Paper

Int’l J. of Aeronautical & Space Sci. 18(2), 175–185 (2017)

DOI: doi.org/10.5139/IJASS.2017.18.2.175

Calculating Dynamic Derivatives of Flight Vehicle with New Engineering

Strategies
Baigang Mi* and Hao Zhan**
School of Aeronautics, Northwestern Polytechnical University, Xi’an 710072, China

Baibing Chen***
Xi’an Modern Control Technology Research Institute, Xi’an 710065, China

Abstract
This paper presents new differential methods for computing the combined and single dynamic stability derivatives of flight
vehicle. Based on rigid dynamic mesh technique, the combined dynamic stability derivative can be achieved by imposing
the aircraft pitching to the same angle of attack with two different pitching angular velocities and also translating it to the
same additional angle of attack with two different rates of angle of attack. As a result, the acceleration derivative is identified.
Moreover, the rotating reference frame is adopted to calculate the rotary derivatives when simulating the steady pull-up
with different pitching angular velocities. Two configurations, the Hyper Ballistic Shape (HBS) and Finner missile model, are
considered as evaluations and results of all the cases agree well with reference or experiment data. Compared to traditional
ones, the new differential methods are of high efficiency and accuracy, and potential to be extended to the simulation of
combined and single stability derivatives of directional and lateral.

Key words: Combined dynamic derivative, Acceleration derivative, Rotary derivative, Computational Fluid Dynamics (CFD)

1. Introduction angle of attack, airspeed, altitude, etc. Thus, the research on

methods to obtain dynamic derivatives has attracted more and
How to obtain the characteristics of flight dynamics is a more attention.
significant step in the process of aircraft design. Due to the The dynamic derivatives can be calculated by several
complicatedly unsteady flow around the flight vehicle, until methods, namely empirical methods, wind tunnel testing or
now it is a huge challenge to determine every aerodynamic computational fluid dynamics (CFD), and each of them has
parameter for flight dynamics analysis, especially for advanced its own advantages and disadvantages. Empirical methods
aircraft during high maneuvers. In order to solve the problem, usually use existing data to estimate the stability derivatives
a common approach is based on the assumption that the of similar configurations. It is quick but not accurate. For
aerodynamic forces and moments react in a linear form for small example, DATCOM [4] is a software using this method which
variations from a given steady –state flight initialization, which contains a document of more than 1500 pages covering
means they can be characterized by a series of aerodynamic detailed methodologies for determining stability and control
derivatives, called stability derivatives [1-3]. Specifically, characteristics in a variety of aircraft configurations; Wind
these dynamic derivatives are used for the indication of flight tunnel testing may be the most reliable tool to analyze
dynamic stability, and present measurements of how changes dynamic characteristic besides flight testing, but it is always
will occur to the forces or moments acting on the vehicle with restricted to be applied due to the limitation of blockage,
a fluctuation of parameters in the flight condition, such as scaling, Reynolds-number effect and support interference [5-

This is an Open Access article distributed under the terms of the Creative Com- * Ph. D Student, Corresponding author: [email protected]
mons Attribution Non-Commercial License (http://creativecommons.org/licenses/by- ** Professor
nc/3.0/) which permits unrestricted non-commercial use, distribution, and reproduc- *** Engineer
tion in any medium, provided the original work is properly cited.

Received: April 18, 2016 Revised: March 5, 2017 Accepted: May 24, 2017
Copyright ⓒ The Korean Society for Aeronautical & Space Sciences 175 http://ijass.org pISSN: 2093-274x eISSN: 2093-2480

(175~185)16-062.indd 175 2017-07-03 오전 9:50:58

 Int’l J. of Aeronautical & Space Sci. 18(2), 175–185 (2017)

6]. Recent years with the development of CFD technique and codeandprovides lots of models
its high-precision andaddressing
robustnessa widehaverangebeen ofvalidated
fluid in many fields.
unsteady aerodynamics theory, it is promising and possible flow problems and its high-precision and robustness have
to acquire dynamic derivatives with high-fidelity parallel beenThe equation
validated describing
in many fields.the unsteady dynamic motion is the 3D Reynolds-averag
CFD methods. The equation describing the unsteady dynamic motion is
Generally, the simulation of dynamic derivative is
an be expressed as:
the 3D Reynolds-averaged N-S equation can be expressed as:
obtained with CFD by imposing a forced sinusoidal motion  1
t 
WdV   F  nds＝  Fv  nds (1)
around the aircraft center of gravity. Based on that, many Re
V V V
scholars have carried out a variety of techniques to identify
where V indicates the control volume, W=(ρ, ρu, ρv, ρw, ρE)T
the stability derivatives. Park [7-8] presented a dual-time where V indicates the control volume, W  (  ,  u,  v,  w,  E )T is the vector of co
is the vector of conserved variables, ρ is the density and u, v
stepping algorithm to solve the unsteady oscillation with
and w is arethethedensity
components
and u ,of vvelocity
and wgiven are bythethe Cartesian of velocity given by the
components
a multigrid DADI method, which is accurate but time-
velocity vector U=(u, v, w)T. The total energy per unit mass
consuming. Oktay [9] employed arbitrary Lagrangian –
Eulerian (ALE) formulation with a dynamically deforming
is E. F denotes
vector U  (u,the v, w)inviscid
T component
. The total energy perof theunitflux is E . F denotes the invi
vectors,
mass
while Fv is the viscous flux vector which contains terms of
mesh algorithm to calculate the dynamic derivatives of the
thetheheat fluxfluxvectors, while forces
and viscous Fv isexerted
the viscous
on theflux vector
body. Re which
is the contains terms of the he
Basic Finner Missile (BFM) model, but the unaffordable
Reynolds number.
cost of mesh deformation became new challenge. Da Ronch
forces exerted onmodel
The turbulence the body. usedReisisthe
theSpalart-Allmaras
Reynolds number. (S-A)
[10-11] developed Fourier integral and linear regression
model [14], which is an one-equation model that solves a
methods to analyze the stability derivatives of Standard The turbulence model used is the Spalart-Allmaras (S-A) model [14], which i
modeled transport equation for the kinematic eddy viscosity.
Dynamic Model (SDM) and the Transonic Cruiser (TCR).
Themodel S-A that model solves was designedtransport
a modeled especially for aerospace
equation for the kinematic eddy viscosity. T
The analyzed results showed good agreement with wind
applications involving wall-bounded flows and has been
tunnel testing but only part of stability derivatives could
designed
shown to give especially
good results for aerospace
for boundary applications involvingtowall-bounded flows and h
layers subjected
be simulated by using these methods. Murman [12] used
adverse pressure gradients.
Cartesian mesh to solve the non-linear, reduced frequency give good results for boundary layers subjected to adverse pressure gradients.
The rigid dynamic mesh technique is adopted to model
approach to simulate derivatives, and the method was rather
flowsThe where the whole
rigid dynamic domainmeshistechnique
varying as is a function
adoptedoftotime model flows where the w
efficient but still not so accurate.
due to unsteady motions, aiming to identify the combined
Even though much work has been done for the computation
and arying as a function
one single of time dueThis
stability derivatives. to model
unsteady is amotions,
special aiming to identify the
of stability derivatives, only the combined form derivatives
case of general dynamic mesh motion wherein the nodes
are identified with most methods, which cannot satisfy the ne single stability derivatives. This model is a special case of general dynam
move rigidly in the whole zone and can be conveniently
requirements of advanced flight vehicle whose aerodynamics
used in cases
wherein theof nodes
single freedom
move rigidlymotions in with only one
the whole body,
zone and can be conveniently
and control system need to be accurately designed with
seen in Fig. 1. Actually, the rigid dynamic mesh technique
detailed single dynamic derivatives. In this case, new CFD
single
is not freedom
a novel motionshowever,
approach, with only one body,
compared to theseen in Fig. 1. Actually, the rig
widely
methods are urgently needed to finely simulate the derivatives.
used dynamic remeshing technique, the grid quality can be
In this work, we seek to detailed calculate all combined technique is not a novel approach, however, compared to the widely used d
guaranteed during the motion. Moreover, the rigid dynamic
and single dynamic derivatives with a series of new CFD
mesh technique the
g technique, can lead
grid to high efficiency
quality as the reason
can be guaranteed that the motion. Moreover,
during
methods based on rigid dynamic mesh technique. With
there is no need to calculate the grid deformation, which can
which the aircraft will be forced to translate or rotate twice
of ic
Originals be usedmesh technique
to quickly
illustrations used in the can(preferably
identify
text leaddynamic
the towith
high efficiency
derivatives
captions as the
with
on a separate thereason
sheet or well that there is no
with different parameters, both the combined and single
separated new
stability derivatives can be easily achieved with the unsteady from thedifferential
the illustration tomethods.
enable scanning).
grid deformation, which can be used to quickly identify the dynamic de
aerodynamic forces and moments. The longitudinal dynamic
derivatives of Hyper Ballistic Shape (HBS) and Finner missile new differential methods.
configurations are further employed to validate effectiveness
To determine the other single derivative, the rotating reference frame is adopted
of the new methods. In addition, another merit of the
proposed methods is that they can be applied to lateral and unsteady in the stationary (inertial) frame steady with respect to the moving fram
directional dynamic derivatives simulation.
rotating frame, it is possible to transform the equations of fluid motion to the movi

2. Numerical method 4

(a) Initial mesh (b) Dynamic mesh after rigid motion

The commercial software, ANSYS Fluent [13], is chosen
to calculate the steady and unsteady cases in this paper. The Fig. 1. Rigid dynamic mesh technique
Fig. 1. Rigid dynamic mesh technique

DOI: doi.org/10.5139/IJASS.2017.18.2.175 176

(a) Rotating to  with q1 (b) Rotating to  with q2

Fig. 2. Differential method to calculate combined stability derivative with two different pitching
(175~185)16-062.indd 176 2017-07-03 오전 9:50:59
angular velocities
    2 0 sin(0 t )  q
   sin( cos(
 t ) t )
 0 0

     0 cos(t )

Substituting Eq. (4) into Eq. (3), the longitudinal pitching moment coef
Substituting Eq. (4) into Eq. (3), the longitudinal pitching moment coefficient can b
Cm Cm 0  (Cm   2Cmq ) 0 sin t  (Cm  Cmq ) 0 cos t
Cm Cm 0  (Cm   2Cmq ) 0 sin t  (Cm  Cmq ) 0 cos t
Baigang Mi Calculating Dynamic Derivatives of Flight Vehicle with New Engineering Strategies
In the
In thecase
casethatthat
t 2tn 2nand
 and 
nt  2/n2, 
t 2 2 ,are
 /(5)
Eq. Eq.respectively
(5) are respectively
expressed as

To determine the other single derivative, the rotating CmCtm2 nt 2nC m 0 Cm 0
rotating frame, it is possible to transform the equations of fluid motion Cmm to CCmqthemq moving k0k frame such that
(6)
reference
rotating frame, frame it is is adopted
possible
rotating frame, it is possible to transform the equations of fluid motion to the moving frame such asto the problem
transform the is unsteady
equations in
of the
fluid motion to the moving frame suchthat that Cm t 2 n  /2  Cm 0
rotating frame, it isframe possible to transform thetoequations of fluid motion to the moving frame such that
0
C   2
C mq �
stationary
rotating
rotating frame, (inertial)
frame, it is possible
it is possible steady to with
transform
to transform therespect the the
equations moving of C
fluid motion 
to Cthe moving frame such that
m
0
steady-state solutions are available. C equations  Cmq of
2
� fluid motion to22the moving
m 
 t 2 n   /2 m 0
C
Cmm
frame such  Cm 0that (7)
steady-state solutions are available. C   C /2  Cm 0
tt 22nn  /2
steady-state
steady-state
frame. For solutions asolutions
steadily are areavailable.
rotating available. frame, it is possible mto transform 
C0m   C2 mq � C

 C 2 (7) C t 2 n  /2  Cm 0
steady-state solutions are available. C  C m 
Cm   Cmq
mq 
m  t02 n    /2 m 0 Cm  k Cmqc /�2V is the non-dimension
m
steady-state
equationssolutions
steady-state ofsolutions Cmmotion areavailable.
 are
2
Cto
available. � moving m t 2 n  /2 m0
0C (7)where 0
t 2 n  /2  Cm 0
the fluid mq the frame such that
3. Dynamic derivative calculation method  C   2
C � m
0


steady-state
3. Dynamic solutions
derivative are calculation
available. method 0 where k   c / 2 V is the non-dimensional m reducedmq frequency  of the applied motion. Here Cm  Cmq
3.3.Dynamic
Dynamicderivative derivativecalculation calculationmethod method

wherewhere k  c / 2V is the non-dimensional reduced is the non-dimensional frequency of the
reduced
represents applied motion.mom He
where k  the c / 2Vpitch - damping
0
3. Dynamic derivative calculation method is the non-dimension
3.3.1 3.Dynamic Dynamicderivative
Traditional derivative
CFD where methods calculation c / 2V method
kcalculation is
methodthe non-dimensional reduced
represents the pitch - dampingwhere
frequencyofofthe
frequency theapplied
moment
applied motion. motion. Here Here Cm  Cmq represents
3.1 Traditional CFD methods k  c coefficient,/ 2V is the non-dimensional which is oft5 ereduced r phase
m efrequency
d derivative,
a s of the out- and moti
3.1
3.1 Traditional
3.Traditional
Dynamic CFD CFD methodsmethods represents the pitch - damping moment coefficient, which is t eapplied r m e d
3.1 Traditionalderivative CFD methodscalculation method the pitch-damping moment coefficient, which is termed
represents 5 the pitch - damping mom
3.1 3.1 Traditional
Traditional
Traditionally CFD CFD
the methods
methods
dynamic
represents derivative is obtained by forcing the aircraft moving sinusoidal-y
Traditionally the dynamic the derivativepitch - isdamping ofobtained phase moment
by forcing coefficient,
derivative, asthe out-of aircraft which
and
phase moving
represents dueis sinusoidal-y
derivative, t e r and
theto pitch
m ethe d to
-duedamping
areason
sthe out-
due moment
reason thatthat 
coefficient,
toare difficult toreason
and which
separate,
q
this isderivative
t e r m
Traditionallythe
Traditionally thedynamic dynamicderivative derivativeisisobtained obtainedbybyforcing forcingthe theof aircraft
aircraft phase moving
moving sinusoidal-y
derivative,
sinusoidal-y and of the
phase derivative,that and 
3.1 Traditionally
Traditional CFD the dynamic
methods derivative is obtained by forcing and the q aircraft
are difficult moving to sinusoidal-y
separate, this derivative is also named
around Traditionally
Traditionally the center the the
of ofdynamic
dynamic phase derivative
gravity, derivative
which derivative, is issimilarisobtained
areobtained
and to wind
difficult
byby due forcing
forcing
tunnel
to tunnel
separate, tothe the aircraft
thisthe
testing, aircraftand
derivative
moving
reason moving
the is also that
sinusoidal-y
aerodynamic sinusoidal-y
named asfor

combined and q
around of dynamic phase derivative, anddynamic Cdue derivative.
2to theMoreover,
as reason that
around
around the thethe center
center center ofofgravity, ofgravity, gravity, which which
which issimilar
isissimilar similar totowind towind wind tunneltesting,
tunnel as testing,
testing, combined
are andand
difficult and
the the tothe aerodynamic
aerodynamic
aerodynamic
separate, this for
derivative. forfor Moreover,
derivative is also named m 
are

as C
difficult mq known
combined
to separate,
in-phase
dynamic derivati
derivative
this derivative
around the center of gravity, which is similar to wind tunnel testing, and the aerodynamic for
around
Traditionally the center the mdynamic ofnof gravity, derivative which isisobtainedsimilar by toforcing
wind tunnel testing, and the aerodynamic is for
cc around
ee ss aa nn ddthem mcenter oo m eeare ss gravity,
n ttdifficult aa rr ee toll iiwhich
nn ee aa rr isffthis
separate, uusimilar tt ii oonnto2sC
nn ccderivative s windooisknown aatunnel
ff also ss ssnamed iitesting,
oo ccin-phase
aa tas
t eeas ffand
ddcombined llare hhthe
ii gg difficult ppaerodynamic
ttdynamic aaalso
rrtoaa separate,
m tt ee rr ssfor
aeecombined
derivative. .. Moreover,
known in-phase derivative also a combined dynamic
m this derivative
dynamic is also namedcomprised as combined of dynamic der
c cethe
es s aircraft a an nd d mmoving mo omme ensinusoidal-y nt st s a ar er e l il naround i ne ea ar r fthe nc ctCitmoionof
uf uncenter ns sgravity, as
o ofmqf a as s os oc ciderivatives
ai aC
derivative
t etmed d flf2ilCgicomprised
gh htknownt p pa ar is arasamof m e et aet erstatic
in-phase sr .s .derivative is also
derivatives
a combined
aCstatic derivative
2
C mq Cmas
dynamic  and derivatives
a rotational cod
c e s a n d m o m e n t s a2 r e l i n e a r f u n c t i o n s o f a s s o c i a t e d f l i g h t p a r a m e t e r s . mq derivative m and a known in-phase derivat
c eFor
which cse sisaexample, na dn d mm
similar om
to othe
wind mCe mnelongitudinal
tntunnel
st 
 saC raemq l ilnpitching
rtesting,
eknown i en aeand ra rin-phase
as ftheufmoment
nuaerodynamic
cntcitoderivative
i no sncoefficient
s o fo f is a saalso
ssorotational
siscoaicanormally
i taet de dderivative
combined fClfidynamic
lgiformulated
hg th t2C p apderivatives
r.aknown
ar mam e as
teaset rein-phase
ascomprised
r.s . of
derivative is also a combined dynamic derivati
ForFor example, thethe longitudinal pitching moment coefficient isnormally
Cis normally normally m formulated
Cmq as
aa. a

For example,
example, the longitudinal
longitudinal pitching
pitching moment
moment a staticcoefficient coefficient
derivative is and a rotational formulated
formulated derivative mq
asas In
forces Forand example, moments theare longitudinal
linear functions pitching moment flight
of associated coefficientmisaIn static
normally
the derivative
above formulated
method, Cm the and combined
a as
rotationala derivativeaderivatives
dynamic Cmqthe
static above method, the combined
 .derivative Cm and a rotational
For
function Forexample, of the the
example, angle thelongitudinal of attack, pitching
longitudinal pitching
, pitching moment
moment angular coefficient
velocity,isisqnormally
coefficient and the formulated
, normally rate formulated of theseastwo asa a
function
parameters.
function
function ofofthe ofthe
Forthe angleexample,
angle anglea static
ofofattack, of derivative
the
attack, attack,
longitudinal Cpitching
, ,pitching and angular
,mpitching
pitching a angular
rotational
angular
moment derivative
velocity,
velocity,
velocity, qcanq, ,C qbe
and , .and
and the thethe
simulated aratestatic
rate rate of ofthese
through ofthese
derivative these
the twoC
two two
m and a rotational
identification derivative
of aerodynamic
fication Cmq .
of through
aerodynamic
function of the angle of attack,  , pitching In angular
the above method,
velocity, q
the
mq
, andcombined the rate dynamic of these derivatives
two can be simulated the forces
identi and mome
In the above method, the combined dynamic In the
derivatives above canmethod,
be simulatedthe combined throu
function
coefficientfunction of
is of
variables,  and q , which is defined as the
normally the angle angle of
formulated of attack,
attack, as  a , ,pitching
functionpitching of angular
theangularangle velocity,
velocity, forces q ,
q ,and
and and the the
moments rate rate of of
after these
these
several two two cycles’ unsteady motion,
variables,
variables,
of attack,α,
variables, 
and
pitching and
and q q, angular qIn
,which ,
whichis which
the above is
isdefined
velocity, defined
defined method,
q, as andas theas the
ficationcombined
rate ofofthese dynamic
aerodynamic but
derivatives
forces canIn
and momentscost
the computation
be
the simulated
above
afterisseveral method,
huge.
through the
cycles’ the
combined identi
unsteady dynamic
ionmotion,
cost is but derivatives
huge. theEven can be simulated
though the harmo
computat
variables,  and q , which is defined as fication of aerodynamic forces andEven momentsthough afterthe harmonic
fication
several ofcycles’
aerodynamicunsteady forcesmotion,and momebut
two variables,
variables,
variables,   and and andq ,,qwhich ,which which is isdefined
isdefined
defined as asas ˆ balance method fication [15] of or time spectral
aerodynamic forces method
and [16] has
moments after been several cycles’ unsteady motion
Cm Cm 0  Cm fication  Cm  ofaerodynamic
 Cmq q  Cmqforces q    (
and
ˆ cost  , q )
moments after several cycles’ unsteady motion, (2)
but the computat
(2) [15] or time
C
 mmC Cmm00C C   C  CC qCqmq qC qC  qˆ ˆion (ˆ (, q,)q ,) q)
is huge. Even though the harmonic balance method been spectral
usedismethod reduce[16]
tohuge. time hascoststhewith t

CC  CC    qmq (2) (2) ion cost [15] Even though spectral harm
     mqmq   ( 
mq
mm C m 0m 0 C mm  C ion to costreduce is huge. Even though thethe harmonic balance method or time meth
m m mq
m m used time costs with unsteady calculation
Cm Cm 0  C   mm
 C    mqmq
 C q  C q  
ˆ (
ˆ   , q ) (2) (2)
CmCm CmC0m0 C  mCm
m
ion ˆC cost C 
m
 C
ishuge.
 mqC qEven
mq
qC
  mq q   ( (
Cthough
mq q  the
mq 
 ,harmonic
q,)q ) balance transformedmethod [15] ion from cost or time is huge.
time spectral Even
domainmethod (2)
to(2)
though the harmonic
[16] hadomain,
frequency balance
so frequency method [15]
it is domain, it is still not a s or time spectr
ˆ ( m, q ) aremqhigh-order
m
where C mq q and been terms used whichto reduce are often timediscardedcosts with duetheto unsteady their small ef
calculation transformed been used fromto time reduce domain
time costs t with t
where C q and ˆ ˆ  (   , q ) are high-order terms which are often discarded due to their small ef
whereCC
where
where q qmq and and and  ( (  ˆ(,q,q) ),are areare high-orderterms
high-order
high-order termswhich
terms whichare are areoftenoftendiscarded been
discardeddue
still not used
a to
duetototheir
systematic reduce theirsmall
method time costs
smallefef
and with
only thethe unsteady
combined calculation
form of transformed from tim
mq
wheremqmq C mq q andbeen 
ˆ their q ) are high-order terms which are often discarded due
been transformed usedto their
to reduce small time ef costs with the ty derivatives are obtained.
unsteady calculation transformed f
where
often where
fects. discarded
Then
CC mqmq qtheq dueand andto
unsteady
(ˆ(used ,small

moment q,)qto)are reduce
are high-order
effects. time
high-order
coefficient Then ocan terms
costs
the terms
frequency
be
with
unsteadywhich
which
simplified
the are
domain, are often
unsteady
as often it is
stability discarded
calculation
discarded
still not due
derivatives adue toaretotheir
systematic theirsmall
obtained. from
small
method ef
timeand domain
ef only theot combined frequency domain, form of itstabili is still not a s
fects. Then the unsteady moment coefficient can be simplified as o frequency domain, it is still not a systematic method and only the combined fo
fects.
fects.
moment ThenThen the theunsteady
coefficient unsteady can moment
moment
be simplified coefficient
coefficient as can canbebesimplified simplifiedasas To deal with o frequency the difficulties domain, itof isestimating still not a the Todynamic
systematic deal with method theand difficulties
only theofcombi estim
fects. Then the unsteady o frequency moment coefficient
domain, it is ty can
still benotsimplified
derivatives a systematicare as as
obtained. method and only the combined form of stability derivatives are obtained.
fects.
Cfects. � Then
C Then  theC the unsteady
 unsteady
  C moment
 moment
   C coefficient
coefficient
q can can be be simplified
simplified as derivatives, a
ty derivatives are obtained. set of new approaches (3) are proposed and
C�mmm C Cmmm000C C   C    qCqmqmq q
(3) (3)(3)
m m mq
CC  ty derivatives are (3) obtained. are proposed and investigated in this
m m� Cm 0m 0 Cmm   CtyC     CC
m
m m
m
Cm � Cm 0  C   
mmderivatives
C
Originals    
mqmq
C are q obtained. To deal with the investigatedof
difficulties in this estimating paper. the dynamic (3) derivatives, aToset deal of withnew the difficulties of estim
approaches
m of illustrations used in the text (preferably with captions on a separate sheet or well
CThe CThe
m CmC
� relationship
m � 0m 0 m
relationship C Cmm
C mCm
between
between  angleC
angle
 mqCmq
mq
qofqofattack attackand andpitching pitching angular To deal with
velocity are the
presented difficulties (3)
of(3)estimating the dynamic derivatives, a set of new
in difficulties
Eq.
The relationship between angle ofattackattack and To deal with the of estimating the dynamic derivatives, a set
The The relationship
relationship between between
To deal angle
angle with of
ofEq.the attack difficultiesand and ofpitching
pitching
pitching
areenable proposed estimatingangular angular
angular thevelocity velocity
velocity
dynamic are
are are
derivatives,
presented presented
presentedpaper. ainin setEq.inEq.
ofEq. new approaches are proposed and investigated in this
pitchingandangular investigated in this
angular separated from inthe illustration to scanning). 3.2  New CFD methods to calculate combined and
The velocity relationship are presented
between angle of(4)attack under and harmonic are velocity
proposed are and presented
investigated in Eq. in this paper. 3.2 New CFD methods to calculate combi
(4)The The
under relationship harmonicbetween
relationship betweenangle
movement. angleofofattack attackand andpitching pitchingangular angular velocity
velocity areare
are
single dynamic derivatives
presented
presented
proposed andinin Eq. Eq.
investigated in this paper.
(4)
(4) (4)
movement.
under under under harmonic
harmonic harmonic are movement.
movement.
movement. proposed and investigated in this paper.
(4) under harmonic movement. Different from traditional CFD metho
(4)
 (4) under
 under sin(harmonic t ) harmonicmovement. movement.
   000 sin(t ) 3.2paper
New CFD the methods to calculate comb
  
 0sin( 0 sin(
  t)
t )cos(  t ) q 3.2 New CFD methods toDifferent calculatefrom combined traditional and single CFDdynamic methods, in this
derivatives
  
        sin(
0  t ) 3.2 New
differential methods CFD methods
3.2 New CFD to
are employed calculate
methods toto combined
calculate combined and single
calculate dynamic
the combined
and single derivatives
dynamic and all the singl
derivatives
(4) calculate the combined
0 0 cos( t )


  
 00
0 cos(
sin(
cos( 0 sin(
02 t)sin(
 tt)) qt )q q 
   
  2 cos(  
t ) t ) 
q q
3.2 New CFD methods to calculate combined (4) and single dynamic derivatives (4) Different from traditional CFD metho
  
 
   2 0
2 0  0 sin( t )  q Different from traditional
and all CFD
the methods,
single dynamic in this paper
(4)
derivatives.(4) the differential
Firstly, based methods
on small are employed to

  
 
2 
0  cos(
00 2sin(
sin( 0
0cos(
 0t t)
t
)sin( ) qt)qtq) q Different from Different traditional from traditional CFD(4)methods, CFD methods,in thistheory,paper thisthe
in the differential
vehicle
paper methods
is differential
the forced to move
methoare
   0sin(  t)  q (4)(4) calculate the combined and all the sing
   
  
 
 2 2sin(
sin(  00 sin( t ) 
sin(
 
t
sin(
) t t)) 
t ) q  q
 Different from traditional CFD methods, in disturbance
this paper the theory,
differential the methods vehicle is areforced
employed to move to with two
    
0
    
0 0 0
 0 sin(0 cos( t ) t) calculate the combined and all the single dynamic derivatives. Firstly, based on small disturbance
 
  
  
 cos( sin( cos(
00 sin(
t)tt ))tt)) calculatekinematic
different thecalculatecombined the combined
parameters. and all Then, theandsingleall
with thedynamic
thesingle derivatives.
difference
dynamic
difference of Firstly, Firstly,
the unsteady
ofderivatives. based
or steady on
basedsmal
aerodo
     0  00cos( 0 
cos(tcalculate ) theory, the vehicle is forced to move

Substituting
  the combined and all the
theory,pitching single
the dynamic
vehiclemoment isthe derivatives.
forced
unsteady to move Firstly,
or steady withbe based
two
aerodynamic on
different small disturbance
kinematic
forces parameters.
and moments, all Then, with the
     0 cos( t(4)
0
   0Eq. cos(  ) t ) into Eq. (3), the longitudinal
Substituting Eq. (4) into Eq. (3), the longitudinal pitching theory, coefficient
thetheory,vehicle cantheis vehicle rewritten
forced isto forcedas
move towith move twobewithquickly
two achieved.
different different
kinematic Now,
kinematic the longitudi
parameters. paramete Th
Substituting
moment Substituting
Substituting coefficient Eq. Eq.Eq. (4) (4)can (4)
into into into
be Eq. Eq. Eq. (3),(3),
(3),
rewritten theas
the the longitudinal
longitudinal
longitudinal pitching pitching
pitching moment moment
moment coefficient
coefficient
thecoefficient
stability can cancan
derivatives be
beberewritten rewritten
rewritten
can beasas as
quickly achieved. difference
Now, the of the unsteady or steady aerod
Substituting Eq. (4) theory, the vehicle
into Eq. (3), the longitudinal is forced to
differencepitching move with
of the unsteady two
moment different coefficient
or coefficient
steady kinematic
aerodynamic parameters.
can be rewritten forces and Then, with the
asmoments, all the stability derivatives can
CmSubstituting Cm 0  (CmEq.
�Substituting Eq. (4) 2
2C (4) into intoEq. Eq.(3), (3),thethelongitudinal
mq ) 0 sin  t  (C m  C mq ) 0 cos  t
longitudinalpitching pitchingmoment moment difference
longitudinal coefficient
of the can
difference
stability can
unsteady be
of betherewritten
derivatives rewritten
unsteady
or steady asor
(5)
are
(5)
as steady aerodynamic
aerodynamic
taken these
forces
as examples new and differential
forces methods
and moments,
tomoments, in detail.
all
all the stability the stab de
C C  ( C 22 Cmq ) 0 sin t  (Cm  Cmq ) 0 cos t (5)
2
(5)(5) be quickly achieved. Now, the longitudi
CC �m C 0 
m m� C m 0m
m m
(C 0(C  m
0
mm  C mq
 C
m 
) )
difference  sin sin 0 t t
 (
ofthe C( C  CC ) )  coscos0 t t
t  unsteady
(Cm  Cmqorbe steady aerodynamic forcesthe andlongitudinal
moments, new all the stability (5)derivatives can as examples to introduce
m mq m mq
Cm � Cm 0  (Cm   mq
2 0
Cmq ) 0 sin
 0 m m
  mq mq
)
0 0
quickly
0 cos  t
achieved. introduce Now, these
quickly stability
differential methods
derivatives in detail.
are taken 1) Calculation of combined aredynamic asderiv
be quickly be achieved. achieved. the Now, (5)(5)the longitudinal stability derivatives astaken ex
CmC
InIn
�m C
the
�the C0 m0case
m
case
(C(mC m
that  
that  2C2mq
ωt=2nπ
Ctmq ) )02sin
n0 sin
and
 tand
t (C( Ctm C
ωt=2nπ+π/2,
m 2mq
(a) Initial nCmq) )
mesh 0 /cos
Eq. 02cos
Eq.
,(5) t t (5) are respectively
are (b)1) Calculation
Dynamic expressed
mesh after of combined
rigidasNow, longitudinal
motiondynamic derivative
stabilitythesederivatives
new differential are taken
methods examples
in detail.
InInthe In thecase the case
casethat that be that t  t
t quickly 
2n2n and 2 n  and and
achieved.
  
t t 2n2Now,t 2 n
n    
/the
these
 / 2 , Eq.
longitudinal
, new
2/ 2, Eq. Eq.(5) (5) (5)
are
differential are respectively
stability
arerespectively
respectively
methods derivatives expressed
expressed
expressed
in detail. areasas taken as as examples to inintroduce
respectively In the case expressed that ast  2n and  t 2n Fig.  / 21.,Rigid Eq. (5) are respectively newexpressed
The differential
these these new
differential method differential
as methods solves inmethods
the combined
detail. The differential method solves the co
detail. derivatives
InInthethecase Ccase that that  C t  t 2
 n 2 n  and and   t
 t 2 n 2 n 
 
 /
 2 / ,
2 Eq.
, Eq. (5) (5) are
dynamic are respectively
respectively
mesh technique expressed
expressed as as 1) Calculation of combined dynamic deri
Cm  Cmq � Cmmmttt 222nnn these Cmmm000 new differential methods in detail.
1) Calculation of combined dynamic (6)
C   Cmq
CC 2t n2 nk CmC0m 0 1) derivative
Calculation of combined (6)(6) derivativedynamic derivative
CmCm  CCmq� �
m
m mq mmt 
Cmt  k2 n 0 C 1) Calculation of combined dynamic (6) The differential method solves the co
Cmmq Cmq � CkmC k0tm20tn2 00n CmC m0
(6)
CmC m CmqCmq� � 0 m0
k 01) Calculation of combined dynamic The differential derivativemethod solves theThe combined differential derivatives
method
(6) (6)solves
by enforcingthe combined the aircraft derivativesrotating by to the
enforcing the air
kk0 0 The differential method solves the combined derivatives by enforcing the aircraft ro
2 Cmm
tt
 2n 
n  /2  Cm
 /2 m00
Cmm   Cmq
mq
2  The
2 differential method solves the combined derivatives by enforcing the aircraft
(7)rotating to the
 00 6 6
6
5
55 6
5
(a) Rotating to  with5 q
5 (b) Rotating to  with q2
5 1
Fig. 2. Differential method toFig. 2. Differential
calculate combinedmethod to calculate
stability combined
derivative stability
with two derivative
different withangular
pitching two different pitching
velocities

angular velocities

177 http://ijass.org

(175~185)16-062.indd 177 2017-07-03 오전 9:51:02

 3. Dynamic derivative
3. Dynamic
calculation
derivative
method calculation method

3.1 Traditional CFD methods

3.1 Traditional CFD methods

Traditionally the dynamic

Traditionally
derivative
the dynamic
is obtained
derivative
by forcing
is obtained
the airc

around the center ofaround

gravity,
thewhich
centerisofsimilar
gravity,to which
wind tunnel
is similar
testing,
to w
Int’l J. of Aeronautical & Space Sci. 18(2), 175–185 (2017)
c e s a n d m o m e nc tess aarned l imn oe m a re nf tusn catrieo nlsi n oe fa r a sf su on cc it ai ot ends
same angle of attack  by two different angular velocities q1 and derivativeq2 , which of is lagshown of wash in and
Fig. indicates
2. the aerodynamic moment increment caused b
derivative of
derivative of lag lag of of wash wash and and indicatesindicates the the aerodynamic
aerodynamic moment moment increment increment caused caused bb
derivativeconverted For
of lag lagexample, of wash wash the indicates
and longitudinal
For example, the pitchingthe longitudinal
aerodynamic moment moment coefficient
pitching
incrementmoment iscaused
normbc
by enforcing the aircraft rotating to the same angle of attack derivative
downwash
ofexpansion angle of attack of of rate downstream and
. Since the form of  is similar to the pitching angularb
applied indicates from thethe aerodynamic
wing. Hence moment
it increment caused
The unsteady moment coefficients are obtained by a Taylor series of angle of attack attack during ratein the
same
same by angle angle of attack
of attack angular  by by two two differentqangular
different angular velocities
1 and qvelocities
qq11 andand
isof qqangle , which
22 , called
which of function is shown
is shown
of inthe
rate  ..of
Fig.
Fig. Since
Since 2.lag of
2. the
the formand
form of of
of indicates isis similar
 pitching similar to the
to the pitching
pitching angular angular v
 ..angle function oftheattack, of ,angle
the ofthe angular
attack, velocity,
, pitching q ,anguand
∆α two different velocities 2, which also the derivative wash
of angle
of angle of of attack rate
attack rate  Since
Since the form
form of  is is similar
similar to the
to thepitching
pitching angular
angular v
is
pitching, shown in
formulated Fig. 2. The unsteady
as coefficients C moment  C
coefficientsm 0by a Taylor seriesthe
are the
the the
free
free
aerodynamic free
free stream
stream stream
stream moment velocity
velocity velocity
velocity is is
increment is
unchanged,
unchanged,
is unchanged,
caused
unchanged, the
theby the
single
the single
the single
unit
single derivative
derivative
change derivative
derivative CC C m  is difficult
C m is difficult be
is
is difficult
difficult toto bt
to
The unsteady moment
Cm asthe 2 aret 2 n obtained expansion during the applied applied

Caerodynamic
mm
derivative
derivativeofofobtained laglag pitching,
The ofofwash unsteady
wash formulated
andand moment
indicates
indicates coefficients
the � m
aerodynamic are
 /2
obtained
moment
moment by
increment
incrementa Taylor caused
causedseries by the
by theexpansion
thethe free
freeunitunit stream
stream
changechange during velocity
velocity
the isis unchanged,
unchanged, the
(7)
the single
single derivative
derivative C C m is
is difficult
difficult to
to
by a Taylor series expansion  mq  during the applied ofthe angle
the
freeof stream
free attack variables,
streamrate
velocity
velocity . Since andisis qthe unchanged,form ofisthe
variables,
, which
unchanged, defined
theisand single
similar
single q as, to derivative
which the is defined
derivative
Cm is asdifficult to
mm

pitching,
pitching, formulated
formulated as
as   C q 0 the free stream velocity is unchanged, the single derivative CCmm isis difficult
difficult to t
Cpitching, �
pitching, C  C
formulated
formulated    C
asas traditional
traditional
traditional
traditional
pitching methods,
methods,
angular methods,
methods, butbut
velocity the
butthe
but differential
differential
the
the q differential
when (8)
differential the method
method method
method
free can
can
stream effectively
effectively
can
can effectively
effectively
velocity address
address address
addressthis
thisthis
problem.
problem.
this problem.
problem.
ofofangle
angleofofattack attackmpitching,
1 rate rate m 0 formulated
Cm1 Cm0  Cm   Cm 1  Cmq q1
  . .Since
mSince
 the
theas form
mform
 of
1 of  mq
 is is
1 similar
similar to to thethe pitching
pitching angular
angular velocity
velocity
traditionalmethods,
traditional q q when
when
methods,but butthe thedifferential (8)
differentialmethod methodcan caneffectively
effectivelyaddress addressthis thisproblem.
problem.
traditional ofCmethods, Cm 0single but Cmotion. the differential
(8) C C Cmm0method can
 effectively
 mtoˆbe (  ,Cqaddress
) q  Cmq this
 problem.
ˆ (  , q
C C C where
 C k c / 2CV qis the non-dimensional reduced isfrequency traditional
unchanged, the the
methods, applied but thederivative
differentialmm C
Here CisC

methodCmmq qdifficult
 Cmq
can qC
effectively address q
this problem.
00     traditional methods, but mthe differential method can effectively address this problem.
C
C m
m 11 C
C mm 
 C
C mm 
 
 
 
 C
C m
m 
 
 
 11  Cmq

derivativeC
derivative
mq q
q 11
of lag
of lag
of wash
of wash and and (8)
indicates
indicates Similar
theSimilar
the Similar
Similar
aerodynamic toto
aerodynamic the
thetoto
m
calculation
thecalculation
the moment calculation
calculation
moment of
increment of (8)
combined
combined
(8) of
of
increment combined
combined
caused dynamic
caused dynamic
by
mq
dynamic
dynamic
by
the derivative,
thederivative,
unit derivative,
derivative,
unit change change we
we we
mq
enforce
enforce
we enforce
enforce the
the the aircraf
aircraft
aircraft
the aircrafttran
aircrafttra
the
the free free streamstream C
Cm 2 mmvelocity �� C
�1 Cm 0mm00 Cisis  C m    C
  Cmmm 
  
the  C
1 single q Similar tothe the calculation (8) (9)
ofcombined
combined dynamicderivative, derivative, weenforce
enforcethe theaircraft
aircraft
Cm1velocity munchanged,
unchanged, the single
mq 11 derivative
2 q derivative CC   isis difficult
difficult to to be
obtained be
Similar obtained
obtainedinto traditional inincalculation methods, of but the differential dynamic method we

C  mC    C 21  C
mq
mq mm Similar to the calculation of combined
(9) dynamic derivative, we enforce the aircraft
2 m0 m
represents m
the 2
pitchmq 2
- damping moment coefficient, Similar
Similar which to to the theiscalculation calculation
t qethis rand m of
e
of d(combined
ˆwhere
combined
two a, qsC )two dynamic
out-
dynamic ˆderivative,
derivative, when we enforce
we enforce theterms
the aircraft
aircraft
C
Cmm22 C Cmm00  C Cmm    C Cmm   ofC q where C with q
are high-order
andtime q (timerates , rates
q )of
terms ofare which
of high-order
are often discar w
of22 angleCmq angleq22of attack of attack raterate  .Since . (9)
Since thethesame
formcan
same same
form ofeffectively
additional
additional
of additional
is
additional isangle
similar address
angle
similar angle
to ofof toattack
 the attack
of
the problem.
attack
pitching (9)
with
(9)
pitchingwith with two
angular angular different
different
different
different
velocity
velocity time time
qwhen rates rates angle
angle angleof
of attack
attack
of attack1 1,,
C C  C    C    C mq
q same
same additional angle
angle
mq of
of attack
attack (9) with two
two
mq
different time rates of
of angle
angle of
of attack
attack 1111
traditional
traditionalmethods, methods, Cmm22 �but
Note � but
Cmmthe
that 00theC
Eqs. differential
m and
differential
(8)  Cm(9)m 
method
method  22  can
have Ccan mqq
already effectively
effectively
22 discarded address
address thethis thisproblem.
higher terms.same
problem. sameSince
Similar additional
the
additional to rate
the ofangle angleof
angle
calculation of ofattackattack
attack
of (9)
combined with
with two
two different
dynamicdifferent time
time
derivative, rates
rates of
of angle
angle of
of attack
attack 
 
Notethat thatEqs. Eqs.of(8)
m
(8) and and
phase (9)
(9)have
mq
already
derivative, alreadydiscarded and thedue higher to terms. same the Since additional the
reason rateangle of angle that of attack of attack
attack  with and two different
different
q time rates rates of of angle
angle of of attack
attack  1
11
Note
Note that Eqs. (8) and (9) have
have
already discarded
discarded
the
the
higher terms.
noting
noting same
noting
Since that
that additional
the that
the the rate the
change
change of angle
change
angle ofof of of
angle of
angle attackangle ofof with
ofdifficult
attack
attack two
attack should
should should bebe betime
transformed transformed
transformed toto thatto that
that that
ofof ofcan addition
additional
additional vev
Similar
Similartotoequals the Note
thecalculation
calculation that Eqs.
of (8)
ofcombined
combined and (9) the
have
dynamic
dynamic the free free
already stream
derivative,
derivative, stream
discarded velocity
we we velocitythe
enforce
enforce is
higheris
unchanged,
unchanged,
terms. we Sincethe
noting enforce the single
the thatsingle fects.
rate thethederivative
of Then
derivative
aircraft
angle
change theof C unsteady
translating
of C
attack
angle is fects.
is of moment
difficult to Then
attack the to thecoefficient
to
be
same
should unsteady
obtained
obtained
additional
be canmoment
in
transformed be
in simplified
coefficient
to asof additiona
be
additiona sim
Note
Note toterms. that Eqs.
pitching
that Eqs. (8) the
angular
(8) and velocity
and (9)
(9) have
have already
when
already ofdiscarded
the free stream the higher is the
higher the
to aircraft
unchanged aircraft
terms. noting translating
translating
Since
then
noting the
we that
that rate
obtain tothe
tothe theofthe angle of
change
change of of attack
ofattack
m 
angle of
angle
 m 
of attack
attack should should be be transformed
transformed to to that
that of of additiona
additiona
whendiscarded the terms. Since the rate of angle
higher Since rate of angle attack equals noting that the change of angle of attack
equals to pitching are angular
difficult velocity
to separate, the free
this derivative stream isisalso unchanged
named angle noting asthen combined
of we
attack
that obtain
the with dynamic
change two different ofderivative.
angle time
of ratesshould
Moreover,
attack of
should anglebe be oftransformed
attack
transformed
to
to
that
that
of
of
additiona
additiona
equals
equals to
to pitching
pitching angular
angular velocity
velocity when
when the
the free
freeof stream
isstream is unchanged
isdifferential
unchanged noting
then
then we
we that
obtain
obtain the change of angle of attack should be transformed to that of additiona
attackrate,rate, given
given Itin incan Fig.Fig. 3.3.
pitching angular velocity when the free stream unchanged rate,
same
sameadditional
additional equalsangle
angleto toofof attack
attackwith
pitching withtwo
angular two traditional
different
different
velocity traditional
when timetime methods,
the methods,
rates
rates
free of but but
angle
stream thethe
angle isofof differential
attack
unchanged method
 ,,method
rate,
,then
1then
rate, 2.2.we we .given
Itcan
given
It
given is
obtain
is isC worth
worth in
effectively
m effectively
in
worth
in
Fig.
�Fig.
Fig.
Cnotingm 03.
3.address
3.
Cmaddress thatthis C
the Cthis �
problem.
mchange m0 CofC
C
problem. q  ofCattack
angle m   Cmq q

 
equals

1we
 q q
pitching angular velocity when the free stream is unchanged 1 rate,
rate, given obtain
given dynamic in Fig.
in Fig. 3.derivatives comprised 3. (10) m mqm

1 
then 1obtain
Cm   2C mq known as in-phase derivative is also should a rate,combined (10) ofvelocity and its
   qq111 rate, given
given
be in
transformedin Fig. Fig. 3. 3. to that (10) of additional
  1  SimilarSimilar to the
to the
calculation
calculation of combined
of combined When
Whendynamic When
Whendynamic the
the the
twotwo
derivative,
the derivative,two
two rates
rates rates
rates of
weof weangle ofenforce
angle
enforce
of angle
(10)
angle ofof the of
attack
attack
the
of attack
aircraft
aircraft
attack 1 and
and
1translating and
22 
translating
1 and are
are
to 2the are
given,
toare given, given,
thegiven, the
thethe the
same
same same
same instant
instantaneo
instantane
instant
noting
notingthat thatthe thechange  11 
change
1
 qqof ofangle angleofofattack attackshould shouldbe betransformed
transformedtotothat thatof ofadditional
(10) additional rate, velocity
velocity
When
When given the and
the
in and The
Fig. two its
two its 3. relationship
rates
rates of of (10)
angle
angle
(10) between The
ofof relationship
attack
attack angle  1111 of and
and attack
between
  2222 are andgiven,
are angle
pitching
given, oftheattack
the angular and
same instanta
same veloc
instant pi
2  q 11 When the two rates of angle of attack  and
(11)of attack 1 and 2 are given, the same instant  are given, the same instanta
2q
a static derivative C and a rotational derivative C When the
When
. the two two rates rates of of angle angle (11)
of attack   and   are given, the same instanta

  qq22 same same additional
madditional angle angle of of attack attack with withtwotwo
angle
angle different
When
 of ofdifferent ofthe
attack
attack time two time rates
rates rates ofof of
angle
angle angle of
of of attack
attack attack  ,and
1,12 .It
1
2 .areisIt worth
isgiven,
2 worth
2
(4)isisunder defined
defined by by (11)
2 2
rate,
rate,givengivenininFig. Fig.  2 

angle
mq
attack is defined by
 3.
23.2  qq22 (11) the
angle
angle
angle
angle same of
of
of
attack
ofinstantaneous
attack
attack
attack is
is
is defined
isdefineddefined
defined
harmonic by
by
by
additional
(11)
(11)(4)
by(11) movement.
under 1harmonic
angle of attack is
movement.
defined
Rearranging Eqs. (8) and (9), we obtain
2 Originals of illustrations used in the text (preferably with
angle captions
ofbe attack on a separate sheet or well
Rearranging Eqs.ofIn (8)
the and above (9),
noting
 noting
and weobtain
method,
and obtain
that 2that
the the
the
change combined
change oftheof angle dynamic
angle ofinstantaneous
of
attack angle
derivatives
attack
 should
 should
1t1
of
1t1
attack
additional can be
transformed be isis defined
transformed defined
simulated to by by
to through
that thatof of additionalthe identi
additional velocity velocity andand its its
When
Whenthe thetwo two rates
ratesofofangle angle of attack
attack 2 areare given,
given, the same
same instantaneous additional
Rearranging
Rearranging Eqs.
Eqs. (8)
(8) and
and(9), 
(9),
(9), we
we obtain by    11tt111   0 sin( t )    0 sin(t )
Rearranging Eqs. (8) and 1 1 we obtain  
Rearranging Eqs. Eqs. (8) (8) and (9), (9), wetheobtain
obtain     1t1t1t1
Rearranging separated
and from we illustration to enable scanning).       1

Cm1 � Cm0  Cm   (Cm  Cmq )q1 (12)
 
1 1    
C 1 CCm0CCm fication
 (C(CmofCaerodynamicCrate,
)q)rate,
q1given given inforces
Fig.
in Fig. and3.moments after several
3. 
cycles’ 1t1unsteady  0 cos(motion, t )  q but (12) the 0 cos( t )
computat q
angle
angleofofattack attackis CCismmmdefined
11defined Cmm00 by Cmm  
by  (Cmm  Cmq mq )q11
mq (12)    
2

2t2t 2
 212ttt122   (12)
(12)   (15)
Cm1 �� C Cm0 C Cm ((C Cm C Cmq))qq1  
   2 2     2
 sin(  t )  (12)
q
(12)      2
 sin(  t )  q
22tt22
C m1 m0 m m mq 1 
     t 0 0
Cm 2 � Cm 0  Cm  ion  (Ccost m  C mq ) q2When When thethough
thetwotwo rates rates of of angle angle of of
attack
attack 1  and
 2 2   
122tt2and  are are
 given, given, the t ) the (13)
samesame instantaneous
instantaneous
hast ) additional additional
 1t1t1 CC
Cmmm222 C
C C
Cmmm000
C 
Cmmm ((C (C is
Cmmm
C Chuge.
Cmq )q)2q2
Even the harmonic
(13)
balance method
Combined
Combined Combined
[15]
2  (15)
with

with
2 (15) or
with
2time
thethe
 0 spectral
the
relation
sin(
relation relation (13)
method
(13) (13) 
 [16]   0 sin( (16)
mq )q2
mq
Cm 2 �� C Cm 0 C Cm ((C Cm C Cmq))qq2 Combined    with
    the
 relation
cos(  (13)
t
(13) )       cos(  t )
C m2 m0 m m mq 2 Combined
Combined
Combined
 withwith
 with the relation  the
the 0 relation
relation  0
ItIt is obvious that that the theused onlytoangle angle
differencesof attackoftime attack iscosts
between defined
is defined by the
these by two equations Combined
Combined are the with with unsteady the therelation relation moment
 2t22t2 Itis is
It obvious obviousbeen
is obvious thatthe
that only
the only differences
only reduce
differences
differences between
between between these
with these two
these twounsteady
two equations
equations
tan
tantan  Combined
calculation
tan are
 tan(
tan( are
the 1t1(16)
tan( with
transformed
(16)
the
unsteady
1t1)unsteady
)
 1 the
unsteady
VV 11/ /VVVVrelation
moment Vfrommoment time domain t
It is obvious that the only differences between these two equations 
 are  the tan(  11tt111))   1 //V
moment
1 
ItIt is is obvious
obvious that the only onlymoment differences between these thesethe two equations
equations are
 the  unsteady
 moment 
equations are thethat the
unsteady differences between
coefficients and two tan
tan
tan  are
(12)  tan(  the tan(
tan( Substituting
unsteady

1t1Eq.
1 t
1 1t )
1 )   
) (13),  VV1 / Vwe caninto
V1 /
moment
1
1 /
V Eq.
V  (4) Substituting
Eq. (3), the Eq. (4) longitudinal
into (17) Eq. (3), pitching
the longitudinal
moment coe p
coefficients and the terms of pitching    1t1
1t1
angular velocities, Subtracting Eq.
tanEq. and  from
tan(  1tt1)the )  V(13),
combined /VVcan

(15) (15)
Combined
Combinedterms with coefficients
coefficients
with
coefficientsthethe
ofrelation
relation
pitching andothe
and
and thethefrequency
termsof
terms
terms
angular of domain,
ofpitching
pitching
pitching
velocities, it
angular is still
angular
angular
Subtracting notEq.
velocities,
velocities,
velocities, a systematic Subtracting
Subtracting
Subtracting
(12) method
Eq.
Eq.tan
  
(12)
(12) (12) from
from
 tan(
22t2t2
only 
from Eq.Eq. Eq. 
(13),
(13), V 1 /we
we can we form
can of stabili
coefficients and and the the termsterms of of pitching
pitching angular angular velocities, velocities, Subtracting Subtracting tan
tan 
tan
Eq.   
(12)tan(
 tan(  tan(
from )) 1
2
2Eq.tt
1
)V
 V
(13),//VV22 we
1
//VVcan

can
tan   tan( 2)
(VV
22
coefficients Eq. (12) from �2Eq. (13), C222m/we
VV CCmqm )�0 Csin m 0  m  mq ) (18)  tt (Cm  Cmq ) 0 cos 
2
tantan     tan( Cm
tan( t2C22) 
2t2t2)m)0  V /V t(C(mC 2Cmq 0 sin
0 cos
from Eq.
finally yield (13), the we can
following
ty finally
derivatives equationyield
  
are  the
to
  t  following
calculate
obtained.
 t the equation
combined to dynamic tan 
derivative:   tan(    V / (16) (16)
tan
tan tan(
tan(1finally
finally
11 )
tfinally
t )    V yield
V yield
/
1yield V/ V the
the following
following equation
equation
1   the following equation to calculate the combined dynamic tan
to2 2 calculate
to 2 2
calculate the thecombined
combined dynamic
dynamic derivative:
tan
derivative:  
derivative:
  tan(
tan( (17)
 (17)
  2
t2
2t 2)  )    V V 2
2/ V / V
finally yield
calculate
finally yield
the combined the following
the following dynamic equation
equation to calculate
derivative:
to calculate the the combined
combined dynamic dynamic Then
Then derivative:
Then
derivative:
Then thethe the the
additional
additional additional
2
additional 2 velocities
velocities
2 velocities
velocities
 are
areare are
Then Thenthe
Then the the additional
Inadditional
additional the case velocities that
velocities
velocities  are
t In
 are the
2aren case and   t 2nt  2n/2 , and
that Eq. (5) t are 2nrespectively
  / 2 , E
CmC1  C  C 2 To deal with the
Combined
Combined difficulties
with with the of
the estimating
relationrelation the dynamic Then the
derivatives, additional a set velocities
of new are
approaches
tan
tan tan(
tan(C2mt22Ct)2) C VC V
2� 2/ V/V CCm11 1CCmmm22 2
m
Then
Then the
the (18) (18)
additional
additional velocities
velocities(14)
(14) (14) are are
C 
 mqCC  Cmm
mq � q1 m C 11 q
qqq  CC (14)    
V V  V V tan( V
tan(   tan( t t ) )   V
t V )  
 t
V t  t (14)
 tan(111t111)  V
 11V11 V    1111 11t 11  (19)
m  mq
C
m 
Cmm C Cmq mq m 1
2q 22m 2
1 qq
m 2 1111
tan( t t1)))VVVC 1t t1 (14) (14)
mq � paper.   VVV111 VVVtan(
  Cm 0 C  Cm 0 (17)
1
qq111 qqare 2 proposed tanand 
tan investigated
tan( tan(1t1)1 t1 ) Vin1/VV this
1/ V  tan( Cm111t11C �1m11t111t  2 n (17)
m
Cm  Cmq � mt  2 n
Then
Then the the additional
additional velocities velocities are 22
are  
V1V V 1 V  tan( V tan(  t ) mq
11t11)  V11t11 k 0 V  t k 0
  
V V  V tan(
tan(
V   
tan( t t )  )   tV V )  Vt t  t
where
where where
where q qq ql  ql / 2//V22V
ql VVisisis the
the
is the
the non-dimensional
non-dimensional
non-dimensional
non-dimensional (a) Initial pitching
pitching
pitching
pitching mesh angular angular
angularvelocity,
angular velocity,
velocity,  22l
(b) l lV isis
is 222 the
Dynamic the theVthe reference
reference
tan(
reference 22 22 22after
mesh  2 t222) 
length length
2V
rigid
length 
 222t 2of
2 motion 22 of
of the
theof the (20)
where isisthethe non-dimensional  pitching V2/angular velocity, is222lthe is reference length the the
 2 2
whereqtqqtql
1t1where
ql//22V
ql / 2V is the non-dimensional
tantan
non-dimensional  tan( tan( tpitching
 2pitching

2 )2 t 2 ) angular
V2/angular
V velocity,
V velocity, 
 
llVV V
is the V tan(
VVreference tan(
reference
tan(  222tt2t22)))length
(19)  VVV
length 22t2t2t2of
2of the (18) (18)
 VV1 1 VV  tan(
tan( velocity,
t)1 )VV   1l 1is 1 the reference length of the computational model,   V V tan( (19)   t )  V   t
Fig. 1. Rigid dynamic mesh For Vtechnique
2 small Vsmall tan( 2t 2)  V 
disturbance, t22 aircraft
 22aircraft aircraft isapproximately
approximately in uniformly
computational
computational
computational
V is the free model, model,
3.2 New
model, andand
and VCFDVVisisthe
is the
methods
the free
free
free stream
stream
tostream
stream calculate velocity.
velocity. combined and singleFor For
For For small
2For small
dynamic disturbance,
smallderivativesdisturbance, 2disturbance,  aircraft
disturbance, aircraft is approximately in the uniformly
2 isis approximately
approximately
is inin the
the
in uniformly
uniformly
the accelerative
accelerative
accelerative
accelerative moti
motm m
and computational
computational
computational
stream
model,
model, velocity.
and
and V Vis isthe
model, and V is the free stream velocity. the
Then free
Then free the stream
the additional velocity.
velocity.
additional velocities
velocities are arethe For
uniformly small
small disturbance,
disturbance,
accelerative aircraft
aircraft
motion. is
is approximately
approximately
Thus, the in
unsteady in the
the uniformly
uniformly accelerative accelerative m m
 VV V V tan(
tan(   t t ) )  V V  t t For small (20)(20)disturbance, aircraft is approximately in the uniformly accelerative m
22  2 22Now, 2   it 2 22 2is clear that subtraction used to identify the
ForFor small small disturbance,
disturbance, aircraft isis approximately approximately in in the the uniformly
uniformly accelerative accelerative m m
Now, Now,
Now,it it itis is isclear clear
clearthat that subtraction
thatsubtraction
Different subtraction
from traditional used
used
used to to identify
to identify
identify
CFD the
thethe
the
methods, stability
stability
stabilityin this derivatives
unsteady
unsteady unsteady
derivatives
derivatives
moment
paper
unsteady moment
moment
the is is
iscoefficients
momentjust
just just
differential
moment the
coefficients
coefficients
the the reason
reason are aircraft
reason
coefficients
methods
coefficients why
are
whyare
only why only
areonly
are
are related
only
relatedrelated
employed
only related
related to toto   
to
to and
and
and
 and
 
and  ,,,expressed
expressed
 ,, expressed
expressed as
as as as
Now,
Now, it
it is
is clear
clear that
that subtraction
subtraction used used to to identify
identify the stability
stability derivatives
derivatives is just is just
the the
reason reason why why and 
ForForsmall
stability
smalldisturbance,
Now,derivatives
disturbance,aircraft
it is clear isthat 
justsubtraction
aircraftisisapproximately

the reason
approximately
 V 1 V V
1  used
why tan(
V
inin

tan(
the
to t 
 differential
the
)
 
1 1 identifyt
uniformly
)
V
uniformly
  V

1 1  11 method t  thet
1 1 stability derivatives
accelerative
accelerativeexpressed
unsteady
unsteadyismoment
unsteady
motion.
motion. Thus,
Thus,
moment
moment
as just thecoefficients
the
coefficients
the coefficientsare
coefficients
reason why are only are only
only relatedrelated
related to toto    and and
(19) (19)  ,, expressed
5 , expressed
expressed as asas 5
is differential
differential method
method is called.
iscalled.
called. In
In comparison
comparison with
with the traditional
traditional small
small unsteady
amplitude
unsteady
amplitude moment
moment oscillation
oscillation coefficients method,
method, are only
are only related related to to  and and   ,, expressed
expressed as as
differential
called. Inmethod
differential
differential method
comparison
methodis
method
is
calculate with
isiscalled.
called.
called.
In
thethe
In
comparison
combined
In
traditional
comparison
comparison
andsmall with
all the
with with the
amplitude
the the traditional
single dynamic
traditional
traditional
small
C
smallC 11C
small
mm
amplitude
derivatives.
C C C mm00Cm
�
m111 amplitude
amplitude 
C  C C
m000m
Firstly,
 
Cm oscillation

C
oscillation
m   


moscillation

  C
basedC
 mmethod, Cm
method,
m 
C 
on
 
method, small
method, 
11 111  disturbance
unsteady
unsteadymoment moment
differential
oscillation coefficients
coefficients method, areareonly only
there 
related
related
 In
is noto2toV  comparison
V V tan(
V
2  and
periodic 
tan(
and
 t
2   
2 
with
) t
 )
V  the

2,2,expressed
calculation V

expressed
t
 2 2 2in  traditional
t 2 astheas
small C Camplitude
mm
� C C m
 oscillation

Cmmm111 � Cmmm000  Cmmm    Cmmm  111 C C m
    C C
m
m 
 
   (20) (20)
(21)
there there
thereis is isnono noperiodic periodic
periodic theory, calculation
the vehicle
calculation
calculation inin
in the the
theis new new
forced
new approach, approach,
to movewhich
approach, with two
which
which indicates
indicates
indicates CCmm1that
different that �� CCkinematic
1that it
itmmit00is isisC aaCahighly highly
m 
highly  
parameters. C
efficient
efficientCm 
efficient 1Then,
1 with the
   
 C new there
there is
is
approach,
is no
no no periodic
periodic
periodic which calculation
calculation
indicates
calculation For
in
that
inForin the
the
small the
it new
is
new
small new
a approach,
highly approach,
approach,
disturbance,
disturbance,
efficient which
which
aircraft which
aircraft
CC
indicates
indicates
indicates
is m C
approximately
is
C
approximately
m 2 2 m C C
that
that
2 m �m 0 0C
C
that 
itit C C is
m00 in in
m 0
is
m
mit  a
C aCthe
 mhighly

is 
highly
m the
 a
 
 
C
uniformly
C
highly efficient

efficient

uniformly
m m  C 
C m
m 
efficient
accelerative
 accelerative

2 2 22 motion.
motion. Thus, Thus, (22)
the the
 CCmmm
2 n  /2 m0
C �C Cmm00(21) C Cmm  C 222derivatives can
m22 
CCmmmand
m m
CCm1m1��CCmm0 0method. CCmmethod. m   CCmm   11 (7) m
22 � C
m
(21)
C
m

m  2
0 method.

method. difference of the unsteady or steady aerodynamic forces C 2
� moments,
C m0
 C m all
  the

angle  stability
C   
CSince C  to Cadditional  with  Cangle ofattacks
method.
method.
method. (a) Rotating to  with q1 (b)
m
2 �
2 Rotating m 0
m
m  qm2  of
m    2
2) Calculation of acceleration unsteadyunsteady
derivative moment moment coefficients
coefficients areare only only Since
Since
related related
m Since
Since tothe
the the
to thethe
additional
madditional
0additional
and additional and  , angle
angle
expressed
, expressed
angle of ofof
2attacks
attacks attacks
as attacks
as at final
atat at final
final
final
at final
times
times
times times
times will
will
willwill will
bebebe the be the
the
be the
same,
same, same,
same, the
the the
transition
transitio
the trans
trans
CCmm2 2��CCmm0 0CCm2) 2)  CCmm  
mCalculation

Calculation of2be
of 2acceleration
quickly
acceleration derivative
achieved.
derivative Now, the longitudinal stability the Since
Since
Since
derivatives
same, the
the
the the additional
(22)
are additional
(22)
additional
transition taken as angle
angle
angle
examples
acceleration of
of
of attacks
attacks
attacks
to introduce
derivative at
at
at final
final
final value times
times
times can will
will
will
be be
be the
be the same,
the same, the
same, the transi
the trans
transi
s the non-dimensional reduced 2) 2) Calculation
2)
The frequency
2) Calculation translation
Calculation
Calculationof
of ofacceleration
the applied
acceleration
acceleration
ofacceleration derivative
motion. derivative
Fig. 2. Differential
acceleration derivative Here
derivativemethod C  C
represents
m to calculate the
combined stabilitySince derivative thewith additional two different angle pitching ofsubtracting
attacks at at Eq. final times will be the same,same, the the trans
trans
Cderivative Since the additional angle of attacks final times will Eq.be the
mq

finite time of aerodynamic loads m1C� m1C

at mC
�the 0  0Cm
mtail Clag m CmC
changes  m 1 1 derivative
in
derivative
derivative
obtained value
value
by value can
subtracting cancan be beobtained
obtained
beEq. obtained (21) byby subtracting
fromby subtracting
Eq. (22): Eq. (21)
(21)
Eq. from
(21)
(21) from
(21) fromEq.
(21) (22):
Eq.(22):
Eq. (22):
(22):
Since
Sincethe theadditional
additional The translation
angle
The translationthese angle ofof acceleration
attacks
attacks at at final
new differential
acceleration derivative
final times
derivative times will
methods C
will
C m in
m  be berepresents
thethe same,
same,
detail. angular
represents the the finite
the
the finite transition derivative
time
transition
time
velocities
derivative
derivative of aerodynamic
acceleration
acceleration
of aerodynamic valuevalue
value canloads cancan be
loads
loads
bebe obtained
obtained
obtainedat the byby subtracting
subtracting
at the by subtracting Eq. (21) from Eq. (22): Eq.
Eq. (21)
(21) from
from Eq.
Eq. (22):
(22):
h - damping moment coefficient, The The
The translation
translation
translation whichacceleration is t e r derivative
acceleration
acceleration m e d aC
derivative
derivative CsmmmC represents
represents
out-
 represents thethe
the
the finite
finite
finite time
time
time derivative ofofaerodynamicaerodynamic value can can loads
be at at
obtained theat thethe by subtracting
subtracting Eq. Eq. (21) (21) from from Eq. Eq. (22):
(22):
CofCmm11aerodynamic loads at the
The translation acceleration derivative C represents finite derivative
time of value be obtained by
C C � C � C  C  C
 m
     C  C     CC C C11maerodynamic
22C Cmm22 loads
(22) (22)
derivative
derivativevalue value can
can be beobtained
obtained 1)by C C C C m m

m
C
inby subtracting
subtracting
Calculation ofEq. Eq. (21)
combined (21)from from dynamicEq. Eq.(22): (22):
derivative C  C
m1 m2
tail
tail lag
lag changes
changes in downwash
downwash
m
converted
2 m 2
converted
m 0 m 0
downstream
m 
downstream
m  m 
from
from the
m  2
the wing.
2
wing.
mm 
C mHence
m 
Hence
�
 �
�11C m m1 1 it
112 itC  mmis
ism22 also also called called the the
CCmmmHence
m
vative, and due tail to tail laglag lag the
changeschanges reasonininindownwash
downwash that converted 
converted and downstream
downstream q from
from the
thethe wing.
wing. Hence
Hence 2
ititCC
itis is 22 also
22m 2is alsocalled called
called thethe
CC
m
tail changes 
tail lag changes indownwash
downwash converted
converted downstream
downstream from from the wing.
Cwing.
C � 
Hence
 
m
m111111   2
also
m2 2it is also called the
the
CCm1m1CCm 2mderivative of lag of The
wash differential
and Since
indicates Since
method the thetheadditional
additional
solves
aerodynamic the angle angle
combined
moment of of attacks
attacks
derivatives
increment atmmfinal
at � final
by
caused  times
11enforcingtimes
 by2 will will be the be
the thesame,
aircraft same, thethe
rotating transition
transition
to the acceleration acceleration
derivative of lag of wash and indicates the aerodynamic moment increment caused
    by 2 the the unit
unit change
change
CCmm ��is also named as combined dynamic derivative. Moreover,
ate, this derivative
2
ˆ ˆ (23)(23) ˆ ˆ
where
where where11     ˆ 1
  l l /2
/ 2
1ll //by 2 V V and
and    ˆ   l l // 2 2VV indicate
indicate the
the8 non-dimensional
non-dimensional rates
rates of
ofangle
angle
11 22VVEq. ˆ 2 22 2ll //22VV indicate
1 12 2 derivative of lag of wash and indicates the aerodynamic 77 moment increment 1caused the and
unit 22 change indicate the non-dimensional rates of an
ˆ1Eq.
   
of angle of attack rate  . derivative
Since derivative
the form value value
of can canbeis be
obtained
similar obtained to bythe by whereangular
subtracting
subtracting
pitching Eq. (21) (21) 1from
velocity from  and
 Eq. q (22): (22):
when 8 8 the non-dimensional rates of ang
of angle of attack rate  . Since the form of  is similar to the pitching angular velocity q when 2 2 
8
as in-phase derivative is also a combined dynamic derivatives comprised of 6 8
of angle of attack rate  . Since 8C the
8m1Cform of  is similar to 3) 3)Calculation
the Calculation
pitching
3) Calculation angular ofofrotary rotary
of velocityrotary derivative
derivative q when
derivative
mC1 m 2Cm 2 3) Calculation of rotary derivative
CmC�m � 7 (23) (23)
1   2  2
1 
7
m and a rotational derivative Cmq . (a) Translating with  (b) Translating with 
88 8
71 2 8
8 8
hod, the combined dynamic
Fig. derivatives
3. Differentialcan be simulated
method Fig.
to calculate through
3. Differential the stability
method
acceleration identi
to calculate acceleration
derivative stability
with two derivative
different rateswith two different
of angle of attackrates of

mic forces and moments after several cycles’ unsteady motion, but the computat angle of attack

DOI: doi.org/10.5139/IJASS.2017.18.2.175
en though the harmonic balance method [15] or time spectral method [16] has 178
22

time costs with the unsteady calculation transformed from time domain t

it is still not a systematic method and only the combined form of stabili

btained. (175~185)16-062.indd 178 2017-07-03 오전 9:51:07

  CCmm 
CCmm1 1 CCmm0 0CCmm   1 1 motion
(21)
(21)
Cm1 Cm0  Cm    Cm  1 motion and
and further
further identify
identify the single
single dynamic
the (21) dynamic derivative
derivative C .
mq .
Cmq

 CCmm 
CCmm2 2 CCmm0 0CCmm   2 2 (22)
(22)
Cm 2 Cm 0  Cm    Cm   2 When
When the
the vehicle
vehicle travels
travels at speed
speed V
at(22) V ,, two
two different
different pitching
pitching angular
angular velocities
velocities
rotating frame, it is possible to transform the equations of fluid motion to the moving frame such that
Since
Sincethethe additional
additionalangle
angleofof attacks
attacks atat final
finaltimes
timeswill
willbe
bethe
thesame,
same, the
the transition
transitionacceleration
acceleration
Since the additional angle of attacks at final times will be theadded same,to
added achieve
tothe the
the pull-up
transition
achieve motions.
motions. As
acceleration
pull-up As aa consequence,
consequence, the
the fields
fields of
of the
the relative
relative
steady-state solutions are available.
derivative
derivativevalue
valuecan
canbe
beobtained
obtainedby Baigang Mi
bysubtracting
subtractingEq.Calculating
Eq.(21)
(21)fromDynamic
fromEq. Derivatives
Eq.(22):
(22): of Flight Vehicle with New Engineering Strategies
derivative value can be obtained by subtracting Eq. (21) from Eq. (22): airplane
airplane are
are curved,
curved, and
and the
the radiuses
radiuses of
of the
the cases
cases are
are
CCm1m1CCmm2 2 3. Dynamic derivative calculation method
CCmm C C V
V (23)
(23)
Cm 1 1m12 2 m 2 rr1  (23)
(23) 1  q (24)
1   2 3.1 Traditional CFD methods q11
whereˆ1ˆ11l1l/ /22VV and
where
where and ˆˆ2 22l2l/ /22VVindicate
and indicate
indicatethe thenon-dimensional
non-dimensionalrates
non- rates
V ofofangle
angleofofattack. attack.
where ˆ1  1l / 2V and ˆ 2   2l /Traditionally 2V indicatethe the dynamic
non-dimensionalderivative
rr2  Vratesisofobtained angle of by forcing the aircraft moving(25)
attack. sinusoidal-y
dimensional rates of angle of attack. 2  q
3)3)Calculation q 2
angular velocity derivative Cmq isofofalso
Calculation
3) Calculation rotary
rotary
The
of derivative
derivative
called
angular
rotary the velocity
derivative rotary derivative,
derivative representing
C is also the called 2the rotary derivative, representing the
3) Calculation of rotary derivative around the centermq of gravity, which is similar to wind tunnel testing, and the aerodynamic for
The angular velocity derivative Cmq is also called the rotary AsAs the steady characteristic of the movement, the motion
88 As the the steadysteady characteristic
characteristic of of the
the movement,
movement, the the motion
motion of of airplane
airplane cancan bebe
ments of aerodynamic forces and moments
derivative, increments
due to pitching
representing of aerodynamic
the c e s a
angular velocity.
increments n d
forces mand
of aerodynamico m
Since e
8 moments
n t s a r
the pitching e
due l i n e
to pitching
of airplane a r f u n
canangular c t i o n
be transformed s o f a
velocity. Since s s o c i a t e
the pitching
to a reference d f l
frame,i g hknown
t parameters.
forces and moments due to pitching angular velocity. asreference
the method
reference frame,
frame, of known
rotatingas
known the
asreference
the method
method frame,of
of rotating
illustrated
rotating reference
in
reference frame,
frame, illustrated
illustrated in
in
For example,the the aerodynamic
longitudinalaerodynamic pitching forces
forces
moment and and moments
moments
coefficientare
are only
isonly function
function
normally ofofpitching
pitching angular
angular
as velocity,
avelocity,expresse
expres
nt produced by the additional aerodynamic moment forcesproduced
always seems by the toadditional
impede aerodynamic
motion, Cforces
is
Fig. 4.always
aerodynamic
Now the seems
forces toand
aerodynamic impede
momentsthe motion,
forces areandonly Cmq is formulated
function
moments of pitching
are only angular velocity, expr
Since the pitching moment produced by the additional mq
aerodynamic
aerodynamicforces forcesand andmoments
momentsare areonly
onlyfunction
function99of ofpitching
pitchingangular
angularvelocity,
velocity,expr
expr
aerodynamic forces always seems to impede functionthe ofmotion,
the angle CC
Cmq of attack, m1m 1 
function CC m 0m,
 0  Cpitching
C
ofpitching  q1q1 angular
angular velocity,
velocity, expressed
q , and as rate of these two
the
as damping derivative. To calculate thisnamed stability
as damping
derivative, derivative.
the effectTo of calculate
the rate of this
angle ofCm1 derivative,
stability  Cm0  Cmqthe
mqmq
 q1effect of the rate of angle of
is named as damping derivative. To calculate this stability CCm1  CCm0 CCmq qq1 (26)
m1 m0 mq 1
derivative, the
 should be eliminated. The first method effect of the
attacktoattain rate
should of angle variables,
thisbegoal of attack
eliminated. 
is to subtract and
should
The first q
be, which
C m method C
from the Cis
mto CC
2m 2defined
m 0m
attain Cas
0 this C q2q2 is to subtract C m from the
goal
Cm 2  Cm0  Cmq  q2mqmq
eliminated. The first method to attain this goal is to subtract CCm 2  CCm0 CCmq qq2 (27)
m2 m0
Cmethod mCmq,,   aSubtracting
Subtracting Eq.Eq. ˆmq((26)
(26) 
2
from
,from) Eq.
Eq. (27),
(27),we weobtain
obtain (2)
ned dynamic derivative Cm from Cmq , the
but combined
here
combineda newdynamic
differential
dynamic m C m 0C
derivative mC
is proposed butCtomhere
directly Cnew q  C
differentialq   
method q is proposed to directly
derivative but mqSubtractingEq.
mq   Eq.(26) (26) from Eq. (27), we obtain
Subtracting
Subtracting from Eq. (27), we obtain
here a new differential method is proposed to directly SubtractingEq. Eq.(26) (26)from fromEq. Eq.(27),
(27),we weobtain
obtain
ate Cmq . ˆ CC m1m 1  CC
calculate Cmq. calculate Cmq . where C mq q and  (  , q )mqmq C C 
are  C  C
m 2m 2
 Cmqhigh-order

q qC 1m
m1
q2 q2C
terms
m2 which are often discarded due to their small ef
(28)
A typical motion only related to pitching angular velocity CCmq 1 C q
m11 Cm 2
1  q2m 2
ypical motion only relatedistothepitching angular–Auptypical
velocitymotion
is theonly
steady related
pull –to uppitching
without
duringangular elevator mq
velocity qq1 the
is  qq2steady pull – up without elevator
steady pull without elevator fects.
angle, Then
becausethe unsteady moment coefficient 1  can 2 be simplified as
Where qˆ1qˆ1q1ql1/l 2/ V
Where
Where 2V and and qˆ qˆ q ql /l 2/ V
and 2V represent
represent
representthe the
the non-dimensional
non-dimensional pitc pit
the pull-up maneuver, the aircraft experiences a steady state Where qˆ1  q1l / 2V and2 2 qˆ2 22 q2l / 2V represent the non-dimensional
because during the pull-up maneuver, angle, the aircraft
becauseexperiences
during thea pull-up
steady state maneuver,
of rotationthe non-dimensional
aircraft
in Where experiencesˆ
q ˆ  q l a
/
pitching2 steady
V andstate
angular ˆ
q of  rotation
q
velocities.
ˆ l / 2 V in represent the non-dimensional p
of rotation in vertical plane, while the Cm angle
� Cm 0of C m  and
attack  Cm   CmqWhere q q11  q11l / 2V and q22  q22l / 2V represent the non-dimensional (3)
velocities.
velocities.
Now the new differential methods can be summarized as
al plane, while the angle velocity
of attackremain
and velocity
vertical
the same. remain
plane, while
the same.
Apparently, thethere
angle
Apparently,
is of
no attack
such thereandisvelocity
ideal no such velocities.
remain the same. Apparently, there is no such
The relationship between velocities.
the
angle flowofdiagram
velocities. attack in Fig.
and 5. Starting
pitching with velocity
angular steady cases, all the
are presented in diagram
Eq.
movement in factual flight. However, this motion can still be NowNowthe thenew newdifferential
differential methods
methods cancanbebesummarized
summarized asasthe
theflow
flow diagramininFig.Fig.5.5
movement in factual flight. However, this
ideal motion
movement can still
in factual
be usedflight.
to compute
However, thethis combined
rotary
motion Now can the
and new
stillsingle
bedifferential
dynamic
used to methods
derivatives
compute can
the be
can summarized
be
rotary calculated as the flow diagram in Fig
used to compute the rotary derivative. Now
Now the the new
new differential methods can be summarized as the flow diagram in Fig
the steps. differential methods can be summarized as the flow diagram in Fig
(4) under harmonic movement. following
steady
steady cases,
cases,allall thethecombined
combinedand andsingle
singledynamic
dynamicderivatives
derivativescan canbebecalculated
calculatedfollowin
followi
Both the steady and unsteady methods can be used to steady cases, all the combined and single dynamic derivatives can be calculated follo
tive. derivative.
simulate the motion. When we use the  unsteady
 sin(t ) approach steady
steadycases, cases,all allthethecombined
combinedand andsingle
singledynamic
dynamicderivatives
derivativescan canbebecalculated
calculatedfollow
follo
0

h the steady and unsteady withmethods can be
rigid dynamic Both
used
mesh theto steady
simulate

technique, andthe
the
 unsteady
motion.
 0 cos(will
aircraft methods
t )When qideally wecanuse be theused to simulate the motion. When we use the
 
4.4.CFDCFDmodeling modeling (4)
    0 sin(
2
rotate around the fixed points in vertical plane, t) 
whilst q the
dy approach with rigid dynamic mesh technique,
unsteady the
approach
aircraft with
willrigid
ideally dynamic
rotate around
mesh technique,
the fixed 4. the
CFD modeling
aircraft will ideally rotate around the fixed
 rotating
   0 sin(reference
t ) 4.
steady method is established based on the 4.CFDCFDmodeling modeling
velocity
      cos(t ) 4.14.1HBS HBSstability
stabilityderivatives
derivatives
frame. In this
in vertical plane, whilst the steady method work,
points the
is established rotating
plane,on
in vertical based whilst
the rotatingof
the steady
0 unsteady
reference
methodframe. 4.1 HBS stability
is established based on derivatives
the rotating reference frame.
approach is applied to the frame instead of the aircraft, 4.1
4.1HBS HBSstability
stabilityderivatives
derivatives
WeWefirstlyfirstlyanalyze
analyzethe theHyper
HyperBallistic
BallisticShapeShape(HBS) (HBS)model model[17] [17]totoverify
verifythe thederivative
derivati
work, the rotating velocity of unsteady
which is readilyapproach
Intransformed
this work,is applied
the rotating
to the
to steady Substituting
velocity
case. frame
Then, instead
ofthe Eq.
unsteady (4)the
of
steady into Eq. (3),We
approach
aircraft, isthe appliedlongitudinal
firstly analyze
to the frame pitching
the Hyper
insteadmoment
Ballistic coefficient
Shape
of the aircraft, (HBS) canmodel
be rewritten
[17] to as verify the deriv
We
We firstly analyze the Hyper Ballistic Shape (HBS) model [17] to verify the deriva
firstly analyze the Hyper Ballistic Shape (HBS) model [17] to verify the deriv
method is adopted to calculate the motion and further methods
methodsininthis thiswork.
work.The Theconfiguration
configurationisisa aslender slenderbody bodyrevolution
revolutionwith withhemisphere
hemisphe
is readily transformed to steady case. which Then, is thereadily Cm � is
steady transformed
method  (C
Cmadopted
0 to m to
steady  case.
2
Cmq )Then,
calculate0 sin themethods
 (Csteady
t the in this) 0work.
m  C mqmethod cosis  tThe
adopted configuration
to calculate is athe slender body revolution (5) with hemisp
identify the single dynamic derivative Cmq. methods
methods in in this
this work.
work. The The configuration
configuration isis aa slender slender body body revolution
revolution with with hemisph
hemisp
two-stage
two-stageexpansion expansionskirt, skirt,shownshownininFig. Fig.6.6.The Themesh meshisisgenerated
generatedbybytransfinite
transfiniteinterpola
interpo
When the vehicle
n and further identify the single dynamic derivative travels
motion andCfurtherat speed V ͚ ,
identify two
In the
different
the single
pitching
dynamic t derivative two-stage expansion skirt, shown in Fig. 6. The mesh is generated by transfinite inter
mq . case that and(a)
2n two-stage C t . 2expansion
Steady / 2 ,skirt,
   with
npull-up Eq. (5)
q1 shown are respectively expressed as
shownin inFig.
Fig. 6. The
Themesh q2is
isgenerated
generatedby bytransfinite
transfiniteinter
mq (b) Steady
angular velocities q1 and q2 are added to achieve the pull-up two-stage expansion skirt, 6.pull-up with
mesh inter
thethealgebraic
algebraicmethod methodininthe theANSYS
ANSYSICEM ICEMCFD CFDsoftware,
software,and andtotal
totalmesh
meshsize sizeofof1.6
1.6m
motions. As a consequence, the (a)of
fields Steady
the pull-up
C
relative withair q C
flow Fig. 4.
the
(b)
Differential
algebraic
Steadymethod pull-up
to
method
with
calculate q in the
rotary
ANSYS
derivative with
ICEM
two different
CFD software,
pitching angular
and total mesh size of 1
velocities
en the vehicle travels at speed V , two different When the pitching
vehicleangular
travels
Cm  Cmq � velocities
at speed q 1, and
m t  2 nV

1
mtwo
0 q 2
Fig.
different
are
the 4. algebraic
Differential
pitching method
angular
method
the algebraic method in the ANSYS
tothecalculate
2 velocities
in ANSYS q rotary
1
and
ICEM q
derivative
are with two and total
ICEM2 CFD software, and total mesh size of 1.
CFD software, (6) mesh size of 1
past the airplane are curved, and the radiuses of the cases k 0 are used
usedtotodifferent
simulate
simulate both
boththe
pitching the static
angularstatic and
anddynamic
dynamiccases,
velocities cases,and andFig.
Fig.7 7isisthe
thesurface
surfacemesh meshofofthe them
Fig. 4. Differential method to calculate rotary derivative with usedtwo todifferent
simulate bothangular
pitching the static and dynamic cases, and Fig. 7 is the surface mesh of
velocities
to achieve the pull-up motions. As a consequence,
added to achieve the fields
the pull-up
of the relativemotions.airAsflow a consequence,
past the used
usedto simulate
tothe fields both
simulate of
both thethe
the static
staticand
relative air dynamic
and flow
dynamicpastcases,the and
cases, andFig.Fig. 77isisthe
thesurface
surfacemesh meshof oft
Dynamic
Dynamictesting testingwas wasperformed
performedininthe theisentropic
isentropiclight lightpiston
pistontunnel
tunnel(ILPT)(ILPT)atatthe the
Dynamic testing was performed in the isentropic light piston tunnel (ILPT) at
ne are curved, and the radiuses of the cases airplane
are are curved, and the radiuses of the cases are Dynamic
Dynamic testing was performed in the isentropic light piston tunnel (ILPT) at t
testing was performed in the isentropic light piston tunnel (ILPT) at
Southampton,
Southampton,which whichisisananintermittent
intermittentwind-tunnel
wind-tunnelfacility facilityproving
provingananopen openjetjettest
testsecse
Southampton, which is an intermittent wind-tunnel facility proving an open jet test
V Southampton,
Southampton, which which is an intermittent wind-tunnel facility
5 is an intermittent wind-tunnel facility proving an open jet test proving an open jet test

r1   exit
exitplane
(24) planediameterdiameter0.21m 0.21mwith witha aflow
flowMach Machnumber numberofofM=6.85
(24) M=6.85for fordurations
durationsofoftypic typi
q1 q1 exit plane diameter 0.21m with a flow Mach number of M=6.85 for durations of ty
exit
exit planeplane diameter
diameter 0.21m 0.21m with with aa flow
flow MachMach number number of of M=6.85 for durations of ty
6 6 M=6.85 for durations of ty
unit
unitReynolds
Reynoldsnumber numbertest testisisapproximately
approximately Re
 Re 0.72 0.7210
10 for
forthe
thecases
casesand andthethecence
unit Reynolds number test is approximately  Re 0.72 106 for the cases and the
V unit Reynolds number test is approximately Re
 0.72  10 6
6 for the cases and the

r2  (25)unit Reynolds number test is approximately (25) Re 0.72 10 for the cases and the

q2 q2 position
positionisis3.24d 3.24dfrom fromthe thehead headposition
positionofofHBS. HBS.
position is 3.24d from the head position of HBS.
position
positionisis3.24d 3.24dfrom fromthe theheadheadposition
positionof ofHBS.
HBS.
the steady characteristic of the movement, Asthethemotion
steady ofcharacteristic
airplane canofbethetransformed movement,tothe Although
Although
motionthe
a Although the steady
steadyaerodynamic
ofFig.5.
airplane aerodynamic forces
forcesand andtomoments
momentstesting testingdata
dataisisnot notoffered
offeredtoto
the Simulation steps of the new methodsanda moments testing data is not offered
steady can be
aerodynamictransformedforces
Although
Although the steady aerodynamic forces and moments testing data is not offered
the steady aerodynamic forces and moments testing data is not offered
nce frame, known as the method of rotating referencereference
frame,frame,
knownillustrated
as the method in Fig.of4.rotating Now CFDCFD
the results,
results,we
reference westill
frame, stillneed
needtotocalculate
illustrated calculate
Fig.the the
4.static
staticcases casestotobetter
betterunderstand
understandthe theflow
flowchara
cha
CFD results, we still need to in calculate Now
the staticthe cases to better understand the flow c
Fig. 5. Simulation steps of the new methods Fig.5. Simulation steps of the CFD
CFD results,
results, we still need to calculate the static cases to better understand the flow cc
new methods we still need to calculate the static cases to better understand the flow
9 provide
9provideinitial initialsolutions
solutionsfor forthe thedynamic
dynamicsimulations.
simulations.Fig. Fig.8 8shows
showsthe thecontours
contoursofofpressu
pres
provide initial solutions for the dynamic simulations. Fig. 8 shows the contours of pr
provide
provide initial initial solutions
solutions for for the
the dynamic
dynamic simulations.
simulations. Fig. Fig. 88 shows
shows the the contours
contours of of prpr
andandthe thestreamlines
streamlinesofofHBS HBSmodel modelatatseveral
severalangles anglesofofattack,attack,we wecan canclearly
clearlysee seethe th
and the streamlines of HBS model at several angles of attack, we can clearly see
and the streamlines of
179 and the streamlines of HBS model at several1010 HBS model at several http://ijass.org
angles of attack, we can clearly see
angles of attack, we can clearly see
10
10
10

(175~185)16-062.indd 179 2017-07-03 오전 9:51:10

 Int’l J. of Aeronautical & Space Sci. 18(2), 175–185 (2017)

4. CFD modeling angles of attack, we can clearly see the recirculating region
around the base of the model. For this configuration, a
4.1 HBS stability derivatives detached shock wave forms ahead of the body which contains
a normal shock The wave angular
perpendicular velocity derivative
to the directionCmq is also called the rotary
of free
We firstly analyze the Hyper Ballistic Shape (HBS) model Thestream, angular velocity derivative Cmq is also called the rotary derivative, represe
two weaker oblique shock waves on both sides and
[17] to verify the derivative computation methods in this a Mach wave far increments
from theofsurface.aerodynamic
The flow forces
behind andthemoments
bow due to pitching angula
work. The configuration is a slender body revolution withincrements shockofconsists
aerodynamic forces and
of supersonic, moments
transonic and due to pitching
subsonic gases angular velocity. Since the
hemisphere-cylinder and two-stage expansion skirt, shown moment produced by the additional aerodynamic forces always seems
and leads to a more complicated aerodynamic phenomena.
in Fig. 6. The mesh is generated by transfinite interpolationmoment produced by the additional aerodynamic forces always seems to impede the motio
Therefore, the unsteady characteristic of the flow is evident
named as damping derivative. To calculate this stability derivative, the
based on the algebraic method in the ANSYS ICEM CFD at high angle of attack, and the steady results may be not so
named as damping derivative. To calculate this stability derivative, the effect of the rate of
software, and total mesh size of 1.6 million cells is used to reliable and are for reference
attack  should only.
be eliminated. The first method to attain this goal
simulate both the static and dynamic cases, and Fig. 7 is theattack 1)should The effect of time step for unsteady
be eliminated. The first method calculation
to attain this goal is to subtract C m
surface mesh of the model. The combined combined dynamic derivative Cm  Cmq , but
dynamicderivative andhere a new differential m
Dynamic testing was performed in the isentropic lightcombined dynamic derivative Cm are
acceleration Cmq ,obtained
but here froma newunsteady
differential method is proposed t
piston tunnel (ILPT) at the University of Southampton, calculate
aerodynamic forces, andCmq one. of the most important factors
which is an intermittent wind-tunnel facility proving ancalculate Cmq .
for unsteady calculation is the time step, so we firstly analyze
open jet test section of nozzle exit plane diameter 0.21m with the effect of timeAstep typical
andmotion
find itsonly related values
reasonable to pitching
for theangular velocity is the stea
a flow Mach number of M=6.85 for durations of typically 0.5s. A typical motion only
simulation in this work. related to pitching angular velocity is the steady pull – up withou
angle, because during the pull-up maneuver, the aircraft experiences
The unit Reynolds number test is approximately Re=0.72×10 6
The longitudinal dynamic pitching cases are chosen as
angle, because during the pull-up maneuver, the aircraft experiences a steady state of ro
for the cases and the center of gravity position is 3.24d from examples. Thevertical
time steps plane,
we usewhilearethe angle
shown in of attack
Rable1 and andthevelocity remain the same.
the head position of HBS. vertical plane, while the angle of attack and velocity remain the same. Apparently, there i
residual variation can be seen in Fig. 9. The results indicate
Although the steady aerodynamic forces and moments that when the ideal
time step movement in factual
is 0.00005s, only 10flight.
steps areHowever,
neededthis motion can still be
testing data is not offered to compare with CFD results, we ideal movement in factual flight. However,
to keep the convergence of the inner iteration, while this motion can still
22 be used to compute t
derivative.
still need to calculate the static cases to better understand more steps are used when dt=0.0008s. The smaller the time
derivative.
the flow characteristics and provide initial solutions for the step, the less inner Both the steady
iterations and unsteady
required methodsand
for convergence, can be used to simulate th
dynamic simulations. Fig. 8 shows the contours of pressure Boththethegreater steadythe
andcomputational
unsteady methods can be used
cost needed, as theto result,
simulate the motion. When w
coefficient and the streamlines of HBS model at several we finally chooseunsteadythe timeapproachstep with rigid dynamic
dt=0.0002s, mesh technique, the aircraft will
inner iteration
unsteady approach
steps n=15 aswith rigid dynamic
calculation mesh technique,
parameters with the the aircraft will ideally rotate around
comprehensive
points in vertical plane, whilst steady method is established based on
points consideration.
in vertical plane, whilst the steady method is established based on the rotating referen
In this
2) Longitudinal work, thedynamic
combined rotatingderivatives
velocity of unsteady approach is applied to the
In this work,
The the
newrotating velocity
differential of unsteady
strategy is used approach is applied
to calculate theto the frame instead of th
which is readily transformed to steady case. Then, the steady metho
longitudinal combined dynamic derivative when the steady
which is readily transformed to steady case. Then, the steady method is adopted to calc
simulations are motionfinished. Two identify
and further differentthe pitching angularderivative C .
single dynamic mq

motionvelocities q1=0.2rad/s,
and further identify theq2single =0.4rad/sdynamicare imposed
derivativeto C the .HBS
mq
model to move When unsteadily to the same
the vehicle travelsadditional
at speed angle V , two of different pitching angul
Fig. 6. HBS configuration
Fig. 6. HBS configuration Fig. 7. Surface attack
mesh ofΔα=0.004rad.
HBS model The unsteady aerodynamic coefficients
When the vehicle travels at speed V , two different pitching angular velocities q1 an
during the calculation
added tosteps achieveare shown in Fig.
the pull-up 10, and As
motions. we can
a consequence, the fields of
added see that as the
to achieve time step
the pull-up increases,
motions. As athe aerodynamic
consequence, theloads
fields of the relative air flow
also exhibit aairplane are curved, and
linear relationship the radiuses
at these angles of of the cases are
attack,
airplane are curved, and the radiuses of the cases
which indicates that all the unsteady cases still satisfy theare
V
small disturbance r1  theory.
V q
r1  According to the 1new method, the combined dynamic
q1
derivatives can be obtained with the transient pitching
V
r2   at the final instantaneous angle of
moment coefficients
V q
r2  attack. Table 2 shows2 the simulation results of the combined
Fig. 7. Surface mesh of HBS model q2
g. 6. HBS configuration Fig. 7. Surface mesh
Originals of HBS
of tables model
used in the text (each on a separate sheet). As the steady characteristic of the movement, the motion of airpl
Table 1. Time steps used for testing As the steady characteristic of the movement, the motion of airplane can be transfor
Table 1. Time steps used for testing
reference frame, known as the method of rotating reference frame, i
Time step (s) 0.00005 0.0001
reference0.0002
frame, known0.0004 0.0008
as the method of rotating reference frame, illustrated in Fig. 4.
9
Inner iteration steps 22 22 22 22 22
(a)   0 (b)   5 9
Table 2. Calculation results of combined stability derivative

DOI: doi.org/10.5139/IJASS.2017.18.2.175 Cmq  cal Cm  Cmq  exp

Angle of attack (degree) Cm 180 Error
0 -17.9 -18.14 1.32%
5 -17.4 -17.58 1.02%
10 -16.7 -16.72 0.12%
15 -25.8 -25.52 1.10%
20 -60.63 -56.05 8.17%
(175~185)16-062.indd 180 2017-07-03 오전 9:51:11
 Tab.Tab.
2 also
2 also
shows
shows
thatthat
the the
calculation
calculation
errorerror
of C mC mq Cat
ofm C at 20 20
mq  is extraordinarily
is extraordinarily
increased,
increased,

while
while
the the
coefficients
coefficients
at this
at this
angle
angle
of attack
of attack
are are
linearly
linearly
varied
varied
withwith
the the
constant
constant
pitching
pitching
angular
angular

velocities,
velocities,
the possible
the possible
reason
reason
for this
for this
phenomenon
phenomenon
maymay
be that
be that
the flow
the flow
fieldfield
is heavily
is heavily
affected
affected
withwith
The angular velocity derivative Cmq is also called the rotary derivative, representing the
the flow
the flow
separation
separation
and and
shock
shock
at this
at this
angle
angle
of attack,
of attack,
and and
the unsteady
the unsteady
calculations
calculations
cannot
cannot
describe
describe
the the
Baigang Mi Calculating Dynamic Derivatives of Flight Vehicle with New Engineering Strategies
increments of aerodynamic forcesflows
complex
complex and moments
flows due to
so accurately.
so accurately.pitching
Thus,
Thus,
the theangular
results
results
mayvelocity.
may be Since
not not theatpitching
credible
be credible such
at such
a large
a large
or larger
or larger
angle
angle
of of
moment produced by theattack.
dynamic derivatives and they maintain good
additional
attack.
consistency
aerodynamic forces always seems to impede the motion, Cmq is at
flows so accurately. Thus, the results may not be credible
with tunnel testing data [17] at all the angles of attack. such a large or larger angle of attack.
named as damping derivative. To calculate
3) Longitudinal
3) Longitudinal this stability
translation
translation derivative,
acceleration
acceleration the effect of the rate of angle of
derivatives
derivatives
With the increment of angle of attack, the absolute value 3) Longitudinal translation acceleration derivatives
of derivative is matched,
attackwhich meansbetheeliminated.
 should improvement
To avoid
To avoid
The theoffirst
the methodTo avoid
interference
interference
to of the
of attain
pitching interference
pitching
thisangular
angular
goal of
isvelocity pitching
to velocity
during
subtract duringangular
C mthe the velocity
unsteady
unsteady
motion,
motion,
twotwodifferent
different
 from the
dynamic damping. during the unsteady motion, two different rates of angle of
Table 2 also showscombined
that the dynamic
calculation ratesrates
derivative
error ofofCangle
of angleof attack
of attack
m  Cmq , but attack
here1 anewrad
0.2 0.2 / s , /
rad
differential
1=0.2rad/s,
method
2  rad
0.4
s ,2 =0.4rad/s0.4rad / sproposed
is
are s are
/are
used used used
totodirectly
to force force
to force
the the the
body bodybody
to to translate
to translate
to to
at α=20 is extraordinarily increased, while the coefficients
o
translate to the same additional angle of attack Δα=0.004rad.
calculate
at this angle of attack Cmq . varied with
are linearly the the
the
same same
additional
constant additional angle
Theangle
of attack
of attack
additional    0.004rad
velocities  0.004rad. The
perpendicular. The
additional
additional
to free velocities
streamvelocities
perpendicular
speed perpendicularto free
to free
pitching angular velocities, the possible reason for this transformed from the rates of angle of attack are respectively
A typical motion only stream
stream
related speed
to speedtransformed
pitching transformed
angular fromfromtheisthe
velocity rates
theratesof angle
steady of pull
angleof
– up attack
ofwithout
attack
are elevator
are
respectively
respectivelyV1 V404
1  t404
1
and
t1 and
phenomenon may be that the flow field is heavily affected ΔV1=404t 1 and ΔV2=808t2, the unsteady pitching moment

with the flow separation and shock at this angle of attack, coefficients during the dynamic processes are given in Fig.
angle, becauseFig. during V2pull-up
the V808
2 
, the
t808 t2 , unsteady
2 maneuver,
the unsteady
thepitching
pitching
aircraft momentmoment
experiences coefficients
coefficients
a steady during
during
state the
of dynamic
the dynamic
rotation processes
in processes
are given
are given
in Fig.
in Fig.
and the unsteady calculations cannot 6. HBS configuration
describe the complex Fig.
11. 7.Similar
Surface to
mesh of HBS
those of model
combined derivative calculation, all
vertical plane, while the11.
angle of attack
Similar
11. Similar toand
to those ofvelocity
thosecombinedremain
of combined the same.
derivative
derivative Apparently,
calculation,
calculation, allthere
all the is no such
coefficients
the coefficients
increase
increase
linearly
linearly
withwith
timetime

ideal movement in factual flight. However, this motion can still be used to compute the rotary
step.step.

derivative. TheThe
acceleration
acceleration
derivatives
derivatives
C m Ccan
m can
be calculated
be calculated
by using
by using
the the
newnew
differential
differential
method
method
and and

Both the steady and numerical

unsteady methods
numerical
results in can
resultsTab. be
in Tab.used
3 show to good
3 show
goodsimulate thewith
prediction motion.
prediction with Whendata
reference we
reference use[18].
data
[18]. Itthe
isItshow
is show
thatthat
the time
the time
raterate
unsteady approach with of
rigid
ofdynamic
angle
angle ofmesh
of attack technique,
attack
may may theadverse
havehave aircraft
adverse impactwillon
impactideally rotatestability
dynamic
on dynamic around thethis
stability
of fixed
of this
configuration
configuration
since
since
mostmost
points in vertical plane, whilst the steady
acceleration
acceleration methodare
derivatives is positive
derivatives established
are positive based
values. on
values.
Also thecan
Also
we rotating
we can reference
findfind
the same frame.
the same
characteristic
characteristic
as the
as last
the last
section,
section,
In this work, the rotatingthe
velocity of unsteady
calculation
the calculation
errors approach
errors
are is applied
increased
are increased
at high to the
at high
angles frame
angles instead
of attack,
of attack, of may
which the may
which aircraft,
be caused
be caused
by the
by same
the same
reason.
reason.
(a)   0 to steady case. Then, the(b)
which is readily transformed

  5 method is adopted to calculate the
steady


4) Longitudinal
4) Longitudinal
rotary
rotary
derivatives
derivatives

motion and further identifyUsing

theUsing
single dynamic
the the
rotating derivative
rotating
reference
reference . when
Cmqframe
frame when
simulating
simulating
the the
steady
steady
pull-up,
pull-up,
we we
chose
chose
fourfour
different
different

pitching
When the vehicle travelspitching
angular
at speed angular
V , velocities
twovelocities
q1 pitching
different q0.05 rad rad
1  0.05 s , / sq,2 velocities
/angular q0.1
2 rad / s ,q/ qs3 ,
0.1rad q0.2
and 3 rad
q0.2 / s /and
rad s and
q4 q0.4
4 rad / s /to
0.4rad s to
1 2 are

detailed
detailed
added to achieve the pull-up analyze
analyze
motions. theanew
As the new
differential
differential
consequence, method
the method
fieldsfor
of obtaining
for
theobtaining
the air
relative rotary
theflow
rotary
derivative.
derivative.
past the

airplane are curved, and the radiuses of the cases are

12 12
V
r1  (24)
q1

(c)   10 (d)   15

V
r2   (25)
q2

As the steady characteristic of the movement, the motion of airplane can be transformed to a

reference frame, known as the method of rotating reference frame, illustrated in Fig. 4. Now the
24
9

(e)   20 (b)   25

Fig. 8. The contours of pressure coefficient at different angles of attack

Fig. 8. The contours of pressure coefficient at different angles of attack

181 http://ijass.org

(175~185)16-062.indd 181 2017-07-03 오전 9:51:12

 The angular velocity derivative Cmq is also called the rotary derivative, representing the

increments of aerodynamic forces and moments due to pitching angular velocity. Since the pitching

moment produced by the additional aerodynamic forces always seems to impede the motion, Cmq is

named
Int’l as damping
J. of Aeronautical derivative.
& Space To175–185
Sci. 18(2), calculate this stability derivative, the effect of the rate of angle of
(2017)

attack  should be eliminated. The first method to attain this goal is to subtract C m from the
the coefficients increase linearly with time step. velocities q1=0.05rad/s, q2=0.1rad/s, q3=0.2rad/s and
combined dynamic derivatives
The acceleration derivative Cm can Cmq ,bebutcalculated
here a new bydifferential
q4=0.4rad/smethodtoisdetailed
proposed to directly
analyze the new differential method
using the new differential method and numerical results in for obtaining the rotary derivative.
calculate
table 3 showCgood mq . prediction with reference data [18]. It is Figure 12 shows the steady moment coefficients of these
show that the time rate of angle of attack may have adverse pitching angular velocities at several angles of attack, as
A typical
impact on dynamic motion only related
stability of thisto pitching angular
configuration since velocity
Originals is the
mostof tables the steady
used in the
linear text–(each
pull up without
relationship on abetween elevator
separate sheet).
aerodynamic coefficients
acceleration derivatives are positive values. Also we can find and angular Table velocities,1. Time rotary derivatives
steps used can be obtained
angle, because during the pull-up maneuver, the aircraft experiences a steady0.3 state of rotation in for testing 0.7 15deg-0.2rad/s
15deg-0.4rad/s
the same characteristic as the last section, the calculation with
Time step (s) steady results
0.00005 of arbitrary
0.0001 two cases.
0.0002 Table
0deg-0.2rad/s
0deg-0.4rad/s
4 shows
0.00040.6 the
0.0008 20deg-0.2rad/s
0.25 20deg-0.4rad/s
vertical plane, while the angle of attack and velocity Inner remain the same. stepsApparently, there 22 is no such 5deg-0.4rad/s
5deg-0.2rad/s
errors are increased at high angles of attack, which may be final
iteration values of
22 stability derivatives obtained
22 from both
22 0.5 the
22
25deg-0.2rad/s
25deg-0.4rad/s
10deg-0.2rad/s
0.2
caused by the same reason. 10deg-0.4rad/s 0.4

ideal movement(e)   20
4) Longitudinal in factual
rotary flight. However,
derivatives

(b) this
 25 motion can still
 be
TableTable
used to0.15 compute
2. Calculation
2. Calculation
theofrotary
resultsresults
of combined combined
stability stability
derivativederivative
0.3

Cm
Cm
0.2
UsingFig.
the rotating
8. The contours reference frameatwhen
of pressure coefficient differentsimulating
angles of attackthe
derivative. Angle of attack
0.1
(degree) Cm  Cmq  cal Cm  Cmq  exp Error
0.1
steady pull-up, we chose four different pitching angular 0.05

Both the steady and unsteady methods can be used to simulate the motion. 0 When we -17.9 use the -18.14 1.32%
0

5 0 -17.4 -17.58 1.02%

-0.1

unsteady approach with rigid dynamic mesh technique, the aircraft will ideally 10 rotate 0
around
100 -16.7
the
Timefixed
200 300
steps
400-16.72 500 0.12%
-0.2
0 100 200 300
Time steps
400

15 -25.8 -25.52 1.10%

points in vertical plane, whilst the steady 0.3 method is established based on the 20rotating
0.7 reference Fig. frame.
-60.6310. Unsteady -56.05
pitching moment
15deg-0.2rad/s 8.17%
coefficients vs time steps
15deg-0.4rad/s
0deg-0.2rad/s
0deg-0.4rad/s
25 0.6 -62.3 -61.2
20deg-0.2rad/s
20deg-0.4rad/s
1.80%
In this work, the rotating velocity of unsteady approach is applied to the frame0.5instead of the aircraft, 25deg-0.2rad/s
0.25
5deg-0.2rad/s
5deg-0.4rad/s 25deg-0.4rad/s
10deg-0.2rad/s
0.2 0.25 0deg-0.2rad/s 0.7 15deg-0.2ra
10deg-0.4rad/s 0.4
15deg-0.4ra
which is readily transformed to steady case. Then, the steady method is adopted
Table to calculate
3.0.3Calculation results ofthe
acceleration derivative0.6
0deg-0.4rad/s
5deg-0.2rad/s 20deg-0.2ra
0.15 0.2 5deg-0.4rad/s 20deg-0.4ra

Cm
Cm

10deg-0.2rad/s 0.5 25deg-0.2ra

0.2 10deg-0.4rad/s 25deg-0.4ra
motion and further identify the single dynamic
0.1 derivative Cmq . Angle0.15of attack (degree) Cm  cal Cm  ref 0.4
0.1

0.05 0
0 2.01 1.92 0.3
0.1

Cm
Cm

When the vehicle travels at speed V0 , two different pitching angular velocities 5 q and q2 are3.33 3.46 0.2
-0.1 1
0.05
-0.2
10 6.01 5.82 0.1
0 100 200 300 400 500 0 100 200 300 400 500
added to achieve the pull-up motions. As a consequence, Time steps
the fields of the relative
0
15air flowTimepast the3.98
steps 4.17 0

20 -11.31 -10.99 -0.1

Originals of tables used in the text-0.05(each on a vs

separate sheet).
airplane are curved, and the radiuses of the cases areFig. 10. Unsteady pitching moment coefficients
25 time steps -14.91 — -0.2

Fig. 9. The effect of time stepFig.9. The effect of time step 0 100 200
Time steps
300 400 500 0 100 200
Time steps
300 400

Table 1. Time steps used for testing

V
r1   0.3
0.25
0.7
Time 0deg-0.2rad/s
step (s) 0.00005
15deg-0.2rad/s
Table 0.0001
4. Calculation
0.7
15deg-0.4rad/s
(24)of0.0002
Fig. 11.results
Unsteady rotary 0.0004
derivative
15deg-0.2rad/s
pitching
0.0008
moment coefficients during plunging motion
q1 0deg-0.2rad/s 0.6 0deg-0.4rad/s 20deg-0.2rad/s
15deg-0.4rad/s
0.25
0deg-0.4rad/s
5deg-0.2rad/s
0.2
Inner iteration steps
5deg-0.2rad/s
5deg-0.4rad/s
22 0.6
20deg-0.4rad/s
25deg-0.2rad/s
22 22
20deg-0.2rad/s
20deg-0.4rad/s 22 22
5deg-0.4rad/s 0.5 25deg-0.2rad/s
10deg-0.4rad/s Angle of attack (degree) Cmq  Diff C mq  Sub Cmq  ref
10deg-0.2rad/s 0.5
25deg-0.4rad/s
10deg-0.2rad/s 25deg-0.4rad/s
V 0.2 10deg-0.4rad/s 0.4
Table 2. Calculation results of-20.43
combined stability derivative
r2  
0.15 0.4
0
0.2
(25) -19.91 -20.19
q2 0.15 0.3
0.3
5 -21.09 -20.73 -20.890.4
Cm
Cm

0.1
Cm
Cm

0.2
0.1
25 Angle of attack (degree) 0.2
10
0.15 Cm  C 
-22.66
mq cal C m 
-22.71
 C mq  exp
-22.71Error
0.3
As the steady characteristic of the0.05movement, the 0.1
motion of airplane0 can0.115be transformed -17.9 to
-30.03 a -18.14
-29.78 1.32%
-29.33
0.05 0 0 0.2
0 5 20
0.1
-17.4
-44.04 -17.58 -43.97
-49.32 1.02%
Cm

reference frame, known as the method of rotating -0.1reference frame, illustrated -0.1 in Fig. 4. Now the
Cm

15deg
0 0deg
10 -16.7 -16.72 0.12%
20deg
-0.05 -0.2
25
-0.2
-49.23 -47.39
5deg
10deg
— 0.1 25deg
0 100 200 300 400 500 0 100 200 300 400 500
Time steps 0 100 9200 300 400 Time steps
500 15 0.050 100 -25.8
200 300 400-25.52500 1.10%
Time steps Time steps 0

0.7
20 -60.63 -56.05 8.17%
0.3 15deg-0.2rad/s
Fig. 10. Unsteady pitching moment coefficients vs time steps25
15deg-0.4rad/s 0
-62.3 motion -61.2 1.80%
-0.1
0deg-0.2rad/s 0.6 20deg-0.2rad/s Fig. 11. Unsteady pitching moment coefficients during plunging
25
0deg-0.4rad/s
5deg-0.2rad/s
20deg-0.4rad/s
25deg-0.2rad/s
Fig. 11. Unsteady pitching
0.1moment
0.2 coefficients
0.3 during
0.4 plunging motion 0.1 0.2 0.3 0
5deg-0.4rad/s 0.5 q(rad/s) q(rad/s)
25deg-0.4rad/s
10deg-0.2rad/s
0.2 10deg-0.4rad/s 0.7
0.4
Table 3.15deg-0.4rad/s
Calculation
results ofresults of acceleration
derivativederivative
0.25 0deg-0.2rad/s 15deg-0.2rad/s
0deg-0.4rad/s
0.2 Table 3. Calculation acceleration
0.3 5deg-0.2rad/s 0.6 20deg-0.2rad/s
15
20deg-0.4rad/s
Cm

0.2 5deg-0.4rad/s
10deg-0.2rad/s 0.5
0.4
25deg-0.2rad/s 26
of attack (degree) Cm  cal Cm  ref
Angle25deg-0.4rad/s
0.2
10deg-0.4rad/s
0.1 0.15
0.15
0.1 0.4 0.3

05
0 2.01 1.92
0 0.3
0.1 0.2 5 3.33 3.46
Cm
Cm

0.1 0.2
Cm

-0.1
Cm

15deg
0
0.05
0deg 10 6.01 20deg 5.82
-0.2 0.1 5deg 0.1 21 25deg
0 100 200 300
Time steps
400 500 0 100 200 300
Time steps
400
0.05
500
0
10deg
15 3.98 4.17
0
-0.1
0
20 -11.31 -10.99
Fig. 10. Unsteady pitching moment
-0.05 coefficients vs time steps 25 -14.91 —
Fig. 10. Unsteady pitching moment coefficients0vs time steps -0.2 -0.1
0 100 200 300 400 500 0 100 200 300 400 500
Time steps 0.1 0.2 0.3 Time
0.4 steps 0.1 0.2 0.3 0.4
q(rad/s) q(rad/s)

0deg-0.2rad/s 0.7 15deg-0.2rad/s

Table 4. Calculation results of rotary derivative
0deg-0.4rad/s
5deg-0.2rad/s 0.6
Fig. 11. Unsteady pitching moment coefficients during plunging motion
15deg-0.4rad/s
20deg-0.2rad/s
DOI: doi.org/10.5139/IJASS.2017.18.2.175
5deg-0.4rad/s 182
20deg-0.4rad/s
25deg-0.2rad/s
10deg-0.2rad/s
10deg-0.4rad/s
0.5
25deg-0.4rad/s
26 attack (degree) C  Diff C  Sub
Angle of Cmq  ref
mq mq
0.4
0.2 0 -20.43 -19.91 -20.19
0.3
0.4 5 -21.09 -20.73 -20.89
Cm

0.2
0.15
0.1 0.3
10 -22.66 -22.71 -22.71
0
15 -30.03 -29.78 -29.33
(175~185)16-062.indd 182 2017-07-03 오전 9:51:15
20 -44.04 -49.32 -43.97
0.2
0.1
m
 0.3 0.7 15deg-0.2rad/s
15deg-0.4rad/s
0deg-0.2rad/s 0.6 20deg-0.2rad/s
0deg-0.4rad/s 20deg-0.4rad/s
0.25
5deg-0.2rad/s 25deg-0.2rad/s
5deg-0.4rad/s 0.5
25deg-0.4rad/s
10deg-0.2rad/s
0.2 10deg-0.4rad/s 0.4

0.15 0.3

Cm
Cm
0.1 Baigang Mi Calculating
0.2
Dynamic Derivatives of Flight Vehicle with New Engineering Strategies
0.1

0.05 0

differential and 0 subtraction methods, all the cases -0.1 show 4.2 Finner missile stability derivatives
-0.2
good consistency with calculation
0 100 200 data in reference [18] 0
300
Time steps
400 500 100 200 300
Time steps
400 500

Generally, the major computational cost for time-domain To further validate the methods, the stability derivatives of
.3 0.7
traditional methods with forced Fig. 10. oscillation
Unsteady pitching
15deg-0.2rad/s moment
is spent oncoefficients
the the Finner
vs time steps missile model are further simulated, which is also
15deg-0.4rad/s

25
0deg-0.2rad/s
0deg-0.4rad/s
0.6
simulations of periodic motions, and
20deg-0.2rad/s
this will not happen
20deg-0.4rad/s
a common case for testing. The configuration and surface
5deg-0.2rad/s 25deg-0.2rad/s
5deg-0.4rad/s 0.5
in the new differential
10deg-0.2rad/s methods to calculate25deg-0.4rad/s
combined0.7and mesh of the model are shown in Fig. 13 and 14 respectively
.2 10deg-0.4rad/s 0.25 0deg-0.2rad/s 15deg-0.2rad/s
0.4
single dynamic derivatives. For the 0deg-0.4rad/s
HBS model, a typical and the computational15deg-0.4rad/s
20deg-0.2rad/s
mesh is also generated by ICEM CFD
5deg-0.2rad/s 0.6
15 0.3 20deg-0.4rad/s
0.2 5deg-0.4rad/s
with a total size of 2.8 million cells. The center of gravity
Cm

calculation of0.2 combined dynamic 10deg-0.2rad/s derivative with time-

10deg-0.4rad/s
0.5 25deg-0.2rad/s
25deg-0.4rad/s
.1
domain method locates at 6.1 times the diameters from the nose along the
0.1 requires about 12 hours of CPU time on 18
0.15 0.4

05
processors with 0 about 1.6 million cells. However, we only use
0.3 longitudinal axis of the model.
0.1

Cm
Cm

0 8 hours to compute the combined and two single dynamic

-0.1
0.2 With the new differential methods, the steady cases
0.05
0 100 200 300 derivatives
400 500 with
-0.2
0 the100 new200 differential
300 400 methods.500
0.1
Hence, should also be firstly calculated under the condition of Mach
Time steps Time steps 0
0
they are really efficient ways to obtain detailed stability number Ma=1.96 and Reynolds number Re=0.187×106 based
-0.1

Fig. 10. Unsteadyderivatives.

pitching moment Moreover,
-0.05coefficients when
vs time the
stepspitching angular velocity on cone base diameter d. Then, all the other simulations to
-0.2
0 100
is replaced with yawing andTimerolling 200 300
steps
400 500
angular velocities, the 0 100 identify
200
Time steps
the combined
300 400 500 and single dynamic derivatives can

Fig.
be 12. Pitching moment
finished. Fig. 15 coefficients
comparesof thedifferent angles of attack
longitudinal during pull-up
combined,
new differential 0.7
methods can be extended to directional and
15deg-0.2rad/s
0deg-0.2rad/s
Fig. 11. Unsteady
lateral stability0.6derivative simulation. 20deg-0.2rad/s
0deg-0.4rad/s pitching moment
15deg-0.4rad/s coefficients during plunging
acceleration, motionand rotary stability derivatives when the angle
5deg-0.2rad/s
Originals of tables used
5deg-0.4rad/s in the text (each on a separate sheet). 20deg-0.4rad/s Fig.of12. Pitching moment coefficients of different angles
attack ranges from 0 degree to 30 degrees, the numerical of attack during pull-up
10deg-0.2rad/s 0.5 25deg-0.2rad/s
10deg-0.4rad/s 25deg-0.4rad/s
data
Fig. matches
12. Pitching moment verycoefficients
well with wind-tunnel
of different testing
angles of attack duringand
pull-up
Table0.4
0.2 1. Time steps used for testing
0.3
Time step (s) 0.00005 0.0001 0.0002 0.00040.4
0.0008
Cm

0.2
Inner iteration steps 220.15
0.1
22 22 22 0.3 22
0 0.2
Table 2. Calculation
0.1
results of combined stability derivative
Cm
Cm

15deg
-0.1 0deg 20deg
5deg 0.1 25deg
-0.2 10deg Fig. 13. Computational model of Finner missile
0 100 200 300 Angle
400 500 attack (degree) 100Cm 200
of 0.050 Cmq  cal  Cmq 500exp
300 Cm400 Error
Time steps Time steps 0

0 -17.9 -18.14 1.32%

Fig. 13. Computational model of Finner missile
Fig. 11. Unsteady pitching moment5 coefficients during plunging
-17.4 motion -17.58 1.02%
0 -0.1
Fig. 13. Computational model of Finner missile
10 0.1
-16.7
0.2
q(rad/s)
0.3
-16.72
0.4
0.12% 0.1 0.2
q(rad/s)
0.3
Fig. 0.4
13. Computational model of Finner missile
15 -25.8 -25.52 1.10%
2
20 -60.63 -56.05 8.17%
25 -62.3 -61.2 1.80%
0.4
26
5
0.3

1 Table 3. Calculation
0.2
results of acceleration derivative
Cm

15deg
0deg 20deg
5deg 0.1 25deg

5
10deg
Angle of attack (degree) Cm  cal Cm  ref
0
0 2.01 1.92
Fig. 14. Surface mesh
0 -0.1 5 3.33 3.46
Fig. 14. Surface mesh
0.1 0.2 0.3 0.4 10 0.1 0.2 6.01
0.3 0.4 5.82
q(rad/s) q(rad/s) Fig. 14. Surface mesh
15 3.98 4.17 -500

20 -11.31 -10.99 -450 Fig. 14. Surface mesh

Fig. 12. Pitching
26 moment25 coefficients of different
-14.91 angles of—attack during -400
-500
pull-up -350
-450
Dynamic derivative

-300
-500
-400
Table 4. Calculation
Table 4. Calculation results
results of rotary of rotary derivative
derivative -250
-450
-350 Exp Cmadat+Cmq
Cal Cmadat+Cmq
-200
derivative

-400
-300 Cal Cmadot
Cal Cmq
Angle of attack (degree) Cmq  Diff Cmq  Sub Cmq  ref -150
-350
-250 Ref[17]
Exp Cmadot
Cmadat+Cmq
derivative

-100
-300 Cal Cmadat+Cmq
0 -20.43 -19.91 -20.19 -200
Cal Cmadot
Dynamic

-50
-250 Cal
-150 ExpCmq
Cmadat+Cmq
5 -21.09 -20.73 -20.89 -200 0
Ref[17] Cmadot
Cal Cmadat+Cmq
-100 Cal Cmadot
10 -22.66 -22.71 -22.71
Dynamic

-15050 Cal Cmq

-50
Ref[17] Cmadot
15 -30.03 -29.78 -29.33 100
-100
0 -5 0 5 10 15 20 25 30 35
alpha(deg)
20 -44.04 -49.32 -43.97 -50
50

25 -49.23 -47.39 — 100

Fig. 15. Longitudinal
-5 0 combined
5 10 and
0
15single
20 dynamic
25 30 derivatives
35
Fig. 15.
50 Longitudinal combined and single dynamic derivatives
alpha(deg)
100
-5 0 5 10 15 20 25 30 35
alpha(deg)
Fig. 15. Longitudinal combined and single dynamic derivatives
27 http://ijass.org
183 Fig. 15. Longitudinal combined and single dynamic derivatives
27

27

(175~185)16-062.indd 183 2017-07-03 오전 9:51:16

 Int’l J. of Aeronautical & Space Sci. 18(2), 175–185 (2017)

reference data [18-20] across the full regime of angle of attack. [3] Pamadi, B., Performance, Stability, Dynamics, and
Dynamic values indicate that the longitudinal dynamic Control of Airplanes, AIAA Press, USA, 2004.
damping improves as the increment of angle of attack, [4] Williams, J. E., “The USAF Stability and Control Digital
while the proportions of two single dynamic derivatives are Datcom”, U. S. Air Force Flight Dynamics Lab., AFFDL-
different from each other. Different from aircraft, traditional TR-79-3032, Dayton, OH, 1979.
wings are not appeared in this configuration, only weak [5] Ronch, A. D., “On the Calculation of Dynamic
downwash flow caused by the blunt body affects the flow Derivatives Using Computational Fluid Dynamics”, Liverpool:
field of fins, therefore, the change of acceleration derivatives the University of Liverpool, UK, 2012.
is reasonable. [6] Greenwell, D. L., “Frequency Effects on Dynamic
Stability Derivatives Obtained from Small-Amplitude
Oscillatory Testing”, Journal of Aircraft, Vol. 35, No. 5, 1998,
5. Conclusion pp.776-783. DOI: 10.2514/2.2369
[7] Park, S. H., Kim, Y. and Kwon, J. H., “Prediction of
In this paper, the new differential strategies to simulate Dynamic Damping Coefficients Using Unsteady Dual-Time
the combined and single dynamic derivatives are presented. Stepping Method”, 40th AIAA Aerospace Sciences Meeting &
Based on rigid dynamic mesh technique, the combined and Exhibit, Reno, Nevada, AIAA 2002-0715, 2002, pp. 1-9. DOI:
acceleration stability derivatives are firstly calculated, while the 10.2514/6.2002-715
rotating reference frame approach is adopted when identifying [8] Park, S. H., Kim, Y. and Kwon, J. H., “Prediction of
the rotary stability derivative. The longitudinal dynamic Damping Coefficients Using the Unsteady Euler Equations”,
derivative calculation results of HBS model and Finner missile Journal of Spacecraft and Rockets, Vol. 40, No. 3, 2003, pp.
model match well with reference and experiment data. 356-362. DOI: 10.2514/2.3970
Different from traditional methods, the new differential [9] Oktay, E. and Akay, H., “CFD Predictions of Dynamic
methods can be used as convenient ways for the production Derivatives for Missiles”, 40th AIAA Aerospace Sciences
of advanced flight vehicle as the reason that lower Meeting & Exhibit, Reno, Nevada, AIAA 2002-0276, 2002, pp.
computational cost is needed and more detailed stability 1-11. DOI: 10.2514/6.2002-276
derivatives can be obtained for further flight dynamics [10] Ronch, A. D., Vallespin, D., Ghoreyshi, M. and Badcock,
analysis. Also the methods can be applied to directional and K. J., “Evaluation of Dynamic Derivatives Using CFD”, AIAA
lateral stability simulation by replacing the pitching angular Journal, Vol. 40, No. 2, 2012, pp. 470-484. DOI: 10.2514/1.
velocity with those of yawing and rolling. J051304
However, the presented work is just the beginning of the [11] Ronch, A. D., Ghoreyshi, M. and Badcock, K. J., “On
process to deeply analyze the dynamic stability characteristics the Generation of Flight Dynamics Aerodynamic Tables
of aircraft. The methods are still based on small-disturbance by Computational Fluid Dynamics”, Progress in Aerospace
theory with linear or weakly nonlinear flow at medium-small Sciences, Vol. 47, No. 8, 2011, pp. 597-620. DOI: 10.1016/j.
angle of attacks. Future research will focuses on improving paerosci.2011.09.001
their efficiency and accuracy and modifying the methods to [12] Murman, S. M., “Reduced Frequency Approach for
better fit the cases at high angle of attack. Calculating Dynamic Derivatives”, AIAA Journal, Vol. 45, No.
6, 2007, pp. 1161-1169. DOI: 10.2514/1.15758
[13] “ANSYS FLUENT”, ANSYS Inc. http:// www.ansys.com
Acknowledgement / Products / Fluids/ ANSYS-Fluent.
[14] Spalart, P. R. and Allmaras, S. R., “A One-Equation
The work was supported by the Aeronautical Science Turbulence Model for Aerodynamic Flows”, 30th AIAA
Foundation of China (Grant No. 2015ZA53013). Aerospace Sciences Meeting and Exhibit, Reno, Nevada,
AIAA-92-0439, 1992, pp. 1-17. DOI: 10.2514/6.1992-439
[15] Jeffrey, P. T., Chad, H. C., Earl, H. D. and Kenneth,
References C. H., “Unsteady Flow Computation Using a Harmonic
Balance Approach Implemented about the OVERFLOW 2
[1] Erkin, B., Dynamics of Atmospheric Flight, John-Wiley Flow Solver”, 19th AIAA Computational Fluid Dynamics,
Press, New York, 1972. San Antonio, Texas, AIAA 2009-4270, 2009, pp. 1-15. DOI:
[2] Robert, F. S., Flight Dynamics, Princeton University 10.2514/6.2009-4270
Press, Princeton, 2004. [16] Gopinath, A. K. and Jameson, A., “Time Spectral

DOI: doi.org/10.5139/IJASS.2017.18.2.175 184

(175~185)16-062.indd 184 2017-07-03 오전 9:51:16

 Baigang Mi Calculating Dynamic Derivatives of Flight Vehicle with New Engineering Strategies

Method for Periodic Unsteady Computations over Two- and of Beijing University of Aeronautics and Astronautics, Vol.
Three- Dimensional Bodies”, 43rd AIAA Aerospace Sciences 41, No. 2, 2015, pp. 222-227. (in Chinese) DOI: 10.13700/j.
Meeting and Exhibit, Reno, Nevoda, AIAA-2005-1220, 2005, bh.1001-5965.2014.0094
pp. 1-14. DOI:10.2514/6.2005-1220 [19] Uselton, B. L. and Uselton, J. C., “Test Mechanism
[17] East, R. A, Hutt, G. R., “Comparison of Predictions and for Measuring Pitch Damping Derivatives of Missile
Experimental Data for Hypersonic Pitching Motion Stability”, Configurations at High Angles of Attack”, AEDC-TR-75-43,
Journal of Spacecraft and Rockets, Vol. 25, No. 3, 1988, pp. Tennessee: AEDC, 1975.
225-233. DOI: 10.2514/3.25975 [20] Shantz, I. and Graves, R. T., “Dynamic and Stability
[18] Xi, K, Yan, C, et al, “Numerical Simulation of Individual Measurements of the Basic Finner at Supersonic Speeds”,
Components of Pitching-Damping Coefficient Sum”, Journal NAVORD Report 4516, 1960.

185 http://ijass.org

(175~185)16-062.indd 185 2017-07-03 오전 9:51:16