Analysis of Three D 00 B Our

DOE/ER/03077-180
Courant
Institute of
Mathematical Sciences
An
Analysis of Three-Dimensional
Transonic Compressors
Antoine Bourgeade
Prepared under Contract DE-AC02-76ER03077 with

the U. S. Department of Energy and
NASA-Ames
Research Center Grant NGR-33-01 6-201.
Research and Development Report

February 1983
New
York University
UNCLASSIFIED
New York University Courant Institute of Mathematical Sciences
Mathematics and Computing
DOE/ER/03077-180
AN ANALYSIS OF THREE-DIMENSIONAL
TRANSONIC COMPRESSORS
Antoine Bourgeade
February 1983
Prepared under Contract No. DE-AC02-76ER03077 with the U. S. Department of Energy and NASA-Ames NGR-33-016-201. No. Research Center Grant
UNCLASSIFIED
Contents
Page
iv
ABSTRACT
I.
II.
INTRODUCTION
THE PARTIAL DIFFERENTIAL EQUATIONS OF MOTION
III.
IV.
V.
GRID GENERATION
FINITE DIFFERENCE SCHEME
10 22
RESULTS AND FIGURES

HOW TO USE THE CODE
28
VI.
49
58
61
VII.
BIBLIOGRAPHY
VIII. LISTING OF THE CODE
t\
AN ANALYSIS OF THREE-DIMENSIONAL TRANSONIC COMPRESSORS
Antoine Bourgeade
February 1983
ABSTRACT
This
presentation
sets
forth
method
for
computing
the
or
three-dimensional transonic flow around the blades of a

of
a
compressor
propeller.
The
method
is
based
on
is
the use of the velocity
potential, on the hypothesis that the flow
inviscid,
irrotational
and isentropic
The
equation
of
the
potential is solved in
a
transformed space
plane
such that the surface of the blade is mapped into
where
the
periodicity
is
implicit.
the
This equation is in a nonconservative form

of
and is solved with
help
finite
difference
method
using
artificial viscosity and artificial time.

A
computer
code is provided and some sample results are given in
order to demonstrate the influence of three-dimensional effects and the

blade's rotation.
-1I.
INTRODUCTION
Many scientists and engineers continue to study the ways and means
of
making better use of the energy

the
resources
at
our
disposal,
even
though
energy
crisis
is
perhaps no longer considered so severe.
Turbines and compressors, which both create and consume energy, have in
recent
years
been
the
subject
of
many theoretical and experimental
studies aimed at improving their design and efficiency.
Although
some
major
improvements
have
been
introduced
lately,
the
a
study of the
transonic flow across a single stage of a turbine, or of

still
remains highly complex.
compressor,
For this reason most

the
of
the theoretical
work done so far has focused on
two-dimensional
cascade
problem
[13-17] or on the mean flow problem [22].
Let
us first discuss the physical background.
A basic
compressor
rotors
and
consists of a succession of
rotors
and
stators.
These
stators are situated on the hub and surrounded by the cowling; they are
composed
shape
of
certain number of blades distributed around the hub,

blades depending on their use.
a
If
the
of the
the hub is cut along a
generator line and transformed into a plane,

these blades is obtained.
The
so-called
"cascade"
of
A propeller is a compressor without cowling.
present study deals with the problem of transonic flow around
compressor or propeller blades.

this
type,
From the mathematical point
of
view,
leads
in
us
to a
system of partial differential equations of mixed

unknowns
the
which
the
of
are
the
geometric
a
and
physical
characteristics
these
compressor.
complex
of
For
to be
three-dimensional analysis
integrated
that
equations
are
too
without
some
simplifications.
First
all,
we
suppose
the fluid we are

a
concerned with is a polytropic and nonviscous gas and that
velocity
-2-
potentlal exists.
This necessitates a further hypothesis, namely, that

is
the variation of the entropy
small
so
that
the
entropy
itself
remains
essentially constant.
equivalent
Thus our system of partial differential

to
a
equations becomes
single
equation,
the
potential
equation, in both the steady and the time-dependent cases.
The
potential equation has already been solved numerically in the

[2,14,15],
three-dimensional case for oblique and swept-wings
However,
because
of
the periodicity of the compressor problem, the square-root
transformation used in those works is not practical here.
We therefore
propose
new transformation, which maps the surface of the blade into

includes periodicity implicitly.
to
a plane and
The scheme we use is similar
the
one
used
by
Jameson
and
Caughey
We
by
[15] in the development of the swept-wing code known as FL022.
solve the finite-difference approximation to the potential

row relaxation.
equation
A typical run consists of 50 iterations on a 48x6x4
grid,
15
followed by 100 iterations on a 96x12x8 refined grid.

on the CDC 6600 and
3
This takes
minutes
minutes on the STAR.
The simplicity
of our grid generation, together with the other hypotheses given above,
limits
our
choice
of
blade
geometries.
Nevertheless, this method

the
enables us to study how the speed of rotation influences

flow
relative
around
the
blades,
and
we have compared our three-dimensional
results with data for two-dimensional cascades.

In Section
2
we shall derive the equations of motion
in
physical
space
and,
in
Section
3,
we shall consider the potential equation,

of
which is obtained after several changes

scheme
is
variable.
The
numerical
presented
5
in Section 4. The numerical results obtained are
shown in Section
together with some Calcomp plots,
while
Section
-3-
provides
manual
on the use of our computer code.
Sections
and
contain
bibliography and a listing of the code.

study
has
This
been by
1
supported
U.
S.
by
NASA
under
Grant
No.
NGR-33-016-201
and
the
Department of Energy under Contract
DE-AC-02-76ERQ3077 and
gratitude
F.
take this opportunity of expressing my sincere
to
Prof. P. Garabedlan for his Invaluable advice and to Dr.

I
Bauer for her constant encouragement.
am
also
Indebted
to
my
entire family for their moral support.
II.
THE PARTIAL DIFFERENTIAL EQUATIONS OF MOTION
This
section
sets
forth
the
various
equations
used
in
our
computational method and in the resulting code.

The
general
equations
of
fluid dynamics for an inviscid gas are
well known [6].

(a)
These are:
the equation of conservation of mass,
(1)
p t
+ pu x +
pv+pwz
ii
+ PyV +
p zw
0,
where
(x,
is
the
density,
(u,v,w)
are
the
velocity
components,
and
y,z) are the coordinates in physical space;
(b)
the equations of conservation of momentum,

p(u
t
+ uu x + vu
+ wu z
+ Px =
=
(2)
p(v t + U v x + vv + wv z ) + P y y
p (w t
+ uw x + vw
+ wwz
= ^ + p z
'
where P is the pressure; and

(c)
the equation of conservation of energy,
(3)
=0 at
S
where
is the entropy.
It
will be recalled that
+ u
dt
3t
3x
+ V
3.3
+ w
3y
3z
is
the material time derivative.
We are neglecting here such external
actions as gravity.
For the following calculations we shall assume that the entropy is
-5constant throughout the fluid; the last equation will therefore not
used.
But then, in order to represent
be
weak shock waves which occur in

the
a
transonic
conditions.
flows,
we
is
have
done
to
replace
Rankine Hugoniot
shock
This
by
permitting
jump in the horizontal
component
of
momentum and by adding artificial viscosity terms
to
the
partial differential equations.

numbers close to
1.
The approximation is adequate for Mach
Let us suppose now that the fluid is a polytropic gas.
Therein:
(4)
P =
A(S)p^
where the function A(S), on the basis

constant, and where
of
our
hypothesis,
becomes
is
the
adiabatic exponent of the gas, i.e.
the ratio of the specific
heats at constant pressure and at constant volume.

If
c
is the
local speed of sound in the gas, we have
(5)
c*-. dp
so that equations (2) become

o
p(u t + uu x + vu
(6)
+ wu z
c p c p
2 2
x
y
= =
p(v t + uv x + vv
p (w t
+ wv z ) + + wwz ) +
+ uw + vw x
= c p z
-6Suppose now
of
t
tliat
the flow is steady,
i.e.
that the
flow is independent
ime
so tliat
(7)
!. =
i!i=
3t
3t
ll9t
^8t
Then from (6) and (7) we obtain
(8)
p(u x +
+ wz
_
v v
2
(u u
+ uvu + uwuz
y
+ vuv +
+ vwv + wuw + wvw + w 2 w
or,
if
we
assume the existence of

,
velocity potential,
<j>
and if we
divide by p/c 2
(9)
2
(<t>
xx+
yy
<,
zz )
- (u 2 *
xx+
yz +
v *
yy
+ w 2 zz
4
2uv<}>
xy
2vw<j)
2wu4i
zx
with u
<j>
=
<J>
w =
(j>
This is the potential
equation.
It
is
hyperbolic for supersonic flow,
u 2 + v2 + w 2 > c2
and elliptic for subsonic flow,
u 2 + v2 + w2 < c 2
Equations (4) and (5) show us that
-7dc z _ dp
(10)
Using the velocity potential, equations (6) become
fJ
,2 Xi + *; + x
2, ^) a
}
.2
7rTJx Y
(ID
(y
<
+ ) + ^y=
y
:j<*2 +
so that
+ ?)
+^rr)
(12)
I
z
(4>2
,(,2
+ 2j z
y~i
c0nstant
throughout
us,
the
fluid.
This is the Bernoulli equation, which enables
knowing the velocity potential, to compute the speed of sound.

Suppose next that the flow is no longer steady with respect to the
initial
frame
of
reference, but that, if we consider a frame turning

a
around the x-axis with
constant speed of rotation,
co
the
flow
is
again independent of time.
Then, instead of (7), we have
/,t N (15)
9p
9t
ojz -
9p
9y
+
,
djy -
9p
9z
9u =
9t
-to
z -
9u + 3y
toy -
9u 9z
9v
~
9t
= -
uz
9v + 9y
coy -
3v9w -
3z
,
= - a)z -
9t
9w 3w + wy 9z 3y
By
introducing the cylindrical coordinates (x,9,R), the equations (13)
become
9p
Ft
3p
3u
F3"'
Ft"
3u
3v _
'
FT
Ft
3v
39"
'
3w Ft
_ ~
~ w
3w
FT
These new equations show us that the flow dependence
on
and
is
characterized by a dependence on
In
(9
- ut) alone.
this time-dependent case a combination of equations (1) and (6)
gives us
(14)
p(u x + u
+ wz
2- (uu x+
c
+ v u v + vv u v u + u w u Q y z x y
+ v Q w v + w Q u w + w Q v w + w^w z z x y
where (uq,vq,wq) are the components of the relative velocity defined by
(15)
Uq = u
vq = v - wz
Wq = w + coy
If
we again assume
the existence of a velocity potential, it
must
now
satisfy the equation,
(16)
2
(<fr
xx+t}> yy-hj, zz )- (u^ xx+
v4>
yy
wg<t,
zz + 2u v
<t,
xy
2v
w Q yz + 2w u
4,
<J>
zx )
This
equation
is
similar
to
equation
(9);
it
is
second-order
nonlinear partial differential equation which is hyperbolic for
u2
+ v + wg
>
2
,
-9-
and elliptic for

2
15
2 2 + VQ + W Q
< c
We still have the relation (10)
for the speed of sound,
so
that
using
the
velocity
potential
and
with the help of (13), we obtain,
instead of (6)
(17)
(_ fj. 2^ 2^ 2 -a) Z ((fr^-h^)

1
+o)y*
+ÎT
and this gives us the new Bernoulli equation,
(1)
y ^x^y 4 *^
"
wz<f,
^z
4rf
Constan t
along each line parallel to the x-axis.
In the next section we
shall
transform
these
equations,
by
change of coordinates, into the system which is solved by our code.
-10III.
We
GRID GENERATION
have obtained the equations to be solved, but we have still to

slip
impose the geometrical constraints due to the
condition
on
the
surface
of
the compressor blade and
to the
periodicity condition.
space
in
In the
this section, we shall describe the
geometrical
which
equations
which
are
will
be
solved.
to
We shall therefore list all the mappings

the
performed
transform
physical
space
onto
the
computational space.
For
the
purpose
9
of
the
periodicity cylindrical
condition,
coordinates
we
begin
by
introducing the angle

respect to the x-axis.
of
To
the
(x,G ,R)
with
accomplish this we use a conformal mapping
(19)
(x,y,z) -
(x
,0
,Z
Q)
defined by
(z+iy)
=
exp(Z +i8),
x = xn
The Jacobian matrix associated with this transformation is

1
(20)
J
l
Q
P
-Q
with elements defined by
P =
cos
Q =
sin
9
, R = exp(Z n )
v
We
also require the inverse matrix, namely,
-111
JJ
z y
-y
z
->->
If
(i,j,k) is the orthonormal basis for physical space,

->
then the basis

*
v,
vectors connected with the new coordinates are
i,
u,
and
where
(21)
u = zj - yk
v = yj + zk
For the surface of the blade, the
representation
of
the
finite
difference
if
slip condition becomes greatly simplified and more accurate
the boundary surface lies on a
coordinate
plane.
The
idea,
cf.
[14],
is to transform the surface of the blade onto a plane which will
constitute the lower boundary of a half-space.
Unfortunately, if
reference
[14]
,
we
apply
the
square-root
transformation
of
the
periodic strip is transformed in such a way that

We therefore prefer
the periodicity condition is difficult to satisfy.

a
transformation
which
would map a periodic strip conformally onto a
half-space,
so that the periodicity condition becomes implicit.
The required transformation can be decomposed into two
successive
mappings.
For
the
purpose
of
simplification we consider only plane

The
first
sections orthogonal to the ZQ-axis.

periodic
strip
(cf.
mapping
transforms
Figure
la)
onto a slit plane (cf.

the
Figure lb).
real
The image of the two lines delimiting the strip is

axis.
We
negative
then apply a square-root transformation which maps this plane
onto a half-plane (cf.
Figure lc).
-12-
B.,
Figure la.
The periodic strip.
Figure lb.
The slit plane,
-13-
<
Figure lc
The half plane with
bump.
-14To Implement these transformations we draw in each plane section a
singular
line
(the
hranch cut) from inside the blade, near the nose,
out to downstream infinity.

has the coordinates (x g ,6
s
)
.
This is actually a half-line whose

We next perform the mapping
origin
(22)
(x ,9,Z
(X.Y.Z)
defined by
2
)
,
(x -x s
+ i(e-e
-n"
logfi +
(X+ * Y)
-z
where
and
are two constants.
N is more precisely the number of

6
blades on the compressor.

Z.
The coordinates x g and

respect
usually depend
9
on
Their
derivatives
with
to Z will be denoted x z and
respectively.
The Jacohian matrix determined by this transformation is

a
-b
a
f
(23)
b
e
whose elements are defined by
Hx x
= H6
e =
ax z - b6 z
- a6 + z
bxz
H = xx + 9 X
The inverse matrix is
-15X
J
X
X
z
~X
x
z
X
e
'
and the basis vectors related to this transformation are

with
now
(4,B,C),
(2 4)
A = Xyi + 9vii
B =
- Oî
+ xû
C = x
u + v
Let
us
return
to
the
potential
equation
(9),
it
can
be
reformulated as
(2 5)
c V
2
4,
(V4>
V) 2
=
<f>
where
9x
k 3z
If
is a notation for the
gradient.
we denote the
final Jacobian matrix
A = J
J2
hv
(a
and, if we use
d
-
-16-
for the derivatives with respect
to
the
new
coordinates,
then
the
derivatives with respect to the initial coordinates are represented hy
Jt
\
J
Ud
'
37
2j
d
.i
J^
l
j
3j
d
j
Thus the Laplacian, which is defined by
v2
=li + li+ li
9x 2
3y 2
3z 2
can be rewritten as
(26)
aij a ik d.jd k +
a ij a j!k d k.
where
the
coefficients
stand for the derivatives of the Jacobian
matrix elements;
ik
= d
a ik
Let us set
(F
.)
= F =
A C A
D
L
= V 2X
= V 2Y
= V 2Z
These notations allow us to simplify the equation (26) to arrive at
(27)
V2 =
Fj k d jdk +
Dk d k
j,k
For the second term of equation (25) let us denote the

of the velocity in the physical space by
components
17-
9<}>
di>
3$
Using the same computation as for the Laplacian, we obtain
(28)
(V^-V) 2 =
\B d k d
+
Jl
I I
C^d^
k,
->-->
where
(B^Bp.B^), the components
of
the velocity in the basis (A,B,C)
are given by
B.
1J+1
and where the coefficients
C.
are defined by
C
1
(V-j).V)
(V^j.V)
Y,
(V^V) 2
After these necessary but
somewhat
tedious
transformations
the
potential equation acquires the useful form
(29)
(c
ti
BiB^d^jO. +
I
1
(c\
- C
dtf -
i.j
This
has
the advantage of being relatively tractable,

In
for an equation
case,
which is, really, quite complicated.

need
only
to
the
the
time-dependent
we
replace
the
cf>.'s
by
components of the relative

of
velocity in the calculation

Moreover,
if
a
of
the
coefficients
equation
(29).
reduced potential is defined as the difference between

the
the true potential and the uniform flow potential corresponding to
-18inlet
speed, we can use it to compute the derivatives d d
.
4>
and
d.<|>
in
equation (29) without changing that equation.

We next describe the grid used in our computational method.
It
is
obtained
with
the help of a system of sheared coordinates.
These are
defined by considering coordinates parallel to the transformed

of
surface
the blade.
If
Y = S(X,Z)
is
the
equation
of
this
surface, the transformation in question is
defined by setting
T =
J = Y - S(X,Z)
K = Z
The Jacobian matrix is

1
-s
1
(30)
-s
and the related basis vectors are
(31)
= A
S YB
J = B
K = C +
S7B
In order to carry out the computations on a finite

last
domain,
these
coordinates are stretched so that the final domain of computation

a
becomes
cube
the
edges
of
which
have
length
1.
After
these
transformations equation (29) remains of the same general type, but its
coefficients
have
to
be
changed
slightly.
Complete
details
are
specified in the listing of the code in Section VIII.
-19It
remains to define the boundary conditions at the

and on the blade surface.
hub,
at
the
cowling,
We impose
to
a
slip condition for the
flow on these boundaries;
this leads
Neumann
problem
for
the
velocity potential, since the normal derivatives at the boundaries must

vanish.
We shall express
the
this condition in our system of coordinates.
The hub and
cowling are defined by equations of the form
Constant
At any boundary points

>
i x
-*
the two tangential vectors are
*-* and
i
>
u,
and v
u is a
normal vector.
The boundary condition at the hub and at the
cowling may thus be expressed by the equation
(32)
3<J>
y<J>
Z(
|
On the other hand
the blade surface is defined by the equation
J =
The two tangent vectors
and K lead to the normal vector
whoso
coordinates in the basis (i,u,v) are
S ve
S7
a = R"(9 K + s xx x)
'
8 =
X X" X
'
Hence the boundary condition on the blade surface is given by
-20-
(33)
ex
P- + Xa
&
|1 +Y
#--!
O rv
+ * + YYH 2 +
(Tz-6y)l3
In
equations (32) and

i.e.
(3 3),
(<j>
,<j>
3)
must be the derivatives of

the
in
the true potential,
the components
of
velocity.
For
the
time-dependent
case,
they
are
replaced,
equation
(33), by the
components of the relative velocity.

In the code the values
of
(ij>
,<j>
jA 3)
ar e
computed
only
when
necessary, and the only derivatives available are
u = i*
ax
v=il,
'
ay
w -
li
az
'
These derivatives are related to the components of the velocity through

the relation
(34)
4>
= A
* 3
W
(3 2)
The boundary conditions
and (33) become
(3 5)
eU + fV + W =
and
(36)
(oa + 6b + ye)U + (-ob + Ba + yf)V + yW =
In
three dimensional space an appropriate vortex
sheet
behind
blade
must be considered.
The shape of the vortex sheet is modeled by

In
our code in the following way.
each plane cross-section, the
upper
and
lower
surfaces of the blade are extended behind the trailing edge
-21by
two lines parallel
to a branch cut (Cf.
Figures 1,4,7).
These two
blade
is
lines represent the vortex sheet in that section, and if the
closed
they
are identical.
For two points situated on opposite sides

are
the
of the vortex
sheet
the
pressures
same,
and
the
normal
velocities are zero; only the tangential components of the velocity may
be different.
Since a shape
is
assumed
for
the
vortex
sheet,
we
require
only continuity of the normal component of the velocity across
the sheet.
Computationally this reduces to a condition like
(37)
(fr
YY
after a jump in
across
section.
this jump.
the
<j>
is removed.
Moreover,
supposed
the jump of
to
the
potential
vortex
sheet
is
be constant in each plane
The Kutta-Joukowsky condition at the trailing edge determines
This amounts to
linearized treatment of the vortex sheet.
-22-
IV.
FINITE DIFFERENCE SCHEME
The success of codes for the design and analysis of

wings
[1,2,3,15]
supercritical
shows how effective the computational fluid dynamics

The first step in this development
[1].
has become for transonic flow.
was
the design of shockless airfoils by the hodograph method
Then the
analysis
introduction of a retarded difference scheme

of
[20]
allowed the
flow at off-design conditions.
This scheme incorporated artificial

zone.
It
viscosity in order to capture shocks in the supersonic

then
was
[14]
improved
to
permit the analysis of three-dimensional wings

[11].
In
and their
design
our
computational
method
we
use
the
last-mentioned scheme, and this section explains how it is incorporated

into our computer code.
All the equations we have described in the

represented
in
the computer code by finite
previous
section
are
difference approximations.
of
The flow conditions at each grid point are determined with the help
the
reduced
potential
C.
The first derivatives of this potential are
calculated by using central differences.

the
of
With these values we
compute
approximate
velocity.
This allows us to determine, with the help

the
Bernoulli's equation (12), whether, at
point
considered,
the
flow is subsonic or supersonic.

If
the
flow
is
subsonic
the
second
derivatives used for the

are
computation of the potential
equation
residual
approximated
by
central differences of the form
(38)
i+l,j,k
2G
i,j,k
2
+ G i-l j,k
)
(AX)
-23-
and
(3 9)
G i+1, j+l,k " G i-H,j-l,k " G i-1, j+l,k + G t-1, j-l,k

'XY
_______
For
supersonic
points,
in
a
on
the other hand,

to
we have to introduce
artificial
viscosity
by
order
capture
weak
shocks.
[15].
This
is
accomplished
using
retarded
difference
of
scheme
Thus we
separate the equation (29) into two groups
terms:
(40)
U 2 -q 2
Gc.
(V
G)
where
first
is the speed and
s a
coordinate in
the
flow
direction.
The
the
term represents the second derivatives in the
flow direction;
derivatives in the other directions form the second term.

derivatives
used
in this
The
second
second
expression
are
computed by using
the
equations (38) and (39).
But for the second derivatives of
first
expression
types:
we
use retarded difference approximations of the
following
(AX)'
a nd
(42)
rovided that the velocity has positive components in the
direction
-24-
and in the
Y direction.
Equations (41) and (42) are
only
first-order
accurate and introduce the truncation errors
AXGyvy
- A VG X U XXX
'
"*"
AY GyvY
For
the
potential
equation
(40),
at supersonic points, these terras
represent a positive artificial
viscosity
which,
when
the
flow
is
aligned with the X direction, reduces to
(q
- c2
AX
G^x
as in the scheme of Murman and Cole [20],
Equations
(38)
to
(42) are used at each iteration of a run.
To
is
describe the iteration process, which is done by row relaxation, it

helpful
to
introduce
an
artificial
time
t,
which increases by the
quantity At at each iteration.

(39) and (41) become
The right-hand sides of equations (38),
G ?+ l,j,k-
^-^
G ij,k -
/"> G i,j,k+ G i-1, j,k
(AX)
G i+l,j+l,k" G i-l,j+l,k " G i+l
j-l ) k
+ G i-1, j-l.k
4AX AY
2G i,j,k- G i,j,k- 2G i-l,j,k+ G i-2,j,k

(AX)
2
where
the
superscripts
and
denote
new
or
old values of the
-2 5-
potenttal
and where w
is
the overrelaxat ion factor.
These expressions
represent approximations to
G YY -
xx
AX
(G Yt + Xt
.
(AX
(o
DG.), G YY XY
fc
I GYt AX
2
G YY + xx
^ AX
G,
>
Hence the equation solved is
(43)
i,j
(cV r
B^GX^
a,
+
?
l
1< C
V
are
0% +
a i GX
it
+6G t
where
the
coefficients
and
determined
by
the
new
approximations (38), (39) and (41).

By considering an orthonormal system of
coordinates (p,r,s), where
the flow direction is still the
direction, we arrive at the equation
(4)
(c 2 -q 2 )G sg +
2 q
2 rr+q Gpp
+6^+
B 2 G rt +
^pt
'
If
= t
1 _
Pi
s
$2
q"
^3
q"
-q-
is a new time coordinate,
then equation (44)
is transformed into
(45)
2 2 2 2 (c -q )G ss + q G rr + q G
(_c
+
q
o 2
)G
TT
+ 5GT
0.
-q
q"
In
order to ensure
of
the
the
convergence
of
the
iteration
scheme
to
a a
solution
steady-state equation, we want equation (45) to be
-26daraped
three-dimensional wave equation.

is
Given
Che
form
of
equation
(45)
this
equivalent to the condition
(46)
(-T2 c -q-
-T)
q
(c
>
>
'
At
subsonic points the damping condition is always satisfied.

points
(41)
it
At
of
supersonic
may
and
no
longer
be
true.
The
choice
approximations
(42) help to ensure a large coefficient g,
but near the sonic line this may not be enough.
One way to ensure that

by adding a term of the
condition
form
(46)
is satisfied, is to increase g,
U st
" e
((
fl G xt +
*2 G yt + *3 G zt>
= ^
|| (*iG Xt +
G Vt+ B G
3
Zt
to equation (43), where

increase
B
j
(3
is
positive
parameter
large
enough
to
by the right quantity.

,
The
mixed derivatives used above, G
Gy,-
G 7r
are obtained by
using approximations of the type
rN G
,
'
Xt
P PN + p i-l,j,k G i,j,k - G i,j,k ~ G i-l,j,k At AX

,
We have described
the difference scheme for interior points.
For
the
boundary
conditions at the hub and cowling, we transform equation value

of
(35) into a difference equation in order to compute the
the
reduced
potential
at these boundaries.
For the blade surface and the
-27vortex sheet we introduce ghost

points
are
if
points
behind
the
boundary.
These
used
to
compute the value of the potential at points on the

The values of the potential
at
blade as
the
(36),
the
they were interior points.

points
ghost
i.e.
behind
tlie
blade are determined by using equation
the slip condition.

is
For the points behind the vortex sheet

by
the
potential
determined
Kut ta-Joukowski condition.
To
an
compute the value of G at the
remaining
boundary
points
we
use
approximation to the outlet velocity.
The program permits not only compressor blades but also propellers
to be modeled.
Since propellers have no cowling, the

Thus,
inlet
speed
is
equal to the outlet speed.
for a propeller run, all the boundary

blade
and
on
the
conditions except those at the hub, on the

sheet are replaced by
vortex
G =
With
program FL022 as the starting-point, we have used the result
stated in this section to write a new program in FORTRAN named CSCDF22.

This program is listed in Section Vltl.
-28V.
RESULTS
In the preceding sections we described our

In
computational
method.
order
to
show
how
tliis
method works, we present in this section
seven examples of runs made with the program CSCDF22.
The first case was run to
test
the
validity
of
our
2
code.
compressor blade had been designed with code K [3].

result of this run.
Figure
shows the
We used the coordinates
of
this
two-dimensional
The profiles were
for
blade
profile
to
create a three-dimensional blade.
identical at each span station.

a
Since we chose to run the program
cascade
configuration and in the compressor case, we were solving a

The inlet speed and the
inlet
two-dimensional cascade problem.
angle
were the same as in the design run, but the blade was markedly cambered
and we chose a gap-to-chord ratio of 1.5.
Figures
3
and
show
representation of the cascade and the grid at the hub which is the same
at each span station.
The pressure distribution is given in Figure

run
5.
The other six examples are all
with
the
same
blade.
The
profile is defined at three different span stations.

is the NACA-0012 profile.
The basic profile
At each span station the chord and

the
thickness
are
the
same
but,
for
realistic axial flow configuration, the
coordinates of the singular points are
different.
The
blade
has
sweep
angle
of
14
degrees
and the dihedral angle decreases from 0.
degrees at the hub to -10.

equal
to
0.
degrees at the tip.

the
The angle of twist
is
degrees
at
hub,
2.
degrees at the cowling, and
1.
degree in between.
between
the
For
the
axial
flow
configuration
the
distance
hub
the
and the axis is 2.
This is also the distance between
the hub and
cowling (or the tip).

blade
is a
The first example with this
run
for
compressor
-29cascade configuration which appears in Figure representation

of
6.
Figures
and
show a
the
grid
and
the
pressure
8.
distribution.
is
the
The
different flow parameters are shown in Figure

number.
M2
Ml
inlet
Mach
is the outline Mach number.

the
DEV is the difference between
the outlet angle and
inlet angle given by ALP.
The four ne Kt runs are also analyses of compressor flow,

in
but
now
an
axial
flow configuration and with eight blades around the hub.

speed
of
For each run the parameter OM determining the
rotation
has
different
values.
These values are 0.0, 0.5,
1.0 and 2.0.
The other
at
flow parameters are identical.

the
Figures
and 10 represent the grid
hub
and
cowling.
It
will
be
noticed how different they are,
although the profile is almost the same at the hub and cowling; but the
gap-to-chord
ratio
is twice as
large at the cowling.
Figure 11 shows
The
how the blade looks in a plane orthogonal to the axis.
different
pressure
13,
distributions
obtained
for each run are given in Figure 12,

flow
14 and 15. the
These figures clearly illustrate how the

of
evolves
as
speed
rotation
increases: the shock on the upper surface

a
weakens and then disappears while, on the other hand,

on the lower surface and intensifies.
shock
appears
The
last
example
is
similar
to
the
L
fifth, except that it is a

the
in
propeller analysis.
case,
and
Figure 16, 17, and
show
geometry
Figure
5,
of
19.
the
by
the
pressure
distribution
19,
is
given
comparing Figures 14 and appearing

at
we
observe that, for example
the
shock
the cowling which is caused by the speed of rotation, is
amplified significantly by the cowling.

These numerical experiments show that our computat ional method, in
spite
of
its
restrictive
hypotheses,
enables
us
to
analyze
the
-30-
th ree-dimens tonal transonic flow around compressor blades.

first
step
This
Is
towards
the
study
of compressors or propellers in three
dimensions.
Considering what happened in the study of
transonic
flow
past swept wings, we venture to suggest that the next steps could be to
improve this method in order to include the
computation
of
the
wave
drag
and
to
design
blades
with
a prescribed
pressure distribution
[4,11].
-31-
1.2
INPUT
CP
OUTPUT
-.4
0.0
__
H _L
.8 __
1.2
MU.707
M2=.534
DEL TH= 35-00

Figure
2
G/Cr .99
-32-
CflSCRDE REPRESENTATION
G/C
1.
50
Figure
-33-
G-RID
ON THE SURFflCE
0.
00
Figure
-34-
PRESSURE DISTRIBUTION
Ml
=
BLADE PROFILE
DEV
5
.71,
M2
.57,
Figure
-32.0.
ALP
35.0
-35-
CASCADE REPRESENTATION
G/C =
1.50
Figure
-36-
GRID ON THE SURFACE
0.00
Figure
-37-
M1
BLADE PROFILE
.74,
M2 =
.70,
DEV =
.2,
ALP =
0.0
Figure
-38-
GRID ON THE SURFRCE
2.
00
Figure
-39-
GRID ON THE SURFACE
4.
00
Figure 10
-40-
CUT IN THE PLRNE
500
Figure 11
-41-
Ml
=
BLADE PROFILE
DEV
-
.75.
M2
71.
.2,
ALP
0.
Figure 12
-42-
M
l
BLADE PROFILE
DEV
=
1. 9,
75,
M2
70,
RLP
0.
Figure 13
-43-
Ml
=
BLADE PROFILE
DEV
=
.75.
M2 =
.70.
3.5.
ALP
0.0
Figure 14
-44-
Ml
=
BLADE PROFILE
DEV
=
.75,
M2 =
.69.
7.0.
RLP
0.0
Figure 15
-45-
GRID ON THE SURFRCE
2.
00
Figure 16
-46-
&RID ON THE SURFnCE
4.
00
Figure 17
-47-
CUT
IN
THE PLANE
500
Figure 18
-48-
Ml
=
BLADE PROFILE
DEV
=
.75,
M2 =
.75.
O.O,
ALP
0.0
Figure 19
-49-
VI.
HOW TO USE THE CODE
This section is intended to serve as a computer

listing
of
guide
code
for
users
of
our
code.
We
explain
the
different
options and give a
the input parameters.

the
Our computer code includes two options which permit

of
solution
it
a
either
the
compressor
problem
or
the propeller problem, and
treats them either in a realistic axial
flow
configuration
or
in
three-dimensional cascade representation.

a choice of two
In the
first option there is
boundary conditions.
In the propeller case the reduced
velocity
potential
is set to zero at infinity; in the compressor case
the boundary condition is given by a Neumann condition at the
cowling.
The first option determines the first mapping to be performed.
For the
of
axial flow configuration the whole space is mapped into a part

space
the
consisting of as many periodic strips as there are blades around

For the cascade
the hub.
configuration
this
first
mapping
is
not
required,
In
since the computation starts directly from a periodic strip.

cases
the
both
periodic
strip
is
mapped
conformally
into
quadrilateral.
For the compressor case the blade shape and the vortex
the
sheet define the bottom of the quadrilateral; the hub and
cowling
determine
infinity.
the
sides;
and
the
upper
edge is defined by the flow at
For the propeller case the side previously determined by the
cowling is defined by the flow at infinity.

The
shape
of
the
blade is defined by giving the coordinates at
different span sections from the hub to the cowling.
As
many
as
six
sections
can
z
be
defined
by giving the Cartesian coordinates at each

If
section with
constant.
two
sections
are
similar,
only
the
coordinates of the first section are read in.
-50We shall now
describe
occur
the
input
data
cards
and
explain
the
parameters
which
in the input.
All the data cards are read in

each
data
card
or
format (8F10. 6).
A title card is inserted before
each series of data cards.
Title Card
1:
NX, NY, NZ, FPLOT, XSCAL, PSCAL, GRIDX, GRIDY.

NX,
NY,
NZ
are
the
number
of mesh cells in each direction of the
transformed space.
in
Thus NZ is the number of mesh
cells
the
radial
direction,
NX the number of mesh cells

,
along the blade and the vortex sheet

of
and NY the
number
of
mesh
cells
in
If
the
third
,
direction
the
computational grid.
NX =
the program stops.

,
The
dimensions in the code are NX+1
NY+2, NZ+3.
FPLOT
is
the
parameter
controlling the generation of plots.

but
no
If
IF FPLOT = 0, a printer plot
Calcomp
FPLOT
=
plot
1,
is
obtained
at
each
span
station.
both a
and
if
printer plot and a Calcomp plot are
generated;
FPLOT = 2, only a Calcomp plot is generated.

XSCAL, PSCAL
determine
the scales of the Calcomp plot.
The pressure
scale is set to PSCAL per inch
in
each
section
plot.
PSCAL
is
equivalent to PSCAL = 0.5.
If PSCAL is
5.
positive, each section is scaled so that the span is

If
PSCAL
is
negative,
each
section
is
scaled
proportionately to the local chord and the maximum chord

is XSCAL/2.
GRIDX, GRIDY
control
the
grid generation.
of
The grid depends largely

so
on
the
geometry
the
compressor,
these
two
-51-
parameters must be chosen carefully so as to acceptable

grid.
obtain
an
GRIDX
has to be greater than -1 and

the
GRIDY is positive.
In most of
cases
we
can
use
GRIDX
and
GRIDY =
1.
Decreasing GRIDX gives more

at
points near the nose and
less
the
trailing
edge.
Increasing
GRIDX
gives
fewer points near the nose and
more at the trailing edge.

the
Increasing
the
GRIDY
increases
number
of
points
near
blade and the vortex
sheet
Title Card
2:
MIT, COV, PI, P2, P3, BETA, FHALF
FCONT.
MIT
COV
is the maximum number of iterations computed.

is the desired accuracy.
If
the maximum
correction
is
less
than
COV,
the
program
will
10.
goes to the next grid or
terminates.
The
same
happen
if
the
maximum
correction is greater than

P1,P2
are
the
relaxation parameters for the subsonic and the

1.
supersonic points respectively PI is between

P2 is less than or equal to P3
is
1,
and
2.,
1.
used
for the boundary values.
P3 is usually set to
but a smaller value improves the convergence for very
cambered blades.
BETA
is the damping parameter controlling the amount of added
d>
T s ... t
It
is normally set between 0. J
and 0.5.
FHALF
indicates whether the mesh is to be refined.

=
For
FHALF
the
program
stops after convergence or after the
-52-
prescribed number of
For FHALF =
1
iterations
have
been
completed.
the mesh will be refined and another input

the
be
card must be read for
parameters
refined
of
the
refined
mesh.
The
mesh
can
only
twice and the
maximum grid is 128x 24x14.

FCONT
indicates how the computation starts.
After
If
each
run
the velocity potential is stored on tape 8.
FCONT =
the
program
If
begins
by
1
,
initializing
the
velocity
potential.
FOONT =
we have a continuation run and
the data stored on tape 8 from a previous run is used as
input and is read in on tape 7.
Title Card
3:
FMACH, OM, AL, PA, ZONE, DC.

FMACH
CM
is the inlet Mach number.
is
the
speed
of
rotation.
This
means
that
the
compressor is turning at the speed of OM revolutions per

60
units of time.
The unit of time depends on the flow

the
conditions.
blade
The unit of length is

by
length
of
the
divided
CHORD
and
the unit of speed is the
inlet speed.
AL PA
is the inlet angle.
determines whether an
axial
flow
configuration
or
a
1
cascade is to be studied.
PA =
for a cascade.
PA =
for an axial flow configuration.

ZONE
is the distance between the axis and the
hub.
In
the
cascade case it is set equal to
1.
-53DC is
the gap-to-chord ratio for the cascade case.
is
For an
blades
axial flow configuration it
the
number
of
around the hub.
Title Card
4:
ZSYM, NC, SWEEP1, SWEEP2, SWEEP, DIHED1, DIHED2, DIHED

ZSYM
indicates
whether
we
have
compressor problem or a
propeller problem.
ZSYM
=
1
ZSYM =
for the compressor case and

Thus,
if ZSYM = 1,
for
the propeller case.
there is no cowling, the outlet speed is the same as the

inlet
speed
and
the
radial direction is stretched to
infinity.
NC
is the number
of
span
stations
at
which
the
blade
section
is
defined.
The blade section is interpolated

stations.
the
If
linearly between two span
NC
<
the
geometry
is
given
by
last
case studied and the
program will run for this ZSYM and
for
the
parameters
given by data cards

any more cards.
through
without the need to read
SWEEP1 .SWEEP 2, SWEEP are the angles of sweep of the singular line at the
hub,
at
the
cowling (or the tip) and at the far field

For a compressor blade SWEEP is not used.
respectively.
,
DIHED1,DIHED2 DIHED are the dihedral angles of the

the
singular
line,
at
hub,
at
the
cowling
(or the tip) and at the far
field respectively.
used
In the compressor case DIHED is not
-54-
Title Card
Z,
Z
5:
XLE,
YLE, CHORD, THICK, ALPHA, FSEC
is the location of the span section,
XLE.YLE
CHORD
THICK
are the coordinates of the leading edge,

is the chord of this section, used to scale the profile,
multiplies
each y coordinate; thus the thickness of the
section is multiplied by THICK.

ALPHA
is the angle of
twist
through
which
the
section
is
rotated.
at each
Changing the value of ALPHA by the same amount
span station introduces a stagger angle,

in
FSEC
indicates whether section coordinates are to be read

from data cards.
be
FSEC =
means no further cards are to FSEC

=
1
read
for
of
this
the
section.
means
the
coordinates
section
will be determined by the
next input cards.
Title Card
6:
YSYM, NU, NL
YSYM
indicates whether the profile is symmetric or not.

=
1
YSYM
for a symmetric profile and YSYM =

1
otherwise.
surface
For
are
YSYM =
read.
only the coordinates of the upper
NU,NL
are the number of points on the upper and lower surfaces

respectively.
Title Card
7:
TRAIL, SLOPT, XSING, YSING
-55-
TRAIL
is the included trailing edge angle in degrees,
or
if
the
angle
between
the
upper
and
lower
surfaces
the
profile is open.
SLOPT
is the slope of the mean camber
line
at
the
trailing
edge.
SLOPT
is used not only for the profile but also
to determine the vortex sheet
behind
the
profile.
If
SLOPT is too big and the gap-to-chord ratio is small the
program may abort.

XSING.YSINT,
are the coordinates of the profile

about
singular
profile
a
as
If
point
is
inside
the
which
the
unwrapped.
that
the
This
point must be chosen in
such
way
mapped
a
coordinates
are
as
smooth
possible
such a
without
pronounced bump near the nose.

we
bump
occurs,
first
move
the bump nearer the nose by positioning
the singular point closer to the upper or lower surface.
Then
the bump is smoothed by placing the singular point
at the right distance
from
the
nose.
Moreover,
the
singular
point
must
be redefined not only for any new
profile but also for any new
configuration,
since
the
geometry changes with the parameter DC.
Title Card
X,
8:
X,Y
are
the coordinates of the upper surface.
There are NU
cards and the coordinates are read from the leading edge
to
the trailing edge.
-56-
Title Card
X,
9:
X,Y
are
the coordinates
of the
lower surface.
are
There are NL
cards and
the
coordinates
read
again
from
the
leading
edge to the trailing edge.
The first data card
repeats the first data card for the upper surface
which
defines the coordinates of the leading edge.
These
cards are followed by (NC-1) series of cards beginning with

5.
Title Card
The program provides
both
graphical
and
printed
output.
The
Calcomp plots were shown in the previous section.

In
the
printed
output
we
can
read
the
listing
surface
of
all the
coordinates given in the input.

are printed
Then the
the tip.
mapped
coordinates
for
the
hub
and
This allows us to ascertain

After
the mean
whether the coordinates of the singular points are well chosen.

the
iteration
history,
with
the
maximum
correction
and
correction to the velocity potential and with the maximum residual

the
and
mean
residual for the difference equations at each iteration, the

on
program gives us a printed plot of the coefficient of pressure

surface.
the
This is followed by a Mach number chart.
These last two are blade
given for each span station.

are printed.
If
Finally the characteristics of the
the
mesh has to be refined, a new series of output is printed
-57-
for the new mesh.
If
not,
the
program
terminates
or
restarts
by
reading a new data deck.
-58VII.
1.
BIBLIOGRAPHY
Supercritical
Wing
Sections,
66,
F.
Bauer, P. Garabediart, D. Korn,
Lecture
Notes
,
in
Economics and Mathematical Systems, Vol.
Springer-Verlag
2.
New York, 1972.

Wing
F.
Bauer, P. Garabedian, A. Jameson, D. Korn, Supercritical
Sections II, Lecture Notes in Economics and Mathematical Systems,

Vol.
3.
108, Springer-Verlag, New York,
1975.
F.
Bauer, P. Garabedian, D.
Korn,
Supercritical
Wing
Sections
III,
150,
4.
F.
Lecture
Notes
in Economics and
Mathematical Systems, Vol.
Springer-Verlag, New York, 1977.
Bauer, P. Garabedian, G. McFadden, The NYU Inverse Swept
Wing
Code, to be published.
5.
J.
Brochet,
J.
P.
Veuillot,
J.
P.
Boisseau, M. Enselme, Calcul de

de
l'lcoulement
1
ridimensionnel dans une roue mobile
souffante,
colloque d 'aereodynamique appliquee de 1'aaaF, November 1981.

Courant
,
6.
R.
K.
0.
Friedrichs, Supersonic Flow and Shock Waves,

New York, 1948.
Interscience-Wiley
7.
C.
Farrell,
Qua si-Three-Dimensional
Full
Potential
Transonic
Blade-to-Blade Code, NASA Lewis Research Report, October 1981.

8. C.
Farrell,
P.
Beauchamp,
System
of
Codes
for
Quasi-Three-Dimensional Shockless Blade Row

Research Report, October 1981.
9. P.
Design,
NASA
Lewis
Garabedian, Estimation of the Relaxation Factor for Small Mesh

Tables Aid.
Size, Math.
10.
P.
Comp., Vol.
10,
pp.
183-185, 1956.
Garabedian, Partial Differential Equations, Wiley,
New
York,
1964.
-5911.
P.
Garabedian,
G.
McFadden, On the Design of Supercritical

20,
Swept
Wings, AIAA Journal, Vol.

12.
pp.
289-291, 1982.
H.
Howard,
A.
Hansen, Three-Dimensional Compressor Flow Theory

Effects,
,
and Real flow
in
pp.
Aerodynamic
365-384.
Design
of
Axial
Flow
Compressors (NASA-SP-36)
13.
D.
Ives,
J.
Liutermoza, Analysis of Transonic Cascade Flow Using

Journal,
Vol.
Conformal Mapping and Relaxation Techniques, AIAA

15,
pp.
647-652, 1977.
14.
A.
Jameson, Iterative Solutions of Transonic Flows over Airfoils

Including
and Wings,
2
Flows
at
Mach
1,
CPAM,
Vol.,
27,
pp.
83-3 09, 1974.
15.
A.
Jameson, D. Caughey, Numerical Calculation of the Flow Past a Wing,

ERDA
Swept
Research
and
Development
Report
No.
COO-3077-140, June 1977.

16.
W.
H.
Jou,
Finite Volume Calculation of Three-Dimensional Flow
around a Propeller, Oral Communication.

17.
E.
Mclntyre,
Design
of
Transonic
Cascades
by
Conformal
Transformation
of the Complex
Characteristics, ERDA Research and

,
Development Report, No.

18.
C.
COO-3077-136
November 1976.
Flow
S.
Morawetz, On the Nonexistence of Continuous Transonic

Profiles,
CPAM,
Vol.
9,
pp.
Past
10,
45-68,
1956 and CPAM, Vol.
pp.
107-131, 1957.
,
19. Ph.
d'
Morice
H.
Viviand, Equations de
conservation
et
condition
irreversibility
pour les ecoulements transsoniques potentiels,

t
Compte Rendus de l'Academie des Sciences, Paris,

pp.
291, serie
B,
235-238, 1980.
-6020. E.
Murman,
J.
Cole, Calculation of Plane Steady Transonic Flows,

9,
AIAA Journal, Vol.

21.
pp.
114-121,
1971. and
M.
H.
Vavra,
'
Aero-Thermodynamics
Flow
in
Turbomachines
Wiley, New York, 1960.

22.
J P.
Veuillot,
Calcul
de
l'ecoulement
moyen dans une roue de

155,
turbomachine axiale, Publication ONERA No.

2 3.
1973.
J.
P.
Veuillot,
Methodes
de
calcul
de
l'ecoulement
tridimensionnel dans une roue de turbomachine, Revue francaise de
Mecanique no.
24.
J.
P.
5-56,
pp.
19-29, 1975.
Veuillot, Calculation of the
Quasi-Three-Dimensional
Flow
in a Turbomachine Blade Row, Engineering for Power, Vol.

1,
99, No.
pp.
53-62, 1977.
-61-
VIII.
LISTING QF
THfc
CODE
PRO G* AM CSCDF?2(lNPUT=512.jUTPUT=512,TAPE5=INPUT ,T APE6 = UUTPL;T, 17 = 5 1 2 T A PE8=512) MA I N R ju TINE WHICH CONTROLS THE COM PUT A T I JN a L PR OCbDURE
r
T A P
tr
KEDUCED VELOCITY POTENTIAL COP M D N b (12 9,2fc,17),50(129,17),E0(17),IV(129,l7) ,ITE1(17) ,ITE2(17) 1#A0 (129) ,AK129),A2(129),A3(129),B0(26)*31(26)jB 2(26),83(2t>),Z<17) ?,C1 (17), C2(17),C3(17)CM17),C5(17),XC(17),XZ(17 ),XZZ(17),YC (17), 3Z(1 7).YZ Z(i7),KSYM,NX,NYNZ,KTElKTE2,I:>irM,SCAL, SCALZ, XK, OMEGA, AL^ <HA, CA,SA ,F 1ACH T, PA, A8LADE CQM MQN / CAl/ P1P2,P3,BETA,FR,kM,DG,GM,NS,U1V1, W1,IPJR,KR,IG,JG>
G
I S
He
1KG DIM ENSIO N X S lkAI L<6)> N P

(
T LE(5 ), SLOP! (6) XP(2 0), YP(2JC 2), DJ (20 0), D2(2C0) 03(200), X(129), Y(129), SV (129 ), SM(129), C 3(12 9), C H0RG(l7), SCL(17), SCD(17), SCMQ7), FIT (3), CGV0(3), P1G( 3), P30(3), 3ETAC(3), FHALF(3) RESI5G1) , CGUNT(501), DESC *3) P 2 TI TLE( iG 5(10 NQ = 2 Go NE = 129
(
2
)
YS
*
,
)
fc
ZS
(
6)
,
XLt(o),
(
El
E 2
E 3
E 5
o ),
IND = 1 Ul =
(
Vl
wl
=
=
J.
0.
IRE A0 = 5 IwR IT-6 KPL 0T = J IPL QT = 1 JIT = REw IND

(
RAD = 57.2 9578 I wk IT, 430) 10 WRI TE wRI Tt (I vRIT, 210) REA D I* EAD,420) TITLE WRI TE (I wRIT,h60) TITLE IR EAD,42G) DESC REA D IK EA 0,410) FNX,FNY,FNZ,FPLOT,<SCAL,PS AL,F ,TY REA D FNX. LT.l. GG TO 200 IE WRI TE (I WRIT, 520) FNX,FNY,FNZ,FPLOT,XSCAL,PSCAL, F,TY NX = FNX NY = FNY NZ = FNZ KPL OT=A8 S(FPLOT) IR EAD,420) DESO REA C
( ( ( (
)
WRI TE NM =
(I
WRIT,530)
20
NM+1
62-
NM NM , P 1 NM , 2 NM , P 3 NM 6 F T A RFA D IR EAD, A10) F IT(NM),COV ),F HALF NM) , FCONT WPI TE (I ^ P I T ,<t50) FIT(NM),CU VO(NM),l10(NM)P2G(NM),P30(NN),tJETAO(N 1M), FHALF (NM) IF (FHAL FdMM .NE.O. AND.NM.LT .3) GO TO 20 FHA LF(3) = 0. REA D IK EAO, <t20) DESC PEA b (1H EAD, <tlC) FMACH,GM,AL ,PAZONEDC WRI TE (I WRIT ,5AC) FMACH,OM,A L, P A, DC DC = (l.-P A)*D C*RAD/360.+PA/DC
(
XK = l./DC ZDN E*l.+ PA* (ZONE-1. OME GA = PA *GM/RAD*6./Z0Nc ALP HA= AL /RAD CAL L GEO M (NDNCNP,ZSXS# YS XLt,YLESLOPT>TRAIL#XP,YP#SWEtPlSWEE IP?., SWEEP ,DIHEDl>DIHtQ2>DIHED ,KSYM,XTEO,CHORDO,ZTIPISYMC,XK,PA,INC
)
2)
ISY M=ISYMO
IF
(
ALPHA. ME .0.
)
)
ISYM=0
CA = COS (ALPHA
SA
IF
SIN(ALPHA (FCCNT.LT.l) GO TO AO REA D (7) NX,NY,NZ,NM,K1,K2,J IT#U1V1W1

MX NX-H MY* NY + 2
MZ = NZ + 3 DO 30 K=1,MZ PEA D (7) (G(I*JK) I=1MX
(
),
J=1,MY)
CON T 1NUE PEA D (7) (E0(<) ,K*K1,K2) CO.M TINUb <c IP,XMAX,ZMAX,KSYM,PA,SY,SCAL,SCALZ,AX, CAL ,32,B3,ZC1,C2,C3,C4,C5) 1AY CH0RD0PA,SwEEPl,SWtEP2,SWtEP,DlHtDl,D CAL XZZ#YCYZYZZ#Z,E1E2,E3.EA*E5,INDZ0 1IHE 2NE) CAL L SU KF (ND,NE,NC,NX,NZ> IS YM,KTEl>KTE2>SCALA0>Z>ZS,XC>YCSL0PTj 1TRA IL, X sYi#NP ITE1,ITE2,IV> S0>XP,YPD1,D2D3XY,INDXK*PA,>Z*YZ> 2A1, C1K SYM IF IND .EO.O) GO TO 190 IF (FCO NT.Gh.l .) GO TO 50 NM = 1 CAL L ES TIM 50 CON TINU E NIT = J I T T = 2 ./SC AL IWRIT,A30) 60 WRI TE FCO NT*C MIT = fit (NM)+JIT cov = CQV O(NM) BET A = BE TAO(NM) MX* NX + 1
30
)
-63-
y = .\y + 2
7 =
Ul
+ 3
j
Y =
\Y
1=3 ?=NZ +^ 'lit iw*

(
20
7C
1
(
2*
4
4
2
7C
kITE Ik? ONT NUc

I
60)
(IV(I>K),K=K1,K2)
i ie
R
1
(IWF
(
ITE 6u
*P
30) 30)
A0(
I
(
RITE I Wi= PO C ONTI NOt RITE (IwR

P
440)
240)
44 0)
I)50(lKl),Su(IUTE2)
iTt
i^P
I
XMAX> AX
Rl TF
WR ^
RITE
D
W
UP
43C 250)
4 4 0)
90
SI IE
1
J =
(IWR
N
9C
C 'INT
UE
PITfc
(IwR
2fcC)
RITE
^ I T E
(UP
(
RITE
10
IwR (IwR
440) 430) 27C

2
SY AY
100
KITE RITE UR ONTI NUE PITE IWP PITE UK

(
<=K ll.k
440) 440) Z(K)XC(K)YC(K)*XZ(K)Y2(K)fXZZ('<)#yZZlK) 280) 440) 430)

2 3
ZMAXt AZ
PI IF
IwR
RITE UP RITE (IWR RITE UP ALL SEC ON PITE UP RITE IWR PITE (UP
(
(
90)
0)
47C) TIME 510) 310)

4
NXjNY,NZ
)
TIKE
4C)
)
FMACHOMAL
RI TE
X =
L =
WK
1
320
NX /2
0. il
K = K1 K2 1= I T El K II) = 1. + .5*SCAL*(A0(Il)*A0(Ii)-S0(IlK)*S0(UfK)) 1 = SC AL* AOdl )*S0( I1K II = xC <K)+AL0G(X(Il)**2*Yl**?)/XK/2. (LX) = 1. + .5*SCAL+(AO(LX)*AO(LX)-S0(L<. <)*iiO(LX,K))
(
)
1-SC AL* AO(LX)*SC(LX,K

(LX
)
xC
(K)+ALQG(X(LX)**2+Yl**2)/XK/2.
= X
(
110
HO PL (<) ONTI NUt
l)-X(LX)
-54-
ZDU*=KTb2 -I
0.
)
s =
DG C
s
120 K=K TE1 KZDUM zo=.5* (z( K+l )-Z( K) = S + L ZC* (C H0R0(K + 1 )+CHORD(K) WC5(K
CNT iNUfc
120 130
BLADE=S
N
P
P P
IT*Ml +
2= P2
(
1=P10(NM)
NM
3=P30(NM) c ALL BOUND L 2=U1+CA V 2=V1+SA (IWP IT>490) NIT,0GIG,JG>KGFR,IR,JR><RGM,R*,F1,U2,V2,NS w RITE c DUNT(NIT) =NIT-1-JIT
R
1
S (NIT) =F R
)
F (N1T.LT .MIT .AND.ABS(DG) .GT.COV. AND. ABS(PG) .LT.1C. GO TO 130 ATE=0. F (NIT.GT .JIT+1) RATE=(A8S(RES(NIT)/RES(JIT+1)))**(1./(CJUNT(NIT) 1- COUNT (J IT + 1) W RITE (ImR IT, 330) (IWP IT, 5 00) RES(J1T+1),RES(NIT),C0UNT(NIT),RATE 4 PITE C ALL SECUN (TIME) w KITE (IWK IT, 310) TIME L X=NX/2+l 150 K=K 1,MZ C C GO TO 150 (K.LT.K TF1.0R.K.GT.KTE2 I F
R
I
)
lITfcl(K)
2=1TE2(K) VELD (k,K,SV,SM,CPX,Y) c HORD(K )=X (Il)-X(LX) FORCF (I1,I2,X,Y,CP,AL,CHQRD(K),XC(K),SCL(K),SCDK),SCM(),A2 c ALL 1, A3,C5(K), PA) (KPLOT. GT.1.AND.K.GT.KTE1) GO TO 140 F w RITE (IWP IT, <30) (Iwk IT,34C) W RITE (IWR IT, 440) FMACH,OM,AL * RITE IWP IT, 350) w RITE IWR IT, 4 40) Z(K),SCL(K),SCD(K),SCM(K).CHORD(K) 140 w RITE , FM ACH (KPLOT. LE.l.) CALL CPLOT 1 1, I 2, SM , AO, S F 1 ,K SPEED (K) c ALL 150 c 0NT1NUE (KTE1,KTE2,CH0RD,SCL,SCD,SCM,Z,XCC5,CL,CD,CMP,CFP,CMY c ALL TOTFO R 1, PA,A8LA0E V LD0. F (A8S(CD J.GT.l.E-6) VLD=CL/CD
J
AtL
RITE (IWP IT, 430) PITE (UR IT, 360) (KSYM.E 0.1) U2CA I F I F (KSYM.E Q.l) V2SA V =ATAN2( V2 ,U2)
i>
-65-
=
1
Aj-AL u;* U2 + V2
= 1
/ F
V2
Wl* Ml
+
M AC
C. 2
)
bGP r/(Q/( Jl-0 .2* C) PI Tt FMACH#QM AL#U# I w p IT ,4 40) 70) RITE I UP IT, RITE (IrfG I T # 40) CL,CD, VLO
=
( (
<t
sue
i
J
4 AC) CMPCMR,CMY*ABLADE RITE NM ) 1 .-F H A L F L PLOT = F & TO 160 (L PLOT. LT 1 36 O./XH /PA J *ZC NF/CHukDO D THREE D I PLO T,$V,S*,CP,X,Y,TITLf,DC,AL,ZONE,U,V,CHORGu>X:>CAL> C ALL
<
IwR IT, iwR IT, F*3L0 T*(

(
60)
(
)
IP SC AL
160
C
I
ONTI NUE (F HALF F

X= NX
+
>|M).tO.C.)
GO
TO
170
NX
N
C
Y=NY + NY
N2 + NZ ALL CCORD
t
N Z =
(NX,NY,NZ,XTEu,ZTIPXMAX,ZMAX,K$YM,PA,SY,SCAL,SCALZ,AX, ,AJ,A1A2A3,8 0B13263>Z>C1>C2C3,C4,C5) (NC,NZ,KTEl,KTE2,CH0RD0,PA,SwEEPi,SWEEP2,Skr.tP,DIht Jl,r> c ALL S INGL II HEC2 DIHE 0,ZS,XLE,YLE,XC,XZ,XZZ,YC,YZ,YZZ,Z,E1,F2,E3,64,E5,IND,ZG
1
Yi AZ
TY iF
2N E)
IT 2A
ALL SU^P (ND, NE,NC,NX,NZ,ISYM,TE1KTE2SCAL,A0,Z>ZS,XC>YC>SL0?T> RAIL , xS, Y St N? fITElITE2#IV*S0*XP,YP#Dl02*D3#X,YlNDXK,PA,XZYZ 1,C1 ,KS Y v GC TG 190 I F 1 N D . t C . C) C ALL REF1N N M = NM + 1
C
i
N IT = C G
TC tO
170
ONTI NUE
SI TE
w
w
(8)
\|X,NYNZ>N1,K1,K2,NIT.U1,V1>W1
((G(I,J,K),I = l,fMX),J=l,MY)
(
RITE
18
D
k
(6/3 90) K=l

)
RITE (6 1*0 C ONTI NUE w RITE (8) E NO F 1LE 8 R EWIN D

&
EG(
)>KK1,K2
TO
10
(
190
RITE 3 I * R I T , u RITE I WRIT, 400) 0102 S TOP 200 C ONTI NUE (0.,0.,999) c ALL PLOT 0101 s TOP
*
't ) (
210 FORMAT 220 FORMAT
(20H0NYU COMPRESSOR CODE) 49H0 IN D IC A T 1 ON OF LOCATION OF BLADE AND VORTEX iHEET,27-i ,K=Ki , K2 , I = , NX I V I t K / 27H0 IN COORDINATE PLANE Y =
( .
(
<
-6623C FORMAT (49H0CHGRDwISE CELL DISTRIBUTION IN TRANSFQRi > 1NU MAPPtc SURFACE COORDINATES AT HU3 ANO TJ.P/J5HO TIP PROFILE 2\> HUB PRGFILE,15H POWER LAW TE LOCATION' >i5H 24C FGn,MAT (15HG (^mON jfmal cell distribution IN TRANSFORMED PLANE/15 H-J 2 5 C FORMAT
) )
FORMAT 270 FORMAT

2
POWER LAw SCALE FACTO-U15H 4dHOS j ANwI Sb CELL DISTRIBUTION AND SINGULAR LINE/13HC XZ X Y SING , 15H SING , 15H , i&H 1 Z 13H YZZ XZZ ,15H yz 2,15H POWER LAW TIP L0CATI0N15H 280 FOP-MAT (15H0 x9H0ITERATIVE SOLUTION) 290 F P A T NY ,15H NZ NX , 15H 300 FOKMAT (13 HO ,15H AN& OF A T T A C <) OMEGA ,15H MACH NO 310 FORMAT (lpHO K l3H 4H , 4H J I CORRECTION , 4H 1CH0IT ERAT I0N,15H 3 2 C FORMAT K ,10H COR.tlOH i t A N R E S . t 1 >4H F A 4H >1H J I PESiCUAL 1 ,10H SONIC PTS) XSPEED ,1GH YSPEEG 2H REL FC1 110H wORK MAX RESIOAL 2,15H MAX R E S I D A L 1 * 1 5 H 1 5 H 330 F k M A T
60
(1SHO
I-
^1
Ni
REDUCTN/CYCLE 340 FORMAT (24HCSECTI0N

15H
:
15H GM MACH NO C HAR AC Tt R I ST IC S / 1 5H0 ,15H ANG OF ATTACK) (/13H SPAN STA1 I0N,12X2HCL,13X2HCD,13X2hCM,10X5HCHGRD) QMEG MACH NO 1 ,15H (22H0dLAUE CH AR AC T ER I ST I C S/ 1 5HG UTV. ANGLE MACH NO 2 ,15H ANG OF ATTACK, 15H 15H ,15H L / j CO CL , 15H (lb HO (/ZX6HCM PITCH, 6X7HC.1 RGLL,9X6HCM YAw, 9X6HAEL ADE (IX, 1 4 H WRITE ON TAPE6) (24H0BAD OATA, SPLINE FAILURE) (8F10.6) 10A8)
)
) )
1H1
(F12.5, 7F15.5)
(1X*8E15.5)
(1HC1JA9)
(18.7115) 1> ,3ZIA)
(
111D,E15.5,3I4,E15.5,3I4,2E10.3,3F10.5,I1C)
(
cE 1 5
4 , 2 F 1 5
15H0C JMPUTING TIME,F10.3, 10H
SECONDS)
</5x3hFNX,llX3hFNY,ilX3HFNZ,10X5HFPLCT,9X5HXSCAL,9x5HP$CAL, 19X5hGPlUX,9x5HGRIDY/lX,dE14.3) FORMAT (/4X7HFIT<NM),ex8HCGvO(NM),8X7HP10(NM),3X7HP2C(NM),dx7HP?G( 5 3C 1NM) #6X9HBETA0(NM) , 6X9HFHALF NM 540 FORMAT (/5X5HFMACH,11X2H0M,14X2HAL, 12X3H PA,12X4H DC /1X>5E15.5)
(
) )
END
-67-
SUB 11.5 2PA,
ur ine
EGM
(ND,NC,NR.ZS,XS,YSXLEYLt:>LUPT,TUAlL,*P, YP,SwEEP
Xl< ,
EP2,
D)
S *
EP,DIHED1,DHED2,DIHED,kSYM,XTE0,CH0RDC,ZTIP,I SYMG,
X L t
GEG DIM
16)
2 C
,
(
TRIG JEFIMTIQN CF 6LAuE SIDN X S N D > 1 * Y $ N 1 , Z S 1 RAIL(l), XP(1), YP(1). NP(1),

( ) ( )
(
>
X(6),
1 , YLE(l), Y(b), Z(6)

)
SlUPT XC(c)
i ),
7T( YC(fc),
DIM
IRE 1WR
SIGN DESC (10),

= 5
=
A(o),
B(6), D(6),
E(6)
RAD REA REA *RI KSY

IF
.29b
I
7fl
)
R E A J , 2 3 C
DESC
IREAD,22C)
(I wR
IT, 270)
)
ZSYM,FNC,SwEEPl,SWEEP2,SWEEP,DIHEDl,DlHED 2,0IHEL ZS YM,FNC,SWEEP1, SWEEP 2, SWEEP, DIHFD1,DI HE D2 , DIHED
SYM
C.LT.3.
=
RETURN
k
NC = SWE SWE SWE
b*t P 1/
AD
=bwEEP2/PAP
iWfcEP/RAC
DIH DIH DIH

51 = 52 = Tl = T2 = ISY
sQIHEDl/ RAD =D1HEU2/ PAD L1HED/RAD

(
S WE P 1
(S*EEP2) (DIHED1) (D1HED2)

1
XTE CHQ ZCN

K=l
1C
PEA D PEA D
IR
EA0,230)
)
DESC
IR AD, 22
(
ZT(K),XL,YL,CHORD,THICKAL.FSEC
ZT(K), XL, YL, CHORD, THICK, AL,FSEC
WPI TE 1 hRlT,2 80) ALP HA= AL /RAD

2T( ZS(
IF IF
K
K
)
=
=
ZT
(K)/ZQNE
AL
UG(ZT(K)+(1.-PA))*PA+(1.-PA)*ZT(<)
GO TG
TO 20
)
(FSEC EQ.O.)
(K
.GT .1)
=
=
(
GG
ZON E(l. -PA)+PA*ZT (1)

ZS( 1 ZT( 1)
)
,-PA)*ZT(l
1.
CDN T1NUE PEA D IR hA0,230) DESC PEA D IR EAD,22C) YSYM,FNU,FNL WRI TE YSYM,FNU,FNL,ZONE 1 wPIT,290) NU* FNU NL = FNL IF YSYM .EQ. 1. NL = NU
( (
(
N= N U+
REA D
NL- 1 IR tA0,230)
(
DESC
-6 b-
Rt A
Rt A
22 C 3G
wRI TF kEA u
IiiR
I
(1RIT,3C0)
(
TRL,SLT,XSING,YSING TRL,3LT,XSING,YSING
DESl
Tt
3C
(
(IwRIT, 310)
i=NL N
DG
Rb'A D
UEA
(
2
,
.2
) )
I
(
.
)
Y P
.
I
(
UPI TF
3C
I wR I
2 6
X P
Y P
CON TIN Ufc L=N L + l

IF
(YS Y.^.GT .0.

(
GO
)
TO
50
PFA U
W&I Tc DC 4C
I
(
REA
23
DESC
IV/PI T, 3
20)
V A L
REA
J = L
WRI TF
-i
= =
IlNL 1READ220) (Ikfll,2fc0)

VAL
UM
VAL.DUM
XP( J) YP( J)
DUh
70
I
40
50
CON TH UE
GO
J = L
TO
DO
J = J
60 -1
J
J
)
iM
XP
YP< WI
-XP II )=-YP(
70
CON TINUt:
T E
IUF
iltit
WftI TE
WRI Tt
PC
IT, 200) IT. 250) (IWR IT, 240)
ZS(K
TPL,SLT, XiING YSING
CON TINGE /ZONE SCA L[ =CHb 0/( X P(l )-xP( L C(K )=EXP(- PA*Z S(K) XLE (K )=XL/ ZONE +(XSING-XP(NL))*C(K)*THICK*SCALt YLE (K)=YL/ ZONE +(YSING-YP(NL))*C(K)*THICK*SCALE XX = XP (NL + XSI NG-XP(NL )*C (K)*THICK yy = YP(NL + YSI NG-YP (NL )*C <K)*THICK CA = CGSULP HA)
>t
>
SA
XC
=
(
SIN (ALP HA
K)=SCAL *(X X*C A*YY*rHICK*SA YC( K)=SCAL E*( Y Y*THICK*CA-XX*SA)

)
DO 90 1 = 1, N XS( I*K = SC ALE* (XP(I)*CA>THICK*YP(1)*SA) YS( I,K)=5C ALE* (THICK*YP(I)*CA-XP(I)*SA)

90
CON TINUt SLO PT(K =T AN( ATAN( TH ICK+SLT) -ALPHA )*C(K TRA IL U )=A TAN(THICK*TAN(TRL/RAD)*C(K))
)
NP( K) =N
CHO RD0=AMA XI (CHORDO, CHORD) IF YSYrt.L E.O. . OR. ALPHA. NE.O.) ISYMO=0 wRI T T (UP. IT, 210) ZS(K) WBI Tt (Iv-F IT, 250) XL, YL, CHORD, THICK. AL K=K + 1
(
-69-
IF
II
L
( {
GO
TO
GC
10
XK = XK
.- pa)
. 1 . 0) 1, N !> NC
)
/CHCRDO+Â)
70 150
DO CO
x
(PA. 120
1
)
Ne
1
< =
(
XS I,K Y(K = AT AN2 rS( IK) ZT(K) **2 + 2T (K Z (K )=A L C6( Y S ( I* K

(K
=
)
(
**2
II.
100 CON
N
(
El = 1. / E2 = 1./
Fl F? El E?
=
=
CuS( Y(1))+T1*SIN(Y(1))) CLS Y(NC)) + T2*SIiM(Y(NC)))

(
=
=
(COS (CDS S1*E

52 *t
(Yd >)*T1-SIN(Y(1)))+E1
(
Y(
M C
)*T2-SIN( Y(NC
)
)*E2
2*ZT (NC
CAL L bP LiF (lNCZ>X,A,3,Ci#Ell>E2,CG.tlND) CAL L IN r pl (1NC>ZS>D1NCZ>X,A,?,C0) CAL L SP LIF d,NCZY.A.6Ci*Fl,ltF2.C0.*IND> CAL L IN TPL (l>NCjZS#EtlNC#Z#Y,A>BC0) DO 110 K1j NC XS( 1K) = D( K YS( I*K = E(K 11C CON TINO E 120 CON TINU E DO 13 j K=l, NC
)
)
)
XC (K)
(
) ( ) )
Y(K )=AT AN2( Y C K , Z T K Z(K = AL 0G( Y C (< )**2 + ZT(K) **2) /2.
)
13C CON TINO E F A = cost Yd EB = SIN( Yd EC = COS y(n: ED = SIN( Y(NC El = Sl/( tA + T 1*5 B) E2 = S2*Z r (nc )/( EC+T2*EC) Fl = (EA* Tl-E b)/ E A + Tl + E F2 = (EC* T2-E 0)/ (EC+T2+E0) CAL L S P LIF 1, NC,Z,X,A,3>D,1,E1,1>E2,0,G.,IND) CAL L IN TPL 1, NCZS#XClNCZ>XA8tD0) CAL L SP LIF 1, NCZ>Y,A,3,D,lFll>F2,t>0.,lNL)) CAL L IN TPL 1, NCZj#YC*lNCZ#YA,BD*0) DC 140 K = l NC X(K )=XL E IK) Y(K = AT AN2( YLE (K)ZT(K)) Z(K =AL 0G( Y LE( K)**2+ZT(K)**2)/2. 14C CON TINO b EA = COS( Yd E6 = SIN( Yd) EC = CCS( Y(NC ED = SIN( Y(NC El = Sl/( EA+E B*T1) E? = S2*Z T(NC )/{ EC+ED*T2)
) )
ti
-70-
F 1 =
f.
A*
F2=
(F
CALL CALL INT PL CALL S^L IF CALL INT ^ L

GO
Z
i-cr )/ (EA+E3+T1) :*t <:--_>)/ EC +E n* T ^ SPL If (lfNC2#X*A#flC*l#61l*t'2#00.lN

[ (
)
(1NC>ZSXIE1NCZX,A,8D,0) (l^CZYA,-i,Ji.FliFZ0U.l (lNCZS>YLEl>NCZrAiL).u)
N
'
TO
= Z S
(
1 I
15C CnNTI NUE

I
)
co
u
1
K = i
(
NC
)
ZS(K) = ZS 160 CGNT I r.uE 17 C CONT NUt

CO 19 C DO It XS( I* K
Y
c
-
-Z
N =
1
)
NC
(
(
1N
Xi
I K )-XC YS(I*< l-YC
=
=
K)
)
I ,
I
IPC CUNT
XX5
X X X X
r
>
NUE
xp(
K.T(
XK*Xi(l>K))*CDS(X"<*YS(l#K))-l
bX 3 (2.*XK*XS(if<))-i:.*XXS-l.) XXS) /?.
= 5
=
I
XX
T E C = A'V A
xl(XTEtXXS)
)
19C C0NT1 NUt ZTIP = Z3 NC RE Tlk

(
200 FORMAT lLPt ,
1 5
210 FORMAT
1H
ANGLE >F10.5-/15HO (16H0PR0FILE AT Z SING H ilNG 1 5 (27HCSECTIQN DfFINlTION AT Z t F 1 o . 5 1 1 b H C .13HTH1C* NtSS R ChORD *lbH YLE
U
XL
r
.
>L
1
ri
If
. i
AL
220 230 240 2^0 260 270
2PHA FORMAT
(8F10.6
(
1GA6 FORMAT 1H1) FORMAT FORMAT F12 .4. 7F15.4 FORMAT (8f 15.5) FORMAT (/5X<tHZSYMjl2X3HFNC>10X6HSWtEPl>*X6HS
) (
( )
HHEDl9XeHDlHED2#lCX5HJlHED/lX8ElD.t)
260 FORMAT (/5X5HZS(K * 12X2HXL#13X2HYL 11X5HCH0R 12X<.HFStC/lX,7E15.5 (/6X<tHYSYMllX3HFNU12X3HFNL/lX,<tti:>. 290
)
(/t>X3HTRL12X3HSLT#HX5HXSIN3>10X5HYS
(/5X5HXP( I )1QX5HY?(I (/6X3HVAL>12X3HDUM)
)
-71(NX NYpNZ*XTtO*Zl ^* X I AX , Z v A X * * S Y * , y A, 3 Y iC JJ R *lY*F,ACAl,Ak:.AjBC*ei/B2jft3^Cl*C?C3fL^*Ct)) 3IKFTOEL AM) jPAN*13E CDLSDIMTES

L
t-
iMj T
X ,
AL
* S
C A
117,
bt
r
A
S
A Y , A
/.
;>H
DIM 1Z(1
=
h
)
NS
JM
4 C
,
(
Aid).
1
)
A >
i
)
) ,
(
A 3
B
1
)
r 1
B2
ti
CHI)..
C 2
C 3
l *
!> (
Q* 2. /NX DY = 1 ./Mr
K Y =
\Y+
N
2
Z0 = K?
C7
K
=
1 =
+ I
i*
1 . /
AEi =
ax
A/
=
i
.
AY=^
=
=
Hi IF
*,W A
X=
. .
(*3
= 5 J
x = .
(
J
j
.
.
1
)
XMA<*0.bZb
-i
373
Z
(7
E
/
M,A
IF
Kb
2
Q
(
.
.
(
M A X=
NZ
SCA
3
L = X
=
5 J
A L
.
IP/
L
1
)
1.
GC1*XNAX*XMAX) C0C001*7MAX)
T=2
SY = S ax = Art i * BX= V? = ID*

(
IS UK
F
/TY (Oi>?5 0.*( T-0.U7)* Aob (T-0.C7)

/ 3
1 1
)/?. +BX / (F + l.
Y)
**2
I
hl w?
bftX
./
AL
Ul
=-8
= =
X/DZ )* *2
3 )/((l.+S73)*XMAX**3) 37j)/12.)*X,MAX**3 * X*AX*XMAX/i2. + S 7 3 +BdX* 3.*C3X-<..*XÂX*Xi>UX)*XMAX*Xi'iAX/bOKT(C6X-XÂX*A
SOîT (?
b X * 3
7.
+
+
S7
A^X
CriX
1.
( 1
CM T((7.
) (
AP3 X=
MX = NX
LX
=
A +
/
',AX)
M
( (
LQ 3C DD = IIF 1.
1
.
1, ^X )* jx
.
DO=-l.+l . /NX
D=
1
.
IF
.NX)
-1
M a
Bl
IF
A= C
(ArjiiuD).Gr.xfAx) ex- DO* DO

SOhiT
(
c-u
io
AS
A)
C = A
Bx*Ai4btJX*(3.*CBX-4,*DD*DD)*DD*D9 DO = ABX*DD+dBX*AS*DD**3
= A3/C D2 = dexMUUM-6.*CttX+lSi.*DD*Di))-12.*0D**<t)*Du/U*C) GO TG 2u
01
1C
IF
A = A
I .
H =
d-
(OC-d*XMAX)/(l.-XMAX))**2
7,
C =
I
/.
<
>
--(/I
:
W
; -
<,-,.)..(+
.
'.,*
/
(
(
.,
<
-I
-'-')
)
;
.
/,
'
= r
=
<-,
..
-:>):(
<',<-:*..>
<
'
>
'
>
; r
(
I
i
=
-
- i
i
: /
i
, = '.- ^
i
'
"
'
..
<
<
A Y
:-iuy4M-].)*u^-i)
i; 1
= a * C /
l
U
J
-3
r <
'J /
H 2
;
i
(
=
=
L*
i * C 1 * v ?
4 Y *
)
i
v
(
j
>
Y*
.
(
(
3
7
* i +
*A
)
->
Z =
- "^ *
.
i C
i-
-!
*
(
. .
i 7 '<)
)
1
)
b 7 3
1
* Z > A X * * 3
;'
L = 1
=
(
i
.
/ *
i
is-i
+w
J
7
>
+ > 7 3
7
i"i
/ i
* Z
(
"!
A X * * j
* 7 hA
2
.
AX/
*,
i,>.,
> /
{.
.- Li
i = A
L +
* C p 7
Ua*
l>
* Z
'
* i
y- 1.
>
11.
4.
-5
'
-1
;
v
(
k = c
)
> 1
>
: ,:=
k-m
/
ic
it-
(k
J/-/, .LÎ")->. *0Z .t-..v Z)

(
6 o
[. i
. i V-
a X
u
.
jo
A=
Cti?-L-L)*l
.7 *
i<
&S=SC^T(a)
C = A
f
As
*
j i:
'(
*'C
-J
?-"*
3
L-
0* L D
* J i>+
P.0= At Z*
C
1
JU +
(
);
7 *
AS*Q-'D**
.
= A S / C
C
J
[)?=VjF 7*
Z*
.
-h
*Cu/+
1
.
-. *
JU*L-D )-!?.
j**^ )+
--
*:
GU
:
(
1(
oC
U
(
IF
A = i
.
I
b =
'.
b-
o*
)/(l.-ZÂX))**2
C = A * *AZ
.-A) D-B * Z A X / C ClA*C/({l. + L-)*Ae3Z) D? = -(AZ + A7)*(OL3-6*ZivIAx)*(3.+D)/{(l.+D)*A*{i.-7ÂX)^*

.
D=( A/+AZ-1 Lu= B * Z A \+

-!
)*(
!i
A3
Z *
IJ
\*.
^0
Z(K)=jCALZ*L)0
C4 (K)
J
Ji*jl*k<'
;,
70
C-5(k)=c<P(-? A*7(K)) C .(^)=CD('<)'*C?(X)

^
Z(K2+1)=Z(K2)+Z(K?)-Z(K2-1)
h c r
i
t-
no
7 3-
vJ
U
)
GL
NO
k rh
,K
TE2>CHGRD0*
^ A
*FF
s,
3 2
.">
A C
fc
'
I.
1HK) 1 MHtU 2 1 NO ZJNb

#
HE&jZS>XLOYLEXOXZ>XZZ*YC>YZ>Y2Z>Z#F-J>E2>
5
J
>
G
D
1
)
E
i
N
M
t Y
E
A
b
/.
T t i
j
N G LL A&
(
LINK
r
(
F
.
LI
TRANSFORATION
(
N
/
JN I)
,
ZS
/ ,
>
X L
1
)
Y L L
t;
C
1
1 ,
X 7
1
i
X
(
ZZ
1
)
<>
'
E 3
b 5
1 C
K = 1
'IC
EM
L
>
[j .
K)
7
I
= j.
1C
MLb Kl = K2 = NZ + ^ KTE 1*
. L
.
T. Z^
.
1)
KTE
K+
X.
L E
ZS
(NO)
KTb2 *
20 CON
a = c
T 1
NUE
H&kDG/Z DNE
TAN(Swfc EP1)
(
TAN Qlh ED 2) (PA .At. 1.0) GG TO 30 F 1 = i / c a s (YLE (1))*T1*SIN(YLE(1)M F2 = 1./ (COS (YLE (NC )*T?*SliM( YLE (NC Tl = (CGi< Yl Ell) )*T1-SIMYLE (i) )+Fi T2 = (CCJ(YL E NC )*T2-SIN(YLE(NC )) )*F2 SI* Sl + Fl S? = S2*F2*E XP (Z MNC 30 CON T1NUE CAL 5 P L I 1* NC*ZSXLEclE2*E3l#Sl*lS2C#G.*lND) KT Fl,<Tt2ZXC. A ,NC.ZSXLFtl,2>t3.0 INTPL C AL K CAL pl (K r E1j<TE2jZXZ*4.NCZS*E1E2E3*E40) I C AL INTPL (K r El#KTE2#ZxZZ#lNC#ZSE2E3jE4#t5#0) CAL SPLIF i NO Z S YL E > El > E2> E3 , 1, T 1 1* T 2, G C . INL CAL INTPL (KT El>KTE2ZYOi>NOZSYLttiE2*E3.0) KT El*KTE2ZYZfl>NCZSjElE2fE3E*C) CAL 1 N T P L CAL INTPL (KT El#KTE2ZYZZlNC#ZStE2E3*EÊ5#0) = B * TA N S U dbp Sl = B*S1 S'* B*S2
=
( 1
si S2 Ti T2
F
= T A N
S Wh
AN
EP2 EDiJ
b>
T= 3 *T AN
H&
Tl = r? = B*T2
>c YC
(
2
Z
)=3.*( XC (3 )-XC (<) + XC( 5 = 3 * YC 3 )-YC(4) )+YC( 5)

) )
N=K T E 2 +
-74-
IF
(r-
.GT.K2)
k = N
,
<\
GD
TO
)/3
50
OG
4C
ZZ=(Z(K)-ZUTE2)
A=EXP(-ZZ)
XC (K
)
XC(KTE2)+S*ZZ+{S2-S)*ll.-A)
)
YC(K) = YC UTh2) + T*ZZ+(T2-T)*ll.-A) XZ(K)=(S+( j2-S *A) /3 +(T?-T )*A) /fl YZC<) = XZZ(K)=-(S <:-S )*A/ (6*6 YZZ(K) =-(T2-T) *A/( b+6) 40 CON! 1NUE
(
I )
5C CONTINUE YC( Kl-1 )=YC YC K2d =YC

( )
(Kid)
(
K2-1
^ F T U* S
END
SUB ROUTINE S U F (NL),NENCNA#NZISYMKTtl,TE2SCAL>AC>Z>Z:>>XC*YC 1SL0 PTTRAlLXSYS*NP#ITtlITE2#IVS0XP*YP>Dl>D2*D3*X,Y#lNDXK,PA, 2xZ, Y Z , A 1 , C 1 , < S Y M INT EkPOLATES CAPPED BLADE SURFACE AT "Pbri POINTS INT EPPQLATION IS LINEAR FOR CYLINDRICAL COORDINATES DIM ENilUN 39(nE1) XS(ND.1) YS(NDd). ZS(i) SL0FU1), TAil_(l)>
)
XC (l)i
YC
I
1
(
>
A
i
,
(
i t
, I
1
(
,
)
Yd),
I
X
<
H (1
)
>
Y P
(
D1
(
D2
1
C
?f
3(1),
)
\l
NE
NP
T E 1
T E 2
Al(l),
31(1
PI- 3.14lt>92b5 SI- .i*SCAL

T = l
./SI
2. /NX LX- N X / 2 + MX = NX + 1 rz = NZ + 3 I vo -1-ISYM-I
D<
SYM-ISY1
VI --1-ISYM CO 1 C K = 1 f 1 ITE 1 (K )=MX ITE Z(K)=MX DO 10 1=1, MX IV( 1*< )=-2 S0( I,K)=0.
I
10 CON TINUE
K=K TE1 K2- 1 K? = K2+1
-7bK2-1
I
R2 =
U
3
VD
P2 R2 Rl
C=
=
=
=
(ZS(K 2 )-Z (< 204C#30 (EXP( Z(K))-EXP(2S(K1)J)/(EXP(ZS(K2))-EXP(Z5(K1)))

)
PA*Ri!
+(l.-PA)*{Z(K)-ZS(Ki))/(ZS(K2)-ZS(<l))
)
1.-R2
Pl*xS( 1K1 )+P2*XS<lK2 R1*YS( 1><] >+P?*YS(l>2) ex =EXP( X K*C )*CCS( XK*D)-1. CJ(- T XP (2. * XK*C )-2.*CX-l. c- (C+Cx) II cc = S C R T (C + C )/SCAL)
o
IF
If
i>
Dl
5C
( ( (
2*'^X
) )
AC( I )*-. 5*DX .LT.-CC A0( I)-.5*DX) .LT.CC)

II 12
11*1+1
12=
I
cc:
NTIMUc IT El(K)i IT E2()=
CC AO( 12 )/CC KK K1 Pa Rl
bO
NP(KrO
X
XA *EXP( YA EXP( KB = ExP< YB EXP( C* Q CC

(
K*XS(1<K))*C0S(XK*YS(1K<))-1 K*XS(lKK))*SlN(XK*YS(lKK)) x K*XS(N>KK) )*CCS(XK*YS(N*KK) )-l x K*XS(N#KK) )*SIN(XK*YS<N#KK)

X
)
)
Cx =SQRT( XA*XA+ YA*YA
(XA+C X)/2. /CC 7C 1 = 1MX = C*A 0( I) X 7C en TINUE

=
SQPKC X/C)
)
AN v
L =
PI
PI
DO
x I
= 1
YI R>
IF
P(XK*XS(IKK))*CUS(XK*YS(IKK))-1. P(XK*XS(IKK))*SIN(XK*YS(I<K))
T
XI**2 + Y 1**2 LQ.O.) GG TO 80

(
AN U" V R
ANGL+ATAN2<(U*YI-V*XI)#(U*XI+V*YI))
T( (R+R
WSCAL
XP YP GO 80 AN U"
v xp
I
)
=R*CDS( .5*ANGL) P*SIN( ,5+ANGL)

90
PI
YP
I)=0.
-76-
CON TINUE ANG L= ATAN SLO* T(KK ANG Ll= ATA N( YA/ X A ANG L 2 = A T A N Y 8 / XB) ANG Li = ANG L-. b* (ANGL1-TP AIKKK) ANG L5r = ANG L-. 5* ANGL2 + TkAIL KK ANG Ll= ANG ll + XK *f3 (1KK) ANG L2=ANG L ?+XK *YS (Nf KK Tl = T A N A N GLl T2 = T A N A N GL2) CAL L SPLI F (1, N>XP*Y?Ol*d2D3l*Tl>l*T2#00.#IND) (II ,I2>X,Y,1,NXP,YP#D1D2D30) CAL L -T P L
( )
('
DEF
X A =
IMTIIN
XS( 1,KK
)
OF
THE
VORTFx SHEET
XI
.25*XA
(
A=S LGPT(KK)*(XA-X1) e=i ./ XA-X1) ANG L=PI+PI

= 1
v = o
MI -1 x xs l.+S CAL*(X(Il)**2-Y(Il)**2)/2./G/0 YV5 SCAL *X(I1)*Y(11) YYS ATAN 2( YYS XXS) /XK
DO
10
1M
5
t=l
/I .0
100 CON
E= l
INUE
05*E XX 1 ALOG (l. + .5*E*SCAL*X(I)**2/0/ti)/XK D=B (XXI -X 1) YY1 YYS + A* (0-1. )/D XX = XP X K*XX1)*CQS (XK*YY1)-1.
(
V=
XP(
K*<X1)*SIMXK*YY1)
)
RS PT(X X**2+YY**2 ANG = ANG L+ATAN2((U*YY-V*XX),(U*XX+V*YY)) U=X V=Y

P = S
KT
(R+*)/5CAL )*U
)
XP(
IF Y(
I
11C CON
X3 =
A = S
)=P*CDS( .5*ANGL) GO TO 100 XP(1 ).oT. x( I *R*SIN( .5*ANGL INUE

) )
N, KK)
)
B= l
OPT( KK) *( XB-Xl /( XB-Xl

)
ANG
U= l v= o
M=
I
0.
2+ 1
(
XXS = 1 .+SCAL*( < 12) **2-Y( 12 )**2 )/2. /Q/a YYS SCAL*X( I2)*Y(I2)
-77-
YYS
DO
= A T
1
.
A.nI
rYb, XXS
/XK
20
/1.0 E = l 120 CONT 1NUE t=l 05*e XX1 = AlOG (l.+.5*F*SCAL*X(I)**2/C/Q)/XK C=B* XXI -XI) YY1 = YYS + A*( D-l . /D XX= E XP X K*ul) *CCS (XK*YY1 )-l. YY=E X P( X k*xxi *^ in( xK^rr i) R-SQ RT( X x**2+ YY**2 ANGL = ANG L + ATAN2({U*Yf-V*xx),(U*XX+\,*YY>)
.
( )
u=
V =
xx YY P = SO KT( (k + X >/5CAL )*0 XP( I )=h-*CC:>( .5*ANGl) XP( I ).LT.X I) IF GO

(
( )
TJ
120
Y(
=R* S I
5* AN G L
13C CONT INUE C=P* CC DG 1 4C I=1MX S0( I >K) =S0( 1 K )+Q* Y( I 14 CQNl INUE KK .EL.K2 IF GC TO 150
)
KK
= K 2
P=R2 GO T
15C CO
IV(
1
I
60 60
1
11*12
>K)=2
160 CONT INUt S0(1 ,K)=SQ(2,<)+S0(2,K)-S0(3,K) SO(M X*K)=SO(NX,K )+SO(NX*K)-SO(NX-l,K) M=I1 -1

00 ZZ =
IF
1
Z
(
70
(K
)
1=2.^1
) )
ZZ.GE.Z(KTE1
1
IV( I,K)
1V0
170 CONT INUE

M = I2 +
DO zz =
IF
1 z
(
60
(K)
1=M#NX
) )
ZZ.GE.Z(KTE1
IV
I>K)=I VO
ieo CONT IMUE K2 = K 2-1 K=K + 1

IF
(
K.Lt.KTE2) GO TO 20
Kl=3
K2 = N Z + 2 190 DO 2 00 I=2NX
zz
IF
= z
(
U)
ZZ.LE.ZS(NC).ANO.ZZ.GE.Z(KTED)
IVC I#K J- IVO
200 CONT INJE K=K + 1
7 ' -
IF
L E
K 2
Li
iVu
N=K
1
U
i
1 l
fi
-
UTE 1
1=2
(
01
IF
i
?c'J
v
If K2
,
N
.
>
I
1
K
<-l
c-
G
.
U
F
.
210
I
>/
(
(
IV
1
,K +
i
I
GT
K+
. .
I j
K
<
) )
= I V 1
IF
+1
G7
C*
I- 1
>
*- 1
G T
1 #
V 1
21C
N TI
M
)
220
0( L X,K (?. / K r CON 1 IN IE

b = S
IF
5/(T-S*S).LT.l.E-u5)
,
L x
= J
2 30 1 = 1 bO( I f M Z)=3 S0( It * Z) =K S0( 1< 1-1)
MX
,*(SG(IM7-l)-S0(lMZ-2))+SO(INZ) >YM*S0(I#MZ)+(1-kSYM)*S0{I>MZ-2)
=
SC( I>Kl + l)
+
23
CON
ITF
IN
f<
LIE
1TE 1(
Z) = K
SYM*ITEl(MZ)
(l-Ki)YM)*ITEl( M7-2)
1-1) ITEKK1 + 1) ITfc 2(M 7 )=N K + 2-1TEKMZ) ITF ?(< 1-1) =ITE?(K1+1)
IK
FT IF N
ENC
SUB INI
U7
AL.
F^T iv
OF
7
)
FS ri".ATE
KEDUCEO POTENTIAL
COM
I> AG
6(1
)
2 9
(
2 t
* 1
1
)
c *
,
i
(
E
)
7
(
I
)
29
, A 1
<? )
A2
1 2 i
A 3
1 2 9
-3
2 6
I 2
(
-)
1
)
, 1
(
T E 1
17
fc
)
2
:
2 6
,?. 2
1c
i3
17
2, CI
3 Z
(
7) >G2
>
(
AHA.
T = 2
(17),C3(17),CM17)>C5(17)XC(17),XZ(17),XZZ(17),YC(17),t Y Z Z 17),KSYM,NX*NY*NZ><7E1KTE2,ISYM.SCALSCALZXK,0MEGA,ALP SA,F MACHTPA,ABLAOE

1
SC AL
+
+ 3
1 =
KY =
^z
=
on
DQ
1.
129
It
1
DO
G(
I
10
COlNi
J=1j K=l, K)=C Not

.
KZ
DO DO
IF
V
1=2, NX
K
(
= K
i
ZMZ
)
LT .2)
GO
TG
DSI
DSK
0(I+1K )-SC( 1-1, K) J (I>K+1 )-S0( ItK-1
-73-
57
*
Al CI
C(
<
*D S *
l)
U
I
)
S<
= A
Y = 5
I
I
XX =
yx
xK
r.
*X*<T*-X**? + Y**2)/{ (T+X**2-Y+*2 )**2+(2.*x*Y)**6)

* Y *
(
= 2 = X
/ X
T->**2-Y**2)/< (T+X**2-Y**2)**2+(2.+X*Y)**2)
FH
A2 8Z
X X
+YX* YX
H=l . /Fh
=
-nn ZU)-YX*YZ(K)
-X X* 5(K)
*C DS
Y A
(
+YX*XZ
(K
A =C
Yl
YC(K)
I
ArAIM?(2.*X*Y>T+<
)
x-Y*Y)/XK
P= A
PA *Y1
Q=A
SIM PA^Yi
F7Z = A* A FYZ = H*FZ Z*3Z F*7 = H*FZ i*AZ FYY r (h*H* YX *Yx +FZ Z*( 6Z*riZ+ Xx*XX
(
FX Y F xx = H*H* A* = F> Y-S
(-XX*YX+F ZZ*(AZ*8Z+XX*YX)) X X*XX + FZ Z*( AZ*AZ+YX*YX x*pxv-sz* FXZ

( )
AY = F YY- S X*F X Y-!>7* FYZ A7 = FYZ-i X*Fx Z-S Z* FZZ BY = AY-S.X *AX-SZ* AZ
V=i
= C
w = C
A*Xx* F /F77-C A* YX-UMEGA*XX/C4(K) A*XX + SA*YX*P/F ZZ-OMEGA* YX/CMK A*XA K +S A* C + P*YZ(K) )/FZZ-OMEGA*YZ(K)/C<(K)
)
( )
G(
KY +
T I
= G(
IK Y-lK)+(AX*U+AY*V+AZ*W)/(BY*ei(KY))
20 CON 30 CON
NLt 1 N lE
Kl = KTE1 K? = KTt2 DG 40 K = Kl K?
<-L
EC K)=0. CON T 1 N U t PPT URN END

(
SUBROUTINE BOUND OFFINES THE BOUNDARY VALUES OF THE VELOCITY POTENTIAL G COMMON G<lZ926fl7)*S0U29*17)*E0(17)*IV(12917)ITEK17)#IT2<17) L#A0(129)A1(129).A2(129),A3(129) B 2 6 >31(2b)82(26)jB3(2t>) 1 1 7
(
)
2>Cl(17)C2(17)>C3(17),CM17)C5(17)XC<m,XZ(17)iXZZ(17)YC(i7)Y
3Z(17)jYZZ(i7)>KSY*l>NX*NY>NZ>KTEl#KTE2>ISYM>SCALjSCALZiXK>a! IE&A,ALP ^HA,CA,iA>F'1ACH*T*PA,ABLA0E
v
-80COM MIN /CAL/ PlP2,P3,6ETA,FR,Pn,DGGM,NS,Ul.Vl.Wl,L -,J*,*^,]C .Ju, 1KG COM MON /SW P/ Gl(129,2fc)G2(l29,26),$X(12-n,S2<129),SxX(l?^>, .<7<I2
l ; ;
19)
bZZ<129 ),R0(129),PH129),C(129),D(129),Il,12,lX,*X,*Y,'-<.Y,Ti,AAC
2,Q1 ,02, MM LX = NX/2+i MX = NX+1 KY = NY+1

PY = NY + 2
MZ NZ+3 DX 2. /NX Tl- DX*DX AAO = 1. /FMA CH**2+.2 ZSY M=1-K5Y Gl = 2. /PI 02* 1./P2 RM = 0. GM = o. NM FR* c
IR
JR=
KR = c DG = 0. IG* JG = KG = NS* G II- 2 12 = NX DO 10 Jl, DO 10 1=1, i^X Gl( 1,J)=G( I, J,l) G2( I, )=G< I,J,1) 10 CON TINUE JT- 1 2C CON TINUE
KL (NZ+5-J T*(NZ+1) )/2

KM = KL + JT KN = KM* JT I*L X G<1 ,l,Kl)= G(I,1,KN)+ZSYM*(G(I,2,KL)-GU,2,KN)) DSI S0(I+1 ,KM)-SO< I-1,KM) DSK -SO(I#K N)-SO(I#KL) SX( I)=A1(I )*DSI SZ( 1) = CKK M)*DSK R*l .0 DO 30 J2, KY G(l #J,KL) G(1>JKN)+ZSYM*(G(2,J,KL)-G(2,J,KN)) G(M X, J,Kl) G(MX#J#KN)+ZSYM*(G(NX,J,KL)-G(NX,J,KN)) YP BO(J)+S 0(I#KM) J.EQ.K Y) R-AMINOU, IV( I, KM)) IF
(
-?1r-*(T-rh'*>.J)** i/(l.-K + {?./XK*YP)**)

1
/=-YPxT
>
/
i
< .*
)*p./X*/n-YP**2)
*
,
<
/ x
J - * 2
>
=
=
r-
H*ii*il(I)*JT
<
h
=
i
<
<!
*
i
* - I
* 4 1
* J T
><
>
v:
;.
1 ,
!*
1
ij-
d
]
'
,-L)
1 > J
] . )
,i;ij*[)f )-Mij-i^ ') = ;(I.j.k' hl/.-i-jGI-L'D^Jl/CK^)

=
=
)
(
UfJjKL)
(
v.
d
i
*>
J
)
+ 3
7
.
C
(
*
.
-3
K M
)
1 *
."-
-G
I ,
= *
v
J J
=o
=
)
i
>
(
UJ.KN)43.*(G(IJfKL)-G(ItJ.. KM))
?
(
] j
t 1
=G
=1
I ,
4 Z s Yy*
-G i
x / e
i
I
-i
i
lx-i
i
:
-.
n
:;.
inu
I
= L X 4
S
1
5,
w
:
,
;
X-
r
i
S C
(
l
i >
)
K K
< = >
y
i
)
)-i
*
i
kl
{!)
i
( (
/:
= :
] K
f (
(
L)=j(I1KM)4ZSYM*(G112*KL)-GI*2KN)J
K r
+
J
I
)
= A
Y = H
(J
e
.
I
(
. T
=
=
<-./*
)(**2'4Y**?)/((l4X**2-Y**2)+*i:+(2.*X*Y) )'*2) T- X*^-V**^l/( (T4X<*2-Y**?)**i:4( 2,*X*Y)**2)

,
=
7
/
!./(/> -x x * =- x x *
=
+ Y M
V!
YX
X
)- Y
)
*YZ
<M
= S 1
'
)*
*
Jl
(
=
=
H* ^ (H*(
41
5 (
I
I
.;
*
)
-JT*-
*Hi
)* JT
k= 1+ i = i 3 I = C
>
U
(
J 1
- G
-i
v
( 1
i
J =
( j
G(
J
K
i
)-G(IJ-li*M)
L
. .
KM)*G{I#JKN)+A*{&{IFJKL)*0GI)-6*DLJ)/(C1(KM)4A)
H
I J)
1 # >
J
J
I
>
L)
)
?
(
J
>
kn)+3.*(o( [JKL)-G(IJi<M) J 1 )-G2( I. J J J4ZSYK* (G( 1

t
r>- t l
=
(
wu
.
".
Ku
1
(
f
J
J
,
l):(CH'<
=
vi)*iJ
)
(I>J>K\|)4AMG( IP>JfKL)4UGl )-6*DGJ)/(;il^)U)
KL
-82-
IF
.1
.L
TO
GO
TO
7C
CON 1
IF
(
Nut
1
tu.-i
liC
KK
L =
I
= ?
3
t>
J=N
K = L
X = A
&( G(
P
(
30(
Yl*
A*
DEN =
yx
= c
(YC()+ATAN2(2.*X*YT+X*X-Y+Y)/XK) T +X*X-Y*Y)**2+(2.*X*Y)**2
/
XX* 2.
.1
(
XK*X* T+ X* X+ Y*Y )/DEN XK*Y* T-X* x-Y* Y /DEN

( (
H=l
G
= C =
./
XXTXX+YX+YX)
) )
P = C !>{K
)*COS( Yl &(K * S I N Y 1
(
AZ
X X
/:
K, )
Y X
* Y Z
BZ = -XX *YZK )+YX*XZ (K) = Aid)* (jC( I*1K )-S0( I-l.K SX( SZ( I) = C1(K)*(S0(I K+1)-S0( I,K-1) GX = Alt I )*<GJK )-G( 1-1 tJ,K) GY = Bl( J)*(S(IJ-1K)-G(I,J+1,K)) GZ = Cl( K) *{ G( I, J, K+l )-G( I, JK-1) U = G X-5 X I) *GY
1
V =G w= G
00 = VO = WO = ul vi =
wl
=
Z-S Z 1 *GY H*( XX*U-YX * V) H*( P*YX+0 *AZ)*U+(P*XX+2*BZ)*V)+G*w h*( P*AZ-C *YX)*U+(P*SZ-0*XX)*V)+P* .9* U1+.1*U .9* V 1 . 1 * V . 9* * 1 + . 1**>
(
)
OD 100 K = K < , M IS* ISI GN (12* (MZ-K)-l IT = ISI &N(1,2* K-KK )-l 1=2NX DO 8
(
X =
A C**2 XX = 2./ XK*X* T + X*X + Y*Y )/0EN YX = 2./ XK*Y*( T -x*x- Y*Y )/DEM Yl* PA* YC(K)+ AT AN2 <2.*X*YT+X*X-Y*Y) /X< G?( 1*1 )=G1(I . 1)
Y = S 0(
)
Gl( If 1 )=G( 1,1 K) G(I tit K)=G(I 3>K )- (YX*Ul-XX*(Vl*C0S(Yl)-Wl*SIN(Yl))/Cij(<))/61(2) G(l tit K)=ZSYM *G I, 1>K SX( I) = Al( I) *< S0( 1 + 1,K)-S0( 1-1, K) SZ( I)* C1(K)*( S0( I. K+IS)-SO (IK-IT) DSI I = S I + 1 # K )-S0( I*K)-30( I*K)+S0(I-1K)+A3{ I )*OSI DSK K = S 0( IK+I SJ-SO (IK)-SO(I,K)+SO( It K-IT)+C3(K)*DSK DSI K = S 0< I+UK + IS)- S0(I-1 <+!$)- SO (H-1,K-[T) + S0(I-1,-IT)
(
SXX(I)=A<:(l)*liSlI
SZ2(I)=C2()*DSKK S*Z{I)=T1*A1(1J*C1(K)*0SIK eo cnMiNiit

DQ
Y =
:
<*C
(
J = <:',<<Y
X=A U
2 2
-i
ATAN2(2.*x*Y,T + X*X-Y*Y)/Xi<) 0tN=(T-K*y-Y'l Y)**?+(^.*X*T')**2 = t./>,Kn*( f+x*x+y*y)/den > YX=2./XX*Y*(T-X*X-Y*Y)/DEN P=C5(K)*CaS(Yl)

Yl
<
?A'MYC('<)
C= C
MN
Y 1
GY=Bl(J)*(G(2tJ-iK)-G(2*J+i*K)) G?(1,J)=61(1J) G (1J)-G(1JK) G(1,JK)=GOJ>)-<XX*01+YX/C4(K)*<?*V1-Q*W1)+SX(2)*GY)/AI{2) G(l,JtK)=Z5YM*b(l,J,K)

NL=h x-i
X*A
(
NX
Y=SO(NX,K)+B0(J) fl=PA*( YC( K) +U AN2(2.*X*Y,T+X*X-Y*Y CEN=(T+X*X-Y*Y)**2+(2.*X*Y)**2 XXsg. /XK*X* (T+X*X+Y*Y /OtN YX=2./XK*Y*(1-X*X-Y*Y)/0EN P*C5(K)*C0S(Yi)
)
/ X
K)
Q=
C5(K)*SIiN(Yl)
)
(
GY*6HJ)*(G(NX,J-1,K)-G(NX,J + 1,K))
G2(MX J =G1 MX, J G1(MX,J) =G(MX, J,K) &(MX, J,K)=G(Nl> J>K)+(XX*U1+YX/C4(K) G( MX, J.<) = IS YM*G (MX, J ,K CONTINUE
)
)
*(
P*Vl-Q* t*l
+ SX
NX
*t> Y
/A
N X
G(1,1,<)=G(1,2,K)+G(2,1>K)-G(2,2,K)*ZSYM G(MX,i,K)=G(MX,2,K )+G (NX , 1,< )-G NX , 2. K) *Z S YM E=G(1TE2(K),KY,K)-G(ITE1(K),KY,<) GU,KY + l,K)*G(MX,KY-l,K)-E*ZSYtf G(MX,KY+1,K)=G(1KY-1,K)+E*ZSYM G(MX,KYK)=G(1KYK)+E*ZSYM
(
IF iS*Ll .GT.O) 100 CTNTINLt

(
CALL YSwEEP
GG
TG
(K)
IF
JT G3
(KSYM.NE.O)
20
110
TG
110 CONTINUE PR=1. 2*FK/ AAQ RM=1.2*rtM/NM/AA0
GM=GM/NM RETURN
END
-84-
SUb F U 1 N t Y S W EEP (K) THE E30ATIJNS FOP G AkE SGLVEU HbPE FOR MIXED SUBSONIC AND SUPEKSJNIC FLOW ,EY 0* PELAXATIJN ,AND BY USING A HT ATED DIFFE K ENCE SCHctfF
I
1, AG (129)>A1< 12 2, CI (17) ,C2< 1 7)
COM MC'N G(129,2 fc,17),SC(129,i7),E0(17),IV(129,17),ITEi(17),ITE2(17) y)#A2(12^),A3(129),BC(26),^l(^t),82(2 6),6 3(2fc)-'(17)

c
32(1
<tHA,
YZZ
<
17
,C3(17),C4(17),C5(17),XC(17),XZ(17),XZZ(17),YC(17)Y KSYM,NX,NlY,NZ>KTEl,KrE2*ISYM,SCAL,SCALZ,XK,0rEGA,ALP
v
CA,SA,FMACH ,T,PA, ABLAOE COM MDN /CtL/ P lP2.P3,BtTAFKF lDG,GM.[N,S,Ul,VlUl.lK,jK,KRio,JG, 1KG COM MJN /bW?/ G K129,26)G?(12 9,26),SX(129),SZ(129),SXX(129),SXZ(l2 19), S Z Z 1 2 9 , R (129),Rl(12^),C(129),D(129),Il,I?.L vlx,KY,MY,Tl,AAG 2,01 , Q 2 , n ^
(
)
L=K
JL=
IF
C
(
D(
(FWAC h.GE.l 1-1) = c. 1-1) = 0.
Jl=3
DO 10 1 = Il 12 kj( I )=1. Rl( I )=1. Gl( I. 1) = G( 1,1, L) Gl( I , J I - 1)=G(I f Jl-1 10 CGN T INUt
J = J 1
13 = 12
2G
CON
TINUfc"
)
BC -Tl*rt 1( J )*C1(K DO fcG 1 = 11,13 AB = -Tl + A kd*b: j) AC = T1*A1 (I )*c (K
k
)
X = A Y= S
*X-Y*Y )**2 + 2.*X*Y )**2 XX = 2./XK *X*( T+X*X + Y*Y )/DEM YX = 2./XK *Y*( T-X*X-Y*Y)/DEN FH = 1 .-kg (I + XX*XX+ YX*YX H = R 0(1)/ FH AZ = - X X * X Z(K)-YX*YZ(K) 8Z = -XX* Y Z(< )+YX*XZ(K)
DEN
(
)
G< 1) 0( I.K = (T + X
B0<
E=H *AZ E = H *BZ XXX = 2. /X K*{ (T+ 3.*X*X+Y*Y)*DtN-4.*X*X*(T+X*X+Y*Y)**2)/DEN**2 YYY = <. /X K*X*Y* (DEN+2.*(T+X*X+Y*Y)*(T-X*X-Y*Y))/DEN**2 AA = P0( I) *Cbi K) Yl = YC(K) +ATAN2(2.*X*Y,T+X*X-Y*Y)/XK
-t b-
l" ?
Tf-c
f-.'LL. hJf' c
A < t
1
LINES
r
:
iHf COFFFiCItNTb
a
Of
I
THt
I
J A C
"]
fi
A\
I>
1 '
i)Eh
>
EL L
AS
Hf
I S
Dc K
VA
IV Pi
win
RESPECT
A = f =
[j
THE
CjCÔINATES
*X +
X X
H*
h*
X y a x * *
<?
A<=h**.*((Yv* ?-X
A Y = H * *
>
+c
<
y *
* Y Y Y
)
>.
X
)
)
X*XX>
(YX**2-XX**Z
* YYY
P=AA*Cdi
PX=-Q*YX
PY=->J*X\
Y i* P A
C=AA*jIMYl*PA
P7=-P*PA-w*F QX=P*PA*YX QY=P*?A*XX GZ=-0+P*FA*F
E
X=-
AX* ^Z
(k
AY* YZ(K)
bY=-(AY*X/(K)- AX*
EZ=-( A*x^Z
(
mo
Z(K)
(K
)
)+
t*Y
FZ=-A*YZZ(<)+2 *xz Z
YXJ=P*>3 + Q*E
Y 7
YJ
x
I-
n+^F
*
<3
j = P * E
ZYJ=P*F-J*A
YXU
= PX*!i + GX*E + P*A YXV=PY+ri+QY*E- P* A
+
+
*E
Q*EY
f
YXW*PZ*d+QZ*E+ Q*fc
Y
ru= h<
=
* A +
ux
*A
-G*E
PY*A + ;:r*F + P* A YYW*PZ*A+QZ*F+ 0*F

YYv
P*E
+Q* X
-o*Ar
+
PY*t-G Y*^ + P*E /X' = PZ*E-QZ*P + P*b ZYU= PX*F-3X*A- P*b Z YV = P Y*F- J Y* A + P*E ZYw=PZ*F-QZ* A+ P*F DGI=G( 1+1, J, L) -G( DGJG(I#J+1#L) -Gl DGKG(I*J#L+1) -Gl
Z
Xv
Q*
A X
-0*Ax
-Q* A Y
-1
I
ItJ-1)
,J
)
DGII=G(I+l>JfL )-G IJfL)-G(IfJfL)+G(I-lJfL) + A2( I )*0G1 DGJJ=G(I*J+1L )-G If Jf L)-G(1JL)+G(lf J-l.L) -i3( J *L)GJ UGKK=G(IfJL+l )-G I, J L )-G( It J, I )+G( I. J, L-l + C3(K *0GK
)
)
DG1
J=
G( I+lf
1
)
J +l
DGIK=G(I+1JL
DGJK = G(
bX=
/ 1
(
+ 1
J + 1
1
L + 1
)-G( 1+ 1, J-li L) + G(I+1J L-l )-G( I-l# J. L+i )+ G( I J- 1 . L + 1 )-&( It J + l L-l

J
G( 1-1
.
+ lt L
G G
J-i
J
L
l
)
)G
I
I .
L-
-1
L-l)
GY=-31
(J
*UG )*u^J
GZ = C1U)*DGK UP = GX-SX;(I)*GY V = GY WB=GZ-SZ(I )*GY
U=UB+C A+xX+SA*YX* P/C4U) )-CA*YX V=VB+S A*XX*P/C4

(
= w b +
C A* u
X I
(K)
SA*(G + P*YZ(<))/CMK)
+
UC = A*U-3*
vO bG
=
=
Y
X J *
+ Y r/J*\/
C*w
ZX J* J+7 YJ*V + P**i

 + wC**? vo= VO-OMEG A * P / C 4 ( K 1*0 = WG + 'JffcG A* -/C^t (K

)
)
AA = D cc = U
F
XX YY
AUf
AA *U +V YX M
(
>.2:*Ow-2.*Cj' 1EGA*(3*U + A*V)))

,
*vo+wo**c J*>xj+Zxj*ZXJ
*
y r
= b * 6
+ Y
Y j
j + ? y
vi
y j
FZZ = C * w + F * p FXY = -&* + Y xj*yyj+zxj*2yj FXZ = * Y X J + FY2 = (-*Y Y J + P*Z Y J

t$
BU = A*UJ+> X J* VO+ZXJ*W0 3V = -B*Uo+ Y YJ*VO+ZYJ**0 BW = Q*VC+P* 40
0U = A*AX-b* AY+YXJ*YXU+YYJ*YX\/+3*YXV +ZXJ*ZXU+7YJ*ZXV+F*2Xw 0V = - A * A Y - *AX+YXj*YYG+YYJ*YYV+y*YYw+ZXj*ZYU+ZYJ*ZYV+P*ZY* 0'*' = rx j*wX + fYJ*t.Y + C*GZ + ZXJ*PX+ZYJ*PY + P*PZ cu- UO* (AX* 3U+AY*&V)+VO*(YXU*BU+YXV*3V+YXW*tf*)+WQ*(ZXU*3U+ZXV*3V+ZX
1
fa
w)
CV
oO* -AY
(
)
*ts
L+
:
AXtbV)+Va*(Yf u *BU+YYV*3y/ + YY"w*oW)+wC*(ZYL*DL+ZYV*bV + Z

Z
1YW* Fl> cw= VC FXX = F FY* = f

F
j>*
k 1
3
(
U + Q Y* 6 V + 3
I
)
*aW
WG*
? X*
6U + P Y* t) V+
P Z * BV
XX +
r
-k
1
J
+ 1
-0(
X.Z
(
f =
SX(
H* X?*rXX+FYY+SZ(i)**2*FZZ-2.*iX(I)*FXY-2.*
(
<
-Z(I)*FY 7 + /'.*
S X
II)* SZ(1)*F FXY =FXY-bZ FYZ =F YZ-bX

AV
=
EV-SX
iiu*riu
(I
* FXZ-SX *FXX *FX7-SZ )*FZZ )*BU-S2 1 )*8w

I
)
(
UU
I
V =
BU*AV
UW = bU*4 w vv = AV*AV v W = Av* ti*

W'W =
i?
W* b W
AZZ AXZ
R=-
1CV) AXT
I
=F XX*t -bU = F/LZ*AA - * =2. *(FX Z* Al-UW AX A*bX <(I)+AZZ*S72(l)+AXZ*3XZ(I))*GY+TlM(AA*Ji-CU*ct *V3+( AA *DW-Cw)*WB)
)
a A * D V
=A5 S
6U * A 1
)
>
)
)
AY =A3S( AV * L (J A/r = A 3 S B * C 1 K A=* 0( 1 )*bE rA*AA/AMAXl(A<T,AYTAZT*l.-SO(I)) AXT = A*A XT AYT = A*A Y AZT = A* AZT
( *l
( )
SUPERS
IF
a y
(0J
* = i k
I-nJC *3 . G t . A A
)
NTS
GO
)
AMD
Tu
31
)
jUÎMC
POINTS
A-
StPA^.ATFL
x*A3
YY
H
Y =
A Z Z * AZ
AX Y= AX Z =A
(
f
K
V *B2 (J 7*CZ K XY*AA-U\/ )*( Ae + AF)

(
a-v
)
i *
AC
A Y 7 =
YZ*AA-VW)*(8C+3C)
-AXX -Q1*(4YY+AZZ) D&II +AYY*0GJ J + AZZ*OG><N+AXY*QGI j
t P = A X X. v = A X X B
i
B*GG
XX
R =A X X *
+ A YZ
*C& JK + A X 7* OG I<+
TO 40 CONT IN UE
N
S = N;> + 1
s i = s i
Ml.
.1)
i
1
IM= l-i Flx(
IMMIM -IFI
AX
X =
X( SI
UU *A^( 1 AYY = M\J *B2< J) A Z Z = w *C2 K) AX f = b. JV* A E AXZ = 8. *SI* J.v*AC AYZ =8. 3C dx a = (F ax *u J-UU *A2 I BYY= F T Y *U 3-VV )*S2( J) bZZ=(F ZZ*Q J-nW )*C2 (K BXY=(F XY*J J-UV )* AB+ AF bXZ=(F XZ*G Q-0* *( AC + AC 3YZ=(F YZ*0 i-M'n )* (BC + bC AQ= AA/ au DEL TAG = 3Xx *DGII+BYY*DGJJ+8ZZ*DGKK+BXY*DGIJ+8YZ*DGJK+BXZ*DGIK DGI I = G i>j L )-G(XM*J*l)-G( IM> JL)+G(IMM*JL) + A3( I)*DGI DGJ J=G I ,j ,L)-G(I,J-lL)-G(i,J-l,L)+Gl(l,J-2)-P3(J)*l&J
'*
DGKK=G
i,
j
j
D G I J = G (i#
(
DGIK=G ij DGJK=G (It j

GS
S A X x *b G
U-G( I JjL-D-GCI* JL-1)+G2{ I J)+C3(K )*DGK L)-G(IMJL)-G(IfJ-l,L) + G(nJ-l,L) iL)-G(If JjL-l)-G(IMf J,L)+G(IM,J,L-1) L)-G(lJ,L-l)-G(l,J-I,U+G(I,J-i,L-l) II+AYY*0GJJ+AZZ*0GKK+AXY*0GIJ+AYZ*DGJK+AXZ*D3IK
.
)
*( AXX + AXX + AXY + AXZ B.5*< Ag-1 PP= ACi* BAX- (l.-SI *3 BM= A3* 3 XX- (l.+SI)*B b=-AQ* 3XX +BXX+Q2* (BYY+3ZZ ) )+( AQ-1. )*( 2.*( AXX+AYY+AZZJ+AXY+AYZ+AXZ
)
) (
40
AQ- 1.)* GSS+AQ*DELTAG+ft CONTIN Lit

R=
IF
(A3 S(R) .LE. BS< FP

I
GO
TO
50
IPJ9 =
K! = K
DC
CGNTIN ut
-8 6-
R=
R-AYT*(GK
+ +
It
J-1)-G(I J-lf L) )-AZT*(Gl<
if J
)-G
I f J
L-l
Rtf = R!*
A8$(R)
NM=N
i"l
b=B -AAT-AfT-AZT
BMBM+A<T
B=l
.
<
tf-6*!*C
1-1)
(
C( D(
)=d*3P
)
B*(K-BI"i*0
1-1
6C
CONTINUE
CG
1
G.
M=
1 1
# I
13 DO 80
=
CGD(1 )-C(I)*CG
GM=Gtf+Ai3S (CG
IF
)
(A3S(CG).Lt.A6S(DG)) GO TO DG=CG
IG=I JG= J KG = K
7C CONTINUE
G2(IJ)=G1(
Gl( If J
)
I* J) G( I, JL) G( I J L)=G( If J >L

=
)-CG
1=1-1 to CONTINUE
J = J + 1
QC
IF (J-KY) 209CfllO con riNut I3=ITE2(K) IF (I2.GT ITE2 (K) ITF2(K) .EQ.HX) 13 = LX IF = I i 13 CO 100 LV= IA6b 1-IA3S I V If K
) (
I
(
kO I = A M I NO Pl( I)=LV
(
)
L V
A B S
IOC CONTINUE GO TO 20 110 CONTINUE I=LX+i
IO=NX+2-I3 IF (K.GT.KTE2)
DO
GC TO 130
120 X=AO( I)
=1 Of
I 3
Y=SO(IfK)
DEN=(T+X*X-Y*Y)**2+(2.*X*Y)**2 XX*2./XK*X*(T+X*X+Y*Y)/DEN YX=2./XK*Y*(T-X*X-Y*Y)/DEN A=l .-R0( I) +XX*XX+YX*YX

HRO(I )/A A7=-XX*X2(<)-YX*YZ(K) BZ=-XX*YZ( K) +YX+XZ (K) A=RO( I )*Cb(K)
Yl*YC<K)+ATAN2(2.*X*YfT+X*X-Y*Y)/XK
-8<i-
p = i
+ca s
^* Yi
(
C=A * s I N P4*y FZZ = A* A

(
1)
Y
r y r =
hX
f x
+FZZ*(8Z*BZ+XX*XX))+1.-K0(I) +FZZ*lAZ*AZ+YX*YX))+i.-fiG(I) X + FZZ*(AZ=MZ+AX*YX) y = H*H* (-XX*

H*r^*
Y
<*
< =
^H*
X X*A X
FXZ = H*FZ / * fyz = H * F Z I*

AX
AY
=
A 7
it
I
Fx f-i
f
= F Y = F =
Y-j x(
Z
1
I
)
)*F X X - > Z )*F XY-S7(

( )
* F X 1
I
I
)*FY7
)*i-ZZ
)
AZ BY
I
I
.s
A
(
*F
A X
XZ-SZ
-S
Z
(
AY-iX
=G
=
(
v-
*AI
i<Y,L )-C( I-1#KYL L +1 )-G2 I, k Y) V = S A*X A* P/C^ (K )-CA*YX-JMEGA*XX/C^(K U = A 1(1)* Cbi+C A *XX + SA*YX*P/C< (K )-Q>lEGA*YX/C4(K) * = C 1(K)* DGK+CA *xZ(K) + $A*(Q + P*YZ(K))/CMK)-DPEGA*YZU)/CMK) 6(1 >ky + ; L = G IKY-ltL)+(AX*U+AY*V+AZ*W)/(PY*Hl(KY)) 120 CGN t i n u e
DG1 DGK
+
.
G(
KY
= 1
c
(
. N E.ITE1 (K) ) GO 1 =G (13>< Y> L )-G IOKY, C( L) = eo (L )+P3*( t-0(
IF
TO
L
L
) )
3i
130 CON INbE

I
IF
1
(I. It .11) -1
^FTUKN
t =
IF
b = E
(IVli*K).NE.l) GO TC
G(L
)
l<tO
14C CON TINUt M*N X+2-I 6(1 ,KY + i,L

G
(
)
)
>
KY
=G( l",KY-l =G( I ,K Y-lj
L
L
)-E
)
+ b
G2( M>KY)*GL (MKY) Gl( M,KY)=G(M>KY>1)
G(M >KY.L)=G(1*KY>L)+E GO TO 130 END
SUBROUTINE VELO (K,LSV,3*>CP> X,Y) CALCULATES SURFACE VELOCITY AND PRESSURE COEFFICIENT C OIMON G 129, 2 6, 17 j S C 129, 1 7 > EO 1 7 I V 1 2 9 1? > I T E 1 1 7 j I T F 2 1 7 l>A0(12n,Al(i2^)A2(12 5)A3(12S<),BC(26)î(2b),a2(2c).^3(2b)Z(17)
1
) ( ) (
) (
<
2.CK17),C2(i7),C3(17),Ct(17),C5(17),xC(17),xZ(17),xZZ(l7),rC(i7),Y 3Z(17)YZZ(17)KSYM>NXiNiY#NZ*KTElKTE2ISYi 'bCALSCAL?XKJ^tGi.ALr

,
-90-
hH
C A
>
;WI- 1aChTPA>ABLA0E
S v
k
)
(
UIMENSIUN
110=
I
I
>
CP(x)
<
Y( 1)
2 N
T1 2
J =
Y + 1
.
01=
? * Frl A C -1**2
.
T 1 = 1
f-
fc
.
i C
H**2
B=C5
CJ
1
U
C
I (
J > I
A=AG(
Q=SG
I J
DEN=(T+A**2-Q**2)**2+(2.*A*G)**2 XV=?,/XK*A*( T+A**2+n**2)/DEN YX=2./XK*Q* T-A**2-0**2) /DEN YX=2./X>,

(
CL.-V v*V Fh= XX*X H = 0.
V UYU YX
*
-A.
IF
(IV(I*K).NE.O)
=
= 1
. /
AZ Z
-x K*x ?(K)-Y X*YZ (K
BiZ =
-Xa*YZ(K)+YX*XZ(K)
I
D>SI*SO(
k s a i r 1>X= Al(
+ 1K >- SO
1-1K
SK=SO(l*K+i)-SC(I#K-l)
SZ Y1
= =
n *n^ 1 *0S i i C1 (< )*DS)
YC(K)+AU,>I 2(2.* A* GT
(
)
A*A-0*0)/XK
P=B*CGi PA*Y1 C = 3*SIN PA*Yi

(
Of
L
I
- A
I l\ (
'
n L A
T
\J
J
(
=Y C
A N2
A2( A3(
I
I
)=XZ )=YZ
(K
IK
)-YX*SZ + XX*SZ
)
10 CONTINUE
RETUPN
ENO
-91-
sua
n t
CP L
(I112*X,},D*FMACH)
IN
t
PLG TS CP aT EjUAL ir:TE-VAL> JIM ENS ION <JDc(3)j LINE 100) PA A iluc / 1h , 3ri+,lHQ/
(
ME
X(l),
"IAPPE
L A
J
N
)
3(1)
J(
K.
1WR 1T
.PI
= c
(
Tr
I k
K I T
b
1
1C
DO lj 1=1* 100 LIN E( I =KJOt FDE N = l . / b I P AÂ X = )

)
CMM Ax CPM AK
DO 2D
A
[=11.12
A A
=
X I
(
1A
X =
,"i
A 8 S
C MM
AX
A.m]NKC*.MAXjD(I))
20 COM a K = AMAXKCPMAX.D(I)) CM A X = C PMAX-CMM A X DC 50 1=11*12

XP<?
AC
(
= E
S
l) E
N*
r<
Kl = Kl =
./AfUX)*A9S(X(I))-Kl.
(
Mn
K 1
LIN E (K 1 )KQDE( 2) K? = 41 K2 = KIN J(K2. 100 LIN E (K 1 )=<JOc (3 K3 = 3o ./C NUX)*(D(I)-CM--1AX)+l K3 = MIN 0(K340) IP IK3 .oE.l) GO TO 30
)
) (
K3 = 1. GO TG
30 LIN E 40 K4* 1 LIN E

JJ
=
(K 3
)=KODt (2)
3
)
(<
4)=JDE(
(
W*I Tfc IwPIT70) LIN E(K 1)=K3DE(1)

IN E (K 2)=KQDfc( 1) LIN E (K 3)=K00E( 1 LIN E (* h =KQDE I 20 CJN TIN UE RET URN
L
) (
XFk AC,
),
JJ, LINE
fO 70
FOR MAT FOR MAT ENO
lx It IX , Fb
3
.
X 3HX/C 4Xj 3-liviC GMP 2F9. 3, I3>2X,iCQAl
^ K
5HM0SGN* 3H QN)
-92-
UT
^t
i pe e d
T
(K
t
THE PCI
COM 1>A0
2, CI
MACH rtUMBEP ARE CUMPUTFO AT EACH GRID TH EN THE "ACH MU^BERS ARE WRITTEN ON THF 01 T F-UT , N G(l 29, ^6 17), 3 0(129,17)* E0(17), IV(129,17)ITEl(17),irE2( 17) 2s>) ,A 1( 12$) A2(129),A3(129),B0(26),B1(25),B2(26), 83(26), Z 1 7
PttC AND
H
(
7) ,C2
,
(
3Z(
3(17),Ct(17)C5(17),XC(17),XZ(17),XZZU7),YC(l7),Y YZ2 17), KS YN.NXNY.NZ,KTEl,KTE2,ISrM,SCAL,SCALZX<,G' EC;^,ALH

(
17) ,C
<.HA
DM
Cl =
A= C
SA,F A C H T SIGN IS (24)

>l
PA, ARLADE
/f-MAC H K)
* * 2
.2
11 12
= =
K Y=
U
(6*30)
(b,<tO) Ff.ACH 1=11, 12
wRI wRI DG DSI
3(I+1K)-S0(I-1,K)
0(1, K + l )-SC( I,K-1) 1)*D3I (K )*OSK
(
DSK
bX = sz =
DO
X = A
J =
<-,<Y
+
I)
Y = S
i,K)
ao(
DEN
XX = Yx =
T+X*X-Y*Y )**?+( 2. *X*Y)**2 /A**X*(T+>*X+Y*Y)/DEN /XK*Y* T-X*X-Y*Y )/OEN

(
FH =
H=
**<^+Y<**2
IF
AZ
Yl
(FH.GT. .16-09) H=1./FH -XX*XZ(K)-YX*YZ (K

)
BZ = -X
=
X*YZ( K) +f X*XZ
)
K)
P =A
YC(K)+ATAN2(2.*X*Y,T+X*X-Y*Y)/XK *C0S(PA*Y1
0=A *SIN( FA*Y1) DGI = G( 1+1,J,K)-G( 1-1 J,<) DGJ = G( I, J+ 1,K )-G( I, J-l K DGK = G( I J, K+l )-G I, J, K-l GX = Al I)*DGI GY = -BK J )*0GJ
( (
1(K)*QG*-SZ*GY + CA*)<ZU>+:>A=MC+P*YZ(K))/C4('<.) UO H*( XX*U-YX*V VC = H*((P*YX+Q*AZ)*U+(R*XX+Q*BZ)*V)+0*W wo= H*((P*AZ-0*YX)*U+(P*BZ-Q*XX)*V)+P*w

)
U=G V=G W=C
X-SX*GY+CA*XX+SA*YX*P/C4(K)
Y+SA*XX*P/C<t
(K )-C
A*YX
Q= U
0*UG + V0*\/0+W0*W0
-93-
IS(J)
= S
jPT(U/(Q1-0.2*(Q-2.*0MEGA*HMYX*u+XX*V))))*]0CC.
(
10 CONTINuc
2
WRITE (6*50) CONTINUE RETURN
IS
J), J=2KY)
30 FORMAT 40 FORMAT 50 FORMAT END
(1H1) (1SH PRINTOUT (2415

)
OF
SPEED, /,13H
FMACH
4. 2,/)
SUbR'JUTINE FOPCF (I1,I2,X,Y,CP,AL, CHORD, xr.,CL,CD,Ct'Xt<,YK,Zl,PA) CALCULATES SECTION FORCE COEFFICIENTS DIMENSION X(l), Y(l), XK(1), YK(1), CP(1) RAD = 57. 29578 ALPHA= AL/R AD
CL*0.
CD
=
0.
CM=0. N=I2-1
DO 1C
X(
1=11,
I
1
DX=(
D-X
)Y
(I
(
) )
WCHORD
)
OY=(Y(
+1
/CHORD/ZI
)
XA=(.5*(X(I+1)+X(I))-XM)/CHQRD
YA=.5*( Y(I+1 )+Y(I
CPA.5*(CP(I+1 )+CP I DX*DX*COS(PA*YA)+XK(I)*SIN(PA*YA)+DY DYDY*<l.+YK(I)*SIN(PA*rA)**2/ZI) YA*((1.-PA)*YA+SIN(PA*YA)/ZI)/ CHORD DCL=-CPA*DX DCD=CPA*DY CL=CL+DCL CQ*CD+DCD 1C C*1 = CM + DCD*YA-DCL*XA DCLCL*COS(ALPHA)-CD*SIN<ALPHA) CD=CL*SIN(ALPHA)+CD*COS(ALPHA) CL=DCL RETURN
(
)
END
-94sua iRJUFINE TOT FOR (KTtl,KTEiCHLl*D^CLi:0#iCh.Z>^C.CJlLCL >CMP,C ins, CMY. PA> ArtL 4C f CAl CULATtS TOTAL FO^Ct COEFFICIENTS DIM ENSIGN C H 8 C ( 1 ) t S C L ( 1 ) S C ( 1 S C 1 ( I Z 1 . X C ( i ) C b ( 1 SPA N=Z(KTE?)-Z(KTE1J
,
>
CL
C.
CD = 0. CMP = w t CM* = 0. CNY = 0.

S =
N=K
TE2 DO 10 DZ = .5*
AZ = .5* DZ = DZ + AZ* AZ + CL = CL +
CD = CD + CMP = CM
1D(K )*( CMR = CM CMY = cm

10
KTE1N ,/Cb(K+l)-l./C?(K) ,/C5U + l)+l./C5(K) .-PA )* (Z(K+1)-Z( K) )/2. .-PA )* (Z(K + 1)+Z(K) 12. *(SCL(K + 1)*CH0kD(K+1 + SCL(K)*CHORD{K) *(SCD( K+l) *CHCRD(K + i)+SCO(K )*CHORD(K 0Z*(C'H0RD(< + 1)*(SCM(K + 1)*CH0RD(K+1)-SCL(K+1)*XC(K*1))+CMi M(K)*CHORD(K)-SCL(K)*XC(K)))
) )
)
S=S A3L
DZ
AZ*DZ*(SCL(K+1)*CHGkD(K+1)+SCL(K)*CHURD(K)) AZ*DZ*(SCD(<+l)*CHOkO(K+l)+SCO(K)*CHJHO(K)) CHJRD(K+1 )+CHORD(K

)
ADF.
CL = CL/ CD- CD/ CMP = CM CMR = (C CMY (C RET URN END
SPAN/S**2
+CMR)
/
(S*SPAN)
+C^Y)/(S+SPAN)
SU6R0UT INE PEF IN HALVES MESH SIZE COMMON &(1292617)*SC(129l7)EO(17)fIV(129,17)ncl(17)*ITt2(17) 1A0(129 ),Al(129)>A2(129),A3(129),dQ(26>,BH26) '?(26)63(26)Z(l7) 2C1(17) ,C2(17)C3(17)#C4{17)*C5(17)*XC(17)>XZ(17)jXZZ(17)jYC<17)#Y 3Z(17), ZZ(17)KSYM,NX,NY,NZ*KTEl*KTE2iISYM,SCALSCALZjXK,QMFc-.A,ALP 4HA,CA, A,FMACH,T,PA, AELADE MX*NX+1 KY=NY+1 fY=NY+2 MZ=NZ+3 MXO=NX/ 2 + 1
!
-9bNZ /2+3
K = *K
l*.Z0 =
KK
= i
DD J=N
JJ
6G
r
h Z
/2 +
= K y
1C
IM X o
11 =
X
I
20 G(
II IF
J J
J t K
I = 1
1-2 l.GT .0)

1
Gj
TL 20
J = J
JJ
If
J-2 J.GT.G)
C
J=
I
GO
x ,
TO
10
DO DO
1>KY>2
= ^
30
G<
J.K)=.5*(G(I+lJK)+b(I-lJ.K))
J
DO
DC;
I-ltMX J2NY2
*
G(I 50 G(I eo CON

<i0
JK)=.5*(G(IJ+1K)+G(I,J-1K))
my k
)
INUE
MZM MZO MZS *MZ 70 COM INUE

DO DC 80
G( IF
1
j =
i,n
I=1#MX J,MZST )=G( IJ#MZK) MZST.tO.l) GG TO 100

=M Z S
J = 1
1
]
MZS DO
DO
MY
= 1*
MX
Jt
9C G(I
IF(
JMZST)*0.5*(G(I#
ZST.fcQ.l
)
MZ)+G( 1J>MZM-D)
NZM MZM-1
GO
TO
iCO
MZS GO 100 CON KK = DO

I
=MZST-1
70
INUE
5
K=KK,MZ
,3
=M
=
0+1
IF II
K.LT.KTE1.0R.K.GT.KTE2)
10 1 = 11* I? S0( I+1K)-S0( I-1#K S0( I><+1 )-S0( IK-1) 1(1) *OSI
)
TO
12G
TEl(K) 12- TE2(K)

DO DSI DSK
SX
=
sz 1(K)*DSK DGI G( I+1*KYK )-G( I-1KY,K DGK G( If KYK + 1 )-G( I*KY#K-1)

R= A
INOUjIVUjK))
-9o-
AC
SCC
=
I f
x.)
X
N=(T +
1.
I
-Y*Y)**2+(2.*X*Y)**2
x**X
ft
MT>X*X+Y*Y)/DEN
x x
-2.1 *(T-X*X-Y*Y)/DtN 1.0- + a x * x+ Y x * Y S/A = -xx *</( <)-YX*YZ(K) = -xx *YZ K)+/X*XZ(K)
(
P*Ct> (K
=
)
YC( * + A r an? A*Cl l( PA *Y1) A*SI N( PA *Yi

)
.*X*f, T+X* X-Y*
)/ X'
Z = Y =
A* A H* h*(
r X
*Y* +P
*(
(
X=H* H* (X A*XX+F Y=H* H* (- xx* y Z=H* FZZ* AZ
Z* Z Z*
bZ**Z+XA*XX AZ*AZ+Ya* YX AZ*BZ + AX*YX

) >
+ 1
.-K
.-(5
+1
)
110
F ZZ* 3Z FXY -SX* FXX-SZ = FYY -sx* FX Y-SZ *FYZ = FY? -sx* FX7-SZ * F ZZ = AY- jX* A X-SZ*A Z Al( I )*0G I+C A*X X+SA*YX*P/C4(K)-QMFGA*YX/C4(K) CKK )*u(. <+C A*X Z(K) + SA*(Q + P*YZ(K))/CMK)-0MEGA*YZ(K)/C4{'<) * S A*X A* P/ C4U)- CA*YX-QMEGA*XX/C4 I <Y f liK >=G( I. KY-1>K)+(AX*U+AY*V+AZ*W)/(BY*61(KY)) K) = &{ 12 K YK -G II .*Y,K
120
11 1-1 0. (IV (IK I.Nc.l)

)
13C
140 150
GG TJ 130 EG(K NX + ? -I IKY + 1*K = &( M.KY-1 K )- M,KY + 1K )=G( 1 KY-1 K +t IV iK I.NE.-l) GO FO 14o IKY ,K) = .>*G(I,KY,K-1) + .2:>*(&(I>KY.K+1)+G(M,KY,K + 1>) IV I,K + D.LT.1) G(IKYK).5*G(I*KY*< + l) + .25*(G(I*KY*K-l') + G(Mf ,K-1 MKY ,K) = G( I> KY,K) IKY -1*K )=.;*(G(IKY,K,)+G( I.KY-2, K) MKY -1K ).5* (G(MKYtK)+G( MKY-?K) GJ Tu 120 I. GT.2 NT1N UE TURM
)
-97-
Sl
RC
LV
CI
If
(MN,3F*F?*FPP*FP'P*i<M*\/MKN*\/Nl 5 P L I F SPLINE FIT wITh ÊSCRIBED END CONDITIONS ENSIJN id), F(l) FP(1), FP(1), Ff>PP(i)
L)
T I Nc
I N
IC
(
'
1 e
IN
K =
IF
1C
K =
1
IN-M (K-l) 13G,ie0*10 M-M) /

)
r<
MS
J =
+ K
DS
IF
S(J )-S(I
(DS) 20*180*20 (F(J )-F(I /DS
)
(
D = IS
2C
UF
IF U=
V =
M-2
30*40,50
|V
3C
i.*{UF-V l)/DS
TO
'
GG 40
U=
v
dO
M I!
-
GG
bC
U=
OJ
1.
VGO
DS*V.1 TO 6
60
1*.
J*,
DS
)- S(
)
70
IF DF:
3
*bi
(
1^0,190,70
(
J) -F
CS
(DS +DS+U)
S
U"
v =
b.* DF-V)
=
)
80 FP
FP
=
U
= V
-0) *D5
DF + .*DF+DS*V -is) (J-N) 63*90*60 (KN-2) 100*110*120 N-2 * VN 6.*VN-V) /U
v =
90
IF IF
v =
v=
ICO
110
GO TO
N
13C
GG TG 13c 120 V DS*VN+F.J P( 130 BD=
)/
(l.+FP(
I)
140 DS S( J )-S( I U = PP(I )-FP(
*v
-98FPPP(
F PP
(
I
)
)=
(V-U)/ DS
I
)
FP{ I)>(F(J)-F(
v
=
)/DS-D5*(V+U+U)/6.
J-I
I-I-K j-n if 150 I = N-K

(
>
1^0.1 50>i<o
FPPP(N)*FPPP(I)
FPP (N
)
= B
FP(N)=Di-+u*(FPP(I)
IND-1 IF (MODE)
3+B)/b.
130>180>160
160 FPPP(J) V = FPP( J 170 I=J

J
= rQ:1
)
=J +K
I
DS=S( J)-S( U=FPP( J

)
FPPP(J)=FPPP(I)
V=U
.t>*DS*(F(I)
F(J)-DS*DS*(J + V>/12.)
IF (J-N) 17J180170 180 CONTINUE
RETURN
END
SUBROUTINE INTPL (MI N I#SI t F #M*N*S#F* FP# FPP# FPPP* MODE INTEkPClATIJN OF CUBIC SPLINE BY TAYLCR SEPIES DIMENSION SHI), FI(l), S(l>, F(l), FP(1). FPP(l), i-PPP(l)
1
K=I ABS (M-M) K= (N-M )/K I-H

MIN=M
NIN*NI D=S (N)-S(M) IF (D*(SI (NI )-SI (MI 10 MIN*NI NIN=MI

dO
10,20.20
KI= IABS(NlN-MIIv) IF (KI) <tO^0,30 30 KI (MN-MIN) /K I

<tO
II=M1N-KI
C=
0.
IF
(MODE) 60*60,50
l.
50
C =
-99-
60
7
1 1
1 1
I
+K
(
S S = S
I=I +
IF IF
J =
I
(1-K)
(0*(i>(
tfO*93*8G
I)-SS
70*70* 90
I-I-K SS= SS-S
( I
FPPPP=C*(FPPP(J)-FPPP(I))/(5(J)-S(I))
l) + .25*S5*Ff'PFP F F= F P P P FF=FPP (I )+ iS*FF/2. FF=FP( A )+.i*SS*FF FI( ID =F( I) +SS*FF IF (II-NIN) 60*100*60 IOC CONTINUE RETURN
(
END
subrgu TINE THREED I PLOT* S V* SM* CP* X Y* T IT LE* OC A L* ZONE*F M2,Q EV * lkJO* XS CAL* PSCAL GENERA TEj THPtE DIMENSIONAL CALCOMP PLOT;> ON C n C 6t00
(
)
HO
COMMON G{129*2 6*17)*SC(12917)*E0(17)*IV(129,17)*ITE1(17).ITE2(17) 1*A0(12 *)*Ai(129)*A2(129)*A3ll29)*BC(26)#31(26),62(26)*B3(26)*Z(17)

2 *C 1
(
)*C2(17)*C3(17)*CM17),C5(17)XC(17)*XZ(17)XZZ(17)YC(17)*Y
i
32(17)* rZZ(17)KSYM,NX*r>if,NZ*KTfcl,KTE2,l5YM,SCAL,SCALZ*XK,G v EGA,ALP

<HA C A*
SA*F1ACH*TPA*ABLA0E
Y(l),
DIMENS ION X(l)

N-l
S\M1)*
SM(D* CF(1), TITLE (10), R(20)
IF (XS CAL.Nt .0.) SCALX.5*A8S(XSCAL)/CH0RD0*ZONE DZ=Z(K TE2)-Z(KTE1

)
IF IF
(PA .tQ.l. DZ=1./C5(KTE2)-1./C5(KTE1) (PS CAL.GE.U.) SCAL>=5./DZ

)
SCALP= -1.00 IF (PS CAL.ME.O.) SCALP=-.5/ABS(PSCAL) TX=3.0 SX=-SC AL>*XC UTE1

)
IF
IP LOT. ME. 1)
(
GO TO
10
+
1000*25hANT0INE BOURGEADF CALL P LOTSdL 1C CONTIN UE IPLDT= C CALL F R AME CALL P LOT (1.25*1. -3) ENCODE (65, 190, R) FMACH*F;i2*DEV*AL CALL S YMBOL (0.0*0.75* . 1 R*0.*65) ENCODE (bO*200*R)
<
337wvH)
-130-
CAlL SYrttlQL CONT INUE

20 CONTINUE
K=
l
50,
.25,
14, R, 0. ,60)
30 CONTINUE
K = K + 1
IF
GT
. <
F E
IF
(K.LT.KTE1)
GO GO
TO TO
70
3 3
U-ITEl(K) I2-ITE2U)
CALL
V E L
(K,K,SV$I I>CP,X,Y)
V
SY=5.*(Z(K)-Z(Krfcl))/(Z(<TE2)-?(KTtl))+2.45 SCP = 5.*(Z(K)-Z(KTFl))/<Z(rt2)-Z(KT::l))+2.75

DO 4C 1= 11 12 X I =iC ALX*X
( ) (
I
I
+ SX
Y(
4
SC ALX*Y
+SY
I
)
CP( I)=SCALP*CP CON! I Nut IF (M.EJ.2) GO N=I2-I 1+1
+SCP
50
Tu
CALL LINt
GO TO 30 50 CONTINUE N= I 2-1 1 + 1 DC 60 1 = 11
X
(
(<(ll)CP(Il).N,i,C,2,0.,l.,0.l.)
>
= X
12 TX
f
60 CONTINUE CALL LINf GO TU 2C 70 CONTINUE

M
(X(ll),f(Il),N,l,0,2,0.,l.,C.,l.)
M+
(M.GT 2) GO TO 20
IF
GO
TO
80
80 CONTINUE
CALL FRAME CALL VELC (KTE1,KTE1,SVSM,CPX,Y)
Il=ITEi(KTEl
12=
I
TE2
KTE
SCALX=1.5*XK/3
DO 90
)
14159263
I'll. 12 X( 1 = 2.*X( I) +SCALX4SX + 1. Yd )=2.*Y( I )*SC ALX+ .5 CP (1 = Y( I )+6.

)
90 CONTINUE
N=I2-I1+1
, 1 .i X( I ) Y( I I , N , 1 , U 2 , CALL LINE CALL LINE (X{I1)iCP{I1)Nj1iOi20.iL ENCODE (60. 210, R) CALL SYMBOL (1.0, 3. 5. .14, P.O. ,60) ENCODE (60.220, ) DC 1 4, R , . , 60 1 . 0, 7 . 75 , CALL SYMBOL CALL PLOT (-1 25,-1. -3)
(
.1.
-101-
LX=NX/2+l
KY= N Y+ 1 NA= PA
KD= (Tt2-ME 1+1 )/2*NA + l-NA)*(KTfc2-<TEl+2

(
KD=l"iAX0(
K = K
K.D
i2)
Ttl
1CC C0N1 I N U c CALL FRAKE CALL PLOT DO 110 J = 2

A
/-
3.
KY
/
<
J>,-3)
G=30(J)+bQ(LXK) 0=30(J)+SQ(LXk)
Tr _ v ^ C A=XC(K) + AL3G( ABS
/ *
[I
IF
(0*LJ.LT.T)
GO
.-. t*SCAL*U*D) TO 12C
/XK
110 CONTINUE 120 CONTINUE
0=B0(KY)+SD(2K) B = l. + Au(2)*A0 12 )-0*0 )/T j=SCal*Ao(2)*j B=XC(K)+ALJG(3*B+0*0)/XK/2.

(
A=-2./
B = <t.
/B
A = 6
IF
(A.Lt.C.)
SSX=1.2/Z0NE*XK SSX=AMIN1(SSXA6)
DO 130 I=2>NX A=AO( I LPL0T=3 J=?KY DO 13 0=BO(J) + SO(1M) X1*1.+.5*SCAL* (A*A-0*Q)
)
Y1=SCAL*A*G X2=XC(*<)+AL0G(Xl*xl + riri)/xK/2.

Y2=YC (K )+ATAN2 (Y1>X1 Y3=bIN( Y* )/C5 (K
)
/XK
X2=SiX*X^ Y3=SSX*Y3
CALL PLOT LPLCT = 2 130 CONTINUE
(X2,Y3>LPL3T)
LPLQT=3
140 J=2KY 0=80( J)+SO(LX,
DO
XI
- 1
K
.-.5*SCAL*D*0 X2*XC(K)+ALQG(ABS(X1) /XK )-ATAN2 (0. XI) /XK Y2 = YC Y3SIN( Y2)/C5(K X2=SSX*X2 Y3=SSX*Y3 Y3=SS**Y3 CALL PLOT (X2#Y3*LPL0T) LPL0T=2 l^C CONTINUE DO 160 J=2,KY
)
-102LPLCT=i
CO
loG 1=2 >N< A=AC(

I
D=dO(J)+SO(I<) X1 = 1. + .5*SCAL* Y1=SCAL*A*H

(
A* A-J*0
X2 =
XC
+A LUG
X1*V 1+ f 1
*Y
XK
/ 2
Y2=YC(KJ+ATAN2(Y1X1)/XK Y3=SIN(Y2)/C5(K) X2=S$X*X2 Y3=SSX*Y3

IF
(
NE
L X
3 C
10 15
Y2=2.*YC(K)-Y2 Y^=SSX*SIM( Y2 /C5 (K CALL PLOT (X2>Y4LPL3T) LPLUT=i 150 CONTINUE

)
CALL PLOT LPL0T=2 16C CONTINUE
X2 Y3 LPLuT
Zl*PA*ZJNt*EX?(Z(<))+( i .- D A)*Z(K) ENCODE (bO,230,R) Zl - 1. 0. 4. 5 , 14, , C . , 60 CALL SfMbLL

(
CALL
PluT
(-3.
,-<t
5,-3
TO
K=K+kO IF (K.LE.KTE2)
K = K-1
GO
100
IF
IF
N=2
(K.FC.KTE2)
(PA.NE.1.0)
GO TO
GO
TO
100 160
DOl./XK
CO
S =
170 1=1, FLOAT( 1) /FLJAT(N)
X0=0.5/Z0NE*S
CALL CUT (XO,OC,SCALPj5V,SM,CP,X,Y,Z,YCITE1,1TE2,KTE1,<TE2,KSYMJ 170 CONTINUE 160 CONTINUE RETURN
FOR CM 16HA LP = 2 00 FOR MAT 210 FDR MAT 220 FOP MAT 230 FOR MAT
19
{
5HM1
=
)
F4 .2. 1H 2*
>
5HM2
F 5
H,
2 x
oH DE
,F5.1,H,2X>
)
,F*.l
(21HPR6SSURE DISTR I BUT ION * 5 X , l^HAL A3E prfjFlLt (<:3H CASCADE REPRESENTATION) (9H G/C =,F5.2) =F5.2) I (25H GRID ON THE iUPFACt
END
UBkQuTINE CLT
(XC,DCj;>CALP,SV>:>''i>CP>X,YZ,YC,ITFl,ITE2,l<Ttl,'<TE2j
IK
1
SYM
HIS SlrJKOJTlNE PLOTS 5ECTIl>n3 OF The CD'IPi* t S SO" CMMJN /S ''/ jl(1292fc)G2(l2926)*SX(129)$Z(129),':>XX(129)>iX2U2
1
C
l *
1
>SZZ(129),SJ(129),*1(129)>C(129)>D(1?9)
01, J2,f^
I
, 1 1 ,
P , L X
'i
v
.
L ,
A A C
D
i
MENS
(
S V
C P
) >
X(l),
Yd), HI),
T t i
Y C
ALL
C
1
F k AM
N
[
Z=KTE2-KTE1+1
LO
=
I
I
k = h
TE1
K K
)
)
KTE2
S
C
I
I
ALL
2C =
= 1
.
VELL
T
E-
(K ,K,
CV
X. Y
T Ed
t:
D
1
10
F
= 1 1 3 , I 2
(X(I).LT.XO) GO TJ 20
10
C
1
ONI INGE = I-i
C
F
ONT INUE = x 0-X(

(
1) )/(<( I-l)-X(I) V(K >=B*(Y(I)+F*(Y(I-1)-Y(I) )-YC(K) )+YC(<) (K

)
(dMI>+F*(CP(I-l)-CP<
3
I)
)*3*SCALP-2.
1.
DG 3 M = 11
I F
(
1=11
),
120
)
0+120-1 XlM) .LT.XO
GO
TJ
jG
ONT INUE
0.
( )
B=
*0
ONT INU X 0-X (M) F =

"1(K
D
R
P
)
(X (M + l )-X (A)
= B
(K)
0( K
=
)
*(Y(M)+F*(Y{M+l)-r(M))-YC(K))+YC(K) (C P(f<) + F*CCP(M+l)-CP<M)))*B*SCALP + 2.

) )
= E XP(Z (K )-Z(KTE2 1(K = ? .**0(K)-1.!J 5C C ONT INU E PL J T I . , - 3 c ALL 1 . (R1(KTE1)>CUTE1)NZ>W0, 2jO.il. c ALL L 1 N E (*l(KTEl),0(KrEl),NZ,l,0,2,G.,l. L I N E c ALL = 1 . /DC PL (6. C . -3) C ALL
) ( ,
)
Dg
x
L=L
t
D G
x
K=kTE1<TE2
(K) *3.*f?0(K)*CGS(SV(K) + 2.*3.1415*265*XL*DC)

)
Y (K
3.*R0(K)*SIN(SV(K)+2i*3tl<15926t*XL*0C)
GO
C C
(X(KTE1),Y(KTE1),NZ>10*20.,1.0.1 K=KTE1KTF2 A = 2. *3.14159265*XL*DC X (K) =3.*R0(K)*C0S(SM{K)+A)

D
ONT INUE ALL LINE

7
Y(K)
70
RC
3.*F<C<K)*SIN(Srt(K
+A
CONTINUE CALL LINE (X(<TE1)*Y(KTE1)NZ*1> 0*2*0,

C
L.)
7 1 N
b
I
A3.*F.XP(Z(KTE1)-Z(KTE2)
00
C(
(
90
I
l>97
T = 2.*3.1<*1:j926
5*FL0AT(I-1)/96.
)=3.*C0S( 7 D I)=3.*SIN(T) X( I = 4*CuS( 7 Yd )-A*SlN(TI 90 CONTINUE X 1 CALL L I r IF (KSYM.EQ. 1. CALL LINE (C 1

)
->
1.
1.
) )
GO
C(
1
TQ
100 ),97, 1, 0>2C
,1. .0
IOC CONTINUE
X0=3.*Xu/ A ENCODE (60110*RU XG -2 .1 4 J > CALL SYMBOL CALL PL07 (-7.,-5.-3) RCTUP
(
R i
. ,
60
<l
11C FQR1AT END
K'c-j^
SECTION
IN
THE
PLANE
This report was prepared as an account of Government sponsored work. Neither the United States, nor the Department, nor any person acting on behalf of the Department:
A.
Makes any warranty or representation, expressed or implied, with respect to the accuracy, completeness, or usefulness of the information contained in this report, or that the use of any information, apparatus, method, or process disclosed in this report may not infringe privately owned rights; or
B.
Assumes the use the use method, report.
any liabilities with respect to of, or for damages resulting from

of any information, apparatus or process disclosed in this
As used in the above, "person acting on behalf of the Department" includes any employee or contractor of the Department, or employee of such contractor, to the extent that such employee or contractor of the Department, or employee of such contractor prepares, disseminates, or provides access to, any information pursuant to his employment or contract with the Department, or his employment with such contractor.
c.l NYU DOE/ER 03077-180 Bourgeade An analysis of threedimensional transonic

This book
may be
kept
FOURTEEN DAYS
A
fine will
be charged for each day the book
is
kept overtime.
N.Y.U. Courant Institute of
Mathematical Sciences
251 Mercer
St.
New
York, N. Y.
10012

Analysis of Three D 00 B Our

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Analysis of Three D 00 B Our

Uploaded by

Copyright:

Available Formats

DOE/ER/03077-180

Prepared under Contract DE-AC02-76ER03077 with

Research Center Grant NGR-33-01 6-201.

Research and Development Report

New York University Courant Institute of Mathematical Sciences

Mathematics and Computing

RESULTS AND FIGURES

VIII. LISTING OF THE CODE

AN ANALYSIS OF THREE-DIMENSIONAL TRANSONIC COMPRESSORS

three-dimensional transonic flow around the blades of a

the use of the velocity

potential, on the hypothesis that the flow

such that the surface of the blade is mapped into

This equation is in a nonconservative form

and is solved with

artificial viscosity and artificial time.

code is provided and some sample results are given in

order to demonstrate the influence of three-dimensional effects and the

making better use of the energy

perhaps no longer considered so severe.

many theoretical and experimental

studies aimed at improving their design and efficiency.

transonic flow across a single stage of a turbine, or of

For this reason most

work done so far has focused on

[13-17] or on the mean flow problem [22].

us first discuss the physical background.

certain number of blades distributed around the hub,

the hub is cut along a

generator line and transformed into a plane,

A propeller is a compressor without cowling.

present study deals with the problem of transonic flow around

compressor or propeller blades.

From the mathematical point

system of partial differential equations of mixed

the fluid we are

concerned with is a polytropic and nonviscous gas and that

This necessitates a further hypothesis, namely, that

the variation of the entropy

Thus our system of partial differential

equation, in both the steady and the time-dependent cases.

potential equation has already been solved numerically in the

three-dimensional case for oblique and swept-wings

the periodicity of the compressor problem, the square-root

transformation used in those works is not practical here.

new transformation, which maps the surface of the blade into

The scheme we use is similar

[15] in the development of the swept-wing code known as FL022.

solve the finite-difference approximation to the potential

A typical run consists of 50 iterations on a 48x6x4

followed by 100 iterations on a 96x12x8 refined grid.

minutes on the STAR.

Nevertheless, this method

enables us to study how the speed of rotation influences

we have compared our three-dimensional

results with data for two-dimensional cascades.

we shall derive the equations of motion

we shall consider the potential equation,

which is obtained after several changes

in Section 4. The numerical results obtained are

together with some Calcomp plots,

on the use of our computer code.