Memorandum
108852
December
1994
Center, Moffett
Field, California
TABLE
OF
CONTENTS
Page
LIST
OF FIGURES
SUMMARY
....................................................................................................................
................................................................................................................................
INTRODUCTION
......................................................................................................................
Background
.....................................................................................................................
Laminarflow
control .............................................................................................
Transition
Previous
Work
Current
GOVERNING
Mean
Work
...............................................................................................................
.................................................................................................................
..................................................................................................................
EQUATIONS
.....................................................................................................
Flow ......................................................................................................................
Coordinate
transformation
.....................................................................................
Thinlayer
approximation
......................................................................................
BoundaryLayer
Linear Stability
Equations
Equations
11
Compressible
stability equations
...........................................................................
Solution of the eigenvalue
problem .......................................................................
13
stability
METHODS
Flow
block
diagonal
BoundaryLayer
Equations
BoundaryLayer
Stability
COMPUTATIONAL
GRID
algorithm
17
...................................................................
ADI algorithm
...........................................................
.............................................................................................
Equations
AND
...............................................................................
BOUNDARY
Grid Configurations
surface
ADI
CONDITIONS
.............................................
..............................................................................................
generator
(WSG)
.................................................
.............................
Stability
Reynolds
STABILITY
DISCUSSION
ANALYSIS
.................................................................................
.................................................................................................
Automation
Validation
.....................................................................................
Number Effects ...............................................................................................
AngleofAttack
Effects
..................................................................................................
iii
P_liiOU_n"Ni
16
17
......................................................................................................................
PulliamChaussee
Wing
equations
.........................................................................................................
BeamWarming
RESULTS
9
10
11
NUMERICAL
Wing
4
6
.........................................................................
Incompressible
Mean
.............................................................................................
...............................................................................................
18
18
18
18
19
19
19
19
20
20
21
21
21
23
F"N.ME_
Reynolds
Number Effects with Angle of Attack ............................................................
Sweep Effects ..................................................................................................................
CONCLUSIONS
AND RECOMMENDATIONS
Conclusions
Recommendations
REFERENCES
.....................................................................
.....................................................................................................................
...........................................................................................................
...............................
.............................................................................................
24
25
26
26
27
28
APPENDIX
A .............................................................................................................................
59
APPENDIX
B .............................................................................................................................
63
APPENDIX
C ............................................................................................................................
85
APPENDIX
D .............................................................................................................................
93
iv
LIST
OF
FIGURES
Page
Figure
1.1
Transition
2.1
General coordinate
transformation
from physical to
computational
space (ref. I0) ............................................................................................
32
2.2
Boundarylayer
33
2.3
Disturbance
4.1
Grid generation
5.1
Automated
6.1
Stability
6.2
Transition
6.3
Boundarylayer
6.4
Chordwise
6.5
Effect
of Reynolds
number
on crossflow
at 48%
6.6
Effect
of Reynolds
number
on crossflow
at 48% semispan
6.7
Effect of Reynolds
number on shear stress in the boundarylayer
at 48%
semispan
for x/c = 10% .....................................................................................................
43
Effect
44
code's
wave
coordinate
orientation
process
stability
of Reynolds
process
system
........................................
analysis
distribution
number
36
.........................................................................................
number
34
35
................................................................................
due to Reynolds
stability
pressure
coordinate
....................................................................................................
validation
front result
........................................................................
on the swept
analysis
automation
system
31
37
...............................................................
38
region ...........................................................................
39
(o_ = 0 ) ..........................................................................
40
on transition
at 48%
semispan
semispan
..............................................
41
42
..............................................
prediction
at 48% semispan
................................................................................
45
Chordwise
pressure distribution
effects at 48% semispan
due to
angle of attack at Re = 6.34 million ..................................................................................
46
6.11
Effect
of angle of attack
on surface
flow patterns
47
6.12
Effect
of angle
on leading
edge
6.13
6.10
of attack
............................................................
flow attachment
at 48% semispan
...................
48
49
6.14
Maximum
crossflow
effect due to angle of attack at 48% semispan
for Re = 6.34 million and 45 sweep ................................................................................
50
at 48% semispan
..............................................................................
51
Maximum
crossflow
effect due to angle of attack at 48% semispan
for Re = 12.68 million and 45 sweep ..............................................................................
52
Higher Reynolds
number effect with angle of attack on
transition
for the 45 sweep ..............................................................................................
53
6.18
Swept
54
6.19
6.15
6.16
6.17
6.20
6.21
6.22
geometry
Effect of leading
at 48% semispan
surface
grids
...........................................................................................
at the lower
case ...................................................................
55
56
Maximum
crossflow
effect due to leading edge sweep
at 48% semispan
for Re = 12.68 million and ct = 0 ........................................................
57
58
vi
PARAMETRIC
STUDY
ON LAMINAR
SUPERSONIC
Joseph
Ames
WINGS
AT
Avila Garcia
Research
Center
SUMMARY
Laminar
eration
flow control
of High Speed
increase
range,
payload,
at supersonic
speeds
Reynolds
the increased
angle
the amount
cost.
stability
of attack,
flow
over
capabilities
as well as lower
fuel requirements,
A parametric
study to predict
the extent
a computational
fluid dynamics
using
The results
can actually
delay
This results
investigated
showed
Therefore,
in viscous
in larger
code
cou
were Reynolds
of attack
for
lift capability,
payload
(CFD)
in angle
higher
will
skin tempera
of laminar
in this study
that an increase
transition.
as well as a reduction
simultaneously.
in the development
of laminar
and sweep.
numbers
can be expected
as a key element
was conducted
layer
angle of attack,
specific
identified
Extending
and altitude
the overall
has been
Transports.
caused
by
in transition,
and range.
INTRODUCTION
Background
Laminar
flow control
boundarylayer
active,
or reactive
control,
namic
control
the extent
of laminar
flow is equivalent
Increasing
or control
of laminar
techniques
are categorized
Reynolds
(ref.
as those
parameters;
for example,
techniques
for example,
are categorized
wall suction,
1). Passive
means
techniques,
of altering
also known
the boundarylayer
pressuregradient,wall
flow
to delaying
is obtained
as natural
laminar
flow through
shaping,
sweep,
by passive,
flow (NLF)
normal
aerody
angle of attack,
and
number.
Active
means;
transition.
outofphase
disturbances
already
trols include
periodic
trol is complex
disturbances
present,
as those
means
of altering
outside
applied
heat transfer.
is reactive
flow control.
are artificially
thus stabilizing
heating/cooling
Reactive
introduced
However,
method.
flow control
is the process
layer
Some
by
to cancel
those
reactive
con
of laminar
flow con
The underlying
principle
of these
stage
in a process
of boundarylayer
ary layer
bances
as one expert
that involves
transition
techniques,
is an area which
include
a viscous
process
freestream
known
vorticity,
The disturbances
in consideration
in the boundary
roughness,
vibrations,
eventually
(viscous)
waves
enter
or high Mach
the amplifications
numbers),
are large
(or TS waves),
enough,
transitional
One thing
of these external
distur
Identifying
and therefore
and
layer
is not usually
For
vortices
nonlinearity
be noted
and therefore
flow studtes
waves
for curved
sets in through
in all laminar
the active,
processes
as described
in
disturbances
into the bound
the critical
Rayleigh
and GtJrtler
include
and sound.
layer
and must be
low amplitude
disturbances,
the amplification
can be modeled
modes responsible
for the amplification
of these disturbances
TollmeinSchlichting
physical
external
that tran
in the boundary
methods
parameters
reliable
flow. These
as receptivity
surface
of disturbances
requires
The transition
process is composed
of several
1). The transition
process begins by introducing
through
instabilities
amplification
(i.e., instabili
streamlines
secondary
(ref.
and tertiary
1).
portion
be approximated
is the introduction
of high levels
by
of
initial nonlinear
disturbances,
which cause a bypass of the _inear disturbance
regime and yield an
almost instantaneous
transition.
An example of such a nonlinear
transition
is attachmentline
contamination,
and is commonly
found in swept wings
edge caused by turbulent
flow from the fuselage.
Previous
Laminar
flow control
began
which
investigated
halted
transonic/supersonic
and maintain
speeds
reasonable
threedimensional
current
existing
of the wing
crossflow
means.
induces
and required
aircraft
performance
instability
The sweepback
a boundarylayer
that eliminated
crossflow.
to obtain
the ability
favorable
The sweep
methods
of natural
and highly
leading
Work
laminar flow
development
in the
at the wing
reached
lower
numbers
of sweeping
to maintain
pressure
local mach
the wing
laminar
gradient
and adverse
pressure
introduced
flow through
near the
trailingedgelikewiseinducescrossflowinstabilitieson thetrailingedgeportion of
the more common
pressure
when
gradient
pressure
which
transition,
use of suction
crossflow
instability,
active
thins
thus extending
Further
development
laminar
this is by
(LFC).
was delayed
fuel efficiency
Reynolds
number,
LFC
(ref. 6).
of about
ten years
by the abundance
of low
1970s that
day.
to consider
A major
laminarflow
control
skin friction
fuel efficiency
skin material
manufacturing
processes
to include strength and smoothness,
as well
in supercomputers
and computing
methods to analyze boundarylayer
stability for
have made
is turbulent
out
in the 1960s
designers
The
of extending
for a period
to the present
of
and
damping
feasibility
caused
such capabilities.
fuel efficiency
types
wings.
the basic
aircraft
These
on swept
wall region,
as high as 30 million
has forced
control
ments in aircraft
as advancements
prediction,
affecting
flows
top requirement.
transition
factor
flow control
the effective
showed
aircraft
to actively
lowering
in LFC
aircraft
gradient
of accomplishing
numbers
began
work
research
to improve
are amplified
of pressure
by attempts
The X21A's
of the current
in LFC research
The need
closer
at Reynolds
necessity
One method
Unlike
a favorable
instabilities
the presence
as laminar
layer,
profile
the wing.
when
flows.
which
the boundary
flow techniques
known
flow control,
are damped
inflectional
more commonly
crossflow
can be reduced.
research
boundarylayer
which
by reducing
low pressuregradient
as active
on the wing
the crossflow
instabilities
control
are categorized
moves
the threedimensional
produce
laminarflow
boundarylayer
controls
TS instabilities,
these crossflow
NLF airfoils
Natural
twodimensional
is applied,
gradients
viscous
drag.
Advance
a more realistic
method
of improving
the amount
of laminar
flow over
aircraft
fuel efficiency.
Turbulent
skin friction
drag
is reduced
by extending
an aircraft.
Until recently,
most studies on laminar flow have been in the subsonic
flow region. Work done in
this subsonic
realm has shown that turbulent
skin friction drag can contribute
as much as 50 percent
of the total aircraft
nificant
(refs.
potential
reduction,
gross
which
weight
the cruise
benefit
allows
skin/structure
range/payload
laminar
Speed
flow
Research
requirements
study
(NLF)
is being
on finite
Program
for supersonic
conducted
swept
(HSRP)
laminar
wings
underway
Transports
ratio by increasing
flow at supersonic
material
speeds
options
(SSTs)
the extent
includes
have
shown
of laminar
aerodynamic
and, therefore,
decreased
sig
flow
heating
aircraft
capability.
Current
A parametric
Supersonic
lifttodrag
of laminar
for more
and increased
on typical
Work
as an effort
to numerically
at supersonic
at NASA
flow control
speeds.
predict
This study
the extent
is one part
to gain an understanding
(SLFC).
of natural
of the High
of the technical
decrease
in the turbulent
skin friction
which,
in turn, decreases
the nature
ing benefits
can be expected
increased
payload,
decreased
cost,
and decreased
compressible
coefficients
addressed
compressible
boundarylayer
leadingedge
sweep
6.34 million
sweeps
stability
consists
study.
study
procedure
cases
These
(COSAL)
range,
initial
number,
angle
of attack,
and
compu
Cp's
number
the angles
of attack
Reynolds
of 0, 5, and
the sweep
number
(ref.
pressure
the
under
of the extent
GOVERNING
12).
an angleofattack
the Reynolds
study,
and a
numbers
sweep
of
of 45 deg. The
study addresses
process
conducted
measure
flow.
transition
The above
by the work
graphical
to turbulent
study,
study addresses
of 45 deg. Finally,
studies.
to predict
number
of attack
sweep
code
at an angle
yields a threedimensional
tion location of laminar
flow solver
cases.
of a Reynolds
total of seven
(CNS)
The Reynolds
addresses
and leadingedge
to control
angleofattack
values
study
on the air
profiles through
(ref. 11). Finally,
The parametric
NavierStokes
boundarylayer
code (WING)
is a signif
cost.
leadingedge
wing sweep. These parameters
tational fluid dynamics
(CFD) flow solver,
dimensional
laminar
there
operating
The parameters
of supersonic
an aircraft,
was substantially
this study.
of laminar
number
the leadingedge
of 0 deg. This yields
automated
This automation
flow by predicting
through
procedure
the transi
EQUATIONS
Mean
Flow
of mass,
momentum,
and energy.
timedependent,
nondimensional
NavierStokes
given in the following
vector form:
_gQ
+t
4
3E
OF
_x +_y4
These
conservation
equations
3G
+z
_Ev+3Fv4
+x
"_y
in Cartesian
oGv
+z
coordinates
the
(X, Y, Z) as
(2.1)
.__
flux vectors
Ev
pu
pv
E=
[
pu
2
pv
pw
puv
+p
F =
puv
F, G, are:
puw
pv 2 + p
pvw
pw
puw
pvw
pw 2 +p
_u(e + p).
_v(e + p)
w(e + p)
= ReI
"lTxx
1:xy
lyx
, Fv = Re I
"l:zx
0
'txz
"l:yy
Gv
Re I
'lTzz
lzy
.[3x
(2.2)
l:y z
_l]z
__y
with
= rZX' = I"l(v:
fix = YwPrI
+ Wy)
+ W'Cx:
(2.3)
2)
The variables
length,
are nondimensionalized
L, the velocity
by the corresponding
components
freestream
by dividing
by the freestream
values,
the spatial
speed
coordinates
of sound,
(x, y, z) by a reference
and viscosity
e, by ( pa2)_.
equations
transformation
To solve
into a generalized
bodyconforming,
the governing
it is necessary
to transform
coordinate
system
numerical
conditions.
algorithm,
independent
This transformation
of an efficient
of the boundary
equations,
curvilinear
(ref.
10) as
resulting
in a grid with
is expressed
as:
>(_,rl,_)
_ : _(x,y,z)
rl = rl(x,y,z)
(2.4)
_ =_(x,y,z)
The chain
rule of partial
differentiation
is applied
to these transformation
equations
as follows:
_y
(2.5)
the metric
determined
terms
matrix
_q
differential
_q
_x
"fly
rlz
dy
;y
;z
dz
y_
in equation
2.5 can be
expressions:
E x]
dy
_z) appearing
yq
y//_gq
(2.6)
(2.7)
Therefore
{
Cx
Cy
Cz
qx
fly
rlz
xn
Yrl
zn
]
(2.8)
(XnZ xzn)
xcz xz
(xczn xnz)
yqzyCzrl
= J](y_z;
 y;z_)
[. YCZn ynz
and the metric
terms
are represented
q
]
xnyxcyn
(xCyxCy)
xCy n xny
as follows:
Cx = J(yqz
Cy = J(z_x
yCzq)
xqz)
Cz = J(xqyxCyq)
qx = J(zcy
yCz)
qy = J(xcz
xz)
_z = J(ycxCx = J(Y@q
(2.9)
xCy)
yqz)
J is the determinant
of the Jacobian
xqy)
of the transformation
a(,n,)

Cx _y
_,,
nx
_x
TIz
_z
_y
(2.1 O)
which
as
J = l/j
1
=1/b(_'rl'_)
_(x,y,z)
 1
z_
=
Applying
l/[x_(YrlZ
this transformation
yq
Y;
z_
z;
(2.11)
equations
 yrlz_)]
2.1 gives
_Q _ _ _e _:_ _ _e_/
+____nn
+_=_ff_+__
+ _9_) or
(2.12)
where
= jI
9U
pu
puU + _xP
pv
pvU + _yp
9_
pwU + _zP
(e + p)U
pV
9W
puV+rlxP
puW+_xp
pvV + fly p
pvW+_yp
pwV + rlz p
pwW + _zP
(e + p)V
(e + p)W
Ev=
jI

_XTXX +_y_xy
+_z_xz
Tlxa:xx + rly'Cxy
+ TIz'Cxz
_x_yx
+_y_yy
+_z'Cyz
qx'l;yx
+ qz_yz
_x_zx
+ _y'_zy
+_z_zz
+_y_y
+ _z_z
_xl3x
+ rlyTyy
Tlx_ x + lly_y
+ lqz_ z
0
_xa;xx
+_y'_xy
+_z_xz
_x_yx
+ _y'l:yy
+ _z'_yz
_x'_zx
+ _y_zy
+ _z'Czz
+ _y_y
+ _z_z
_x_x
(2.13)
vector
were given
in equation
2.3 and
U=xU+ yV+ w
(2.14)
V = qxU+rlyV+qzW
W =_xU+_yV+_zW
Thinlayer
dependent
tries.
approximation
threedimensional
To alleviate
governing
some
equations.
high Reynolds
number
Large
NavierStokes
of this large
the body
surface
ing be normal
be mapped
to this surface.
equations,
approximation
where
centrated
CPU requirement,
This thinlayer
flows,
amounts
the boundary
It should
to a coordinate
The resulting
be noted
surface
particularly
complex
approximation
is applied
yields
the following
to the present
study
study
mean
geometo the
involving
of viscosity
only
are con
approximation
requires
_ = _min)
in the bodynormal
o3G _ 1 (OS']
at
Re I
J
the time
a thinlayer
is applicable
layer
to solve
that
direction
direction
This
flow equations:
(2.15)
where
0
_t
(4 2 +;_
and
+ ;2)[0.51a(u2
+_(;x
(_ are given
+v 2 +w2);
by equation
mean
about
+;yV;
+ _zW_)
Equations
to obtain
an accurate
boundarylayer
boundarylayer
code uses a conical
assumes a polar coordinate
system
valid
(a2);]l
2.13.
flow solution
Id
Pr(7  1)
u + ;yV + _zW)(;xU;
BoundaryLayer
the NavierStokes
for pressure
isobars
along
flow approximation
as shown in figure
constant
percent
chord
distribution
solution,
over
the
code to provide
the
WING was used. This
of trapezoidal
planform.
It
boundarylayer
equations
compressible
by the fundamental
flow,
equation:
(2.16)
_(pru) + __
3 (pw) + _zz(1)rv)
3
= 0
rmomentum
laminar
continuity,
equation:
au
waU
au
PU_r+PT3o+PV_zzP
w:
r
I0
3(
au'_
=_zz_!Lt_z)
(2.17)
0momentum
equation:
3w
w Ow
3w
uw _
OU_r + P r_+ PV_z  P r
Energy
l _)p
3 (cOw)
r 30 +_zz[._z
(2.18)
I. ( l r l)3Iu2+w22
Pvazaz [Vr0z
(2.19)
equation:
[.
3H
w 3H
3H
f'uLr+ 07 +
The following
boundary
conditions
0 [ _ 3H
at y=O;u=O,v=Vw,
W=O,(3_yH /
at y _
where
y is the distance
normal
The symbol
boundarylayer
Furthermore,
He , w _
to the wall,
3 represents
w indicates
the boundarylayer
direction,
component
in the radial
Linear
The
edge)
the boundarylayer
and the subscript
e is used to
(r) direction,
w is the velocity
component
Compressible
Stability
AnaLysis
component
as a perfect
Stability
in the bodynormal
gas, Sutherland's
direction
law is used
(COSAL)
code
is used to analyze
the stability
theory
13). Note,
ible threedimensional
for simplicity.
The derivation
derivation
by deriving
will be completed
for [a and
Equations
dimensional
boundary
of the threedimensional
(ref.
quanti
thickness,
edge quantities.
u is the velocity
3; u ) Ue, H _
of the linear
stability
the incompressible
with the derivation
equation
flow (p = constant)
of the compressible
condition
stability
equations.
Incompressible
expressed
stability
by the following
equationsnonlinear
The threedimensional
NavierStokes
incompressible
flow is
equations:
_U
+u.
3t
viscous
Vu='Vp+
p
11
vV2u
(2.20)
V.u
The fluid motion
where,
is then decomposed
The x, y, z Cartesian
directions,
equations
neglected.
and higher
Finally,
dynamic
velocities
by a reference
(2.22)
p(x,y,x,t)
= P(x,y,z)
+ p(x,y,z,t)
(2.23)
flow velocities
are oriented
following
linearized
terms
powers
and pressures
disturbance
direction.
of the original
is applied
Ue, density
respectively
in the x, y, z directions.
and products
similitude
velocity
as follows:
+ fi(x, y, z, t)
The basic
away,
disturbance
= U(x,y,z)
subtracted
into a steady
(2.21)
u(x,y,x,t)
coordinates
respectively,
=0
These
nonlinear
are substituted
NavierStokes
equations
of the perturbation
where
all lengths
by 9e, pressure
and spanwise
disturbances
terms,
being
are scaled
very
into
are then
small,
by a reference
are
length
1,
the
equations:
!
 t UVu
bt
+ u. VU = Vp+
V2_
R
(2.24)
V. fi = 0
where
R is a characteristic
Furthermore,
a "quasiparallel"
flows
(z) direction
flow assumption
number
defined
as:
flow is assumed,
coordinate
in the y direction
boundarylayer
spanwise
Reynolds
(2.25)
which
point
since,
at high Reynolds
are much
can therefore
smaller
numbers,
be represented
This
assumption
is applicable
in the streamwise
(y) direction.
U(y)
and W(y)
The linear
(PDEs),
disturbance
equations
normal
as follows:
components
(2.26)
(x) or
The quasiparallel
U = U(U(y),O,W(yi)
where
to
solution
applies:
12
separable
partial
respectively.
differential
equations
(2.27)
(P(Y) J
where
(ref.
for a given
fi,O,_,_
frequency
Substituting
tions 2.24
of the disturbance
eigenfunctions
wave
vector,
that determine
k, as shown
the structure
equations
the following
set of ordinary
NavierStokes
differential
equations
of the disturbance
disturbance
fi(o_2 + 132)
dW_,d__y_
i131_ +1
(2.28)
_(0_2 + _2 )]
[d26v
(2.29)
 *(Or2 + _2 )]
(2.30)
conditions
as y >oo (freestream);
an eigenvalue
This solution
problem
conditions
exists
can be expressed
(2.31)
are applied:
y = 0 (wall);
fi(y) >0,
and equations
and a solution
exists
by the following
_(y)
Now
exists
the following
six arbitrary
>0, _'(y)
dispersion
real parameters:
13
>0
therefore
relation:
= 0
2.28 through
co = co(o_,_)
where
equa
(ODEs):
R [d2y
the following
2.3
(co).
Next,
in figure
(2.32)
which
then form
an eigenvalue
problem.
Compressible
stability equationsequations
are an extension
of the above
The fluid motion
done
is decomposed
equations
respectively,
2.33 through
The resulting
linear
NavierStokes
equations,
disturbance,
as was
as follows:
= U(x,y,z)
+ fi(x, y, z, t)
(2.33)
p(x,y,x,t)
= P(x,y,z)
+ _(x,y,z,t)
(2.34)
x(x,y,x,t)
= T(x,y,z)
+ _(x,y,z,t)
(2.35)
x, y, z is used again
and neglecting
higher
the yaxis
pressure
powers
the basic
, _
Next,
NavierStokes
terms
and products
Finally, assuming
the basic flow is locally parallel as was done above
assumption
of equation
2.26, the linearized
compressible
NavierStokes
separable,
permitting
the following
normal mode solution:
to the
of the instanta
and temperature.
compressible
away
effects.
is normal
the x, y, z components
the compressibility
in which
equations
flow stability
u(x,y,x,t)
system
equation
viscous compressible
equations.
NavierStokes
coordinate
tions.
into a steady
nonlinear
velocity,
The threedimensional
derived incompressible
equa
of the original
non
of the perturbation
terms.
in the quasiparallel
flow
equations
become
,(y)/
=. _(y)texp[i(otx
+ 13z  cot)]
(2.36)
f)(Y) /
Here,
the quantities
with tildas
denote
, "_(Y) J
complex
disturbance
amplitudes.
Substituting
equations
2.26 and 2.36 into the linearized
compressible
yields the following
system of ordinary differential
equatio:ls:
(A D 2 + B D + C)_ :0
14
NavierStokes
equations
(2.37)
where
D represents
"d/dy"
and
is the vector
defined
by
(2.38)
given
by
A_._
1 dBo
go
dTo T6
0
B
i(k
1)/_
2(_
1)M2cj(o_U_
+ [3W o)
(0_2 + [32 )
1 dBo T'
Bo dTo
R
Bo_.
2 dgo
1 dg o (otU + _W6)
]to dT o
Bo dTo T;
2(y
I)M2c(otW_
(or2 + [32 )
15
 6U_))
1 dBo
dToT6
JR
(tUo+13Wo
_T
+_o) xic,2+132)1J
[(_u
iR /
o + f_wo)
[{ccU
laoTo
,
+'R a.o%1_2+_2)
[
o + 13Wo
_m)+(Ot2
_to dTo
+_2)/X
'
"[iRoT o/Tog
 2i(y
I)M20(aUo
+ 13Wg)lJ
(
/'_"o
1
i.,_M2
(_U o
, f
iR / BT(7
q3Wo
 I)M2 1
+co) J
C
l_(_U
o + DW o
120 O
[(aUo
O
0
m)+(R_ +ff_)(7
d21ao
+l_Wo).71
dT
_o
l I_[
o I
1 d_
I)M2 ((
Bo dTo
__L%
_"o a.o
dTo (_Uo+{3%)}
To+ I_ug+_Wo')
d.ol
aToJ
U'o 2
(o_W_  13Uo) ]
d"
Po
dTo
+ d_
(OtWo'_[gUo)]
dTo
+Wo 2 ) _ d._2o
dT
(TO )2
o
_ dgo T"'
dTo
ol
_[
iRo
o]
(ccU o+_W
I_T
o_)+(ot
}
2 +[32)1
(2.39)
The boundary
conditions
for equation
2.37 are
y = 0; 01  (I)2= 4 = 5 = 0
(2.40)
y _
The above
found
boundary
conditions
expressed
by the dispersion
nents oc, 13 with the complex
oo; i = 2 = 04
and equations
derivation
2.372.39
represented
= 05
") 0
re_resent
earlier.
an eigenvalue
This eigenvalue
problem
problem
can also be
parameters
16
as was
comporeal
(O_r,o_i,_3r,_3i
,cor,coi )
Solution
of the eigenvalue
problemThe eigenvalue
problem can be solved
four of the six parameters
mentioned above and finding the other two parameters
equations
2.282.31
flow, or 2.372.39
for compressible
by specifying
by using
flow.
In order to
parameter
Therefore,
the disturbance
amplification
or decays
is represented
by the complex
or
to be
compo
as follows:
level measurement
(Nfactor
or N) is obtained
'_t
N =
for transition
and is represented
as
(2.41)
COi
c Re(Vg)
ds
Note,
to another
the Nfactor,
The Nfactor
number.
downstream
The group
along
ity. Transition
is then predicted
at an Nfactor
from previous
studies
wings
on swept
the curve
the thinlayer
approximation
nal implicit
NavierStokes
factorization
algorithm
scheme
the change
in frequency
length
(X) must be
on comparison
with empirical
velocdata
17).
in the compressible
These
finitedifference
and Chaussee
wave
Flow
options
equations.
algorithm
by Pulliam
yields
METHODS
Mean
There
tangent
of 8 to 10 based
NUMERICAL
solve
velocity
location.
is then integrated
(2.42)
(ref. 19).
17
NavierStokes
schemes
and Warming
(CNS)
code
to
requirement
is needed
to capture
the boundarylayer
viscous
effects
occurring
in the present study. Unlike explicit methods that yield stiff problems
and restrict the time step to
very small values for stability,
implicit methods generally
avoid such stiffness problems
and allow
the use of a larger time step without loss of accuracy
(ref. 19).
BeamWarming
order
accurate
or factored
algorithm
block
ADI algorithm
to reduce
produces
the process
The BeamWarming
or fourthorder
accurate
to a set of onedimensional
a 5 5 block
tridiagonal
system
algorithm
an space.
is first
The equations
problems
for each
or second
are spatially
time iteration.
split
The
dimension
diagonal
the NavierStokes
ARC3D
computer
ADI algorithm.
sides.
formations
which
pentadiagonal
ADI algorithm
equations
greatly
The second
is known
uses a fourthorderaccurate
In this algorithm
matrices
simplify
need be inverted
block
scheme
partial
differential
The compressible
equations
dures.
A global
boundarylayer
finitedifference
eigenvalue
operator
by special
Diagonal
similarity
trans
times
The PulliamChaussee
less
diagonal
Equations
BoundaryLayer
secondorder
used to
smoothing
above.
although
algorithm
at the PulliamChaussee
described
BoundaryLayer
parabolic
taken
numerical
are diagonalized
process.
algorithm
The boundarylayer
code
2.162.19.
This method
basic
unstable,
expensive
tions
is unconditionally
is maintained.
This scheme
righthand
in threedimensions
terms the stability
stability
formulation
search
procedure
Stability
equations
in references
Equations
2.37 are solved
includes
by the COSAL
two eigenvalue
is available
A local eigenvalue
search is used when a good guess for the eigenvalues
imately 10 times faster than the global procedure
(ref. 14).
18
2023.
code
search
using
proce
is available;
this is approx
COMPUTATIONAL
GRID
AND
Wing Grid
The computational
eration
code
required
grids
named
wing
by the above
codes
was developed
wings
instructions
several
to quickly
wing
direction
It should
surface
chord
be noted
surface
airfoil
ordinates,
to obtain
ordinates
for
airfoil
grid generation
geometries.
This code
airfoil
grid gen
section
through
is
the
(ref. 27).
shape.
code
to allow
of spanwise
final spacing
a set of
and appendix
taper
C has
method
of creat
ratio or aspect
in the spanwise
above
singleelement
A contains
number
mentioned
generates
Appendix
was designed
desired
surface
The
developed
B contains
WSG
direction
initial
ratio;
spacing
(wingroot
in
spac
file.
necessary
file requires
the process,
refer
The surface
wing
appendix
sweep;
(wingtip
ordinates
grid (VG)
codes.
in this study.
from a code
HYPGEN
various
execution,
or quarter
developed
The algebraic
and taper
preprocessing
ing singleelement
the spanwise
generate
sweep
for program
edge
(WSG)
from an algebraic
the threedimensional
grid generator
generator
required
leading
grid is generated,
volume
with specified
or WSG
generator
were generated
Wing
surface
CONDITIONS
Configurations
surface
BOUNDARY
to the instructions
grid generation
to obtain
the abovementioned
listed
in appendix
code requires
of figure
airfoil
ordinate
input
explanation
of
A.
execution
Finally,
Volume
block
and checks
points
along
grid generator
grids.
equations
sweptforward
surface
grid generator
the spanwise
direction
The threedimensional
HYPGEN
accomplishes
consisting
check
computational
relations
being
used
is a cell volume
checks
computation
The second
19
wake,
grids
stretching
in the
routine
wing
threedimensional
grid generation
volume grid over the generated
this by solving
of two orthogonality
code currently
wing edges.
surface
generation
analyze
geometries
HYPGEN
cannot
the threedimensional
hyperbolic
tetrahedron
grid
check.
by HYPGEN
test is a Jacobian
code
single
after
a grid is
decomposition,
computation
and uses a
Boundary
The
farfield
wing's
Conditions
eliminates
effects
and adiabatic.
A symmetry
that could
yield
leadingedge
or
flow contami
AUTOMATED
In order
analysis
to conduct
process
the following
parametric
STABILITY
number
study
analysis
process
was created
it was necessary
of manhours
the pressure
ANALYSIS
required
to substantially
to obtain
it was necessary
using
a script
automate
a transition
the
prediction.
to perform
validation.
The
that combined
the three
codes
as illustrated
in figure 5.1. A copy of the script can
begins after a file is generated
from the mean flow
distributions
of the selected
span stations
for a specific
wing
geometry.
The pressure distributions
are supplied
tion at a time, which computes
the boundarylayer
to the boundarylayer
code (WING), one span staparameters
and profiles. Next the boundarylayer
outputs
AnaLysis
are supplied
to the Compressible
condition.
Stability
frequency
scan
requires
WING
for an average
by setting
to measure
COSAL
processor
Cray
postprocessing
codes
of frequencies.
run is approximately
and other
(COSAL)
the distur
Note that for each span station the stability analysis requires that the
for a spectrum
of frequencies
between 0 and 40,000 Hz to determine
the
This is accomplished
code
YMP
span stations
is needed
in the developed
Finally,
30 seconds
an outer
and since
on the wing,
per case.
loop
the
a total average
The actual
turnaround
automation
noted that the stability analysis must be run with 64bit precision (e.g.,
needed accuracy of the eigenvalue
search routine used in COSAL.
2O
to run a set of
script.
Also,
Cray YMP)
it should
due to the
be
RESULTS
As mentioned
in section
decreases
reduction
will allow
extending
laminar
tures,
allowing
for increased
addressed
wing
type aircraft,
the range
sweep.
speeds
sweeps
selection
this study
of angle
of attack
number
consist
to validate
of the F16XL
wing
cess, compared
result
study
the automation
transition
focused
Furthermore,
the surface
number,
on High
angle
Speed
tempera
of approximately
of attack,
and
Civil Transport
Automation
process
validation
process
a full parametric
region
decrease
This drag
number
45,000
(HSCT)
of 1.12 million
feet is used.
The
of 45 and 60 deg.
case, using
previously
the number
study
the newly
obtained
developed
manually,
of man hours
required
to a matter
Number
was conducted,
of laminar
Validation
of section
Reynolds
Before
fuel requirements.
are Reynolds
is limited
dimensional
transition
front dropped from hours
time dropped from days to a matter of hours.
had a reasonable
significantly
of skin material.
of the automated
and decreased
is being
Stability
In order
in the present
Since
on a Mach
flow over
range/payload
The parameters
laminar
DISCUSSION
in turn, decreases
flow at supersonic
leadingedge
leadingedge
1, extending
which,
AND
automated
as shown
to obtain
of minutes,
was used.
stability
in figure
a single
The results
pro
6.1. As a
three
turnaround
Effects
it was necessary
to establish
a baseline
case that
of changing
the vari
ous parameters
could be distinguished.
To achieve a fair amount of laminar flow, maintain
supersonic cruise conditions
at 40,000 to 50,000 feet altitude, and achieve a free stream Mach number of
1.5, a Reynolds
changing
laminar
figure
number
length.
of 1.27 million
The critical
transition
of the Reynolds
region
of the chosen
signifies
prediction
number
study
was increased.
wing
of the wing
transition
number
showed
This is illustrated
laminar
lengths,
flow
by
of
in
5 and
is no longer
disturbance
number
baseline
the portion
sweptwing
The results
(refs.
Nfactor
as a conservative
value
be noted
based
parametric
on previous
results near the tip and root of the wing are not valid due to the conical flow assumption
used in the
boundarylayer
code (WING). Tip effects also eliminate
the potential
for laminar flow near the wing
tip region. The analysis was therefore
only limited to the gray area shown in figure 6.3. Further
investigation
into the conical flow assumption
showed that for this configuration
the flow was not
21
show
figure
6.6 indicates
ferent
Reynolds
Only
number
cases
number
are lower
cases
cases
follow
of the crossflow
However,
number
Reynolds
of 10 percent
profiles
trend
stress
to illustrate
the same
occur
shear
that, at a given
number
is shown
exactly
stresses
the relationship
for
in the bound
Reynolds
number
of the movement
of the
crossflow
inflection
point of figure 6.5 to the shear stress (fig. 6.7) due to Reynolds
number effects.
Finally, the crossflow
boundarylayer
profile results (figs. 6.5 and 6.6) show that changes in
Reynolds
Next,
figure
number
do not affect
stability
curves
6.8. This
figure
the magnitude
of the transition
is a plot of chordwise
of the maximum
results
crossflow
at the 48 percent
x/c versus
frequency
semispan
velocities.
station
are shown
number
in
study
at
disturbance
level (Nfactor)
of 8. Basically, this plot shows the x/c locafrequencies
yield the disturbance
level of 8, and it is defined that the x/c
22
x/c value
of approximately
12 percent
the results
and a frequency
for a frequency
show
of 20,000
of 14,000
Hz. Therefore,
shifts forward
Reynolds
num
to an x/c of approximately
of the linearstability
theory's
transition
prediction
and the boundarylayer
profiles, its is revealed that a decrease in Reynolds
number yields higher shear stresses in the boundary
layer which act to damp out the crossflow
instabilities
and delay
transition.
AngleofAttack
Recall
in the wing
assumption
made in the boundarylayer
diction results will only be addressed
following
Boundarylayer
stability
angles
of attack
solutions. Therefore,
the boundarylayer
at the midsemispan
(48 percent semispan)
moves
Although
back
of 14,000
to study
flow
and transition
prestation for all the
span is shown
favorable
gradient
pressure
6.10.
The result
continues
back
produce
more delay
revealed
this trend
occurs
up to approximately
80 percent
chord
transi
of transition,
at 48 percent
is increased
chord.
at a
of attack,
in the delay
chord
that
Hz.
than others.
cases
show
of 12,000
in angle
in transition
case
location
case results
to 15.75 percent
shows
transition
frequency
in angle of attack
distribution
at a critical
in
only moves
this shows
of attack
at the leadingedge
gradient
chord
are shown
frequency
pressure
in figure
able pressure
is 14,000
18.5 percent
angles
semispan
frequency
Hz. Therefore,
at 48 percent
of 12.25 percent.
why an increase
analysis
which
disturbance
rdc value
In order
curves
to approximately
tion moves
of the transition
of three
at an approximate
transition
curves
consists
at the boundarylayer
of 0 deg indicate
critical
cases.
figure
occurs
Effects
semi
a stronger
favor
wing's
three
dimensional
crossflow
is expected to further destabilize
with increase in angle of attack, due to the
presence
of the stronger pressure gradients.
Therefore,
it is expected that the prediction
in transition
would move further forward. However,
this trend does not occur.
Next,
study
the surface
flow patterns
the wing
as the angle
of attack
of the different
These
wing
surface,
wing
the leadingedge
noted
that because
increased,
including
surface.
reveal
a separation
the attachment
velocities
cases
are shown
occurring
(fig. 6.11 ) to
edge
of
is
the leadingedge
point,
on to the lower
the crossflow
patterns
increases
angleofattack
surface
of the wing
point rotates
below
at the leadingedge
23
as angle of attack
increases.
the leadingedge
as angle
location
are reduced.
point
It should
of attack
below
also be
is
It was found
F16XL
angle
from
a previous
parametric
of attack
increased
study
crossflow
on the leadingedge
velocity
at a given
of the attachment
also rotate
point
forward.
profile
curves
are plotted
cases
the maximum
crossflow
exhibit
higher
up to approximately
the findings
increasing
maximum
wing's
angle
This translates
is increased
5 percent
case at approximately
The results
of the angle
prediction.
8 percent
number
increases
the shear
flow instabilities
which
that maximum
crossflow
idate this possibility,
and downstream
3 percent
Crossflow
Reynolds
show
chord
boundarylayer
number
Further
downstream
results
the maximum
angleofattack
increase
in laminar
for the
Effects
study
show
of the Reynolds
with Angle
that maximum
number
in the boundary
angleofattack
crossflow
the maximum
flow as the
crossflow
in the
crossflow
may directly
(fig 6.5(b))
thereby
Reynolds
influence
the
in Reynolds
damping
findings indicate
location.
To val
number
of attack.
are shown
is larger
crossflow
as in the previous
exhibit
of Attack
that a decrease
cross
10 deg case.
by maximum
st_ldy show
layer
case
cases
increase
stresses
of 12.7 million
edge,
the lower
another
and remains
chord
chord
chord
Number
crossflows
10 deg angleof
of attack
The results
The
in
of attack
the maximum
case.
a plot of
is shown
angles
Reynolds
of the
this trend
of attack
chord
transition
represent
angles
shows
of attack,
Results
of attack
angles
to better
are larger
angle
crossflow.
of attack
In order
After
of this study
crossflows
chord.
It
higher
5 percent
versus
as the
the leadingedge.
values.
flow is larger
crossflow
"(W/Uoo)max"
(5 and 10 deg)
attack
decreased
underneath
Although,
crossflow
profiles reveal that the crossflow
velocity components
attack cases near the leadingedge.
However,
further downstream
angleofattack
line of the
location
investigate
the above findings, boundarylayer
crossflow
profile
for the three angleofattack
cases at approximately
midsemispan
figure
attachment
wing
lower
in figure
6.15.
24
of attack
of 0, 5, and
The results
angles
from
of attack
these crossflow
angleofattack
study
cases.
profiles
edge.
These
(fig. 6.13).
the same
However,
as in the previous
as in the Reynolds
higher
Reynolds
earlier
lower
numbers
Reynolds
the boundary
layer
ther upstream
mum
results
crossflows
Transition
(fig. 6.15)
stability
study,
curves
influence
away
Reynolds
edge.
number
angles
to move
forward
6.17.
to investigating
It was necessary
be misinterpreted
avoid
the effects
by other
sweeping
the wing
requirements,
as well as maintain
figure
into lower
shear
leads to a suspected
results
for the
to the
stresses
transition
(fig. 6.16)
show
in
fur
that maxi
results
is increased,
therefore
validate
cone,
Effects
which
would
to shear
aspect
of sweep
surface
the same
in the
cross
location.
ratio constant
it was necessary
as shown
that maximum
cause
shock
the baseline
would
not
waves
clipped
and distort
delta
wing
to
the flow.
to obtain
lengths,
as shown
in
in figure
6.18.
Flow
wing
profiles
in the wing's
almost
study.
of the crossflow
transition
aspect
changes
which
the crossflow
These
Sweep
In addition
points
this translates
case,
reveals
of attack.
of figure
on the predicted
(fig. 6.16)
angleofattack
Therefore,
In this region,
study
number
the inflection
are further
at the higher
is now predicted
Reynolds
are higher
boundarylayer
number
number
lower
traces
case appears
21 percent
sweep.
"x/c"
sweep
wing
edge
of the wing.
of the crossflow
except
at 1 percent,
from about
Although,
effects
results
of the maximum
shows
chord
case maximum
chord. Overall,
below
6.19,
30 percent
as mentioned
due to sweep
Furthermore,
percent
cases
show
of the semispan
earlier,
(fig. 6.20)
show,
crossflow
"(W/Uoomax"
crossflow
is slightly
are stronger
versus
larger
at 3 percent
will be
up to
streamwise
location
at 1
crossflows
are larger after 2 percent chord and slightly fluctuate
after 10 percent
these results show that maximum
crossflow
is larger for the higher 60 deg swept
wing.
Next,
stability
curves
of the transition
results
that transition
occurs
at approximately
10 percent
approximately
2 percent
and a frequency
of chord
when
earlier
semispan
case, transition
of approximately
the wing
is swept
25
are shown
in the angleofattack
an x/c of 12 percent
approximately
at 48 percent
and a frequency
is predicted
20,000
in figure
study.
to occur
Hz. Transition
6.22.
The results
of 14,000
Hz for
at an x/c of
moves
forward
these
results
CONCLUSIONS
A parametric
study
was successfully
layer
stability
to predict
completed
code.
the extent
using
The study
AND
RECOMMENDATIONS
of laminar
a computational
was conducted
fluid dynamics
wings
at supersonic
code coupled
to gain understanding
with a boundary
of the technical
speeds
requirements
in
Research
Program
and sweep were
Conclusions
the boundarylayer
stability analysis has reduced the time required
transition
front location from hours to a matter of minutes. Further
of the Reynolds
of laminar
the overall
number
flow over
increasing
the extend
study show
that a decrease
crossflow
instabilities.
Essentially,
viscous wall region when Reynolds
therefore
turnaround
in Reynolds
to the effects
of laminar
The results
can actually
means
result
of the combined
be delayed
of delaying
is delayed
is an increase
downstream
effects
in angle
the maximum
in the laminar
An advantage
the maximam
of Reynolds
was accomplished
where
increases
of Reynolds
number
on
flow.
with an increase
transition
to the point
on the wing.
further
number
from
the crossflow
boundarylayer
profile is moved closer to the high
number is decreased,
damping
out the crossflow
instabilities
and
front prediction
to
number
of attack
crossflow
and angle
lbr specific
by decreasing
crossflow
is lower
of attack
Reynolds
the Reynolds
show
control
for the
that transition
numbers.
This
so that transition
angle of attack.
a reduction
flow (NLF)
are lower
number
and therefore
laminar
velocities
the
near
The
in the viscous
drag
increase
due to lift (caused by the increase in angle of attack) can partially be recovered
by the viscous drag reduction
due to the increase
in the laminar flow over the wing. The results basically
show
that if maximum
transition
crossflow
occurs,
crossflow
can further
is decreased
be delayed.
is a key to transition
increasing
the crossflow
Finally,
prediction.
instability
the results
As the wing
and therefore
26
where
transition
of the _weep
is swept
allowing
is predicted
study again
back
show
an increase
an earlier
to occur,
that maximum
in the crossflow
transition
prediction.
then
Recommendations
In the future,
supersonic
a leadingedge
speeds
A total drag
computed.
total drag.
on the extent
calculation
This would
Investigation
layer
code
reveal
uses a conical
replaced
with a threedimensional
research
Finally,
active
of laminar
the actual
should
flow conditions
of Reynolds
being
flow assumption
boundarylayer
information
of bluntness
at
flow.
effect
methods
be conducted
applied
and sweptwing
number,
show
of attack,
should
and sweep
angle
configurations
for swept
Furthermore,
wings
boundary
and should
it is recommended
solutions
in place
the results
of this Supersonic
laminarflow
laminar
control
Natural
methods
Laminar
in order
27
Flow Study
to establish
design.
should
be
that future
of the
be combined
an optimum
method
be
on
solutions.
supersonic
ing supersonic
study
of the numerical
which
boundarylayer
shape
with
of achiev
REFERENCES
Malik,
M. R.: Stability
and Aeronautics,
Bushnell,
AD686178),
Srokowski,
Aug.
A. J.: Mass
1977.
Flow
Including
Pfenninger,
Kaynak,
11.
Kaups,
Using
12.
Malik,
Center,
White,
Flows.
Review
Reynolds
Sponsored
Flow
of Laminar
RP1035,
AIAA
to the X21
vol.
I, Sept.
Paper 771222,
Academic
Press,
New
Airplane.
York
J. Aircraft,
M. R." COSAL:
J. Aircraft,
A BlackBox
Dimensional
Fluid Flow.
Studies.
for Reducing
NASA
Quiet
Skin
TMX2894,
Tunnel
March
for Transition
Northrop
LFC Research
compiled
Between
1949 and
by C. T. D'Aiutolo,
NASA
T.: Compressible
Wings.
of Concepts
Number
Control,
Apr. 67,
F. M.: Viscous
NASA
Design.
as Applied
for Future
an Euler/NavierStokes
Prediction
in Three
CR 165925, 1982.
13.
of High
and Navy
L.: A General
on Laminar
Research
U.; Hoist,
Flow
by Suction
W.: USAF
Langley
of Parallel
Recommendations
I. E.: Development
AIAA
Shear Layer
(available
as
Beckwith,
in Astronautics
on Attainment
and Suction.
Control
Fischer,
Progress
and Bibliography
Gradient
Kosin,
1967. Workshop
10.
Pressure
W. O.: Stability
Research.
.
M. H.: Survey
Betchov,
R.; and Criminale,
1967.
Friction,
1974.
of Transition
Traveling
Flow Requirement
vol. 2, Sept.Oct.
Design.
1969.
in Air Using
R. E.: Laminar
Control
Flow Control
1979.
vol.
Morkovin,
M. V.: Critical Evaluation
with Emphasis
on Hypersonically
NTIS
Theory
Laminar
vol.
of Transonic
Approach.
NASA
F;oundary
Separated
TM88311,
Layers
July
with Suction
Wing
1986.
on Swept
Compressible
Boundary
Stability
Layers.
28
Analysis
High
Code
Technology
Series
in Mech.
for Transition
Corp.,
NASA
Engineering.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
Buning, P.G.; Chan,W. M.; Renze,K. J.; Sondak,D.; Chiu,I. T.; andSlotnick, J. P.:
OVERFLOW User'sManual,Version1.6.NASA AmesRes.Ctr., Moffett Field,
Calif., 1991.
30.
Gray,W. E.: The Effect of Wing Sweepon LaminarFlow. RAE TM Aero 255, 1952.
31.
3O
External
Disturbances
Freestream
Freestream
Freestream
Velocity
Sound
Entropy
Spots
Particulates
Surface
Roughness
Vibrations
Receptivity
Control
Parameters
Pressure
"Slow"
Linear
Amplification
TS Instability
I
___
Crossflow
Instability
GSrtler Instability
/
/
Gradient
Wall Temp.
Mach #
Angle
of Attack
Sweep
Geometry/Curvature
IN
Bypass
3D
"_
onlinear
Space /
ime Disturbance.,/
"Fast"
Secondary
Tertiary
I
Figure
TURBULENCE
1.1. Transition
3]
and
Instabilities
flow chart.
= _ (xoy,z,t)
= _ (x,y,z,t)
= _"(x,y,z,t)
PHYSICAL
DOMAIN
Y
COMPUTATIONAL
DOMAIN
Figure
2. 1. General
coordinate
transformation
from physical
32
to computational
space
(ref.
10).
9\
\
\
33
Free
Stream
C,
Z,w,p
Vg
P
k
= Disturbance
= Magnitude
of the Disturbance
in the X direction.
= Magnitude
of the Disturbance
in the Y direction.
Up=
Potential
Vg=
Group
wave
= Phase
Yg = Angle
Wave
Vector.
flow velocity.
velocity
(direction
and speed
of Disturbance
energy).
Angle
k makes
flow Up.
wave orientation
34
system.
NACA64A010
_R
S3D or VG
Point
edistribution
._
g check
sf
Geometry Inputs
Aspect Ratio or Taper
Ratio
LE or 1/4 Chord
sweep
Desired number of
spanwise cuts.
Final & intial spacing
in the span dir.
Volume
Grid
Parameters
Figure
35
process.
_I Cp Data
!
I
I
Co e" io )"
I
/
/
I
I B.L Param./Profiles
p'
0
u_
l Stability
} STAB'OUTt
JTransition
Parameters
0) = 0 to 40 KHz
I TRANSITION
RESULTSFRONT
36
F16XL
Crossflow
Stability
Calculated
with Autom
_r_i
.........
_:,,_,_i_!iiiiiii_i',ii!_ii!!iii'_iilg
i_gi_.:._i_!_i_iiigii_i{i_i_i_i_i_iigi_i!i_
:.'.:_:
:._i":_!"_
*
::_:_::_.:g_
::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
!3_,'.'_
::::::::::::::::::::::::::::::::::::::::::::::
==========================================================================
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
::i;_i::i:@i_::_!_ii::::::_i
ii!::::
::::iiii::i!ii._ii_:::_iii;ii_!_i
iii_i!
!i!_
i!!!ii::
::!_ili!_
_i_i;ii_i!i
.>:8 ::2: :_'_:: _:::: 3 8._ 8._ _.::::: :;:_.:::::;
_:_::::::::_ &':::::::_i:::::: :::.:::
_::_i:: ;_::_
!.!.__i.i,.%!.i._.i.i,
i._.!_.:.
!.!.%_ .!_.
i.i.:!4!:_.%!_.i_.i._._.i!.!.!._!.iJ!_ ._$_..ii..::!
.:_._;.!.!.i.._
Transition
CC = 2.0 deg.
H = 44000 ft.
F16XL
Crossflow
Front in Black
NonLaminar
Stability
Figure
Flow in _ _;_,::_._,_
::y':
.........
Calcul
6. 1. Stability
automation
37
validation.
Re = 6.34 Million
Re = 12.68 Million
Figure
6,2.
Transition
front result
38
due to Reynolds
number.
Stability
Figure
Analysis
6.3.
Boundarylayer
Region
stability
39
analysis
in _ii_,_
_ij,
region.
0.3
0.2
0.1
0
Q,.
x
0.1
ii
0.2
0.3
 q=13%

q=19%
_
q=33%
_
q=48%
'
q=61%
q=72%
0.4
....
q=80%
.....
n=87%
0.5
0
0.2
0.4
0.6
_c
0.8
RE = 12.68 Million
0.3
T:......
0.2
0.1
0
Q" 0.1
0
!
0.2
0.3
0.4
0.5
Figure
6.4.
Chordwise
pressure
40
distribution
(a = 0).
BOUNDARY
LAYER
PROFILE
X/C
12.68 Million
0%
BOUNDARY
0.006
LAYER
PROFILE
X/C
1%
0.006
0.005
........................
i.............
!.............
:_
.............
i...........
0.004
,_..
0.004
0,003
0.003
0.002
0.002
0.001
0.001
0
0.1
0.08
0.06
0.04
0.02
0
W/Uo_ (Crossflow component)
BOUNDARY
LAYER
PROFILE
0.02
X/C
0.1
5%
0.006
i i! !ii! !ili
0.004
>
0.003
0.06
0.04
0.02
o
W/U =o(Crossflow component)
BOUNDARY
0.0o6
0.005
0.08
=,
;>
LAYER
PROFILE
0.02
X/C
=10%
.
0.003
0.002
0.001
0
0
0.1
0.08
0.06
0.04
0.02
W/U_ (Crossflow component)
BOUNDARY
0.006
LAYER
PROFILE
, ............
0.02
X/C
0.1
=16%
0.08
BOUNDARY
0.06
0.04
0.02
W/Uoo (Crossflow component)
LAYER
PROFILE
XJC
0.02
=21%
0.006
_ , , , _ , , ,
0.005
0.004
=
>
0.003
0.002
0.001
0.1
0.08
0.06
0.04
W/Uoo (Crossflow
0.02
component)
0.02
0.1
0.08
0.06
0.04
WAJ_ (Crossflow
0.02
component)
41
0.02
1.2
:
0.8
_0
0.6
0.4
0.2
0
0.1
0.08
W/Uoo
0.06
0.04
(Crossflow
0.02
0.02
component)
Figure 6.6. Effect of Reynolds number on crossflow at 48% semispan for x/c = 10%.
42
1.2
0.8
_0
>
0.6
0.4
0.2
0
0
0.1
0.2
Shear Stress
0.3
0.5
[Ibf/ft 2]
43
0.4
Stability Transition
Curves
0.2
i
:
i
:
:
JI,
RE=
RE=
6.34
12.68 Million
Million
0.15
C)
_0.1
X
Transition
Critical
Predicion
Frequency
0.05
0
5000
10000
15000
Frequency
Figure
6.8. Effect
of Reynolds
number
44
20000
25000
[Hz]
on transition
at 48% semispan.
, , Stability,
Transition
Curves,
,
,
0.25 i
.!
4
0.2
0.15_
X
0.1
Angle
of _.ttack
"*
0.05 '
Oi
deg.
"_Sideg.
......
I........
,0,eg.I
i
6000
9000
12000
Frequency
Figure
6.9. Effect
of angle
of attack
on transition
prediction
sweep.
45
15000
18000
[Hz]
at 48% semispan
and 45"
1] = 48%
0.5
'lJl,lllllltlt,ll,I
I
I
o.. 0
I i,
Angle of Attack
O
_5
deg.
deg.
10 deg.
1
0
0.2
0.4
0.6
X/C
0.8
Figure 6.10. Chordwise pressure distribution effects at 48% semispan due to angle of attack at
Re = 6.34 million.
46
LRE 45WING
a=.O.deg.
c_=.5.deg.
.(z=.10 deg.
Figure
6.11.
Effect
of angle
of attack
4?
on surface
flow patterns.
(Z = 0 deg.
._
_ ___.___::..:_..:..,
_"
'_
48%S
an
Flow
__
hment
O_ 5 deg.
48% Span
Flow
.=nt
O_ 10 deg.
48% Span
Flow
Figure
6.12.
Effect
of angle
of attack
on leading
48
at 48% semispan.
Angle of Attack
O deg.
.e 5 deg.
= 10 deg.
BOUNDARY
LAYER
PROFILE
oooo
i
@ X/C
= 0%
.........
BOUNDARY
LAYER
_"_ ooooi
PROFILE
@ X/C
o,oo5
_L..........................................................................................
_
o.oo5
...........
0.004
_ .....................................................................................
0.004
................................................................................
E o.oo3_
_ E 0.003
>
.i>
0.001
= 1%
0.001
......
0.16
0.12
W/U_o
BOUNDARY
0.08
0.04
(Crossfiow
component)
LAYER
PROFILE
= 5%
,
_" ...................................................................................
0.004
" ............................................................................
BOUNDARY
_
"_
I
0.005
0.12
0.08
W/U_
@ X/C
0.006
_
0.16
0.006
0.005
0.004
'_
_
0.04
(Crossflow
LAYER
component)
PROFILE
@ X/C
= 10%
.............
ii i
iiiii
ii
ooo3
oooOOO
OOOoooO
0
0.16
',
0.12
W/Uoo
BOUNDARY
.AYER
'
0.08
0.04
(Crossflow
component)
PROFILE
d i
''
0.16
"
0.003
!.
X/C
16%
BOUNDARY
....i
_i
0.001
PROFILE
_'> 0.003
X/C
21%

0.002
0.001
0
0.12
W/U_o
Effect
LAYER
........
i
i
6.13.
0.04
component)
0.004
Figure
0.08
(Crossflow
......
i o.oo_
....... J
::
0.16
0.12
W/U_
_
i
0.004
"
of angle
0.08
0.04
(Crossflow
component)
of attack
on crossflow
0.16
i
0.12
W/U=
profiles
at 48% semispan
49
_
0.08
(Crossflow
_
0.04
,
0
component)
0.15
0.1
E
0.05
I*o
Is
0.05
J_lJl]l]l
0.05
ir]J
0.05
O.1
deg.I
deg. !
O.15
0.2
0.25
X/C
Figure 6.14.
sweep.
Maximum
crossflow
of attack
5O
at 48% semispan
and 45"
Angle
of Attack
_O
BOUNDARY
LAYER
PROFILE
o.oo8_
0.007
deg.
e
s.
10 deg.
XIC
0%
deg.
BOUNDARY
0.008
, , , _ ,
:....... q
LAYER
_,
PROFILE
X/C
1%
i..........................................................................
!.......
0.007
4
0.006
0.006 F.............................................................................
_.......
0.005
0.005
,.:.
0.004
__
r;1
1"
0.003_.................
:...................
_...................
_..................
_......
4
0.002 iE......i..................
_
o.oo1
0.003
0.001
0.16
0.12
W/U=,
BOUNDARY
0.008
,.
.......
0.08
(Crossflow
LAYER
0.04
component)
PROFILE
X/C
0.16
5%
_ o.oo5_
0,12
W/Uoo
BOUNDARY
0.008
._
_, 0.004
>
"'
>
0.08
(Orossflow
LAYER
0,04
component)
PROFILE
X/C
10%
0.004
0.003
0.003
0.001
!...
::
:.1
0.16
0.12
0.08
W/Uo_ (Crossflow
BOUNDARY
0.008
o.oo o.oo
LAYER
......
0.04
component)
PROFILE
!
@
,
X/C
0.16
16%
BOUNDARY
0.008
b
i
....
0.12
0.08
W/Uoo (Crossflow
LAYER
.
PROFILE
0.04
component)
@
,
X/C
l
=
,
21%
.................
..........................................................
i
0.007
0.006_.................
_...................
!...................
_...............
........
_
0.005
::
0.004
i .................
i...................
i...................
i..................... ._
0.005
0.004
0.003
,. ................. _...................
..................................
0.003
i .......
0.002
o.oo
...................
:.............
!_ l.........
0
0.16
Figure
0.006
0.12
0.08
W/Uoo (Crossflow
6. 15. Effect
of angle
0.04
component)
of attack
on crossflow
0.002
0.001
0
0.16
at 48%
51
semispan
0.12
W/U_
for Re
0.08
(Crossflow
= 12.68
milfion
0.04
component)
and
45 sweep.
Angle of Attack
O deg.
5 deg.
0
.........
0.05
' '
_'
0.05
0.05
10 deg.
O.1
O.15
0.2
0.25
X/C
Figure
45"
6.16.
Maximum
crossflow
effect
due
to angle
of attack
sweep.
52
at 48%
semispan
for
Re = 12.68
milfion
and
0.05
0.04
0.03
F
0
X
0.02
Angle
0.01
of Attack
.........................................................................
5
''_0
deg.
deg.
10 deg.
I ...............
F
E
5000
IIIll,l
II1[
12000
19000
6. 17. Higher
Reynolds
number
effect
with
angle
of attack
53
il,!
26000
Frequency
Figure
33000
40000
[Hz]
on transition
for the
45"
sweep.
45WING
Leading
Aspect
Edge Sweep
,, 45 deg.
Ratio  1.45
60WING
Leading
Edge
Sweep
Trailing
Edge Sweep
Aspect
Ratio  1.45
 60 deg.
= 36.2 deg.
Figure
6.18. Swept
geometry
.54
surface
grids.
Figure
6. 19. Effect
Leading
Edge
Sweep
= 45 deg.
Leading
Edge
Sweep
= 60 deg.
of sweep
on surface
flow patterns
at the lower
55
Reynolds
number
case.
Sweep
45 deg.
KBOUNDARY
LAYER
PROFILE
0.006
0"005
_ .........................................................................
X/C
60 deg.
0%
BOUNDARY
LAYER
0.006
..................
! , . .
L
L
PROFILE
X/C
1%
; ............
0"C105
.......................................................................................
0.004b...................................................................................
J o.oo4 ...........................................
,:.
= 0.003 _..................................................................................
4 _ 0.003 .....................................................
:........................
>
;
L ...........................
k
_
0.002
:..............
!
:.......................................
i
i
>.
_
..
1 F............
!............................................
T.............
i ...........
i
I
o_,,,i
.........
......
L
0.1
0.08
BOUNDARY
0006
0.06
0.04
0.02
W/U_
(Crossflow component)
LAYER
PROFILE
X/C
O.C
0.1
0.06
0.04
W/U=o (Crossflow
0.02
component)
PROFILE
XlC
0.0:
10%
i,
0.004 _............
i ..............
t..............
i..............
i..........................
_
i
0.08
BOUNDARY
LAYER
0.006
..................
5%
I ....................
t"
o
0.002
o.oo_
_............
i..............
i..............
i...............
0.004
4 ._
_ o.oo_
0.002
0.002
ooo,
............ ............
................
0
0.1
0.08
BOUNDARY
0.006 i
0.06
0.04
0.02
W/Uoo (Crossflow component)
LAYER
PROFILE
X/C
0.0
16%
0.1
0.08
BOUNDARY
0.006
'
0.06
0.04
W_J=o (Crossflow
LAYER
0.02
component)
PROFILE
X/C
0.0;
...................
2i%
J
0.003
>
;
_
0.002
!
;
:"
;
..........................
!
;
::
::
':
i ...............................................
0.004
"
0.001
0.02
component)
6.20.
0.C04...........................
0.C01
0
'
0.1
Effect
0.08
0.06
0.04
W/Uo_ (Crossllow
of leading
edge
sweep
0.003
0.002
Figure
4 '_
_ _.
on crossflow
O.C
profiles
56
l`
o0.1
at 48%
' ' i
0.06
0.04
W/Uoo (Crossflow
semispan
(Re
= 6.34
' '__"_=_'''"
0.02
component)
million
and
0.0_
a = 0).
i
0.08
..................
_..............................
;..............................
'..............................
_............................
E
r
iiiiI
0.04
....................................................................................................................................................
m
Sweep
0.02
,..._.........
.&.45
deg.
_60
deg.
0
0
0.05
0.1
0.15
0.2
0.25
x/c
Figure
6.21. Maximum
crossflow
and a = 0".
5?
at 48% semispan
for Re = 12.68
million
Stability Transition
0.2
Curves
I
0.15
0.1
Sweep
0.05
45 deg.
60 deg.
..................
J
1
....... L____L
5000
___L_J
....
10000
!..............
15000
Frequency
Figure
6.22.
Effect
of leading
edge
sweep
L__J......_L_
.....
JL__[
on transition
at 48%
58
20000
L__
25000
[Hz]
semispan
for Re
= 12.68
milfion
and
a  O"
APPENDIX
59
WING
NACA
The following
6 or 6aseries
SURFACE
will describe
airfoil.
GRID
CREATION
the process
PROCEDURE
used to generate
a surface
Steps
output
GOING
input
file to get an
DEPEND
ON WHETHER
THE POINTS:
YOU
ARE
3)
TO REMEMBER
ABOUT
REDISTRIBUTING
ON VG:
upper surface
wing surface grid
grid. The output
is the number
of
then
file
61
PRI__
FILMED
Read
Swap indices
to[PGA] and
To select the
in the PICK
MG Plot3d
<PICK
PICK
A SECTION
A POINT>
^PICK
Select
the entire
wing patch
by using
Now redistribute
the points by going
select "REDISTRIBUTE
SECTION".
A LINE
to "GDP"
button.
and under
this menu
Remember
now to swap back the indices
5) Now to put the new airfoil distribution
in the proper format to read into the
surface grid generator use the prograta
"S3d_airf.f".
6) Finally, check the spacing with the program
"sf.f' to make sure the
stretching
condition
of sf=l.3 is met.
IV. GENERATING
THE
SURFACE
GRID
The following
Finally,
are inputs
"WingSurf_new.f"
Grid Generator:
62
grid.
to generate
an
APPENDIX B
63
program
WingSurf_new
C
C
C
Joseph
Date:
Jan.
A.
Garcia
This
program
13,
1992
C
C
C
PROGRAM:
clipped
using
will
delta
an
generate
wing
with
Airfoil
surface
grid
64A010
sections
NACA
Potential
Analytical
for
Description
C
C
C
MODIFICATIONS:
To
no
MODI:
an
input
longer
use
normalized
has
been
the
desired
wise
point
name
span_dist2.f
desired
from
the
Airfoil
another
airfoil
called
using
for
Visual
chordwise
and
again
NACA
use
64A010,
(VG)
is
to
now
which
an
by
by
be
as
have
Also
develop
modified
will
to
the
destribution.
which
spacing
Grid
point
distribution
but
sixseries.f
coordinates
modified
point
Description
code
VG
span
program
to
have
the
to
this
code.
input
C
C
MOD2:
THIS
IS
TIP
MOD
SHAPE
[10/5/91]
PORTION
TO
OF
THE
EXTEND
THE
SWEEP
INTO
THE
WING
C
c
MOD3:
This
one
that
modifies
the
with
equal
will
code
to
leading
now
allow
edge
require
for
and
taper
trailing
Aspect
ratio
edge
Ratio
(AR)
of
sweeps
input.
c
c
MOD4:
This
mod
to
be
out
each
stretching
distribution.
will
allow
able
to
input
having
create
1/4
this
to
sweep
surface
any
a
and
grid
sweep
spanwise
taper
subroutine
generation
clipped
point
ratio,
will
wing
with
destribution
instead
be
code
delta
used
for
the
Vinokur
to
determine
either
input
the
c
c
MOD5:
This
as
mod
is
either
to
allow
LE
the
sweep
or
user
1/4
to
chord
sweep
sweep.
C
c
This
mod
MOD6:
code
to
assigned
edge
of
This
mod
will
allow
be
able
this
to
aspect
ratio
the
wing
surface
create
as
wing
any
"AR"
grid
sweep
which
wing
will
generation
with
an
sweep
the
trailing
necessary.
c
c
MOD7:
TE_sweep
the
done
for
all
span,
was
A_R,
to
the
TR,
have
the
"WingSurf_gen"
various
wing
LE_sweep,
give
inputs
Qrt
sweep
along
as
the
with
necessary.
C
C
MODS:
This
mod
of
the
was
wing
done
to
with
the
sweep
all
of
the
tip
zero
section
LE_sweep.
c
c
MOD9:
This
mod
points
will
to
cluster
match
the
those
of
zero
the
thickness
swept
trailing
edge
wing.
c
c
MODI0:
This
section
mirroring
mod
will
using
the
cluster
the
points
the
Vinokur
off
zero
thickness
streching
the
wing.
WintTip
routine
and
not
just
c
c
INPUTS:
Quarterchord
sweep
surface
grid
airfoil
ordinates
or
angle
(GAMMA),
dimensions
(jmax,
file
taper
lmax),
named
ratio
and
(lamda)
normalized
"airfXXX.ord"
(airf127.ord
airf200.ord).
c
c
OUTPUT:
PLOT3Dformat
surface
grid
of
the
wing
c
*W*WW*W*WW****WW*WW*WW****WWWWWWWWW*WWWWW*WWWWWWWWWWWWW**WW**W**WW
parameter
(jdim=500,kdim=100,1dim=10,idim=500)
dimension
x(jdim,
+
+
ldim),y(jdim,kdim,
z_U(idim),
s(150),
CHARACTER*30
i000
kdim,
x_U(idim),
ldim),z(jdim,
x_L(idim),
kdim,
z_L(idim),
ldim),
yy(kdim),
t(100),w(50),IDM(jdim),JDM(kdim),KDM(idim)
OUTFILE,name,INFILE
FORMAT(A)
REAL
GAMA,
lambda,
X,
dely,
yspan,
Qrt_sweep,
dw0w,
+
INTEGER
t_22,
t_10,
t_ll,
t_12,
t_13,
t_23,
t_24,
t_25,
Chord,
delx,
dely_t,
dl,
d2,
stotin,
dtl,
dwlw,
jmax,
delx_te,
counter,
imax,
llmax,
jmax_te,
kk,
tmax,
jj,MG,
IGRID,
LE_length,
dely_wt,
kmax_w,
jmax_te_U,
y_edg
TE_sweep
delwk
deltp2,
thrdspan,
kmax_t,
npts_U,
sw_type,
t_21
sweep,
TE_length,Chord_t
AR,
dt2t,
jmax_u,
Chord_r,
dtlt,
dw3w,
count,
t_15,
span,
LE_sweep,
dt2,
dw2w,
kmax,
t_14,
AR_type,
sf
jmax_t
npts_L,tr_testl
tr_test2,
cont_testl
wmax
c
c
c
c
Taper
ratio
lambda
SS$$$$$$SS$$$SS$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
SSSS$$$SSSS$$SSSSSSSSS$$$S$$$$$SSSSSS$$$SSSS$$$$$
Qrt_sweep
1/4
chord
sweep
in
DEGREES
GAMA
1/4
chord
sweep
in
RADS
sweep
Leading
Edge
sweep
in
RADS
LE_sweep
Leading
Edge
sweep
in
DEGREES
TE_sweep
Trailing
Edge
sweep
in
DEGREES
sweep_te
Trailing
Edge
sweep
in
RADS
dtlt
Initial
dt2t
Final
c
c
sf
TE
TE
Strecthing
Wake
Wake
spacing
spacing
factor
tip
tip
(1.3)
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$S$$$$$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
c
c
set
sf
1.3
I. 0
dl
0.3
d2
0.005
0.5*Chord_r
ngrid
Chord_r
TE_length
c
default
parameters

..............................................
c
WRITE(*,
+type
' (a,$)
"i"
read
if
')'If
NOT
you
type
"0":
.eq.
i)
KNOW
what
you
want
your
TAPER
RATIO
to
taper
was
be
'
(*,*)tr_testl
if(tr_testl
then
continue
else
WRITE(*,'
+ecified
(a,$)')'You
(.84):
must
now
specify
'
66
span
since
no
sp
read
(*,*)span
goto
1
endif
WRITE(*,'(a,$)')'If
read
(*,*)tr
test2
if(tr_test2
the
.eq.
goto
else
taper
I)
ratio
is
1 type
"I"
or
"0"
if
not:'
then
continue
endif
WRITE(*,
read
' (a,$)
(*,*)
')
' INPUT
taper
rat{o:
'
Aspect
Ratio
lambda
c
1
PRINT*,'If
you
plan
read
(*,*)AR_type
IF(AR_type
.eq.
to
specify
i)
THEN
WRITE(*,'(a,$)')'INPUT
+rd:
'
read
Aspect
Ratio
type
desired
or
normalized
if
by
not:'
root
cho
(*,*)AR
if(tr_testl
lambda
.ne.
) then
(2*span/AR
else
 1.0
continue
endif
WRITE(*,'(a,$)')
read
(*,*)sw_type
'If
if(sw_type
.eq.
WRITE(*,'(a,$)')
read
(*,*)
sweep=
Sweep
i)
is
based
on
LE
type
"I"
or
I/4C:'
then
' INPUT
LE
Sweep
[deg]:
'
LE_sweep
= AR*(l+lambda)/2
GAMA
= ATAN(
(span*TAN(sweep)
Qrt_sweep
= GAMA*(180/3.141592654)
TE_sweep=
TE_sweep=
.25*(lambda
ATAN(
(span*TAN(sweep)
TE_sweep*(180/3.141592654)
if(TE_sweep
PRINT*,
.it.
'YOUR
CHOICE
PRINT*,'AND
THE
PRINT*,'SO
IF
read(*,*)
0.0
OF
BL
WANT
 Chord_r))/span
lambda)/span
..... TE_SWEEP'
YEILDS
"WING"
TO
DOES
CONTINUE
NOT
ANYWAYS
TAKE
THIS'
TYPE
cont_testl
.eq.
continue
) then
else
PRINT*,'
OK
!!!!!!
TRY
AGAIN
!!!!!!!![!'
STOP
endif
else
continue
endif
else
WRITE(*,
' (a,$)
(*,*)
')
' INPUT
) then
INPUT
CODE
YOU
if(cont_testl
GAMA
if
LE_sweep*(3.141592654/180)
span
read
"0"
1/4
Chord
Qrt_sweep
= Qrt_sweep*(3.141592654/180)
67
Sweep
[deg]
'
else
0:'
span
= AR*(l+lambda)/2
sweep
= ATAN(
(span*TAN(GAMA)
LE_sweep=
sweep*(180/3.141592654)
TE_sweep=
ATAN(
TE_sweep=
TE_sweep*(!80/3.141592654)
endif
.25*(lambda
(span*TAN(sweep)
1 +
 Chord_r))/span
lambda)/span
ELSE
WRITE(*,'(a,$)')'If
read
Sweep
is
based
on
LE
type
"i"
or
"0"
if
i/4C:'
(*,*)sw_type
if(sw_type
WRITE(*,'
read
.eq.
(a,$)')
(*,*)
sweep=
i)
then
' INPUT
Sweep
[deg]:
'
LE_sweep*(3.141592654/180)
WRITE(*,'(a,$)')'INPUT
read
LE
LE_sweep
(*,*)
TE_sweep=
TE_sweep
if
Delta
wing
then
use
deg:
'
TE_sweep
TE_sweep*(3.141592654/180)
c
if(tr_testl
.ne.
lambda
= span*(
if(lambda
.it.
PRINT*,'YOU
span
) then
TAN(TE_sweep>
0.0)
then
CHOSEN
TO
 Chord_r)/(TAN(TE_sweep)
TAN(sweep)
(0.0
+
PRINT*,'SPAN
PRINT*,'
MUST
LARGE
BE =
!!!!!!
 TAN(sweep)
SPAN
FOR
or > ',span
TRY AGAIN
THESE
+ Chord_r
SWEEPS'
)
!!!!!!!!!!'
STOP
else
continue
endif
else
continue
endif
c
span
GAMA
= ATAN(
(lambda
 Chord_r)/(TAN(TE
(span*TAN(sweep)
sweep)
+
 TAN(sweep))
.25*(lambda
 Chord_r))/span
Qrt_sweep=
GAMA*(180/3.141592654)
AR
= 2*span/(l+lambda)
TE_sweep=
TE_sweep*(180/3.141592654)
else
WRITE(*,'(a,$)')
read
(*,*)
GAMA
=
WRITE(*,'
read
' INPUT
1/4
Chord
Qrt_sweep*(3.141592654/180_
(a,$) ') 'INPUT
TE_sweep
(*,*)
TE_sweep=
Sweep
[deg]:
'
Qrt_sweep
if
Delta
wing
then
use
TE_sweep*(180/3.141592654)
if(tr_testl
lambda
.ne.
=
) then
span*(TAN(TE_sweep)
else
 TAN(sweep)
) +
Chord_r
continue
endif
if(
span
TE_sweep
=
sweep=
deg:
TE_sweep
(0.75"(1
ATAN(
.eq.

0.0)
then
lambda))/TAN(G_A)
(span*TAN(TE_sweep)
LE_sweep=
sweep*(180/3.141592654)
TE_sweep=
TE_sweep*(180/3.141592654)
68
lambda)/span
'
AR
2*span/(l+lambda)
else
sweep
span=
(0.25*(lambda
(.25*TAN(TE_sweep)
sweep=ATAN(
 Chord
TAN(GAMA))/(.75)
r)
(span*TAN(GAMA)
LE_sweep=
sweep*(180/3.141592654)
TE_sweep=
TE_sweep*(180/3.141592654)
AR
)/(
(TAN(GAMA)
.25*(lambda
 TAN(sweep))
 Chord_r))/span)
2*span/(l+lambda)
endif
endif
ENDIF
'span=
'LE_sweep=
' , LE_sweep
'TE_sweep=
' , TE_sweep
'Qrt_sweep=
'AR=
'Taper
' , span
',
' , Qrt_sweep
AR
ratio=
' , lambda
WRITE
*,'(a,$)')'INPUT
read
*,*)kmax_w
WRITE
*, '(a,$)')'INPUT
* t *)dl
initial
*, '(a,$)')'INPUT
* t *)d2
final
read
WRITE
read
how
many
point
in
spacing
spacing
the
in
in
spanwise
spanwise
the
[25]:
dir.
spanwise
dir[.005]:
#################################################################
CALL
vinokur
( s, kmax_w,
span, dl, d2 )
#################################################################
i=0
do
yy(i)
k =
4
C
s(i)
+ 1
if(ABS(yy(i)
continue
 span)
.it.
0.001)
kmax_w
= k
#################################################################
This
C
C
i=l, kmax_w
=
k
section
will
set
the
spanwise
outer
boundary
for
the
tip zero
section.
#################################################################
MODIO
C
C
dely_wt
dwlw
Print*,'
dw3w
=
=
dely_wt=
= 0.
wmax
c2/25/93
do
C
15
(yy(kmax_w)
 yy(kmax_wl)
',dely_wt
1
dw0
dwlw
dw2w
0.
thrdspan
0.3*span
thrdspan
1.0*span
jj
I,i00
deltp2
=
if(dw2w
if(dw3w
dw0
=
.20*thrdspan
.it.
deltp2)
.it.
dw0*sf
wmax
dw3w
 yy(kmax_wl)
(yy(kmax_w)
then
thrdspan)
= wmax
then
= dw0
69
)*Chord_r
)*Chord_r
[.05]:
'
'
dw2w
= dw3w
else
 dw3w/sf
continue
endif
15
Continue
kmax=
kmax_w
(wmax
i)
c
c
if(yy(i)
kmax=
c
c
c
cc
.le.
.5*span)
then
kmax_w
+ (kmax_w
*, 'kmax=
endif
 k)
' ,kmax
*, 'kmax_w=
' ,kmax_w
goto
3
#################################################################
c
c
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
c
c
This
and
will
open
then
the
read
it
spanwise
into
an
ordinate
data
file
ceated
array
c
c
c
c
c
read(21,*)
k=
0
do
if(ABS(yy(i)
'old'
, form='
formatted'
yy(i)
1
 span)
print
*, kmax_w=
continue
.it.
'
0.001)
kmax_w
kmax_w
goto
3
\\\\\\\\\\\\\\\\\\\\\
status=
i=l,kmax
c4
.crv',
kmax
read(21,*)
k = k +
'span2
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
c
c
2
WRITE(*,'(a,$)
read
')
' INPUT
GAMA
Qrt_sweep
TE_sweep
lambda
WRITE(*,
read
LE
TE
'
sweep
=
sweep
= sweep
1.0
'(a,$)')
' INPUT
or
length
[y/Cr]
'
(*,*)LE_length
span
= LE_Iength*COS(GAMA)
WRITE(*,'(a,$)')
read
' INPUT
Aspect
Ratio
normalized
pc.int
in
by
root
Chord:
'
(*,*)AR
span
= AR*(I
=
lambda)/2.0
2*span/(l+lambda)
*,
'span=
*,
'LE_sweep=
*,
'Qrt_sweep=
*,
'TE_sweep=
*,
'AR=
WRITE(*,
read
[deg]:
sweep*(3.141592654/180)
LE_sweep
AR
Sweep
(*,*)sweep
',
',
span
', LE_sweep
', Qrt_swee_
', TE_sweep
AR
'(a,$)')'INPUT
how
many
the
spanwise
[kmax_w]
'
dir.
'
(*,*)kmax_w
WRITE(*,
' (a,$)
read
(*,*)dl
')
' INPUT
initial
70
spacing
in
the
spanwise
WRITE(*,'
(a,$)')
read
(*,*)d2
' INPUT
final
spacing
in
the
spanwise
dir.
#################################################################
CALL
vinokur(s,kmax_w,
span,dl,d2)
#################################################################
k
do
=
6
0
i=l,kmax_w
yy(i)
= s(i)
k = k + 1
if(ABS(yy(i)
 span)
if(yy(i)
.le.
kmax
print
endif
c
kmax_w
*, 'kmax=
0.001)
kmax_w
then
(kmax_w
 k)
' ,kmax
print
*,'k max_w=
continue
6
c
.!t.
.5*span)
' ,kmax_w
#################################################################
kmax_w
yspan
do 2
c
c
=
=
i
0.75*kmax
span/(kmax_wl)
=l,kmax
yy(i)
c2
: yspan*(il)
continue
c
c
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
This
by
will
the
open
the
SIXSERIES
airfoil
code
ordinate
ref
__
and
data
then
file
read
ceated
it
into
an
array
c
c3
open(20,file='airf.ord',status='old',form='formatted')
3
WRITE(*,'(a,$)')'
READ(*,1000)infile
ENTER
grid
AIRFOIL
ORDINATE
open(20,file=infile,status='old',form='formatted')
read(20
i000)
name
read(20
npts_U
read
20
read
20
read
20
read
20
read
20
read
20
(x_U(i),z_U(i),i=l,npts_U)
npts_L
(x_L(i),z_L(i),i=l,npts_L)
jmax
te
delx_te
TE_length
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
jmax
= npts_U
(npts_Ll)
PRINT*,'HEY!!!
jmax
jmax_U
= npts_U
llmax
= 1
c
dely
kmax_w
0.6*kmax
kmax_t
kmax
',jmax
span/(.6*kmaxl)
 kmax_w
c
c
kmax
!.2*kmax_w
c
MOD9a
c
c
do
50
k=l,kmax_w
71
INFILE
NAME:
'
PRINT*,'
k=
',k
c
c
c
This
will
add
zero
thick
****
MOD9:
for
Starting
section
behind
upper
surface
the
from
the
Tip
of
the
the
"WingTrailing
Edge"
****
wing
c
c
TE_sweep
c
ATAN(
(span*TAN(sweep)i
PRINT*,'***********
y(j,k,l)=
lambda)/span
',y(j,k,l)
c
MOD9
c
c
Chord
(I
Chord_t
yy(k)*(TAN(TE_sweep)
(i
PRINT*,'
Chord=
PRINT*,'
Chord_t=
delx_te
dtlt
dt0
=
=
dt0
NOTE:
=
j
x_U(npts_Ul)
x_U(npts_Ul)
',delx_te
.eq.
HI
i)
THEN
i!!'
dtlt
0.
i,i00
This
is
sometimes
conditions
the
in
grid
if(
0.12*TE_length
delwk
0.13*TE_length
HI
2!!
.It.
PRINT*,'
=
to
avoid
subroutine
delwk
dt2t
dt0
change
Vinokur
spacing.
PRINT*,'
HI
delwk=
delwk)
',delwk
then
3!!'
dt0*sf
tmax
tmax
dt3t
dt0
dt2t
dt3t
PRINT*,'#1
dt3t/sf
dtlt=',dtlt,
' dt2t=',dt2t
else
continue
endif
7
continue
dt2t
PRINT*,
=
'#I
dt3t
dt3t/sf
dtlt=',dtlt,
'
dt2t=',dt2t
c
ELSE
CONTINUE
ENDIF
c2/93
dt2t
dt2
)*Chord
)*Chord_t
dt2t
do
delx_te
PRINT*,'
tmax
)
TAN(sweep))
dtl
IF
c
',Chord_t
(x_U(npts_U)
delx_te=
dtl
TAN(sweep))
',Chord
(x_U(npts_U)
PRINT*,'
yy(kmax_w)*(TAN(TE_sweep)
PRINT*,'#2
=
=
dt3t
dt3t/sf
dt2t
dtl=',dtl,'
dt2=',dt_
72
certain
that
distorts
jmax
te
tmax
jmax_te
c
PRINT*,'HEY
!!!
',dt3t
PRINT*,
',tmax
1
te
jmax_te=
PRINT*,'dt3t=
'tmax=
2*jmax
',jmax_te
#################################
CALL
vinokur(t,tmax,TE_length,dtl,dt2)
#################################
jj
tmax
if(tr_test2
.eq.
TE_sweep
i)
then
GAMA
else
continue
endif
do
i0
j=
jj
l,jmax
jj
PRINT*,
't(jj)=
y(j,k,l)
x(j,k,l)
',t(jj),
=
PRINT*,'
Chord_r
'
jj=
',jj
yy(k)
+
x(j,k,l)=
z(j,k,l)
i0
te
1
y(j,k,l)*(TAN(TE_sweep))
',x(j,k,l),'
j=
t(jj)
',j
0.0
Continue
c
c
c
This
starting
will
compute
from
****
the
the
for
upper
root
the
surface
trailing
upper
of
the
wing
edge.
surface
****
c
c
c
c
c
MOD9
c
c
do
npts_U
jmax
npts_U
20
j=jmax
te
te
l,jmax_U
jmax
te
c
i
=i
y(j,k,l)
1
=
yy(k)
if(tr_test2
.eq.
Chord
x(j
k,l)
I)
then
1.0
Chord
x_U(i)
(y(j,k,I)*TAN(GAMA))
else
TE_sweep
Chord
=
=
(I
ATAN(
+
(span*TAN(sweep)I
y(j,k,l)*(TAN(TE_sweep)
x(j,k,l)
=
endif
Chord
x_U(i
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
z(j,k,l)
c
Chord
TAN(sweep))
y(j,k,l
*TAN(sweep)
\\\\\\\\\\
\\\\\\\\\\\\\\\\\\\\\\\\
\\\\\\\\\\
\\\\\\\\\\\\\\\\\\\\\\\\
z_U(i)
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
2O
lambda)/span

continue
c
c
This
starting
will
compute
from
the
the
root
lower
leading
surface
edge
c
count=
73
of
the
wing
MOD9
do
30
j=jmax_U+l,jmax
do
30
j=jmax_U
count
y(j,k,l)
jmax
count
yy(k)
jmax
te
te
+
l,jmax
jmax
te
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
if(tr_test2
.eq.
x(j,k,l)
i)
Chord
then
x_L(count)
(y(j,k,I)*TAN(GAMA))
else
TE_sweep
Chord
ATAN(
(I
(span*TAN(sweep)i
y(j,k,l)*(TAN(TE_sweep)
x(j,k,l)
Chord
x_L(count)
lambda)/span
TAN(sweep))
y(j,k,l)*TAN(sweep)
endif
c
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
This
convert
section
will
it
to
read
the
in
proper
the
ordinate
values
to
of
define
the
airfiol
the
wing
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
z(j,k,l)
30
Chord
and
z_L(count)
continue
c
c
This
will
add
****
a
for
zero
the
thick
lower
section
surface
behind
****
the
"Wing
Trailing
Edge"
MOD9
jj
= 1
do
40
j=
jmax
jmax
do
40
j=
jmax
jmax_te_U
jj
y(j,k,l)
jj
yy(k)
.eq.
=
else
t(jj)
TE_sweep
ATAN(
x(j,k,l)
Chord_r
z(j,k,l
Continue
4O
l,jmax
i,jmax
I)
+
2*jmax
te
then
y(j,k,I)*TAN(GAMA)
(span*TAN(sweep)I
+
endif
if(tr_test2
x(j,k,l)
te
lambda)/span
y(j,k,l)*TAN(TE_sweep)
)
t(jj)
0.0
c
c
5O
contlnue
c
c
c
do
This
i00
will
kk
kmax_w
kk
k=
kmax_w+l,
add
****
for
zero
the
***
MODI0
***
kmax
thick
section
cff
upper
surface
****
c
c
MODI0
kk
kk
i
74
the
"Wing
TipTrailing
Edge"
dely_wt
dwlw
Print*,'
 y(l,kmax_wl,l)
)*Chord_r
 y(l,kmax_wl,l)
)*Chord_r
',dely_wt
= dwlw
dw3w
0.
dw2w
0.
thrdspan
thrdspan
55 jj =
do
(y(l,kmax_w,l)
(y(l,kmax_w,l)
dely_wt=
wmax
= 1
dw0
c2/25/93
=
=
0.3*span
= 1.0*span
i,i00
deltp2
=
if(dw2w
if(dw3w
dw0
=
.2*thrdspan
.it.
deltp2)
.it.
dw0*sf
wmax
= wmax
dw3w
dw0
dw2w
dw3w
then
thrdspan)
+
then
 dw3w/sf
else
continue
endif
55
Continue
dw2w
dw3w
 dw3w/sf
dtl
= dely_wt
dt2
dt2
= deltp2
= dw2w
#################################
CALL
vinokur(w,wmax,thrdspan,dtl,dt2)
#################################
kk
kk
C
C
C
MOD9
C
C
C
jj
= npts_U
tmax
do
60
jj
j=
=
+++++++++++++++
y(j,k,l)
l,jmax
jj MODI0
te
1
++++++++++++++++++
= yy(kmax__w)
(yy(kmax_w)
 yy(kk))
y(j,k,l)
= w(kk)
+ yy(kmax_w)
++++++++++++++++++++++++++++++++++++++++
IF(tr_test2
.eq.
I)
THEN
MOD9
x(j,k,l)=x_U(i)
+ y(j,k,I)*TAN(GAMA)
x(j,k,l)=x(j,kmax_w,l)
(y(j,k,l)
y(j,kmax_w,I))*TAN(GAMA)
ELSE
TE_sweep
Chord_t
= ATAN(
:
(i
(span*TAN(sweep)
lambda)/span
+ y(j,k_max_w+l,l)*(TAN(TE_sweep)
C
C
C
C
******************************************************************
THIS
IS A MOD
TO EXTEND
THIS
ZERO
THICKNESS
75
THE
LE_SWEEP
SECITON
INTO
TAN(sweep))
C
C
MOD9
C
C
x(j,k,l)=Chord_t
x(j,k,l)=t(jj)
x_U(i)
+ y(j,k,l)*TAN(sweep)
Chord_t*x_U(npts_U)
Chord_r
+ y(j,k,l)*TAN(sweep)
C
ENDIF
C
z(j,k,l)
Continue
60
0.0
This
will
add
********
zero
thickness
For
the
Upper
i=
npts_U
section
surface
off
the
"Wing
Tip"
chord
*******
MOD9
i
do
do
70
70
npts_U
j=
j=
i
jmax
te
jmax_te_U
=
+++++++++++++++
jmax
te
+
+
l,jmax
l,jmax_U
1
U
U
+
jmax
te
1
MODI0
++++++++++++++++++
y(j,k,l)
= yy(kmax_w)
y(j,k,l)
= w(kk)
(yy(kmax_w)
 yy(kk))
+ yy(kmax_w)
+++++++++++++++++++++++++++++++++++++++
IF(tr_test2
.eq.
Chord
x(j,k,l)
TE_sweep
Chord_t
= Chord_t
ELSE
ATAN(
i)
(i
THEN
1.0
* x_U(i)
(y(j,k,I)*TAN(GAMA))
(span*TAN(sweep)i
lambda)/span
+ y(j,kmax_w+l,l)*(T_(TE_sweep)
 TAN(sweep))
C
******************************************************************
THIS
IS
MOD
THIS
TO
ZERO
EXTEND
THE
THICKNESS
LE_SWEEP
INTO
SECITON
*W**W**W*W******W**WWW*W*W*WW***W_*WW*WWWWWWWW*WWWWWW**WW**W*WWWW*
x(j,k,l)=
Chord_t
* x_U(i)
y(j,k,l)*TAN(sweep)
ENDIF
$$$$$$$$$$$$$$$$SSSSSSSSSSSSSSSS_:SSSSSS$$SSSSSSSSSSSSSSS$$SSS$SS$
z(j,k,l)
0.0
$$$$$$$$$$$SSSSSSSSSSS$$SSSSSSS$_$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
C
7O
continue
C
C
This
will
add
********
For
zero
the
thickness
Lower
sect:.on
surface
counter
76
off
_'******
the
"Wing
tip"
chord
MOD9
C
C
C
do
80
j=
jmax_U
do
80
j=
jmax
l,jmax
jmax
te
jmax
U
te
l,jmax
jmax
te
counter
counter
+++++++++++++++
MODI0
++++++++++++++++++
y(j,k,l)
= yy(kmax_w)
y(j,k,l)
= w(kk)
(yy(kmax_w)
 yy(kk))
+ yy(kmax_w)
++++++++++++++++++++++++++++++++++++++++
IF(tr_test2
.eq.
Chord_t
x(j,k,l)
TE_sweep
= Chord_t
= ATAN(
Chord_t
(i
i)
THEN
1.0
* x_L(counter)
ELSE
(span*TAN(sweep)i
(y(j,k,I)*TAN(GAMA))
lambda)/span
+ y(j,kmax_w+l,l)*(TAN(TE_sweep)
)
TAN(sweep))
******************************************************************
THIS
IS
MOD
THIS
TO
ZERO
EXTEND
THE
THICKNESS
LE_SWEEP
INTO
SECITON
**WWWWWW*WWW***W*WWWW*WW**WWWWW**W*WWWWW***WWWWWW**W***WW*WW*WW***
x(j,k,l)=
Chord_t
* x_L(counter)
+ y(j,k,l)*TAN(sweep)
ENDIF
$$$$SSSSSSSSSSSS$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
z(j,k,l)
80
= 0.0
continue
C
C
This
will
add
****
zero
for
thick
the
lower
section
off
surface
the
"Wing
Tip
Trailing
Edge"
****
MOD9
i
= npts_L
do
90
jj = 1
j= jmax
do
90
j=
jj
y(j,k,l)
y(j,k,l)
te
jmax
te
i,
jmax
+
+
jmax
1
te
I,
jmax
jmax
i = i
+++++++++++++++
 jmax
= jj
MODI0
jmax_te
1
++++++++++++++++++
= yy(kmax_w)
+ (yy(kmax_w)
= w(kk)
+ yy(kmax_w)
 yy(kk))
++++++++++++++++++++++++++++++++++++++++
IF(tr_test2
x(j,k,l)=x_L(i)
x(j,k,l)=x(j,kmax
TE_sweep
Chord_t
w,l)
+
ELSE
= ATAN(
=
.eq.
i)
THEN
+ y(j,k,I)*TAN(GAMA)
(I
(y(j,k,l)
y(j,kmax_w,I))*TAN(GAMA)
(span*TAN(sweep)i
lambda)/span
+ y(j,kmax_w+l,l)*(TAN(TE_sweep)
C
C
C
******************************************************************
THIS
IS
A MOD
TO
EXTEND
THE
77
LE_SWEEP
 TAN(sweep))
INTO
THIS
ZERO
x(j,k,1)=Chord_t
x(j,k,1)=t(jj)
THICKNESS
x_L(i)
SECITON
y(j,k,1)*T__,l(sweep)
Chord_t*x_U(npts_U)
Chord_r
y(j,k,1)*TAN(sweep)
C
ENDIF
z(j,k,l)
Continue
9O
0.0
i00
continue
C
C
write
grid
WRITE(*,'(a,$)')'
READ(*,1000)outfile
ENTER
WRITE(*,'(a,$)')'
IF
grid
YOU
FILE
WANT
NAME:
b_LTI
'
GRID
OUTPUT
TYPE
i:
READ(*,*)MG
C
C
change
'binary'
to
'unformatted'
to
run
on
CRAY
or
VAX
OPEN(UNIT=7,FILE=outfile,STATUS='new',
form='unformatted')
MOD9
C
C
PRINT*,'HEY
jmax
= jmax
2
+
!!! jmax_te=
jmax_te
PRINT*,'HEY
!!!
jmax
IDM(1)
JDM(1)
KDM(1)
IF(MG
kmax
llmax
.ne.
WRITE
(7)
i)
jmax,
WRITE(7)
jmax=
',jmax_te
',jmax
THEN
kmax,
llmax
(((X(J,K,L),
J=jmax,
l,l),
K=l,kmax),
L=l,llmax),
(((Y(J,K,L),
J=jmax,
l,l),
K=l,kmax),
L=l,llmax),
(((Z(J,K,L),
J=jmax,
l,l),
K=l,kmax),
L=l,llmax)
ELSE
NGRID
WRITE(7)
WRITE(7)
DO
110
NGRID
(IDM(IGRID),JDM(IGRID),}_M(IGRID),IGRID=I,NGRID)
IGRID=
I,NGRID
WRITE(7)
(((X(I,J,K),
I=IDM(IGRID),I,I),J=I,JDM(IGP.ID)),K=I,KDM(IGRID)),
(((Y(I,J,K),
I=IDM(IGRID),I,I),J=I,JDM(IGkID)),K=I,KDM(IGRID)),
(((Z(I,J,K),
ii0
I=IDM(IGRID),I,I),J=I,JDM(IGLID)),K=I,KDM(IGRID))
CONTINUE
ENDIF
stop
78
end
subroutine
vinokur(s,lmax,smax,dsle,ds2e)
c
c
stretches
at
Vinokur
points
the
on
boundaries
surface
is
so
satisfied.
that
Taken
specified
from
spacing
NASA
CR
3313
by
made
to
better
(1980).
c
c
In
this
version,
match
The
four
i.
deltas
is
2.
deltas
from
the
distinct
resulting
iterations
deltas
iterations
are
values
are
to
summarized
the
requested
values.
below:
series
3.
set
equal
the
to
last
the
desired
iteration
value.
is
corrected
from
Taylor
expansion.
deltas
is
calculated
is
calculated
from
linear
fit
quadratic
between
the
first
two
guesses.
4.
deltas
three
desired
after
guesses,
if
value.
three
from
indeed
If
it
quadratic
doesn't,
it
fit
between
will
takes
pass
the
the
through
value
first
the
calculated
swipes.
c
c
Additionally,
for
approximate
this
y=sin(x)/x
and
version
uses
y=sinh(x)/x
solution
was
also
the
approximate
rather
than
taken
from
inverse
a
NASA
Newton
CR
3313.
c
c
common
/io/
dimension
input,kopy,
s(200),
default
dl(4,2),d2(4,2)
c
C
......
for
an
IRIS
2500,
emax=87.0
c
dsavg=smax/float(imaxl)
c
21
write(*,103)dsavg
PRINT*,
'dsle=
',dsle
PRINT*,
'ds2e=
',ds2e
PRINT*,
'smax=
',smax
21
c
c
dsavg=0.001
dsle=dsavg
call
realval(l,l,dsle,q,q,*21,*101)
if(dsle.ge,
22
dsavg=0.01
c22
write(*,104)dsavg
c
c
smax.or.dsle.lt.
0.0)go
to
21
0.0)go
to
21
ds2e=dsavg
call
realval(l,l,ds2e,q,q,*22,*101)
if(ds2e.ge.(smaxdsle).or.ds2e.lt.
if(dsle.eq.0.0.and.ds2e.eq.0.0)then
kase=0
dsle=dsavg
ds2e=dsavg
nlast=4
else
if(dsle.eq.0.0)then
kase=l
79
solution
iteration.
The
nlast=l
c
23
write(6,106)
continue
23
call
realval(0,l,slop,no,no,*23,*101)
if(slop.lt.0.0.or.slop.gt.l.0)go
to
23
to
24
dsle=slop
else
if(ds2e.eq.0.0)then
kase=2
nlast=l
c 24
24
write(6,106)
continue
call
realval(0,l,slop,no,no,*24,*101)
if(slop.lt.0.0.or.slop.gt.l.0)go
ds2e=slop
else
kase=0
nlast=4
end
if
dssl=0.0
dss2=0.0
do
6 n=l,nlast
if(n.le.2)then
dsl=dsle0.5*dssl
ds2=ds2e+0.5*dss2
dl(n,l)=dsl
d2(n,l)=ds2
PRINT*,'dI(I,I)=
PRINT*,'dI(I,2)=
',dl(l,l)
',dl(l,2)
PRINT*,'dI(2,1)=
',dl(2,1)
PRINT*,'dI(2,2)=
',di(2,2)
else
if(n.eq.3)then
dsl=dl(l,2)*(dl(2,1)dl(l,l))/(dl(2,2)dl(l,2))+dl(l,l)
PRINT*,'d2(I,I)=
',d2(l,l)
PRINT*,'d2(I,2)=
',d2(i,2)
PRINT*,'d2(2,!)=
',d2(2,1)
PRINT*,'d2(2,2)=
',d2(2,2)
ds2=d2(l,2)*(d2(2,1)d2(l,l))/(d2(2,2)d2(l,2))+d2(l,l)
C
PRINT*,'dsI=
',dsl
PRINT*,'ds2=
',ds2
PRINT*,'nlast=
PRINT*,'HELP!!!'
if(dsl.lt.0.0)dsl=0.5*aminl(dl(l,l),dl(2,1))
',nlast
if(ds2.1t.0.0)ds2=0.5*aminl(d2(l,l),d2(2,1))
dl(n,l)=dsl
d2(n,l)=ds2
else
if(n.eq.4)then
denom=(dl(l,l)dl(2,1))*(dl(2,1)dl(3,1))*(dl(3,1)dl(l,l))
all=dl(2,1)dl(3,1)
a21=dl(3,1)**2dl(2,1)**2
a31=dl(2,1)*dl(3,1)*(dl(2,1)dl(3,1))
al2=dl(3,1)dl(l,l)
a22=dl(l,l)**2dl(3,1)**2
a32=dl(3,1)*dl(l,l)*(dl(3,1)dl(l,l))
al3=dl(l,l)dl(2,1)
a23=dl(2,1)**2dl(l,l)**2
a33=dl(l,l)*dl(2,1)*(dl(l,l)dl(2,1))
bl=(all*dl(l,2)+al2*dl(2,2)+al3*dl(3,2))/denom
8O
b2=(a21*dl(l,2)+a22*dl(2,2)+a23*dl(3,2))/denom
b3=(a31*dl(l,2)+a32*dl(2,2)+a33*dl(3,2))/denom
disc=(b2*b24.*bl*b3)
if(disc.lt.0.0)go
to
ddl=(b2+sqrt(disc))/(2.*bl)
dd2=(b2sqrt(disc))/(2.*bl)
dd3=dl(3,1)
if(abs(ddidd3).it.abs(dd2dd3))then
dsl=ddl
else
dsl=dd2
end
if
denom=(d2(l,l)d2(2,1))*(d2(2,1)d2
3,!))*(d2(3,1)d2(l,l))
all=d2
a21=d2
(3, i) *'2d2
a31=d2
a12=d2
(2, i) **2
a22=d2(l,l)**2d2(3,1)**2
a32=d2(3,1)*d2(l,l)*(d2(3,1)d2(l,l)
a13=d2(l,l)d2(2,1)
a23=d2(2,1)**2d2(l,l)**2
a33=d2(l,l)*d2(2,1)*(d2(l,l)d2(2,1)
bl=(ail*d2(l,2)+a12*d2(2,2)+a13*d2(3
2))/denom
b2=(a21*d2(l,2)+a22*d2(2,2)+a23*d2(3
2))/denom
b3=(a31*d2(l,2)+a32*d2(2,2)+a33*d2(3,2))/denom
disc=(b2*b24.*bl*b3)
if(disc.le.0.0)go
to
ddl=(b2+sqrt(disc))/(2.*bl)
dd2=(b2sqrt(disc))/(2.*bl)
dd3=d2(3,1)
if(abs(ddldd3).it.abs(dd2dd3))then
ds2=ddl
else
ds2=dd2
end
if
if(dsl.lt.0.0.or.ds2.1t.0.0)go
end
if
calculate
to
constants
s0=smax/float(imaxl)/dsl
sl=smax/float(Imaxl)/ds2
b=sqrt(s0*sl)
a=sqrt(s0/sl)
if(kase.eq.l)then
b=sl
else
if(kase.eq.2)then
b=s0
end
if
calculate
based
on
value
of
if(bl.)i,2,3
C
is
real
if(b.lt.0.26938972)then
pi=4.*atan(l.)
x=pi*(l.
b
+ 6.794732"b*'4
+
b**2
13.205501"b*'5
else
81
(l.+pi**2/6.)*b**3
+ ii.726095"b*'6)
c=l.
x=
b
sqrt
*
*
(6.*c)*
+0.15"c
(i.
0. 053337753"c*'4
end
0.057321429"c*'2
0.07584513_*c*'5)
0.057321429"c*'2
+0.048774238"c*'3
if
go
to
c
x
is
zero
x=O.
go
to
c
c
is
imaginary
if(b.lt.2.7829681)then
c=bl.
*
x=
sqrt(6.*c)*(l.
0.15"c
+0.0077424461"c*'4
0.024907295"c*'3
0.0010794123"c*'5)
else
v=alog(b)
w=l./b
0.028527431
x= v + (l.+l./v)*alog(2.*v)
+ 0.24902722"w
+
*
*
end
2.6294547"w*'3
0.02041793
1.9496443"w*'2
8.56795911"w*'4
if
c
c
distribute
points
along
boundary
c
4
continue
if(kase.eq.l.or.kase.eq.2)then
s(l
) = 0.0
s(imax)
do
smax
i=2,1maxi
j= imax+li
xi=float(il)/(imaxl)
if(b.gt.l.
ul=l.
else
OOOl)then
tanh(x/2.*(xil.))/tanh(x/2.)
if(b.lt.O.9999)then
ul=l.
tan
(x/2.*(xil.))/tan
(x/2.)
else
end
ul=
if
xi*(l..5*(bl.)*(l.xi)
_(2.xi))
u2=sinh(xi*x)/sinh(x)
if(kase.eq.l)then
fact=abs(dsle)
else
s(j)
= ( (l.fact)*(l.ul)
if(kase.eq.2)then
fact*(l.u2)
fact*
) *smax
fact=abs(ds2e)
s(i)
(l.fact)*
ul
end
if
continue
else
do
i=l,lmax
xi=float(i!)/float(Imaxl)
cnum=x*(xi0.5)
cden=x/2.
if(b.lt.O.9999)then
cc=tan(cnum)/tan(cden)
u=0.5*(l.+cc)
82
u2
*smax
else
if(b.ge.O.9999.and.b.le.l.
OOOl)then
u=xi*(l.+2.*(bl.)*(xiO.5)*(l.xi))
else
if(b.gt.l.
OOOl)then
cc=tanh(cnum)/tanh(cden)
u=0.5*(l.+cc)
end
if
s(i)=u*smax/(a+(l.a)*u)
end
if
c
if(imax.ge.4)then
dssl=(
s(4)
+4.*s(3)
5.*s(2)
+2.*s(1))
dss2=(2.*s(imax)5.*s(imaxl)+4.*s(imax2)
end
s(imax3))/2.
if
c
esl=s(2)s(1)
es2=s(imax)s(imaxl)
if(n.ne.4)then
dl(n,2)=esldsle
d2(n,2)=es2ds2e
end
if
continue
esmin=
c
l.Oe+08
esmax=l.Oe+08
do
j=2,1max
stmp=s
(j) s (jl)
if(stmp.lt.esmin)then
jnj
=j
esmin=stmp
if
end
if(stmp.gt.esmax)then
jxj
=j
esmax=stmp
end
7
if
continue
write(6,105)esl,es2,jnji
jnj,esmin,
jxjl,jxj,esmax
c
i01
return
103
format(/,6x,
*
104
'enter
/,6x,'(default
format(/,6x,
*
105
/,6x,
delta
=
'enter
'(default
format(/,6x,
'
*6x,
'minimum
*6x,
'
6x,
'
at
'
spacing
beginning
0.=
of
arclength',
autospacing)',t59,'>',$)
end
0.=
of
arclength',
autospacing)',t59,'>',$)
at
beginning:',gl2.5,/,
end:',gl2.5,/,
'maximum
format(6x,
',g12.5
'computed
at
'
delta
=
*6x,
106
',g12.5
enter
(between
spacing
(i=',i3,',',i3,'):',g12.5,/,
spacing
the
(i=',i3,',
degree
O.
of
(tanh)
end
83
',i3,'):',g12.5)
stretching',/,
and
I.
(sinh)
)'t59,'>',$)
APPENDIX
PAO_
85
_f
_.D,N_
NOT
FILMED
program
Joseph
Date:
sf
A. Garcia
Jan
1993
**********************W**WWWWW**WWWW*****W**WW**********
This
program
input
of
then
will
check
points
and
the
critical
the
streching
factor
flag
the user
when
value
of 1.3
in
the
this
(sf)
for
"sf"
case.
is
parameter(ii=201,jj=201,kk=3)
dimension
character*20
common
c
c ....
i000
read
xx(ii),yy(ii),del_x(ii),del__v(ii),del_r(ii)
ident
/corner/
in
xl,yll,ylu,xi,yil,yiu,x4,y41,y4u,x5,y51,y5u
locals
format(A)
CHARACTER*30
WRITE(*,'
infile
(a,$)')'
ENTER
FILE
NAME
'
READ(*,1000)infile
open(30,file=infile,
read(30,1000)
ident
read(30,*)
idim
write(*,*)
'idim
read(30,*)
(xx(i),yy(i),i=l,idim)
write(*,*)
xx(1)
write(*,*)
xx(2)
do
20
c
20
',idim
i=l,idim
del_x(i)
= xx(i)
del__y(i)
del_r(i)
= yy(i)
 yy(il)
= sqrt((del_x(i))**2
 xx(il)
continue
do
25
sf
sf
= del_r(i)/del_r(il)
if
i=3,idim
(xx(i)xx(il))/(xx(il)xx(i2))
(sf.lt.l.0)
sf =
endif
if
i,sf,xx(i)
(sf.gt.l.3)
write(*,*)
endif
continue
stop
end
then
1.0/sf
write(*,*)
25
status='old',form='formatted')
then
.....
(del__v(i))**2)
given
larger
Programairf_2dsurf
Joseph A. Garcia
Date: Jan 1993
This program will read the output of the sixseries code
and then create an upper surface curve of the airfoil
with a zero thickness trailing edge section to be
used as the 2d surface grid on VISUAL GRID for
redistribution.
parameter(idim=200,jdim=200,kdim=5)
dimension
x_U(idim),z_U(jdim),x_te(100),z_te(100)
integer
real
npts,
npts_0
delx_te
character*20
I000
FORMAT
name,
WRITE(*,'(a,$)')
READ(*,1000)
airfoil,wing
READ(*,1000)
WRITE(*,
'Enter
the
input
'Enter
the
airfoil
the
number
file
name
:'
infile
WRITE(*,'(a,$)')
in
consideration:
' (a,$)')
'Enter
pts
in
the
zero
npts_0
format(I3)
open(25,file=infile,status='old',form='formatted')
read(25,1000)
name
read(25,*)
npts
read(25,*)
(x_U(i),z_U(i),i=l,npts)
c
delx_te
c
c
c
.5/npts_0
WRITE(*,'(a,$)')
'Enter
READ(*,1000)
an
output
file
wing
open(26,file='airf.crv',status='old',form='formatted')
WRITE(26,1000)
name
WRITE(26,40)
npts
WRITE(26,50)
do
2O
20
+ npts_0+l
(x_U(i),z_U(i),i=l,npts)
i=l,npts_0
x_te(i)
z_te(i)
continue
0.0
WRITE(26,50)
40
5O
'
airfoil
READ(*,10)
i0
infile,
(A)
delx_te*(il)
(x_te(i),z_te(i),i=i,npts_0+l)
format(I4,1x,'Upper
format(el4.8,3x,
Coordinates:)
el4.8)
stop
END
88
name
'
section:'
Program
S3d_airf
c
c
By:
Joseph
A.
Garcia
c
c
This
"airf.ord"
program
cut
will
create
WingSurf
WingSurf
airfoil
generator
surface
grid
ordinate
from
file
the
modified
first
on
S3d
c
c
Date:
Jan
22,
1993
c
parameter(idim=200,jdim=200,kdim=5)
dimension
x_U(idim),z_U(jdim),
x_te(20),
,IDM(5),JDM(5),KDM(5),X(idim,
,Y(idim,
jdim, kdim),Z(idim,
INTEGER
REAL
npts,
npts_0,
delx_te,
ii,IGRID,
TE_Ingth,
character*20
te(20)
form_test
delwk
name,wing,
Defaults
jdim, kdim)
jdim, kdim)
infile,airfoil,outfile,
formm
c
npts_0
25
i000
FORMAT(A)
WRITE(*,'(a,$)
READ(*,1000)
') 'Enter
infile
WRITE(*,'(a,$)')'If
READ(*,*)form_test
if
the
file
( form_test
.eq.
formm
= 'formatted'
input
is
file
fomatted
name
type
:'
1
or
if
unform.:'
l)then
else
formm
'unformatted'
endif
WRITE(*,
'(a,$)')
READ(*,1000)
WRITE(*,
' (a,$)
READ(*,10)
FORMAT(f3.1)
i0
'Enter
the
airfoil
in
consideration:
') 'What
do
you
want
to
use
as
TE_ingth
PRINT*,'formm=
',formm
open(7,file=infile,status='old',form=for_)
IF
( form_test
PRINT*,'FORMATTED'
READ(7,*)
15
.eq.
l)then
NGRID
READ(7,*)
DO
(IDM(IGRID),JDM(IGRID),KDM(IGRID),IGRID=I,NGRID)
IGRID=
I,NGRID
READ(7,*)
+
(((X(I,J,K),
+
+
I=I,IDM(IGRID)),J=I,JDM(IGRID)),K=I,KDM(IGRID)),
(((Y(I,J,K),
+
+
+
15
I=I,IDM(IGRID)),J=I,JDM(IGRID)),K=I,KDM(IGRID)),
(((Z(I,J,K),
I=I,IDM(IGRID)),J=I,JDM(IGRID)),K=I,KDM(IGRID))
CONTINUE
ELSE
PRINT*,'UNFORMATTED'
READ(7)
READ(7)
DO
20
'
airfoil
NGRID
(IDM(IGRID),JDM(IGRID),KDM(IGRID),IGRID=I,NGRID)
IGRID=
I,NGRID
READ(7)
89
the
TE
sec
ingth:'
(((X(I,J,K),
+
+
I=I,IDM(IGRID)),J=I,JDM(IGRID)),K=I,KDM(IGRID)),
(((Y(I,J,K),
+
+
I=I,IDM(IGRID)),J=I,JDM(IGRID)),K=I,KDM(IGRID)),
(((Z(I,J,K),
I=I,IDM(IGRID)),J=I,JDM(IGRID)),K=I,KDM(IGRID))
CONTINUE
20
ENDIF
C
WRITE(*,'
(a,$)')
'Enter
READ(*,1000)
outfile
the
output
file
name
open(26,file=outfile,status='new',form='formatted')
WRITE(26,1000)
airfoil
C
C
This
section
into
the
C
C
id
will
put
airfoil
the plot3d
mg coordinates
ordinate
format
C
C
npts
ii
do
30
ii
IDM(1)
= npts
i=l,npts
= ii  1
j=l
k=l
3O
x_U(i)
=X(ii,j,k)
z_U(i)
continue
=Z(ii,j,k)
Writing
out
the
WRITE
(26,40)
WRITE(26,50)
ordinates
npts
(x_U(i),z_U(i)
WRITE(26,50)
WRITE
airfoil
,i=l,npts)
(x_U(i),z_U(i),i=l,npts)
(25,50)
(((X(I,J,K),
I=I,IDM(1)),J=I,JDM(1)),K=I,KDM(1)),
(((Z(I,J,K),
I=I,IDM(1)),J=I,JDM(1)),K=I,KD_4(1))
WRITE(26,55)
(x_te(i),z_te(i)
,i=l,npts_0)
WRITE
(26,45)
npts
WRITE
(26,55)
(x_U(i),z_U(i),i=l,npts)
WRITE
(26,55)
(x_U(i)
WRITE
(26,55)
(x_te
,z_U(i),i=l,:]pts)
(i) , z_te
(i) , i=l,npts_0)
delx_te
= x_U(npts)
 x_U(nptsl)
C
C
This
section
wing
wake
npts_0
C
C
will
estimate
# pts
= NINT(TE_ingth/delx
te)
AAA^AA_AAA^AAA_A^AAAA_AAAA^A^AA^AA
npts_0
= 0
dt0
= delx_te
dt2t
do
35
needed
section
0.
delwk
0.1*TE_ingth
i,I00
9O
in
the
'
dr0
if(
dt2t
= dt0*l.2
.it.
npts_0
=
dt3t
= dt0
dt2t
dt3t
else
delwk)
npts_0

then
1
dt3t/l.2
continue
endif
35
continue
A_AAAA_A_A_A_A_AAAAA_A_AA_
WRITE(26,60)
npts_0
WRITE(26,65)
delx_te
WRITE(26,70)
TE_ingth
40
format(I4,1x,
45
format(I4,1x,'Lower
50
format(el4.8,3x,el4.8)
55
format(el4.8,3x,el5.8)
60
format(I4,1x,
65
format(el4.8,1x,
70
format(f3.1,1x,
'Upper
Coordinates')
Coordinates')
'= Number
'= Delta
'= Length
of
pionts
X
of
stop
END
9!
in
in
the
the
TE
the
TE
zero
zero
TE
zero
section')
section')
section')
APPENDIX
PAGE
93
BLANK
NOT
FILME_
#!
/bin/sh
JN='LRE60WINGI_caseI'
SN:'COSAL_iI8000'
MK='make4.eagle'
#
COMPILE
THE
CODES
#
if test
then
mv
fi
Tran_rpt_n8.p3d.old
Tran_rpt_n8.p3d.old
if test
then
mv
fi
s
s
Tran_rpt_n8.p3d.older
Tran_rpt_n8.p3d.new
Tran_rpt_nS.p3d.new
Tran_rpt_n8.p3d.old
cd
/uO/rfa/jgarcia/stab_src_dir
make
f SMK
cp
mv
mv
cosal_4.exe
wing.exe
stabin.exe
/uO/rfa/jgarcia/$JN/$SN
/uO/rfa/jgarcia/$JN/$SN
/u0/rfa/jgarcia/$JN/$SN
mv
getstab.exe
/uO/rfa/jgarcia/$JN/$SN
cp
interp_n5_8_p3d.exe
/u0/rfa/jgarcia/$JN/$SN
mv plot3d_tran.exec
/u0/rfa/jgarcia/$JN/$SN
cd
/u0/rfa/jgarcia/$JN/$SN
chmod
+x
*.exe
mkdir
jobl
cd jobl
#cp
/u0/rfa/jgarcia/stab_run_dir/run*
cp
/uO/rfa/jgarcia/stab_run_dir/run*
/uO/rfa/jgarcia/$JN/$SN/jobl
if test
s Tran_frnt.p3d.old
then
mv
fi
Tran_frnt.p3d.old
if test
then
mv
fi
s
s
Tran_frnt.p3d.old
Tran_rpt_n8.p3d.old
Tran_rpt_n8.p3d.old
if test
then
Tran_frnt.p3d.older
Tran_frnt.p3d.new
Tran_frnt.p3d.new
if test
then
mv
fi
s
/u0/rfa/jgarcia/$JN/$SN
Tran_rpt_n8.p3d.older
Tran_rpt_n8.p3d.new
mv
Tran_rpt_nS.p3d.new
Tran_rpt_n8.p3d.old
fi
mv
cosal.time
cosal.time.old
#
#
THIS
IS WHERE
THE
LOOP
FOR
THE
SPECIFIED
SPAN
STATION
STARTS
#
#for
case
in fort.73
fort.74
fort.75
fort.76
fort.77
fort.78
fort.79
fort.80
fort.81
fort.82
fort.83
fort.84
fort.85
95
fort.86
fort.87
fort.88
fort.89
fort.90
fort.91
fort.93
fort.94
fort.95
fort.96
#
#for case in fort.74
fort.75
fort.77
fort.79
fort.92
#
for
case
fort.85
#
do
# CLEAN
in
fort.74
fort.75
fort.77
fort.79
fort.81
fort.83
fort.87
UP
THE
OUTPUT
FILES
#
rm
rm
rm
stab.out
fort.7
cosal.out
rm
rm
rm
rm
int_nl0.out
int_n8.out
wing.out
fort.2
#
#
EXECUTE
THE
INPUT
FILE
#
nice
../stabin.exe<../$case
#
#
EXECUTE
THE
B.L.
CODE
#
nice
../wing.exe<
fort.2
>
wing.out:
#
#
#
THIS
IS
THE
START
OF
THE
STAB
CODE
ANALYSIS
LOOP
#
#for
run
in runl
run25
#
for
run
in runl
run2
run3
run4
runll
runl2
runl3
runl4
runl5
run21
run22
run23
run5
runl6
run6
runl7
run7
run8
run9
runl0
runl8
runl9
run20
#
#for
runl0
run
in
runll
run0
runl2
runl
run2
run3
run4
runl3
runl4
runl5
run5
runl6
run6
runl7
run7
run8
runl8
run9
#
do
#
touch
cosal.time
echo
"Timing
information
for
running
cosal_4.exe:"
>>
cosal.time
date
>> cosal.time
##
/bin/time
nice
../cosal_4.exe
< ../$run
> cosal.out
cosal.time
/bin/time
nice
../cosal_4.exe
< _run
> cosal.out
2>>
cosal.time
date
>> cosal.time
2>>
#
#
APPEND
THE
STAB.OUT
INFO,
cosal.out
>>
T_KEN
#
../getstab.exe
<
stab.out
96
FROM
THE
COSAL.OUT
FILE
mv
cosal.out
done
#
#
THIS
cosal.out.bak
IS
THE
END
OF
THE
LOOP
#
nice
../interp_n5
8 93d.exe<../$case
#
if test
then
"$case"
"fort.73"
mkdir
stat_c3al
cp stab.out
stat_c3al/stab_sl.out
cp
int_n8.out
stat_c3al/int_nS.sl
cp
fi
int_nl0.out
stat_c3al/int_nl0.sl
#
if test
then
mkdir
cp
cp
cp
fi
"$case"
"fort.74"
stat_c3a2
stab.out
int_n8.out
int_nl0.out
stat_c3a2/stab_s2.out
stat_c3a2/int_n8.s2
stat_c3a2/int_nl0.s2
#
if test
then
mkdir
cp
cp
cp
fi
"$case"
"fort.75"
stat_c3a3
stab.out
int_nS.out
int_nl0.out
stat_c3a3/stab_s3.out
stat_c3a3/int_n8.s3
stat_c3a3/int_nl0.s3
#
if test
then
mkdir
cp
cp
cp
fi
"$case"
"fort.76"
stat_c3a4
stab.out
int_n8.out
int_nl0.out
stat_c3a4/stab_s4.out
stat_c3a4/int_nS.s4
stat_c3a4/int_nl0.s4
#
if test
then
"$case"
"fort.77"
mkdir
stat_c3a5
cp
stab.out
stat_c3a5/stab_s5.out
cp
int_nS.out
stat_c3a5/int_n8.s5
cp
int_nl0.out
stat_c3a5/int_nl0.s5
fi
#
if test
then
"$case"
"fort.78"
mkdir
stat_c3a6
cp stab.out
stat_c3a6/stab_s6.out
cp
int_n8.out
stat_c3a6/int_n8.s6
cp
int_nl0.out
stat_c3a6/int_nl0.s6
fi
97
#
if test
then
"$case"
"fort.79"
mkdir
stat_c3a7
cp stab.out
stat_c3a7/stab_s7.out
cp int_n8.out
stat_c3a7/int_n8.s7
cp
int_nlO.out
stat_c3a7/int_nlO.s7
fl
#
if test
then
"$case"
"fort.80"
mkdir
stat_c3a8
cp stab.out
stat_c3a8/stab_s8.out
cp
int_n8
out
stat_c3a8/int_n8.s8
cp
int_nlO.out
stat_c3a8/int_nlO.s8
fi
#
if test
then
"$case"
"fort.81"
mkdir
stat_c3a9
cp
stab.out
stat_c3a9/stab_s9.out
cp
int_n8
out
stat_c3a9/int_n8.s9
cp
int_nlO.out
stat_c3a9/int_nlO.s9
fi
#
if test
"$case"
= "fort.82"
then
mkdir
stat_c3alO
cp stab.out
stat_c3alO/stab_slO.out
cp int_n8
out
stat_c3alO/int_n8.slO
cp int_nlO.out
stat_c3alO/int_nlO.slO
fi
#
if test
"$case"
then
mkdir
stat_c3all
cp
cp
cp
fi
#
"fort.83"
stab.out
stat_c3all/stab_sll.out
int_n8
out
stat_c3all/int_n8.sll
int_nlO.out
stat_c3all/int_nlO._ll
if test
"$case"
then
mkdir
stat_c3al2
cp
cp
cp
fi
#
"fort.84"
stab.out
stat_c3al2/stab_sl2.out
int_n8
out
stat_c3al2/int_nS.sl2
int_nlO.out
stat_c3al2/int_nlO.sl2
if test
then
"$case"
"fort.85"
mkdir
stat_c3al3
cp
stab.out
stat_c3al3/stab_sl3.out
cp
int_n8
out
stat_c3al3/int_n8.sl3
98
cp
fi
int_nlO.out
stat_c3al3/int_nlO.sl3
#
if test
"$case"
then
mkdir
stat_c3al4
cp
cp
cp
fi
"fort
86"
stab.out
stat
c3al4/stab_sl4.out
int
n8.out
stat_c3al4/int_n8.sl4
int_nlO.out
stat_c3al4/int_nlO.sl4
#
if test
"$case"
then
mkdir
stat_c3al5
cp
cp
cp
fi
stab.out
int_n8.out
int_nlO.out
"fort
87"
stat_c3al5/stab_sl5.out
stat
c3al5/int_n8.sl5
stat_c3al5/int_nlO.sl5
#
if test
then
mkdir
cp
cp
cp
fi
"$case"
"fort
88"
stat_c3al6
stab.out
int_n8.out
int_nlO.out
stat_c3al6/stab_sl6.out
stat_c3al61int_n8.sl6
stat_c3al6/int_nlO.sl6
#
if test
then
"$case"
"fort
89"
mkdir
stat_c3al7
cp stab.out
stat
c3al7/stab_sl7.out
cp
int_n8.out
stat_c3al71int_n8.sl7
cp
int
nlO.out
stat_c3al7/int
nlO.sl7
fi
#
if test
then
"$case"
"fort
90"
mkdir
stat_c3al8
cp stab.out
stat_c3al8/stab_sl8.out
cp
int_n8.out
star
c3al81int_n8.sl8
cp
int
nlO.out
stat_c3al8/int_nlO.sl8
fi
#
if test
then
"$case"
"fort
91"
mkdir
stat_c3al9
cp stab.out
stat_c3al9/stab_sl9.out
cp
int
n8.out
stat_c3al91int
n8.s19
cp int_nlO.out
stat_c3al9/int_nlO.sl9
fi
#
if test
then
mkdir
"$case"
"fort
92"
stat_c3a20
99
cp
cp
cp
fi
stab.out
int_n8.out
int_nlO.out
stat_c3a20/stab_s20.out
stat_c3a20/int_n8.s20
stat_c3a20/int_nlO.s20
#
if test
then
"$case"
"fort.93"
mkdir
stat_c3a21/
cp stab.out
stat_c3a21/stab_s21.out
cp
int_n8.out
stat
c3a21/int_n8.s21
cp
int_nlO.out
stat_c3a21/int_nlO.s21
fi
#
if test
then
"$case"
"fort.94"
mkdir
test_stat_c3a22/
cp
stab.out
test_stat
c3a22/stab_s22.out
cp
int_n8.out
test_stat_c3a22/int_n8.s22
cp
int_nlO.out
test_stat_c3a22/int_nlO.s22
fi
#
if test
"$case"
= "fort.95"
then
mkdir
stat_c3a23/
cp stab.out
stat_c3a23/stab_s23.out
cp int_n8.out
stat_c3a23/int_n8.s23
cp
int_nlO.out
stat_c3a23/int_nlO.s23
fi
#
if test
"$case"
= "fort.96"
then
mkdir
stat
c3a24/
cp
cp
cp
fi
stab.out
stat_c3a24/stab_s24.out
int_n8.out
stat_c3a24/int_n8.s24
int
nlO.out
stat
c3a24/int_nlO.s24
#
done
rm cosal.out
rm
int
nlO.out
rm
int_n8.out
rm
stab.out
rm
rm
rm
wing.out
fort.2
fort.7
100
Form
REPORT
DOCUMENTATION
PAGE
Approved
oue No ozo4olee
Public reporling burden for this collection of information is estimated to average I hour per response, including the time for rewewing instructions, searching existing data sources,
gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this
collection of information, including suggestions for reducing this burden, to Washington Headquarters
Services, Directorate for information Operations and Reports, 1215 Jefferson
Davis Highway, Suite 1204, Arlington, VA 222024302,
and to the Office of Management and Budget, Paperwork Reduction Project (07040188),
Washington. DC 20503.
1.
AGENCY
USE
ONLY
(Leave
blank)
2.
REPORT
DATE
December
4.
TITLE
AND
3.
REPORT
1994
TYPE
Technical
SUBTITLE
Parametric
AND
DATES
:5.
Study on Laminar
FUNDING
NUMBERS
Speeds
6.
COVERED
Memorandum
53707
AUTHOR(S)
PERFORMING
ORGANIZATION
NAME(S)
AND
8o
ADDRESS(ES)
PERFORMING
REPORT
Ames Research
Center
9.
ORGANIZATION
NUMBER
SPONSORING/MONITORING
AGENCY
A94146
NAME(S)
AND
10.
ADDRESS(ES)
SPONSORING/MONITORING
AGENCY
National Aeronautics
and Space Administration
Washington, DC 205460001
11.
SUPPLEMENTARY
Unclassified
13.
DISTRIBUTION/AVAILABILITY
Subject
NASA TM108852
STATEMENT
12b.
Field, CA 940351000;
DISTRIBUTION
CODE
 Unlimited
Category
ABSTRACT
NUMBER
NOTES
Point of Contact:
12a.
REPORT
(Maximum
02
200
words)
Laminar flow control has been identified as a key element in the development of the next generation of High
Speed Transports. Extending the amount of laminar flow over an aircraft will increase range, payload, and altitude
capabilities as well as lower fuel requirements,
skin temperature,
and therefore the overall cost. A parametric
study to predict the extent of laminar flow for finite wings at supersonic
speeds was conducted using a
computational
fluid dynamics (CFD) code coupled with a boundary layer stability code. The parameters
investigated in this study were Reynolds number, angle of attack, and sweep. The results showed that an increase
in angle of attack for specific Reynolds numbers can actually delay transition. Therefore, higher lift capability,
caused by the increased angle of attack, as well as a reduction in viscous drag, due to the delay in transition, can
be expected simultaneously.
This results in larger payload and range.
14.
SUBJECT
Laminar,
15.
TERMS
Supersonic,
NUMBER
OF
PAGES
11l
Wings
16.
PRICE
CODE
20.
LIMITATION
A06
17.
SECURITY
OF
CLASSIFICATION
REPORT
Unclassified
NSN
7540012805500
18.
SECURITY
OF
THIS
CLASSIFICATION
PAGE
19.
SECURITY
OF
CLASSIFICATION
OF
ABSTRACT
ABSTRACT
Unclassified
Standard
Prescribed
Form
by
ANSI
298
Std.
(Rev.
Z3918
289)