You are on page 1of 11

SIMULATION OF THE F/A-18D "FALLING LEAF"

Paul T. Jaramillo* and John Ralston†


Bihrle Applied Research, Inc.
Hampton, Virginia 23666

Abstract I. Introduction
This paper discusses a phenomena which has been The purpose of this paper is to describe the key features
encountered on several of the today's strike and fighter of a type of motion called the "falling leaf" and to
aircraft. The motion can be described as violent, with delineate the basic requirements for its simulation. At
angle-of-attack and sideslip angles traversing as much present, there exist few works on this matter. The
as 100 deg within a few seconds. The motion, unlike a development of an operational definition of a new
spin, does not involve a continuous unidirectional problem or phenomena is the first step to understanding
heading angle change, but rather, displays an in-phase it and solving it. In the case of the falling leaf, it
roll and yaw oscillation about a zero or small non-zero appears that it may be just one of several closely related
mean. This motion is produced by a complex interplay motions in a spectrum of out-of-control motions
between dynamic and aerodynamic moments, and running from steady spins on one extreme to the falling
involves strong all-axis coupling. The details of this leaf on the other. Furthermore, preliminary
phenomena are discussed and six degrees-of-freedom investigations show that the recovery techniques may
simulations of the motion presented. The simulations vary considerably across this spectrum and therefore,
have been evaluated against flight data. such a problem definition is of practical importance.

Nomenclature II. System Modeling


b wingspan Aircraft flight simulation involves the numerical
c mean aerodynamic chord calculation of the state of a dynamical system which is
ei, quaternion components acted upon by external forcing functions, namely the
F vector of aerodynamic and propulsive force aerodynamic, gravitational and propulsive forces and
Fi aerodynamic and propulsive force component moments. As such, a typical flight simulation model
g gravity has two major components: a set of relations which
Iii moments and products of inertia describe its dynamical and kinematical aspects, and a
L rolling moment set of relations which describe its aerodynamical
M pitching moment aspects. In addition, modern aircraft often contain
m mass complex control systems which, in effect, forms a third
N yawing moment major component of the numerical model. In the
q% dynamic pressure following sub-sections, these three components will be
P roll rate briefly described.
Q pitch rate
R yaw rate 2.1 Simulation Model
U x axis velocity component 2.1.1 Aerodynamic Relations
V y-axis velocity component
The aerodynamic database used in this study is
V translational velocity vector
essentially a subset of the F-18C/D simulation data base
VAS vertical airspeed
version 3.0, currently in use at the Manned Flight
W z-axis velocity component
Simulation (MFS) Center, Patuxent Naval Air Station.
α angle-of-attack In this study, the high Mach functionality, for some of
β sideslip the aerodynamic components with more than three
Ω rotational velocity vector dimensions, was omitted because of the low speed
_____________________________________ flight regime examined as well as the need to reduce
*Research Engineer, Member AIAA the database size for manipulation in the parametric
†Engineering Manager study. The six aerodynamic components and their
This paper is declared a work of the U.S. Government and is not functionalities are depicted in Tables 1 and 2. The
subject to copyright © Protection in the United States. database itself is the evolution of an original
McDonnell Douglas F-18 simulation that has been the the significance of dynamic data mechanization
subject of much revision and validation since the first schemes, wherein methods of combining wind axes and
comparisons with flight were initiated at MFS in the body axes damping terms in the simulation were
1980’s.1-7 The primary goals of these efforts have been evaluated against a wide range of flight test data for
to improve the database’s low to mid speed range their ability to model dynamic motions.8, 9 The
fidelity, through the use of flight and wind tunnel data, mechanization method recommended as a result of this
although many other flight regimes and store study, a technique developed by Kalviste, was
configurations were also examined. Structurally, the subsequently used in the simulation of the motions
model was unified to make the original fragmented described in the following discussion.10, 11
database continuous in angle-of-attack from -90° to During the course of the work, some adjustments were
+90°, and sideslip effects through 30° were added to made to more closely reproduce falling leaf motions.
most of the increments. Further, rotary damping terms, These changes primarily affected the high-angle-of-
were also added to the simulation database. This data attack pitching moment characteristics but some static
set was recently used in a study investigating coefficients were estimated to 45 degrees of sideslip.
Table 1. Aerodynamic Force Increments

Increment Functionality Multiplier Remarks


DRAG
CD basic α, Mach
CD fs CL, Mach full scale correction
DCD tf Mach trainer canopy
DCD clt Mach center-line tank
DCD int Mach interdiction loading
DCN int α DCD int = DCD int + DCN int*SIN(α)
DCD lef α, Mach δlef leading edge flap
DCD tef α, Mach δtef trailing edge flap
DCD dhl, dhr α, δhl & δhr separate DCD dh = (DCD dhl + DCD dhr) / 2
DCN rot α, Sign(β)*Ωb/2V, |β| DCD rot*SIN(α)
DCN q α, Mach Qosc * c/2V DCD q = DCN q*SIN(α)
SIDE FORCE
CY basic α, β
DCY tf α, |β| Sign(β)
DCY clt α, |β| Sign(β)
DCY int α, |β| Sign(β)
DCY lef α, |β| Sign(β)*δlef / 25
DCY tef α, |β|, Mach Sign(β)*δtef / 20
DCY dal, dar α, δal & δar separate DCY ddf = (DCY dal - DCY dar)*2
DCY dr α, |δr|, β*Sign(δr) Sign(δr)
DCY ddh α, δh δdh δh = (δhl + δhr)/2, δdh = δhl - δhr
DCY rot α, Sign(β)*Ωb/2V, |β| Sign(β)
DCY p α, Mach Posc * b/2V
DCY r α, Mach Rosc * b/2V
LIFT
CL basic α, Mach
CL beta α, |β|
DCN int α DCL int = DCN int*COS(α)
DCL lef α, Mach δlef
DCL tef α, Mach δtef
DCL dal, dar α, δal & δar separate DCL ddf = DCL dal + DCL dar
DCL dhl, dhr α, δhl & δhr separate DCL dh = (DCL dal + DCL dar)/2
DCN rot α, Sign(β)*Ωb/2V, |β| DCL rot = DCN rot*COS(α)
DCL ad α, Mach α dot*c/2v alpha-dot
DCN q α, Mach Qosc * c/2V DCL q = DCN q*COS(α)

2
Table 2. Aerodynamic Moment Increments
Increment Functionality Multiplier Remarks
ROLLING MOMENT
Cl_basic α, β
DCl_off α static offset (non-zero at zero beta)
DCl_tf α, |β| Sign(β)
DCl_aim9 α Kaim9 (= -1,0,1) -1=Rt on, 1=Lt on, 0=both on/off
DCl_clt α, |β| Sign(β)
DCl_int α, |β| Sign(β)
DCl_ibpyl α β
DCl_lef α, |β| Sign(β)*(1-δlef / 34)
DCl_tef α, |β|, Mach Sign(β)*(1-δtef / 20)
DCl_dal, dar α, δal & δar separate DCl_ddf = DCl_dal - DCl_dar
DCl_da_b α, β, |δdf| Sign(δdf) δdf = ( δal - δar ) / 2
DCl_dr α, |δr|, β*Sign(δr) Sign(δr)
DCl_ddh α, δh, β δdh δh = (δhl + δhr)/2, δdh = δhl - δhr
DCl_rot α, Sign(β)*Ωb/2V, |β| Sign(β)
DCl_p α, Mach Posc * b/2V
DCl_r α, Mach Rosc * b/2V
PITCHING MOMENT
Cm_basic α, Mach
DCm_beta α, |β|
DCm_clt α
DCm_int α
DCm_lef α, Mach δlef
DCm_tef α, Mach δtef
DCm_dal, dar α, δal & δar separate DCm_ddf = DCm_dal + DCm_dar
DCm_dr α, |δr| Sign(δr)
DCm_dhl, dhr α, δhl & δhr separate δdh DCm_dh=(DCm_dhl + DCm_dhr)/2
DCm_rot α, Sign(β)*Ωb/2V, |β|
DCm_ad α, Mach α_dot*c/2v
DCm_q α, Mach Qosc*c/2V
YAWING MOMENT
Cn_basic α, β
DCn_off α
DCn_tf α, |β| Sign(β)
DCn_aim9 α Kaim9 (= -1,0,1) -1=Rt on, 1=Lt on, 0=both on/off
DCn_clt α, |β| Sign(β)
DCn_int α, |β| Sign(β)
DCn_ibpyl α β
DCn_lef α, |β| Sign(β)*(1-δlef / 34)
DCn_tef α, |β|, Mach Sign(β)*(1-δtef / 20)
DCn_dal, dar α, δal or δar DCn_ddf = DCl_dal - DCl_dar
DCn_da_b α, β, |δdf| Sign(δdf) δdf = ( δal - δar ) / 2
DCn_dr α, |δr|, β*Sign(δr) Sign(δr)
DCn_ddh α, δh, β δdh δh = (δhl + δhr)/2, δdh = δhl - δhr
DCn_rot α, Sign(β)*Ωb/2V, |β| Sign(β)
DCn_p α, Mach Posc * b/2V
DCn_r α, Mach Rosc * b/2V

3
2.1.2 Automatic Flight Control System other kinematical schemes for large-angle simulations
The flight control system used in the following full-up because of their ability to perform all possible aircraft
simulation examples shown in this paper, was built to orientations in a singularity free manner. This is in
represent the up and away inner loop of the V8.3.3 F- contrast to the popular Euler angle schemes which have
18 control law mechanization.12 Because of the flight numerical difficulties when the aircraft’s pitch angle
regime examined, the outer loop functions, as well as reaches plus or minus 90 degrees as in a vertical climb
the power approach logic and higher order structural or dive.
filters, were not modeled in this study. Quaternions, of course, have their own share of
nuances, not the least of which is a general lack of
2.1.3 Dynamic and Kinematic Relations usefulness of the quaternion components at the physical
level. The simulation software provides relations
The dynamic and kinematic relations used in the six
which allow the Euler angles to be calculated from the
degrees-of-freedom simulation are given in Eqs. (1)
quaternion components. The second nuance is of a
through (6). These are employed in a fourth-order
mathematical nature. In particular, since the attitude
numerical integration scheme in order to propagate the
equations employ four equations to calculate only three
motion forward in time. The first set of equations,
degrees-of-freedom, a constraint relationship is
commonly referred to as the force equations, utilize
implied. Equation (4) provides this relationship. In
quaternions to resolve the gravitational vector into the
particular, Eqs. (3) must be numerically propagated in
aircraft-carried axes. The forces, Fi, contain the
such a manner so as to preserve the relationship
aerodynamic and propulsive components. Mass was
between the quaternion components given by Eq. (4).
always assumed to remain constant on a given
Several techniques have been developed to accomplish
simulation run.
this and in this work a constraint of the algebraic type
& = FX + 2 g (e e − e e ) + R V − Q W was employed.13
U 1 3 0 2
m
e& 0  e 0 − e1 − e2 − e3   0 
& = FY + 2 g (e e + e e ) + P W − R U (1)
V  e&   − e3 e 2   P 
2 3 0 1  1 1  e1 e0
m &  =   (3)
e 2  2 e 2 e3 e0 − e1  Q 
& = FZ + g {1 − 2 (e 2 + e 2 ) } + Q U − P V
W  
m
1 2 e& 3  e3 − e2 e1 e 0  R 

The moment equations are given in Eqs. (2) below. e 20 + e12 + e 22 + e32 = 1 (4)
These equations do not account for any gyroscopic
effects associated with the engines, mass shifting Finally, the navigation equations given in Eq. (5) allow
effects due to fuel sloshing or mass loss due to fuel the calculation of the trajectory of the aircraft through
consumption. the atmosphere. For the present work, only the last
equation was utilized since the numerical model is not
 L I N  I xx − I yy  I xz affected by the aircraft’s spatial position with the
P& =  + xz  +  1 +  PQ
 I xx I zz   I zz  I xx exception of its altitude. The third component is
therefore modified as in Eq. (6) in order to provide the
 I yy − I zz I 2xz   I xx I zz necessary information for the standard atmosphere
+  −  Q R 
2
 I xx I xx I zz   I xx I zz − I xz model.

{
& = M + (I − I ) P R + R2 − P2 I
Q zz xx ( ) } xz
1 (2)
I yy
X
&   1 −2 (e2 + e2 ) 2(e e − e e ) 2(e e + e e )U
&  
2 3 1 2 0 3 1 3 0 2
 
Y = 2(e1e2 + e0e3 ) 1 − 2 (e1 + e3 ) 2(e2e3 − e0e1 )V (5)
2 2

  I − I yy Z
&  2  
I 2xz    2(e1e3 − e0e2 ) 2(e2e3 + e0e1 ) 1 − 2 (e1 + e2 )W
2
& =  N + I xz L  +  xx
R + P Q
 I xx I zz 
 I zz I xx   I zz
& = −Z
&
H (6)
 I yy − I zz I  I xx I zz
+ − 1 xz Q R 
 I xx  I zz  I xx I zz − I 2xz III. OVERVIEW OF THE FALLING LEAF
The attitude equations shown in Eqs. (3) are necessary 3.1 General Characteristics of Falling Leaf Motions
to update the quaternion components utilized in Eqs. Table 3 illustrates some general kinematical
(1) and (5). The use of quaternions is preferred over characteristics of the falling leaf in relation to other out-

4
of-control motions. The first row of comments in the Table 4 summarizes the main characteristics of falling
table compares the most prominent rotational velocity leaf motions in terms of some key variables. In
components of the falling leaf in relation to steady and general, the motions can be divided into two general
oscillatory spins. The second row compares the categories: slow falling leaf motions and fast falling
translational velocity vector through the aerodynamic leaf motions. The primary differences in going from
angles. The fact that many of the falling leaf motions the slow to the fast motions (darker highlight) can be
tend to look like oscillatory spins has led some to seen in the movement of the angle-of-attack trace to
naming these motions as low- or intermediate- mode higher values, the biasing of the yaw rate and a
oscillatory spins. Although this view is certainly decrease in the period of the motion, i.e. increase in the
acceptable, distinctions between these motions are frequency of oscillation. Simulations indicate that fast
important, because, unlike oscillatory spins experienced falling leaf motions do not divide along such distinct
by this aircraft, the motions discussed in this paper do borders but rather occur in a spectrum which primarily
not share the characteristic of being relatively easy to involves further increases in the minimum angle-of-
recover from. In fact, due to the yaw rate behavior attack of the motion while maintaining the same angle-
during the falling leaf, spin arrows displayed while in of-attack range, i.e. approximately 60 degrees between
spin recovery mode have given pilots intermittent and the minimum and maximum angles. Secondary
confusing information during such motions. The differences include small increases in the magnitudes of
recognition that a “spectrum” of out-of-control motions the yaw and pitch rates. These differences are
exist, which at once recognizes the similarities and the indicated by the lighter highlight.
differences between these motions, may serve to make
control system designers aware of the broader spectrum
of out-of-control motions which must be dealt with.
Table 3. Spectrum of Post-Stall Gyrations
FALLING LEAF OSCILLATORY SPIN STEADY SPIN
In-Phase, Periodic Roll and Yaw In-Phase, Periodic Roll and Yaw Rates Steady Roll and Yaw Rates At
Rates Fluctuating about Small or Fluctuating about Large Non-Zero Means. Large Non-Zero Values.
Zero Means.
Large α and β Fluctuations Yielding Moderate α Fluctuations but β Fluctuations Small α or β Fluctuations
Large peak α& and β& . Angle-of- may be Large: Moderate α& and β& Aircraft Yielding Minimal α& and β& .
Attack May Dip Below Stall Angle. remains Stalled throughout Motion Aircraft Remains in a Deep Stall

Table 4. Falling Leaf as a Sub-Spectrum of Post-Stall Gyrations


STATE SLOW FALLING LEAF FAST FALLING LEAF HIGH ALPHA FAST FALLING LEAF
α (deg) -5 to +60 +20 to +70 +30 to +85
β (deg) -40 to +40 -40 to 40 -40 to +40
P (deg/sec) -120 to +150 -90 to +130 -90 to +130
Q (deg/sec) -30 to +30 -35 to +35 -50 to +50
R (deg/sec) -50 to +50 -10 to +60 -10 to +75
λ (sec) 7 4.7 4.5
α dot (deg/sec) 70 70 70
β dot (deg/sec) 100 100 100
H dot (fps) 350 325 325

Whereas the shift in the angle-of-attack range may sub-spectrum of falling leaf motions within the larger
impact the effectiveness of the controls, the higher spectrum of post-stall gyrations, raises questions as to
rotational rates about the pitch and yaw axes result in whether or not one recovery system can work for all
higher momentum and energy levels. This can impact motions. It is even possible that a recovery system or
the ease of- or time required for- recovery due to a technique capable of stopping one type of motion, may
difference in the level of energy which must be actually aggravate other motions in the spectrum.
dissipated to attenuate the motion. The existence of a

5
Figure 1 provides and a rare, high-quality flight-test become visible in the pitch rate trace, and it can clearly
record of a slow falling leaf motion. The first four be seen as lagging behind the roll and yaw traces. This
graphs provide the aerodynamic angles, rotational is due to the influence of the inertial pitch acceleration
velocity components, Euler angles and speed in that component, which is highest when the roll and yaw
order. The next four graphs show the rudder, left and rates are at their peak. Since the rate is the integral of
right aileron, left and right horizontal stabilizers, and the acceleration, this lag is expected; and is also an
altitude, again in that order. Although the motion indication of the strength of the inertial pitch
depicted in this figure cannot be considered as having acceleration term.
reached a “steady state” condition, there are several Figure 3 depicts a simulation of a motion which is
characteristics which are evident and become more being termed a “high-alpha fast falling leaf” in this
clearly defined during sustained motions. paper. Like the motion depicted in Figure 2, the angle-
First, note that the aircraft spends most of the time in a of-attack range is not precisely fixed, but in this case
very high sideslip condition or in a high angle-of-attack slowly drifts upward even once the motion has reached
coupled with a high sideslip condition. Simulation a repeatable pattern beyond about 35 seconds. The
databases of many military aircraft may cover a large relationships between the states are not appreciably
angle-of-attack range but tend to lack high sideslip different from the fast falling leaf shown in Figure 2.
data. Without high sideslip data coverage, at least in Subtle changes are noted, however, such as a slight
the basic static coefficients, there is little chance of change in the pitch rate behavior as well as a small
simulating the motion. In addition, most of the controls increase in the pitch and yaw rates. The consequences
show strong dependence on both of the aerodynamic of this seemingly small shift can be seen in the trace of
angles so that the database must also include high the magnitude of the angular momentum vector which
sideslip controls data. The controls data requirements is shown in the last graph of Figures 2 and 3. A
are necessary, not only for properly simulating the correlation can be made between the angle-of-attack
motion, i.e. automatic flight control systems may play a range of the motion and the area below the angular
large role in the motion, but also to allow a credible momentum curve. In particular, as the angle-of-attack
analysis of recovery options. The other distinguishing drifts towards higher values, the average angular
feature can be seen in the overplots of the roll and yaw momentum increases. Since this momentum must be
rates. In particular, these two states display an in-phase overcome to stop the motion, this feature should be
behavior throughout the motion. expected to affect recovery.
Figure 2 provides a time history of a simulated motion
referred to in this paper as a “fast falling leaf.” The 3.2 Simulation vs. Flight
motion begins as a slow falling leaf similar to that This section addresses the accuracy to which the falling
shown in Figure 1 until the controls are held neutral at leaf motion could be simulated using the model and
30 seconds. The oscillations are seen to speed up for methodology described in section 2. In doing so, it also
about two cycles then slow down again for another half establishes the extent to which conclusions based on
a cycle. At around 40 seconds, the motion finally such work can be considered as reliable.
begins to sustain itself in a higher frequency oscillatory Unfortunately, there existed very few flight-test data of
mode. The period of motion decreases from just under sufficient quality against which the accuracy of the
8 seconds in the slow pattern to just under 5 seconds in simulation could be measured. Figure 4 compares a
the fast motion. The relationships between the state few features of Flight 12-9, seen previously in Figure 1,
variables settle down into distinct patterns, although in and a simulation run driven only by the recorded pilot
true non-linear fashion, the states never exactly repeat stick and rudder inputs. Considering the nature of the
themselves. Notice that the angle-of-attack range of the motion, the agreement is excellent in terms of the
motion shows some variation with time even in the magnitudes and phasings between the states; and the
“steady state” portion of the motion. fact that the simulation showed a similar recovery.
A strong biasing of the yaw rate also distinguishes this Figure 5, provides an example of the ability of the
type of motion from the slow motion depicted in Figure model to produce fast falling leaf motions.
1. Although the yaw rate actually reverses itself during Unfortunately, a set of high-quality data for this type of
the motion, the bias results in a near continual heading motion was not available for this study. However,
change resembling an oscillatory spin. This is clearly reference 14 provided a hand-rendered time history of a
seen in the trace of Psi, the Euler heading angle shown motion experienced by an F-18A, which, at the time
in the third graph Figure 2. This behavior led some to was called an “intermediate yaw rate oscillatory spin.”
naming this motion a low- or intermediate- mode
oscillatory spin. In the sustained motion, a pattern does

6
Figure 1. Flight 12-9: Slow Falling Leaf

7
Figure 2. Simulation of a Fast Falling Leaf

8
Figure 3. Simulation of a High-Alpha Fast Falling Leaf

9
Figure 4. Flight 12-9 vs. Simulation

Figure 5. F-18A Flight Record vs. Simulation


To prepare Figure 5, the data traces of the flight record
were retraced and scanned into the computer. The

10
angular rates were inverted to change the flight record 4. Hess, R. A., “Effect of Various Stores on the
from a left motion to a right motion. The middle 25 Aerodynamics of an F/A-18B Aircraft Aerodynamic
seconds of the motion were selected to emphasize the Identified from Flight-Test Data,” Rept No. SCT-6612-
most salient points. Any attempt to simulate the entire 020-2, February 1989.
flight would have been very difficult given the quality 5. O’Conner, Cornelius, “NATC F-18A/B
of the data, the sensitivity of the motion to the initial Aerodynamic Math Model Modifications Incorporated
conditions, and the lack of key configurational data. during the Phase I Model Unification Effort,” BAR 89-
The correlation of the simulation with the available 5, March, 1989.
states recorded during this flight, and general
comparisons with falling leaf traces from other aircraft, 6. O’Conner, Cornelius, “NATC F-18A/B
indicate that the simulation provides a reasonably Aerodynamic Math Model Modifications Incorporated
accurate portrayal of this type of motion. during the Phase II Model Unification Effort,” BAR
90-13, September, 1990.
In evaluating the accuracy of the simulation, it must be
kept in mind that practically the entire simulation 7. Ralston, John N. and Avent, J. L., “Evaluation
database is excited by these motions. In particular, the and Modification of F/A-18B Aerodynamic Data Base
angle-of-attack ranges between slightly negative angles for Improved Departure Modeling, BAR 92-2, Mar.
to over 70 degrees while the sideslip angle ranges 1992.
between plus and minus 40 degrees. The wide range of 8. Ralston, John N., O’Connor, C. J. and Avent, J.
angular rates also excites a large portion of the L., “Evaluation of the NAWC/AD F/A-18 C/D
rotational and dynamic database. In sum, the accuracy Simulation Including Data Base Coverage and
of the simulation model is sufficient to allow a detailed Dynamic Data Implementation Techniques,” BAR 95-
analysis of the motion; and results should be considered 3, Dec. 1995.
to have a high degree of reliability.
9. O’Connor, C. J., Ralston, John N., and Timothy
IV. Concluding Remarks Fitzgerald, “Evaluation of the NAWC/AD F/A-18 C/D
Simulation Including Data Base Coverage and
This work has illustrated that very complex motions Dynamic Data Implementation Techniques,” AIAA
such as the falling leaf can be modeled with sufficient Paper 96-3365, May 1996.
accuracy given a comprehensive aerodynamic database
and an accurate automatic flight control system model. 10. Kalviste, Juri, "Use of Rotary Balance and Forced
The minimal database requirements include high Oscillation Test Data in a Six Degrees-of-Freedom
sideslip data in the static coefficients: at least up to 45 Simulation," AIAA Paper 82-1364, Aug. 1982.
degrees, and preferrably up to 60 degrees. Accurate 11. Kalviste, Juri, "Math Modeling of Aero Data for
simulation also requires a strong complement of rotary Aircraft Dynamic Motion" AIAA 94-AFM-26-7,
balance and forced oscillation data. In order to provide August 1994.
a credible analysis of recovery, data on controls 12. Glaser, B., Hess R. K., Trame, L. W., “F/A-18
effectiveness at high sideslip is also essential. V8.33 Control Law Mechanization,” McDonnell
Douglas Memo 338-5788, May 1984.
Acknowledgments
13. Bach, Ralph, Russell Paielle, "Direct Inversion of
The authors wish to express their gratitude to Bill Rigid Body Rotational Dynamics," NASA TM-102798,
McNamara and Chad Miller at NAWC/AD Manned 1990.
Flight Simulation Facility for their support of this work.
14. McNamara, William, “F/A-18A Intermediate Yaw
References Rate Oscillatory Spin Requiring Manual SRM
Recovery,” Memorandum, October 9, 1987.
1. Hess, R. A., “Simulation Checking Using an
Optimal Prediction Evaluation (SCOPE) User Guide,”
SCT Rept. 4522-270-2, September 1987.
2. Hess, R. A., “Subsonic F/A-18A and F/A18A/B
Aerodynamic Identified from Flight-Test Data,” Rept
No. SCT-4522-220-1, July 1987.
3. Ralston, John N., “Analysis of F-18A/B Low
Speed, High Angle-of-Attack Aerodynamic Data Base,
Bar 88-7, September 1988.

11

You might also like