another impact on the fuselage
contribution is the Munk effect, which shifts the aerodynamic center forward. The
vertical tail also has a significant contribution. The size and location of the vertical
tail determines the amount of contribution it has. Eq [24] shows the contribution of
the vertical tail, Ref [5].
20
 Eq [23]
 Eq [24]
3.2.11 Change in Yawing Moment due to Change in Aileron Deflection, C
noa
With aileron deflections, the lift increases on the aileron with positive camber,
downward deflection, and the lift decreases on the aileron with negative camber,
upward deflection. An increase in lift will cause an increase in induced drag, and a
decrease in lift will cause a decrease in induced drag. Higher drag on one wing will
cause the aircraft to yaw. This type of yawing moment, called an adverse yawing
moment, is undesirable because it tends to yaw the aircraft out of an intended turn.
Therefore, either pilot input or differential ailerons are used to prevent the aircraft
from yawing.
3.2.12 Change in Yawing Moment due to Change in Rudder Deflection, C
nor
This derivative depends largely on the size of the vertical tail in relation to the
wings. The lift curve slope of the vertical tail also plays a large role. The location of
the vertical tail will determine the moment arm. Also, the size of the rudder will
influence the yawing moment. Eq [25] is used to estimate the derivative, Ref [5].
 Eq [25]
3.3 Perturbed State
A perturbed state flight condition is defined as one for which all motion
variables are defined relative to a known steady state flight condition, Ref [5]. It can
21
be thought of as the aircraft‟s motion varying from the steady state condition. These
variations are increases or decreases in velocity in any direction, or acceleration in
any direction or motion.
The longitudinal motion is influenced by a change in velocity in the u and w
direction if the velocity is broken down into three components, u, v, and w. Pitch
rate, q, also affects the longitudinal derivatives. It is assumed that the pitch rate has a
negligible influence on the lateraldirectional motion.
Some of the lateraldirectional perturbation influences are the v component of
a velocity change, yaw rate, and roll rate. Small roll rate perturbations cause non
symmetrical changes in local angles of attack over the lifting surface. It can easily be
visualized how these changes in angle of attack will influence the lateraldirectional
derivatives. Small yaw rate perturbations cause changes in the local velocity on the
lifting surfaces. They also cause changes in the angle of attack of the vertical tail.
There are also perturbation influences from changes in the flight angles. For
example, these flight angles can be the sideslip angle or angle of attack.
22
4 AAA
Advanced Aircraft Analysis, AAA, is a software program developed by
DARcorporation that is an aircraft stability, control, and design tool. It is widely
accepted as the industry standard and is used in 45 countries. The program follows
the Dr. Roskam method of preliminary aircraft design. Although the program will
model any size aircraft, the lower limit of the Reynolds number for calculations is
3e6. This means modeling of smaller aircraft may be inaccurate to some degree.
The software works for preliminary design all the way up to Class II cost
analysis. Geometric and flight characteristics of the aircraft being modeled are input,
and with a simple click of „calculate‟, outputs are produced. This program helps
eliminate errors in calculations and spreadsheets. If the designer does not understand
how something is being determined, there is a help button that shows all of the
equations used for each variable. This helps eliminate the black box feeling that
comes from most modeling software programs. If this help is not enough, the books
the program is based on are Airplane Design IVIII, Airplane Flight Dynamics and
Automatic Flight Controls, Parts I and II, by Dr. Jan Roskam, and Airplane
Aerodynamics and Performance, by Dr. C.T. Lan and Dr. Jan Roskam, Ref [8].
AAA is used before the aircraft is ever built, but can also be used to model
production vehicles. Using AAA, the stability and control derivatives can be found
and compared with actual flight test data. Doing this helps validate the software,
allowing the designer to be more comfortable with the results AAA produces when
designing an aircraft from scratch.
23
5 VORSTAB
VORSTAB is a software program that was developed by Dr. Edward Lan for
NASA. It has been used on several NASA research projects. The acronym name
stands for Vortex Stability. Using vortex flow effect, the lateraldirectional and
longitudinal stability derivatives are determined for an aircraft. The program follows
the PrandtlGlauert equation, Eq [26], in subsonic flow, Ref [9]. The PrandtlGlauert
equation is the linearized fullpotential equation, Eq [27], using small velocity
perturbations assumptions, Ref [9].
 Eq [26]
 Eq [27]
These are a few assumptions that VORSTAB uses for calculations.
 Assumes thin wing and therefore thickness effect is not accounted for in
calculations.
 The boundary layer separation is not accounted for, and therefore the flow
stays attached to the body.
 The wake aft of the wing is flat and does not increase in the zdirection.
The designer of the input model also makes assumptions based on the aircraft that
is being modeled and the flight characteristics. A detailed description of the program
can be found in Ref [10].
24
5.1 Creating a Model
VORSTAB runs an input file that the designer creates. There are 83 groups
that are used to describe the aircraft and the test cases that are to be run. These
groups break down a step by step procedure for creating the model and allow the
designer to trouble shoot more easily. Not every group is used to describe the aircraft
that is being modeled, and many groups are repeated several times. Detailed
descriptions of the groups are found in Ref [11].
To create a model of the aircraft the designer needs to have detailed
schematics of the geometry. Typically, this geometry comes from a CAD model.
Using the CAD model the designer can find all of the geometry needed for
VORSTAB. The software does not require that a fuselage be present in the model.
The first step in creating the model is to specify the number of lifting surfaces that are
to be modeled. These lifting surfaces include wings, horizontal tail, vertical tail, v
tail, and canards. The level flight geometry of each lifting surface is input into
VORSTAB. If this is not done, the program assumes that the geometry is at zero
angle of attack, when it might actually be more or less. For example, when the
Meridian sits on the ground, the wings and tail are at a higher angle of attack when
compared to level flight. The input lifting surface geometry required includes airfoil
characteristics (thickness and camber), location in terms of distance from reference
point (example: nose of aircraft), dihedral, twist, location of control surfaces, and
much more. Also needed is the lift coefficient, drag coefficient, and pitching moment
coefficient with respect to angle of attack. This step is repeated for each lifting
25
surface. Next, the fuselage data is input into VORSTAB. The geometry needed for
this includes crosssection shape, camber, axial locations from a reference point, and
much more. The more crosssections used in this model, the more accurate the results
will be. It must be made sure that the lifting surfaces are not inside of the fuselage.
The fuselage model is also created in level flight for the same reason as the lifting
surfaces. Scaling the input file is very important. The entire aircraft is scaled to a
size where the largest crosssectional radius of the fuselage is no larger than a
measurement of 1 in any unit system.
Next, the type of case that is to be run is described. This includes the flight
condition, control surface deflection angles, angle of attack, and more. With the input
file complete, VORSTAB can run. If there is an error in the input file, the output file
will stop at the group with the problem. This is a great advantage to find simple
problems in the input file that the designer overlooked. VORSTAB also creates a file
that can be opened with TECplot to see a graphical representation of the model. This
graphical representation allows the designer to visually determine if something looks
wrong with the model created.
26
6 FLUENT
FLUENT is a commercially sold computational fluid dynamics program that
is widely accepted for its high fidelity. The program has many different capabilities
and functions. Both 2demensional and 3demensional cases can be tested with
laminar, invisid, or turbulent flow. Depending on the model and flow type, the
designer chooses which flow field type is best. For large models, over 200,000 cells
and a coupled solver, a standard PC with single memory cannot process the model.
Super computers with multiple processors must be used to solve these larger models.
The following are the steps followed to run a FLUENT model.
Step 1: Create a GAMBIT model with a mesh for the flow field. GAMBIT is a
software program that allows the designer to easily create a geometric model of the
object being tested. The geometry of the object and the flow field around the model
are created. The size of the mesh will dramatically affect the accuracy of the results
and the processing time.
Step 2: Export the mesh so that it can be imported into FLUENT.
Step 3: With FLUENT opened, select either 2D or 3D, depending on the mesh
created in GAMBIT. There is also 2ddp and 3ddp. The “dp” stands for double
precision, and is more accurate than just the standard 2d or 3d. Case run time is
larger for this method, but the accuracy of the results is important to the validity of
the model.
27
Step 4: Import the mesh and run a check to see if there are errors in the mesh.
The size of the mesh can also be seen to verify that there are not too many cells to run
the test on a standard PC.
Step 5: Choose the type of solver: segregated or coupled. Coupled solver
requires much more computer memory. The „solver‟ is the method that the program
uses to solve the equations.
 Segregated solver solves the equations one by one. Momentum, continuity,
energy (if compressible flow), turbulent factor, and then check for
convergence.
 Coupled solver solves the same equations as the segregated solver, but does
this simultaneously instead of one by one.
Step 6: Choose the type of flow: laminar, invisid, or turbulent. If choosing
turbulent flow select the turbulent model to use. It is difficult to determine the most
accurate turbulent model without testing several different ones. The accuracy
depends on the mesh shape and flow pattern. Research has shown that the most
accurate method for aerodynamics is SpalartAllmaras.
Step 7: Set the fluid properties and boundary conditions. The fluid properties
include but are not limited to density, viscosity, velocity, wall friction, and much
more.
Step 8: Select the solution controls and discretization methods. This depends on
the shape of the body and type of flow. A description of each solution and
28
discretization method should be read before the designer chooses which one is best
for their model. The FLUENT online help menu should be used for these questions.
 SIMPLE and SIMPLEC are good choices for noncomplicated flow problems
(such as laminar flow).
 PISO is used for transient flows.
Step 9: Set the residuals and turn on what parameters are to be monitored. For
example, the lift and drag of the body can be monitored.
Step 10: Iterate until the solution converges. The convergence criteria is
determined by setting the residuals.
These steps can slightly vary from mesh to mesh, but in general these ten
steps allow the user to run a FLUENT model.
29
7 YAK54
The YAK54 is a remote control aircraft that the Aerospace Engineering
Department at the University of Kansas owns. This aircraft is a 1/3 scale version of
the Russian Yakovlev Yak54 acrobatic airplane. An autopilot system was installed
into this aircraft to help test the robustness and capabilities of the system before the
same system was installed on the larger and more expensive Meridian. Figure 2
shows a model of the YAK54 used by the University of Kansas, Ref [12].
Figure 2: YAK54
From the aircraft itself, detailed measurements were taken to create an exact
CAD model of the aircraft that was used during flight tests. Taking measurements
directly from the aircraft allowed abnormalities due to manufacturing to modeled and
30
accounted for. Table 1 shows some of the measurements taken from the YAK54. To
create the CAD model, Unigraphics 6.0 was used. Figure 3 shows the final CAD
model of the YAK54. The flight conditions used in AAA can be found in Table 2,
Ref [13].
Table 1: YAK54 Lifting Surface Dimensions
Wing
S = 10.9 ft
2
b = 7.9 ft
c
̅
= 1.45 ft
ì =
0.46
AR = 5.77
c
r
= 1.9 ft
c
t
= 0.874 ft
Horizontal Tail
S = 2.3 ft
2
b = 3 ft
c
̅
= 0.767 ft
ì =
0.81
AR = 3.91
c
r
= 0.844 ft
c
t
= 0.684 ft
Vertical Tail
S = 1.6 ft
2
b = 1.42 ft
c
̅
= 1.21 ft
ì =
0.35
AR = 1.25
c
r
= 1.67 ft
c
t
= 0.584 ft
31
Figure 3: Unigraphics CAD Model of YAK54
Table 2: YAK54 Flight Conditions
Flight Conditions
Altitude (AGL), h (ft) 400
Mach Number 0.106
TAS, U
1
(ft/sec) 118
Dynamic Pressure, q (lbs/ft
2
) 16.4
C.G. location, fraction c
̅
(in) 25.5
The importance of creating a VORSTAB model of the YAK54 was to gain
experience with the program and see how accurate the program was since this vehicle
had been flight tested. The need for a VORSTAB model also came from AAA
overestimating the drag on vehicles with low Reynolds numbers. After flight tests,
Edmond Leong tuned the AAA results to develop more accurate stability and control
derivatives. This will all be discussed later.
32
7.1 AAA Modeling of the YAK54
AAA was used to determine the YAK54‟s dimensionless longitudinal and
lateraldirectional stability and control derivatives. Table 3 shows the trimmed
condition values of the aircraft, Table 4 shows the longitudinal derivatives, and Table
5 shows the lateraldirectional derivatives, all results were taken from Ref [13].
Table 6 shows the stability requirements for the YAK54 produced by AAA, Ref
[13].
Table 3: YAK54 AAA Moment of Inertia and Trimmed Values
Mass Data
Weight (lbs) 28
Ixx
B
(slugft
2
) 1.09
Iyy
B
(slugft
2
) 2.11
Izz
B
(slugft
2
) 3.04
Ixz
B
(slugft
2
) 0.05
Steady StateTrimmed
C
L1
0.147
C
D1
0.053
C
Tx1
0.052
C
m1
0.002
C
mT1
0.0009
Angle of attack, o
1
(deg)
1.79
Elevator Deflection Angle, o
e1
(deg)
0.69
33
Table 4: YAK54 AAA Longitudinal Derivatives
Longitudinal Coefficients and Stability
DerivativesStability Axes
Derivatives (1/deg) (1/rad)
C
Du
0.0011 0.0630
C
Do
0.0015 0.0859
C
txu
0.1546 8.858
C
Lu
0.0017 0.0974
C
Lo
0.0792 4.538
C
Lodot
0.0337 1.9309
C
Lq
0.0899 5.1509
C
mu
0.0004 0.0229
C
mo
0.0065 0.3724
C
modot
0.078 4.469
C
mq
0.1484 8.503
C
mTu
0.0000 0.0000
C
mTo
0.0005 0.0286
Longitudinal Control and Hinge Moment
DerivativesStability Axes
Derivatives (1/deg) (1/rad)
C
Doe
0.0000 0.0000
C
Loe
0.0066 0.3782
C
moe
0.0153 0.8766
34
Table 5: YAK54 AAA LateralDirectional Derivatives
LateralDirectional Coefficients and Stability DerivativesStability
Axes
Derivatives (1/deg) (1/rad)
C
l
0.0004 0.0229
C
lp
0.0067 0.3839
C
lr
0.0009 0.0520
C
y
0.0060 0.3438
C
yp
0.0001 0.0060
C
yr
0.0041 0.2349
C
n
0.0017 0.0974
C
nT
0.0001 0.0057
C
np
0.0003 0.0172
C
nr
0.0020 0.1150
LateralDirectional Control and Hinge Moment Derivatives
Stability Axes
Derivatives (1/deg) (1/rad)
C
loa
0.0061 0.3495
C
lor
0.0003 0.0172
C
yoa
0.0000 0.0000
C
yor
0.0034 0.1900
C
noa
0.0020 0.1146
C
nor
0.0017 0.0974
35
Table 6: YAK54 AAA Stability Requirements
Type of
Stability
Corresponding
Derivatives
Criterion Derivative (deg
1
) Stable/Unstable
Forward Speed C
Txu
 C
Du
< 0 0.1557 Stable
Sideslip C
y
< 0 0.3462 Stable
Vertical Speed C
Lo
> 0 4.7625 Stable
Angle of Attack C
mo
< 0 0.3813 Stable
Angle of
Sideslip
C
n
> 0 0.0954 Stable
Roll Rate C
lp
< 0 0.3817 Stable
Pitch Rate C
mq
< 0 10.040 Stable
Yaw Rate C
nr
< 0 0.1161 Stable
Lateral C
l
< 0 0.0257 Stable
7.2 VORSTAB Modeling of the YAK54
A YAK54 input file was created for VORSTAB. This input file can be found
in Appendix A. The input file modeled the entire aircraft include all three lifting
surfaces and the fuselage. Table 7 shows the lift, drag, pitching moment coefficients,
and how those three coefficients are affected by pitch rate. Due to the assumption
that the flow stays attached to the body, there is not a stall point for this model.
Figure 4 through Figure 7 shows Table 7 in graphical form.
Table 7: YAK54 VORSTAB Longitudinal Coefficient and Pitch Rate (1/rad)
o (deg) C
L
C
D
C
M
C
Lq
C
Dq
C
Mq
5 0.376 0.025 0.205 8.95 0.000 10.6
0 0.013 0.020 0.219 8.84 0.000 10.1
5 0.438 0.021 0.317 8.36 0.000 8.99
10 0.820 0.034 0.409 8.63 0.000 10.2
15 1.14 0.075 0.476 8.22 0.000 9.95
20 1.40 0.142 0.548 7.63 0.000 9.43
36
0.600
0.400
0.200
0.000
0.200
0.400
0.600
0.800
1.000
1.200
1.400
1.600
5 0 5 10 15 20
C
L
(
N
.
D
.
)
o (deg)
Figure 4: YAK54 VORSTAB Lift Curve Slope
0.000
0.020
0.040
0.060
0.080
0.100
0.120
0.140
0.160
5 0 5 10 15 20
C
D
(
N
.
D
.
)
o (deg)
Figure 5: YAK54 VORSTAB Drag Curve
37
0.600
0.500
0.400
0.300
0.200
0.100
0.000
5 0 5 10 15 20
C
M
(
N
.
D
.
)
o (deg)
Figure 6: YAK54 VORSTAB Pitching Moment Coefficient Curve
15.00
10.00
5.00
0.00
5.00
10.00
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
CL_q CM_q
Figure 7: YAK54 VORSTAB Longitudinal Derivatives
38
The longitudinal derivatives produced by VORSTAB follow the expected
trends. For example, the pitching moment curve has a negative slope. This slope
makes the aircraft naturally want to pitch back to the stability axes. These values
were also within the range that one would expect to see. The drag, C
Do
, is lower than
the AAA results predicted. This is expected since AAA overestimates the drag at
lower Reynolds number. AAA approximation was 0.0313 and VORSTAB estimates
it to be 0.013, this is a significant change. The drag is known to be lower than the
AAA results due the fact that AAA says the aircraft must be trimmed at a higher
throttle setting, approximately 45%, and from the flight tests it is trimmed at,
approximately 27% throttle setting. The magnitudes of these values might be off
slightly off, but the coefficients follow the correct trend. Table 8 shows the lateral
directional derivatives due to sideslip, roll rate, and yaw rate. Figure 8 through
Figure 10 depict these lateraldirectional derivatives in graphical form.
Table 8: YAK54 VORSTAB LateralDirectional Derivatives (1/rad)
o
(deg)
C
y_
C
l_
C
n_
C
y_p
C
l_p
C
n_p
C
y_r
C
l_r
C
n_r
5
0.403 0.021 0.187 0.013 0.213 0.045 0.433 0.011 0.196
0
0.497 0.030 0.220 0.046 0.218 0.023 0.486 0.032 0.229
5
0.531 0.050 0.249 0.016 0.233 0.070 0.531 0.082 0.250
10
0.506 0.065 0.264 0.018 0.198 0.078 0.558 0.105 0.251
15
0.437 0.064 0.262 0.010 0.171 0.086 0.570 0.108 0.238
20
0.363 0.068 0.249 0.001 0.154 0.090 0.563 0.111 0.211
39
0.6
0.5
0.4
0.3
0.2
0.1
0
0.1
0.2
0.3
0.4
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
Cy_beta Cl_beta Cn_beta
Figure 8: YAK54 VORSTAB LateralDirectional Derivatives due to Sideslip
0.250
0.200
0.150
0.100
0.050
0.000
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
Cy_p Cl_p Cn_p
Figure 9: YAK54 VORSTAB LateralDirectional Derivatives due to Roll Rate
40
0.300
0.200
0.100
0.000
0.100
0.200
0.300
0.400
0.500
0.600
0.700
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
Cy_r Cl_r Cn_r
Figure 10: YAK54 VORSTAB LateralDirectional Derivatives due to Yaw Rate
Similar to the longitudinal derivatives, the lateraldirectional derivatives
follow an expected trend. All of the coefficients have the correct positive or negative
value, except for rolling moment coefficient due to yaw rate at an angle of attack of 
5˚. This coefficient would always be positive when the aircraft has flow that is
attached to the fuselage, because the fuselage contribution outweighs the vertical tail
contribution. Vertical tail contribution can be negative or positive. Again, the
magnitudes might be slightly off, but the in the expected range. Both longitudinal
and lateraldirectional derivatives change when the control surfaces are deflected.
Table 9 shows the lateraldirectional derivatives due to aileron deflection, Table 10
shows the longitudinal derivatives due to elevator deflection, and Table 11 shows the
lateraldirectional derivatives due to rudder deflection. These control surface
deflection derivatives can be seen in Figure 11 through Figure 19.
41
Table 9: YAK54 VORSTAB LateralDirectional Due to Aileron Deflection (1/rad)
o (deg) o
a
C
y
C
l
C
n
0 10 0.0009 0.0240 0.0054
5 10 0.0013 0.0221 0.0027
10 10 0.0041 0.0243 0.0003
15 10 0.0077 0.0224 0.0040
20 10 0.0103 0.0176 0.0073
0 5 0.0054 0.0120 0.0019
5 5 0.0070 0.0110 0.0035
10 5 0.0083 0.0121 0.0049
15 5 0.0093 0.0111 0.0061
20 5 0.0099 0.0087 0.0071
0 0 0.000 0.000 0.000
5 0 0.000 0.000 0.000
10 0 0.000 0.000 0.000
15 0 0.000 0.000 0.000
20 0 0.000 0.000 0.000
0 5 0.0033 0.0121 0.0056
5 5 0.0029 0.0110 0.0051
10 5 0.0026 0.0121 0.0044
15 5 0.0022 0.0110 0.0036
20 5 0.0010 0.0086 0.0020
0 10 0.0213 0.0244 0.0231
5 10 0.0189 0.0221 0.0209
10 10 0.0178 0.0242 0.0193
15 10 0.0164 0.0219 0.0172
20 10 0.0125 0.0171 0.0128
42
0.0250
0.0200
0.0150
0.0100
0.0050
0.0000
0.0050
10 5 0 5 10
C
y
(
N
.
D
.
)
aileron deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 11: YAK54 VORSTAB Sideforce Coefficient due to Aileron Deflection
0.0300
0.0200
0.0100
0.0000
0.0100
0.0200
0.0300
10 5 0 5 10
C
l
(
N
.
D
.
)
aileron deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 12: YAK54 VORSTAB Rolling Moment Coefficient due to Aileron Deflection
43
0.0100
0.0050
0.0000
0.0050
0.0100
0.0150
0.0200
0.0250
10 5 0 5 10
C
n
(
N
.
D
.
)
aileron deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 13: YAK54 VORSTAB Yawing Moment Coefficient due to Aileron Deflection
The aileron deflection is expected to have a very small effect on the sideforce
coefficient. The rolling moment controls are in close proximity to the fuselage and
this can cause a rolling moment. Therefore, the VORSTAB results are not to be
assumed as wrong. Conventional ailerons will usually produce a negative yawing
moment, but not always. A positive aileron deflection is defined as one that produces
a positive rolling moment, and this is seen in the results produced. VORSTAB shows
a positive yawing moment at almost all aileron deflections. The negative yawing
moment produced by the ailerons‟ deflection will yaw the aircraft out of an intended
turn and the VORSTAB results say the opposite. Also, the yawing moment does not
have symmetric results with respect to negative and positive aileron deflections.
Opposite aileron deflections should result in the same magnitude, but should have
opposite sign convention. VORSTAB asks for a sideslip angle to be put in the input
44
file and this might be the cause of the asymmetric results. Without the source code it
would be difficult to determine the source for the error. Therefore, these two
asymmetrical results should be assumed wrong and ignored.
Table 10: YAK54 VORSTAB Longitudinal Due to Elevator Deflection (1/rad)
o (deg) o
e
C
L
C
D
C
M
0 12 0.0114 0.0199 0.2153
5 12 0.4363 0.0211 0.3123
10 12 0.8181 0.0350 0.4039
15 12 1.142 0.0757 0.4707
20 12 1.401 0.1430 0.5432
0 6 0.0121 0.0199 0.2169
5 6 0.4374 0.0211 0.3148
10 6 0.8191 0.0347 0.4064
15 6 1.143 0.0754 0.4736
20 6 1.402 0.1427 0.5458
0 0 0.0129 0.0199 0.2186
5 0 0.4384 0.0212 0.3173
10 0 0.8202 0.0345 0.4089
15 0 1.144 0.0752 0.4764
20 0 1.403 0.1425 0.5484
0 6 0.0136 0.0199 0.2202
5 6 0.4394 0.0212 0.3197
10 6 0.8213 0.0343 0.4119
15 6 1.146 0.0749 0.4793
20 6 1.404 0.1422 0.5511
0 12 0.0143 0.0199 0.2219
5 12 0.4404 0.0212 0.3221
10 12 0.8227 0.0341 0.4155
15 12 1.147 0.0747 0.4821
20 12 1.405 0.1420 0.5537
45
0.0000
0.2000
0.4000
0.6000
0.8000
1.0000
1.2000
1.4000
1.6000
10 5 0 5 10
C
L
(
N
.
D
.
)
elevator deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 14: YAK54 VORSTAB Lift Coefficient due to Elevator Deflection
0.0000
0.0200
0.0400
0.0600
0.0800
0.1000
0.1200
0.1400
0.1600
10 5 0 5 10
C
D
(
N
.
D
.
)
elevator deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 15: YAK54 VORSTAB Drag Coefficient due to Elevator Deflection
46
0.6000
0.5000
0.4000
0.3000
0.2000
0.1000
0.0000
10 5 0 5 10
C
m
(
N
.
D
.
)
elevator deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 16: YAK54 VORSTAB Pitching Moment Coefficient due to Elevator Deflection
The lift and drag coefficients are not seen to change dramatically due to
elevator deflection, but this deflection will change the angle of attack. This change in
angle of attack will then change the lift and drag dramatically. The pitching moment
due to elevator deflection should be negative and the VORSTAB results show this.
Also, the change in pitching moment seems small, but the change is large enough to
change the aircraft‟s angle of attack.
47
Table 11: YAK54 VORSTAB LateralDirectional Due to Rudder Deflection (1/rad)
o (deg) o
r
C
y
C
l
C
n
0 10 0.0597 0.0033 0.0310
5 10 0.0591 0.0033 0.0307
10 10 0.0586 0.0032 0.0305
15 10 0.0576 0.0030 0.0300
20 10 0.0570 0.0029 0.0297
0 5 0.0300 0.0016 0.0156
5 5 0.0297 0.0017 0.0155
10 5 0.0295 0.0016 0.0153
15 5 0.0290 0.0015 0.0151
20 5 0.0287 0.0015 0.0149
0 0 0.0000 0.0000 0.0000
5 0 0.0000 0.0000 0.0000
10 0 0.0000 0.0000 0.0000
15 0 0.0000 0.0000 0.0000
20 0 0.0000 0.0000 0.0000
0 5 0.0300 0.0016 0.0156
5 5 0.0297 0.0017 0.0155
10 5 0.0295 0.0016 0.0153
15 5 0.0290 0.0015 0.0151
20 5 0.0287 0.0015 0.0149
0 10 0.0597 0.0033 0.0310
5 10 0.0591 0.0033 0.0307
10 10 0.0586 0.0032 0.0305
15 10 0.0576 0.0030 0.0300
20 10 0.0570 0.0029 0.0297
48
0.0600
0.0400
0.0200
0.0000
0.0200
0.0400
0.0600
10 5 0 5 10
C
y
(
N
.
D
.
)
Rudder deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 17: YAK54 VORSTAB Sideforce Coefficient due to Rudder Deflection
0.0033
0.0023
0.0013
0.0003
0.0007
0.0017
0.0027
10 5 0 5 10
C
l
(
N
.
D
.
)
Rudder deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 18: YAK54 VORSTAB Rolling Moment Coefficient due to Rudder Deflection
49
0.0320
0.0220
0.0120
0.0020
0.0080
0.0180
0.0280
10 5 0 5 10
C
n
(
N
.
D
.
)
Rudder deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 19: YAK54 VORSTAB Yawing Moment Coefficient due to Rudder Deflection
A positive rudder defection will produce a positive sideforce as expected, and
viceversa. The rolling moment is a direct result of the sideforce, and therefore, a
positive rudder deflection will also result in a positive rolling moment. A negative
yawing moment is produced by a positive rudder deflection.
The VORSTAB stability requirements are shown in Table 12. As seen from
this table, the YAK54 is stable in all modes. The stability requirements were only
found at the 0˚ angle of attack, and could be unstable at different positions. From
reviewing the previous results from VORSTAB, it can be assumed that the aircraft is
stable in all flight conditions.
50
Table 12: YAK54 VORSTAB Stability Requirements
Type of Stability
Corresponding
Derivatives
Criterion
Derivative
(rad
1
)
Stable/Unstable
Sideslip C
y
< 0 0.4968 Stable
Vertical Speed C
Lo
> 0 4.8762 Stable
Angle of Attack C
mo
< 0 1.1314 Stable
Angle of Sideslip C
n
> 0 0.2200 Stable
Roll Rate C
lp
< 0 0.2178 Stable
Pitch Rate C
mq
< 0 10.090 Stable
Yaw Rate C
nr
< 0 0.2289 Stable
Lateral C
l
< 0 0.0304 Stable
As seen from these results produced by VORSTAB, the YAK54 is a stable
aircraft with a large amount of control. This is expected since it is a remote
controlled aircraft. The lift curve slope of the results was found to be 4.8762 rad
1
.
This derivative was found from 0˚ to 5˚ angles of attack, since this was the most
linear section of the curve. The aircraft is known to not stall at these angles also.
Characterizing the drag and pitching moment due to angle of attack into a single
number is more difficult than lift, since these two coefficients are not linear curves.
The same range of angles of attack was used for the other two derivatives. The drag
derivative, C
Do
, is 0.0144 rad
1
and the pitching moment derivative, C
mo
, is 1.1314
rad
1
. All other derivatives were determined at 0˚ angle of attack. The results from
VORSTAB and AAA will be discussed and compared in more detail in the following
section.
51
7.3 Method Comparison YAK54
Both AAA and VORSTAB are considered as an accurate approximation for
the control derivatives. Using a computational fluid dynamic method such as
VORSTAB should produce results closer to that of the actual data than results from
DATCOM and AAA. As stated previously, the VORSTAB results were found over
an angle of attack range of 0˚ to 5˚. Table 13 shows the comparison data for the
YAK54 for AAA and VORSTAB. The tuned AAA results were derived after
several flight tests were conducted, and the system identification warranted an
improved model. These turned values were taken from Ref [14] and explanation for
deriving them can be found there.
52
Table 13: YAK54 AAA and VORSTAB Stability and Control Comparison
Stability and Control Derivatives
Derivative
(1/rad)
AAA
AAA
Tuned
VORSTAB
% Difference
AAA
% Difference
AAA Tuned
C
Lo
4.5380 4.538 4.8762 7.45 7.45
C
Do
0.0859 0.0859 0.0144 83.24 83.24
C
mo
0.3724 0.3724 1.1314 203.81 203.81
C
Lq
5.1509 5.1509 8.8411 71.64 71.64
C
mq
8.5030 16.1064 10.090 18.66 37.35
C
l
0.0229 0.022 0.0304 32.75 238.18
C
lp
0.3839 0.5858 0.2178 43.27 62.82
C
lr
0.0520 0.0743 0.0323 37.88 56.53
C
y
0.3438 0.2707 0.4968 44.50 83.52
C
yp
0.0060 0.0194 0.0455 858.33 334.54
C
yr
0.2349 0.2531 0.4865 107.11 92.22
C
n
0.0974 0.1052 0.2200 125.87 109.13
C
np
0.0172 0.0387 0.0231 34.30 40.31
C
nr
0.1150 0.289 0.2289 99.04 20.80
C
Loe
0.3782 0.3782 0.0071 98.12 98.12
C
Doe
0.0000 0.0000 0.0000 0.00 0.00
C
moe
0.8766 1.2289 0.0158 98.20 98.71
C
loa
0.3495 0.3707 0.1388 60.29 62.56
C
lor
0.0172 0.0219 0.0188 9.30 14.16
C
yoa
0.0000 0.0000 0.0379  
C
yor
0.1948 0.2228 0.3441 76.64 54.44
C
noa
0.1146 0.0088 0.0646 156.37 834.09
C
nor
0.0974 0.1404 0.1789 83.68 27.42
Several of the differences between the AAA and VORSTAB results fall
within the error ranges given by Dr. Roskam. The differences that are in the 100 of
percents do not fall within the error range though. The tuned AAA results improved
53
in some cases and not in others. These derivatives, C
nr
, C
nor
, C
l
, C
mq
, and C
moe
were
tuned from the flight test and the others were taken from the AVL model. The tuned
yawing moment derivatives show an improvement in the VORSTAB results over
AAA. More flight tests need to be conducted and parameter system identification
need to be conducted to improve all of the derivatives.
Both methods show that the aircraft is stable. For stability and control
derivatives, there are typical ranges that can be expected, and through his vast
experience, Dr. Jan Roskam developed ranges for the control derivatives. These
ranges are for conventional aircraft, and the ranges are a function of Mach number.
Table 14 shows the Dr. Roskam, Ref. [1], typical ranges compared to the AAA and
VORSTAB results.
54
Table 14: YAK54 Stability and Control Derivatives Typical Ranges
Stability and Control Derivatives
Derivative
(1/rad)
VORSTAB AAA
Typical Range
Within Range
(Yes/No)
Dr. Roskam VORSTAB AAA
C
Lo
4.8762 4.5380 1.0 to 8.0 Yes Yes
C
Do
0.0144 0.0859 0.0 to 2.0 Yes Yes
C
mo
1.1314 0.3724 4.0 to 1.0 Yes Yes
C
Lq
8.8411 5.1509 0.0 to 30.0 Yes Yes
C
mq
10.09 8.5030 90.0 to 0.0 Yes Yes
C
l
0.0304 0.0229 0.1 to 4.0 Yes Yes
C
lp
0.2178 0.3839 0.1 to 0.8 Yes Yes
C
lr
0.0323 0.0520 0.0 to 0.6 Yes Yes
C
y
0.4968 0.3438 0.1 to 2.0 Yes Yes
C
yp
0.0455 0.0060 0.3 to 0.8 Yes Yes
C
yr
0.4865 0.2349 0.0 to 1.2 Yes Yes
C
n
0.2200 0.0974 0.0 to 4.0 Yes Yes
C
np
0.0231 0.0172 0.5 to 0.1 Yes Yes
C
nr
0.2289 0.1150 0.0 to 1.0 Yes Yes
C
Loe
0.0071 0.3782 0.0 to 0.6 Yes Yes
C
Doe
0.0000 0.0000 Negligible Yes Yes
C
moe
0.0158 0.8766 0.0 to 4.0 Yes Yes
C
loa
0.1388 0.3495 0.0 to 0.4 Yes Yes
C
lor
0.0188 0.0172 0.04 to 0.04 Yes Yes
C
yoa
0.0379 0.0000 Negligible Yes Yes
C
yor
0.3441 0.1948 0.0 to 0.5 Yes Yes
C
noa
0.0646 0.1146 0.08 to 0.08 Yes Yes
C
nor
0.1789 0.0974 0.15 to 0.0 No Yes
The YAK54 VORSTAB results fall within all of the typical ranges that Dr.
Roskam developed except for the yawing moment coefficient due to rudder
deflection, C
nor
. These ranges are just estimations and many aircraft‟s derivatives do
55
not fall into the ranges. The ranges just give the designer an idea of things to expect.
Also, the sideforce moment coefficient due to aileron deflection, C
yoa
, is not
negligible. This can happen when the rolling moment controls are in close proximity
to a vertical surface (fuselage or vertical tail). The ailerons, on the YAK54, are in
close proximity to the fuselage, and this causes the coefficient to be a nonnegligible
value. The AAA results all fall within the typical ranges. From these results, it is
concluded that both programs obtained valid results. These results should be
compared to the more flight test data to validate both software programs. A model
should be created with the new flight test data and then compared to the AAA and
VORSTAB results.
7.4 Linearized Model of the YAK54
For the information and data on the state space model please refer to
Appendix B.
56
8 MantaHawk
The MantaHawk is a remote control aircraft designed at the University of
Kansas, by the AE 721 graduate design course in the fall 2009. The design team
consisted of Emily Arnold, Robert Burns, Dustin Grorud, Katrina Legursky, Rick
Riley, Dave Royer, and Jonathan Tom. This small UAV was developed in
conjunction with the CReSIS Meridian project. MantaHawk‟s size and weight were
well suited for a seaborn launch with a much shorter range than the other Meridian
Antarctic mission. The goal of the team was to help improve the current research
going on at the University of Kansas and within CReSIS itself.
After much was research conducted by the team, a wingbody configuration
was chosen as the most suitable configuration for the particular mission. The design
of the MantaHawk followed the Dr. Roskam method using AAA ,and AVL for some
of the stability and control derivatives that AAA did not produce. Class II design
resulted in an empty weight of 32 lbs for the vehicle with a design cruise speed of 70
kts. The vehicle had a payload capacity of 15 lbs plus 7 lbs of fuel. This aircraft was
a prototype and had its problems. Soon after takeoff, the aircraft pitched downward
aggressively, and this caused the airplane to go into a nose dive. With post flight
analysis, the team determined that a negative pitching moment was much higher than
predicted and without preflight trim there was not enough control to fly. Figure 20
shows the MantaHawk designed by the AE 721 class, Ref [15]. The design and flight
test of the MantaHawk can be found in Ref [15].
57
Figure 20: MantaHawk
8.1 AAA Modeling of the MantaHawk
Unlike the YAK54, the MantaHawk was designed at the University of
Kansas using AAA software. The MantaHawk‟s AAA model estimated the stability
and control derivatives, but this vehicle was never a proven platform. Table 15 and
Table 16 shows the stability and control derivatives produced by AAA, Ref [15].
Table 17 shows the AAA stability requirements for the MantaHawk, Ref [15].
58
Table 15: MantaHawk AAA Longitudinal Derivatives
Longitudinal Coefficients and
Stability DerivativesStability Axes
Derivatives (1/rad)
C
Du
0.0000
C
Do
0.0615
C
Dq
0.0000
C
Txu
0.0000
C
Lu
0.0021
C
Lo
3.9630
C
Lq
5.0920
C
mu
0.0000
C
mo
0.4034
C
mq
4.7340
C
mTu
0.0000
C
mTo
0.2050
C
Doel
0.4209
C
Loel
0.4784
C
moel
1.0630
59
Table 16: MantaHawk AAA LateralDirectional Derivatives
LateralDirectional Coefficients and Stability
DerivativesStability Axes
Derivatives (1/rad)
C
l
0.0817
C
lp
0.4892
C
lr
0.0407
C
y
0.1877
C
yp
0.1249
C
yr
0.0897
C
n
0.0359
C
nT
0.0007
C
np
0.0169
C
nr
0.0219
C
yoel
0.0012
C
loel
0.3588
C
noel
0.0055
Table 17: MantaHawk AAA Stability Requirements
Type of Stability
Corresponding
Derivatives
Criterion
Derivative
(rad
1
)
Stable/Unstable
Sideslip C
y
< 0 0.1877 Stable
Vertical Speed C
Lo
> 0 3.9630 Stable
Angle of Attack C
mo
< 0 0.4034 Stable
Angle of Sideslip C
n
> 0 0.0359 Stable
Roll Rate C
lp
< 0 0.4892 Stable
Pitch Rate C
mq
< 0 4.7340 Stable
Yaw Rate C
nr
< 0 0.0219 Stable
Lateral C
l
< 0 0.0817 Stable
According to AAA, the MantaHawk is stable in all modes and should not have
a problem in flight. These results gave the flight test team confidence going into the
60
first flight. Shortly after liftoff, the aircraft nosed into the ground. The wreck raised
concern and prompted a VORSTAB model to be produced to determine if the vehicle
was actually stable in all modes.
8.2 VORSTAB Modeling of the MantaHawk
The MantaHawk was a flying wing that used three different airfoil shapes
throughout the wing. There were also large wingtips that were used to help prevent
wingtip vortices and wingtip stall. Since it was a flying wing, the VORSTAB model
did not include a fuselage, but rather just one lifting surface and one control surface.
A symmetrical deflection of the control surface, elevon, imitates an elevator input,
and an asymmetrical deflection imitates a rudder or aileron input. The VORSTAB
input file for the MantaHawk can be found in Appendix A. Table 18 shows the
longitudinal derivatives for the MantaHawk. Figure 21 through Figure 25 depict
Table 18 in graphical form.
Table 18: MantaHawk VORSTAB Longitudinal Derivatives (1/rad)
o (deg) C
L
C
D
C
m
C
Lq
C
Dq
C
mq
5 0.1683 0.0119 0.0167 1.9478 0.0000 1.6498
0 0.0377 0.0068 0.0006 1.9645 0.0000 1.6625
5 0.2603 0.0092 0.0159 1.9365 0.0000 1.6498
10 0.4760 0.0154 0.0217 1.8694 0.0000 1.6052
15 0.6123 0.0366 0.0271 1.7585 0.0000 1.5442
20 0.6709 0.0763 0.0256 1.6138 0.0000 1.4615
61
0.3
0.2
0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
5 0 5 10 15 20
C
L
(
N
.
D
.
)
o (deg)
Figure 21: MantaHawk VORSTAB Lift Coefficient
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
5 0 5 10 15 20
C
D
(
N
.
D
.
)
o (deg)
Figure 22: MantaHawk VORSTAB Drag Coefficient
62
0.02
0.015
0.01
0.005
0
0.005
0.01
0.015
0.02
0.025
0.03
5 0 5 10 15 20
C
M
(
N
.
D
.
)
o (deg)
Figure 23: MantaHawk VORSTAB Pitching Moment Coefficient
1.5
1.55
1.6
1.65
1.7
1.75
1.8
1.85
1.9
1.95
2
5 0 5 10 15 20
C
L
_
q
(
1
/
r
a
d
)
o (deg)
Figure 24: MantaHawk VORSTAB Lift Coefficient due to Pitch Rate
63
1.7
1.65
1.6
1.55
1.5
1.45
5 0 5 10 15 20
C
M
_
q
(
1
/
r
a
d
)
o (deg)
Figure 25: MantaHawk VORSTAB Pitching Moment Coefficient due to Pitch Rate
As seen from the pitching moment curve, there is a positive slope. This
means that the aircraft is naturally unstable. When the aircraft is pitched upward, it
will not pitch back downward, but rather continue pitching upward. The other
longitudinal derivatives were as expected. With this information the flight test team
would have exercised more caution during the testing of the vehicle. The drag, C
Do
,
estimated by VORSTAB was 0.0068 and AAA estimated it to be 0.022. This is also
a very significant difference and should be investigated. In reality, the drag is most
likely between these two software programs. VORSTAB only determines the
pressure drag and not the viscous drag. This is why the VORSTAB drag results are
lower than what flight test data should show. The overall lift coefficient for the
aircraft is rather small, but the drag is very low for this aircraft. Pitching moment
coefficient due to pitch rate is typically large, because it is proportional to the square
64
of the moment arm of the horizontal tail. The MantaHawk is a flying wing, so the
moment arm is not as large as a conventional aircraft and this is why the derivative is
not very large. Table 19 shows the lateraldirectional derivatives. Figure 26 through
Figure 28 depicts these derivatives in graphical form.
Table 19: MantaHawk VORSTAB LateralDirectional Derivatives (1/rad)
o
(deg)
C
y
C
l
C
n
C
yp
C
lp
C
np
C
yr
C
lr
C
nr
5 0.098 0.016 0.021 0.082 0.287 0.072 0.036 0.011 0.004
0 0.117 0.048 0.018 0.054 0.287 0.0003 0.047 0.020 0.008
5 0.090 0.082 0.009 0.071 0.283 0.057 0.053 0.055 0.003
10 0.034 0.107 0.002 0.061 0.261 0.0996 0.061 0.084 0.006
15 0.033 0.091 0.011 0.033 0.165 0.086 0.056 0.078 0.016
20 0.060 0.064 0.015 0.004 0.083 0.065 0.056 0.054 0.020
0.1400
0.1200
0.1000
0.0800
0.0600
0.0400
0.0200
0.0000
0.0200
0.0400
0.0600
0.0800
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
Cy_beta Cl_beta Cn_beta
Figure 26: MantaHawk VORSTAB LateralDirectional Derivatives due to Sideslip
65
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
0.05
0.1
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
Cy_p Cl_p Cn_p
Figure 27: MantaHawk VORSTAB LateralDirectional Derivatives due to Roll Rate
0.02
0
0.02
0.04
0.06
0.08
0.1
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
Cy_r Cl_r Cn_r
Figure 28: MantaHawk VORSTAB LateralDirectional Derivatives due to Yaw Rate
66
At high angles of attack, aircrafts do not always follow the expected trends for
the derivatives. This is observed on several of these derivatives. The sideforce
coefficient due to sideslip is typically negative due to the defined positive angle of
sideslip. VORSTAB estimates that at high angles of attack this derivative becomes
positive. The yawing moment due to sideslip should be positive for the same reasons
as sideforce should be negative. This derivative becomes negative at high angles of
attack. Due to the definition of sideforce coefficient due to roll rate it would be
assumed that the derivative should be negative, but at high angles of attack the
moment arm will change signs. This will result in the derivative changing signs to
positive. It can be expected for the sign to change for the yawing moment derivative
due to roll rate as the angle of attack changes from positive to negative. The effect of
this derivative is rather small on the airplane dynamic stability, and since this is the
case the sign convention does not matter. Yawing moment coefficient due to yaw
rate changes signs due to the fact that as the angle of attack increase the moment arm
goes from positive to negative. All other derivatives were as expected.
With the control surfaces deflected, the longitudinal and lateraldirectional
derivatives change. Table 20 shows the effects of an elevator input sent to the
elevons on the longitudinal derivatives. This type of input results in a symmetrical
deflection of the elevons. Table 21 shows the effects of the aileron or rudder input
sent to the elevons on the lateral directional derivatives. There is a resulting
asymmetrical deflection of the elevons. An asymmetrical deflection imitates an
67
aileron and a rudder input both. All values are given in radians. Graphically, these
tables are shown in Figure 29 through Figure 34.
Table 20: MantaHawk VORSTAB Longitudinal Derivatives due to Symmetrical
Deflection
o (deg) o
el
(deg) C
L
C
D
C
m
0 12 0.0896 0.0125 0.0561
5 12 0.1251 0.0107 0.0827
10 12 0.3413 0.0127 0.0888
15 12 0.5118 0.0252 0.0806
20 12 0.6034 0.0534 0.0602
0 6 0.0250 0.0088 0.0267
5 6 0.1930 0.0091 0.0477
10 6 0.4132 0.0126 0.0529
15 6 0.5705 0.0296 0.0485
20 6 0.6432 0.0622 0.0407
0 0 0.0377 0.0068 0.0006
5 0 0.2603 0.0092 0.0159
10 0 0.4760 0.0154 0.0217
15 0 0.6123 0.0366 0.0271
20 0 0.6709 0.0763 0.0256
0 6 0.1079 0.0085 0.0298
5 6 0.3354 0.0109 0.0201
10 6 0.5317 0.0215 0.0025
15 6 0.6433 0.0484 0.0120
20 6 0.6871 0.0939 0.0154
0 12 0.1840 0.0126 0.0687
5 12 0.3969 0.0158 0.0476
10 12 0.5768 0.0299 0.0210
15 12 0.6643 0.0645 0.0004
20 12 0.7001 0.1117 0.0086
68
0.2
0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
12 10 8 6 4 2 0 2 4 6 8 10 12
C
L
(
N
.
D
.
)
elevon deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 29: MantaHawk VORSTAB Lift Coefficient due to Symmetrical Deflection
0
0.02
0.04
0.06
0.08
0.1
0.12
12 10 8 6 4 2 0 2 4 6 8 10 12
C
D
(
N
.
D
.
)
elevon deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 30: MantaHawk VORSTAB Drag Coefficient due to Symmetrical Deflection
69
0.08
0.06
0.04
0.02
0
0.02
0.04
0.06
0.08
0.1
12 10 8 6 4 2 0 2 4 6 8 10 12
C
m
(
N
.
D
.
)
elevon deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 31: MantaHawk VORSTAB Pitching Moment Coefficient due to Symmetrical
Deflection
These three longitudinal derivatives due to the elevator input have a value and
sign convention that is expected. A large symmetrical elevon deflection would
increase the magnitude of the pitching moment and as a result the lift and drag would
also increase. These trends are observed in the VORSTAB results.
70
Table 21: MantaHawk VORSTAB LateralDirectional Derivatives due to Asymmetrical
Deflections
o (deg) o
el
(deg) C
y
C
l
C
n
0 10 0.0062 0.0304 0.0010
5 10 0.0050 0.0301 0.0011
10 10 0.0021 0.0296 0.0024
15 10 0.0003 0.0287 0.0035
20 10 0.0020 0.0273 0.0043
0 5 0.0033 0.0152 0.0001
5 5 0.0023 0.0151 0.0003
10 5 0.0008 0.0149 0.0010
15 5 0.0003 0.0144 0.0015
20 5 0.0011 0.0137 0.0019
0 0 0.0000 0.0000 0.0000
5 0 0.0000 0.0000 0.0000
10 0 0.0000 0.0000 0.0000
15 0 0.0000 0.0000 0.0000
20 0 0.0000 0.0000 0.0000
0 5 0.0032 0.0152 0.0008
5 5 0.0017 0.0152 0.0001
10 5 0.0004 0.0149 0.0005
15 5 0.0005 0.0144 0.0009
20 5 0.0008 0.0129 0.0010
0 10 0.0059 0.0304 0.0021
5 10 0.0029 0.0304 0.0007
10 10 0.0006 0.0298 0.0004
15 10 0.0007 0.0279 0.0010
20 10 0.0010 0.0230 0.0008
71
0.008
0.006
0.004
0.002
0
0.002
0.004
0.006
0.008
10 5 0 5 10
C
y
(
N
.
D
.
)
Elevon deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 32: MantaHawk VORSTAB Sideforce Coefficient due to Asymmetrical
Deflection
0.04
0.03
0.02
0.01
0
0.01
0.02
0.03
0.04
10 5 0 5 10
C
l
(
N
.
D
.
)
Elevon deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 33: MantaHawk VORSTAB Rolling Moment Coefficient due to Asymmetrical
Deflection
72
0.005
0.004
0.003
0.002
0.001
0
0.001
0.002
10 5 0 5 10
C
n
(
N
.
D
.
)
Elevon deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 34: MantaHawk VORSTAB Yawing Moment Coefficient due to Asymmetrical
Deflection
A positive aileron/rudder input should result in a positive sideforce. At high
angles of attack this is not seen in the VORSTAB results. At extreme angles of attack
the forces and moments are usually not what is expected or desired, unless the aircraft
is designed to fly at these angles. The MantaHawk was not designed to fly at these
angles though. Similar to the sideforce, a positive asymmetrical elevon deflection
will result in a positive rolling moment. Typically, a positive rudder input will result
in a negative yawing, and a positive aileron input will result in a negative yawing
moment but can be positive. At low angles of attack the yawing moment is negative,
as expected, but at higher angles it becomes positive. This is most likely due to the
change in drag over the airfoils at extreme angles of attack. The result of this positive
73
value will actually help the intended turn. Table 22 shows the stability requirements
for the MantaHawk.
Table 22: MantaHawk VORSTAB Stability Requirements
Type of Stability
Corresponding
Derivatives
Criterion
Derivative
(rad
1
)
Stable/Unstable
Sideslip C
y
< 0 0.1146 Stable
Vertical Speed C
Lo
> 0 2.5500 Stable
Angle of Attack C
mo
< 0 0.1756 Unstable
Angle of Sideslip C
n
> 0 0.0172 Stable
Roll Rate C
lp
< 0 0.2866 Stable
Pitch Rate C
mq
< 0 1.6625 Stable
Yaw Rate C
nr
< 0 0.0077 Stable
Lateral C
l
< 0 0.0458 Stable
Like the YAK54, the MantaHawk‟s stability requirements were determined
at 0˚ angle of attack. At other angles of attack, the aircraft may become unstable in
modes other than angle of attack stability. The vertical speed and angle of attack
stability requirements were both found over the angle of attack range of 0˚ to 5˚ from
the VORSTAB results. As seen from the AAA data, it was predicted that the aircraft
was stable in all modes and VORSTAB predicted an unstable mode.
8.3 Method Comparison MantaHawk
The results from AAA and VORSTAB are different in the terms that C
mo
is
unstable as predicted by VORSTAB. AAA‟s results say that the MantaHawk is
stable in all modes. A comparison of the two methods can be seen in Table 23.
74
Table 23: MantaHawk AAA and VORSTAB Comparison
Stability and Control Derivatives
Derivative (1/rad)
AAA VORSTAB % Difference
C
Lo
3.9630 2.5500 35.65
C
Do
0.0615 0.0275 55.28
C
mo
0.4034 0.1756 143.53
C
Lq
5.0920 1.9645 61.42
C
mq
4.7340 1.6625 64.88
C
l
0.0817 0.0458 43.94
C
lp
0.4892 0.2866 41.41
C
lr
0.0407 0.0199 51.11
C
y
0.1877 0.1146 38.95
C
yp
0.1249 0.0539 56.85
C
yr
0.0897 0.0472 47.38
C
n
0.0359 0.0172 52.09
C
np
0.0169 0.0003 98.22
C
nr
0.0219 0.0077 64.84
C
Doel
0.4209 0.0168 96.01
C
Loel
0.4784 0.6703 40.11
C
moel
1.0630 0.2908 72.64
C
loel
0.3588 0.1745 51.37
C
yoel
0.0012 0.0363 2925.00
C
noel
0.0055 0.0093 69.09
VORSTAB produces results that are lower than most of the AAA derivatives.
This means that the aircraft is not going to have as much control as expected. Most of
the differences between AAA and VORSTAB fall within the error ranges that Dr.
Roskam describes that AAA will have. Pitching moment coefficient due to angle of
attack has the largest variance and this is the most critical derivative, since this was
75
the unstable mode. VORSTAB should be used as a method for checking if there is a
possibility for one of the modes to be unstable. Even if VORSTAB is incorrect this
will give designers a chance to plan the flight test with this possibility.
The same typical ranges that applied to the YAK54 apply to the MantaHawk
as well. These ranges again were developed by Dr. Roskam, Ref [5], for what
approximately an aircraft can expect the control derivatives to be. These ranges are
for conventional aircrafts, and the ranges are a function of Mach number. Table 24
shows the VORSTAB MantaHawk results compared to typical ranges from Dr.
Roskam, Ref [5].
76
Table 24: MantaHawk Stability and Control Derivatives Typical Ranges
Stability and Control Derivatives
Derivative
(1/rad)
VORSTAB AAA
Typical
Ranges
Within Range
(Yes/No)
Dr. Roskam VORSTAB AAA
C
Lo
2.55 3.963 1.0 to 8.0 Yes Yes
C
Do
0.0275 0.0615 0.0 to 2.0 Yes Yes
C
mo
0.1756 0.4034 4.0 to 1.0 No Yes
C
Lq
1.9645 5.092 0.0 to 30.0 Yes Yes
C
mq
1.6625 4.734 90.0 to 0.0 Yes Yes
C
l
0.0458 0.0817 0.1 to 4.0 Yes Yes
C
lp
0.2866 0.4892 0.1 to 0.8 Yes Yes
C
lr
0.0199 0.0407 0.0 to 0.6 Yes Yes
C
y
0.1146 0.1877 0.1 to 2.0 Yes Yes
C
yp
0.0539 0.1249 0.3 to 0.8 Yes Yes
C
yr
0.0472 0.0897 0.0 to 1.2 Yes Yes
C
n
0.0172 0.0359 0.0 to 4.0 Yes Yes
C
np
0.0003 0.0169 0.5 to 0.1 Yes Yes
C
nr
0.0077 0.0219 0.0 to 1.0 Yes Yes
C
Loe
0.6703 0.4784 0.0 to 0.6 No Yes
C
Doe
0.0168 0.4209 Negligible No No
C
moe
0.2908 1.0630 0.0 to 4.0 Yes Yes
C
loa
0.1745 0.3588 0.0 to 0.4 Yes Yes
C
lor
0.1745 0.3588 0.04 to 0.04 No Yes
C
yoa
0.0363 0.0012 Negligible No No
C
yor
0.0363 0.0012 0.0 to 0.5 Yes Yes
C
noa
0.0093 0.0055 0.08 to 0.08 Yes Yes
C
nor
0.0093 0.0055 0.15 to 0.0 Yes Yes
Not all of the control derivatives fall within the expected ranges. This does
not necessarily mean that the aircraft is unstable. These are estimated ranges and an
unconventional aircraft will not necessarily fall in the range. Also, the rudder
77
deflection and aileron deflection control derivatives are the same because there is
only one input for those control surfaces. The input simulates both a rudder input and
an aileron input. This means that the control derivatives might fall into the range for
the rudder or aileron deflection, but not the other deflection. For example, the rolling
moment coefficient due to aileron deflection falls into the typical range Dr. Roskam
gives, but it does not fall into the typical range for the rudder deflection. The drag
coefficient due to elevator deflection is usually ignored. Therefore, there is no cause
for alarm that there is an increase in drag when the control surfaces are deflected. As
stated earlier, the MantaHawk is unstable in the pitching mode, and this is observed
since the pitching moment coefficient due to change in angle of attack, C
mo
, does not
fall into the typical range Dr. Roskam describes.
8.4 Linearized Model of the MantaHawk
For the information and data on the state space model please refer to
Appendix B.
78
9 Meridian UAV
The Center for Remote Sensing of Ice Sheets (CReSIS) is a science and
technology center established by the National Science Foundation (NSF) in 2005.
The foundation has the mission of developing new technologies and computer models
to measure and predict the response of sea level change to the mass balance of ice
sheets in Greenland and Antarctica. CReSIS is developing sophisticated sensors and
a longduration uninhabitant aerial vehicle (UAV), called the Meridian.
The KUAE, in close collaboration with national and international partners, has
designed and developed the Meridian UAV for the Center for Remote Sensing of Ice
Sheets (CReSIS). The Meridian has a gross takeoff weight of 1,100 lbs, a wingspan
of 26.4 ft, and a range of 950 nautical miles. Vehicle cruise speed is 100120 kts,
endurance exceeds 9 hours, and available payload exceeds 120 lbs and 300W
consumed power. As such, this is an extraordinarily valuable scientific research
platform, since most civilian UAV research currently limits payloads to a few pounds
and a fraction of the power. Added benefits of greater than two cubic feet of payload
volume and eight wing hard points provide a sensor platform with little competition.
This data was taken from Ref [16]. Table 25 shows some of the important salient
characteristics and Figure 35 shows the design of the Meridian UAV both from Ref
[17]. Information about the design on the Meridian can also be found in Ref [17].
79
Table 25: Meridian Characteristics
Characteristic Wing VTail
Area 69.6 ft
2
10 ft
2
Span 26.4 ft 5.9 ft
MGC 31.7 ft 1.75 ft
Aspect Ratio 10 3.5
Quarter Chord Sweep Angle 0 deg 26.3 deg
Taper Ratio 1 0.5
Thickness Ratio 18% 12%
Dihedral Angle 5 deg 50 deg
Incidence Angle 0 deg 0 deg
Aileron Chord Ratio 0.24
Aileron Span Ratio 0.601.00
Flap Chord Ratio 0.2
Flap Span Ratio 0.10
0.599
Characteristic Fuselage
Maximum Length 14.8 ft
Maximum Height 2.9 ft
Maximum Width 3.5
Figure 35: Unigraphics CAD Model of the Meridian UAV
80
The design of the Meridian was completed using AAA and Dr. Roskam‟s
method. The AAA software was also used to find the stability and control derivatives
of the Meridian. The tail of the Meridian is unconventional and uses a vtail. It is
also a fully moving tail.
9.1 AAA Modeling of the Meridian
Similar to the YAK54 and MantaHawk, the AAA file produced stability and
control derivatives for the Meridian UAV. The longitudinal derivatives are found in
Table 26 and the lateraldirectional derivatives are found in Table 27, both Ref [18].
As seen in Table 28 the aircraft is found to be stable in all modes.
81
Table 26: Meridian AAA Longitudinal Derivatives
Longitudinal Coefficients and Stability DerivativesStability
Axes
Derivatives (1/rad)
C
Du
0.0000
C
Do
0.1409
C
Txu
0.0713
C
Lu
0.0109
C
Lo
5.1648
C
Lodot
0.7407
C
Lq
4.6179
C
mu
0.0028
C
mo
0.6207
C
modot
2.9826
C
mq
13.973
C
mTu
0.0294
C
mTo
0.3419
Longitudinal Control and Hinge Moment Derivatives
Stability Axes
Derivatives (1/rad)
C
Dore
0.0117
C
Lore
0.4149
C
more
1.6709
82
Table 27: Meridian AAA LateralDirectional Derivatives
LateralDirectional Coefficients and Stability DerivativesStability
Axes
Derivatives (1/rad)
C
l
0.0776
C
lp
0.5546
C
lr
0.1099
C
y
0.4789
C
yp
0.1465
C
yr
0.3217
C
n
0.1386
C
nT
0.0007
C
np
0.0351
C
nr
0.1338
LateralDirectional Control and Hinge Moment Derivatives
Stability Axes
Derivatives (1/rad)
C
loa
0.2316
C
lore
0.0253
C
yoa
0.0000
C
yore
0.3681
C
noa
0.0134
C
nopc
0.1481
83
Table 28: Meridian AAA Stability Requirements
Type of Stability
Corresponding
Derivatives
Criterion
Derivative
(rad
1
)
Stable/Unstable
Sideslip C
y
< 0 0.4789 Stable
Vertical Speed C
Lo
> 0 5.1648 Stable
Angle of Attack C
mo
< 0 0.6207 Stable
Angle of Sideslip C
n
> 0 0.1386 Stable
Roll Rate C
lp
< 0 0.5546 Stable
Pitch Rate C
mq
< 0 13.973 Stable
Yaw Rate C
nr
< 0 0.1338 Stable
Lateral C
l
< 0 0.0776 Stable
9.2 VORSTAB Modeling of the Meridian
Several different Meridian VORSTAB models were created with changes to
the geometry of the fuselage. The reasons for these changes were due to the fact that
VORSTAB calculates such a high downwash as a result of the shape and camber of
the fuselage, and these changes to the model were to improve the results. Typically,
the fuselage does not result in a large amount of downwash and can almost be
completely ignored. The downwash on the Meridian fuselage is due to the large
amount of convergence from the nose to the tail and the negative camber or
downward swoop of the empennage. From this downwash, the effectiveness of the
tails is almost blanketed out. In reality, there is not as much downwash as
VORSTAB is predicting due to the fact that the flow will separate from the fuselage.
9.2.1 Model 1: Half Scale Model of Meridian
The first model created was of the exact shape of the entire Meridian, but
scaled down to half of the size. This is a requirement for VORSTAB that the largest
84
radius of the fuselage be no larger than 1, for any unit system. The downwash was
so large for this model that VORSTAB did not obtain any valid results. Table 29
shows the longitudinal derivatives obtained, and Table 30 shows the lateral
directional derivatives. As seen from the results, there is negative lift and zero drag at
multiple angles of attack. These results were known to be invalid.
Table 29: Meridian VORSTAB Model 1 Longitudinal Derivatives (1/rad)
o (deg) C
L
C
D
C
m
C
Lq
C
mq
5 2.863 0.1402 0.9534 33.592 17.037
0 3.1363 0 1.3823 34.603 17.109
5 3.6006 0 0.2483 35.110 17.069
10 2.8592 0 1.5042 35.041 16.463
15 2.882 0 1.1976 34.609 16.118
20 2.6188 0 1.5477 33.746 15.799
Table 30: Meridian VORSTAB Model 1 LateralDirectional Derivatives (1/rad)
o
(deg)
C
y
C
l
C
n
C
yp
C
lp
C
np
C
yr
C
lr
C
nr
5 17.49 16.48 10.04 3.735 6.656 2.925 2.917 1.700 1.357
0 1.203 2.280 2.412 0.215 1.161 1.129 4.541 0.659 2.095
5 0.588 0.655 2.364 1.738 0.581 2.120 6.194 0.602 2.787
10 9.515 2.671 2.697 1.041 0.041 1.589 8.032 0.489 3.677
15 12.611 3.054 3.485 2.411 0.072 2.185 10.99 0.510 5.079
20 12.672 3.177 3.183 2.825 0.052 2.210 13.07 0.523 5.717
After several discussions with Dr. Lan about the possible cause for the
problem, he offered advice on what he believed to be the problem and how it could
possibly be fixed. VORSTAB predicts that all of the flow over the top of the fuselage
will become downwash, but in reality not all of it will. There will be some flow
separation that does not result in downwash. The vtail are in the path of this flow
wake and rendering them ineffective. In the VORSTAB output file, the downwash
85
on the fuselage can be seen as extremely high pressure on the upper surface. Dr. Lan
recommended changing the fuselage to help eliminate some of the downwash and
obtain valid results. The VORSTAB input file can be found in Appendix A.
9.2.2 Model 2: Larger Empennage
The best way to help eliminate some of the downwash was to eliminate the
large amount of convergence, and also, to estimate the crosssections of the fuselage
with a circular crosssection. This method replaced the fuselage body with a fuselage
wake surface. Having circular crosssections allowed for much faster modifications
to the model. The forward section of the fuselage size did not change and the aft
sections radius was increased. The size of the increase of the aft fuselage sections
was changed multiple times. The final increase was to approximately half of the
forward section size. Table 31 shows the diameter used for this approximation as
well as the right side view of the fuselage diameter for the first model. Both models
are to half scale. Figure 36 shows the increase in size of the empennage.
86
Table 31: Fuselage Diameter for Meridian Models 1 and 2
Station x (ft)
Model 1
Diameter (ft)
Model 2
Diameter (ft)
1 0 0 0
2 0.3458 0.3225 0.3225
3 0.75 0.5183 0.5183
4 0.7917 0.5183 0.5183
5 1.042 0.9879 0.9879
6 1.321 1.238 1.238
7 1.608 1.317 1.317
8 2.058 1.375 1.375
9 2.438 1.388 1.388
10 2.883 1.376 1.376
11 3.521 1.292 1.292
12 4.146 1.118 1.118
13 4.642 0.9458 0.9458
14 5.25 0.7542 0.7542
15 5.771 0.6125 0.7542
16 6.592 0.4525 0.7542
17 7.463 0.4121 0.7542
18 8.125 0.3325 0.6708
19 8.563 0.0417 0.5042
20 8.604 0 0.2042
1
0
1
5 5.5 6 6.5 7 7.5 8 8.5 9
X (ft)
Z
(
f
t
)
Increased Upper Emp
Increased Lower Emp
Original Upper Emp
Origianl Lower Emp
Figure 36: Meridian Empennage Models 1 and 2
87
Model 2‟s results were much better than the first model, but the drag was very
high. The high drag was a result of a few different things. Increasing the fuselage
size added some to the drag. Also, there is still camber in the empennage that results
in some downwash that increases the drag. For this model the drag should be
ignored. Like all VORSTAB models this model does not show that the aircraft stalls,
because there is no flow separation. The input file for this model can be found in
Appendix A. Table 32 shows the longitudinal derivatives and Figure 37 through
Figure 40 show Table 32 in graphical form.
Table 32: Meridian VORSTAB Model 2 Longitudinal Control Derivatives and Affected
by Pitch Rate
o (deg) C
L
C
D
C
m
C
L,q
C
D,q
C
m,q
5 0.5876 0.4912 0.0352 6.9660 0.0000 8.4348
0 0.0978 1.4931 0.1877 7.1682 0.0000 8.5662
5 0.5335 2.8528 0.2621 7.2364 0.0000 8.5967
10 0.9949 4.6480 0.3414 7.1232 0.0000 8.3508
15 1.3140 6.6175 0.4179 6.7959 0.0000 7.6702
20 1.6382 8.6688 0.4923 6.5872 0.0000 7.8880
88
1
0.5
0
0.5
1
1.5
2
5 0 5 10 15 20
C
L
(
N
.
D
.
)
o (deg)
Figure 37: Meridian VORSTAB Model 2 Lift Coefficient due to Angle of Attack
0
1
2
3
4
5
6
7
8
9
10
5 0 5 10 15 20
C
D
(
N
.
D
.
)
o (deg)
Figure 38: Meridian VORSTAB Model 2 Drag Coefficient due to Angle of Attack
89
0.6
0.5
0.4
0.3
0.2
0.1
0
5 0 5 10 15 20
C
M
(
N
.
D
.
)
o (deg)
Figure 39: Meridian VORSTAB Model 2 Pitching Moment Coefficient due to Angle of
Attack
10
8
6
4
2
0
2
4
6
8
10
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
clq cdq cmq
Figure 40: Meridian VORSTAB Model 2 Longitudinal Derivatives due to Pitch Rate
90
The lift coefficient is appropriate in magnitude, but the zero lift angle of attack
is positive. For this type of aircraft, the zero lift angle of attack should be negative.
This lift would result in the aircraft cruising at a positive angle of attack. It is obvious
that the drag is too high. All of the other derivatives have an appropriate value and
trend. For example, the pitching moment coefficient shows the aircraft is inherently
stable.
The lateraldirectional control derivatives due to sideslip and roll rate for
Model 2 can be seen in Table 33. The graphical form of this table is shown in Figure
41 and Figure 42. Table 34 shows the lateraldirectional control derivatives due to
yaw rate, and Figure 43 shows this in graphical form.
Table 33: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to Sideslip
and Roll Rate
o (deg) C
y,
C
l,
C
n,
C
y,p
C
l,p
C
n,p
5 0.5775 0.1071 0.0877 0.1552 0.4765 0.2220
0 0.7695 0.1083 0.0991 0.1010 0.2913 0.0312
5 0.9597 0.1283 0.0837 0.1356 0.4538 0.5138
10 0.9603 0.1203 0.0493 0.0569 0.2312 0.3247
15 0.7403 0.1146 0.0258 0.0288 0.1242 0.2432
20 0.4882 0.1077 0.0029 0.0167 0.0801 0.2357
91
0.2
0
0.2
0.4
0.6
0.8
1
1.2
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
Cyb Clb Cnb
Figure 41: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to Sideslip
0.6
0.5
0.4
0.3
0.2
0.1
0
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
cyp clp cnp
Figure 42: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to Roll
Rate
92
Usually, the sideforce coefficient due to sideslip is negative, but the
VORSTAB results show a positive value. Sideforce coefficient due to roll rate has a
positive and negative typical range, but it is usually negative due to the moment arm.
This is seen in the VORSTAB results. The rest of the derivatives are as expected.
Table 34: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to Yaw
Rate
o (deg) C
y,r
C
l,r
C
n,r
5 0.1537 0.2550 0.0949
0 0.2091 0.2263 0.0748
5 0.2432 0.1840 0.0483
10 0.2017 0.1126 0.0030
15 0.1741 0.0672 0.0244
20 0.1548 0.0429 0.0549
0.3
0.2
0.1
0
0.1
0.2
0.3
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
cyr clr cnr
Figure 43: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to Yaw
Rate
93
The lateraldirectional derivatives due to yaw rate have values that are
expected, except for the rolling moment coefficient. This derivative is normally a
positive value, since the wingfuselage contribution is always positive and outweighs
the vertical tail contribution. The vertical tail contribution can be either positive or
negative. This vertical tail contribution outweighs the wingfuselage contribution for
this aircraft since it has a negative value.
Deflecting the control surfaces affects the stability and control derivatives.
Table 35 through Table 37 show the control derivatives due to different control
surface deflections. These three tables are also graphical depicted in Figure 44
through Figure 52. The Meridian has a vtail so there is not a conventional rudder or
elevators. The vtail imitates a rudder input with an asymmetrical deflection of the
tails and imitates an elevator input with a symmetrical deflection of the tails. These
types of control surfaces are called rudevators.
94
Table 35: Meridian VORSTAB Model 2 Lateral Directional Derivatives due to Aileron
Deflection
o (deg) o
a
C
y
C
l
C
n
0 10 0.0099 0.0460 0.0121
5 10 0.0100 0.0462 0.0131
10 10 0.0091 0.0417 0.0122
15 10 0.0066 0.0306 0.0087
20 10 0.0048 0.0222 0.0061
0 5 0.0050 0.0230 0.0060
5 5 0.0050 0.0231 0.0066
10 5 0.0045 0.0209 0.0061
15 5 0.0033 0.0153 0.0043
20 5 0.0024 0.0111 0.0031
0 0 0.0000 0.0000 0.0000
5 0 0.0000 0.0000 0.0000
10 0 0.0000 0.0000 0.0000
15 0 0.0000 0.0000 0.0000
20 0 0.0000 0.0000 0.0000
0 5 0.0049 0.0229 0.0060
5 5 0.0050 0.0230 0.0066
10 5 0.0045 0.0208 0.0061
15 5 0.0033 0.0152 0.0043
20 5 0.0024 0.0111 0.0031
0 10 0.0098 0.0456 0.0121
5 10 0.0100 0.0458 0.0131
10 10 0.0090 0.0413 0.0122
15 10 0.0066 0.0303 0.0087
20 10 0.0047 0.0220 0.0061
95
0.015
0.01
0.005
0
0.005
0.01
0.015
10 5 0 5 10
C
y
(
N
.
D
.
)
aileron deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 44: Meridian VORSTAB Model 2 Sideforce Coefficient due to Aileron Deflection
0.06
0.04
0.02
0
0.02
0.04
0.06
10 5 0 5 10
C
l
(
N
.
D
.
)
aileron deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 45: Meridian VORSTAB Model 2 Rolling Moment Coefficient due to Aileron
Deflection
96
0.015
0.01
0.005
0
0.005
0.01
0.015
10 5 0 5 10
C
n
(
N
.
D
.
)
aileron deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 46: Meridian VORSTAB Model 2 Yawing Moment Coefficient due to Aileron
Deflection
The sideforce coefficient due to aileron deflection can be normally be
neglected, unless the rolling moment controls are near a vertical surface. This is not
the case for the Meridian, but the contribution is very small and almost negligible.
VORSTAB does not assume that it should be zero, so it will always calculate a value.
Also, the yawing moment coefficient is usually negative since is tends to yaw the
aircraft out of an intended turn. VORSTAB shows that this derivative actually shows
that it will yaw the aircraft into the intended turn. As expected, a positive aileron
deflection produces a positive rolling moment.
97
Table 36: Meridian VORSTAB Model 2 Longitudinal Derivatives Affected by
Symmetrical Tail Deflections
o (deg) o
re
C
L
C
D
C
m
0 12 0.1130 1.6814 0.1278
5 12 0.5193 3.1227 0.2052
10 12 0.9911 4.8756 0.3255
15 12 1.2988 6.8064 0.3564
20 12 1.6229 8.9060 0.4301
0 6 0.1009 1.5270 0.1756
5 6 0.5381 2.9049 0.2805
10 6 0.9918 4.6793 0.3290
15 6 1.3111 6.6600 0.4062
20 6 1.6353 8.7221 0.4806
0 0 0.0916 1.3830 0.2122
5 0 0.5431 2.7613 0.3009
10 0 1.0035 4.5399 0.3764
15 0 1.3224 6.4777 0.4515
20 0 1.6465 8.4944 0.5257
0 6 0.0814 1.2854 0.2526
5 6 0.5545 2.6096 0.3462
10 6 1.0140 4.3693 0.4186
15 6 1.3327 6.2676 0.4928
20 6 1.6567 8.2347 0.5659
0 12 0.0733 1.1596 0.2847
5 12 0.5645 2.4716 0.3862
10 12 1.0234 4.1793 0.4564
15 12 1.3422 6.0375 0.5303
20 12 1.6660 7.9323 0.6017
98
0.4
0.2
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
12 10 8 6 4 2 0 2 4 6 8 10 12
C
L
(
N
.
D
.
)
elevator deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 47: Meridian VORSTAB Model 2 Lift Coefficient due to Elevator Input
0
1
2
3
4
5
6
7
8
9
10
12 10 8 6 4 2 0 2 4 6 8 10 12
C
D
(
N
.
D
.
)
elevator deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 48: Meridian VORSTAB Model 2 Drag Coefficient due to Elevator Input
99
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
12 10 8 6 4 2 0 2 4 6 8 10 12
C
m
(
N
.
D
.
)
elevator deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 49: Meridian VORSTAB Model 2 Pitching Moment Coefficient due to Elevator
Input
A positive elevator input, shows increasing pitching moment and lift. This is
expected since the elevators are used to increase the pitch and create lift. As stated
earlier the drag is going to be neglected because of its high magnitude.
100
Table 37: Model 2 Lateral Directional Control Derivatives Affected by Asymmetrical
Tail Deflection
o (deg) o
re
C
y
C
l
C
n
0 10 0.0236 0.0019 0.0102
5 10 0.0203 0.0022 0.0089
10 10 0.0162 0.0022 0.0074
15 10 0.0116 0.0022 0.0056
20 10 0.0101 0.0023 0.0051
0 5 0.0114 0.0010 0.0049
5 5 0.0103 0.0011 0.0045
10 5 0.0082 0.0011 0.0038
15 5 0.0059 0.0011 0.0028
20 5 0.0051 0.0011 0.0026
0 0 0.0000 0.0000 0.0000
5 0 0.0000 0.0000 0.0000
10 0 0.0000 0.0000 0.0000
15 0 0.0000 0.0000 0.0000
20 0 0.0000 0.0000 0.0000
0 5 0.0114 0.0010 0.0049
5 5 0.0103 0.0011 0.0045
10 5 0.0082 0.0011 0.0038
15 5 0.0059 0.0011 0.0028
20 5 0.0051 0.0011 0.0026
0 10 0.0220 0.0021 0.0094
5 10 0.0203 0.0022 0.0089
10 10 0.0162 0.0022 0.0074
15 10 0.0116 0.0022 0.0056
20 10 0.0101 0.0023 0.0051
101
0.03
0.02
0.01
0
0.01
0.02
0.03
10 5 0 5 10
C
y
(
N
.
D
.
)
tail deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 50: Meridian VORSTAB Model 2 Sideforce Coefficient due to Rudder Input
0.0025
0.002
0.0015
0.001
0.0005
0
0.0005
0.001
0.0015
0.002
0.0025
10 5 0 5 10
C
l
(
N
.
D
.
)
tail deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 51: Meridian VORSTAB Model 2 Rolling Moment Coefficient due to Rudder
Input
102
0.015
0.01
0.005
0
0.005
0.01
0.015
10 5 0 5 10
C
n
(
N
.
D
.
)
tail deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 52: Meridian VORSTAB Model 2 Yawing Moment Coefficient due to Rudder
Input
A positive rudder input will produce a positive sideforce, and a positive
sideforce will usually generate a positive rolling moment. This type of rudder
deflection will also generate a negative yawing moment. This is all seen in the
VORSTAB results.
9.2.3 Model 3: Zero Camber Fuselage
The third model created was the same as the second model, but the camber
was taken out of the fuselage. Removing the camber removed the amount of
downwash that was produced from the large negative camber in the empennage.
Having zero camber in the fuselage allows for the assumption that the downwash
produced by the fuselage is negligible. The input file for this model can be found in
Appendix A. Table 38 shows the longitudinal control derivatives and Figure 53
103
through Figure 56 depicts Table 38 in graphical form. Table 39 shows the lateral
directional control derivatives and Figure 57 through Figure 60 depicts Table 39 in
graphical form.
Table 38: Meridian VORSTAB Model 3 Longitudinal Derivatives at due to Pitch Rate
o (deg) C
L
C
D
C
m
C
L_q
C
D_q
C
m_q
5 0.7157 1.6110 0.2014 6.9744 0.0000 8.4373
0 0.2017 0.6312 0.0819 7.1767 0.0000 8.5688
5 0.4238 0.0982 0.0467 7.2448 0.0000 8.5992
10 0.9047 0.0703 0.0596 7.1331 0.0000 8.3595
15 1.2318 0.3564 0.1686 6.8019 0.0000 7.6632
20 1.5543 1.1494 0.2437 6.5935 0.0000 7.8832
1
0.5
0
0.5
1
1.5
2
5 0 5 10 15 20
C
L
(
N
.
D
.
)
o (deg)
Figure 53: Meridian VORSTAB Model 3 Lift Coefficient due to Angle of Attack
104
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
5 0 5 10 15 20
C
D
(
N
.
D
.
)
o (deg)
Figure 54: Meridian VORSTAB Model 3 Drag Coefficient due to Angle of Attack
0.3
0.25
0.2
0.15
0.1
0.05
0
0.05
0.1
0.15
0.2
0.25
5 0 5 10 15 20
C
M
(
N
.
D
.
)
o (deg)
Figure 55: Meridian VORSTAB Model 3 Pitching Moment Coefficient due to Angle of
Attack
105
10
8
6
4
2
0
2
4
6
8
10
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
CL_q CD_q CM_q
Figure 56: Meridian VORSTAB Model 3 Longitudinal Coefficients due to Pitch Rate
Similar to Model 2, these longitudinal derivatives are within the appropriate
range and have the expected sign convention. The drag is still high, but is much
better than the second model. Lift is negative at zero degree angle of attack, and
should be positive.
Table 39: Meridian VORSTAB Model 3 LateralDirectional Derivatives
o (deg) C
y
C
l
C
n
C
yp
C
lp
C
np
C
yr
C
lr
C
nr
5 0.668 0.104 0.051 0.161 0.476 0.213 0.067 0.260 0.062
0 0.823 0.103 0.077 0.107 0.292 0.024 0.126 0.232 0.046
5 0.994 0.123 0.071 0.145 0.466 0.523 0.186 0.195 0.032
10 0.941 0.119 0.069 0.059 0.236 0.328 0.223 0.118 0.022
15 0.708 0.116 0.056 0.030 0.126 0.244 0.213 0.069 0.005
20 0.472 0.108 0.032 0.018 0.081 0.235 0.183 0.045 0.026
106
0.005
0
0.005
0.01
0.015
0.02
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
Cy_b Cl_b Cn_b
Figure 57: Meridian VORSTAB Model 3 LateralDirectional Derivatives due to Sideslip
0.6
0.5
0.4
0.3
0.2
0.1
0
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
Cy_p Cl_p Cn_p
Figure 58: Meridian VORSTAB Model 3 LateralDirectional Derivatives due to Roll
Rate
107
0.3
0.2
0.1
0
0.1
0.2
0.3
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
Cy_r Cl_r Cn_r
Figure 59: Meridian VORSTAB Model 3 LateralDirectional Derivatives due to Yaw
Rate
The sideforce coefficient due to sideslip is usually negative, but the
VORSTAB results show a positive value. Sideforce coefficient due to roll rate has a
positive and negative typical range, but it is usually negative due to the moment arm.
This is seen in the VORSTAB results. The rest of the derivatives are as expected.
The rolling moment and yawing moment coefficients due to roll rate have extreme
fluctuations from one angle of attack to another. This does not necessarily mean that
it is wrong, but rather the angle of attack has a large influence on the control of the
aircraft.
The control surface deflections affect the stability and control derivatives in
both the longitudinal mode and the lateraldirectional mode. Table 40 shows the
affects of the aileron on the lateraldirectional derivatives, Table 41 shows the affect
of a symmetrical tail deflection on the longitudinal derivatives, and Table 42 shows
108
the affect of an asymmetrical tail deflection on the lateraldirectional derivatives.
These tables are depicted in Figure 60 through Figure 68.
Table 40: Meridian VORSTAB Model 3 LateralDirectional Derivative due to Aileron
Deflection
o (deg) o
a
(deg) C
y
C
l
C
n
0 10 0.0099 0.0460 0.0120
5 10 0.0099 0.0462 0.0131
10 10 0.0090 0.0420 0.0124
15 10 0.0066 0.0310 0.0089
20 10 0.0048 0.0225 0.0063
0 5 0.0050 0.0230 0.006
5 5 0.0050 0.0231 0.0066
10 5 0.0045 0.0210 0.0062
15 5 0.0033 0.0155 0.0044
20 5 0.0024 0.0113 0.0031
0 0 0.0000 0.0000 0.0000
5 0 0.0000 0.0000 0.0000
10 0 0.0000 0.0000 0.0000
15 0 0.0000 0.0000 0.0000
20 0 0.0000 0.0000 0.0000
0 5 0.0050 0.02291 0.006
5 5 0.0050 0.02302 0.00657
10 5 0.0045 0.02095 0.00619
15 5 0.0033 0.01544 0.00442
20 5 0.0024 0.01121 0.00313
0 10 0.0098 0.04558 0.012
5 10 0.0099 0.04581 0.01315
10 10 0.0090 0.04169 0.01239
15 10 0.0066 0.03071 0.00885
20 10 0.0047 0.02231 0.00627
109
0.015
0.01
0.005
0
0.005
0.01
0.015
10 5 0 5 10
C
y
(
N
.
D
.
)
aileron deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 60: Meridian VORSTAB Model 3 Sideforce Coefficient due to Aileron Deflection
0.06
0.04
0.02
0
0.02
0.04
0.06
10 5 0 5 10
C
l
(
N
.
D
.
)
aileron deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 61: Meridian VORSTAB Model 3 Rolling Moment Coefficient due to Aileron
Deflection
110
0.015
0.01
0.005
0
0.005
0.01
0.015
10 5 0 5 10
C
n
(
N
.
D
.
)
aileron deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 62: Meridian VORSTAB Model 3 Yawing Moment Coefficient due to Aileron
Deflection
A sideforce coefficient due to aileron deflection is usually negligible, and
VORSTAB shows a very small value. This is expected since there are no rolling
moment controls close to a vertical surface. Positive aileron deflections result in a
positive rolling moment and yawing moment.
111
Table 41: Meridian VORSTAB Model 3 Longitudinal Derivatives due to Symmetrical
Tail Deflections
o (deg) o
re
(deg) C
L
C
D
C
m
0 12 0.2132 0.5057 0.0970
5 12 0.4096 0.0814 0.1028
10 12 0.8894 0.0849 0.0019
15 12 1.2173 0.4285 0.1098
20 12 1.5409 1.2620 0.1896
0 6 0.2042 0.5916 0.0916
5 6 0.4230 0.0948 0.0496
10 6 0.9016 0.0730 0.0474
15 6 1.2290 0.3648 0.1572
20 6 1.5518 1.1718 0.2335
0 0 0.1932 0.6789 0.0488
5 0 0.4349 0.1091 0.0024
10 0 0.9135 0.0631 0.0956
15 0 1.2400 0.3106 0.2016
20 0 1.5616 1.0614 0.2731
0 6 0.1836 0.8198 0.0111
5 6 0.4476 0.1225 0.0482
10 6 0.9250 0.0564 0.1420
15 6 1.2500 0.2607 0.2421
20 6 1.5706 0.9528 0.3088
0 12 0.1707 0.9102 0.0400
5 12 0.4568 0.1337 0.0935
10 12 0.9356 0.0534 0.1847
15 12 1.2591 0.1949 0.2786
20 12 1.5787 0.8708 0.3404
112
0.4
0.2
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
12 10 8 6 4 2 0 2 4 6 8 10 12
C
L
(
N
.
D
.
)
tail deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 63: Meridian VORSTAB Model 3 Lift Coefficient due to Elevator Input
0
0.2
0.4
0.6
0.8
1
1.2
1.4
12 10 8 6 4 2 0 2 4 6 8 10 12
C
D
(
N
.
D
.
)
tail deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 64: Meridian VORSTAB Model 3 Drag Coefficient at due to Elevator Input
113
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
0.05
0.1
0.15
12 10 8 6 4 2 0 2 4 6 8 10 12
C
M
(
N
.
D
.
)
tail deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 65: Meridian VORSTAB Model 3 Pitching Moment Coefficient due to Elevator
Input
Elevator input deflection will increase the lift and pitching moment. This is
the purpose of the positive elevator input. The drag coefficient increases on some
angles of attack and decreases on the others. When the lift increases the drag should
always increase. This is due to the definition of drag being a function of lift.
VORSTAB shows that at angles of attack of 10˚ or higher the drag will decrease with
positive elevator deflections. As stated previously, the drag is much higher than it
should be and is not accurate.
114
Table 42: Meridian Model 3 LateralDirectional Derivatives Affected by Asymmetrical
Tail Deflections
o (deg) o
re
(deg) C
y
C
l
C
n
0 10 0.0215 0.0021 0.0083
5 10 0.0226 0.0020 0.0092
10 10 0.0229 0.0018 0.0098
15 10 0.0162 0.0020 0.0070
20 10 0.0145 0.0022 0.0064
0 5 0.0109 0.0011 0.0042
5 5 0.0117 0.0010 0.0048
10 5 0.0109 0.0010 0.0046
15 5 0.0082 0.0010 0.0035
20 5 0.0073 0.0011 0.0033
0 0 0.0000 0.0000 0.0000
5 0 0.0000 0.0000 0.0000
10 0 0.0000 0.0000 0.0000
15 0 0.0000 0.0000 0.0000
20 0 0.0000 0.0000 0.0000
0 5 0.0109 0.0011 0.0042
5 5 0.0123 0.0009 0.0051
10 5 0.0109 0.0010 0.0046
15 5 0.0082 0.0010 0.0035
20 5 0.0073 0.0011 0.0033
0 10 0.0215 0.0021 0.0083
5 10 0.0239 0.0019 0.0098
10 10 0.0209 0.0020 0.0088
15 10 0.0162 0.0020 0.0070
20 10 0.0145 0.0022 0.0064
115
0.03
0.02
0.01
0
0.01
0.02
0.03
10 5 0 5 10
C
y
(
N
.
D
.
)
tail deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 66: Meridian VORSTAB Model 3 Sideforce Coefficient due to Rudder Input
0.0025
0.002
0.0015
0.001
0.0005
0
0.0005
0.001
0.0015
0.002
0.0025
10 5 0 5 10
C
l
(
N
.
D
.
)
tail deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 67: Meridian VORSTAB Model 3 Rolling Moment Coefficient due to Rudder
Input
116
0.015
0.01
0.005
0
0.005
0.01
0.015
10 5 0 5 10
C
n
(
N
.
D
.
)
tail deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 68: Meridian VORSTAB Model 3 Yawing Moment Coefficient due to Rudder
Input
As expected, a positive rudder input results in a positive sideforce and rolling
moment. It also results in a negative yawing moment.
9.2.4 Model 4: Wings and VTail
Since the problem with the downwash came from the fuselage, it was
removed from the forth model. This type of model helps the designer see what effect
the fuselage is having on the results. The wings and the vtail were kept in their exact
location but were extended to the centerline. Appendix A contains the input file for
this model. Table 43 shows the longitudinal control derivatives, and Figure 69
through Figure 72 depicts Table 43 in graphical form.
117
Table 43: Meridian VORSTAB Model 4 Longitudinal Derivatives
o (deg) C
L
C
D
C
m
C
Lq
C
Dq
C
mq
5 0.7969 0.0012 0.0309 3.6132 0.0000 3.7654
0 0.3916 0.0155 0.0995 3.7241 0.0000 3.7942
5 0.0736 0.0244 0.1135 3.7503 0.0000 3.7654
10 0.6302 0.0202 0.1031 3.6904 0.0000 3.6798
15 0.9934 0.0431 0.1256 3.5476 0.0000 3.5400
20 1.2843 0.0581 0.1517 3.3238 0.0000 3.3255
1
0.5
0
0.5
1
1.5
5 0 5 10 15 20
C
L
(
N
.
D
.
)
o (deg)
Figure 69: Meridian VORSTAB Model 4 Lift Coefficient due to Angle of Attack
118
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
5 0 5 10 15 20
C
D
(
N
.
D
.
)
o (deg)
Figure 70: Meridian VORSTAB Model 4 Drag Coefficient due to Angle of Attack
0.16
0.14
0.12
0.1
0.08
0.06
0.04
0.02
0
5 0 5 10 15 20
C
M
(
N
.
D
.
)
o (deg)
Figure 71: Meridian VORSTAB Model 4 Pitching Moment Coefficient due to Angle of
Attack
119
5
4
3
2
1
0
1
2
3
4
5
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
CL_q CD_q CM_q
Figure 72: Meridian VORSTAB Model 4 Longitudinal Derivatives due to Pitch Rate
As seen from the pitching moment graph, the wings and tail become unstable
at a certain angle of attack. This is seen by the slope of Figure 71 becoming positive.
With a positive slope the aircraft wants pitch away from the stable point. The drag
and lift results are as expected, except for the angle of zero lift being around 4˚. With
the airfoils used on the wings and tail and their geometric configuration it would be
expected that the angle of zero lift would be at a negative angle of attack. There
could be many causes for these results. A possible cause is the high camber of the
airfoils is producing downwash. From these results, it can be seen that the fuselage
contributes significantly to the drag in the VORSTAB model. The lateraldirectional
control derivatives are shown in Table 44. Figure 73 through Figure 75 depicts Table
44 in graphical form.
120
Table 44: Meridian VORSTAB Model 4 LateralDirectional Control Derivatives
o
(deg)
C
y
C
l
C
n
C
yp
C
lp
C
np
C
yr
C
lr
C
nr
5 0.009 0.0021 0.0013 0.1703 0.473 0.1837 0.1107 0.252 0.087
0 0.011 0.0019 0.0016 0.1229 0.296 0.0226 0.1291 0.232 0.067
5 0.014 0.0024 0.0005 0.1719 0.479 0.4940 0.1475 0.198 0.031
10 0.016 0.0022 0.0006 0.1145 0.311 0.3801 0.1561 0.132 0.003
15 0.014 0.0019 0.0006 0.0587 0.140 0.2353 0.1528 0.071 0.013
20 0.011 0.0017 0.0005 0.0435 0.090 0.2169 0.1461 0.043 0.029
0.2
0
0.2
0.4
0.6
0.8
1
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
Cy_b Cl_b Cn_b
Figure 73: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to Sideslip
121
0.6
0.5
0.4
0.3
0.2
0.1
0
0.1
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
Cy_p Cl_p Cn_p
Figure 74: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to Roll
Rate
0.3
0.25
0.2
0.15
0.1
0.05
0
0.05
0.1
0.15
0.2
5 0 5 10 15 20
(
1
/
r
a
d
)
o (deg)
Cy_r Cl_r Cn_r
Figure 75: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to Yawing
Moment
122
The sideforce coefficient due to sideslip, C
y
, is unstable at all angles of
attack. This can be neglected because having the fuselage will dramatically affect
this. The yawing moment coefficient due to yaw rate, C
nr
, is also unstable at higher
angles of attack. A stable aircraft should naturally want to stop yawing or slow down.
When it is unstable the yawing rate increases and can lead to a spin. This will also be
influenced by the fuselage and could be stable if it were present. Table 45 shows the
lateraldirectional derivatives due to an aileron deflection, Table 46 shows the
longitudinal derivatives due to a symmetrical deflection of the rudevators, and Table
47 shows the lateraldirectional derivatives due to an asymmetrical deflection of the
rudevators. These tables can be seen in graphical form in Figure 76 through Figure
84.
123
Table 45: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to Aileron
Deflection
o (deg) o
a
(deg) C
y
C
l
C
n
0 10 0.0102 0.0456 0.0117
5 10 0.0103 0.0458 0.0127
10 10 0.0100 0.0441 0.0131
15 10 0.0075 0.0336 0.0098
20 10 0.0055 0.0247 0.0071
0 5 0.0051 0.0228 0.0059
5 5 0.0052 0.0229 0.0064
10 5 0.0050 0.0221 0.0065
15 5 0.0037 0.0168 0.0049
20 5 0.0027 0.0124 0.0035
0 0 0.0000 0.0000 0.0000
5 0 0.0000 0.0000 0.0000
10 0 0.0000 0.0000 0.0000
15 0 0.0000 0.0000 0.0000
20 0 0.0000 0.0000 0.0000
0 5 0.0051 0.0227 0.0059
5 5 0.0051 0.0228 0.0064
10 5 0.0050 0.0220 0.0065
15 5 0.0037 0.0167 0.0049
20 5 0.0027 0.0123 0.0035
0 10 0.0101 0.0452 0.0117
5 10 0.0102 0.0454 0.0127
10 10 0.0099 0.0437 0.0131
15 10 0.0074 0.0333 0.0098
20 10 0.0054 0.0245 0.0071
124
0.015
0.01
0.005
0
0.005
0.01
0.015
10 5 0 5 10
C
y
(
N
.
D
.
)
aileron deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 76: Meridian VORSTAB Model 4 Sideforce Coefficient due to Aileron Deflection
0.06
0.04
0.02
0
0.02
0.04
0.06
10 5 0 5 10
C
l
(
N
.
D
.
)
aileron deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 77: Meridian VORSTAB Model 4 Rolling Moment Coefficient due to Aileron
Deflection
125
0.015
0.01
0.005
0
0.005
0.01
0.015
10 5 0 5 10
C
n
(
N
.
D
.
)
aileron deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 78: Meridian VORSTAB Model 4 Yawing Moment Coefficient due to Aileron
Deflection
The sideforce coefficient should be very small without the fuselage present,
and with it present all of the derivatives will change dramatically.
126
Table 46: Meridian VORSTAB Model 4 Longitudinal Derivatives due to Symmetrical
Tail Deflection
o (deg) o
re
(deg) C
L
C
D
C
m
0 12 0.4094 0.0159 0.0305
5 12 0.0563 0.0244 0.0451
10 12 0.6136 0.0201 0.0363
15 12 0.9769 0.0430 0.0588
20 12 1.2689 0.0580 0.0892
0 6 0.4007 0.0159 0.0641
5 6 0.0649 0.0244 0.0791
10 6 0.6218 0.0201 0.0693
15 6 0.9854 0.0431 0.0932
20 6 1.2769 0.0580 0.1216
0 0 0.3916 0.0155 0.0995
5 0 0.0736 0.0244 0.1135
10 0 0.6302 0.0202 0.1031
15 0 0.9934 0.0431 0.1256
20 0 1.2843 0.0581 0.1517
0 6 0.3826 0.0159 0.1341
5 6 0.0821 0.0244 0.1473
10 6 0.6380 0.0202 0.1343
15 6 1.0008 0.0432 0.1554
20 6 1.2909 0.0581 0.1785
0 12 0.3740 0.0160 0.1675
5 12 0.0900 0.0245 0.1784
10 12 0.6450 0.0202 0.1621
15 12 1.0072 0.0433 0.1816
20 12 1.2965 0.0581 0.2012
127
0.6
0.4
0.2
0
0.2
0.4
0.6
0.8
1
1.2
1.4
12 10 8 6 4 2 0 2 4 6 8 10 12
C
L
(
N
.
D
.
)
tail deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 79: Meridian VORSTAB Model 4 Lift Coefficient due to Elevator Input
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
12 10 8 6 4 2 0 2 4 6 8 10 12
C
D
(
N
.
D
.
)
tail deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 80: Meridian VORSTAB Model 4 Drag Coefficient due to Elevator Input
128
0.25
0.2
0.15
0.1
0.05
0
12 10 8 6 4 2 0 2 4 6 8 10 12
C
m
(
N
.
D
.
)
elevator deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 81: Meridian VORSTAB Model 4 Pitching Moment Coefficient due to Elevator
Input
The lift and pitching moment both increase with a positive pitching moment,
as expected. Drag stays rather steady when in reality it will change a small amount.
129
Table 47: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to
Asymmetrical Deflection
o (deg) o
re
(deg) C
y
C
l
C
n
0 10 0.0246 0.0032 0.0102
5 10 0.0251 0.0031 0.0106
10 10 0.0252 0.0029 0.0107
15 10 0.0218 0.0032 0.0092
20 10 0.0198 0.0031 0.0084
0 5 0.0127 0.0016 0.0053
5 5 0.0129 0.0015 0.0054
10 5 0.0119 0.0016 0.0050
15 5 0.0110 0.0016 0.0047
20 5 0.0100 0.0016 0.0043
0 0 0.0000 0.0000 0.0000
5 0 0.0000 0.0000 0.0000
10 0 0.0000 0.0000 0.0000
15 0 0.0000 0.0000 0.0000
20 0 0.0000 0.0000 0.0000
0 5 0.0128 0.0016 0.0053
5 5 0.0123 0.0016 0.0052
10 5 0.0119 0.0016 0.0050
15 5 0.0110 0.0016 0.0047
20 5 0.0100 0.0016 0.0043
0 10 0.0248 0.0032 0.0103
5 10 0.0238 0.0033 0.0100
10 10 0.0226 0.0033 0.0095
15 10 0.0218 0.0032 0.0092
20 10 0.0198 0.0031 0.0084
130
0.03
0.02
0.01
0
0.01
0.02
0.03
10 5 0 5 10
C
y
(
N
.
D
.
)
tail deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 82: Meridian VORSTAB Model 4 Sideforce Coefficient due to Rudder Input
0.004
0.003
0.002
0.001
0
0.001
0.002
0.003
0.004
10 5 0 5 10
C
l
(
N
.
D
.
)
tail deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 83: Meridian VORSTAB Model 4 Rolling Moment Coefficient due to Rudder
Input
131
0.015
0.01
0.005
0
0.005
0.01
0.015
10 5 0 5 10
C
n
(
N
.
D
.
)
tail deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 84: Meridian VORSTAB Model 4 Yawing Moment Coefficient due to Rudder
Input
Again, these derivatives will be greatly influenced by having a fuselage
presents in the model.
9.2.5 Meridian VORSTAB Conclusions
The results from VORSTAB varied greatly from model to model. These
variations make it very difficult to draw any conclusions from VORSTAB. Since the
first model is the only one that has the exact configuration it would be expected that it
is the most accurate, but this was not the case. The forth model‟s results with the
closest to the AAA results. Table 48 shows the results compared to Dr. Roskam‟s
typical ranges, Ref [5]. These ranges are for conventional aircraft, and the ranges are
a function of Mach number. The only results that are not shown in this table are the
132
first model. Model 1‟s results were not valid at all, so they were completely ignored.
The first model was used as a stepping stone to create the other models.
133
Table 48: Meridian VORSTAB Stability and Control Derivatives Dr. Roskam’s Typical
Ranges
Stability and Control Derivatives
Derivative
(1/rad)
Model
2
Model
3
Model
4
Typical
Ranges
Within Range (Yes/No)
Dr. Roskam Model 2 Model 3 Model 4
C
Lo
7.2337 7.1675 5.3308 1.0 to 8.0 Yes Yes Yes
C
Do
15.580 6.1068 0.1020 0.0 to 2.0 No No Yes
C
mo
0.8522 0.4038 0.1609 4.0 to 1.0 No No Yes
C
Lq
7.1682 7.1767 3.7241 0.0 to 30.0 Yes Yes Yes
C
mq
8.5662 8.5688 3.7942 90.0 to 0.0 Yes Yes Yes
C
l
0.1083 0.1031 0.1060 0.1 to 4.0 Yes Yes Yes
C
lp
0.2913 0.292 0.2960 0.1 to 0.8 Yes Yes Yes
C
lr
0.2263 0.2316 0.2320 0.0 to 0.6 No No No
C
y
0.7695 0.8228 0.6383 0.1 to 2.0 No No No
C
yp
0.1010 0.1065 0.1229 0.3 to 0.8 Yes Yes Yes
C
yr
0.2091 0.1261 0.1291 0.0 to 1.2 No No No
C
n
0.0991 0.0773 0.0940 0.0 to 4.0 Yes Yes Yes
C
np
0.0312 0.0242 0.0226 0.5 to 0.1 Yes Yes Yes
C
nr
0.0748 0.046 0.0666 0.0 to 1.0 Yes Yes Yes
C
Loe
0.0974 0.0922 0.0858 0.0 to 0.6 Yes Yes Yes
C
Doe
0.9313 1.3450 0.0040 Negligible No No No
C
moe
0.3867 0.3600 0.3304 0.0 to 4.0 Yes Yes Yes
C
loa
0.2628 0.2625 0.2606 0.0 to 0.4 Yes Yes Yes
C
lor
0.0117 0.0123 0.0178 0.04 to
0.04
Yes Yes Yes
C
yoa
0.0565 0.0567 0.0583 Negligible No No No
C
yor
0.1303 0.1248 0.1463 0.0 to 0.5 Yes Yes Yes
C
noa
0.0692 0.0688 0.0670 0.08 to
0.08
Yes Yes Yes
C
nor
0.0559 0.0479 0.0612 0.15 to 0.0 Yes Yes Yes
There are multiple derivatives that do not fall into the typical ranges. All of
the models have a high drag coefficient except the forth. The forth model also has the
134
most derivatives that fall into the typical ranges described. Table 49 shows a
comparison of the second, third, and fourth model next to AAA. The forth model
predicted several of the derivatives the best.
Table 49: Meridian AAA and VORSTAB Comparison
Stability and Control Derivatives
Derivative
(1/rad)
Model 2 Model 3 Model
4
AAA % Diff.
Model 2
% Diff.
Model 3
% Diff.
Model 4
C
Lo
7.2337 7.1675 5.3308 5.1648 40.06 38.78 3.21
C
Do
15.58 6.1068 0.102 0.1409 10957.49 4434.14 27.61
C
mo
0.8522 0.4038 0.1609 0.6207 37.30 34.94 74.08
C
Lq
7.1682 7.1767 3.7241 4.6179 55.23 55.41 19.36
C
mq
8.5662 8.5688 3.7942 13.973 38.69 38.68 72.85
C
l
0.1083 0.1031 0.106 0.0776 39.56 32.86 36.60
C
lp
0.2913 0.292 0.296 0.5546 47.48 47.35 46.63
C
lr
0.2263 0.2316 0.232 0.1099 305.91 310.74 311.10
C
y
0.7695 0.8228 0.6383 0.4789 260.68 271.81 233.28
C
yp
0.101 0.1065 0.1229 0.1465 31.06 27.30 16.11
C
yr
0.2091 0.1261 0.1291 0.3217 35.00 60.80 59.87
C
n
0.0991 0.0773 0.094 0.1386 28.50 44.23 32.18
C
np
0.0312 0.0242 0.0226 0.0351 11.11 31.05 164.39
C
nr
0.0748 0.046 0.0666 0.1338 44.10 65.62 50.22
C
Loe
0.0974 0.0922 0.0858 0.4149 76.52 77.78 79.32
C
Doe
0.9313 1.345 0.004 0.0117 8059.83 11395.73 65.81
C
moe
0.3867 0.36 0.3304 1.6709 76.86 78.45 80.23
C
loa
0.2628 0.2625 0.2606 0.2316 13.47 13.34 12.52
C
lor
0.0117 0.0123 0.0178 0.0253 146.25 148.62 170.36
C
yoa
0.0565 0.0567 0.0583 0   
C
yor
0.1303 0.1248 0.1463 0.3681 135.40 133.90 139.74
C
noa
0.0692 0.0688 0.067 0.0134 616.42 613.43 600.00
C
nor
0.0559 0.0479 0.0612 0.1481 137.74 132.34 141.32
135
The fourth model best predicted the Meridian stability and control derivatives
when compared to AAA. There are several derivatives that are extremely far off, but
the fuselage will have an impact on most of these. After modeling these three
different aircraft it can be concluded that VORSTAB works best with traditional style
aircraft, such as the YAK54. The Meridian is not a typical style of aircraft and the
VORSTAB results were not what were expected. These results should not be
completely ignored, because they brought up possible issues that the Meridian might
have. For example, the downward swoop of the fuselage could result in downwash
and flow separation. This downwash and flow separation possible could render the
tails ineffective and the aircraft unresponsive to tail inputs. It is recommended that
several more wing and tail models be created to determine if it predicts the stability
and control derivatives well as seen from the Meridian results.
9.3 Linearized Model of the Meridian
For the information and data on the state space model of the Meridian please
refer to Appendix B.
9.4 FLUENT Modeling of the Meridian
As learned from the VORSTAB model of the Meridian, the main concern is the
downwash created over the fuselage that could blank out the vtails. The first flight
tests also resulted in very high drag. These two problems, downwash and drag, drove
the need for a FLUENT model to be created. A 3D model was desired, but due to
time constraints this was not done. Therefore, a 2D side profile of the fuselage was
136
created to see the downwash that was produced from the high camber. A complete 3
D model of the Meridian should eventually be created.
9.4.1 GAMBIT Mesh Generation
A large farfield was created around the axialcrosscut of the fuselage. The
farfield was broken into six individual sections; forward (above and below) of the
fuselage, aft (above and below) of the fuselage, above the fuselage, and below the
fuselage. Above and below the fuselage had a height of 20 maximum thicknesses of
the fuselage, and forward and aft had a length of 20 lengths of the fuselage. This
created a large enough area where the flow could be appropriately modeled. The size
of the farfields depends on the flow being modeled and the shape of the object. A
very fine mesh is desirable everywhere, but this would take a large amount of
computer memory and time to run. Different farfields allow the designer to create
finer meshes in the more critical sections of the flow. For the 2D Meridian fuselage
the critical flow section was the boundary layer of the Meridian walls. This is where
the flow can separate. The flow at a large distance from the fuselage is not as critical
as the flow near the walls. The following are the steps taken to create the six farfields
and their meshes:
Step 1: Import the axial crosssection of the fuselage (points created as a text file).
The tip of the nose will be on the (0, 0, 0) point. Figure 85 shows the Meridian cross
section.
137
Figure 85: Meridian Fuselage Axial CrossSection for FLUENT
Step 2: Create a line that extend from the tip of the nose (labeled point M) to 20 times
the length of the fuselage forward, labeled point A.
Step 3: Repeat step 2, but extend from the tip of the tail (labeled point N) to aft of the
fuselage, labeled point B.
Step 4: Copy the point, on the upper surface, where the nose cone meets the fuselage
cowling (labeled point K) up to 20 times the maximum thickness minus the current
height of the point, called point C.
Step 5: Repeat step 4, but use the point on the lower surface where the nose cone
meets the lower fuselage cowling (labeled point L) called point D.
Step 6: Copy the point N up the same height as the point created in step 4, and then
copy the same point to the height created in step 5, labeled points E and F
respectively.
Step 7: Copy the points A and B to the heights created in step 4 and step 5 labeled
points G, H, I and J respectively. There should be 3 points forward of the fuselage, 3
138
points aft, 2 points above and below the nose cone, and 2 points above and below the
tip of the tail.
Step 8: Use the NUMS tool to create a curved line on the upper surface of the nose
cone. Repeat on the lower surface of the nose cone, upper surface of the fuselage,
and lower surface of the fuselage. This creates four individual lines.
Step 9: Next use the line tool to create a line from point M to point A.
Step 10: Use the line tool to create 3 lines from point N to points B, E and F.
Step 11: Use the line tool to create a line from point K to point C. Then repeat with
point L to point D.
Step 12: Create lines AG, AH, GC, HD, CE, DF, EI, FJ, BI, and BJ. Figure
86 shows the Meridian axial crosssection and the newly created lines to create the
farfields. The Meridian crosssection is the small white shape in the middle.
Figure 86: Meridian Fuselage with Farfield Divisions
Step 13: Create the six different farfield and the Meridian body face. Table 50 shows
the corresponding lines to each face.
139
Table 50: Meridian Farfield Edges
Face Name Edges
Farfield 1 AG, GC, CK, KM, AM
Farfield 2 AH, HD, DL, LM, AM
Farfield 3 CE, EN, NK, CK
Farfield 4 LN, NF, DF, DL
Farfield 5 EI, IB, NB, EN
Farfield 6 FJ, JB, NB, FN
Meridian MK, KN, ML, LN
Step 14: Create „Edge Meshes‟ using Table 51. The small intervals should be closer
to the Meridian wall, on all lines that have first lengths.
Table 51: Meridian Edge Meshes
Edge Interval Count Properties
AG, AH, CK, DL, EN, FN, BI, BJ 100 First Length = 0.25
CG, DH, EI, FJ, NB 100 First Length = 0.25
CE, DF, KN, LN 70 Successive Ratio = 1
KM, LM 8 Successive Ratio = 1
MA 92 First Length = 0.20
Step 15: Create „face meshes‟ in the six farfields. The meshes is denser towards the
Meridian body. Figure 87 shows the meshes concentrated towards the fuselage body.
The lower image is a zoomed in image of the upper picture.
140
Figure 87: Meridian Farfield Meshes
Step 16: Create five groups of edges. Group 1 as edge AG and AH, Group 2 as CG,
CE and EI, Group 3 as DH, DF and FJ, Group 4 as BI and BJ, and Group 5 as MK,
KN, ML and LN. With the groups created, their corresponding boundary conditions
were set. Groups 1, 2 and 3 were velocity inlets, Group 4 was a pressure outlet, and
Group 5 was a wall.
These boundary conditions work because the farfields are so big that the flow
can be considered unaffected by the fuselage body or totally recovered from the flow
141
disturbances. The boundary conditions can be modified in FLUENT if they are not
what were desired. With the GAMBIT model fully created export the 2D mesh and
import it into FLUENT.
9.5 FLUENT Model Generation
The FLUENT model can be run in many different methods depending on what
the goals the designer is trying to achieve. For the Meridian, the goal was to observe
the boundary layer on the fuselage, and observe the flow separation and/or downwash
that were produced from the high convergence and downward camber. Several
different turbulent methods were tested to verify which one was the most accurate.
For the Meridian, the results did not very greatly from one method to the next.
Therefore, the method chosen was the one that has the most accurate results in
aerodynamics. The following are the setup criteria for the FLUENT model.
 Solver – Segregated, because of limited memory
 Viscous – Turbulent, SpalartAllmaras; set the turbulence factors to 0.1 for an
initial guess.
 No Energy Equation because the flow is not compressible
 Materials – Set the density and viscosity to the values that the Meridian will
fly in. At 1,000 m the approximate values are 1.112 kg/m
3
and 1.758x10
5
kg/ms respectively.
 Operating Conditions – Set this to zero to work in terms of absolute pressure.
If left at the default setting the pressures reported would be in terms of gauge
pressure.
142
 Boundary Conditions – The types of boundary conditions chosen in GAMBIT
were the desired ones for this model, but the values had to set.
o Velocity Inlets – 61.7 m/s in the xdirection and 0.0 m/s in the y
direction. For angles of attack the velocity components could be
changed. The model did not converge doing this, so a new mesh was
created for each angle of attack chosen. To create the new mesh the
same method as before was used, but the imported Meridian cross
section was imported at the correct angle of attack.
o Pressure Outlet – 0 Pascals for the gauge pressure
o Wall – Set the wall roughness height to 3x10
5
m and roughness
constant to 0.5. These were the values used by AAA when designing
the aircraft.
 Solution – SIMPLE for the pressurevelocity coupling, PRESTO! for the
pressure discretization, and SecondOrder Upwind for the other
discretizations. These were determined using the online FLUENT help, Ref
[19].
 Residuals – 1x10
6
 Monitor – Lift and drag
 Iterate until converged
The mesh generation and FLUENT runs were repeated for several different
angles of attack. Multiple angles of attack allow the designer to see the changes in
the flow as the aircraft body changes orientation.
143
9.5.1 Meridian FLUENT Results
The angles of attack checked for the Meridian fuselage were 4˚, 0˚, 2˚, 5˚, and
10˚. A large amount of lift was produced by the large camber in the fuselage. The
down camber on the tail lowered the total pressure on the upper surface and also
results in extra lift. The large boundary layer that built up from this negative camber
increases the drag. The lift and drag were determined in terms of wing area, 3.35 m
2
.
Using the wing area as a reference, allows the designer to compare this lift and drag
to the lift and drag with the rest of the aircraft. Table 52 shows the lift and drag found
at all angles of attack.
Table 52: Meridian Fuselage CrossSection FLUENT Lift and Drag
o (deg) C
l_f
C
d_f
L/D
4 0.1080 0.0651 1.658
0 0.4030 0.0286 14.10
2 0.5479 0.0414 13.22
5 0.6456 0.0917 7.039
10 0.8458 0.2573 3.287
As seen from the lift to drag results, the Meridian fuselage will generate a
large amount of lift. At low angles of attack, the lift to drag ratio is close to what the
design of the overall aircraft was. Typically, the lift produced by the fuselage can be
neglected due the small value, but the Meridian‟s lift is too large to neglect. The
overall magnitude of the fuselage lift is not extremely high, but when compared to the
amount of drag the fuselage produces it is considered high. The highly cambered
wings will produce much more lift. Also, the 3D effects on the fuselage will change
these numbers some. Most likely the actual numbers will be less than this.
144
The FLUENT results can best be shown in the form of pictures. The
downwash, flow separation, and boundary layer can all easy be seen in the velocity
magnitude, velocity angle, total pressure, stream function, vorticity profiles. Figure
88 shows the velocity magnitude of the air around the Meridian. From left to right,
top to bottom the figures are in an order of 0˚, 2˚, 5˚, 10˚, and 4˚.
145
Figure 88: Meridian FLUENT Velocity Magnitude Profile for All Angle of Attacks
It can be seen, as the angle of attack increase the boundary layer increases
also. Basically, the boundary layer is the thickness of air that is not traveling at free
stream velocity. The average thickness is directly related to the Reynolds Number of
146
the flow and the length of the surface. A Cartesian coordinate system is used for the
boundary layer definitions. The xcoordinate is parallel with the surface at each
point, and the ycoordinate is normal to the surface. Eq [28] shows an estimate for
the average thickness of the boundary layer, and Eq [29] show the definition of
Reynolds Number, Ref [20].
 Eq [28]
 Eq [29]
There are three common measurements of the boundary layer thickness, o.
One method is defined as the distance from the wall where the velocity reaches 99
percent of the free stream velocity, u = 0.99U. This method is somewhat arbitrary
since any percentage could have been picked. The reason 100 percent of the free
stream velocity was not chosen is that it is difficult to pinpoint the exact location
where this happens. It is appropriate to assume 99 percent is close enough for
accuracy. The second method is the displacement thickness, o*, and there is no
arbitrariness in this method. Displacement thickness is the theoretical distance that
the wall would have to move outward, in a frictionless flow, to maintain the same
mass flux as the authentic flow. Eq [30] shows the displacement thickness, Ref [20].
The upper boundary goes to infinity, because u/U goes to 0 exponentially fast in the y
direction, as y goes to infinity.
 Eq [30]
o “u” is a function of the variable “y”.
147
The third method is the momentum thickness, u, and is defined as the amount
of momentum, pU
2
u, loss due to the boundary layer. This momentum thickness can
be determined using Eq [31], Ref [20]. Again in this equation “u” is a function of
the variable “y”.
 Eq [31]
The type of definition does not matter when it comes to boundary layer. Each
method has its benefits and downfalls. For example, the second and third methods
are difficult to develop an equation for the velocity, u. As stated earlier the first
method chooses an arbitrary number, but this is the easiest to use with experimental
data. FLUENT depicts the velocity magnitude by varying the color. This is depicted
by the blue area around the fuselage. The boundary layer flow stays attached to the
fuselage and follows the contour of it. Boundary layers will build up in the y
direction as the flow travels over the body. The size of the boundary layer is
extremely thick at high angles of attack. For a typical aircraft wing the boundary
layer thickness is on the order of one centimeter. At higher angles of attack, 5˚ and
10˚, and large fuselage curvatures the flow separates completely from the fuselage.
This is seen as the blue area that does not follow the fuselage contour, but instead
starts flowing turbulently. The direction of the flow can be shown by viewing the
velocity angle in Figure 89. The orientation of the angles is 0˚ to the right and 90˚
straight up.
148
Figure 89: Meridian FLUENT Velocity Angle Profile for All Angles of Attack
A change in direction is expected around the fuselage. If the contour of the
fuselage has smooth transitions and small curvatures the flow should follow the
fuselage shape. The changes in flow direction that are not expected are the ones at
149
angles of attack 5˚ and 10˚. Separation occurs at these angles of attack near the
empennage. Similar to the velocity magnitude, the velocity angle depicts its changes
with changes in color. A smooth change is depicted with a slow transition in color,
and a sudden change with a dramatic change in color. This separation bubble is
where the flow can travel in many directions and possible have complete flow
reversal. The direction of the velocity for angles of attack 4˚, 0˚, and 2˚ follows the
contour of the fuselage pretty well. There are a few sudden changes, but this is
expected when the flow encounters a disturbance in the path. For example, the flow
changes direction abruptly at the nose cone of the aircraft. The flow stops at this
point and begins to travel around that spot.
At higher angles of attack, 5˚ and 10˚, there are unexpected turns in the flow.
The pressure gradient along the surface determines how the flow travels over a
surface. A favorable pressure gradient will be upstream of the highest point, and this
will cause the streamlines of the flow to converge and a decrease in pressure. An
adverse pressure gradient is the opposite of the favorable pressure gradient.
Downstream of the highest point the streamlines diverge and result in an increase in
pressure. These pressure gradients not only affect the pressure, but also the velocity.
A favorable pressure gradient will increase the free stream velocity, and an adverse
pressure gradient will decrease the free stream velocity. Figure 90 depicts the
velocity profiles along a curved surface, Ref [20].
150
Figure 90: Velocity Profile in Boundary Layer with Favorable and Adverse Pressure
Gradient (From Ref [20])
The inflection point is where the c
2
u/cy
2
= 0, or in other words where the
curvature of the velocity profile changes signs. A strong enough adverse pressure
gradient can cause the flow to separate from the wall. If the flow continues with an
adverse pressure gradient the flow will reverse directions. The backward flow will
meet the forward flow at some point and begins to flow forward again. Where these
two flows, backwards and forwards, meet is where the stress vanishes. Figure 91
shows a representation of the flow separating from an adverse pressure gradient and
then flow reversal happening, Ref [20]. At low Reynolds numbers, a large steady
vortex will form behind this separation point. This is where the flow swirls in an
area, but it does not shed off the body. This is seen at both the higher angles of
attack. Separation is a function of the flow geometry and whether the boundary layer
151
is laminar or turbulent. A blunt body or sudden change in shape can lead to a steep
pressure gradient, and therefore leads to separation happening quickly.
Figure 91: Velocity Profile with Flow Separation at Point S (dashed line u = 0, From
Ref. [20])
The flow separation is very apparent at the high angles of attack with the
velocity angles flow turning red. The dark red and dark blue are flow reversal points.
To avoid separation and higher drag a structure should have a gradual change in size
and shape. At the higher angles of attack the geometry changes more suddenly and
this causes separation to happen more quickly. Figure 92 shows the total pressure in
the flow.
152
Figure 92: Meridian FLUENT Total Pressure Profile for All Angles of Attack
The total pressure is directly related to the flow velocity and the aircraft shape
and angle of attack. Certain angles of attack will affect the flow field pressure much
more than others. This is seen at angles of attack of 10˚ and 4˚. The colors move
153
away from the red color to a yellow or orange color. These changes in color are a
reduction in the total pressure due to the fuselage contour and flow velocity. This is
one of the reasons for the very low lift to drag ratio at these two angles of attack. The
total pressure is lower where the boundary layer is located and the where the flow
separates from the body due to the small velocity. Total pressure is a function of the
velocity squared, Eq [32].
 Eq [32]
At the 4˚ angle of attack the boundary layer on the lower surface is very
thick. This is due to the sudden change in shape on the cowling. It is similar to the
flow encountering a step, and this causes the flow to separate very quickly. The flow
path can be seen in Figure 93 showing the stream functions. Stream functions have a
constant value along a streamline. These stream functions can be used to analytically
determine the velocity and its direction.
154
Figure 93: Meridian FLUENT Stream Function Profile at All Angles of Attack
Stream functions are defined to be constant along a stream line, and the Av
between two streamlines is the volume of the flow between those streamlines. This
can be thought of as the volume flowing through a streamtube bounded by those two
155
streamlines. The units for stream function are kg/s. A stream function normally has
units of kg/sm, but this model is 2D and has no depth. There is also a unitless form
of stream function and is defined by Eq [33] (without the bar), Ref [21]. To
determine the velocity components from the stream function Eq [34] and Eq [35]
should be used, Ref [21].
 Eq [33]
 Eq [34]
 Eq [35]
At the high angle of attack, 10˚, it is seen that the stream function travels in a
complete closed path around the empennage. This is due to the flow separating from
the fuselage and traveling in this closed path. The rest of the flow travels over the top
of this circulation.
Vorticity is defined as twice the angular velocity, Ref [20]. This is found by
calculating the average rotation rate of two perpendicular lines and then multiplying it
by two. Irrotational flow will have a magnitude of zero vorticity. This is why the
dark blue area is very close to zero in Figure 94. There is high vorticity around the
abrupt curved spots on the fuselage. The boundary layer also is a lighter shade of
blue, because the flow in this area has rotation.
Using Stokes‟ theorem circulation can be written as Eq [36], Ref [20]. As
seen from this equation circulation is related to the rotation of the flow. Higher
angular velocity will produce more circulation around a finite area. Circulation can
be thought of as the strength of vorticity or vortex. Lift is related to circulation using
156
Eq [37], Ref [6]. This is known as the KuttaJoukowski theorem. On most
computational fluid dynamics programs circulation is used as a step to calculate the
forces on a body. When a body is generating lift the circulation is finite. This means
that the lift is directly related to the boundary layer vorticity.
 Eq [36]
 Eq [37]
157
Figure 94: Meridian FLUENT Vorticity Magnitude Profile at All Angles of Attack
KuttaJoukowski Theorem is used to estimate the lift produced by an airfoil or
wing. The lift produced by a wing is a slightly manipulated form of Eq [37], because
it is in terms of per unit of span. Therefore, that equation is multiplied by the amount
158
of span that generates lift. This theorem explains why the fuselage generates lift. For
example, for an airfoil to generate lift the lower surface must have higher pressure
than the upper surface. In order for this to happen in a uniform flow the upper surface
velocity must increase and the lower surface velocity must decrease. This difference
in pressures or velocities over a length will produce the lift. If a line of circulation, l‟,
is used to define an area of integration then circulation can be determined for this
airfoil or wing section. This is depicted in Figure 95, Ref [6] to help explain how
circulation will produce lift.
Figure 95: Circulation around an Airfoil Producing Lift (From Ref [6])
9.6 Method Comparison Meridian
Since time and resources did not allow for a 3D FLUENT model to be
created the stability and control derivatives from FLUENT were not able to be
determined. Lift and drag that was produced by the fuselage, in a 2D since, was
found. Digging into the VORSTAB output files the theoretical fuselage lift and drag
can be found. AAA does not calculate the lift of the fuselage, but rather the
159
contribution of the wing and fuselage combination. Since AAA uses this
combination the individual wing lift was subtracted from the wingfuselage lift to
determine the individual fuselage lift. This resulted in an extremely small lift, and
was negligible when compared to the VORSTAB and FLUENT results. The
comparisons lift and drag values for the Meridian fuselage are shown in Table 53.
The VORSTAB results shown in this table are for Models 2 and 3, because they are
the only results that VORSTAB will generate fuselage data. Model 1 calculated
downwash so high that the pressure on the fuselage was too high to continue
calculating the fuselage data. The forth model did not have a fuselage present for
calculations.
Table 53: Meridian Fuselage Lift and Drag Method Comparison
Method
o = 0˚ o = 5˚ o = 5˚ and 4˚
(FLUENT)
C
l
C
d
L/D C
l
C
d
L/D C
l
C
d
L/D
FLUENT 0.403 0.029 14.09 0.548 0.092 5.975 0.108 0.065 1.659
VORSTAB
Model 2 0.010 1.460 0.007 0.019 2.827 0.007 0.003 0.486 0.005
Model 3 0.001 0.598 0.001 0.009 0.072 0.119 0.006 1.606 0.004
As seen from these results the FLUENT fuselage seems to produce much
more lift and less drag than the other two methods. The 3D effects will reduce the
lift, but also they will also help the flow to stay attached to the fuselage and avoid
separation. This would reduce the drag. Therefore, the lift to drag ratio for FLUENT
might be correct, but the individual values might be off. Model 2 and Model 3 for the
VORSTAB drag results are very high. This is due to the high downwash produced in
160
by VORSTAB. Fuselages do not typically produce much lift. These VORSTAB lift
to drag ratios are nowhere near the expected value.
Similar to the YAK54 and MantaHawk the stability and control derivatives
from both methods, VORSTAB and AAA, can be compared to one another. Table 54
shows the comparison for the Meridian stability and control derivatives.
161
Table 54: Meridian AAA and VORSTAB Comparison
Longitudinal Coefficients and Stability Derivatives
Derivative (1/deg) AAA
VORSTAB
Model 1 Model 2 Model 3 Model 4
C
Lo
5.1648 0.0929 7.2337 7.1675 5.3308
C
Do
0.1409 0.0000 15.5804 6.1068 0.1020
C
mo
0.6207 0.2268 0.8522 0.4038 0.1609
C
Lq
4.6179 34.603 7.1682 7.1767 3.7241
C
mq
13.973 17.109 8.5662 8.5688 3.7942
C
l
0.0776 2.2804 0.1083 0.1031 0.1060
C
lp
0.5546 1.1605 0.2913 0.2920 0.2960
C
lr
0.1099 0.6586 0.2263 0.2316 0.2320
C
y
0.4789 1.2031 0.7695 0.8228 0.6383
C
yp
0.1465 0.2153 0.101 0.1065 0.1229
C
yr
0.3217 4.5412 0.2091 0.1261 0.1291
C
n
0.1386 2.4115 0.0991 0.0773 0.0940
C
np
0.0351 1.1294 0.0312 0.0242 0.0226
C
nr
0.1338 2.0948 0.0748 0.0460 0.0666
C
Loe
0.4149  0.0974 0.0922 0.0858
C
Doe
0.0117  0.9313 1.3450 0.0040
C
moe
1.6709  0.3867 0.3600 0.3304
C
loa
0.2316  0.2628 0.2625 0.2606
C
lor
0.0253  0.0117 0.0123 0.0178
C
yoa
0.0000  0.0565 0.0567 0.0583
C
yor
0.3681  0.1303 0.1248 0.1463
C
noa
0.0134  0.0692 0.0688 0.0670
C
nor
0.1481  0.0559 0.0479 0.0612
Model 1 produced such a high amount of downwash that the program could
not calculate all the derivatives, and the ones it could were done with inaccurate
downwash. Model 4 does not contain the fuselage so the results only show what the
162
wings and vtail produce in terms of stability and control derivatives. The
VORSTAB results are unstable in many different modes as discussed earlier and the
AAA results are stable in every mode.
The same typical ranges that applied to the YAK54 and MantaHawk apply to
the Meridian as well. These ranges again were developed by Dr. Roskam, Ref [5] for
what a conventional aircraft can expect the control derivatives to be approximately.
These ranges are a function of Mach number. Table 55 shows the VORSTAB and
AAA Meridian results compared to typical ranges from Dr. Roskam, Ref [5].
163
Table 55: Meridian VORSTAB and AAA Stability and Control Derivatives Typical
Ranges
Stability and Control Derivatives
Model
2
Model
3
Model
4
AAA
Typical
Ranges
Within Range (Yes/No)
Dr.
Roskam
Model
2
Model
3
Model
4
AAA
C
Lo
7.234 7.168 5.331 5.165 1.0 to 8.0 Yes Yes Yes Yes
C
Do
15.58 6.107 0.102 0.141 0.0 to 2.0 No No Yes Yes
C
mo
0.852 0.404 0.161 0.621 4.0 to 1.0 No No Yes Yes
C
Lq
7.168 7.177 3.724 4.618 0.0 to 30.0 Yes Yes Yes Yes
C
mq
8.566 8.569 3.794 13.97 90.0 to 0 Yes Yes Yes Yes
C
l
0.108 0.103 0.106 0.078 0.1 to 4.0 Yes Yes Yes Yes
C
lp
0.291 0.292 0.296 0.555 0.1 to 0.8 Yes Yes Yes Yes
C
lr
0.226 0.232 0.232 0.110 0.0 to 0.6 No No No Yes
C
y
0.770 0.823 0.638 0.479 0.1 to 2.0 No No No Yes
C
yp
0.101 0.107 0.123 0.147 0.3 to 0.8 Yes Yes Yes Yes
C
yr
0.209 0.126 0.129 0.322 0.0 to 1.2 No No No No
C
n
0.099 0.077 0.094 0.139 0.0 to 4.0 Yes Yes Yes Yes
C
np
0.031 0.024 0.023 0.035 0.5 to 0.1 Yes Yes Yes Yes
C
nr
0.075 0.046 0.067 0.134 0.0 to 1.0 Yes Yes Yes No
C
Loe
0.097 0.092 0.086 0.415 0.0 to 0.6 Yes Yes Yes Yes
C
Doe
0.931 1.345 0.004 0.012 Negligible No No No No
C
moe
0.387 0.360 0.330 1.671 0.0 to 4.0 Yes Yes Yes Yes
C
loa
0.263 0.263 0.261 0.232 0.0 to 0.4 Yes Yes Yes Yes
C
lor
0.012 0.012 0.018 0.025 0.04 to
0.04
Yes Yes Yes Yes
C
yoa
0.057 0.057 0.058 0.000 Negligible No No No Yes
C
yor
0.130 0.125 0.146 0.368 0.0 to 0.5 Yes Yes Yes No
C
noa
0.069 0.069 0.067 0.013 0.08 to
0.08
Yes Yes Yes Yes
C
nor
0.056 0.048 0.061 0.148 0.15 to 0 Yes Yes Yes No
164
As seen from the table not all of the control derivatives fall within the typical
ranges. This does not necessarily mean that the aircraft is unstable. The rudder input
is a symmetrical deflection of the vtail and unsymmetrical deflection is an elevator
input. In some cases the AAA results do not fall into the expected range, but the
VORSTAB results do. This does not mean that VORSTAB was correct in that case
and that AAA was incorrect. The Meridian might just have higher values in some
derivatives than a typical aircraft has.
From these results it can be determined that VORSTAB is most affective with
less complex aircraft. The Meridian was very difficult to model and achieve accurate
results. This does not mean that VORSTAB is not a high fidelity tool or inaccurate,
but rather should be used on aircraft that have a simpler design. The fuselage model
was changed many times until results that were closer to the desired results were
achieved. Basically it was very similar to guessing and checking. This is not
necessarily bad, because many times the meshes of a FLUENT model have to be
remade until the results are what was desired or expected. It is very difficult to
determine if these programs are producing the correct results without comparing their
results with actual fight tests. After that comparison one can draw a better conclusion
about the validity of the results.
165
10 Conclusions and Recommendations
After analyzing the three aircraft with AAA and VORSTAB conclusions
about both software programs were drawn. AAA is a valid method for quickly
estimating the stability and control derivatives, even at a low Reynolds number. It is
a rather affordable software program and easy to use. A downfall to the program is
that a low fidelity method is used to derive the derivatives. A downfall of VORSTAB
is that it is very difficult to make a VORSTAB model and requires someone with
experience or guidance. The VORSTAB model also takes a considerable longer
amount of time to create than AAA. VORSTAB uses a high fidelity method to derive
the stability and control derivatives. For low Reynolds number, AAA overestimates
estimates the drag and this was one of the main reason that a VORSTAB model was
created for all of the vehicles. These are the obvious downfalls and benefits to both
programs.
For the YAK54, VORSTAB showed that the drag, C
Do
, was lower than AAA
predicted, as expected. This is a major advantage of VORSTAB is that the drag
estimation is closer to the actual drag. From the flight test data, VORSTAB estimated
many derivatives better than the AAA program did. Some of the derivatives from
VORSTAB are obvious that they are wrong and these should be ignored. For
example, the yawing moment due to aileron deflection has results that are not
symmetric, but opposite, for positive and negative aileron deflections. It is known
that a positive and negative aileron deflection should result in the same magnitude for
this derivative. These improvements from VORSTAB can be significant in creating a
166
valid YAK54 model. From these results, both software programs are a valid
engineering tool and both should be used for deriving the derivatives.
Similar to the YAK54, VORSTAB predicted a lower drag than AAA for the
MantaHawk. Without flight test it cannot be determined if VORSTAB or AAA
predicted the drag best. The MantaHawk wrecked early into the first flight test and
hardly any data was collected. From the small amount of data and explanation from
the flight test team the vehicle wrecked from a pitching moment error. AAA did not
predict this problem and said the vehicle should be stable in all modes. Most of
AAA‟s data is based on conventional aircraft and not a flying wing. VORSTAB
predicted that the MantaHawk was stable in every mode but the pitching moment
stability. It also estimated all of the derivatives to be much lower than AAA results.
This can have a significant effect on the control of the aircraft. Even if VORSTAB
was incorrect about the instability it would give the designers an idea of something
that should be investigated. Therefore, a recommendation for VORSTAB would be
to create a model for preliminary design and use it to determine if there is a possible
instability issue. Then the flight test team can use precaution when design the flight
tests and flying the aircraft. The instability mode can be tested in the flight tests, and
having the pilot aware of possible instability will dramatically improve the safety.
From the Meridian results, VORSTAB was found to be completely invalid.
The complex shape of the fuselage, vtail configuration, and highly cambered wings
were difficult for VORSTAB to handle. Since VORSTAB worked for both of the
other aircraft, it was determined that this software should only be used on aircraft that
167
have an extremely conventional configuration and a smaller amount of camber. The
MantaHawk did not have a conventional shape, but there was not a large camber.
From the VORSTAB output file, the highly cambered fuselage, a large amount of
downwash was produced. This was the source of most of VORSTAB‟s problems. A
wing and tail model should be created to model the drag if there is a complex fuselage
body. Once the flight tests have been repeated several times and the stability and
control derivatives have been determined a VORSTAB model should be created.
Then modifications can be made to this VORSTAB model until the expected results
are achieved. This would give the designer a general idea for ways to improve the
models in the future.
FLUENT was used to determine if there was a possible downwash problem
over the fuselage and if the fuselage was producing a large amount of drag. At high
angle of attack the flow separated from the fuselage and even began to have complete
flow reversal at 10˚ angle of attack. Since this was a 2D model the 3D effects will
help prevent the flow from separating. Tufting should be added to the empennage
and flight tested to determine if there is flow separation and turbulent flow. After this
investigation, a trip strip could possible be added to the empennage in order to
reenergize the flow and prevent separation. These trip strips can later be removed if it
is found that the separation is not a problem. The drag was determined to be much
lower than VORSTAB was estimating and the flight tests were showing. The high
drag was a result on having an unpainted surface, cowling with large gaps, no fairing,
and more. With improvements made to the aircraft the drag has reduced
168
dramatically. It is recommended that a complete 3D FLUENT model be created of
the Meridian. This model will be difficult to create, so it should be created in steps.
The wings and tails will not be that difficult to create and should be done first. Then
the fuselage should be created. It is a very complex geometry so the CAD model
should be used to import the fuselage. If Tgrid is available it should be used to
import the complete model and create the mesh. All of these recommendations will
help with future and current design projects.
169
11 References
[1] Roskam, J., “Airplane Flight Dynamics and Automatic Flight Controls,”
DARcorporation, Lawrence, KS, 2003.
[2] Lan, E., and Roskam, J., “Airplane Aerodynamics and Performance,”
DARcorporation, Lawrence, KS, 2003.
[3] Roskam, J., “Airplane Design Part VI: Preliminary Calculation of Aerodynamic,
Thrust and Power Characteristics,” DARcorporation, Lawrence, KS, 2004.
[4] “Advanced Aircraft Analysis,” DARcorporation Website,
[http://www.darcorp.com/Software/AAA/ , Retrieved 6 January, 2010].
[5] Anderson, J. D., “Modern Compressible Flow,” Linearized Subsonic Flow, 3
rd
ed., New York, New York, 2003, pp. 324333.
[6] Lan, E. C., “VORSTABA Computer Program for Calculating LateralDirectional
Stability Derivatives with Vortex Flow Effect,” NASA CR172501, January 1985.
[7] Lan, E. C., “User‟s Manual for VORSTAB Code (Version 3.2),” Department of
Aerospace Engineering at the University of Kansas, Revised 2006.
[8] Jager, R., “Test and Evaluation of the Piccolo II Autopilot System on a OneThird
Scale Yak54,” Department of Aerospace Engineering at the University of
Kansas, April 2008.
[9] Keshmiri, S., Leong, E., Jager, R., and Hale, R., “Modeling and Simulation of the
Yak54 Scaled Unmanned Aerial Vehicle Using Parameter and System
Identification,” AIAA20086900, August 2008.
[10] Leong, E., “A and B Matrix Construction Using Roskam Method,” MatLab
Code, September 2007.
[11] Arnold, E., Burns, R., and Grorud, D., “AE 721: Interim Final Review of a
Small Scale UAV,” The University of Kansas Aerospace Engineering
Department, December 2009.
[12] Sweeten, B., “The Meridian UAV Flight Performance Analysis Using
Analytical and Experimental Data,” AIAA20091899.
[13] Donovan, W., “The Meridian Critical Design Review,” June 25, 2007,
Technical Report CReSIS TR 123.
[14] Leong, H. and Jager, R., “Development of a Pilot Training Platform for UAVs
Using a 6DOF Nonlinear Model with Flight Test Validation,” AIAA20086368.
[15] Royer, D., Keshmiri, S., Sweeten, B., “Modeling and Sensitivity Analysis of the
Meridian Unmanned Aerial Vehicle,” AIAA Conference Submitted.
[16] Keshmiri, S., Royer, D., Tom, J., “The Meridian System Identification Flight
Tests in Dugway, UT,” Fall 2009.
[17] “Spatial Discretization,” Fluent Inc., September 2006,
[http://my.fit.edu/itresources/manuals/fluent6.3/help/html/ug/node992.htm].
[18] Kundu, P. K., and Cohen, I. M., “Fluid Mechanics,” 4
rd
ed., Burlington,
Massachusetts, 2008.
[19] Anderson, J. D., “Fundamentals of Aerodynamics,” 4
rd
ed., New York, New
York, 2007, pp. 404437.
170
[20] Mixon, B., D. and Chudoba, B., “The Lockheed SR71 Blackbird – A Senior
Capstone ReEngineering Experience,” AIAA2007698.
[21] Malmuth, N., D., “ Theoretical Aerodynamics in Today‟s Real World,
Opportunities and Challenges,” AIAA20055059.
[22] Alford, L. and Altman, A., “Wavelike Characteristics of Low Reynolds Number
Aerodynamics,” AIAA20044973.
[23] Chudoba, B. and Smith, H., “A Generic Stability and Control Methodology For
Novel Aircraft Conceptual Design,” AIAA20035388.
171
Appendix A
The following files are posted with my thesis on the KU website.
 For the input file of the YAK54 go to YAK_input and for the output file go to
YAK_output.
 For the input file of the MantaHawk go to MantaHawk_input and for the output
file go to MantaHawk_output.
 For the input file of the Meridian Model 1 go to Meridian1_input and for the
output file go to Meridian1_output.
 For the input file of the Meridian Model 2 go to Meridian2_input and for the
output file go to Meridian2_output.
 For the input file of the Meridian Model 3 go to Meridian3_input and for the
output file go to Meridian3_output.
 For the input file of the Meridian Model 4 go to Meridian4_input and for the
output file go to Meridian4_output.
172
Appendix B
YAK 54:
The linear state space model is developed using the Roskam method and a
program written by Edmond Leong, Ref [22]. It is developed in the form of
x˙=Ax+Bu. The program uses the control derivatives found from VORSTAB. The
following A and B matrices were determined from this method for both longitudinal
and lateraldirectional modes.
=
0 1.0000 0 0
0.0208 0.2101  2.3397  0.0020
0.2657  0.9979 0.2024  0.0042 
6.9963 0 29.3521 0.4156 
_ long A
=
0
0.0338 
0.0002 
0
_ long B
=
0.0621  1.7468 0 0.0090 
0.9995  0.0151  0.0592  0.0000 
0 0 0 1.0000
0.0218 0.6102  0 0.1664 
_ lat A
=
1.4499  0.5780
0.0104 0.0011 
0 0
0.3604 3.1788
_ lat B
173
The longitudinal dynamics should have two complex conjugate roots. These
two modes are Phugoid and Short Period. Table 56 shows the longitudinal mode
analysis summary and also compares it to the results that AAA control derivatives
produce, Ref [13]. Table 57 shows the lateraldirectional mode analysis summary
and also compares it to the results that AAA control derivatives produce, Ref [13].
Table 56: YAK54 Longitudinal Modes
Eigenvalues Damping Frequency
(rad/sec)
Mode
VORSTAB
0.281 1.0 0.281 ?
0.426 1.0 0.426 ?
0.342+1.60i 0.209 1.63 ?
0.3421.60i 0.209 1.63 ?
AAA
0.205+0.133i 0.839 0.24 Phugoid
0.2050.133i 0.839 0.24 Phugoid
9.87+1.73i 0.985 10 Short Period
9.871.73i 0.985 10 Short Period
Table 57: YAK54 LateralDirectional Modes
Eigenvalues Damping Frequency
(rad/sec)
Mode
VORSTAB
2.12e5 1.0 2.12e5 Spiral
0.149 1.0 0.149 Roll
0.0473+1.323i 0.0358 1.32 Dutch Roll
0.04731.323i 0.0358 1.32 Dutch Roll
AAA
0.0105 1 0.0105 Spiral
16.7 1 0.0599 Roll
1.22+65.3i 0.18 6.65 Dutch Roll
1.2265.3i 0.18 6.65 Dutch Roll
174
It is obvious from the results that VORSTAB is not correct. This is most
likely due to the fact that the state space model was created with a mixture of the
VORSTAB derivatives and the AAA derivatives. VORSTAB does not give every
derivative necessary to make the model and this is why some of the AAA derivatives
were used. The coupling of the two programs does not work.
Several flight tests were conducted to perform system identification on the
YAK54. The only two modes tested were the Dutch Roll mode and Phugoid mode.
These tests resulted in a damping ratio for the Phugoid mode of 0.575 and 0.303 for
the Dutch Roll, Ref [13].
MantaHawk:
The linear state space model is developed using the Roskam method and a
program written by Edmond Leong, Ref [22]. It is developed in the form of
x˙=Ax+Bu. The program uses the control derivatives found from VORSTAB. The
following A and B matrices were determined from this method for both longitudinal
and lateraldirectional modes.
=
0 1.0000 0 0
0 0.0425  0.1245  0.0137 
0 0.9997 0.0926  0.0063 
32.1740  0 43.6531 0.1495 
_ long A
=
0
1.2486 
0.0176 
0.0521 
_ long B
175
=
0.0035  0.1791 0 0.0001 
0.9999  0.0031  0.2723 0.0001 
0 0 0 1.0000
0.0120 0.6085  0 0.1728 
_ lat A
=
0.0875  0.0875 
0.0010 0.0010
0 0
2.2078 2.2078
_ lat B
The longitudinal dynamics has only one complex conjugate root and two
different real roots. There are typically two roots that are defined as Phugoid and
Short Period modes. When the center of gravity is ahead of the neutral point it will
cause Phugoid and Short Period roots to approach real roots (xaxis), but it still does
not explain the two different real roots, Ref [&^%]. Table 58 shows the longitudinal
mode analysis summary. Table 59 shows the lateraldirectional mode analysis
summary.
Table 58: MantaHawk Longitudinal Modes
Eigenvalues Damping Frequency (rad/sec) Mode
VORSTAB
0.0714 1.0 0.0714 ?
0.449 1.0 0.449 ?
0.0467+0.699i 0.0666 0.701 ?
0.04670.699i 0.0666 0.701 ?
176
Table 59: MantaHawk LateralDirectional Modes
Eigenvalues Damping Frequency
(rad/sec)
Mode
VORSTAB
4.27e5 1.0 4.27e5 Spiral
0.533 1.0 0.533 Roll
0.177+0.581i 0.291 0.607 Dutch Roll
0.1770.581i 0.291 0.607 Dutch Roll
Again, these results are not accurate since a mixture of both AAA and
VORSTAB derivatives were used to create the state space model. The vehicle did
not go under enough flight tests to verify any of the results.
Meridian:
The linear state space model is developed using the Roskam method and a
program written by Edmond Leong, Ref [22]. It is developed in the form of
x˙=Ax+Bu. The program uses the control derivatives found from VORSTAB. The
following A and B matrices were determined from this method for both longitudinal
and lateraldirectional modes of the Meridian model with the large empennage,
Model 2.
=
0 1.0000 0 0
0 0.0383  0.5584  0.0014 
0 0.9996 0.0673  0.0016 
32.1740  0 6.7443 0.0232
_ long A
=
0
0.1899 
0.0008 
1.4761
_ long B
177
=
0.0094  0.1876 0 0.0039 
0.9999  0.0060 0.1589 0.0001 
0 0 0 1.0000
0.0423  0.3085  0 0.0545 
_ lat A
=
0.1066  0.1320
0.0010  0.0004
0 0
0.0333 0.7485
_ lat B
The longitudinal dynamics has only two complex conjugate roots. Also, one
of the roots is unstable, since its eigenvalue is positive. Table 60 shows the
longitudinal mode analysis summary and also compares it to the results that AAA
control derivatives produce, Ref [18]. Table 61 shows the lateraldirectional mode
analysis summary and also compares it to the results that AAA control derivatives
produce, Ref [18].
Table 60: Meridian Model 2 Longitudinal Modes
Eigenvalues Damping Frequency (rad/sec) Mode
VORSTAB
0.0518+0.208i 0.241 0.215
0.05180.208i 0.241 0.215
0.093+0.73i 0.126 0.736
0.0930.73i 0.126 0.736
AAA
0.116 0.255 Phugoid
0.116 0.255 Phugoid
0.435 7.58 Short Period
0.435 7.58 Short Period
178
Table 61: Meridian Model 2 LateralDirectional Modes
Eigenvalues Damping Frequency (rad/sec) Mode
VORSTAB
0.0316 1.0 0.0316
0.232 1.0 0.232
0.103+0.473i 0.213 0.484 Dutch Roll
0.103+0.473i 0.213 0.484 Dutch Roll
AAA
1 109 sec Spiral
1 0.154 sec Roll
0.144 4.14 Dutch Roll
0.144 414 Dutch Roll
Removing the camber, Model 3, resulted in the following matrix A and B
from the VORSTAB results.
=
0 1.0000 0 0
0 0.0383  0.3381  0.0014 
0 0.9996 0.0668  0.0016 
32.1740  0 41.1159 0.0232
_ long A
=
0
0.1769 
0.0007 
2.1317 
_ long B
=
0.0058  0.1461 0 0.0030 
0.9999  0.0002 0.1589 0.0001 
0 0 0 1.0000
0.0433  0.2938  0 0.0546 
_ lat A
=
0.0913  0.1311
0.0010 0.0004
0 0
0.0349 0.7478
_ lat B
179
The longitudinal dynamics has two complex conjugate roots. These two roots
are Phugoid mode and Short Period mode. The lateraldirectional mode has 3 stable
roots, when the spiral mode is usually unstable. Table 62 shows the longitudinal
mode analysis summary and also compares it to the results that AAA control
derivatives produce, Ref [18]. Table 63 shows the lateraldirectional mode analysis
summary and also compares it to the results that AAA control derivatives produce,
Ref [18].
Table 62: Meridian Model 3 Longitudinal Modes
Eigenvalues Damping Natural Frequency Mode
VORSTAB
5.64e3+0.196i 0.0287 0.196
5.64e30.196i 0.0287 0.196
0.0353+0.601i 0.0586 0.602
0.03530.601i 0.0586 0.602
AAA
0.116 0.255 Phugoid
0.116 0.255 Phugoid
0.435 7.58 Short Period
0.435 7.58 Short Period
180
Table 63: Meridian Model 3 LateralDirectional Modes
Eigenvalues Damping Frequency (rad/sec) Mode
VORSTAB
0.0245 1.0 0.0245
0.257 1.0 0.257
0.111+0.436i 0.246 0.45
0.1110.436i 0.246 0.45
AAA
1 109 sec Spiral
1 0.154 sec Roll
0.144 4.14 Dutch Roll
0.144 4.14 Dutch Roll
For the same reason as the other two aircrafts this state space models are not
accurate. Flight tests were performed in Dugway, UT. During these flight tests the
Dutch Roll and Short Period modes were initiated, so system identification could be
performed. From the flight tests and analysis of that data from the flight test team the
Dutch Roll damping ratio of 0.2163 and natural frequency of 3.5341 rad/sec. AAA
was closer to the flight test data than the VORSTAB results. Also, as expected the
results from the model without camber had results closer to the flight test data. The
Short Period had a damping ratio of 0.5973 and a natural frequency of 3.6723 rad/sec.
All of the flight test results and analysis were taken from Ref [23]. Similar to the
Dutch Roll the model without the camber had better results. For complex geometry it
would be wise to use both AAA and VORSTAB as a means to determine the stability
and control derivatives. Aircraft that have noncomplex or standard geometry, like the
YAK54, VORSTAB‟s results should be more accurate.
The Thesis Committee for Benjamin Sweeten certifies that this is the approved Version of the following thesis:
CFD ANALYSIS OF UAVs USING VORSTAB, FLUENT, AND ADVANCED AIRCRAFT ANALYSIS SOFTWARE
Committee:
_____________________________ Dr. Shahriar Keshmiri, Chairperson _____________________________ Dr. Ray Taghavi _____________________________ Dr. Saeed Farokhi _____________________________ Dr. Richard Hale
Date approved:____________________
ii
Acknowledgements
I would like to thank NASA and the National Science Foundation for the contributions to the University of Kansas Aerospace Engineering Department. Without their support and funding, this research would have not taken place. Also, thank you to Dr. Shahriar Keshmiri for his leadership and guidance during my graduate and undergraduate career. Dr. Keshmiri helped in every aspect of my research, from selecting classes, funding, and research topics. Thank you to Dr. Richard Hale for his guidance on the Meridian UAV. He has high expectations of his students, and is willing to put out the same amount of effort if not more. Thank you to my professors that I have had throughout my education. They have given the knowledge needed for my career path. Thank you to Dr. Lan, for allowing me to use his program and giving me guidance during the learning process of this program. Thank you to Andy Pritchard, the Aerospace Engineering Department‟s airplane and power plant mechanic. Andy worked with me on many projects and research. He offered his guidance and experience to everyone, and he was a major influence on my graduate career. Thank you to all my fellow colleagues and students for their help. In particular Dave Royer, Jonathan Tom, and Bill Donovan for answering and helping with all questions that were asked of them. Without their hard work the Meridian would have never flew. I would like to thank my family for offering their support and anything I needed during my college career. Lastly, I would like to thank my fiancée Jessica Donigan for her love and support. She dealt with me in a very difficult time and has always been there for me. Thank you to everyone that I may
iii
have neglected to mention. iv . There have been many people that gave me guidance and support during my graduate and undergraduate career.
built. and the Meridian. The stability and control derivatives are critical for the flight of the vehicle. when this is known to not be the case after several flight tests. These UAVs are the 1/3 scale YAK54. The results found from the high fidelity computation fluid dynamics programs were then compared to the values found from the Advance Aircraft Analysis. software. and many methods can be used to estimate them prior to flight testing. Its results showed the aircraft was unstable in several different modes. The YAK54 results from both programs were very close to each other and also to the flight test results. The topic of this research is using high fidelity computer software. VORSTAB and FLUENT. due to the complexity of the aircraft design. the MantaHawk.Abstract The University of Kansas has long been involved in the research and development of uninhabited aerial vehicles. to determine the flying qualities of three different UAVs. A major problem with the current design of these UAVs is that very little effort was put into the aerodynamics. AAA. but is a very useful tool for design.100 lb UAV has been designed. Currently a 1. The results from the other two UAVs varied largely. Flight test data was also used to help determine how well each program estimated the stability and control derivative or flying qualities. AAA is not considered to be as accurate as CFD. and flown from the University. v . VORSTAB had a very difficult time handling the complex body of the Meridian. UAVs.
This software is a widely accepted program that is known to produce very accurate results. vi .From these results it was determined that VORSTAB. If the aircraft is a traditional style aircraft with noncomplex geometry VORSTAB will return highly accurate results that are better than AAA. FLUENT was used to determine the possible downwash issue over the Meridian fuselage. It is rather simple to determine if the VORSTAB results are valid or not. These results are all discussed throughout the report in detail. while a high fidelity program. and this can outweigh the improved results. and the input file can be easily improved to increase the accuracy of the results. It is always a smart idea to use both software programs to check the results with one another. The benefit of AAA is that a model can be created rather quickly and the results will typically be within an acceptable error range. The major problem is that it is very time consuming to make a model and requires someone with a large amount of knowledge about the software to do so. FLUENT results showed a possibility for a large boundary layer near the tail and flow separation at high angles of attack. A VORSTAB model can be very time consuming to make. has difficulty handling aircraft with complex geometry.
............................1 2 Literature Review......................................................................................Table of Contents Page # Acknowledgements ......................3 Theoretical Aerodynamics in Today‟s Real World.................................................. 24 6 FLUENT ............................................ 6 3 Stability and Control Derivatives................................... 77 9 Meridian UAV ............................................................................................................................................................................... 8 3..... ix List of Tables ................................ 73 8............................56 8............23 5..................................................1 Creating a Model.....22 5 VORSTAB .............................................................. 51 7...............3 Method Comparison MantaHawk ....................................v List of Figures ........................................................................................................................................................................................................................................................ 141 9................ 20 4 AAA ...............................................................8 3.......... 14 3...............................169 vii ..............................................4 The Lockheed SR71 Blackbird – A Senior Capstone ReEngineering Experience ............................3 2................3 Linearized Model of the Meridian .......................................... Opportunities and Challenges .1 AAA Modeling of the Meridian ............................................. 80 9................................................................ 135 9............................................. 135 9.............................................1 Wavelike Characteristics of Low Reynolds Number Aerodynamics .............................................................................2 VORSTAB Modeling of the MantaHawk ............1 AAA Modeling of the YAK54 ...................................... 83 9................................................................................................................. 3 2............................................................................................................................................2 A Generic Stability and Control Methodology For Novel Aircraft Conceptual Design ..........xv 1 Introduction ....................................165 11 References ......................................................... 35 7.4 Linearized Model of the MantaHawk ...................... 60 8.................2 VORSTAB Modeling of the Meridian ...1 Longitudinal Motion .................. 158 10 Conclusions and Recommendations ........................ iii Abstract ........ 4 2....................4 FLUENT Modeling of the Meridian .......................................................... 32 7............................................................................................3 Perturbed State ................. 57 8...............................................................................4 Linearized Model of the YAK54 .............. 5 2... xiii List of Symbols .........................................................29 7..................................................................2 VORSTAB Modeling of the YAK54 .......................................................................................78 9.............................................26 7 YAK54................................1 AAA Modeling of the MantaHawk .............................................................................................................................................................................................................................2 LateralDirectional Motion ................................................................................................3 Method Comparison YAK54.............................................6 Method Comparison Meridian ...............................5 FLUENT Model Generation ................................................................................................................................................................ 55 8 MantaHawk ..................................................................................................
...................................................................................................................Table of Contents Continued Appendix A ...................................................................................172 viii .............171 Appendix B ..............
... 65 Figure 29: MantaHawk VORSTAB Lift Coefficient due to Symmetrical Deflection 68 Figure 30: MantaHawk VORSTAB Drag Coefficient due to Symmetrical Deflection ................... 57 Figure 21: MantaHawk VORSTAB Lift Coefficient...........................................................List of Figures Figure 1: EarthFixed and BodyFixed Axes System ................................ 45 Figure 15: YAK54 VORSTAB Drag Coefficient due to Elevator Deflection ....................................................................................................................................... 42 Figure 12: YAK54 VORSTAB Rolling Moment Coefficient due to Aileron Deflection ................ 64 Figure 27: MantaHawk VORSTAB LateralDirectional Derivatives due to Roll Rate .............................................................................................................................................................................................................................................................................................................................................. 63 Figure 26: MantaHawk VORSTAB LateralDirectional Derivatives due to Sideslip..... 37 Figure 7: YAK54 VORSTAB Longitudinal Derivatives ........................................................ 65 Figure 28: MantaHawk VORSTAB LateralDirectional Derivatives due to Yaw Rate ............................................ 61 Figure 22: MantaHawk VORSTAB Drag Coefficient............................................... 45 Figure 16: YAK54 VORSTAB Pitching Moment Coefficient due to Elevator Deflection .......................................................... 43 Figure 14: YAK54 VORSTAB Lift Coefficient due to Elevator Deflection ................................ 36 Figure 6: YAK54 VORSTAB Pitching Moment Coefficient Curve ................................................................................................ 49 Figure 20: MantaHawk ........................ 46 Figure 17: YAK54 VORSTAB Sideforce Coefficient due to Rudder Deflection ..................... 31 Figure 4: YAK54 VORSTAB Lift Curve Slope ................................................................................................................ 37 Figure 8: YAK54 VORSTAB LateralDirectional Derivatives due to Sideslip ........................................................................................................................ 39 Figure 9: YAK54 VORSTAB LateralDirectional Derivatives due to Roll Rate .............................................................. 62 Figure 24: MantaHawk VORSTAB Lift Coefficient due to Pitch Rate ........................................................................................................................................... 39 Figure 10: YAK54 VORSTAB LateralDirectional Derivatives due to Yaw Rate ............................. 9 Figure 2: YAK54 ................................................................. 68 ix ..................... 42 Figure 13: YAK54 VORSTAB Yawing Moment Coefficient due to Aileron Deflection ... 48 Figure 18: YAK54 VORSTAB Rolling Moment Coefficient due to Rudder Deflection . 62 Figure 25: MantaHawk VORSTAB Pitching Moment Coefficient due to Pitch Rate ..................................................................... 36 Figure 5: YAK54 VORSTAB Drag Curve............................. 48 Figure 19: YAK54 VORSTAB Yawing Moment Coefficient due to Rudder Deflection ..... 40 Figure 11: YAK54 VORSTAB Sideforce Coefficient due to Aileron Deflection ......................................................................................................................................................................................................... 61 Figure 23: MantaHawk VORSTAB Pitching Moment Coefficient .............. 29 Figure 3: Unigraphics CAD Model of YAK54 ........................
............................................................. 91 Figure 43: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to Yaw Rate ............................................................................... 91 Figure 42: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to Roll Rate ............................................................................................................................................................................................ 92 Figure 44: Meridian VORSTAB Model 2 Sideforce Coefficient due to Aileron Deflection ............................................................................................. 96 Figure 47: Meridian VORSTAB Model 2 Lift Coefficient due to Elevator Input ..................................... 95 Figure 46: Meridian VORSTAB Model 2 Yawing Moment Coefficient due to Aileron Deflection ............................................. 89 Figure 40: Meridian VORSTAB Model 2 Longitudinal Derivatives due to Pitch Rate ............... 72 Figure 35: Unigraphics CAD Model of the Meridian UAV ................................................................................................................................................... 99 Figure 50: Meridian VORSTAB Model 2 Sideforce Coefficient due to Rudder Input .......................................................................................... 88 Figure 38: Meridian VORSTAB Model 2 Drag Coefficient due to Angle of Attack.................................................................................................................................................................................................................................... 98 Figure 48: Meridian VORSTAB Model 2 Drag Coefficient due to Elevator Input ........................... 103 Figure 54: Meridian VORSTAB Model 3 Drag Coefficient due to Angle of Attack ................... 101 Figure 52: Meridian VORSTAB Model 2 Yawing Moment Coefficient due to Rudder Input ................................................................ 89 Figure 41: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to Sideslip........... 98 Figure 49: Meridian VORSTAB Model 2 Pitching Moment Coefficient due to Elevator Input..................... 104 x ......................................... 79 Figure 36: Meridian Empennage Models 1 and 2 ............................................................................................................... 69 Figure 32: MantaHawk VORSTAB Sideforce Coefficient due to Asymmetrical Deflection ......... 86 Figure 37: Meridian VORSTAB Model 2 Lift Coefficient due to Angle of Attack..................................... 102 Figure 53: Meridian VORSTAB Model 3 Lift Coefficient due to Angle of Attack.......................................................................................................... 88 Figure 39: Meridian VORSTAB Model 2 Pitching Moment Coefficient due to Angle of Attack.....................................................................List of Figures Continued Figure 31: MantaHawk VORSTAB Pitching Moment Coefficient due to Symmetrical Deflection ..................................................................................................................... 101 Figure 51: Meridian VORSTAB Model 2 Rolling Moment Coefficient due to Rudder Input ................................................................................................................................................................................................................................. 71 Figure 33: MantaHawk VORSTAB Rolling Moment Coefficient due to Asymmetrical Deflection ...... 71 Figure 34: MantaHawk VORSTAB Yawing Moment Coefficient due to Asymmetrical Deflection ............................................................................................................................................................................ 95 Figure 45: Meridian VORSTAB Model 2 Rolling Moment Coefficient due to Aileron Deflection ...............................................................
........................... 109 Figure 62: Meridian VORSTAB Model 3 Yawing Moment Coefficient due to Aileron Deflection ........... 107 Figure 60: Meridian VORSTAB Model 3 Sideforce Coefficient due to Aileron Deflection ........................................ 104 Figure 56: Meridian VORSTAB Model 3 Longitudinal Coefficients due to Pitch Rate ......................... 112 Figure 65: Meridian VORSTAB Model 3 Pitching Moment Coefficient due to Elevator Input..................................................................................................................................................................................................................................................................................... 124 xi ................................. 105 Figure 57: Meridian VORSTAB Model 3 LateralDirectional Derivatives due to Sideslip.............................. 106 Figure 58: Meridian VORSTAB Model 3 LateralDirectional Derivatives due to Roll Rate ................................................................................................. 112 Figure 64: Meridian VORSTAB Model 3 Drag Coefficient at due to Elevator Input ......................... 115 Figure 67: Meridian VORSTAB Model 3 Rolling Moment Coefficient due to Rudder Input ............................................................... 115 Figure 68: Meridian VORSTAB Model 3 Yawing Moment Coefficient due to Rudder Input .................................................................................................................... 117 Figure 70: Meridian VORSTAB Model 4 Drag Coefficient due to Angle of Attack ............................................................................................... 109 Figure 61: Meridian VORSTAB Model 3 Rolling Moment Coefficient due to Aileron Deflection ............................................................................................................................................................................... 119 Figure 73: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to Sideslip........................................................................................................................................ 113 Figure 66: Meridian VORSTAB Model 3 Sideforce Coefficient due to Rudder Input ........................................................ 106 Figure 59: Meridian VORSTAB Model 3 LateralDirectional Derivatives due to Yaw Rate ........................................ 118 Figure 72: Meridian VORSTAB Model 4 Longitudinal Derivatives due to Pitch Rate ............................................................................................................................................................................................................................................................................................................................................... 110 Figure 63: Meridian VORSTAB Model 3 Lift Coefficient due to Elevator Input ...................................... 116 Figure 69: Meridian VORSTAB Model 4 Lift Coefficient due to Angle of Attack........................................................................................... 118 Figure 71: Meridian VORSTAB Model 4 Pitching Moment Coefficient due to Angle of Attack........................................................................................................ 121 Figure 76: Meridian VORSTAB Model 4 Sideforce Coefficient due to Aileron Deflection ............................................................................................................................................................................................................................................................................................................................ 120 Figure 74: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to Roll Rate ......................List of Figures Continued Figure 55: Meridian VORSTAB Model 3 Pitching Moment Coefficient due to Angle of Attack................................................................................. 121 Figure 75: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to Yawing Moment ..................................................................................................................................
.......................................... 124 Figure 78: Meridian VORSTAB Model 4 Yawing Moment Coefficient due to Aileron Deflection ............................................................................ 137 Figure 86: Meridian Fuselage with Farfield Divisions ............... 157 Figure 95: Circulation around an Airfoil Producing Lift ................................................ 150 Figure 91: Velocity Profile with Flow Separation at Point S (dashed line u = 0) ........................................................................................................................................................................................................ 151 Figure 92: Meridian FLUENT Total Pressure Profile for All Angles of Attack .................... 131 Figure 85: Meridian Fuselage Axial CrossSection for FLUENT ................. 125 Figure 79: Meridian VORSTAB Model 4 Lift Coefficient due to Elevator Input ........................................................................................... 127 Figure 80: Meridian VORSTAB Model 4 Drag Coefficient due to Elevator Input ................................................................. 130 Figure 83: Meridian VORSTAB Model 4 Rolling Moment Coefficient due to Rudder Input .............................................................................................List of Figures Continued Figure 77: Meridian VORSTAB Model 4 Rolling Moment Coefficient due to Aileron Deflection ........................................................................................................................................ 130 Figure 84: Meridian VORSTAB Model 4 Yawing Moment Coefficient due to Rudder Input ........................................................ 138 Figure 87: Meridian Farfield Meshes .......................................................................................................................................................................................................................... 158 xii .......................... 145 Figure 89: Meridian FLUENT Velocity Angle Profile for All Angles of Attack ....................... 152 Figure 93: Meridian FLUENT Stream Function Profile at All Angles of Attack ............................................ 140 Figure 88: Meridian FLUENT Velocity Magnitude Profile for All Angle of Attacks ... 154 Figure 94: Meridian FLUENT Vorticity Magnitude Profile at All Angles of Attack ........................ 127 Figure 81: Meridian VORSTAB Model 4 Pitching Moment Coefficient due to Elevator Input..... 128 Figure 82: Meridian VORSTAB Model 4 Sideforce Coefficient due to Rudder Input .................................... 148 Figure 90: Velocity Profile in Boundary Layer with Favorable and Adverse Pressure Gradient............................
..List of Tables Table 1: YAK54 Lifting Surface Dimensions ......... 83 Table 29: Meridian VORSTAB Model 1 Longitudinal Derivatives (1/rad) ............................................................................................................................. 92 xiii .... 59 Table 18: MantaHawk VORSTAB Longitudinal Derivatives (1/rad) ............. 73 Table 23: MantaHawk AAA and VORSTAB Comparison . 52 Table 14: YAK54 Stability and Control Derivatives Typical Ranges .............. 60 Table 19: MantaHawk VORSTAB LateralDirectional Derivatives (1/rad) ........................................... 50 Table 13: YAK54 AAA and VORSTAB Stability and Control Comparison .............................................................. 35 Table 8: YAK54 VORSTAB LateralDirectional Derivatives (1/rad) .................................................................................. 81 Table 27: Meridian AAA LateralDirectional Derivatives ............................... 70 Table 22: MantaHawk VORSTAB Stability Requirements ........................... 44 Table 11: YAK54 VORSTAB LateralDirectional Due to Rudder Deflection (1/rad) .............................................................................................................................................................................................. 41 Table 10: YAK54 VORSTAB Longitudinal Due to Elevator Deflection (1/rad) ................................................... 86 Table 32: Meridian VORSTAB Model 2 Longitudinal Control Derivatives and Affected by Pitch Rate ....................................................................................... 76 Table 25: Meridian Characteristics ........................................ 82 Table 28: Meridian AAA Stability Requirements .................. 33 Table 5: YAK54 AAA LateralDirectional Derivatives........ 64 Table 20: MantaHawk VORSTAB Longitudinal Derivatives due to Symmetrical Deflection ....... 87 Table 33: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to Sideslip and Roll Rate....... 30 Table 2: YAK54 Flight Conditions .................................. 32 Table 4: YAK54 AAA Longitudinal Derivatives............................................................................................................................................................ 34 Table 6: YAK54 AAA Stability Requirements .. 58 Table 16: MantaHawk AAA LateralDirectional Derivatives .................................................. 54 Table 15: MantaHawk AAA Longitudinal Derivatives ...................................................... 74 Table 24: MantaHawk Stability and Control Derivatives Typical Ranges........................... 84 Table 31: Fuselage Diameter for Meridian Models 1 and 2 ................................................. 84 Table 30: Meridian VORSTAB Model 1 LateralDirectional Derivatives (1/rad)................ 47 Table 12: YAK54 VORSTAB Stability Requirements ................................................................................................................................................................................................ 31 Table 3: YAK54 AAA Moment of Inertia and Trimmed Values................................................................................. 59 Table 17: MantaHawk AAA Stability Requirements .. 90 Table 34: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to Yaw Rate ..................................................................................................................... 35 Table 7: YAK54 VORSTAB Longitudinal Coefficient and Pitch Rate (1/rad) .... 38 Table 9: YAK54 VORSTAB LateralDirectional Due to Aileron Deflection (1/rad) ...................................................................................... 67 Table 21: MantaHawk VORSTAB LateralDirectional Derivatives due to Asymmetrical Deflections ................................................... 79 Table 26: Meridian AAA Longitudinal Derivatives ......
......................................................... 108 Table 41: Meridian VORSTAB Model 3 Longitudinal Derivatives due to Symmetrical Tail Deflections ........................................................................ 129 Table 48: Meridian VORSTAB Stability and Control Derivatives Dr............. 143 Table 53: Meridian Fuselage Lift and Drag Method Comparison ..... 161 Table 55: Meridian VORSTAB and AAA Stability and Control Derivatives Typical Ranges ................... 175 Table 59: MantaHawk LateralDirectional Modes .......................................... 179 Table 63: Meridian Model 3 LateralDirectional Modes ............................................................................... 177 Table 61: Meridian Model 2 LateralDirectional Modes ........ 117 Table 44: Meridian VORSTAB Model 4 LateralDirectional Control Derivatives ............................................................................................................................................................ 176 Table 60: Meridian Model 2 Longitudinal Modes ............................................................................................................................................................................................................................................................ 105 Table 40: Meridian VORSTAB Model 3 LateralDirectional Derivative due to Aileron Deflection ..... 178 Table 62: Meridian Model 3 Longitudinal Modes ..................................................................................................... 139 Table 52: Meridian Fuselage CrossSection FLUENT Lift and Drag ...................................................... Roskam‟s Typical Ranges.......................... 103 Table 39: Meridian VORSTAB Model 3 LateralDirectional Derivatives ............................................................................ 111 Table 42: Meridian Model 3 LateralDirectional Derivatives Affected by Asymmetrical Tail Deflections ....... 126 Table 47: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to Asymmetrical Deflection ........................................................................................................................ 123 Table 46: Meridian VORSTAB Model 4 Longitudinal Derivatives due to Symmetrical Tail Deflection ..... 100 Table 38: Meridian VORSTAB Model 3 Longitudinal Derivatives at due to Pitch Rate ....................... 173 Table 57: YAK54 LateralDirectional Modes ....................................................................................... 120 Table 45: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to Aileron Deflection .................................... 159 Table 54: Meridian AAA and VORSTAB Comparison ................................... 134 Table 50: Meridian Farfield Edges ........................................................... 180 xiv .......................................................................................... 94 Table 36: Meridian VORSTAB Model 2 Longitudinal Derivatives Affected by Symmetrical Tail Deflections ..................................................................................... 139 Table 51: Meridian Edge Meshes ..........List of Tables Continued Table 35: Meridian VORSTAB Model 2 Lateral Directional Derivatives due to Aileron Deflection ................................................... 163 Table 56: YAK54 Longitudinal Modes ............................................................................................................ 133 Table 49: Meridian AAA and VORSTAB Comparison ............................................................. 114 Table 43: Meridian VORSTAB Model 4 Longitudinal Derivatives ............................................................................ 97 Table 37: Model 2 Lateral Directional Control Derivatives Affected by Asymmetrical Tail Deflection ............................................................. 173 Table 58: MantaHawk Longitudinal Modes ......................................................
....................................................... Variation of Airplane Pitching Moment due to Thrust ....................................................... Aspect Ratio .............................. CTxu .. 1/rad Cmq ...................................... 1/rad Cna........... 1/rad CmT ............... 1/rad Cnr ..................... 1/rad CLq .......... Cy .............. CL_ih ........ Variation of Airplane Yawing Moment with Rate of Change of Yaw Rate ... 1/rad CL ............... 1/rad Cla ............... 1/rad Cm ....................................................................... Variation of Airplane Rolling Moment with Rate of Change of Yaw Rate ................................ Variation of Airplane Yawing Moment with Rudder Deflection ................... 1/rad Clp ................. b . Variation of Airplane Lift with Elevator Deflection................................................ Variation of Airplane Pitching Moment with Pitch Rate ......... 1/rad Clr ............ Variation of Airplane Yawing Moment with Rate of Change of Roll Rate ........... Variation of Airplane Rolling Moment with Angle of Sideslip .......................................Variation of Airplane Drag with Speed ............................................... Variation of Airplane Lift with Angle of Attack ................ 1/rad Cme ................... Cm ........ Variation of Airplane Pitching Moment with Angle of Attack .................. 1/rad CDu ............................................................................ Variation of Airplane Pitching Moment due to Thrust with Angle of Attack ..................... Pitching Moment Coefficient about Aerodynamic Center ....... ft CD . 1/rad Clr ..... 1/rad Cnp ............................................................  xv ............................................................. ft ct . Variation of Airplane Drag with Pitch Rate ......... Variation of Airplane Pitching Moment with Elevator Deflection . Variation of Airplane Rolling Moment with Rudder Deflection ................................................................ 1/rad CLu .................. Root Chord ....... Variation of Airplane Yawing Moment with Angle of Sideslip .................................................................. Pitching Moment Coefficient .... 1/rad CL. Lift Coefficient . Cm_ac ...................... 1/rad Cn ............................ Cl ........... ft cr ...................................... CD ........................ 1/rad CnT ....... Cn... Variation of Airplane Lift with Differential Stabilizer Angle .................................................................... 1/rad Cnr .... 1/rad CLe ................ Variation of Airplane Rolling Moment with Aileron Deflection .........................................Sideforce Coefficient ............. Variation of Airplane Yawing Moment with Aileron Deflection................................................................................................... Yawing Moment Coefficient .................Variation of Airplane Rolling Moment with Rate of Change of Roll Rate ............ Mean Geometric Chord ........................................................................................... Variation of Airplane Lift with Pitch Rate ....................................................................... 1/rad CDq ................................. Variation of Airplane Drag with Elevator Deflection.................................................................................... Rolling Moment Coefficient ..................... CmT .................................... Speed of Sound .... Drag Coefficient ........... Variation of Airplane Yawing Moment due to Thrust with Sideslip ......................................................................................................................................... ft c̃ . Variation of Airplane Thrust in X Direction...... 1/rad CTx .......................................... ft/sec A or AR .................... Variation of Airplane Lift with Dimensionless Speed ..... Wing Span .................... 1/rad CDe ...... 1/rad Cl.......................................List of Symbols Symbol Description Units Normal a ....... Tip Chord .................................................... Variation of Airplane Thrust in Xaxis with respect to speed . Variation of Airplane Drag with Angle of Attack ......
. d/d ........................... Variation of Airplane Sideforce with Aileron Deflection........................ w ................................................ Three Dimensional .............................................. Two Dimensional Double Precision ................ Area....................... Roll Rate . ft xcg .........c and Airplane c.................................... Distance from Center of Gravity and Reference Point ..... ftlbs M∞....................................................................................................................... rad/sec Ps ................................................................................... ft/sec U ......................... Yawing Moment .................................................................................................................... 1/rad Cyr ............................................................................................................................................ slugsft2 Ixz ............................... Oswald‟s Efficiency Factor ... rad/sec q =0................................................................................ Flow Distance Along Body............................................................ Airplane Products of Inertia about XYZ................................... boundary layer .... lbs/ft2 q .................... ....................................................................................................................... ftlbs N.............. 2ddp ..................... ft/sec v .............. lbs e . ft2 Re . ft N ........................................... ... deg Ixx............Distance from Aerodynamic Center and Reference Point .................................... Volume Coefficient of Tail . Sideforce ... 3ddp .............................................. Free Stream Mach Number ......................................................................................................................... Lift .................................................................................................................. Pitching Moment .......................... 3d ...  xvi .............. Fy ............5V2 ................................ Variation of Airplane Sideforce with Angle of Sideslip.................. M .................................................................. Axial Distance Between Vertical Tail a............................................................................................... lbs L... rad/sec u ..... Horizontal Tail Incidence Angle......................................................................................................................................................................... lbs/ft2 PT .................. ft 2d ...................................................................................................................................................................................... D ........................ Reynolds Number .......................................................... Izz.... Three Dimensional Double Precision ..... 1/rad Cya............................................... ft xac............................... Velocity.................................. Axial Station Along Fuselage ......... Height ...................................... zcomponent of Velocity ........................................................................................................................................................................ycomponent of Velocity .............xcomponent of Velocity ............ Downwash Gradient ........ Static Pressure........................... Two Dimensional ........................................................................................................................................................ Drag ...................................... Axial Free Stream Velocity ................................................ Airplane Moments of Inertia about XYZ .....g...... Variation of Airplane Sideforce with Rate of Change of Yaw Rate......................Sidewash Gradient .............................................................. lbs h .......................................................................................... ft/sec Ṽ ........................D........ ft xvs ............................................... slugsft2 L .............................. Total Pressure ...................................... Yaw Rate ................................ Aircraft Dynamic Pressure .. ft/sec S .......... Nondimensional................................ x .......................................... Pitch Rate ............ Variation of Airplane Sideforce with Rudder Deflection ...... 1/rad d/d.. Iyy...... ft ih ...................................................................................... p ...............Normal Continued Cy................ 1/rad Cyp .. 1/rad Cyr ... lbs/ft2 r........... ft/sec V ................................................................... Variation of Airplane Sideforce with Rate of Change of Roll Rate ........................................
..................Change in ............ Angle of Attack Effectiveness Factor ............. rad/sec Stream Function ......................... Computational Fluid Dynamics CReSIS ................................................................ Fuselage h.................. Coefficient at Zero Deflection r ...........................  .. True Airspeed UAV ........................................................................................................................................... Horizontal Tail o...................................................................................... Angle of Attack...................................................................14 .................................el.......................  ......................................................................... Rudder v......... Wing Fuselages 1............................................................. Personal Computer TAS ................................................................................................................a............................................................................. Computer Aided Design CFD ................................................................................. Dynamic Pressure Ratio ........................................................................................................ Uninhabited Aerial Vehicle Subscripts a ............... Control Surface Deflection .......................................... deg ......... Kinematic Viscosity ................................................... Wing wf ............................................................................................................. Vertical Tail w.................. deg _dot ............... Circulation .................... Sideslip ................... Rate of Change of Angle of Attack ..................... Aileron B ............................................................... Elevon f ... Displacement Thickness .................... Angular Velocity ...................................................................... Taper Ratio ................................................................. deg ̃ ................................................................................................................................................................................................... Steady State Condition xvii ....................... Center for Remote Sensing of Ice Sheets KUAE ................................................. University of Kansas Aerospace Engineering NSF .................. Average Boundary Layer Thickness...........................................................................  Acronyms AAA ........... National Science Foundation PC......................................................................... Advanced Aircraft Analysis CAD ..... Elevator el .............................................. Body Stability Axis System e .................. 3.........  ................. * ........................  ........................................ deg/sec2 ........................ lbs/ft3 ................................................................. m2/sec ............................  ........................................ e................. m2/sec .................................. Density .... Momentum Thickness.................Greek ..............................................................r ....................................
Also. There are several equations that can be used to estimate some of the derivatives. This software is expensive to purchase. Therefore. due to the changes in Reynolds number and other flow characteristics. but it still requires an expert to create a 3D full aircraft CFD model.1 Introduction The shape and design of an aircraft can dramatically influence how the aircraft handles and is controlled. Using an experienced wind tunnel expert and a highly accurate tunnel can minimize these problems. Computational Fluid Dynamic software is much more accurate than it once was and is becoming more user friendly. Also. but can be used over and over again. Over the past couple of decades computer simulation has become much more prevalent. better methods have to be used before investing millions of dollars on an aircraft. but will be very expensive. Wind tunnel tests. and this can have a dramatic change on the results. since the results do not always scale up as easily. These equations are just estimations and can be magnitudes off. many different test cases can be run to determine flying qualities in various situations. There are also several different programs that are readily available that can produce 1 . The mesh generation for a model can be difficult and requires a great deal of experience. Air will flow over a smaller body differently than a larger body. are a method that results in derivatives that are highly accurate. The stability and control derivatives are essential for flight simulation and handling qualities. The problems with wind tunnel tests are that it is very expensive and can be very time consuming. the wind tunnel models are scaled down to fit in the tunnel. but not all of them.
100 lb aircraft that was designed to fly in the Polar Regions. FLUENT is a very high fidelity CFD program. to generate the derivatives quickly. High level CFD programs can be very expensive and time consuming when performing aerodynamic analysis. 2 . The 1/3 scaled YAK54 model was a production aircraft purchased by the University of Kansas. but requires a large amount of experience and time. This is why engineers prefer to use engineering level programs. The stability and control derivatives found from each software program were compared to each other and conclusions about the software were drawn. The main goal of this research was to use high fidelity CFD programs to test the validity of these engineering level programs. such as AAA.high fidelity results. Three different computer programs were used to determine the stability and control derivatives on three different UAVs. Advanced Aircraft Analysis (AAA) and VORSTAB were used on all three aircraft and FLUENT was also used on the Meridian. and some of these programs can be purchased at a reasonable price. The Meridian is a 1. Two of the aircrafts were being designed and built to fly while the other one was already a production aircraft. and the Meridian and MantaHawk were designed at the University of Kansas.
the literature review topics consisted of stability and control analysis software.2 Literature Review It is a wise idea to examine current and past research going on in the field of study. using high fidelity CFD programs. acoustic disturbances. The research that was conducted in this paper is aerodynamic analysis. and sinusoidal wave theory. or has previously been examined in the past. and the CFD software that was used in this research. low Reynolds number aerodynamics. the drag variation can be modeled as a sinusoidal wave along the span. These patterns include a drag increase greater than the rate of increasing lift. to determine the stability and control derivatives of UAVs with low Reynolds numbers. A brief summary of each article will be given and then the conclusion drawn from these papers. This gives the researcher a chance to see what is currently going on. but due to the low Reynolds number the flow can travel in the spanwise direction. Therefore. and variances in drag across the span of the lifting surface. boundary layer theory.1 Wavelike Characteristics of Low Reynolds Number Aerodynamics Lifting surfaces will demonstrate several uncharacteristic flow patterns when flown at a low Reynolds number. 2. It can also give the researcher ideas on topics and experiments to conduct. Using velocity potential theory. Spanwise flow can usually be ignored at higher velocities. These results were then comparedt to the research conducted by Guglielmo 3 . This has a great effect on the design of micro or small UAVs since they have rather small Reynolds numbers.
and Selig, where the drag magnitude was observed to be happening in a wave form along the span. The goal of this research was not to exactly match the Guglielmo and Selig data, but to demonstrate that the drag magnitude and flow can be modeled using sinusoidal wave theory. This goal was successfully accomplished even though it did not match the trend observed by Guglielmo and Selig. All of this can be found in detail in Ref [1]. This research shows how the low Reynolds number can affect the flow around the aircraft‟s lifting surfaces. At low speeds the drag magnitude can vary along the span of the lifting surface, and in turn this can dramatically affect the other stability and control of the aircraft. If the flow is traveling at different speeds and in different directions (spanwise) the aircraft will not react how it typically would at higher speeds. The control surfaces would not have the same impact when the flow is varying. 2.2 A Generic Stability and Control Methodology For Novel Aircraft Conceptual Design Stability and control is the most serious requirement for flight safety, and yet there is not a standard or reliable method for determining stability and control in the design phase. There are several methods used and many are considered acceptable within the industry. A major weakness, of most methods, is the design and sizing of the control effectors. Currently very simple methods are used for the sizing, and are done so in the cruise, landing, and takeoff conditions of the flight envelope. This
4
research shows that sizing should be done so in the grey areas of the flight envelope, where nonlinear aerodynamics prevail. A method for generating stability and control was designed over a four year period and is called AeroMesh. This method is capable of handling both
conventional and unconventional and symmetric or asymmetric flight vehicles. Design constraints and various flight conditions are first implemented into the program. An input file is then created for a CFD program called VORSTAB. This CFD software will estimate the stability and control derivatives as well as determine the size, position, and hinge lines of the control effectors. A 6 degreeoffreedom model is then used to determine stability and control in the trimmed and untrimmed condition. This 6DOF model uses control power to determines the stability and control derivatives in the trimmed and untrimmed conditions. All information was taken from Ref [2]. This research shows just how important a high fidelity CFD program can be do the design of the control effectors and their sizing. The design criteria are at the extremes of the flight envelope, so the control effectors are designed at the point of nonlinear aerodynamics. Using VORSTAB can help eliminate the use of simple methods that are low fidelity. 2.3 Theoretical Aerodynamics in Today’s Real World, Opportunities and Challenges CFD has revolutionized the aerodynamic industry, but it still faces many challenges in predicting and controlling various flows. These flows include UAV
5
low Reynolds number, high angle of attack, boundary layer transition, threedimensional separation, and others. Since this is the case it is wise to combine theoretical, computational, and experimental approaches when analyzing the flow. These problems that a typical CFD program, has with the flow, is discussed and ways to analytically solve these problems are given. Multiple approaches are applied to the flow to find solutions. Using all three of the solution methods allows the users to see the short comings of each method. It is a very important and critical skill set to know and understand how to set up a problem up from the beginning, and then make approximations using mathematical and physicsbased models. This principle should then be applied to a modern computational method. All information was taken from Ref [3]. It can be seen that not only a CFD program should be used during the design process, but also other methods. The research in this report covers both analytical methods, AAA, and high fidelity methods, VORSTAB and FLUENT. Understanding how to set up problems is very important due to the high complexity of modern CFD programs. A small error in the input can dramatically influence the results. It is also very important for the user to be able to interpret the results, and this skill set comes from understanding the theoretical methods. 2.4 The Lockheed SR71 Blackbird – A Senior Capstone ReEngineering Experience At the University of Texas at Arlington, the senior aerospace class reengineered the Lockheed SR71 Blackbird in a two part design course. Currently in
6
A CAD model of the aircraft was created. but also the Dr. All information was taken from Ref [4]. 7 . The Roskam method is outlined in an eight book series about aircraft design. it is very rare for a company to start a design from scratch. This second method is the same as using the AAA software. and the results were then compared to one another. but rather add to or modify previous research. VORSTAB was the primary method for obtaining the derivatives. There were no changes made to the SR71 model.the aerospace industry. but rather the aircraft was reanalyzed. and from there the stability and control derivatives were determined in a two different ways. Roskam method was used. The reengineering of the SR71 by the senior design class at the University of Texas at Arlington is very similar to the topic of this thesis. The VORSTAB results were off by an order of magnitude. Both methods derivatives were then compared to the actual SR71 data. All other aspects of the design process were completed ranging from aircraft systems to flight performance. but followed the correct trends with the exception of the yawing moment coefficient due to sideslip and yawing moment coefficient due to roll rate. Several methods of analysis were used to determine the stability and control derivatives.
1 Longitudinal Motion Longitudinal motion is due to the forces and moments of the aircraft in the xz plane in terms of the stability axes system. and any other surface on the aircraft. 8 . fuselage. tails. 3. it will be assumed that the reader has this basic knowledge. The flow around an entire aircraft is too complex to allow formulas to determine the derivatives. lift force. as well as the deflection of control surfaces. longitudinal motion and lateraldirectional motion. The aircraft orientation also affects these three. To define and understand the stability and control derivatives one must have a basic understanding of aerodynamic principles.3 Stability and Control Derivatives The stability and control derivatives come from the aerodynamic forces and moments acting on upon the aircraft components. These forces and moments include drag force. These forces and moment will be discussed in subsections of this chapter. Coupling between longitudinal and lateraldirectional dynamics is assumed to be zero for stability and control derivative estimation. The aircraft forces and moments are broken into two distinct directional motions. and pitching moment. Wind tunnel tests or high fidelity computational fluid dynamics should be used to estimate the control derivatives with a high level of accuracy. Figure 1 from Ref [5]. These components are defined as the wings.
and induced drag is the drag produced due to the production of lift. and dynamic pressure. angle of attack. The following are influences on the drag: airplane wetted area. Eq [1] and Eq [2] are two equations for determining drag. skin friction. Eq [1] Eq [2] 9 .1.Figure 1: EarthFixed and BodyFixed Axes System (From Ref [5]) 3. speed. There are two types of drag that contribute to the entire aircraft drag. Parasite drag is drag due to the shape of the aircraft when there is zero lift produced. CD Drag is the force that acts in the opposite direction of motion. Ref [5] and Ref [6] respectively.1 Drag Coefficient. control surface deflection.
3. CL The lift is defined as the force acting on a surface that is perpendicular to oncoming flow in the upward direction.4 Lift Coefficient. Lift is what keeps the aircraft in the air. CD. The aircraft wings. This means that it is also perpendicular to the drag force vector. fuselage. when trim drag is important. and therefore it is important for the overall drag. but high fidelity computer simulations will still calculate this derivative. 3.1. It is acceptable to neglect the change in drag due to these control surface deflections in low speeds. Typically.1. drag increases as the angle of attack increases. This derivative.3 Change in Airplane Drag due to Change in Elevator Deflection. or in other terms. The elevator deflection might be used to trim the aircraft. 3. It also increases as the Mach number increases. CD This coefficient represents variation of the drag coefficient with angle of attack. The horizontal tail is affected by 10 . can be estimated by differentiating Eq [2] which then produces Eq [3].2 Change in Airplane Drag due to Change in Angle of Attack. the drag changes as the angle of attack moves away from the steady state condition.1. this increase in drag cannot be neglected. However. described by the derivative CDe. Ref [6]. This derivative is determined at an angle of attack of zero. and tails all can produce lift. CDe Eq [3] Deflection of the elevator produces an incremental change in aircraft drag. The fuselage does not produce much lift. and is usually found as a wingfuselage combination.
Methods for estimating these lift coefficient values can be found in Ref [7]. the CL increases. the flow. As the angle of attack increases. 11 . the lift will begin to decrease because the aircraft has reached a stall point. The angle of attack also has an effect on every lifting surface. the aircraft lift will increase until the stall point.1. 3. or jet exhaust. Once a certain angle of attack is reached. propeller slip stream.5 Change in Airplane Lift due to Change in Angle of Attack. As the subsonic speed increases. the downwash or flow off the wings could blanket out the tail and render them ineffective. Eq [6] can be used to determine the lift coefficient due to angle of attack. At certain angles of attack. Ref [5]. CL Similar to the drag.the fuselage boundary layer. The angle of attack of the stall point is dependent upon the geometry of the aircraft. the downwash effect has to be accounted for. and the altitude. Eq [4] Eq [5] The wingfuselage combination and horizontal tail lift coefficients can be broken into components. Ref [5]. but around transonic region it begins to decrease. This is why an aircraft has a maximum angle of attack. Ref [5]. Eq [6] is found by partial differentiation of the firstorder Taylor series of the lift Eq [7]. the lift is affected by the angle of attack of the aircraft. This allows the aircraft to climb. The following equations can be used to determine the overall lift of the aircraft. Therefore. calculated for zero angle of attack. when determining the horizontal tail lift.
CLe Eq [6] Eq [7] The deflection of the elevator will change the camber of the horizontal tail airfoils and therefore change the lift of those airfoils. This is also found by partial differentiation of the firstorder Taylor series of lift Eq [7]. the center of pressure location changes cause the change in the pitching moment. The effect of the elevator deflection on the total aircraft lift coefficient can be found in Eq [8]. Ref [5]. Highly cambered airfoils usually have higher lift. or rotate upwards and downwards.1. The point that the aircraft rotates about is typically defined as the center of gravity. the lift of the horizontal tail will increase or decrease. Therefore. Aerodynamic center is the point about which the pitching moment coefficient does not vary with angle of attack. Most aircraft are inherently stable as long 12 . Center of pressure is defined as the point at which the pitching moment coefficient is equal to zero. Cm Eq [8] This is defined as the aerodynamic force that creates a moment that causes the aircraft to pitch. These two points create the pitching moment coefficient. 3.6 Change in Airplane Lift due to Change in Elevator Deflection. Ref [5]. Lifting forces create this resultant force that causes the aircraft to pitch.7 Pitching Moment Coefficient. 3. depending on the camber direction of the horizontal tail and the direction of deflection.1. As the angle of attack changes the center of pressure location will change.
For example. Ref [5]. if the aircraft that is statically stable is pitched upward it naturally returns to steady state and pitches down or viceversa if pitched downward. The horizontal tail has a large affect on this since it is used to pitch the aircraft. the center of pressure can move forward and aft as the angle of attack changes. This results in a changing moment arm and an increasing or decreasing pitching moment. If it was not stable. Elevator deflection and angle of attack can dramatically change the pitching moment. 13 . Cm is called the static longitudinal stability derivative which should be negative for an inherently longitudinally stable aircraft. This derivative.as the center of gravity is ahead of the aerodynamic center. Cm As stated previously. Eq [9] Eq [10] 3. The angle of attack also changes the lift on the aircraft.8 Change in Airplane Pitching Moment due to Change in Angle of Attack. Eq [11] can be used to estimate this derivative. The following equations define the pitching moment coefficient and estimation of this value. and therefore changes the aerodynamic force that creates the pitching moment. Horizontal tail incidence angle can dramatically affect this derivative due to the lift it creates on the tail.1. the aircraft would want to continue pitching upward and could flip over. Ref [5].
The sideslip angle can be thought of as the directional angle of attack. and the yawing motion is referred to as the directional motion. The effectiveness of the elevator is basically due to the volume coefficient of the horizontal tail. Ref [5]. The forces and moments that are defined in the lateral 14 .1. The larger the size of the elevator is. a fully moving horizontal tail has just as much effect as the incidence of the horizontal tail.9 Change in Pitching Moment due to Change in Elevator Deflection. and the angle of attack effectiveness of the elevator. This is the angle of directional rotation from the aircraft centerline to the direction of the wind. Eq [12]. Ref [5]. Eq [13] is used to estimate the derivative. . Cme This derivative is referred to as the longitudinal control power derivative and is typically negative. where sideslip plays a large role in lateraldirectional motion. For example. 3. These two motions are results of control surface deflections and sideforces. the more effect it has on the pitching moment. Eq [11] 3.2 LateralDirectional Motion Eq [12] Eq [13] The rolling motion is referred to as the lateral motion.
The control surfaces that affect the rolling moment are the aileron (lateral control surface) and rudder (directional control surface). yawing moment. Cl Eq [14] Eq [15] This derivative is often referred to as the airplane dihedral effect.2. Cl The rolling moment is the aircraft‟s rotation about the xaxis in the stability coordinate system. This will cause a higher lift on that wing and in turn cause the airplane to roll. If the aircraft is at a sideslip and has a dihedral angle on the wings.directional motion are sideforce. and airspeed. and those are sideslip. Similar to longitudinal control. 3. Rolling moment derivative due to sideslip can be estimated by summing the dihedral effect of 15 . 3. Several different things can cause and influence the rolling moment. the moment reference center (usually center of gravity). one of the wings will be hit with more air than the other.2. angle of attack. Elevator deflection influence can usually be ignored since the deflections are symmetrical and theoretically cancel each other out.2 Change in Airplane Rolling Moment due to Change in Sideslip. there are several variables that affect these forces and moments.1 Rolling Moment. deflection of control surfaces. Ref [5]. and rolling moment. Eq [14] is the dimensional form of the rolling moment and Eq [15] shows the first order Taylor series form of the rolling moment. The reason for this is because the airplane dihedral angle can have a huge influence on the rolling moment especially when at a sideslip.
These aileron deflections will also produce a yawing moment. but due to the typical location of the vertical tail and rudder a rolling moment is produced. Ref [5]. The resultant sideforce is typically located above the center of gravity and 16 . The vertical tail will also see a higher sideforce when the aircraft is at a sideslip.4 Change in Airplane Rolling Moment due to Change in Rudder Deflection. the free stream air will encounter the rudder and produce a sideforce.3 Eq [16] Change in Airplane Rolling Moment due to Change in Aileron Deflection.the individual components of the aircraft. the wings‟ location on the fuselage can affect the direction that the aircraft will want to roll. Flow separation can also occur with large aileron deflections. With the rudder deflected. This produces a rolling moment by decreasing the lift on the right wing due to the negative camber of the aileron. 3. Cla A positive aileron deflection is defined as the right aileron up and the left aileron down. Clr The purpose of the rudder is to produce a yawing moment. and this can reduce the effectiveness of the ailerons. For a detailed explanation and ways to estimate the individual components of the dihedral effect refer to Ref [5]. and this is why most ailerons are deflected differentially. This differential deflection will help minimize the yawing moment that is produced. Eq [16]. 3. and increasing the lift on the left wing due to the positive camber of the aileron. For example. There are many factors that play into the individual components‟ dihedral effect.2.2.
will produce this rolling moment. because there is more surface area for the sideslip free stream to contact. and control surface deflection. The wings‟ contribution depends on the dihedral angle. angle of attack. Eq [17] shows how to estimate this derivative. For an unsymmetrical aircraft.6 Change in Airplane Sideforce due to Change in Sideslip. a sideforce could be produced from the side that has a larger amount of surface area being hit by the free stream air. this derivative can be broken down into individual components. Ref [5]. and symmetry of aircraft. control surface deflection. Fuselage contribution depends on its shape and 17 . sideslip. Cy Eq [17] This is the aerodynamic force that causes the aircraft to yaw and can cause a rolling moment if above or below the center of gravity. the sideforce should equal zero for a symmetrical aircraft. 3. Cy Eq [18] Eq [19] Similar to the effect sideslip has on rolling moment. 3. the wings‟ contribution is generally negligible. This derivative is usually positive. A larger dihedral will produce a large sideforce. however. but at high angles of attack it can switch signs because of the vertical tail moment arm location changes. Eq [18] calculates the dimensional sideforce and Eq [19] shows the first order Taylor series. The sideforce is a result of sideslip.5 Sideforce Coefficient.2.2. Ref [5]. With zero angle of attack.
This happens by the increase in lift on one side and a decrease on the opposite side. This derivative can be estimated using methods found in Ref [7]. 3. The purpose of a rudder deflection is to create a sideforce that will produce a yawing moment. Large fuselages have more contact surface area to produce larger sideforces. Cyr The rudder has a large influence on the sideforce. If these rolling moment controls are close to a vertical surface the sideforce cannot be neglected. Ref [5]. can produce a sideforce. This sideforce depends on the size of the vertical tail in relation to the wings. These changes in lift are actually changes in pressure which. The lift curve slope of the vertical tail also plays into the influence of sideforce. 3. Wind tunnel tests have to be completed to measure this in a reliable fashion.size.2. Cya This contribution to the sideforce is very small and more often than not negligible. Depending on the location it will also produce a small rolling moment. if close to a vertical surface.7 Change in Airplane Sideforce due to Change in Aileron Deflection. The vertical tail can produce a large sideforce due to the large moment arm from the center of gravity and the size of the tail.2. The sideforce contribution of the rudder can be determined using Eq [20]. Eq [20] 18 .8 Change in Airplane Sideforce due to Change in Rudder Deflection.
sideslip.9 Yawing Moment. but it depends on the shape and the amount of projected side area forward or aft of the center of gravity. sideslip.10 Change in Yawing Moment due to Change in Sideslip. can play a large role. and location of the moment reference center. The vertical tail also has a significant contribution. Eq [23]. and control surface deflections. Ref [5]. For a symmetrical aircraft the yawing moment is equal to zero for zero angle of attack. speed. The same things that influence the rolling moment influence the yawing moment. control surface deflections. 3. Another impact on the fuselage contribution is the Munk effect. Eq [21] is the dimensional form of the yawing moment and Eq [22] is the first order Taylor series. Ref [5]. Eq [24] shows the contribution of the vertical tail. The fuselage. The derivative can be estimated by summing the components of the aircraft. Ref [5].2. on the other hand. Those influences are angle of attack. The wings‟ influence can be neglected since the flow is usually in line with the airfoil. Cn The aircraft yawing moment is the rotation about the zaxis in the stability coordinate system.2. Cn Eq [21] Eq [22] This derivative is referred to as the static directional stability and plays a large role in Dutch roll and spiral dynamics. 19 . which shifts the aerodynamic center forward.3. The size and location of the vertical tail determines the amount of contribution it has.
called an adverse yawing moment. Therefore. Also. 3. It can 20 .2. and a decrease in lift will cause a decrease in induced drag. The location of the vertical tail will determine the moment arm. upward deflection.11 Change in Yawing Moment due to Change in Aileron Deflection. and the lift decreases on the aileron with negative camber. Ref [5]. The lift curve slope of the vertical tail also plays a large role. Ref [5]. An increase in lift will cause an increase in induced drag. Eq [23] Eq [24] 3. the lift increases on the aileron with positive camber. Eq [25] is used to estimate the derivative. downward deflection.3 Perturbed State Eq [25] A perturbed state flight condition is defined as one for which all motion variables are defined relative to a known steady state flight condition. Cnr This derivative depends largely on the size of the vertical tail in relation to the wings. Cna With aileron deflections. Higher drag on one wing will cause the aircraft to yaw. the size of the rudder will influence the yawing moment. This type of yawing moment. 3.2.12 Change in Yawing Moment due to Change in Rudder Deflection. either pilot input or differential ailerons are used to prevent the aircraft from yawing. is undesirable because it tends to yaw the aircraft out of an intended turn.
It can easily be visualized how these changes in angle of attack will influence the lateraldirectional derivatives. and roll rate. Small roll rate perturbations cause nonsymmetrical changes in local angles of attack over the lifting surface. It is assumed that the pitch rate has a negligible influence on the lateraldirectional motion. Some of the lateraldirectional perturbation influences are the v component of a velocity change. For 21 . They also cause changes in the angle of attack of the vertical tail. v. yaw rate. There are also perturbation influences from changes in the flight angles. or acceleration in any direction or motion. example. Small yaw rate perturbations cause changes in the local velocity on the lifting surfaces. Pitch rate. q. u.be thought of as the aircraft‟s motion varying from the steady state condition. also affects the longitudinal derivatives. and w. these flight angles can be the sideslip angle or angle of attack. These variations are increases or decreases in velocity in any direction. The longitudinal motion is influenced by a change in velocity in the u and w direction if the velocity is broken down into three components.
by Dr.T. 22 . is a software program developed by DARcorporation that is an aircraft stability. If this help is not enough. It is widely accepted as the industry standard and is used in 45 countries. but can also be used to model production vehicles. the stability and control derivatives can be found and compared with actual flight test data. and design tool. the lower limit of the Reynolds number for calculations is 3e6. Although the program will model any size aircraft. allowing the designer to be more comfortable with the results AAA produces when designing an aircraft from scratch. Using AAA. C. AAA. If the designer does not understand how something is being determined. This helps eliminate the black box feeling that comes from most modeling software programs. and Airplane Aerodynamics and Performance. Parts I and II. AAA is used before the aircraft is ever built. there is a help button that shows all of the equations used for each variable. by Dr. The program follows the Dr. Lan and Dr. This program helps eliminate errors in calculations and spreadsheets. control. Geometric and flight characteristics of the aircraft being modeled are input. outputs are produced. The software works for preliminary design all the way up to Class II cost analysis. and with a simple click of „calculate‟. Doing this helps validate the software.4 AAA Advanced Aircraft Analysis. Roskam method of preliminary aircraft design. Jan Roskam. the books the program is based on are Airplane Design IVIII. Ref [8]. Jan Roskam. This means modeling of smaller aircraft may be inaccurate to some degree. Airplane Flight Dynamics and Automatic Flight Controls.
5
VORSTAB
VORSTAB is a software program that was developed by Dr. Edward Lan for
NASA. It has been used on several NASA research projects. The acronym name stands for Vortex Stability. Using vortex flow effect, the lateraldirectional and longitudinal stability derivatives are determined for an aircraft. The program follows the PrandtlGlauert equation, Eq [26], in subsonic flow, Ref [9]. The PrandtlGlauert equation is the linearized fullpotential equation, Eq [27], using small velocity perturbations assumptions, Ref [9]. Eq [26]
These are a few assumptions that VORSTAB uses for calculations.
Eq [27]
Assumes thin wing and therefore thickness effect is not accounted for in calculations. The boundary layer separation is not accounted for, and therefore the flow stays attached to the body. The wake aft of the wing is flat and does not increase in the zdirection. The designer of the input model also makes assumptions based on the aircraft that is being modeled and the flight characteristics. A detailed description of the program can be found in Ref [10].
23
5.1
Creating a Model VORSTAB runs an input file that the designer creates. There are 83 groups
that are used to describe the aircraft and the test cases that are to be run. These groups break down a step by step procedure for creating the model and allow the designer to trouble shoot more easily. Not every group is used to describe the aircraft that is being modeled, and many groups are repeated several times. descriptions of the groups are found in Ref [11]. To create a model of the aircraft the designer needs to have detailed schematics of the geometry. Typically, this geometry comes from a CAD model. Using the CAD model the designer can find all of the geometry needed for VORSTAB. The software does not require that a fuselage be present in the model. The first step in creating the model is to specify the number of lifting surfaces that are to be modeled. These lifting surfaces include wings, horizontal tail, vertical tail, vtail, and canards. The level flight geometry of each lifting surface is input into VORSTAB. If this is not done, the program assumes that the geometry is at zero angle of attack, when it might actually be more or less. For example, when the Meridian sits on the ground, the wings and tail are at a higher angle of attack when compared to level flight. The input lifting surface geometry required includes airfoil characteristics (thickness and camber), location in terms of distance from reference point (example: nose of aircraft), dihedral, twist, location of control surfaces, and much more. Also needed is the lift coefficient, drag coefficient, and pitching moment coefficient with respect to angle of attack. This step is repeated for each lifting Detailed
24
surface. Next, the fuselage data is input into VORSTAB. The geometry needed for this includes crosssection shape, camber, axial locations from a reference point, and much more. The more crosssections used in this model, the more accurate the results will be. It must be made sure that the lifting surfaces are not inside of the fuselage. The fuselage model is also created in level flight for the same reason as the lifting surfaces. Scaling the input file is very important. The entire aircraft is scaled to a size where the largest crosssectional radius of the fuselage is no larger than a measurement of 1 in any unit system. Next, the type of case that is to be run is described. This includes the flight condition, control surface deflection angles, angle of attack, and more. With the input file complete, VORSTAB can run. If there is an error in the input file, the output file will stop at the group with the problem. This is a great advantage to find simple problems in the input file that the designer overlooked. VORSTAB also creates a file that can be opened with TECplot to see a graphical representation of the model. This graphical representation allows the designer to visually determine if something looks wrong with the model created.
25
invisid. The geometry of the object and the flow field around the model are created. The following are the steps followed to run a FLUENT model.6 FLUENT FLUENT is a commercially sold computational fluid dynamics program that is widely accepted for its high fidelity. the designer chooses which flow field type is best. Super computers with multiple processors must be used to solve these larger models. GAMBIT is a software program that allows the designer to easily create a geometric model of the object being tested. and is more accurate than just the standard 2d or 3d. but the accuracy of the results is important to the validity of the model. 26 . The size of the mesh will dramatically affect the accuracy of the results and the processing time. Both 2demensional and 3demensional cases can be tested with laminar. Step 2: Export the mesh so that it can be imported into FLUENT. Depending on the model and flow type. There is also 2ddp and 3ddp. a standard PC with single memory cannot process the model. Step 3: With FLUENT opened. or turbulent flow. over 200. Case run time is larger for this method. The “dp” stands for double precision.000 cells and a coupled solver. depending on the mesh created in GAMBIT. select either 2D or 3D. The program has many different capabilities and functions. For large models. Step 1: Create a GAMBIT model with a mesh for the flow field.
solves the equations one by one. viscosity. The accuracy depends on the mesh shape and flow pattern. Step 5: Choose the type of solver: segregated or coupled. Coupled solver. The size of the mesh can also be seen to verify that there are not too many cells to run the test on a standard PC. Step 8: Select the solution controls and discretization methods. The „solver‟ is the method that the program uses to solve the equations. or turbulent. Coupled solver requires much more computer memory. This depends on the shape of the body and type of flow. velocity. and then check for convergence. A description of each solution and 27 . If choosing turbulent flow select the turbulent model to use. Step 7: Set the fluid properties and boundary conditions. turbulent factor. continuity.Step 4: Import the mesh and run a check to see if there are errors in the mesh. It is difficult to determine the most accurate turbulent model without testing several different ones. The fluid properties include but are not limited to density. and much more. Step 6: Choose the type of flow: laminar. wall friction.solves the same equations as the segregated solver. but does this simultaneously instead of one by one. invisid. Research has shown that the most accurate method for aerodynamics is SpalartAllmaras. Momentum. Segregated solver. energy (if compressible flow).
The FLUENT online help menu should be used for these questions. determined by setting the residuals. For example. The convergence criteria is 28 .discretization method should be read before the designer chooses which one is best for their model. PISO is used for transient flows. Step 9: Set the residuals and turn on what parameters are to be monitored. SIMPLE and SIMPLEC are good choices for noncomplicated flow problems (such as laminar flow). Step 10: Iterate until the solution converges. the lift and drag of the body can be monitored. These steps can slightly vary from mesh to mesh. but in general these ten steps allow the user to run a FLUENT model.
detailed measurements were taken to create an exact CAD model of the aircraft that was used during flight tests.7 YAK54 The YAK54 is a remote control aircraft that the Aerospace Engineering Department at the University of Kansas owns. Ref [12]. Figure 2 shows a model of the YAK54 used by the University of Kansas. Taking measurements directly from the aircraft allowed abnormalities due to manufacturing to modeled and 29 . This aircraft is a 1/3 scale version of the Russian Yakovlev Yak54 acrobatic airplane. Figure 2: YAK54 From the aircraft itself. An autopilot system was installed into this aircraft to help test the robustness and capabilities of the system before the same system was installed on the larger and more expensive Meridian.
25 1.45 ft c̅ = 0.accounted for.584 ft2 ft ft ft ft ft2 ft ft ft ft 30 . Unigraphics 6.67 0.844 cr = 0.0 was used.684 ct = Vertical Tail S= b= c̅ = = AR = cr = ct = 1.46 = 5. Table 1: YAK54 Lifting Surface Dimensions Wing 10.42 1.3 S= 3 b= 0.6 1. Ref [13]. Figure 3 shows the final CAD model of the YAK54.35 1.874 ft ct = Horizontal Tail 2.91 AR = 0. To create the CAD model.9 ft2 S= 7.9 ft b= 1.9 ft cr = 0.77 AR = 1.81 = 3.767 c̅ = 0. The flight conditions used in AAA can be found in Table 2. Table 1 shows some of the measurements taken from the YAK54.21 0.
Edmond Leong tuned the AAA results to develop more accurate stability and control derivatives. This will all be discussed later. q (lbs/ft2) C.5 The importance of creating a VORSTAB model of the YAK54 was to gain experience with the program and see how accurate the program was since this vehicle had been flight tested.106 118 16. The need for a VORSTAB model also came from AAA overestimating the drag on vehicles with low Reynolds numbers. After flight tests. fraction c̅ (in) 400 0. h (ft) Mach Number TAS.Figure 3: Unigraphics CAD Model of YAK54 Table 2: YAK54 Flight Conditions Flight Conditions Altitude (AGL). 31 .G.4 25. location. U1 (ft/sec) Dynamic Pressure.
0009 1.052 0.1 AAA Modeling of the YAK54 AAA was used to determine the YAK54‟s dimensionless longitudinal and lateraldirectional stability and control derivatives. Table 3: YAK54 AAA Moment of Inertia and Trimmed Values Mass Data Weight (lbs) IxxB (slugft2) IyyB (slugft2) IzzB (slugft2) IxzB (slugft2) Steady StateTrimmed CL1 CD1 CTx1 Cm1 CmT1 Angle of attack.04 0.053 0. Table 3 shows the trimmed condition values of the aircraft. 1 (deg) Elevator Deflection Angle. e1 (deg) 0.7.002 0. Table 6 shows the stability requirements for the YAK54 produced by AAA. and Table 5 shows the lateraldirectional derivatives. Table 4 shows the longitudinal derivatives. all results were taken from Ref [13].69 28 1.11 3.09 2.147 0.05 32 .79 0. Ref [13].
0630 CDu CD Ctxu CLu CL CLdot CLq Cmu Cm Cmdot Cmq CmTu CmT 0.1484 0.503 0.078 0.0974 4.9309 5.0015 0.0229 0.0000 0.0017 0.3724 4.0153 0.0286 Longitudinal Control and Hinge Moment DerivativesStability Axes (1/deg) (1/rad) Derivatives 0.469 8.0004 0.0065 0.8766 33 .0011 0.858 0.1509 0.0859 8.3782 0.0066 0.Table 4: YAK54 AAA Longitudinal Derivatives Longitudinal Coefficients and Stability DerivativesStability Axes (1/deg) (1/rad) Derivatives 0.1546 0.0899 0.0337 0.0000 0.0005 0.0792 0.0000 CDe CLe Cme 0.0000 0.538 1.
0057 0.0061 0.0060 0.0009 0.0974 34 .0067 0.0060 0.0017 0.0974 0.0041 0.1900 0.0000 0.0172 0.0020 0.3839 0.0020 0.0520 0.0003 0.Table 5: YAK54 AAA LateralDirectional Derivatives LateralDirectional Coefficients and Stability DerivativesStability Axes (1/deg) (1/rad) Derivatives 0.3495 Cla Clr Cya Cyr Cna Cnr 0.3438 0.1150 LateralDirectional Control and Hinge Moment DerivativesStability Axes (1/deg) (1/rad) Derivatives 0.0004 0.0229 Cl Clp Clr Cy Cyp Cyr Cn CnT Cnp Cnr 0.0172 0.0017 0.0003 0.0001 0.1146 0.2349 0.0034 0.0001 0.0000 0.
Table 7: YAK54 VORSTAB Longitudinal Coefficient and Pitch Rate (1/rad) CL CD CM CLq CDq CMq (deg) 5 0 5 10 15 20 0.000 0.000 10.1161 0.3813 0. pitching moment coefficients.Type of Stability Forward Speed Sideslip Vertical Speed Angle of Attack Angle of Sideslip Roll Rate Pitch Rate Yaw Rate Lateral Table 6: YAK54 AAA Stability Requirements Corresponding Criterion Derivative (deg1) Derivatives CTxu .84 8.7625 0.219 0.000 0.820 1.476 0.40 0.63 8.95 9.99 10.36 8.013 0.548 8.000 0.075 0. drag.3462 4.025 0.317 0.034 0.95 8.CDu <0 0.409 0.2 VORSTAB Modeling of the YAK54 A YAK54 input file was created for VORSTAB.376 0.43 35 . there is not a stall point for this model.14 1.021 0.438 0. Figure 4 through Figure 7 shows Table 7 in graphical form.6 10.205 0.1 8. The input file modeled the entire aircraft include all three lifting surfaces and the fuselage.1557 Cy CL Cm Cn Clp Cmq Cnr Cl <0 >0 <0 >0 <0 <0 <0 <0 0.0257 Stable/Unstable Stable Stable Stable Stable Stable Stable Stable Stable Stable 7.3817 10. Due to the assumption that the flow stays attached to the body.0954 0.142 0.040 0.2 9. Table 7 shows the lift.020 0.000 0. This input file can be found in Appendix A.22 7.000 0.63 0. and how those three coefficients are affected by pitch rate.
100 CD (N.020 0.) 0.600 1.160 0.140 0.120 0.600 5 10 15 20 (deg) Figure 4: YAK54 VORSTAB Lift Curve Slope 0.400 1.1.400 0.000 5 0 5 (deg) 10 15 20 Figure 5: YAK54 VORSTAB Drag Curve 36 .040 0.800 CL (N.080 0.D.200 0 0.000 5 0.D.200 0.) 0.600 0.200 1.000 0.060 0.400 0.
300 0.00 15.000 5 0.00 5 10 15 20 5 10.400 0.) 0.200 CM (N.D.00 (deg) CL_q CM_q Figure 7: YAK54 VORSTAB Longitudinal Derivatives 37 .600 (deg) Figure 6: YAK54 VORSTAB Pitching Moment Coefficient Curve 10.500 0.100 0 5 10 15 20 0.00 0 5.00 (1/rad) 0.0.00 5.
486 0.171 0.196 0.570 0.111 Cn_r 0.249 Cy_p 0. Table 8 shows the lateraldirectional derivatives due to sideslip. The drag is known to be lower than the AAA results due the fact that AAA says the aircraft must be trimmed at a higher throttle setting.250 0.021 0. The drag.001 Cl_p 0.082 0.213 0.The longitudinal derivatives produced by VORSTAB follow the expected trends.187 0. the pitching moment curve has a negative slope.437 0.198 0. approximately 45%.238 0.403 0.032 0.264 0.211 38 .023 0.090 Cy_r 0.229 0.251 0.506 0.011 0. and yaw rate. For example.262 0.045 0. Figure 8 through Figure 10 depict these lateraldirectional derivatives in graphical form. This is expected since AAA overestimates the drag at lower Reynolds number.363 Cl_ 0.016 0. roll rate. These values were also within the range that one would expect to see. This slope makes the aircraft naturally want to pitch back to the stability axes.086 0.078 0.154 Cn_p 0.233 0.065 0.013. but the coefficients follow the correct trend. approximately 27% throttle setting.531 0.105 0.220 0.108 0.0313 and VORSTAB estimates it to be 0.013 0.046 0.068 Cn_ 0. and from the flight tests it is trimmed at.531 0.249 0.064 0.018 0.218 0.070 0. this is a significant change.563 Cl_r 0.050 0.558 0. CDo.497 0. AAA approximation was 0. The magnitudes of these values might be off slightly off.010 0. Table 8: YAK54 VORSTAB LateralDirectional Derivatives (1/rad) (deg) 5 0 5 10 15 20 Cy_ 0.030 0.433 0. is lower than the AAA results predicted.
100 (1/rad) 0.4 0.2 0.0.000 5 0.250 (deg) Cy_p Cl_p Cn_p Figure 9: YAK54 VORSTAB LateralDirectional Derivatives due to Roll Rate 39 .4 0.050 0 5 10 15 20 0.5 0.150 0.3 0.1 0 (1/rad) 5 0.2 0.3 0.200 0.6 0 5 10 15 20 (deg) Cy_beta Cl_beta Cn_beta Figure 8: YAK54 VORSTAB LateralDirectional Derivatives due to Sideslip 0.1 0.
700 0. because the fuselage contribution outweighs the vertical tail contribution.200 0. All of the coefficients have the correct positive or negative value. Table 10 shows the longitudinal derivatives due to elevator deflection.100 0.300 (1/rad) 0. 40 . Again. This coefficient would always be positive when the aircraft has flow that is attached to the fuselage. Vertical tail contribution can be negative or positive.0. These control surface deflection derivatives can be seen in Figure 11 through Figure 19.100 0.300 0 5 10 15 20 (deg) Cy_r Cl_r Cn_r Figure 10: YAK54 VORSTAB LateralDirectional Derivatives due to Yaw Rate Similar to the longitudinal derivatives.400 0. the lateraldirectional derivatives follow an expected trend. and Table 11 shows the lateraldirectional derivatives due to rudder deflection. the magnitudes might be slightly off. but the in the expected range. except for rolling moment coefficient due to yaw rate at an angle of attack of 5˚.200 0.000 5 0.600 0.500 0. Both longitudinal and lateraldirectional derivatives change when the control surfaces are deflected. Table 9 shows the lateraldirectional derivatives due to aileron deflection.
0231 0.0189 0.0103 0.0221 0.000 0.Table 9: YAK54 VORSTAB LateralDirectional Due to Aileron Deflection (1/rad) Cy Cl Cn (deg) a 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 10 10 10 10 10 5 5 5 5 5 0 0 0 0 0 5 5 5 5 5 10 10 10 10 10 0.000 0.000 0.0209 0.0240 0.0027 0.0013 0.0172 0.0176 0.0221 0.0010 0.0033 0.0087 0.0121 0.0086 0.0213 0.0243 0.000 0.0099 0.0171 0.0070 0.000 0.0054 0.0026 0.000 0.0051 0.0178 0.0036 0.0121 0.000 0.0029 0.0244 0.0125 0.000 0.000 0.0020 0.0035 0.0019 0.0041 0.0120 0.000 0.0128 41 .0242 0.0040 0.0110 0.0121 0.0083 0.0111 0.0071 0.000 0.000 0.0061 0.0077 0.0054 0.0003 0.0093 0.0110 0.0219 0.0073 0.000 0.0193 0.0110 0.0224 0.0164 0.000 0.0056 0.0049 0.0022 0.0009 0.0044 0.000 0.
0100 0 5 10 0.0000 10 5 0.0.) 0.0000 10 5 0.0050 0.D.0300 aileron deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 12: YAK54 VORSTAB Rolling Moment Coefficient due to Aileron Deflection 42 .0200 0.D.0050 0 5 10 Cy (N.0150 0.0100 Cl (N.0100 0.0250 aileron deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 11: YAK54 VORSTAB Sideforce Coefficient due to Aileron Deflection 0.0300 0.0200 0.) 0.0200 0.
Therefore.D. VORSTAB shows a positive yawing moment at almost all aileron deflections.0050 0.0050 0. The negative yawing moment produced by the ailerons‟ deflection will yaw the aircraft out of an intended turn and the VORSTAB results say the opposite. but should have opposite sign convention.0000 10 5 0. the VORSTAB results are not to be assumed as wrong. VORSTAB asks for a sideslip angle to be put in the input 43 .0250 0.0.0200 0. but not always. The rolling moment controls are in close proximity to the fuselage and this can cause a rolling moment.0100 Cn (N. the yawing moment does not have symmetric results with respect to negative and positive aileron deflections.) 0. Also.0100 aileron deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha 0 5 10 Figure 13: YAK54 VORSTAB Yawing Moment Coefficient due to Aileron Deflection The aileron deflection is expected to have a very small effect on the sideforce coefficient. and this is seen in the results produced. A positive aileron deflection is defined as one that produces a positive rolling moment. Conventional ailerons will usually produce a negative yawing moment.0150 0. Opposite aileron deflections should result in the same magnitude.
0114 0.0211 0.4764 0.8202 1. these two 44 .147 1.4089 0.143 1.5432 0.0211 0. Table 10: YAK54 VORSTAB Longitudinal Due to Elevator Deflection (1/rad) CL CD CM (deg) e 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 12 12 12 12 12 6 6 6 6 6 0 0 0 0 0 6 6 6 6 6 12 12 12 12 12 0.0199 0.0143 0.5484 0.401 0.0341 0.5458 0.5511 0.0752 0.142 1.1420 0.1430 0.4404 0.5537 Therefore.1422 0.0212 0.4363 0.8227 1. asymmetrical results should be assumed wrong and ignored.1427 0.4394 0.0749 0.3123 0.4821 0.4374 0.4155 0.3221 0.2186 0.403 0.8191 1.0129 0.0121 0.0345 0. Without the source code it would be difficult to determine the source for the error.file and this might be the cause of the asymmetric results.4707 0.0747 0.0347 0.4736 0.405 0.0199 0.4039 0.3173 0.402 0.0136 0.4384 0.0199 0.2219 0.0343 0.3197 0.2202 0.3148 0.2153 0.0199 0.146 1.404 0.8181 1.8213 1.2169 0.4793 0.0212 0.0212 0.144 1.0757 0.0350 0.1425 0.4119 0.4064 0.0754 0.0199 0.
1200 0.6000 1.0000 CL (N.D.1000 CD (N.) 0.D.2000 1.1.4000 1.0000 10 5 0 elevator deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha 5 10 Figure 15: YAK54 VORSTAB Drag Coefficient due to Elevator Deflection 45 .6000 0.1400 0.8000 0.0200 0.0000 10 5 0 elevator deflection (deg) 5 10 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 14: YAK54 VORSTAB Lift Coefficient due to Elevator Deflection 0.4000 0.1600 0.0800 0.2000 0.) 0.0600 0.0400 0.
46 . the change in pitching moment seems small. This change in angle of attack will then change the lift and drag dramatically. but the change is large enough to change the aircraft‟s angle of attack.) 0.6000 elevator deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 16: YAK54 VORSTAB Pitching Moment Coefficient due to Elevator Deflection The lift and drag coefficients are not seen to change dramatically due to elevator deflection.4000 0.D.0. Also. but this deflection will change the angle of attack.1000 0 5 10 0.0000 10 5 0. The pitching moment due to elevator deflection should be negative and the VORSTAB results show this.2000 Cm (N.3000 0.5000 0.
Table 11: YAK54 VORSTAB LateralDirectional Due to Rudder Deflection (1/rad) Cy Cl Cn (deg) r 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 10 10 10 10 10 5 5 5 5 5 0 0 0 0 0 5 5 5 5 5 10 10 10 10 10 0.0597 0.0591 0.0586 0.0576 0.0570 0.0300 0.0297 0.0295 0.0290 0.0287 0.0000 0.0000 0.0000 0.0000 0.0000 0.0300 0.0297 0.0295 0.0290 0.0287 0.0597 0.0591 0.0586 0.0576 0.0570 0.0033 0.0033 0.0032 0.0030 0.0029 0.0016 0.0017 0.0016 0.0015 0.0015 0.0000 0.0000 0.0000 0.0000 0.0000 0.0016 0.0017 0.0016 0.0015 0.0015 0.0033 0.0033 0.0032 0.0030 0.0029 0.0310 0.0307 0.0305 0.0300 0.0297 0.0156 0.0155 0.0153 0.0151 0.0149 0.0000 0.0000 0.0000 0.0000 0.0000 0.0156 0.0155 0.0153 0.0151 0.0149 0.0310 0.0307 0.0305 0.0300 0.0297
47
0.0600
0.0400
0.0200
Cy (N.D.)
0.0000 5 0 5 10
10
0.0200
0.0400
0.0600 Rudder deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 17: YAK54 VORSTAB Sideforce Coefficient due to Rudder Deflection
0.0027 0.0017 0.0007
Cl (N.D.)
10
5
0.0003 0.0013 0.0023 0.0033
0
5
10
Rudder deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 18: YAK54 VORSTAB Rolling Moment Coefficient due to Rudder Deflection
48
0.0280
0.0180
0.0080
Cn (N.D.)
10
5
0.0020
0
5
10
0.0120
0.0220
0.0320 Rudder deflection (deg)
0º alpha 5º alpha 10º alpha 15º alpha 20º alpha
Figure 19: YAK54 VORSTAB Yawing Moment Coefficient due to Rudder Deflection
A positive rudder defection will produce a positive sideforce as expected, and viceversa. The rolling moment is a direct result of the sideforce, and therefore, a positive rudder deflection will also result in a positive rolling moment. A negative yawing moment is produced by a positive rudder deflection. The VORSTAB stability requirements are shown in Table 12. As seen from this table, the YAK54 is stable in all modes. The stability requirements were only found at the 0˚ angle of attack, and could be unstable at different positions. From reviewing the previous results from VORSTAB, it can be assumed that the aircraft is stable in all flight conditions.
49
since these two coefficients are not linear curves. Characterizing the drag and pitching moment due to angle of attack into a single number is more difficult than lift. since this was the most linear section of the curve. Cm. The drag derivative.8762 1.0304 Stable Stable Stable Stable Stable Stable Stable As seen from these results produced by VORSTAB. The same range of angles of attack was used for the other two derivatives. This derivative was found from 0˚ to 5˚ angles of attack. 50 . This is expected since it is a remote controlled aircraft.8762 rad1. CD.Table 12: YAK54 VORSTAB Stability Requirements Corresponding Derivative Type of Stability Criterion Stable/Unstable Derivatives (rad1) <0 0.0144 rad1 and the pitching moment derivative.090 0. The lift curve slope of the results was found to be 4.4968 Stable Sideslip Cy Vertical Speed Angle of Attack Angle of Sideslip Roll Rate Pitch Rate Yaw Rate Lateral CL Cm Cn Clp Cmq Cnr Cl >0 <0 >0 <0 <0 <0 <0 4.1314 rad1. is 0. The results from VORSTAB and AAA will be discussed and compared in more detail in the following section.1314 0. the YAK54 is a stable aircraft with a large amount of control. is 1.2200 0.2289 0. All other derivatives were determined at 0˚ angle of attack. The aircraft is known to not stall at these angles also.2178 10.
The tuned AAA results were derived after several flight tests were conducted. Table 13 shows the comparison data for the YAK54 for AAA and VORSTAB. These turned values were taken from Ref [14] and explanation for deriving them can be found there. As stated previously. 51 . the VORSTAB results were found over an angle of attack range of 0˚ to 5˚. Using a computational fluid dynamic method such as VORSTAB should produce results closer to that of the actual data than results from DATCOM and AAA.3 Method Comparison YAK54 Both AAA and VORSTAB are considered as an accurate approximation for the control derivatives. and the system identification warranted an improved model.7.
0455 0.0229 0.0000 1.87 34.0379 0. The differences that are in the 100 of percents do not fall within the error range though.1509 8.45 83.27 37.5858 0.0387 0.31 20.09 27.0231 0.1146 0.00 98.4865 0.538 0.0859 0.52 334.0000 0.88 44.2228 0.2289 0.0188 0.1064 0.50 858.3724 5.33 107.0194 0.13 40.16 54.75 43.3707 0.0859 0.3441 0.1314 8.22 109.5380 0.37 83.0520 0.1150 0.022 0.53 83.3724 5.1388 0.0172 0.0304 0.20 60.64 156.71 62.2200 0.0158 0.12 0.8762 0.8766 0.0974 0.2289 0.1052 0.24 203.0323 0.66 32.1789 % Difference AAA 7.82 56.3839 0.0060 0.2178 0.3782 0.35 238.30 76.44 834.80 98.0088 0.Table 13: YAK54 AAA and VORSTAB Stability and Control Comparison Stability and Control Derivatives Derivative (1/rad) CL CD Cm CLq Cmq Cl Clp Clr Cy Cyp Cyr Cn Cnp Cnr CLe CDe Cme Cla Clr Cya Cyr Cna Cnr AAA 4. The tuned AAA results improved 52 .54 92.0000 0.3438 0.0646 0.2707 0.18 62.0743 0.2349 0.3495 0.64 18.090 0.56 14.1948 0.04 98.2531 0.42 Several of the differences between the AAA and VORSTAB results fall within the error ranges given by Dr.0000 0.8411 10.68 % Difference AAA Tuned 7.0071 0.11 125.30 99. Roskam.0144 1.64 37.0172 0.24 203.81 71.0219 0.45 83.0000 0.29 9.4968 0.81 71.1404 VORSTAB 4.12 0.289 0.0974 AAA Tuned 4.1509 16.5030 0.3782 0.00 98.
there are typical ranges that can be expected. Roskam. For stability and control derivatives. Both methods show that the aircraft is stable. [1]. Jan Roskam developed ranges for the control derivatives.in some cases and not in others. These ranges are for conventional aircraft. More flight tests need to be conducted and parameter system identification need to be conducted to improve all of the derivatives. Cmq. Cnr. Ref. and the ranges are a function of Mach number. Cl. Table 14 shows the Dr. 53 . These derivatives. and Cme were tuned from the flight test and the others were taken from the AVL model. and through his vast experience. The tuned yawing moment derivatives show an improvement in the VORSTAB results over AAA. Dr. Cnr. typical ranges compared to the AAA and VORSTAB results.
0 to 0.0 4.0158 0.0 to 0.1509 8. Roskam developed except for the yawing moment coefficient due to rudder deflection.2349 0.0000 0.8 0.1 to 4.4968 0.1146 0.3441 0.0 0.0455 0.3724 5.0 to 1.0 0.0 0.0231 0.0 to 8.1150 0.8766 0.2289 0.0 to 1.1 to 0.0229 0.0 0.8762 0.1 to 2. Roskam 1.0 to 4.2200 0.04 to 0.6 0.0 Within Range (Yes/No) VORSTAB AAA Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes The YAK54 VORSTAB results fall within all of the typical ranges that Dr.08 0.1314 8.4865 0.0172 0.5 0.0646 0.1 0.0 0.15 to 0.0 0.3495 0. Cnr.08 to 0.0071 0.0974 0.0 to 4.0 90.0172 0.0974 Typical Range Dr.1948 0.0379 0.0 0.5 to 0.8411 10.0 to 30.04 Negligible 0.0 0.0060 0.0000 0.0000 0.0 to 0.1789 AAA 4.0144 1.2 0.0188 0. These ranges are just estimations and many aircraft‟s derivatives do 54 .3438 0.0 to 0.5030 0.0 to 1.0323 0.0859 0.0304 0.2178 0.3782 0.3839 0.0 to 2.5380 0.3 to 0.Table 14: YAK54 Stability and Control Derivatives Typical Ranges Stability and Control Derivatives Derivative (1/rad) CL CD Cm CLq Cmq Cl Clp Clr Cy Cyp Cyr Cn Cnp Cnr CLe CDe Cme Cla Clr Cya Cyr Cna Cnr VORSTAB 4.4 0.6 Negligible 0.1388 0.8 0.0520 0.09 0.0 to 0.
7. and this causes the coefficient to be a nonnegligible value. Also. it is concluded that both programs obtained valid results. The ailerons. are in close proximity to the fuselage. on the YAK54. A model should be created with the new flight test data and then compared to the AAA and VORSTAB results.not fall into the ranges. This can happen when the rolling moment controls are in close proximity to a vertical surface (fuselage or vertical tail). is not negligible. The ranges just give the designer an idea of things to expect. From these results. the sideforce moment coefficient due to aileron deflection. The AAA results all fall within the typical ranges. These results should be compared to the more flight test data to validate both software programs. Cya. 55 .4 Linearized Model of the YAK54 For the information and data on the state space model please refer to Appendix B.
This aircraft was a prototype and had its problems. With post flight analysis. Soon after takeoff. Roskam method using AAA . After much was research conducted by the team. Class II design resulted in an empty weight of 32 lbs for the vehicle with a design cruise speed of 70 kts. 56 . a wingbody configuration was chosen as the most suitable configuration for the particular mission. The design of the MantaHawk followed the Dr. Katrina Legursky. The design team consisted of Emily Arnold. the team determined that a negative pitching moment was much higher than predicted and without preflight trim there was not enough control to fly. MantaHawk‟s size and weight were well suited for a seaborn launch with a much shorter range than the other Meridian Antarctic mission.and AVL for some of the stability and control derivatives that AAA did not produce. Figure 20 shows the MantaHawk designed by the AE 721 class. and this caused the airplane to go into a nose dive. Ref [15]. Rick Riley. by the AE 721 graduate design course in the fall 2009. The design and flight test of the MantaHawk can be found in Ref [15]. Robert Burns. the aircraft pitched downward aggressively. The vehicle had a payload capacity of 15 lbs plus 7 lbs of fuel. The goal of the team was to help improve the current research going on at the University of Kansas and within CReSIS itself. This small UAV was developed in conjunction with the CReSIS Meridian project. Dustin Grorud. and Jonathan Tom. Dave Royer.8 MantaHawk The MantaHawk is a remote control aircraft designed at the University of Kansas.
but this vehicle was never a proven platform. the MantaHawk was designed at the University of Kansas using AAA software. Table 15 and Table 16 shows the stability and control derivatives produced by AAA. Ref [15].Figure 20: MantaHawk 8. The MantaHawk‟s AAA model estimated the stability and control derivatives. Ref [15].1 AAA Modeling of the MantaHawk Unlike the YAK54. Table 17 shows the AAA stability requirements for the MantaHawk. 57 .
0000 0.0920 0.0615 0.0000 0.0000 0.Table 15: MantaHawk AAA Longitudinal Derivatives Longitudinal Coefficients and Stability DerivativesStability Axes (1/rad) Derivatives 0.0000 CDu CD CDq CTxu CLu CL CLq Cmu Cm Cmq CmTu CmT CDel CLel Cmel 0.2050 0.0630 58 .9630 5.4034 4.0021 3.4209 0.7340 0.0000 0.4784 1.
0007 0.Table 16: MantaHawk AAA LateralDirectional Derivatives LateralDirectional Coefficients and Stability DerivativesStability Axes (1/rad) Derivatives 0.3588 0.0219 0.7340 0.4892 4.9630 0.0359 0.0407 0.1877 0.0817 Stable Stable Stable Stable Stable Stable Stable According to AAA. the MantaHawk is stable in all modes and should not have a problem in flight.0817 Cl Clp Clr Cy Cyp Cyr Cn CnT Cnp Cnr Cyel Clel Cnel 0.0055 Table 17: MantaHawk AAA Stability Requirements Corresponding Derivative Type of Stability Criterion Stable/Unstable Derivatives (rad1) <0 0.0012 0.4892 0.1249 0.4034 0.0359 0.0897 0.1877 Stable Sideslip Cy Vertical Speed Angle of Attack Angle of Sideslip Roll Rate Pitch Rate Yaw Rate Lateral CL Cm Cn Clp Cmq Cnr Cl >0 <0 >0 <0 <0 <0 <0 3.0219 0.0169 0. These results gave the flight test team confidence going into the 59 .
and an asymmetrical deflection imitates a rudder or aileron input.0000 0.0006 0.first flight.0217 0.6123 0.0000 0.2603 0.0000 1.6709 0.9365 1.6052 1. Table 18 shows the longitudinal derivatives for the MantaHawk. There were also large wingtips that were used to help prevent wingtip vortices and wingtip stall. A symmetrical deflection of the control surface.0068 0.6138 0. Table 18: MantaHawk VORSTAB Longitudinal Derivatives (1/rad) CL CD Cm CLq CDq Cmq (deg) 5 0 5 10 15 20 0. 8.0167 0.0256 1. The wreck raised concern and prompted a VORSTAB model to be produced to determine if the vehicle was actually stable in all modes.4615 60 .0092 0.0154 0. Figure 21 through Figure 25 depict Table 18 in graphical form.4760 0.9645 1.0377 0.5442 1.0366 0. Since it was a flying wing. the aircraft nosed into the ground.0000 0. imitates an elevator input.2 VORSTAB Modeling of the MantaHawk The MantaHawk was a flying wing that used three different airfoil shapes throughout the wing. Shortly after liftoff.0763 0.0159 0.9478 1. but rather just one lifting surface and one control surface.0119 0.6498 1.8694 1.6625 1.1683 0.0000 0. the VORSTAB model did not include a fuselage. The VORSTAB input file for the MantaHawk can be found in Appendix A.6498 1.0271 0.7585 1.0000 0. elevon.
5 0.D.09 0.1 0 0.08 0.03 0.1 0 5 0.4 CL (N.6 0.0.7 0.2 0.02 0.) 0.07 0.2 0.06 CD (N.D.05 0.01 0 5 0 5 (deg) 10 15 20 Figure 22: MantaHawk VORSTAB Drag Coefficient 61 .) 0.8 0.04 0.3 0.3 5 10 15 20 (deg) Figure 21: MantaHawk VORSTAB Lift Coefficient 0.
8 1.75 1.) 0.6 1.0.D.85 CL_q (1/rad) 1.95 1.005 0.03 0.015 CM (N.015 0.7 1.55 1.02 0.5 5 0 5 (deg) 10 15 20 Figure 24: MantaHawk VORSTAB Lift Coefficient due to Pitch Rate 62 .01 0.01 0.02 (deg) 0 5 10 15 20 5 Figure 23: MantaHawk VORSTAB Pitching Moment Coefficient 2 1.025 0.005 0 0.9 1.65 1.
6 1. The other longitudinal derivatives were as expected.1.022.55 1. but the drag is very low for this aircraft. This is also a very significant difference and should be investigated. estimated by VORSTAB was 0.7 (deg) Figure 25: MantaHawk VORSTAB Pitching Moment Coefficient due to Pitch Rate As seen from the pitching moment curve. This means that the aircraft is naturally unstable. This is why the VORSTAB drag results are lower than what flight test data should show.5 0 5 10 15 20 CM_q (1/rad) 1. it will not pitch back downward. there is a positive slope. because it is proportional to the square 63 . Pitching moment coefficient due to pitch rate is typically large. CDo. the drag is most likely between these two software programs. The drag. When the aircraft is pitched upward.0068 and AAA estimated it to be 0.45 5 1. In reality. VORSTAB only determines the pressure drag and not the viscous drag. With this information the flight test team would have exercised more caution during the testing of the vehicle.65 1. The overall lift coefficient for the aircraft is rather small. but rather continue pitching upward.
061 0.008 0.018 0.098 0.053 0.117 0. The MantaHawk is a flying wing.056 0.034 0.048 0.0000 (1/rad) 5 0.011 0.078 0.0200 0.020 0. Table 19: MantaHawk VORSTAB LateralDirectional Derivatives (1/rad) (deg) 5 0 5 10 15 20 Cy 0.006 0.016 0.283 0.1400 (deg) Cy_beta Cl_beta Cn_beta Figure 26: MantaHawk VORSTAB LateralDirectional Derivatives due to Sideslip 64 .082 0.002 0.0600 0.015 Cyp 0.0200 0 0.0800 0.060 Cl 0.107 0.0996 0.of the moment arm of the horizontal tail.287 0.072 0.057 0.0400 0.090 0.033 0.084 0.071 0.055 0.004 0.036 0. so the moment arm is not as large as a conventional aircraft and this is why the derivative is not very large.1200 0.083 Cnp 0.054 Cnr 0.056 Clr 0.021 0.065 Cyr 0.082 0.054 0.0003 0.009 0.086 0.011 0.004 Clp 0.165 0. Figure 26 through Figure 28 depicts these derivatives in graphical form.033 0.020 0.016 0.047 0.1000 0.003 0.091 0.261 0.061 0. Table 19 shows the lateraldirectional derivatives.0400 0.0800 5 10 15 20 0.064 Cn 0.0600 0.287 0.
0.1 0.05 0 5
(1/rad)
0.05 0.1 0.15 0.2 0.25 0.3 0.35
0
5
10
15
20
(deg)
Cy_p Cl_p Cn_p
Figure 27: MantaHawk VORSTAB LateralDirectional Derivatives due to Roll Rate
0.1
0.08
0.06
(1/rad)
0.04
0.02
0 5 0.02 (deg)
Cy_r Cl_r Cn_r
0
5
10
15
20
Figure 28: MantaHawk VORSTAB LateralDirectional Derivatives due to Yaw Rate
65
At high angles of attack, aircrafts do not always follow the expected trends for the derivatives. This is observed on several of these derivatives. The sideforce coefficient due to sideslip is typically negative due to the defined positive angle of sideslip. VORSTAB estimates that at high angles of attack this derivative becomes positive. The yawing moment due to sideslip should be positive for the same reasons as sideforce should be negative. This derivative becomes negative at high angles of attack. Due to the definition of sideforce coefficient due to roll rate it would be assumed that the derivative should be negative, but at high angles of attack the moment arm will change signs. This will result in the derivative changing signs to positive. It can be expected for the sign to change for the yawing moment derivative due to roll rate as the angle of attack changes from positive to negative. The effect of this derivative is rather small on the airplane dynamic stability, and since this is the case the sign convention does not matter. Yawing moment coefficient due to yaw rate changes signs due to the fact that as the angle of attack increase the moment arm goes from positive to negative. All other derivatives were as expected. With the control surfaces deflected, the longitudinal and lateraldirectional derivatives change. Table 20 shows the effects of an elevator input sent to the elevons on the longitudinal derivatives. This type of input results in a symmetrical deflection of the elevons. Table 21 shows the effects of the aileron or rudder input sent to the elevons on the lateral directional derivatives. asymmetrical deflection of the elevons. There is a resulting
An asymmetrical deflection imitates an
66
aileron and a rudder input both. All values are given in radians. Graphically, these tables are shown in Figure 29 through Figure 34.
Table 20: MantaHawk VORSTAB Longitudinal Derivatives due to Symmetrical Deflection CL CD Cm (deg) el (deg) 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 12 12 12 12 12 6 6 6 6 6 0 0 0 0 0 6 6 6 6 6 12 12 12 12 12 0.0896 0.1251 0.3413 0.5118 0.6034 0.0250 0.1930 0.4132 0.5705 0.6432 0.0377 0.2603 0.4760 0.6123 0.6709 0.1079 0.3354 0.5317 0.6433 0.6871 0.1840 0.3969 0.5768 0.6643 0.7001 0.0125 0.0107 0.0127 0.0252 0.0534 0.0088 0.0091 0.0126 0.0296 0.0622 0.0068 0.0092 0.0154 0.0366 0.0763 0.0085 0.0109 0.0215 0.0484 0.0939 0.0126 0.0158 0.0299 0.0645 0.1117 0.0561 0.0827 0.0888 0.0806 0.0602 0.0267 0.0477 0.0529 0.0485 0.0407 0.0006 0.0159 0.0217 0.0271 0.0256 0.0298 0.0201 0.0025 0.0120 0.0154 0.0687 0.0476 0.0210 0.0004 0.0086
67
8 0.6 0.02 0 12 10 8 6 4 2 0 2 4 6 8 10 12 elevon deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 30: MantaHawk VORSTAB Drag Coefficient due to Symmetrical Deflection 68 .06 0.D.) 0.12 0.) 0.2 elevon deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha 0 2 4 6 8 10 12 Figure 29: MantaHawk VORSTAB Lift Coefficient due to Symmetrical Deflection 0.1 0.1 0.04 CD (N.D.2 0.08 0.0.1 0 12 10 8 6 4 2 0.5 0.3 0.4 CL (N.7 0.
04 0.D.06 0.08 elevon deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha 12 2 4 6 8 10 12 Figure 31: MantaHawk VORSTAB Pitching Moment Coefficient due to Symmetrical Deflection These three longitudinal derivatives due to the elevator input have a value and sign convention that is expected.06 0. A large symmetrical elevon deflection would increase the magnitude of the pitching moment and as a result the lift and drag would also increase. 69 .) 0 10 8 6 4 2 0 0.08 0.0.1 0.02 Cm (N.04 0. These trends are observed in the VORSTAB results.02 0.
0000 0.0001 0.0035 0.0000 0.0003 0.0019 0.0149 0.0304 0.0000 0.0005 0.0005 0.0059 0.0011 0.0151 0.0033 0.0304 0.0008 70 .0008 0.0000 0.0008 0.0000 0.0011 0.0001 0.0007 0.0152 0.0000 0.0149 0.0009 0.0010 0.0050 0.0015 0.0023 0.0017 0.0000 0.0000 0.0301 0.0000 0.0304 0.0279 0.0008 0.0230 0.0152 0.0003 0.0043 0.0137 0.0129 0.0020 0.0021 0.0298 0.0010 0.0296 0.0000 0.0010 0.0000 0.0004 0.0000 0.0032 0.0024 0.0287 0.0029 0.0010 0.Table 21: MantaHawk VORSTAB LateralDirectional Derivatives due to Asymmetrical Deflections Cy Cl Cn (deg) el (deg) 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 10 10 10 10 10 5 5 5 5 5 0 0 0 0 0 5 5 5 5 5 10 10 10 10 10 0.0062 0.0010 0.0007 0.0003 0.0000 0.0006 0.0152 0.0273 0.0144 0.0021 0.0144 0.0000 0.0000 0.0004 0.
006 0.04 Elevon deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha 10 0 5 10 Figure 33: MantaHawk VORSTAB Rolling Moment Coefficient due to Asymmetrical Deflection 71 .01 0.004 0.D.01 Cl (N.008 0.002 0 5 10 10 0.0.03 0.008 Elevon deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 32: MantaHawk VORSTAB Sideforce Coefficient due to Asymmetrical Deflection 0.006 0.02 0.) 0 5 0.04 0.D.) 0 5 0.004 0.02 0.03 0.002 Cy (N.
unless the aircraft is designed to fly at these angles.003 0.001 0 10 5 0.) 0. but at higher angles it becomes positive.D.0. This is most likely due to the change in drag over the airfoils at extreme angles of attack. At high angles of attack this is not seen in the VORSTAB results.004 0.001 0 5 10 Cn (N.005 Elevon deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 34: MantaHawk VORSTAB Yawing Moment Coefficient due to Asymmetrical Deflection A positive aileron/rudder input should result in a positive sideforce. The MantaHawk was not designed to fly at these angles though. At low angles of attack the yawing moment is negative. a positive rudder input will result in a negative yawing. and a positive aileron input will result in a negative yawing moment but can be positive. a positive asymmetrical elevon deflection will result in a positive rolling moment. Typically. At extreme angles of attack the forces and moments are usually not what is expected or desired. as expected. The result of this positive 72 .002 0.002 0. Similar to the sideforce.
73 .5500 0.3 Method Comparison MantaHawk The results from AAA and VORSTAB are different in the terms that Cm is unstable as predicted by VORSTAB. 8. it was predicted that the aircraft was stable in all modes and VORSTAB predicted an unstable mode.0458 Stable Unstable Stable Stable Stable Stable Stable Like the YAK54. the aircraft may become unstable in modes other than angle of attack stability. the MantaHawk‟s stability requirements were determined at 0˚ angle of attack. At other angles of attack. AAA‟s results say that the MantaHawk is stable in all modes.1146 Stable Sideslip Cy Vertical Speed Angle of Attack Angle of Sideslip Roll Rate Pitch Rate Yaw Rate Lateral CL Cm Cn Clp Cmq Cnr Cl >0 <0 >0 <0 <0 <0 <0 2.0172 0.6625 0. Table 22: MantaHawk VORSTAB Stability Requirements Corresponding Derivative Type of Stability Criterion Stable/Unstable Derivatives (rad1) <0 0.0077 0.value will actually help the intended turn.1756 0. Table 22 shows the stability requirements for the MantaHawk. As seen from the AAA data. A comparison of the two methods can be seen in Table 23.2866 1. The vertical speed and angle of attack stability requirements were both found over the angle of attack range of 0˚ to 5˚ from the VORSTAB results.
00 69.0169 0.0363 0.0003 0.38 52.5500 0.09 VORSTAB produces results that are lower than most of the AAA derivatives.1756 1.94 41.9645 1.Table 23: MantaHawk AAA and VORSTAB Comparison Stability and Control Derivatives Derivative (1/rad) CL CD Cm CLq Cmq Cl Clp Clr Cy Cyp Cyr Cn Cnp Cnr CDel CLel Cmel Clel Cyel Cnel AAA 3.0199 0.42 64.0920 4.84 96.0077 0.0359 0.7340 0.0275 0.9630 0.0615 0. Most of the differences between AAA and VORSTAB fall within the error ranges that Dr.95 56. Roskam describes that AAA will have.1745 0.41 51.0055 VORSTAB 2.22 64.09 98.1877 0.0630 0.65 55.53 61. since this was 74 .6703 0.37 2925.0219 0.85 47.88 43.0539 0.11 72.4034 5.3588 0. Pitching moment coefficient due to angle of attack has the largest variance and this is the most critical derivative.0172 0.0458 0.0817 0.0012 0.0472 0.0168 0.0407 0.1146 0.01 40.1249 0.4892 0.0093 % Difference 35.64 51.6625 0.28 143.11 38.2908 0.0897 0. This means that the aircraft is not going to have as much control as expected.4784 1.4209 0.2866 0.
and the ranges are a function of Mach number. These ranges are for conventional aircrafts. Roskam. These ranges again were developed by Dr. Ref [5]. for what approximately an aircraft can expect the control derivatives to be. The same typical ranges that applied to the YAK54 apply to the MantaHawk as well. Ref [5].the unstable mode. 75 . Roskam. Table 24 shows the VORSTAB MantaHawk results compared to typical ranges from Dr. VORSTAB should be used as a method for checking if there is a possibility for one of the modes to be unstable. Even if VORSTAB is incorrect this will give designers a chance to plan the flight test with this possibility.
8 0.092 4.0 to 1.0 to 2.4892 0. These are estimated ranges and an unconventional aircraft will not necessarily fall in the range.0615 0.0897 0.0 to 1.1 0.08 to 0.0 to 30.15 to 0.2 0.08 0.1877 0.734 0.4034 5.0 to 0.6703 0.0 0.0 0.1249 0.1745 0.0003 0.04 to 0. the rudder 76 .0363 0.8 0.0 0.2908 0.0093 AAA 3.6 Negligible 0.9645 1.0055 0.0472 0.0 0.0219 0.0012 0.1146 0.0 to 4.1745 0.0275 0.6 0.0172 0.0077 0.0 to 0.0407 0.4209 1.963 0.0168 0.3588 0. Also.0055 Typical Ranges Dr. Roskam 1.0359 0.0 to 1.0539 0.0 to 0.1 to 4.0093 0.0 0.0012 0. This does not necessarily mean that the aircraft is unstable.4784 0.Table 24: MantaHawk Stability and Control Derivatives Typical Ranges Stability and Control Derivatives Derivative (1/rad) CL CD Cm CLq Cmq Cl Clp Clr Cy Cyp Cyr Cn Cnp Cnr CLe CDe Cme Cla Clr Cya Cyr Cna Cnr VORSTAB 2.0817 0.3 to 0.0199 0.1 to 0.4 0.0 to 4.0 to 8.1756 1.1 to 2.0169 0.2866 0.55 0.0 4.3588 0.5 0.0 to 0.0 90.0 Within Range (Yes/No) VORSTAB AAA Yes Yes Yes No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No No Yes Yes No No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes Yes No Yes Yes Yes Not all of the control derivatives fall within the expected ranges.0 to 0.0363 0.04 Negligible 0.5 to 0.0630 0.0 0.6625 0.0458 0.0 0.0 0.
8. the rolling moment coefficient due to aileron deflection falls into the typical range Dr. and this is observed since the pitching moment coefficient due to change in angle of attack. Therefore. does not fall into the typical range Dr. Roskam gives. The drag coefficient due to elevator deflection is usually ignored. there is no cause for alarm that there is an increase in drag when the control surfaces are deflected. but it does not fall into the typical range for the rudder deflection. The input simulates both a rudder input and an aileron input. but not the other deflection. the MantaHawk is unstable in the pitching mode. C m. For example. 77 . This means that the control derivatives might fall into the range for the rudder or aileron deflection. Roskam describes.deflection and aileron deflection control derivatives are the same because there is only one input for those control surfaces. As stated earlier.4 Linearized Model of the MantaHawk For the information and data on the state space model please refer to Appendix B.
As such. since most civilian UAV research currently limits payloads to a few pounds and a fraction of the power. Information about the design on the Meridian can also be found in Ref [17]. and a range of 950 nautical miles. This data was taken from Ref [16]. The KUAE. The foundation has the mission of developing new technologies and computer models to measure and predict the response of sea level change to the mass balance of ice sheets in Greenland and Antarctica.100 lbs. and available payload exceeds 120 lbs and 300W consumed power. Added benefits of greater than two cubic feet of payload volume and eight wing hard points provide a sensor platform with little competition.9 Meridian UAV The Center for Remote Sensing of Ice Sheets (CReSIS) is a science and technology center established by the National Science Foundation (NSF) in 2005. endurance exceeds 9 hours. in close collaboration with national and international partners. Table 25 shows some of the important salient characteristics and Figure 35 shows the design of the Meridian UAV both from Ref [17]. a wingspan of 26. Vehicle cruise speed is 100120 kts.4 ft. 78 . called the Meridian. this is an extraordinarily valuable scientific research platform. has designed and developed the Meridian UAV for the Center for Remote Sensing of Ice Sheets (CReSIS). The Meridian has a gross takeoff weight of 1. CReSIS is developing sophisticated sensors and a longduration uninhabitant aerial vehicle (UAV).
9 ft 3.4 ft 31.599 Fuselage 14.Table 25: Meridian Characteristics Characteristic Wing Area Span MGC Aspect Ratio Quarter Chord Sweep Angle Taper Ratio Thickness Ratio Dihedral Angle Incidence Angle Aileron Chord Ratio Aileron Span Ratio Flap Chord Ratio Flap Span Ratio Characteristic Maximum Length Maximum Height Maximum Width 69.00 0.6 ft 26.100.24 0.5 26.5 12% 50 deg 0 deg Figure 35: Unigraphics CAD Model of the Meridian UAV 79 .75 ft 3.7 ft 10 0 deg 1 18% 5 deg 0 deg 0.2 0.8 ft 2.5 2 VTail 10 ft2 5.3 deg 0.9 ft 1.601.
80 . The tail of the Meridian is unconventional and uses a vtail. As seen in Table 28 the aircraft is found to be stable in all modes. 9. Roskam‟s method. the AAA file produced stability and control derivatives for the Meridian UAV. It is also a fully moving tail.1 AAA Modeling of the Meridian Similar to the YAK54 and MantaHawk. The longitudinal derivatives are found in Table 26 and the lateraldirectional derivatives are found in Table 27.The design of the Meridian was completed using AAA and Dr. both Ref [18]. The AAA software was also used to find the stability and control derivatives of the Meridian.
1648 0.3419 Longitudinal Control and Hinge Moment DerivativesStability Axes (1/rad) Derivatives CDre CLre Cmre 0.1409 0.7407 4.973 0.6179 0.6709 81 .0117 0.0000 0.0109 5.6207 2.0294 0.0028 0.4149 1.9826 13.0713 0.Table 26: Meridian AAA Longitudinal Derivatives Longitudinal Coefficients and Stability DerivativesStability Axes (1/rad) Derivatives CDu CD CTxu CLu CL CLdot CLq Cmu Cm Cmdot Cmq CmTu CmT 0.
1099 0.0007 0.1481 82 .0776 0.3681 0.4789 0.1465 0.2316 0.0000 0.3217 0.0351 0.1338 LateralDirectional Control and Hinge Moment DerivativesStability Axes (1/rad) Derivatives Cla Clre Cya Cyre Cna Cn 0.Table 27: Meridian AAA LateralDirectional Derivatives LateralDirectional Coefficients and Stability DerivativesStability Axes (1/rad) Derivatives Cl Clp Clr Cy Cyp Cyr Cn CnT Cnp Cnr 0.0253 0.5546 0.1386 0.0134 0.
1 Model 1: Half Scale Model of Meridian The first model created was of the exact shape of the entire Meridian. The reasons for these changes were due to the fact that VORSTAB calculates such a high downwash as a result of the shape and camber of the fuselage.4789 Stable Sideslip Cy Vertical Speed Angle of Attack Angle of Sideslip Roll Rate Pitch Rate Yaw Rate Lateral CL Cm Cn Clp Cmq Cnr Cl >0 <0 >0 <0 <0 <0 <0 5.1338 0.1386 0. From this downwash. The downwash on the Meridian fuselage is due to the large amount of convergence from the nose to the tail and the negative camber or downward swoop of the empennage. and these changes to the model were to improve the results.6207 0.973 0. Typically.1648 0. 9.Table 28: Meridian AAA Stability Requirements Corresponding Derivative Type of Stability Criterion Stable/Unstable Derivatives (rad1) <0 0. but scaled down to half of the size.2. the fuselage does not result in a large amount of downwash and can almost be completely ignored. This is a requirement for VORSTAB that the largest 83 . the effectiveness of the tails is almost blanketed out. there is not as much downwash as VORSTAB is predicting due to the fact that the flow will separate from the fuselage.2 VORSTAB Modeling of the Meridian Several different Meridian VORSTAB models were created with changes to the geometry of the fuselage.0776 Stable Stable Stable Stable Stable Stable Stable 9. In reality.5546 13.
the downwash 84 .203 2.677 10 12.735 6.041 0.6188 0.072 2.2483 1.095 0 0.925 2.863 3.3823 0. but in reality not all of it will.118 15. These results were known to be invalid.656 2.882 2.671 2.9534 1.515 2.541 0.280 2.717 20 After several discussions with Dr. and Table 30 shows the lateraldirectional derivatives.592 34.364 1.215 1.069 16. There will be some flow separation that does not result in downwash.357 5 1. Table 29: Meridian VORSTAB Model 1 Longitudinal Derivatives (1/rad) CL CD Cm CLq Cmq (deg) 5 0 5 10 15 20 2. In the VORSTAB output file. Table 29 shows the longitudinal derivatives obtained.738 0.581 2.radius of the fuselage be no larger than 1.04 3.5042 1. there is negative lift and zero drag at multiple angles of attack.037 17.99 0.210 13.8592 2.611 3.603 35.799 Table 30: Meridian VORSTAB Model 1 LateralDirectional Derivatives (1/rad) Cyp Clp Cnp Cyr Clr Cnr Cy Cl Cn (deg) 17.110 35.183 2.825 0.588 0.041 1.49 16.609 33.602 2.489 3.109 17.129 4. The vtail are in the path of this flow wake and rendering them ineffective.194 0.411 0.041 34.463 16.510 5.185 10.07 0.032 0.177 3.161 1. VORSTAB predicts that all of the flow over the top of the fuselage will become downwash.1402 0 0 0 0 0 0.48 10.079 15 12. The downwash was so large for this model that VORSTAB did not obtain any valid results.697 1. As seen from the results. for any unit system.6006 2.054 3.787 5 9.917 1. Lan about the possible cause for the problem.659 2.589 8.700 1.485 2.672 3.412 0. he offered advice on what he believed to be the problem and how it could possibly be fixed.1363 3.523 5.1976 1.120 6.746 17.052 2.5477 33.655 2.
The VORSTAB input file can be found in Appendix A. Lan recommended changing the fuselage to help eliminate some of the downwash and obtain valid results. The forward section of the fuselage size did not change and the aft sections radius was increased. The size of the increase of the aft fuselage sections was changed multiple times. 9. 85 .on the fuselage can be seen as extremely high pressure on the upper surface. Figure 36 shows the increase in size of the empennage. Having circular crosssections allowed for much faster modifications to the model.2 Model 2: Larger Empennage The best way to help eliminate some of the downwash was to eliminate the large amount of convergence. and also. This method replaced the fuselage body with a fuselage wake surface. to estimate the crosssections of the fuselage with a circular crosssection. The final increase was to approximately half of the forward section size. Dr.2. Both models are to half scale. Table 31 shows the diameter used for this approximation as well as the right side view of the fuselage diameter for the first model.
5042 0.4121 0.5 7 7.042 1.75 0.5 8 8.375 1.4525 0.3458 0.6125 0.5 6 6.5183 0.608 2.521 4.375 1.238 1.5183 0.9458 0.3225 0.Table 31: Fuselage Diameter for Meridian Models 1 and 2 Station 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 x (ft) 0 0.604 Model 1 Diameter (ft) 0 0.7542 0.592 7.317 1.058 2.463 8.9458 0.7542 0.7542 0.7542 0.9879 1.563 8.3325 0.118 0.321 1.25 5.9879 1.5 9 Increased Low er Emp Original Upper Emp Origianl Low er Emp 1 X (ft) Figure 36: Meridian Empennage Models 1 and 2 86 .376 1.0417 0 Model 2 Diameter (ft) 0 0.146 4.5183 0.292 1.3225 0.238 1.388 1.376 1.317 1.118 0.2042 1 Increased Upper Emp Z (ft) 0 5 5.7917 1.5183 0.125 8.438 2.642 5.883 3.771 6.292 1.7542 0.388 1.6708 0.
3414 0.4348 8.0000 0.1232 6.5876 0.Model 2‟s results were much better than the first model.7959 6. Also.6175 8. Table 32 shows the longitudinal derivatives and Figure 37 through Figure 40 show Table 32 in graphical form.0000 0. there is still camber in the empennage that results in some downwash that increases the drag.4923 6.1682 7. Increasing the fuselage size added some to the drag.6688 0.6382 0. For this model the drag should be ignored. but the drag was very high. because there is no flow separation.4179 0. Table 32: Meridian VORSTAB Model 2 Longitudinal Control Derivatives and Affected by Pitch Rate CL CD Cm CL.q (deg) 5 0 5 10 15 20 0.2364 7.6480 6.5335 0.1877 0.0978 0.4912 1.3140 1.5662 8. The input file for this model can be found in Appendix A.8528 4.3508 7.4931 2.0352 0.5872 0.6702 7.0000 0.q CD.q Cm.9660 7.8880 87 .5967 8.0000 0. Like all VORSTAB models this model does not show that the aircraft stalls.0000 0. The high drag was a result of a few different things.2621 0.0000 8.9949 1.
2 1.5 0 5 10 15 20 1 (deg) Figure 37: Meridian VORSTAB Model 2 Lift Coefficient due to Angle of Attack 10 9 8 7 CD (N.5 0 5 0.5 1 CL (N.) 6 5 4 3 2 1 0 (deg) 5 0 5 10 15 20 Figure 38: Meridian VORSTAB Model 2 Drag Coefficient due to Angle of Attack 88 .D.) 0.D.
D.1 0 5 10 15 20 0.6 (deg) Figure 39: Meridian VORSTAB Model 2 Pitching Moment Coefficient due to Angle of Attack 10 8 6 4 (1/rad) 2 0 2 0 4 6 8 10 (deg) clq cdq cmq 5 10 15 20 5 Figure 40: Meridian VORSTAB Model 2 Longitudinal Derivatives due to Pitch Rate 89 .3 0.) 0.2 CM (N.0 5 0.5 0.4 0.
1283 0.4882 0.0288 0. All of the other derivatives have an appropriate value and trend.1552 0.The lift coefficient is appropriate in magnitude.0029 0.0991 0.9597 0.1077 0.1146 0. the zero lift angle of attack should be negative. The graphical form of this table is shown in Figure 41 and Figure 42.2357 90 .4765 0. For this type of aircraft.2432 0. Table 34 shows the lateraldirectional control derivatives due to yaw rate.9603 0. The lateraldirectional control derivatives due to sideslip and roll rate for Model 2 can be seen in Table 33.0312 0.1010 0.7403 0.5138 0. It is obvious that the drag is too high.2220 0.0258 0.0837 0.0493 0.1083 0.7695 0.5775 0.0877 0.2913 0. Cl.2312 0. Table 33: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to Sideslip and Roll Rate Cy. For example. Cn.4538 0. This lift would result in the aircraft cruising at a positive angle of attack.1071 0.1242 0.1356 0.3247 0. but the zero lift angle of attack is positive.0569 0. the pitching moment coefficient shows the aircraft is inherently stable.0801 0. and Figure 43 shows this in graphical form. 5 0 5 10 15 20 0.1203 0.p Cl.0167 0.p (deg) Cy.p Cn.
5 0.4 0.6 0.1.3 0.8 (1/rad) 0.2 (1/rad) 0.2 1 0.4 0.1 0 5 10 15 20 0.2 0 5 0.2 0 5 (deg) Cyb Clb 10 15 20 Cnb Figure 41: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to Sideslip 0 5 0.6 (deg) cyp clp cnp Figure 42: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to Roll Rate 91 .
but the VORSTAB results show a positive value.1 (1/rad) 0 0 0.1548 0.1840 0.Usually.0949 0.2091 0. Sideforce coefficient due to roll rate has a positive and negative typical range.0030 0.2017 0.1126 0.3 0.0244 0.0748 0.r Cn.0549 0.2432 0. the sideforce coefficient due to sideslip is negative.0483 0.2263 0.r (deg) 5 0 5 10 15 20 0.0672 0. The rest of the derivatives are as expected.r Cl.2550 0. This is seen in the VORSTAB results.1741 0. Table 34: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to Yaw Rate Cy.2 0.3 (deg) cyr clr cnr Figure 43: Meridian VORSTAB Model 2 LateralDirectional Derivatives due to Yaw Rate 92 .1537 0. but it is usually negative due to the moment arm.2 0.0429 0.1 5 10 15 20 5 0.
This vertical tail contribution outweighs the wingfuselage contribution for this aircraft since it has a negative value. This derivative is normally a positive value. Deflecting the control surfaces affects the stability and control derivatives. since the wingfuselage contribution is always positive and outweighs the vertical tail contribution. These types of control surfaces are called rudevators. Table 35 through Table 37 show the control derivatives due to different control surface deflections. The Meridian has a vtail so there is not a conventional rudder or elevators. The vtail imitates a rudder input with an asymmetrical deflection of the tails and imitates an elevator input with a symmetrical deflection of the tails. except for the rolling moment coefficient. These three tables are also graphical depicted in Figure 44 through Figure 52. 93 . The vertical tail contribution can be either positive or negative.The lateraldirectional derivatives due to yaw rate have values that are expected.
0131 0.0000 0.0000 0.0456 0.0209 0.0100 0.0230 0.0066 0.0220 0.0099 0.0153 0.0417 0.0033 0.0066 0.0000 0.0061 0.0098 0.0000 0.0222 0.0045 0.0462 0.0121 0.0050 0.0000 0.0060 0.0000 0.0306 0.0000 0.0087 0.0090 0.0122 0.0066 0.0050 0.0043 0.0000 0.0000 0.0229 0.0121 0.0460 0.0043 0.0061 0.0303 0.0000 0.0033 0.0000 0.0000 0.0458 0.0049 0.Table 35: Meridian VORSTAB Model 2 Lateral Directional Derivatives due to Aileron Deflection Cy Cl Cn (deg) a 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 10 10 10 10 10 5 5 5 5 5 0 0 0 0 0 5 5 5 5 5 10 10 10 10 10 0.0045 0.0111 0.0000 0.0066 0.0208 0.0000 0.0000 0.0230 0.0031 0.0047 0.0111 0.0091 0.0152 0.0024 0.0087 0.0100 0.0061 0.0050 0.0131 0.0231 0.0031 0.0061 94 .0122 0.0024 0.0048 0.0060 0.0413 0.
) 0 10 5 0.04 0.D.015 0.02 0 5 10 0.06 aileron deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 45: Meridian VORSTAB Model 2 Rolling Moment Coefficient due to Aileron Deflection 95 .005 Cy (N.06 0.01 0.0.01 0.02 Cl (N.015 aileron deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 44: Meridian VORSTAB Model 2 Sideforce Coefficient due to Aileron Deflection 0.005 0 5 10 0.) 0 10 5 0.04 0.D.
96 . VORSTAB shows that this derivative actually shows that it will yaw the aircraft into the intended turn. a positive aileron deflection produces a positive rolling moment.015 aileron deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 46: Meridian VORSTAB Model 2 Yawing Moment Coefficient due to Aileron Deflection The sideforce coefficient due to aileron deflection can be normally be neglected.D.01 0. the yawing moment coefficient is usually negative since is tends to yaw the aircraft out of an intended turn. so it will always calculate a value. but the contribution is very small and almost negligible.0. VORSTAB does not assume that it should be zero.005 0 5 10 0.) 0 10 5 0.01 0. As expected. Also. This is not the case for the Meridian. unless the rolling moment controls are near a vertical surface.015 0.005 Cn (N.
1227 4.2988 1.5645 1.9911 1.Table 36: Meridian VORSTAB Model 2 Longitudinal Derivatives Affected by Symmetrical Tail Deflections CL CD Cm (deg) re 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 12 12 12 12 12 6 6 6 6 6 0 0 0 0 0 6 6 6 6 6 12 12 12 12 12 0.6229 0.9918 1.3224 1.0733 0.3422 1.5659 0.6660 1.2805 0.4777 8.4301 0.5431 1.6567 0.0234 1.5399 6.0035 1.1756 0.9323 0.2052 0.1009 0.1278 0.5193 0.9060 1.6096 4.0375 7.3564 0.3764 0.1793 6.0814 0.3009 0.3462 0.2347 1.2676 8.4062 0.8064 8.0140 1.7613 4.4186 0.6465 0.5270 2.8756 6.2854 2.3862 0.3111 1.1596 2.5545 1.3255 0.3830 2.6017 97 .2847 0.6793 6.5303 0.5257 0.3693 6.2122 0.4716 4.3290 0.0916 0.4564 0.1130 0.3327 1.4944 1.6600 8.9049 4.6353 0.4515 0.4806 0.2526 0.6814 3.5381 0.4928 0.7221 1.
6 0.2 2 4 6 8 10 12 0.8 1.) 5 4 3 2 1 0 12 10 8 6 4 2 0 2 4 6 8 10 12 elevator deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 48: Meridian VORSTAB Model 2 Drag Coefficient due to Elevator Input 98 .4 0.D.2 0 12 10 8 6 4 2 0 0.2 1 0.8 CL(N.D.1.6 1.4 elevator deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 47: Meridian VORSTAB Model 2 Lift Coefficient due to Elevator Input 10 9 8 7 6 CD (N.4 1.) 0.
shows increasing pitching moment and lift.) 0.2 0.3 0 2 4 6 8 10 12 Cm (N.7 elevator deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 49: Meridian VORSTAB Model 2 Pitching Moment Coefficient due to Elevator Input A positive elevator input. 99 .4 0.6 0.5 0. As stated earlier the drag is going to be neglected because of its high magnitude.D.0 12 10 8 6 4 2 0. This is expected since the elevators are used to increase the pitch and create lift.1 0.
0000 0.0203 0.0051 100 .0011 0.0056 0.0010 0.0000 0.0056 0.0116 0.0000 0.0019 0.0022 0.0220 0.0101 0.0011 0.0026 0.0103 0.0236 0.0059 0.0114 0.0021 0.0103 0.0000 0.0022 0.0011 0.0011 0.0038 0.0022 0.0074 0.0023 0.0082 0.0023 0.0000 0.0203 0.0051 0.0022 0.0051 0.0028 0.0011 0.0000 0.0010 0.0101 0.0000 0.0000 0.0000 0.0000 0.0162 0.0026 0.0089 0.0082 0.0011 0.0059 0.0000 0.0000 0.0011 0.0022 0.0051 0.0089 0.0022 0.0000 0.0049 0.0011 0.0045 0.0162 0.Table 37: Model 2 Lateral Directional Control Derivatives Affected by Asymmetrical Tail Deflection Cy Cl Cn (deg) re 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 10 10 10 10 10 5 5 5 5 5 0 0 0 0 0 5 5 5 5 5 10 10 10 10 10 0.0000 0.0038 0.0116 0.0094 0.0045 0.0049 0.0074 0.0000 0.0102 0.0114 0.0028 0.
03 0.01 0 5 10 0.002 0.D.0025 0.0005 0.03 tail deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 50: Meridian VORSTAB Model 2 Sideforce Coefficient due to Rudder Input 0.0015 0.001 0.0015 0.0025 tail deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha 0 5 10 Figure 51: Meridian VORSTAB Model 2 Rolling Moment Coefficient due to Rudder Input 101 .) 0 10 5 0.0005 Cl (N.D.0.) 0 10 5 0.001 0.02 0.02 0.002 0.01 Cy (N.
Table 38 shows the longitudinal control derivatives and Figure 53 102 .005 0 5 10 0. 9.D.3 Model 3: Zero Camber Fuselage The third model created was the same as the second model.2. This is all seen in the VORSTAB results. and a positive sideforce will usually generate a positive rolling moment. but the camber was taken out of the fuselage.015 tail deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 52: Meridian VORSTAB Model 2 Yawing Moment Coefficient due to Rudder Input A positive rudder input will produce a positive sideforce. Having zero camber in the fuselage allows for the assumption that the downwash produced by the fuselage is negligible.0.005 Cn (N.) 0 10 5 0.015 0.01 0.01 0. The input file for this model can be found in Appendix A. This type of rudder deflection will also generate a negative yawing moment. Removing the camber removed the amount of downwash that was produced from the large negative camber in the empennage.
3595 7.5543 1.0000 0.0703 0.0000 0.D.0596 0.5 0 5 0.0000 0.3564 1.6312 0.8832 2 1.0467 0.2448 7. Table 39 shows the lateraldirectional control derivatives and Figure 57 through Figure 60 depicts Table 39 in graphical form.6632 7.0000 0.5992 8.0000 0.1494 0.4238 0.1686 0.6110 0.5 1 CL (N.2318 1.0000 8.1767 7.through Figure 56 depicts Table 38 in graphical form.0819 0.5688 8.0982 0.7157 0.8019 6.5 0 5 10 15 20 1 (deg) Figure 53: Meridian VORSTAB Model 3 Lift Coefficient due to Angle of Attack 103 .9047 1.1331 6.2014 0. Table 38: Meridian VORSTAB Model 3 Longitudinal Derivatives at due to Pitch Rate CL CD Cm CL_q CD_q Cm_q (deg) 5 0 5 10 15 20 0.9744 7.) 0.2017 0.5935 0.2437 6.4373 8.
2 0.2 1 CD (N.05 0 CM (N.D.15 0.2 0.15 0.3 5 10 15 20 (deg) Figure 55: Meridian VORSTAB Model 3 Pitching Moment Coefficient due to Angle of Attack 104 .) 0.25 0.) 5 0.25 0.2 0 5 0 5 (deg) 10 15 20 Figure 54: Meridian VORSTAB Model 3 Drag Coefficient due to Angle of Attack 0.4 0.6 0.8 1.4 1.D.05 0 0.6 1.1 0.1 0.8 0.1.
472 0.116 0. these longitudinal derivatives are within the appropriate range and have the expected sign convention. but is much better than the second model.466 0.119 0. The drag is still high.062 0.145 0.223 0.069 0.126 0.051 0.10 8 6 4 (1/rad) 2 0 2 0 4 6 8 10 (deg) CL_q CD_q CM_q 5 5 10 15 20 Figure 56: Meridian VORSTAB Model 3 Longitudinal Coefficients due to Pitch Rate Similar to Model 2.022 0.941 0.260 0.708 0.005 0.244 0.232 0.823 0.236 0. Table 39: Meridian VORSTAB Model 3 LateralDirectional Derivatives Cyp Clp Cnp Cyr Clr Cnr (deg) Cy Cl Cn 5 0 5 10 15 20 0.328 0.046 0.126 0.071 0.523 0.213 0.032 0.108 105 .183 0.292 0.104 0.123 0.195 0.081 0.118 0.067 0.024 0.994 0.077 0.213 0.018 0.186 0.026 0.476 0.032 0.056 0. Lift is negative at zero degree angle of attack.161 0.045 0.030 0.059 0. and should be positive.668 0.235 0.069 0.107 0.103 0.
6 (deg) Cy_p Cl_p Cn_p Figure 58: Meridian VORSTAB Model 3 LateralDirectional Derivatives due to Roll Rate 106 .015 (1/rad) 0.01 0.2 (1/rad) 0 5 10 15 20 0.005 0 5 0.3 0.005 (deg) Cy_b Cl_b Cn_b 0 5 10 15 20 Figure 57: Meridian VORSTAB Model 3 LateralDirectional Derivatives due to Sideslip 0 5 0.0.02 0.1 0.5 0.4 0.
1 (1/rad) 0 0 0.0. Table 41 shows the affect of a symmetrical tail deflection on the longitudinal derivatives. The rest of the derivatives are as expected. Sideforce coefficient due to roll rate has a positive and negative typical range. The control surface deflections affect the stability and control derivatives in both the longitudinal mode and the lateraldirectional mode.1 0.2 0. and Table 42 shows 107 . but the VORSTAB results show a positive value. but rather the angle of attack has a large influence on the control of the aircraft. The rolling moment and yawing moment coefficients due to roll rate have extreme fluctuations from one angle of attack to another. This does not necessarily mean that it is wrong. Table 40 shows the affects of the aileron on the lateraldirectional derivatives.2 0. but it is usually negative due to the moment arm. This is seen in the VORSTAB results.3 (deg) Cy_r Cl_r Cn_r 5 5 10 15 20 Figure 59: Meridian VORSTAB Model 3 LateralDirectional Derivatives due to Yaw Rate The sideforce coefficient due to sideslip is usually negative.3 0.
0050 0.0000 0.0113 0.0066 0.006 0.04169 0.01544 0.04558 0.0047 0.006 0. Table 40: Meridian VORSTAB Model 3 LateralDirectional Derivative due to Aileron Deflection Cy Cl Cn (deg) a (deg) 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 10 10 10 10 10 5 5 5 5 5 0 0 0 0 0 5 5 5 5 5 10 10 10 10 10 0.02095 0.0155 0.0050 0.0460 0.02302 0.00313 0.0050 0.0000 0.0230 0.0000 0.0000 0.0225 0.0131 0.0120 0.0000 0.0000 0.0000 0.0048 0.0062 0.0033 0.0024 0.0033 0.0089 0.0045 0.01239 0.01315 0.0000 0.0024 0.0420 0.0050 0.0000 0.0310 0.0000 0.00885 0.0044 0.02291 0.02231 0.0000 0.00442 0.0124 0.0099 0.012 0.00627 108 .03071 0.0099 0.0231 0.04581 0.0099 0.0462 0.0045 0.01121 0.0000 0.0031 0. These tables are depicted in Figure 60 through Figure 68.the affect of an asymmetrical tail deflection on the lateraldirectional derivatives.0066 0.00657 0.0000 0.0063 0.0000 0.0066 0.0098 0.0210 0.0090 0.00619 0.0000 0.0090 0.
02 0 5 10 0.02 Cl (N.015 0.015 aileron deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 60: Meridian VORSTAB Model 3 Sideforce Coefficient due to Aileron Deflection 0.) 0 10 5 0.0.06 0.005 0 5 10 0.06 aileron deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 61: Meridian VORSTAB Model 3 Rolling Moment Coefficient due to Aileron Deflection 109 .04 0.D.01 0.04 0.005 Cy (N.D.) 0 10 5 0.01 0.
D. This is expected since there are no rolling moment controls close to a vertical surface. 110 .015 0.015 aileron deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 62: Meridian VORSTAB Model 3 Yawing Moment Coefficient due to Aileron Deflection A sideforce coefficient due to aileron deflection is usually negligible.005 0 5 10 0.0.01 0. and VORSTAB shows a very small value. Positive aileron deflections result in a positive rolling moment and yawing moment.01 0.) 0 10 5 0.005 Cn (N.
5518 0.2173 1.0730 0.1718 0.5616 0.3648 1.8198 0.1572 0.3088 0.2421 0.2620 0.0019 0.6789 0.4230 0.1225 0.1847 0.2591 1.1028 0.5057 0.0948 0.0631 0.1932 0.9135 1.2400 1.2786 0.Table 41: Meridian VORSTAB Model 3 Longitudinal Derivatives due to Symmetrical Tail Deflections CL CD Cm (deg) re (deg) 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 12 12 12 12 12 6 6 6 6 6 0 0 0 0 0 6 6 6 6 6 12 12 12 12 12 0.0849 0.1420 0.0024 0.4096 0.2132 0.5787 0.5409 0.5706 0.9356 1.0956 0.0814 0.2290 1.1091 0.0482 0.4568 0.1836 0.1896 0.1337 0.0111 0.4349 0.2500 1.1707 0.1949 0.3404 111 .8708 0.0614 0.0534 0.9102 0.4285 1.0400 0.9250 1.2731 0.0564 0.8894 1.5916 0.0916 0.9016 1.2016 0.2607 0.0935 0.0474 0.0496 0.2335 0.4476 0.0970 0.0488 0.1098 0.2042 0.3106 1.9528 0.
4 1.6 1.2 0 2 4 6 8 10 12 0.6 0.D.6 0.4 1.8 CD (N.2 0 12 10 8 6 4 2 0 2 tail deflection (deg) 10º alpha 4 6 8 10 12 0º alpha 5º alpha 15º alpha 20º alpha Figure 64: Meridian VORSTAB Model 3 Drag Coefficient at due to Elevator Input 112 .4 0.1.2 1 CL (N.4 tail deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 63: Meridian VORSTAB Model 3 Lift Coefficient due to Elevator Input 1.8 0.8 1.) 0.2 1 0.2 0 12 10 8 6 4 20.D.4 0.) 0.
the drag is much higher than it should be and is not accurate.05 0 12 CM (N. As stated previously.0. VORSTAB shows that at angles of attack of 10˚ or higher the drag will decrease with positive elevator deflections.25 0.2 0.15 2 4 6 8 10 12 0. When the lift increases the drag should always increase. The drag coefficient increases on some angles of attack and decreases on the others.1 0. 113 . This is the purpose of the positive elevator input.) 10 8 6 4 2 0.4 tail deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 65: Meridian VORSTAB Model 3 Pitching Moment Coefficient due to Elevator Input Elevator input deflection will increase the lift and pitching moment.1 0.05 0 0.3 0.15 0.D. This is due to the definition of drag being a function of lift.35 0.
Table 42: Meridian Model 3 LateralDirectional Derivatives Affected by Asymmetrical Tail Deflections Cy Cl Cn (deg) re (deg) 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 10 10 10 10 10 5 5 5 5 5 0 0 0 0 0 5 5 5 5 5 10 10 10 10 10 0.0011 0.0019 0.0123 0.0098 0.0092 0.0083 0.0215 0.0020 0.0000 0.0010 0.0145 0.0000 0.0215 0.0010 0.0000 0.0000 0.0226 0.0209 0.0010 0.0020 0.0162 0.0000 0.0000 0.0035 0.0070 0.0064 114 .0000 0.0229 0.0162 0.0000 0.0064 0.0046 0.0009 0.0048 0.0021 0.0010 0.0020 0.0082 0.0021 0.0082 0.0109 0.0117 0.0018 0.0239 0.0020 0.0000 0.0083 0.0033 0.0000 0.0073 0.0011 0.0073 0.0098 0.0011 0.0000 0.0109 0.0088 0.0010 0.0022 0.0046 0.0042 0.0070 0.0000 0.0042 0.0000 0.0000 0.0145 0.0000 0.0022 0.0033 0.0035 0.0051 0.0109 0.0011 0.0109 0.
01 Cy (N.001 0.0005 Cl (N.0015 0.D.0025 0.002 0.D.) 0 10 5 0.02 0.002 0.0015 0.0005 0.0.01 0 5 10 0.0025 tail deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha 0 5 10 Figure 67: Meridian VORSTAB Model 3 Rolling Moment Coefficient due to Rudder Input 115 .03 0.02 0.03 tail deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 66: Meridian VORSTAB Model 3 Sideforce Coefficient due to Rudder Input 0.001 0.) 0 10 5 0.
and Figure 69 through Figure 72 depicts Table 43 in graphical form.4 Model 4: Wings and VTail Since the problem with the downwash came from the fuselage.0.01 0.015 tail deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 68: Meridian VORSTAB Model 3 Yawing Moment Coefficient due to Rudder Input As expected. 9. Appendix A contains the input file for this model.015 0. Table 43 shows the longitudinal control derivatives. The wings and the vtail were kept in their exact location but were extended to the centerline. 116 .01 0. a positive rudder input results in a positive sideforce and rolling moment.005 Cn (N. This type of model helps the designer see what effect the fuselage is having on the results.2.005 0 5 10 0.D.) 0 10 5 0. It also results in a negative yawing moment. it was removed from the forth model.
6302 0.0202 0.1135 0.7241 3.Table 43: Meridian VORSTAB Model 4 Longitudinal Derivatives CL CD Cm CLq CDq Cmq (deg) 5 0 5 10 15 20 0.0309 0.5 0 5 10 15 20 1 (deg) Figure 69: Meridian VORSTAB Model 4 Lift Coefficient due to Angle of Attack 117 .6132 3.0000 0.3916 0.0155 0.6904 3.3238 0.7654 3.0431 0.0000 0.7942 3.D.5476 3.1517 3.5400 3.0244 0.2843 0.7654 3.9934 1.0995 0.) 0 5 0.6798 3.0000 0.5 1 0.7969 0.0000 3.0000 0.7503 3.5 CL (N.0012 0.0581 0.3255 1.1256 0.0736 0.1031 0.0000 0.
D.02 0.06 0.1 0.16 (deg) Figure 71: Meridian VORSTAB Model 4 Pitching Moment Coefficient due to Angle of Attack 118 .14 0.12 0.) 0.03 0.01 0 5 0 5 (deg) 10 15 20 Figure 70: Meridian VORSTAB Model 4 Drag Coefficient due to Angle of Attack 0 5 0.05 0.D.02 0.04 CD (N.04 0.06 0 5 10 15 20 CM (N.08 0.07 0.) 0.0.
A possible cause is the high camber of the airfoils is producing downwash.5 4 3 2 (1/rad) 1 0 1 2 3 5 0 5 10 15 20 4 5 (deg) CL_q CD_q CM_q Figure 72: Meridian VORSTAB Model 4 Longitudinal Derivatives due to Pitch Rate As seen from the pitching moment graph. 119 . With a positive slope the aircraft wants pitch away from the stable point. This is seen by the slope of Figure 71 becoming positive. There could be many causes for these results. The lateraldirectional control derivatives are shown in Table 44. except for the angle of zero lift being around 4˚. the wings and tail become unstable at a certain angle of attack. Figure 73 through Figure 75 depicts Table 44 in graphical form. it can be seen that the fuselage contributes significantly to the drag in the VORSTAB model. With the airfoils used on the wings and tail and their geometric configuration it would be expected that the angle of zero lift would be at a negative angle of attack. From these results. The drag and lift results are as expected.
0226 0.0006 0.031 0.014 0.0017 0.0005 0.1837 0.132 0.4 0.6 (1/rad) 0.1107 0.016 0.009 0.1145 0.252 0.090 0.011 0.0013 0.0587 0.1528 0.071 0.1229 0.1475 0.2 (deg) Cy_b Cl_b Cn_b 0 5 10 15 20 Figure 73: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to Sideslip 120 .043 0.311 0.232 0.1719 0.473 0.Table 44: Meridian VORSTAB Model 4 LateralDirectional Control Derivatives Cyp Clp Cnp Cyr Clr Cnr Cy Cl Cn (deg) 0.067 0.0435 0.0024 0.014 0.198 0.029 1 0.2169 0.2 0 5 0.0022 0.0006 0.0019 0.087 5 0 5 10 15 20 0.2353 0.0005 0.0021 0.479 0.1561 0.0016 0.1703 0.296 0.1291 0.4940 0.013 0.8 0.0019 0.1461 0.3801 0.011 0.003 0.140 0.
3 0.25 0.4 0.1 (1/rad) 0 5 10 15 20 0.2 0.1 0.15 0.2 0.3 0 5 10 15 20 (deg) Cy_r Cl_r Cn_r Figure 75: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to Yawing Moment 121 .2 0.1 0 5 0.05 0.15 0.1 0.5 0.05 0 (1/rad) 5 0.0.6 (deg) Cy_p Cl_p Cn_p Figure 74: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to Roll Rate 0.
The sideforce coefficient due to sideslip. Cy. is unstable at all angles of attack. This will also be influenced by the fuselage and could be stable if it were present. Table 45 shows the lateraldirectional derivatives due to an aileron deflection. The yawing moment coefficient due to yaw rate. Cnr. These tables can be seen in graphical form in Figure 76 through Figure 84. Table 46 shows the longitudinal derivatives due to a symmetrical deflection of the rudevators. When it is unstable the yawing rate increases and can lead to a spin. This can be neglected because having the fuselage will dramatically affect this. A stable aircraft should naturally want to stop yawing or slow down. 122 . is also unstable at higher angles of attack. and Table 47 shows the lateraldirectional derivatives due to an asymmetrical deflection of the rudevators.
0049 0.0245 0.0000 0.0456 0.0052 0.0065 0.0027 0.0050 0.0037 0.0228 0.0074 0.0229 0.0000 0.0000 0.0117 0.0452 0.0454 0.0100 0.0000 0.0000 0.0000 0.0123 0.0098 0.0037 0.0051 0.0127 0.0035 0.0054 0.0247 0.0102 0.0000 0.0059 0.0221 0.0220 0.0000 0.0071 0.Table 45: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to Aileron Deflection Cy Cl Cn (deg) a (deg) 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 10 10 10 10 10 5 5 5 5 5 0 0 0 0 0 5 5 5 5 5 10 10 10 10 10 0.0000 0.0050 0.0000 0.0168 0.0059 0.0000 0.0051 0.0102 0.0228 0.0127 0.0437 0.0131 0.0000 0.0075 0.0167 0.0027 0.0049 0.0131 0.0065 0.0336 0.0441 0.0000 0.0064 0.0000 0.0101 0.0071 123 .0051 0.0103 0.0035 0.0117 0.0227 0.0098 0.0458 0.0099 0.0000 0.0064 0.0124 0.0055 0.0333 0.
) 0 10 5 0.01 0.D.005 0 5 10 0.04 0.015 0.0.02 Cl (N.04 0.D.01 0.015 aileron deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 76: Meridian VORSTAB Model 4 Sideforce Coefficient due to Aileron Deflection 0.005 Cy (N.06 0.02 0 5 10 0.) 0 10 5 0.06 aileron deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 77: Meridian VORSTAB Model 4 Rolling Moment Coefficient due to Aileron Deflection 124 .
005 0 5 10 0.015 aileron deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 78: Meridian VORSTAB Model 4 Yawing Moment Coefficient due to Aileron Deflection The sideforce coefficient should be very small without the fuselage present. 125 .01 0.005 Cn (N.) 0 10 5 0.D. and with it present all of the derivatives will change dramatically.0.01 0.015 0.
0581 0.1816 0.6218 0.0693 0.1256 0.Table 46: Meridian VORSTAB Model 4 Longitudinal Derivatives due to Symmetrical Tail Deflection CL CD Cm (deg) re (deg) 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 12 12 12 12 12 6 6 6 6 6 0 0 0 0 0 6 6 6 6 6 12 12 12 12 12 0.1216 0.6302 0.0244 0.0791 0.0430 0.3826 0.2769 0.1675 0.4094 0.1621 0.0736 0.0201 0.0581 0.2689 0.0244 0.0202 0.0201 0.1517 0.2843 0.0202 0.1784 0.0821 0.9769 1.0202 0.0900 0.3740 0.3916 0.0159 0.0363 0.0995 0.0433 0.6136 0.0305 0.6380 1.0431 0.1554 0.9934 1.0580 0.2012 126 .1135 0.0892 0.0431 0.0155 0.1785 0.0581 0.0245 0.1031 0.6450 1.0649 0.0932 0.0244 0.0451 0.2965 0.0072 1.0563 0.4007 0.0159 0.0008 1.0432 0.0588 0.0160 0.0641 0.0244 0.1343 0.1341 0.1473 0.0580 0.0159 0.9854 1.2909 0.
4 0.2 1 0.D.) 0.02 0.2 0.4 1.03 0.6 CL (N.04 CD (N.01 0 12 10 8 6 4 2 0 2 4 6 8 10 12 tail deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 80: Meridian VORSTAB Model 4 Drag Coefficient due to Elevator Input 127 .05 0.) 0.1.4 2 4 6 8 10 12 0.D.06 0.07 0.6 tail deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 79: Meridian VORSTAB Model 4 Lift Coefficient due to Elevator Input 0.2 0 12 10 8 6 4 2 0 0.8 0.
128 .05 0 2 4 6 8 10 12 0.0 12 10 8 6 4 2 0. Drag stays rather steady when in reality it will change a small amount.D.15 0.1 Cm (N.) 0. as expected.25 elevator deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 81: Meridian VORSTAB Model 4 Pitching Moment Coefficient due to Elevator Input The lift and pitching moment both increase with a positive pitching moment.2 0.
0016 0.0031 0.0016 0.0102 0.0052 0.0218 0.0000 0.0095 0.0084 0.0000 0.0127 0.Table 47: Meridian VORSTAB Model 4 LateralDirectional Derivatives due to Asymmetrical Deflection Cy Cl Cn (deg) re (deg) 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 10 10 10 10 10 5 5 5 5 5 0 0 0 0 0 5 5 5 5 5 10 10 10 10 10 0.0084 129 .0218 0.0198 0.0123 0.0110 0.0016 0.0119 0.0043 0.0054 0.0031 0.0100 0.0016 0.0129 0.0016 0.0033 0.0198 0.0000 0.0103 0.0033 0.0000 0.0016 0.0092 0.0047 0.0016 0.0119 0.0110 0.0000 0.0251 0.0000 0.0000 0.0248 0.0047 0.0050 0.0092 0.0031 0.0252 0.0000 0.0100 0.0000 0.0000 0.0100 0.0128 0.0053 0.0016 0.0050 0.0043 0.0029 0.0246 0.0000 0.0000 0.0032 0.0107 0.0106 0.0238 0.0032 0.0032 0.0000 0.0016 0.0032 0.0226 0.0000 0.0053 0.0015 0.0000 0.
) 0 10 5 0.02 0.01 0 5 10 0.003 0.03 0.001 0.004 tail deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha 0 5 10 Figure 83: Meridian VORSTAB Model 4 Rolling Moment Coefficient due to Rudder Input 130 .03 tail deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 82: Meridian VORSTAB Model 4 Sideforce Coefficient due to Rudder Input 0.003 0.D.002 0.001 Cl (N.01 Cy (N.002 0.) 0 10 5 0.004 0.D.0.02 0.
01 0.2. The only results that are not shown in this table are the 131 .5 Meridian VORSTAB Conclusions The results from VORSTAB varied greatly from model to model. and the ranges are a function of Mach number. but this was not the case.) 0 10 5 0.005 Cn (N.015 0. These variations make it very difficult to draw any conclusions from VORSTAB. Roskam‟s typical ranges.005 0 5 10 0.01 0. these derivatives will be greatly influenced by having a fuselage presents in the model. Since the first model is the only one that has the exact configuration it would be expected that it is the most accurate. These ranges are for conventional aircraft. Table 48 shows the results compared to Dr. Ref [5].0. The forth model‟s results with the closest to the AAA results.D.015 tail deflection (deg) 0º alpha 5º alpha 10º alpha 15º alpha 20º alpha Figure 84: Meridian VORSTAB Model 4 Yawing Moment Coefficient due to Rudder Input Again. 9.
first model. The first model was used as a stepping stone to create the other models. so they were completely ignored. Model 1‟s results were not valid at all. 132 .
5 to 0.2337 15.0 to 0.1010 0.0 to 1.0 Within Range (Yes/No) Model 2 Yes No No Yes Yes Yes Yes No No Yes No Yes Yes Yes Yes No Yes Yes Yes No Yes Yes Yes Model 3 Yes No No Yes Yes Yes Yes No No Yes No Yes Yes Yes Yes No Yes Yes Yes No Yes Yes Yes Model 4 Yes Yes Yes Yes Yes Yes Yes No No Yes No Yes Yes Yes Yes No Yes Yes Yes No Yes Yes Yes There are multiple derivatives that do not fall into the typical ranges.4038 7.0479 Model 4 5.04 to 0.0123 0. Roskam 1.9313 0.2913 0.08 0.0 0.0858 0.1068 0.2316 0.7241 3.6 Negligible 0.1 to 0.0922 1.0312 0.15 to 0.0666 0.1 0.0 to 0.3450 0.5688 0.0974 0.1 to 2.0226 0.6 0.0 0.0 0.3600 0.0670 0.1291 0.08 to 0.0773 0.Table 48: Meridian VORSTAB Stability and Control Derivatives Dr.8 0.2606 0.0040 0.0583 0.1031 0.1609 3.046 0.2091 0.1675 6.1 to 4.2320 0.1682 8.0748 0.7695 0.0 to 8.0 4.8228 0.2 0.0 0.0567 0.0 to 1.0688 0.0 to 30.0 to 4.04 Negligible 0.0178 0.0 to 1.0559 Model 3 7.0940 0.1463 0.0612 Typical Ranges Dr.0117 0.3867 0.3308 0.5662 0.1020 0.3 to 0.0 to 0.1261 0.0 0.0692 0.8522 7.0 0.2263 0.0 0. The forth model also has the 133 .5 0.0242 0.1229 0.0 to 2.1767 8.0991 0.2628 0.0 to 0.6383 0.8 0.0 0.3304 0.7942 0.1083 0. Roskam’s Typical Ranges Stability and Control Derivatives Derivative (1/rad) CL CD Cm CLq Cmq Cl Clp Clr Cy Cyp Cyr Cn Cnp Cnr CLe CDe Cme Cla Clr Cya Cyr Cna Cnr Model 2 7.0565 0.0 to 4.0 to 0.1065 0.2960 0.2625 0.1248 0. All of the models have a high drag coefficient except the forth.1060 0.292 0.1303 0.0 90.580 0.4 0.
05 65.34 148.1291 0.3681 0.0612 AAA 5.2913 0.2628 0.36 72.69 39.0117 1.34 % Diff.23 31.0253 0 0.8522 7.11 59.0858 0.1099 0.0479 Model 4 5.7241 3.52 8059.60 46.32 134 .00 28.83 76.86 47.91 260.90 613.00 141.3304 0.1083 0.most derivatives that fall into the typical ranges described.45 13.232 0.4149 0.2263 0.81 80.30 55.1767 8.3217 0.0312 0.0117 0.61 74.0991 0.50 11.10 233.35 310. third.14 34.30 60.0688 0. Model 3 38.68 32.10 76.6709 0.1065 0.2316 0.68 31.0123 0.1648 0.74 600.28 16.21 27.5546 0.1463 0.86 13.80 44.0567 0.345 0. Table 49 shows a comparison of the second.296 0.94 55.2091 0.0666 0.0242 0.292 0.4789 0. Model 2 40.0692 0.39 50.23 38.0565 0.48 305.0583 0.0776 0.1338 0.2606 0.74 271.004 0.3308 0.62 133.5662 0.56 47.23 12.11 44.1386 0.81 27.43 132. and fourth model next to AAA.973 0.06 35.0226 0.0748 0.18 164. Model 4 3.1031 0.106 0.85 36.0974 0.1675 6.42 137.78 11395.1229 0.2316 0.2625 0.067 0.8228 0. Table 49: Meridian AAA and VORSTAB Comparison Stability and Control Derivatives Derivative (1/rad) CL CD Cm CLq Cmq Cl Clp Clr Cy Cyp Cyr Cn Cnp Cnr CLe CDe Cme Cla Clr Cya Cyr Cna Cnr Model 2 7.0559 Model 3 7.62 77.0773 0.0178 0.87 32.22 79.7942 0.52 170.1261 0.7695 0.1481 % Diff.1609 3.08 19.2337 15.36 139.094 0.102 0.0351 0.0922 1.1248 0.1068 0.0134 0.63 311.49 37.6383 0.4038 7.1409 0.40 616.32 65.6207 4.41 38.1465 0.5688 0.3867 0.47 146.1682 8.9313 0.78 4434. The forth model predicted several of the derivatives the best.101 0.046 0.06 10957.73 78.58 0.1303 0.36 0.6179 13.25 135.74 % Diff.
drove the need for a FLUENT model to be created. the downward swoop of the fuselage could result in downwash and flow separation. 9. It is recommended that several more wing and tail models be created to determine if it predicts the stability and control derivatives well as seen from the Meridian results. because they brought up possible issues that the Meridian might have. These two problems. For example.The fourth model best predicted the Meridian stability and control derivatives when compared to AAA.3 Linearized Model of the Meridian For the information and data on the state space model of the Meridian please refer to Appendix B. The Meridian is not a typical style of aircraft and the VORSTAB results were not what were expected. There are several derivatives that are extremely far off. but due to time constraints this was not done. 9. a 2D side profile of the fuselage was 135 . These results should not be completely ignored. After modeling these three different aircraft it can be concluded that VORSTAB works best with traditional style aircraft. Therefore. A 3D model was desired. such as the YAK54. The first flight tests also resulted in very high drag. the main concern is the downwash created over the fuselage that could blank out the vtails. This downwash and flow separation possible could render the tails ineffective and the aircraft unresponsive to tail inputs.4 FLUENT Modeling of the Meridian As learned from the VORSTAB model of the Meridian. but the fuselage will have an impact on most of these. downwash and drag.
4. The tip of the nose will be on the (0. and below the fuselage. and forward and aft had a length of 20 lengths of the fuselage. above the fuselage. 9. The size of the farfields depends on the flow being modeled and the shape of the object. 0. Above and below the fuselage had a height of 20 maximum thicknesses of the fuselage. This is where the flow can separate.1 GAMBIT Mesh Generation A large farfield was created around the axialcrosscut of the fuselage.created to see the downwash that was produced from the high camber. 0) point. The farfield was broken into six individual sections. aft (above and below) of the fuselage. For the 2D Meridian fuselage the critical flow section was the boundary layer of the Meridian walls. The flow at a large distance from the fuselage is not as critical as the flow near the walls. A complete 3D model of the Meridian should eventually be created. A very fine mesh is desirable everywhere. but this would take a large amount of computer memory and time to run. forward (above and below) of the fuselage. 136 . Figure 85 shows the Meridian crosssection. This created a large enough area where the flow could be appropriately modeled. The following are the steps taken to create the six farfields and their meshes: Step 1: Import the axial crosssection of the fuselage (points created as a text file). Different farfields allow the designer to create finer meshes in the more critical sections of the flow.
and then copy the same point to the height created in step 5. Step 5: Repeat step 4. labeled point A. where the nose cone meets the fuselage cowling (labeled point K) up to 20 times the maximum thickness minus the current height of the point. but extend from the tip of the tail (labeled point N) to aft of the fuselage. but use the point on the lower surface where the nose cone meets the lower fuselage cowling (labeled point L) called point D. called point C. labeled points E and F respectively. 3 137 . There should be 3 points forward of the fuselage. Step 7: Copy the points A and B to the heights created in step 4 and step 5 labeled points G. I and J respectively. labeled point B. H. Step 4: Copy the point. Step 6: Copy the point N up the same height as the point created in step 4.Figure 85: Meridian Fuselage Axial CrossSection for FLUENT Step 2: Create a line that extend from the tip of the nose (labeled point M) to 20 times the length of the fuselage forward. Step 3: Repeat step 2. on the upper surface.
Figure 86: Meridian Fuselage with Farfield Divisions Step 13: Create the six different farfield and the Meridian body face. Table 50 shows the corresponding lines to each face. Step 8: Use the NUMS tool to create a curved line on the upper surface of the nose cone. Figure 86 shows the Meridian axial crosssection and the newly created lines to create the farfields. and 2 points above and below the tip of the tail. DF. EI. The Meridian crosssection is the small white shape in the middle. Step 10: Use the line tool to create 3 lines from point N to points B. GC. Then repeat with point L to point D. E and F. FJ. 2 points above and below the nose cone.points aft. CE. This creates four individual lines. BI. and BJ. Step 12: Create lines AG. HD. upper surface of the fuselage. and lower surface of the fuselage. Step 11: Use the line tool to create a line from point K to point C. 138 . AH. Step 9: Next use the line tool to create a line from point M to point A. Repeat on the lower surface of the nose cone.
LN Step 14: Create „Edge Meshes‟ using Table 51. LM MA 100 100 70 8 92 Properties First Length = 0.Table 50: Meridian Farfield Edges Face Name Farfield 1 Farfield 2 Farfield 3 Farfield 4 Farfield 5 Farfield 6 Meridian Edges AG. NK. DL. EN. KN. CK. BJ CG. KM. FJ. Table 51: Meridian Edge Meshes Edge Interval Count AG. Figure 87 shows the meshes concentrated towards the fuselage body. LM. AM AH. LN KM. DF. The small intervals should be closer to the Meridian wall. FN MK. FN. CK LN. HD. EI. EN.20 Step 15: Create „face meshes‟ in the six farfields. IB. AM CE. CK. DL. 139 . BI.25 Successive Ratio = 1 Successive Ratio = 1 First Length = 0. DL EI.25 First Length = 0. DF. on all lines that have first lengths. GC. EN FJ. DH. ML. NB. The meshes is denser towards the Meridian body. NB. KN. NF. AH. JB. The lower image is a zoomed in image of the upper picture. NB CE.
Group 3 as DH. 2 and 3 were velocity inlets. With the groups created. and Group 5 was a wall. Group 4 was a pressure outlet. KN. ML and LN.Figure 87: Meridian Farfield Meshes Step 16: Create five groups of edges. These boundary conditions work because the farfields are so big that the flow can be considered unaffected by the fuselage body or totally recovered from the flow 140 . their corresponding boundary conditions were set. Groups 1. and Group 5 as MK. Group 4 as BI and BJ. Group 2 as CG. DF and FJ. Group 1 as edge AG and AH. CE and EI.
disturbances. Operating Conditions – Set this to zero to work in terms of absolute pressure.5 FLUENT Model Generation The FLUENT model can be run in many different methods depending on what the goals the designer is trying to achieve.112 kg/m3 and 1. and observe the flow separation and/or downwash that were produced from the high convergence and downward camber. The following are the setup criteria for the FLUENT model. SpalartAllmaras. If left at the default setting the pressures reported would be in terms of gauge pressure. At 1. For the Meridian.758x105 kg/ms respectively.1 for an initial guess. the results did not very greatly from one method to the next. The boundary conditions can be modified in FLUENT if they are not what were desired. No Energy Equation because the flow is not compressible Materials – Set the density and viscosity to the values that the Meridian will fly in. Solver – Segregated. With the GAMBIT model fully created export the 2D mesh and import it into FLUENT. Several different turbulent methods were tested to verify which one was the most accurate.000 m the approximate values are 1. the method chosen was the one that has the most accurate results in aerodynamics. the goal was to observe the boundary layer on the fuselage. 9. set the turbulence factors to 0. Therefore. 141 . For the Meridian. because of limited memory Viscous – Turbulent.
0 m/s in the ydirection. Residuals – 1x106 Monitor – Lift and drag Iterate until converged The mesh generation and FLUENT runs were repeated for several different angles of attack. Ref [19]. but the imported Meridian crosssection was imported at the correct angle of attack. but the values had to set. o Velocity Inlets – 61. For angles of attack the velocity components could be changed.7 m/s in the xdirection and 0. Solution – SIMPLE for the pressurevelocity coupling. Multiple angles of attack allow the designer to see the changes in the flow as the aircraft body changes orientation. 142 . The model did not converge doing this.5. and SecondOrder Upwind for the other discretizations. Boundary Conditions – The types of boundary conditions chosen in GAMBIT were the desired ones for this model. These were determined using the online FLUENT help. These were the values used by AAA when designing the aircraft. PRESTO! for the pressure discretization. To create the new mesh the same method as before was used. so a new mesh was created for each angle of attack chosen. o Pressure Outlet – 0 Pascals for the gauge pressure o Wall – Set the wall roughness height to 3x105 m and roughness constant to 0.
2573 3.8458 0. 5˚. At low angles of attack.1 Meridian FLUENT Results The angles of attack checked for the Meridian fuselage were 4˚. the lift produced by the fuselage can be neglected due the small value. allows the designer to compare this lift and drag to the lift and drag with the rest of the aircraft.9. Typically.658 4 0.0286 14.5. The overall magnitude of the fuselage lift is not extremely high. The down camber on the tail lowered the total pressure on the upper surface and also results in extra lift. A large amount of lift was produced by the large camber in the fuselage. 3.039 5 0. but when compared to the amount of drag the fuselage produces it is considered high. the Meridian fuselage will generate a large amount of lift. and 10˚. Most likely the actual numbers will be less than this.0917 7.5479 0. Using the wing area as a reference. but the Meridian‟s lift is too large to neglect.10 0 0.22 2 0. Table 52 shows the lift and drag found at all angles of attack. The lift and drag were determined in terms of wing area. Also.0414 13. 0˚.4030 0. the lift to drag ratio is close to what the design of the overall aircraft was.1080 0.35 m2. The large boundary layer that built up from this negative camber increases the drag.6456 0. 143 . The highly cambered wings will produce much more lift.0651 1.287 10 As seen from the lift to drag results. 2˚. Table 52: Meridian Fuselage CrossSection FLUENT Lift and Drag Cl_f Cd_f L/D (deg) 0. the 3D effects on the fuselage will change these numbers some.
velocity angle. and boundary layer can all easy be seen in the velocity magnitude. total pressure. 2˚. The downwash.The FLUENT results can best be shown in the form of pictures. From left to right. flow separation. 10˚. top to bottom the figures are in an order of 0˚. vorticity profiles. and 4˚. 144 . stream function. Figure 88 shows the velocity magnitude of the air around the Meridian. 5˚.
Figure 88: Meridian FLUENT Velocity Magnitude Profile for All Angle of Attacks It can be seen. as the angle of attack increase the boundary layer increases also. the boundary layer is the thickness of air that is not traveling at freestream velocity. Basically. The average thickness is directly related to the Reynolds Number of 145 .
as y goes to infinity. Displacement thickness is the theoretical distance that the wall would have to move outward. A Cartesian coordinate system is used for the boundary layer definitions. o “u” is a function of the variable “y”. Eq [28] shows an estimate for the average thickness of the boundary layer. Ref [20]. because u/U goes to 0 exponentially fast in the y direction. The upper boundary goes to infinity. The xcoordinate is parallel with the surface at each point. . Ref [20]. One method is defined as the distance from the wall where the velocity reaches 99 percent of the free stream velocity. The second method is the displacement thickness. The reason 100 percent of the free stream velocity was not chosen is that it is difficult to pinpoint the exact location where this happens. and the ycoordinate is normal to the surface. This method is somewhat arbitrary since any percentage could have been picked. It is appropriate to assume 99 percent is close enough for accuracy. Eq [30] 146 .the flow and the length of the surface. in a frictionless flow. and Eq [29] show the definition of Reynolds Number. Eq [28] Eq [29] There are three common measurements of the boundary layer thickness. u = 0.99U. to maintain the same mass flux as the authentic flow. Eq [30] shows the displacement thickness. and there is no arbitrariness in this method. *.
The size of the boundary layer is extremely thick at high angles of attack. As stated earlier the first method chooses an arbitrary number. loss due to the boundary layer. This is depicted by the blue area around the fuselage. This is seen as the blue area that does not follow the fuselage contour. . Ref [20]. For a typical aircraft wing the boundary layer thickness is on the order of one centimeter. The boundary layer flow stays attached to the fuselage and follows the contour of it. 5˚ and 10˚. Each method has its benefits and downfalls. This momentum thickness can be determined using Eq [31].The third method is the momentum thickness. and is defined as the amount of momentum. U2. but this is the easiest to use with experimental data. For example. Eq [31] The type of definition does not matter when it comes to boundary layer. The direction of the flow can be shown by viewing the velocity angle in Figure 89. The orientation of the angles is 0˚ to the right and 90˚ straight up. but instead starts flowing turbulently. 147 . At higher angles of attack. Again in this equation “u” is a function of the variable “y”. the second and third methods are difficult to develop an equation for the velocity. FLUENT depicts the velocity magnitude by varying the color. and large fuselage curvatures the flow separates completely from the fuselage. Boundary layers will build up in the ydirection as the flow travels over the body. u.
If the contour of the fuselage has smooth transitions and small curvatures the flow should follow the fuselage shape.Figure 89: Meridian FLUENT Velocity Angle Profile for All Angles of Attack A change in direction is expected around the fuselage. The changes in flow direction that are not expected are the ones at 148 .
velocity profiles along a curved surface. The flow stops at this point and begins to travel around that spot. The pressure gradient along the surface determines how the flow travels over a surface. but also the velocity.angles of attack 5˚ and 10˚. and this will cause the streamlines of the flow to converge and a decrease in pressure. Similar to the velocity magnitude. This separation bubble is where the flow can travel in many directions and possible have complete flow reversal. the flow changes direction abruptly at the nose cone of the aircraft. A smooth change is depicted with a slow transition in color. At higher angles of attack. and an adverse pressure gradient will decrease the free stream velocity. These pressure gradients not only affect the pressure. Separation occurs at these angles of attack near the empennage. there are unexpected turns in the flow. Downstream of the highest point the streamlines diverge and result in an increase in pressure. Figure 90 depicts the 149 . For example. and a sudden change with a dramatic change in color. Ref [20]. and 2˚ follows the contour of the fuselage pretty well. the velocity angle depicts its changes with changes in color. 0˚. The direction of the velocity for angles of attack 4˚. A favorable pressure gradient will be upstream of the highest point. There are a few sudden changes. An adverse pressure gradient is the opposite of the favorable pressure gradient. A favorable pressure gradient will increase the free stream velocity. 5˚ and 10˚. but this is expected when the flow encounters a disturbance in the path.
Where these two flows. Separation is a function of the flow geometry and whether the boundary layer 150 . Figure 91 shows a representation of the flow separating from an adverse pressure gradient and then flow reversal happening. Ref [20]. but it does not shed off the body. The backward flow will meet the forward flow at some point and begins to flow forward again. This is seen at both the higher angles of attack. At low Reynolds numbers. or in other words where the curvature of the velocity profile changes signs. a large steady vortex will form behind this separation point. A strong enough adverse pressure gradient can cause the flow to separate from the wall. If the flow continues with an adverse pressure gradient the flow will reverse directions.Figure 90: Velocity Profile in Boundary Layer with Favorable and Adverse Pressure Gradient (From Ref [20]) The inflection point is where the 2u/y2 = 0. backwards and forwards. meet is where the stress vanishes. This is where the flow swirls in an area.
To avoid separation and higher drag a structure should have a gradual change in size and shape. and therefore leads to separation happening quickly. [20]) The flow separation is very apparent at the high angles of attack with the velocity angles flow turning red. Figure 91: Velocity Profile with Flow Separation at Point S (dashed line u = 0. At the higher angles of attack the geometry changes more suddenly and this causes separation to happen more quickly. Figure 92 shows the total pressure in the flow. The dark red and dark blue are flow reversal points.is laminar or turbulent. From Ref. 151 . A blunt body or sudden change in shape can lead to a steep pressure gradient.
The colors move 152 . Certain angles of attack will affect the flow field pressure much more than others.Figure 92: Meridian FLUENT Total Pressure Profile for All Angles of Attack The total pressure is directly related to the flow velocity and the aircraft shape and angle of attack. This is seen at angles of attack of 10˚ and 4˚.
The flow path can be seen in Figure 93 showing the stream functions. These changes in color are a reduction in the total pressure due to the fuselage contour and flow velocity. Eq [32] At the 4˚ angle of attack the boundary layer on the lower surface is very thick. 153 . Stream functions have a constant value along a streamline. Total pressure is a function of the velocity squared. The total pressure is lower where the boundary layer is located and the where the flow separates from the body due to the small velocity. It is similar to the flow encountering a step. These stream functions can be used to analytically determine the velocity and its direction. This is due to the sudden change in shape on the cowling. This is one of the reasons for the very low lift to drag ratio at these two angles of attack. and this causes the flow to separate very quickly. Eq [32].away from the red color to a yellow or orange color.
and the between two streamlines is the volume of the flow between those streamlines. This can be thought of as the volume flowing through a streamtube bounded by those two 154 .Figure 93: Meridian FLUENT Stream Function Profile at All Angles of Attack Stream functions are defined to be constant along a stream line.
There is high vorticity around the abrupt curved spots on the fuselage. Ref [21]. Ref [20]. Circulation can be thought of as the strength of vorticity or vortex. it is seen that the stream function travels in a complete closed path around the empennage. Vorticity is defined as twice the angular velocity. but this model is 2D and has no depth. A stream function normally has units of kg/sm. Ref [21]. Higher angular velocity will produce more circulation around a finite area. 10˚. The rest of the flow travels over the top of this circulation. Eq [33] Eq [34] Eq [35] At the high angle of attack. To determine the velocity components from the stream function Eq [34] and Eq [35] should be used. Irrotational flow will have a magnitude of zero vorticity. The boundary layer also is a lighter shade of blue. This is found by calculating the average rotation rate of two perpendicular lines and then multiplying it by two. There is also a unitless form of stream function and is defined by Eq [33] (without the bar). Ref [20]. because the flow in this area has rotation. Using Stokes‟ theorem circulation can be written as Eq [36]. The units for stream function are kg/s. As seen from this equation circulation is related to the rotation of the flow. This is due to the flow separating from the fuselage and traveling in this closed path. This is why the dark blue area is very close to zero in Figure 94. Lift is related to circulation using 155 .streamlines.
Ref [6].Eq [37]. On most computational fluid dynamics programs circulation is used as a step to calculate the forces on a body. When a body is generating lift the circulation is finite. Eq [36] Eq [37] 156 . This is known as the KuttaJoukowski theorem. This means that the lift is directly related to the boundary layer vorticity.
The lift produced by a wing is a slightly manipulated form of Eq [37]. because it is in terms of per unit of span. that equation is multiplied by the amount 157 .Figure 94: Meridian FLUENT Vorticity Magnitude Profile at All Angles of Attack KuttaJoukowski Theorem is used to estimate the lift produced by an airfoil or wing. Therefore.
If a line of circulation. This is depicted in Figure 95. In order for this to happen in a uniform flow the upper surface velocity must increase and the lower surface velocity must decrease. for an airfoil to generate lift the lower surface must have higher pressure than the upper surface.6 Method Comparison Meridian Since time and resources did not allow for a 3D FLUENT model to be created the stability and control derivatives from FLUENT were not able to be determined. but rather the 158 . Ref [6] to help explain how circulation will produce lift. Figure 95: Circulation around an Airfoil Producing Lift (From Ref [6]) 9. Digging into the VORSTAB output files the theoretical fuselage lift and drag can be found. l‟. For example. This difference in pressures or velocities over a length will produce the lift. was found. is used to define an area of integration then circulation can be determined for this airfoil or wing section. This theorem explains why the fuselage generates lift. AAA does not calculate the lift of the fuselage. Lift and drag that was produced by the fuselage. in a 2D since.of span that generates lift.
006 0. the lift to drag ratio for FLUENT might be correct. The VORSTAB results shown in this table are for Models 2 and 3. The comparisons lift and drag values for the Meridian fuselage are shown in Table 53. because they are the only results that VORSTAB will generate fuselage data.598 0. This resulted in an extremely small lift.548 0.contribution of the wing and fuselage combination.029 14.009 2. The 3D effects will reduce the lift.004 0. This is due to the high downwash produced in 159 .010 Model 3 0. but also they will also help the flow to stay attached to the fuselage and avoid separation.005 1. Model 2 and Model 3 for the VORSTAB drag results are very high. Since AAA uses this combination the individual wing lift was subtracted from the wingfuselage lift to determine the individual fuselage lift.486 0.092 5. Model 1 calculated downwash so high that the pressure on the fuselage was too high to continue calculating the fuselage data. This would reduce the drag. but the individual values might be off. Table 53: Meridian Fuselage Lift and Drag Method Comparison = 0˚ = 5˚ = 5˚ and 4˚ (FLUENT) Method Cl Cd L/D Cl Cd L/D Cl Cd L/D FLUENT VORSTAB Model 2 0.606 0.119 0.003 0.007 0.975 0. The forth model did not have a fuselage present for calculations.001 0. and was negligible when compared to the VORSTAB and FLUENT results.460 0.09 0.007 0.072 0.108 0.403 0.001 1.019 0.827 0. Therefore.065 1.659 As seen from these results the FLUENT fuselage seems to produce much more lift and less drag than the other two methods.
160 . Fuselages do not typically produce much lift. Table 54 shows the comparison for the Meridian stability and control derivatives.by VORSTAB. can be compared to one another. Similar to the YAK54 and MantaHawk the stability and control derivatives from both methods. These VORSTAB lift to drag ratios are nowhere near the expected value. VORSTAB and AAA.
3304 0.1605 0.8522 7.1463 0.0000 0.6709 0.0565 0.0922 1.6179 13.2153 4.1083 0.1020 0.2263 0.1481 Model 1 0.0460 0.0123 0.5662 0.Table 54: Meridian AAA and VORSTAB Comparison Longitudinal Coefficients and Stability Derivatives Derivative (1/deg) CL CD Cm CLq Cmq Cl Clp Clr Cy Cyp Cyr Cn Cnp Cnr CLe CDe Cme Cla Clr Cya Cyr Cna Cnr AAA 5.8228 0.0117 0.5804 0.1229 0.0773 0.2913 0.7942 0.2606 0.0559 6.3681 0.0940 0.3450 0.1068 0.2625 0.0991 0.0583 0.1386 0.2268 34.1409 0.1099 0.1291 0. and the ones it could were done with inaccurate downwash.4789 0.1338 0.1065 0.3600 0.1682 8.0567 0.0974 0.0929 0.0479 Model 4 5.3217 0.1675 15.0253 0.0612 Model 1 produced such a high amount of downwash that the program could not calculate all the derivatives.1609 3.0948 VORSTAB Model 2 Model 3 7.5412 2.0226 0.1248 0.4038 7.1060 0.2804 1.5546 0.4149 0.1261 0.2091 0.5688 0.0000 0.7695 0.0748 0.0242 0.6207 4.3867 0.1648 0.2316 0.2960 0.6383 0.0776 0.2320 0.109 2.0178 0.0351 0.0040 0.7241 3.0858 0.2031 0.1031 0.2316 0.2920 0.603 17.6586 1.0670 0.0134 0.1767 8.1303 0.4115 1.1294 2.101 0.0117 1.2337 7.9313 0.0666 0.0688 0.2628 0.0312 0.1465 0.3308 0.973 0. Model 4 does not contain the fuselage so the results only show what the 161 .0692 0.
The same typical ranges that applied to the YAK54 and MantaHawk apply to the Meridian as well. 162 . Roskam. The VORSTAB results are unstable in many different modes as discussed earlier and the AAA results are stable in every mode. Roskam. Ref [5]. Ref [5] for what a conventional aircraft can expect the control derivatives to be approximately. These ranges are a function of Mach number. These ranges again were developed by Dr.wings and vtail produce in terms of stability and control derivatives. Table 55 shows the VORSTAB and AAA Meridian results compared to typical ranges from Dr.
134 0.724 3.012 0.025 0.0 to 4.024 0.404 7.097 0.123 0.852 7.102 0.0 to 0.077 0.086 0.823 0.0 to 1.177 8.415 0.234 15.125 0.023 0.6 Negligible 0.296 0.232 0.094 0.330 0.0 0.04 to 0.291 0.165 0.931 0.061 AAA 5.147 0.130 0.0 to 30.103 0.292 0.5 to 0.331 0.148 Typical Ranges Dr.0 0.322 0.161 3.387 0.0 4.069 0.263 0.04 Negligible 0.621 4.0 0.4 0.671 0.360 0.012 0.035 0.1 to 4.770 0.226 0.046 0.15 to 0 Within Range (Yes/No) Model Model Model AAA 2 3 4 Yes Yes Yes Yes No No Yes Yes Yes Yes No No Yes No Yes Yes Yes Yes No Yes Yes Yes No Yes Yes Yes No No Yes Yes Yes Yes No No Yes No Yes Yes Yes Yes No Yes Yes Yes No Yes Yes Yes Yes Yes Yes Yes Yes Yes No No Yes No Yes Yes Yes Yes No Yes Yes Yes No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes No Yes No Yes Yes Yes Yes No Yes No 163 .0 to 0.107 0.031 0.069 0.1 to 2.0 to 0 0.0 to 1.092 1.555 0.168 6.368 0.261 0.569 0. Roskam 1.479 0.0 to 0.018 0.5 0.0 to 2.0 0.2 0.0 0.3 to 0.618 13.1 0.004 0.067 0.078 0.209 0.0 to 8.08 to 0.232 0.794 0.075 0.566 0.058 0.638 0.110 0.126 0.101 0.013 0.1 to 0.012 1.263 0.0 0.8 0.067 0.97 0.146 0.108 0.141 0.139 0.345 0.0 90.106 0.0 to 1.107 0.Table 55: Meridian VORSTAB and AAA Stability and Control Derivatives Typical Ranges Stability and Control Derivatives Model Model Model 2 3 4 CL CD Cm CLq Cmq Cl Clp Clr Cy Cyp Cyr Cn Cnp Cnr CLe CDe Cme Cla Clr Cya Cyr Cna Cnr 7.0 to 0.0 to 4.000 0.6 0.099 0.057 0.129 0.08 0.048 5.0 0.232 0.8 0.168 8.58 0.057 0.056 7.
This does not necessarily mean that the aircraft is unstable. The fuselage model was changed many times until results that were closer to the desired results were achieved. 164 . This does not mean that VORSTAB was correct in that case and that AAA was incorrect. From these results it can be determined that VORSTAB is most affective with less complex aircraft. It is very difficult to determine if these programs are producing the correct results without comparing their results with actual fight tests. because many times the meshes of a FLUENT model have to be remade until the results are what was desired or expected. but rather should be used on aircraft that have a simpler design. Basically it was very similar to guessing and checking. In some cases the AAA results do not fall into the expected range.As seen from the table not all of the control derivatives fall within the typical ranges. The Meridian was very difficult to model and achieve accurate results. This does not mean that VORSTAB is not a high fidelity tool or inaccurate. but the VORSTAB results do. The Meridian might just have higher values in some derivatives than a typical aircraft has. After that comparison one can draw a better conclusion about the validity of the results. This is not necessarily bad. The rudder input is a symmetrical deflection of the vtail and unsymmetrical deflection is an elevator input.
A downfall of VORSTAB is that it is very difficult to make a VORSTAB model and requires someone with experience or guidance. AAA overestimates estimates the drag and this was one of the main reason that a VORSTAB model was created for all of the vehicles. but opposite. These improvements from VORSTAB can be significant in creating a 165 . was lower than AAA predicted. Some of the derivatives from VORSTAB are obvious that they are wrong and these should be ignored.10 Conclusions and Recommendations After analyzing the three aircraft with AAA and VORSTAB conclusions about both software programs were drawn. VORSTAB uses a high fidelity method to derive the stability and control derivatives. AAA is a valid method for quickly estimating the stability and control derivatives. A downfall to the program is that a low fidelity method is used to derive the derivatives. It is known that a positive and negative aileron deflection should result in the same magnitude for this derivative. These are the obvious downfalls and benefits to both programs. CDo. VORSTAB showed that the drag. as expected. From the flight test data. even at a low Reynolds number. The VORSTAB model also takes a considerable longer amount of time to create than AAA. It is a rather affordable software program and easy to use. the yawing moment due to aileron deflection has results that are not symmetric. This is a major advantage of VORSTAB is that the drag estimation is closer to the actual drag. VORSTAB estimated many derivatives better than the AAA program did. For example. for positive and negative aileron deflections. For low Reynolds number. For the YAK54.
The MantaHawk wrecked early into the first flight test and hardly any data was collected. VORSTAB predicted a lower drag than AAA for the MantaHawk. VORSTAB predicted that the MantaHawk was stable in every mode but the pitching moment stability. The complex shape of the fuselage. Even if VORSTAB was incorrect about the instability it would give the designers an idea of something that should be investigated. The instability mode can be tested in the flight tests. VORSTAB was found to be completely invalid. From these results. Therefore. it was determined that this software should only be used on aircraft that 166 . This can have a significant effect on the control of the aircraft. AAA did not predict this problem and said the vehicle should be stable in all modes. and having the pilot aware of possible instability will dramatically improve the safety.valid YAK54 model. both software programs are a valid engineering tool and both should be used for deriving the derivatives. a recommendation for VORSTAB would be to create a model for preliminary design and use it to determine if there is a possible instability issue. From the small amount of data and explanation from the flight test team the vehicle wrecked from a pitching moment error. Similar to the YAK54. and highly cambered wings were difficult for VORSTAB to handle. Then the flight test team can use precaution when design the flight tests and flying the aircraft. Without flight test it cannot be determined if VORSTAB or AAA predicted the drag best. Most of AAA‟s data is based on conventional aircraft and not a flying wing. From the Meridian results. vtail configuration. It also estimated all of the derivatives to be much lower than AAA results. Since VORSTAB worked for both of the other aircraft.
With improvements made to the aircraft the drag has reduced 167 . a large amount of downwash was produced. but there was not a large camber. Then modifications can be made to this VORSTAB model until the expected results are achieved. A wing and tail model should be created to model the drag if there is a complex fuselage body. a trip strip could possible be added to the empennage in order to reenergize the flow and prevent separation. This was the source of most of VORSTAB‟s problems.have an extremely conventional configuration and a smaller amount of camber. Since this was a 2D model the 3D effects will help prevent the flow from separating. The drag was determined to be much lower than VORSTAB was estimating and the flight tests were showing. FLUENT was used to determine if there was a possible downwash problem over the fuselage and if the fuselage was producing a large amount of drag. The high drag was a result on having an unpainted surface. At high angle of attack the flow separated from the fuselage and even began to have complete flow reversal at 10˚ angle of attack. Tufting should be added to the empennage and flight tested to determine if there is flow separation and turbulent flow. The MantaHawk did not have a conventional shape. These trip strips can later be removed if it is found that the separation is not a problem. Once the flight tests have been repeated several times and the stability and control derivatives have been determined a VORSTAB model should be created. and more. the highly cambered fuselage. cowling with large gaps. After this investigation. From the VORSTAB output file. This would give the designer a general idea for ways to improve the models in the future. no fairing.
Then the fuselage should be created. so it should be created in steps. If Tgrid is available it should be used to import the complete model and create the mesh. This model will be difficult to create. It is a very complex geometry so the CAD model should be used to import the fuselage. The wings and tails will not be that difficult to create and should be done first.dramatically. All of these recommendations will help with future and current design projects. 168 . It is recommended that a complete 3D FLUENT model be created of the Meridian.
2010]. Thrust and Power Characteristics. [13] Donovan..” Fluent Inc..fit.. “Modeling and Sensitivity Analysis of the Meridian Unmanned Aerial Vehicle. Jager.. KS. J. “A and B Matrix Construction Using Roskam Method.” DARcorporation. S.” Department of Aerospace Engineering at the University of Kansas. pp... 2007. J..” AIAA20086900.” AIAA Conference Submitted. Technical Report CReSIS TR 123. E. “Development of a Pilot Training Platform for UAVs Using a 6DOF Nonlinear Model with Flight Test Validation. J. “Airplane Aerodynamics and Performance. C. 169 .” Linearized Subsonic Flow.3/help/html/ug/node992. [19] Anderson. E..” Department of Aerospace Engineering at the University of Kansas. Sweeten. Burns.11 References [1] Roskam. and Jager. New York. “Fundamentals of Aerodynamics. Lawrence.. S. R. KS.” MatLab Code. P.” Fall 2009. “Test and Evaluation of the Piccolo II Autopilot System on a OneThird Scale Yak54. Keshmiri. [8] Jager.” DARcorporation. April 2008.2).” DARcorporation Website. 404437. and Cohen. 3rd ed.. “AE 721: Interim Final Review of a Small Scale UAV. [2] Lan. [18] Kundu. 2004. M. W.” AIAA20086368. Revised 2006. “Modern Compressible Flow. 2008. UT. [6] Lan. “VORSTABA Computer Program for Calculating LateralDirectional Stability Derivatives with Vortex Flow Effect. R. “The Meridian Critical Design Review. 2007... S. “Modeling and Simulation of the Yak54 Scaled Unmanned Aerial Vehicle Using Parameter and System Identification. [http://www. Lawrence... [14] Leong. [3] Roskam. 2003. Burlington..” DARcorporation..com/Software/AAA/ . [9] Keshmiri.. Lawrence. January 1985. and Grorud. KS.. [10] Leong. [5] Anderson. December 2009. D. D. J... K. J. Tom. New York. “The Meridian UAV Flight Performance Analysis Using Analytical and Experimental Data..” AIAA20091899.edu/itresources/manuals/fluent6.” June 25. [16] Keshmiri. [http://my. New York.” The University of Kansas Aerospace Engineering Department..” 4rd ed. [11] Arnold. September 2007... B.. [15] Royer.darcorp. [7] Lan. I. H.. R.htm]... [17] “Spatial Discretization. and Roskam. D. “User‟s Manual for VORSTAB Code (Version 3. 2003.. E. E. “Airplane Design Part VI: Preliminary Calculation of Aerodynamic. R. [12] Sweeten. C. Leong. B. and Hale. D. pp. 324333.” 4rd ed. 2003... “The Meridian System Identification Flight Tests in Dugway. Massachusetts.” NASA CR172501. E.. September 2006. J. R. August 2008. “Fluid Mechanics. “Airplane Flight Dynamics and Automatic Flight Controls. Retrieved 6 January. Royer. E. New York. [4] “Advanced Aircraft Analysis. D.
.. H. and Altman. B. [23] Chudoba. A. Opportunities and Challenges. D. L. “A Generic Stability and Control Methodology For Novel Aircraft Conceptual Design. 170 . D. [21] Malmuth. N. and Smith. B. “The Lockheed SR71 Blackbird – A Senior Capstone ReEngineering Experience.” AIAA20044973.[20] Mixon.” AIAA20035388..... and Chudoba. “Wavelike Characteristics of Low Reynolds Number Aerodynamics. [22] Alford. B. “ Theoretical Aerodynamics in Today‟s Real World.” AIAA20055059.” AIAA2007698.
For the input file of the Meridian Model 3 go to Meridian3_input and for the output file go to Meridian3_output. For the input file of the Meridian Model 2 go to Meridian2_input and for the output file go to Meridian2_output. For the input file of the MantaHawk go to MantaHawk_input and for the output file go to MantaHawk_output. For the input file of the Meridian Model 1 go to Meridian1_input and for the output file go to Meridian1_output. For the input file of the Meridian Model 4 go to Meridian4_input and for the output file go to Meridian4_output. 171 .Appendix A The following files are posted with my thesis on the KU website. For the input file of the YAK54 go to YAK_input and for the output file go to YAK_output.
The program uses the control derivatives found from VORSTAB.0104 .2024 0.0.3397 .0.2101 0.4499 . Ref [22]. The following A and B matrices were determined from this method for both longitudinal and lateraldirectional modes.5780 172 .0.3521 .0000 .0.4156 29.6102 0 .0.0.0020 .0000 0 0 .0.0151 1.9963 . It is developed in the form of x˙=Ax+Bu.0011 0.0002 B _ long .9979 .0338 0 A _ lat .2657 A _ long 0.0.1.0.0042 .0.Appendix B YAK 54: The linear state space model is developed using the Roskam method and a program written by Edmond Leong.0.0621 3.1788 0 B _ lat .1664 1.0.0.0592 0 0.0000 .0090 0 0 .2.9995 .0.0.7468 0.3604 0 0.0208 0 0 1.0. 0 6.0218 0 .
0105 Spiral 0.133i 0.60i 0.87+1.0 0.205+0.839 0.0473+1.281 ? 0.60i AAA 0.63 ? 0.0 0.7 0.323i AAA 1 0.149 Roll 0.24 Phugoid 0.63 ? 0.0358 1. Ref [13]. These two modes are Phugoid and Short Period.209 1.342+1.65 Dutch Roll 1. Table 57 shows the lateraldirectional mode analysis summary and also compares it to the results that AAA control derivatives produce.04731.985 10 Short Period 9.3i 0.0 0.871.22+65.12e5 1.73i Table 57: YAK54 LateralDirectional Modes Eigenvalues Damping Frequency Mode (rad/sec) VORSTAB 1.0599 Roll 16.24 Phugoid 0.133i 0.32 Dutch Roll 0.323i 0.18 6.2050.3i 173 . Table 56: YAK54 Longitudinal Modes Eigenvalues Damping Frequency Mode (rad/sec) VORSTAB 1.2265.32 Dutch Roll 0. Table 56 shows the longitudinal mode analysis summary and also compares it to the results that AAA control derivatives produce.209 1.18 6.0105 1 0.426 ? 0.0 2.0358 1.3421.The longitudinal dynamics should have two complex conjugate roots.839 0. Ref [13].281 1.65 Dutch Roll 1.12e5 Spiral 2.985 10 Short Period 9.149 0.73i 0.426 0.
0. Ref [13]. It is developed in the form of x˙=Ax+Bu.0926 A _ long . . Several flight tests were conducted to perform system identification on the YAK54.1495 43.0.0137 . The program uses the control derivatives found from VORSTAB.32. Ref [22].0. MantaHawk: The linear state space model is developed using the Roskam method and a program written by Edmond Leong. The coupling of the two programs does not work.0063 .1.2486 0 0 .0176 B _ long .575 and 0.0000 0 174 .0.1740 0. The only two modes tested were the Dutch Roll mode and Phugoid mode.0.9997 0 . These tests resulted in a damping ratio for the Phugoid mode of 0. The following A and B matrices were determined from this method for both longitudinal and lateraldirectional modes.0.0.0521 .1245 0 0 . This is most likely due to the fact that the state space model was created with a mixture of the VORSTAB derivatives and the AAA derivatives.0425 0 1.303 for the Dutch Roll.It is obvious from the results that VORSTAB is not correct.6531 .0. VORSTAB does not give every derivative necessary to make the model and this is why some of the AAA derivatives were used.
0010 0.0.04670.0000 .0.0120 0 0 0 0.2078 0 0 B _ lat 0. summary. Ref [&^%].0.0875 The longitudinal dynamics has only one complex conjugate root and two different real roots. There are typically two roots that are defined as Phugoid and Short Period modes.1791 .0666 0.699i 1.2723 . When the center of gravity is ahead of the neutral point it will cause Phugoid and Short Period roots to approach real roots (xaxis).0001 . Table 58 shows the longitudinal mode analysis summary.0001 0 .0035 2.0010 .0031 .0.0467+0.0666 0.701 ? ? ? ? Table 59 shows the lateraldirectional mode analysis 175 .701 0.0 0.0.0.449 0.0 1. but it still does not explain the two different real roots.2078 2.699i 0. A _ lat .9999 0 0.0. Table 58: MantaHawk Longitudinal Modes Eigenvalues Damping Frequency (rad/sec) Mode VORSTAB 0.0714 0.0714 0.449 0.6085 0.0.1728 1.0875 .0.
0016 .1770.0000 0 1.27e5 Spiral 4.0008 B _ long .291 0. The program uses the control derivatives found from VORSTAB.Table 59: MantaHawk LateralDirectional Modes Eigenvalues Damping Frequency Mode (rad/sec) VORSTAB 1.0.0. The vehicle did not go under enough flight tests to verify any of the results.9996 0 A _ long . these results are not accurate since a mixture of both AAA and VORSTAB derivatives were used to create the state space model.581i 0.0 0.0673 0.0.7443 .291 0. It is developed in the form of x˙=Ax+Bu.533 0. Meridian: The linear state space model is developed using the Roskam method and a program written by Edmond Leong.0.607 Dutch Roll 0.1899 0 176 .0 4.581i Again.533 Roll 0.1740 0. Model 2.0. The following A and B matrices were determined from this method for both longitudinal and lateraldirectional modes of the Meridian model with the large empennage. Ref [22].27e5 1.177+0.0014 . 0 .0.5584 .32.0232 6.4761 .0383 0 0 0 1.607 Dutch Roll 0.0.
05180. one of the roots is unstable.736 0. Ref [18].0004 .9999 .0. .3085 0 0 0.0518+0.0.435 0.1320 .116 0.58 7.0000 A _ lat .241 0.73i 0.0423 0 . Also.116 0.093+0. Table 60 shows the longitudinal mode analysis summary and also compares it to the results that AAA control derivatives produce.255 0.0930.0039 0 .215 0.0.58 Phugoid Phugoid Short Period Short Period 177 .435 0.208i 0.126 0.0545 1. since its eigenvalue is positive.0. Ref [18].1589 0. Table 61 shows the lateraldirectional mode analysis summary and also compares it to the results that AAA control derivatives produce.0094 0.1066 The longitudinal dynamics has only two complex conjugate roots.73i AAA 0.7485 0.0.0.0001 .126 0.0.0333 0 0 B _ lat 0. Table 60: Meridian Model 2 Longitudinal Modes Eigenvalues Damping Frequency (rad/sec) Mode VORSTAB 0.0.0060 0 0.215 0.241 0.1876 .255 7.208i 0.736 0.0.0010 0.
0. 0 .0.0.0 1.0232 41.0913 .0016 .232 0.0. Model 3.0.0004 0.0.232 0.7478 0.0.0.1589 0.1311 .0000 0 0 A _ lat .1461 .1317 .0.2938 .0.0001 0.484 109 sec 0.3381 .144 0.0007 B _ long .14 414 Dutch Roll Dutch Roll Spiral Roll Dutch Roll Dutch Roll Removing the camber.0.0000 0 .9996 0 A _ long .103+0.0058 178 .1769 0 0 .0316 0.0668 0.213 1 1 0.Table 61: Meridian Model 2 LateralDirectional Modes Eigenvalues Damping Frequency (rad/sec) Mode VORSTAB 0.154 sec 4.213 0.0002 0 0.473i AAA 1.144 0.9999 .2.103+0.0.0 0.0.484 0.0.32.0010 0.1159 . resulted in the following matrix A and B from the VORSTAB results.0383 0 0 0 1.0349 0 0 B _ lat 0.0546 1.473i 0.0.0030 0.0014 .0433 0 .1740 0.0316 0.
0586 0. The lateraldirectional mode has 3 stable roots.The longitudinal dynamics has two complex conjugate roots.196 0. Ref [18].601i 0.602 0.602 0.116 0.58 7. Table 62: Meridian Model 3 Longitudinal Modes Eigenvalues Damping Natural Frequency Mode VORSTAB 5. Table 63 shows the lateraldirectional mode analysis summary and also compares it to the results that AAA control derivatives produce.255 7.0586 0. when the spiral mode is usually unstable. Table 62 shows the longitudinal mode analysis summary and also compares it to the results that AAA control derivatives produce. These two roots are Phugoid mode and Short Period mode.601i AAA 0.435 0.03530.196i 5.58 Phugoid Phugoid Short Period Short Period 179 .116 0.196i 0.0287 0.255 0.435 0. Ref [18].0353+0.0287 0.64e30.64e3+0.196 0.
5973 and a natural frequency of 3.6723 rad/sec. AAA was closer to the flight test data than the VORSTAB results.436i AAA 1.144 0. Also. 180 .246 0. so system identification could be performed. The Short Period had a damping ratio of 0.154 sec 4.0 1.257 0.1110.144 0. Flight tests were performed in Dugway.45 109 sec 0.436i 0. like the YAK54. Similar to the Dutch Roll the model without the camber had better results. as expected the results from the model without camber had results closer to the flight test data.14 Spiral Roll Dutch Roll Dutch Roll For the same reason as the other two aircrafts this state space models are not accurate. For complex geometry it would be wise to use both AAA and VORSTAB as a means to determine the stability and control derivatives.45 0.0245 0. Aircraft that have noncomplex or standard geometry.111+0.257 0.14 4.Table 63: Meridian Model 3 LateralDirectional Modes Eigenvalues Damping Frequency (rad/sec) Mode VORSTAB 0.0 0. From the flight tests and analysis of that data from the flight test team the Dutch Roll damping ratio of 0. During these flight tests the Dutch Roll and Short Period modes were initiated. UT. All of the flight test results and analysis were taken from Ref [23]. VORSTAB‟s results should be more accurate.5341 rad/sec.2163 and natural frequency of 3.0245 0.246 1 1 0.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue listening from where you left off, or restart the preview.