DTIC ADA346161: Global Wall Interference Correction and Control for the National Wind Tunnel Complex Transonic Test Section,

17 Pages·1.1 MB·English

by  Defense Technical Information Center

4
Mm ml Mm Mm
AIAA 97-0095
Global Wall Interference Correction
and Control for the NWTC Transonic
Test Section
W. L Sickles and F. W. Steinle Jr.
Sverdrup Technology, Inc., AEDC Group
Arnold Engineering Development Center
Arnold Air Force Base, Tennessee 37389
irxnuunviu \ »^
35th Aerospace Sciences
Meeting and Exhibit
January 6-10, 1997 / Reno, NV
For permission to copy or republlsh, contact the American Institute of Aeronautics and Astronautics
1801 Alexander Bell Drive, Suite 500, Reston, VA 22091
DBimwmON STATE«f-^
QUALrf? msFRmm ft
Approved for public release;
Distribution Unlimited
Global Wall Interference Correction and Control
for the NWTC Transonic Test Section*
W. L Sickles** and F. W. Steinte Jr.*
Sverdrup Technology, Inc., AEDC Group
Arnold Engineering Development Center
Arnold Air Force Base, TN 37389-6001
Abstract c Pitching-moment coefficient
m
c Normal-force coefficient
As part of the currently postponed National N
Wind Tunnel Complex (NWTC) program, a compu- C Pressure coefficient
P
tational fluid dynamic (CFD) study of a large trans-
M Mach number
port configuration was performed using the AEDC
chimera-overset grid XAIR Euler code. Wall poros-
R Wall resistance [Eq. (2)]
ity, porosity control for the reduction of wall interfer-
ence, correction approach, and the impact of refer- a Angle of attack
ence point selection on wall interference at tran-
AC C - C
A Aoo AT
sonic conditions (Mach Number 0.85) were studied.
These effects were computationally investigated ACD CD~ - CDT
using a clean-wing model representative of the MD-
11 aircraft which spanned 80 percent of the test ACL CL~ - CLT
section width. Global wall corrections were deter- AC Cunoo - C x
m m
mined by a streamtube calculation of the interfer-
ence flow field. Results of these computations ACN CN°° _ CNT
showed that global corrections for Mach number at AC C -C [Eq.(1)]
P Poo PT
the 1/4 chord, 50-percent semi-span location with
e flow angle
angle-of-attack correction based on matching wing
lift gave very low interference. Investigation of con- T wall open area ratio, percent
trol of porosity showed that very few zones were
required to achieve superior results for global cor- Subscripts
rections and for model pressure distribution. Resid-
T tunnel condition
ual blockage gradient and flow-curvature effects
are shown to be treatable by very small changes in
oo free-air condition
side-wall divergence angle and model geometric
changes that are equivalent to shearing the coordi- 1.0 Introduction
nate system aft of the wing-fuselage juncture.
The facilities for the National Wind Tunnel Com-
Nomenclature plex (NWTC) were being designed with the expec-
tation that the maximum design Reynolds number
C Axial-force coefficient would be achieved by testing models with spans
A
up to 80 percent of the applicable test-section
CQ Drag coefficient
dimension. A consequence of the relatively large
C Lift coefficient model-to-tunnel size ratio is that interference pro-
L
* The research reported herein was performed by the Arnold Engineering Development Center (AEDC), Air Force Materiel
Command. Work and analysis for this research were performed by personnel of Sverdrup Technology, Inc., AEDC Group, technical
services contractor for AEDC. Further reproduction is authorized to satisfy needs of the U. S. Government.
** Senior Member, AIAA.
Approved for public release; distribution unlimited.
•Associate Fellow, AIAA.
This paper is declared a work of the U. S. government and DUG QUALITF IM0?3ICj^D ö
not subject to copyright protection in the United States. .,
American Institute of Aeronautics and Astronautics
duced by the wind tunnel walls may be unaccept- All computations were solved for flows about a
ably large. The 80 percent span ratio is believed to generic representation of an MD-11 clean-wing
be the largest value that can be tested with any configuration. All in-tunnel calculations were per-
reasonable chance of obtaining high-quality data. formed for a model that is scaled to span 80 per-
Data quality was of paramount importance for the cent of the tunnel width. The test-section geometry
NWTC; therefore, wall correction and wall control of this tunnel was 16 ft wide by 13 ft high with no
strategies were being investigated as an integral corner fillets. The scaled model was 13.4 ft in
part of the development process. The preliminary length, and the model nose was positioned on cen-
criteria for wall interference at Mach number 0.85 terline 16.3 ft downstream of the test-section
were established for the uncorrected levels of entrance. The test section was 42 ft long. The
blockage (AM < 0.003), angle of attack (Act < 0.15 scaled full-span configuration resulted in a solid
deg at C = 1.0), force coefficients (AC < 0.005 blockage ratio of approximately 0.60 percent.
N N
and AC < 0.0003), and pitching-moment coeffi-
A
For all flow conditions under consideration, lat-
cient corrections (AC < 0.01). The intent of these
m
eral symmetry was assumed about the center plane
criteria was to have the level of wall interference
of the model. Therefore, only half of the flow field
sufficiently low that reliable corrections could be
was modeled. Figure 1 shows a portion of the near-
made to the data to support meeting the overall
field grid system for the MD-11. The entire system
uncertainty in the measurement of one drag count,
contains nine grids with 1.1 x 106 grid points.
(C = 0.0001). The purpose of the CFD evalua-
D
tions of wall interference was to support validation
3.0 Wall Interference Procedure
of both the reasonableness of the interference and
the development of the design features of the test In the current investigation, wall interference
section. The current paper summarizes the results effects were predicted by two different methods,
obtained and documented in Ref. 1. constant conditions and global corrections.
2.0 Computational Procedures 3.1 Method 1: Constant Conditions
Wall interference estimates were based on In the constant conditions method, the wall
Euler CFD flow-field solutions which were per- interference effects are predicted at a constant
formed using the AEDC chimera-overset grid XAIR geometric angle of attack and upstream Mach
code.2"4 number by taking the difference between two CFD
wmsHHHifflnBMlBrnRaF
Fig. 1. Chimera inviscid grid system.
American Institute of Aeronautics and Astronautics
flow-field solutions. The first solution is a free-air ture effects of the tunnel walls by determining
flow-field calculation where the outer boundary of adjusted remote flow conditions. The objective of
the computational domain extends well beyond the this procedure is to find equivalent free-air flow con-
location of the tunnel walls. At the outer boundary ditions that best match the tunnel data or, in this
of this domain, free-stream conditions are pre- case, the in-tunnel calculations. The adjusted,
scribed. The upstream and downstream outer equivalent conditions are typically determined by
boundaries are approximately four body lengths selecting a blockage correction (AM) and flow-
upstream and two body lengths downstream of the angle correction (Aa) to the free-air remote condi-
model, respectively. The vertical and horizontal tions which force the free-air and in-tunnel calcula-
planes of the outer boundary are approximately tions to match at a given reference point (or points)
five tunnel heights from the centerline axis of the on the model. In the present context, this procedure
model. The second solution is an in-tunnei involves a minimum of two additional calculations.
flow-field calculation at the same flow conditions
where a wall boundary condition (Sec. 4.0) is pre- The first additional flow-field calculation deter-
scribed at the tunnel walls. Here, the flow at the mines the adjustments to angle of attack and Mach
test section entrance is prescribed as uniform free- number (AM and Aa). This is done by performing
stream flow, while the flow at the test section exit is an inverse computation of the interference flow
specified by an extrapolated boundary condition. A field. The inverse computation involves a tunnel-
pair of flow-field solutions must be computed and empty flow-field or streamtube solution with the
compared to determine the wall interference at interference pressure distribution specified as the
each selected flow condition. To minimize any pos- boundary condition at the tunnel boundaries. This
sible difference due to grid resolution, meshes interference pressure distribution is determined by
internal to the tunnel are identical. The local effects subtracting the free-air pressure distribution from
of wall interference on the model are examined by the tunnel pressure distribution, both of which are
subtracting the calculated local pressure coefficient calculated at the tunnel walls from the aforemen-
for the in-tunnel solution(C x) from the corre- tioned free-air and in-tunnel calculations. The
p
sponding free-air solution(C ) as resulting solution of the inverse flow field is interpo-
poo
lated onto the phantom model surface to provide
ACP - CPoo-CPT (1) disturbances (AM and Aa) at selected reference
points. A complete parametric analysis of the
and the global effects on forces and moments selection of the reference point location is beyond
directly follow by pressure integration of the individ- the scope of this investigation. Typically, the flow
ual solutions and subtracting the result. corrections are determined at a given spanwise
location. At the given spanwise location, a different
It is important to note that the wall interference reference point is selected for determining AM and
procedure outlined above determines corrections Aa. The reference points are selected as the 1/4-
to the model forces and moments at the fixed tun- chord location for AM and the 3/4-chord location for
nel free-stream flow conditions and at a fixed geo- Aa. The selection of these chordwise locations is
metric angle of attack. The result from this compu- based on classical theory.5 The 3/4 chord for Aa is
tation can be compared to the NWTC requirement the classical point of choice to minimize pitching-
for acceptable level of wall interference. Difference moment residuals. Transonic flow features may
in viscous flow development obviously will affect require some other choice. Two spanwise loca-
the results. However, the method could include an tions, 25-percent and 50-percent semispan, were
integral boundary-layer technique to make a vis- selected to investigate different correction strate-
cous correction gies. Initial computations were performed using the
corrections that were determined from both span-
3.2 Method 2: Global Corrections wise locations. As will be shown in the results, the
50-percent location yielded superior blockage cor-
A more widely used wall interference procedure
rections and was selected as the location of choice
is to globally correct for blockage and flow-curva-
for the remaining computations.
American Institute of Aeronautics and Astronautics
The second additional flow-field calculation is was a reasonable assumption for this study. A
performed to determine the free-air flow field at the boundary condition was selected that has been
adjusted remote conditions (M + AM and a + Aa). shown to be representative of the wall behavior in
Using Eq. (1), the residual interference is deter- the NASA/ARC 11-ft Tunnel and was used for in-
mined by comparing the free-air flow-field calcula- tunnel calculations. For this boundary condition,
tion at adjusted conditions and the in-tunnel calcu- the cross-flow behavior at the wall is prescribed as
lation at the original conditions (appropriately scal- dC„ _2_
ing load and pressure coefficients for the difference (2)
de TR
in dynamic pressure and adjusting stability-axis
where R is the wall resistance or porosity parame-
forces for differences in angle of attack).
ter, x is wall porosity or wall open area ratio, and 0
is the local flow angle at the wall. For all in-tunnel
The process to determine the optimum remote
flow-field calculations, R = 19. To simulate test sec-
conditions is an iterative procedure where the
inverse and free-air calculations are repeated until tion walls with global uniform porosity, the value of
x was constant over all walls. To simulate variable
the residual interference is minimized. In the cur-
porosity, % was varied locally on all walls. This
rent investigations, only one inverse computation
was required to obtain an accurate blockage or boundary condition was incorporated into the
AEDC XAIR/chimera code and implemented by two
Mach number correction. Additional adjustments
methods. The first method calculates 8 at each iter-
were made to the remote angle of attack to best
ation in the flow-field solution, determines the wall
match the wing lift between the in-tunnel and free-
pressure from Eq. (2), and then updates and
air calculations. This is similar to the technique
used by Rizk.6 The lift matching process typically imposes the internal energy as the boundary condi-
took one additional free-air computation. However, tion (along with extrapolated values of density and
because the interference on the model is spatially momentum). The second method calculates Cp,
determines 0 from the flow-field solution, and
variant, residual wall interference still exists in the
updates and specifies the appropriate component
lift contribution from the other model components
and in the other force and moment coefficients at of momentum as the boundary condition. The sec-
these adjusted flow conditions. The residual is pre- ond method was utilized when variable porosity
sumed to be significantly less than the total correc- was investigated, while the first method was used
tion determined for the original, unadjusted flow for all uniform porosity computations.
conditions.
5.0 Results
One advantage of this procedure is that differ-
ences in pressure gradient typically are much less The in-tunnel and free-air flow-field calculations
than for Method 1. Hence, incremental corrections about the model were performed on the AEDC
to account for viscous flow development differ- Convex 3800. These inviscid computations typi-
ences are expected to be substantially smaller. cally required 12 CPU hours to converge. Steady-
state convergence was considered met when all
4.0 Wall Boundary Conditions calculated model forces and moments coefficients
remained unchanged to five significant digits.
The porous slotted wall behavior was simulated
by assuming and imposing a linear-homogeneous 5.1 Wall-Interference Computations at
boundary condition representation at the test sec- Constant Conditions (Method 1)
tion walls. Provided that a slotted wall has suffi-
cient slots, the local effect of the discrete slots rap- Baseline wall interference computations for
idly diminishes, and the behavior of the wall is several wall configurations and flow conditions
essentially uniform and continuous at a short dis- were performed. This task involved making a num-
tance from walls.7 The NWTC slotted wall configu- ber of in-tunnel flow-field calculations at different
ration was designed to have sufficient slots and to globally uniform, fixed values of wall porosity for
behave in this manner. Therefore, homogeneity transonic speed conditions and the corresponding
American Institute of Aeronautics and Astronautics
free-air flow-field calculations. The purpose of pressure distributions for a = 4 deg are shown in
these calculations was to quantify the level of wall Figs. 2-4. The local pressure distributions show a
interference that can be expected for walls with significant longitudinal and spanwise variation in
globally uniform porosity. All calculations were per- the increment between the free-air and in-tunnel
formed, as well as the determination of wall inter- calculations. The wing-pressure differences
ference increments, at the fixed transonic-speed between the in-tunnel and free-air calculations
conditions of M„, = 0.85, a = 0 and 4 deg. In addi- appear to be minimized at 4 percent (Fig. 2). As the
tion, in-tunnel computations were made at discrete walls are opened for the transonic conditions, the
porosities of x = 4 , 6 , and 9 percent. effect on the model wing pressure distribution can
be seen. The pressure level increases, the shock
Table 1 shows the baseline computations, the strength decreases, and the shock moves forward.
flow conditions, and the resulting forces and
moments wall interference increments using Figure 5 summarizes the calculated wall inter-
Method 1. The comparisons between the in-tunnel ference force and moment increments for all the
calculations and the corresponding free-air model inviscid baseline calculations and shows the varia-
Table 1. Wall Interference Increments for tion of those increments with porosity. The base-
Baseline Calculations (Constant line calculations show that there is a significant
Conditions Method) variation in force and pitching-moment increments
with uniform porosity and that wail interference var-
M„ MT deg d«eTg. percent ACD ACL ACm ies significantly with model attitude. Because of the
0.85 0.85 0 0 4 -0.0015 0.0017 0.0090 gradient of wall interference over the model, mini-
0.85 0.85 0 0 6 -0.0011 0.0107 -0.0010 mizing all forces and moments simultaneously can-
0.85 0.85 0 0 9 -0.0007 0.0178 -0.0094 not be accomplished with a uniform porosity. With
0.85 0.85 4 4 4 -0.0012 0.0021 0.0181
uniform porosity and no corrections for blockage or
0.85 0.85 4 4 6 0.0012 0.0164 0.0041
0.85 0.85 4 4 9 0.0031 0.0286 -0.0091
o
o
ü
Ü< -0.04 YT. 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 < -0.04 0.1 0.2 0.3 0.4 0.5 6 0.7 0.8 0.9 1.0
fuselage centerline
fuselage centerline -0.08
-0.08
X/CorX/L XA//iC* ourr XA//lL_
Fig. 2. Model pressures at fixed remote condi- Fig. 3. Model pressures at fixed remote condi-
tions, M = 0.85, a = 4 deg, x = 4 percent, tions, M = 0.85, a = 4 deg, x = 6 percent,
globally uniform. globally uniform.
American Institute of Aeronautics and Astronautics
flow curvature, the interference increments do not
meet the NWTC wall interference criteria for tran-
sonic flow conditions.
Controllable walls were considered as a
method to meet the stringent NWTC data quality
specifications. As seen in the baseline computa-
tions, uniform porosity does not provide the desired
data quality for transonic flows. Variable porosity
was investigated to show the possible benefits of
controllable walls in meeting these data quality
specifications. Two variable-porosity methods
were computationally investigated using the linear-
.2 ■ /'"»■—■- . ■ ■ ■ ■
homogeneous boundary conditions. In the first
.8 I Free-Air
Tunnel — method, it was assumed that there is an infinite
number of control segments over the walls. With
this method, the porosity varies locally over the test
section walls to best match the local wall behavior
determined from the free-air computation. This
. ,.—^-. ,a i— i r 'i method will be referred to as distributed porosity
< -0.04 • 0.1 0.2 0.3 0.4 oSnpB 0.7 0.8 0.9 1.0 control and represents the absolute best that can
fuselagei centerline l j
-0.08
be done with variable porosity alone. In the second
X/CorX/L
method, limited control was imposed. The wall was
Fig. 4. Model pressures at fixed remote condi-
segmented, and the porosity was fixed laterally or
tions, M = 0.85, a = 4 deg, x = 9 percent,
vertically over the patch and allowed to vary lin-
globally uniform.
early in the longitudinal direction. The longitudinal
variation was constrained to be piecewise continu-
0.032 0.032
-•— uniform x, 0°, M = 0.85 —•— uniform %, 0°, M = 0.85
0.028 e» dseisgtmribeuntteedd Tx,, 00°°,, MM == 00..8855 0.028 o• dseisgtmribeuntteedd Tx,, 00°°,,M M == 00..8855 T = 9% ■
T = 9% -*- uniform x, 4°, M = 0.85 —■- uniform x, 4°, M = 0.85
0.024 □D dseisgtmribeuntteedd Tx,, 44°°,, MM == 00..8855 0.024 Qa dseisgtmribeuntteedd tx,, 44°°,, MM == 00..8855
/
0.020 0.020 ■
7I / x = $ %f
0.016 0.016 ■
"o5 0.012 o 0.012 ■
I 8
_i 0.008 51 0.008 ■
o
tj y
^ 0.004 0.004
T = 4% /i "BO /
0 0 .T = 4% •
e
-0.004 -0.004 ■ ■
-0.008 -0.008 ■ ■
-0.012 -0.012 ■
-0.016 -0.016
-0.012 -0.008 -0.004 0 0.004 0.008 0.012 0.016 0.020 -0.0016 -0.0008 0 0.0008 0.0016 0.0024 0.0032
ACm(cm~-('mT) ACD(CD--CDT)
Fig. 5. Wall interference increments at fixed remote conditions.
6
American Institute of Aeronautics and Astronautics
ous between patches. This method will be referred
to as segmented porosity control and represents a
more realistic implementation of variable wall con-
trol. Both of these were utilized to investigate vari-
able porosity for the two transonic flow cases.
From the free-air baseline calculations, ideal
values of the local porosity for the transonic cases
were determined. This was accomplished by inter-
polating the free-air flow field onto a surface equiv-
alent to the tunnel boundary and then calculating
the local pressure and flow angle distributions to
determine the ideal local cross-flow properties.
Using Eq. (2), the local values of porosity were
determined. Practical considerations (e.g., the
porosity cannot be negative) made it necessary to
constrain the porosity such that 10% > x > 0% (solid
wall). This distribution of porosity was imposed as
the boundary conditions and the in-tunnel flow-field
solutions were calculated. The results of these cal-
culations were compared to the corresponding free-
< -0.04 -P 0-1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
air solutions. The force increments are shown in -0.08 fuselage centerline
Table 2. The wall interference is significantly X/CorX/L
reduced over the uniform porosity baseline cases, Fig. 6. Model pressures at fixed remote condi-
as can be seen in Fig. 5. The model pressure com- tions, M = 0.85, a = 4 deg, distributed
parisons for a = 4 deg are shown in Fig. 6. The porosity.
results from these calculations show a significant
reduction in wall interference by incorporating vari-
SP 0.50
able porosity control. Figure 7 shows the corre- •a Free-Air •
0.25 T=6% •
sponding wall cross-flow distribution (flow angle vs. Distributed T. ' L
0 ■
pressure). The wall cross-flow distribution for the -0.25 ■ lower wall
distributed porosity cases and the free air shows 2 -0.50
-0.06 -0.04 -0.02 0.02 0.04
good agreement except for a slight overshoot in
pressure coefficient at the model center. This over-
shoot is caused by an acceleration of the flow from
the upstream solid wall patch. The results of the dis-
tributed porosity computations show a significant
benefit to utilizing variable porosity.
A segmented controllable wall was investigated
by developing a patchwork of controllable seg-
-1.25 upper wall
Table 2. Wall Interference Increments for Variable
-1.50
Porosity Calculations (Constant Correc- -0.06 -0.04 -0.02 0 0.02 0.04
Cp
tions Method)
g> 0.50
Free-Air
tf °-25
M„„ *h dae«g, d«eTg. perXc, ent ACD ACL ACm <ra o ■ Distributed T " • >*M■ &~~2 f
0.85 0.85 0 0 dist -0.0008 -0.0010 0.0038 -0.25 side wall
g -0.50
0.85 0.85 0 0 seg -0.0008 0.0018 0.0023 -0.06 -0.04 -0.02 0.02 0.04
0.85 0.85 4 4 dist -0.0010 0.0019 0.0067
0.85 0.85 4 4 seg -0.0008 0.0040 0.0064
Fig. 7. Wall signature comparison at M = 0.85, a =
dist = distributed porosity variation
4 deg, distributed porosity.
seg = segmented porosity variation
American Institute of Aeronautics and Astronautics
merits. The goal was to simulate a wall where the 5.2 Wall Interference Computations with Global
flow through the slots is throttled by using movable, Corrections (Method 2)
plenum-side cut-off plates. The cut-off plates would
For selected uniform and segmented porosity
allow the porosity to be linearly varied in the flow
calculations, adjustments to free-airflow-field angle
direction. In addition, the goal was to develop a seg-
of attack and Mach number were determined to find
mented wall that could be feasibly implemented.
equivalent free-air conditions that best match the
The patchwork was developed by examining the
wall distributions and decid-
ing where the most and
z(in.) from
least wall control is needed. tunnel centerline
The most control is needed
in regions of highest gradi-
ents, e.g., model wing,
while the least control is in
the regions of minimum gra-
dients, e.g., upstream and
downstream. Figure 8
shows the patchwork used
for the computations. This
configuration shows six .100
80
.60
streamwise-controllable ^*0
i 2°
wseaclltsio annsd o onn teh eo nto tph ea nbdo tstoidme -300 -250 ■20« -150 .100' ^ 40 tunyn(einl .c) efrnotemr line
wall. Implementation of this modexl (riont.)a tfiroonm c enter 50 100 150 200 250 --1 0-08 0
configuration would
Fig. 8. Porosity segmentation for segmented porosity runs.
require six controllable
segments along all or selected slots on the top and
side walls and one control segment for all or
selected bottom wall slots.
The longitudinal porosity variation was deter-
mined by anchoring the porosity at the upstream
and downstream ends to be equal to the interpo-
lated free-air porosity and linearly interpolating over
the length of the segment. This constrains the dis-
tribution to be piecewise continuous between seg-
ments. The wall interference results of these calcu-
lations are also shown in Table 2 as well as in Fig
5. The wall interference increments are comparable
to the ideal distributed porosity results. The model
pressure comparison in Fig. 9 and corresponding
wall signature in Fig. 10 are almost identical to the
distributed porosity results. These results show that "Upper distributed x
limited segmented control can reduce wall interfer- lower distributed x
upper segmented x 75% wing semispan
ence as well as infinite control, provided the seg- 0.04 lower segmented x
o 4- ^\
mentation is strategically placed. Additional work
-0.04 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
would be required to determine the optimum place- fuselage centerline
-0.08
ment and minimum number of control segments, »CorX/L
the solution to the pressure overshoot in the wall Fig. 9. Model pressures at fixed remote conditions,
signature, and an active wall control strategy. M = 0.85, a = 4 deg, segmented porosity.
8
American Institute of Aeronautics and Astronautics
in-tunnel calculations. The purpose of these com- the procedure described in Sec. 3.0. Five cases
putations was to determine the residual interfer- were selected and are shown in Table 3. These
ence after the computations have been corrected cases include two porosity distributions at a = 0 deg
for first-order blockage and flow-curvature effects. (a uniform porosity, t = 6 percent, and the seg-
The flow-field adjustments were determined using mented porosity distribution) and three porosity dis-
tributions at a = 4 deg (two uniform porosities, x = 4
2> 0.50
■D« 0.25 FrTe e= -6A%ir ■• and 6 percent, and the segmented distribution
°> 0 Segmented x ■ porosity).
* -0.25 lower wall \
-0.50'— Figure 11 shows representative streamtube
-0.06 ■0.04 -0.02 0.02 0.04
results for the M = 0.85, a = 4 deg, and % = 6 per-
0.50 cent case interpolated onto a phantom surface of
Free-Air —
0.25 Segmented T — the MD-11 wind/body/horizontal tail configuration.
0 Shown are the Mach number and flow curvature
-0.25 effects of the walls. The contour plots show a sig-
g> -0.50 nificant spanwise and longitudinal gradient in the
<
S -0.75 Table 3. Wall Interference Increments at Adjusted Free-
-1.00 Stream Conditions (Global Corrections Method)
-1.25 upper wall Ref.
-1.5-00.'0—6 -0.04 -0.02 —I 0.■0 2,„ ,1M 0.04 Moo MT a«,, deg d<e*Tg. percent ACD ACL ACm peLrocce.,n t
0.84800 0.85 -0.1000 0 6 -0.0011 0.0108 -0.0011 25
0.50 0.84900 0.85 -0.1000 0 6 -0.0010 -0.0023 0.0084 50
Free-Air —
0.25 T = 6% 0.85048 0.85 -0.0294 0 seg -0.0007 -0.0006 0.0036 50
Segmented-c —
0 0.84700 0.85 3.8300 4 6 -O.0027 -0.0103 0.0235 25
< -0.25 side wall 0.84850 0.85 3.8400 4 6 -O.0014 -0.0041 0.0177 50
.2 -0.50 0.84940 0.85 3.9182 4 4 -0.0023 -0.0075 0.0243 50
"- -0.06 -0.04 -0.02 0.02 0.04 0.85096 0.85 3.9320 4 seg -0.0006 0.0004 0.0074 50
Tunnel forces and moments adjusted to the dynamic pressure and angle
Fig. 10. Wall signature comparison at M = of attack of the free-air conditions,
0.85, a = 4 deg, segmented porosity. seg = segmented porosity variation
a. Blockage b. Flow angle (radians)
Fig. 11. Interference flow-field solution.
9
American Institute of Aeronautics and Astronautics