51st AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition

07 - 10 January 2013, Grapevine (Dallas/Ft. Worth Region), Texas

AIAA 2013-1080

Eciency of an Inertial Particle Separator

Dominic Barone, Jane Hawkins, and Eric Loth
Department of Mechanical and Aerospace Engineering, University of Virginia, Charlottesville, VA

Philip Snyder

Downloaded by UNIVERSITY OF MINNESOTA on October 10, 2013 | http://arc.aiaa.org | DOI: 10.2514/6.2013-1080

Rolls-Royce North American Technologies, Inc., Indianapolis, IN

An Inertial Particle Separator (IPS) is one method of reducing the amount of sand and
dust ingested by gas turbine engines. This is important because sand and dust often limit
the useful life of engines operating operating in austere environments. In order to further
improve the eciency of IPS systems, an in depth study of the multiphase ow dynamics
has been undertaken. An experimental method was chosen to ll the void of available data
and provide useful information for model validation. Using a facility at the University of
Virginia, experiments have been conducted using three dierent IPS geometries. Surface
ow visualization of each geometry is conducted using an oil-streak method to identify key
ow structures. Separation eciencies were calculated for a A4 Coarse, Arizona Test Dust
(ISO 12103-1) over a range of operating conditions for each geometry.

Nomenclature
Symbols

ow split

change

uncertainty
m

mass ow

eciency

dynamic viscoscity

density
domain response time
D
w

relative velocity vector
added mass coecient
c
drag coecient
CD
D
characteristic domain length
d
diameter
f
Stokes drag correction factor
drag force
FD
m
mass
domain Stokes Number
StD
u
ow velocity

Corresponding Author: PhD Candidate, Member AIAA, 122 Engineers Way, P.O. Box 400746 Charlottesville, VA 22904,
phone: (434) 218-2271, email: barone@virginia.edu
Student, Member AIAA
Professor, Associate Fellow AIAA
Rolls-Royce Associate Engineering Fellow / Consultant Engineer, Member AIAA

1 of 10
Copyright 2013 by Dominic Barone. Published by the American Institute of Aeronautics and Astronautics, Inc., with permission.

American Institute of Aeronautics and Astronautics

I.

Introduction

hen a turbine engine ingests sand or dust, the components of the engine may undergo highly damaging
W
erosion, and the engine life can be signicantly reduced . This presents a major readiness, cost, and
maintenance issue for military and civilian missions during low altitude operations in sandy areas such as

the Middle East 4 . Sand and dust separation devices traditionally fall into the three categories: Vortex Tube
Separators (VTS), Inlet Barrier Filters (IBF), and Inertial Particle Separators (IPS). In contrast to VTS and
IBF systems that are normally applied as airframe installed options, IPS systems are currently designed as
integral to the engine. In many cases an integrated design can result in a more compact system with lower
total pressure losses.
The basic features of an approximately axisymmetric IPS design 6 are shown in Fig. 1. Particle-laden
air enters the annular inlet and is bifurcated into two ow paths, each of which is then bifurcated by a set
of splitters, with the main core ow rapidly turning while the scavenge ow continues in an approximately
axial direction. In an installed IPS system, the mass ow of air to the core ow is driven by the turboshaft
engine and the split to the scavenge ow is driven by some type of blower or ejector system.

Scavenge

To E
n

gine

Scavenge
(a) IPS patent gure 6

(b) Dual IPS ow path adapted from Snyder 6

Outer Surface Geometry (OSG)

Scavenge Flow

Co

Inflow
Hub Geometry

low

F
re

Downloaded by UNIVERSITY OF MINNESOTA on October 10, 2013 | http://arc.aiaa.org | DOI: 10.2514/6.2013-1080

15

Splitter Geometry

(c) Single path geometry and nomenclature

Figure 1. Inertial particle separator geometry

A simplied schematic that identies the outer set of core and scavenge paths (after the rst bifurcation)
is shown in Fig. 1(c). Within the separator, the airow makes a hub-side turn while the scavenge ow
continues with very little ow turning. While very small drag-dominated particles will tend to follow the
streamlines, moderate size particles with sucient inertia follow the scavenge path which prevents them
from following the core airow around this sharp corner. Large particles that are inertia-dominated will only
be weakly eected by the ow and will be controlled more by their initial trajectories and any bouncing

2 of 10
American Institute of Aeronautics and Astronautics

Recirculation Zone

Figure 2. Qualitative Mach contours and streamlines to illustrate the recirculation zone present in the scavenge
ow.

The particular region of interest being addressed here is a recirculation zone that exists along the outer
surface of the ow path near the splitter, Fig 2. It is believed that this ow feature has a signicant
impact on separator eciency; however, the details of the ow structure are poorly resolved by averaged
ow solution techniques. The direct application of full three dimensional direct numerical ow simulation
for the geometries is highly resource intensive and often out of scope for application to the problem. While
computational simulations can allow quantitative results, such unsteady three-dimensional turbulent ow
solutions performed for an IPS geometry are computationally expensive.
Similarly, the particle dynamics of an irregularlyshaped polydisperse mixture that includes signicant wall
1
reections are not well modelled. Thus, prediction of sep0.9
aration eciency and an understanding of the controlling
0.8
ow phenomena inuencing the particulate motion asso0.7
ciated with the IPS are only denitively obtained with an
0.6
13
experimental approach. As expressed by Jiang et al. ,
0.5
experimental data is needed to validate the computational
0.4
models.
0.3
To date, previous test facilities have only been deMcAnnaly14
0.2
signed to quantify the overall performance of IPS deDuffy15
Breitman16
signs 1417 , Fig. 3. Duy 15 in particular did an extensive
0.1
Vittal17
survey of three dierent concepts over a wide range of
0
0.05
0.1
0.15
0.2
0.25
0.3
operating conditions. These studies have looked at the

overall eciency and aerodynamics of the particle separators with regard to installed performance, which can be Figure 3. Previous IPS eciency data relevant to
used to verify the end result of a model, but provides little current IPS research
data for model validation.
The goal of this work is to begin the work required to understand the ow physics present in an IPS and

Downloaded by UNIVERSITY OF MINNESOTA on October 10, 2013 | http://arc.aiaa.org | DOI: 10.2514/6.2013-1080

o of solid surfaces. In general, the hub, splitter, and Outer Surface Geometry (OSG) are ideally designed
to ensure that most of the particles enter the scavenge ow leg, so that the core ow is clean of damaging
particulate. The scavenged particles are then ejected from the aircraft via the scavenge ow and the cleaned
air proceeds to the turbine engine.
For design of a high eciency low pressure loss IPS, a fundamental understanding of the multiphase
ow physics is required; particularly, the relationship of the airow bifurcation characteristic eects on dust
particles. Understanding of both the uid ow eld and the particulate motion in the splitter region is
critical to achieve designs that maximize particle separation eciency. A signicant amount of eort has
gone in to the computational studies 5,713 . While these studies have identied features like the recirculation
zone using solutions to the Reynolds Averaged Navier-Stokes equations, they have focused primarily on the
average ow eld and do not account for dynamic ow structures and the eect that these structures have
on particles.

3 of 10
American Institute of Aeronautics and Astronautics

particularly that of the recirculation zone. Oil-streak visualization is employed to qualify the position of the
recirculation zone along the outer surface geometries. Eciency measurements of the IPS geometries are
compared to show the eect of the recirculation zone. This work will begin to ll the void of experimental
data with a simple well dened geometries with experiments over a range of conditions.

Methodology

A facility was designed to create the specic branching ow conditions present in an IPS system, a detailed
description can be found in previous work 1 . The system is capable of several measurements, including overall
sand separation eciency, surface ow visualization, and quantitative optical internal ow diagnostics. This
work will focus on separation eciency and surface ow visualization; optical measurements will follow with
planned experiments to quantify the uid ow eld and particle paths through the IPS. A schematic of the
tunnel is shown in Fig. 4. The main components of the tunnel (in ow order) are 1) the particle deliver
system (PDS), 2) the tunnel intake, 3) the IPS test section, 4) core and scavenge ow diusers, 5) core and
scavenge particle lters, and 6) core and scavenge centrifugal blowers (not shown).
The facility has the ability to independently control the mass ow rate through the scavenge and
core ow paths using two independently controlled
'46.)0+
+0.8+4:

centrifugal blowers. Each blower is controlled by a
":56+1
"
variable frequency drive, allowing ease of control of
mass ow split between the core ow and scavenge

ow. The scavenge ow split, , is dened as the

ratio of scavenge air to the total amount of air that
enters the inlet, Eq. (1).


m
s
=
m
c+m
s

26'/+

(1)



34+
.,,75+4
2-.2+
.4,039


The mass ow rate through the system is set using
"
#+56

 
a closed loop proportional-integral-derivative (PID)
"+)6.32
controller. This controller sets a constant mass ow

rate through the core ow leg, then adjusts the scav")'8+2-+ 

P
enge ow to achieve the desired ow split.
.,,75+4
The facility features a modular test section capable of performing measurements on dierent IPS
P
geometries. Three dierent OSGs have been created
34+
.06+4
")'8+2-+
to test the eects of the recirculation zone on sep.06+4
aration eciency. The three geometries are shown
in Fig. 5, a detailed table of characteristic points is
given in the appendix. The test section is 8 inches
wide and is designed for Mach 0.4 ow at the throat Figure 4. Facility schematic, dimensions are in inches
(x 6 inches). These three OSGs have been selected to produce dierences in the location of the recirculation zone.












Downloaded by UNIVERSITY OF MINNESOTA on October 10, 2013 | http://arc.aiaa.org | DOI: 10.2514/6.2013-1080

II.












Figure 5. Test section geometries

4 of 10
American Institute of Aeronautics and Astronautics



A.

Flow Visualization

To determine the location of the recirculation zone, oil-streak surface visualization was used. Oil-steak surface
visualization describes ow direction and local shear strength, including ow separation regions, which in
turn can infer characteristics of the ow away from the surface. Surface oil streaks can be obtained by rst
suspending particles in an oil base and applying the mixture to the test article surface,with a ne brush
shortly before tunnel operation. The oil mixture will then streak as the air moves within the owpath 18 . For
this study, kerosene is used as the oil, titanium dioxide as the particles, and oleic acid is used as a surfactant.
A photograph is then taken of resultant oil streaks for analysis and as a record.

Downloaded by UNIVERSITY OF MINNESOTA on October 10, 2013 | http://arc.aiaa.org | DOI: 10.2514/6.2013-1080

B.

Eciency Measurement

Overall IPS eciency is the target of this research, so it is intuitive to measure particle separation eciency
in the research facility. For each eciency measurement both barrier lters are removed from the tunnel
and weighed; the PDS, lled with particles, is also weighed. Everything is then re-installed and the tunnel
ow rates are brought up to the desired operating condition. Once the steady airow operating condition is
reached, the PDS is turned on and the tunnel is allowed to run for predetermined length of time in order
to have enough particle buildup in the barrier lters to achieve an accurate measurement. The PDS is then
turned o, and after particle delivery has stopped, the air system is turned o as well. Both barrier lters
and the PDS are then weighed in order to determine the change in mass of each component. Generally, a
small fraction of particles are lost during the experiments. On a typical run, about 100 grams of particles
are delivered into the airstream and 1-2 grams are not picked up in the lters. This lost mass is dened
herein as the residual mass, mr . This can be traced to particles that deposit on the surfaces of the tunnel,
very small particles that pass through the lters, or particles that are lost to the ambient air between the
PDS and the tunnel intake. While these are small losses, it is important to quantify their eect.
For this work the eciency of the IPS is dened as the mass of particles entering the scavenge ow ms
path relative to the total number of particles entering the intake, mi .

ms
ms
=
mi
mc + ms

(2)

Here the particles entering the scavenge ow is determined by the change in mass of the scavenge lter, ms .
The total mass entering the intake is measured using the sum of the particles collected in the scavenge and
core lters. This measurement can be checked by using the total mass of particles being delivered to the
system mP DS , resulting the the residual mass measurement. The residual mass, mr , in conjunction with
the mass balances uncertainty, mb , can be used to calculate the uncertainty of the eciency measurement,
, as shown in Eq. 3.


2
2


ms
mc

2
2
(mr + mb ) +
(mr 2 + mb 2 )
(3)
=
2
2
(mc + ms )
(mc + ms )
C.

Test Dust

Particle motion in an IPS is composed mainly of two mechanisms, uid forces (primarily drag) and the
particles inertial forces (primarily bouncing). Small particles which have a small relative mass, and therefore
small inertial forces, will be dominated by uid ow, whereas particles with large relative inertia will not
be signicantly aected by the uid forces and their path through the IPS will be governed by bouncing.
Typical sand and dust entering the engine will have particles covering the entire spectrum of particle sizes,
ranging from totally uid dominated particles to particles that are entirely bounce dominated.
The Stokes Number can be used to determine how the uid and inertial forces will inuence the motion
of a particle. The Stokes Number is a non-dimensional parameter which is the ratio between a particle
response time and a domain uid response time. The particle response time, or how fast a particle reacts to
the uid forces imposed upon it, can be determined by the ratio of the eective particle momentum to the
acting uid forces Eq. 4.
p


me |w|
(p + c air ) dp 2
=
FD
18air f
5 of 10

American Institute of Aeronautics and Astronautics

(4)

The domain uid response time is simply the time it takes the uid to move across the domain, Eq. 5
D

D
uD

(5)

Taking the ratio of these two time constant results in the domain Stokes Number, Eq. 6
P
(p + c air ) dp 2 uD
=
D
18air f
D

(6)

It is important to note that this domain Stokes Number, is a generalized value to give a particles characteristic Stokes Number for the entire ow eld. There are other Stokes Numbers, such as the turbulent Stokes
Number that relate how a particle will react to small perturbations.
If a particles response to the uid is dominated by its own momentum forces, particle will have a
longer response time relative to the uid response time, resulting in a Stokes Number much larger than 1.
Conversely, if a particles response time is much shorter than the uids, it will have a Stokes Number much
less than 1. A particle whos Stokes number is on the order of 1 is likely to have its motion inuenced by
both bouncing and uid forces.
1
Volume Distribution, %/bin
Cumulative Distribution

0.8
0.6
0.4
0.2
0

10
Diameter, m

100

(a) Particle size distribution

(b) Image taken with a confocal

microscope at 100x

Figure 6. A4 Coarse Test Dust

1000

100

10
D

Overall IPS eciency measurements will be conducted using A4 Coarse, Arizona Test Dust (ISO
12103-1) from Powder Technology Inc. This is a
standard test dust that can approximate the conditions seen by the system when installed on an aircraft. The test dust is composed primarily of Silica,
SiO2 , and Alumina, Al2 O3 . Particles cover a range
from 0200m; the volume distribution is shown in
Fig. 6(a). Fig. 7 shows the range of Stokes Numbers
present in the test dust for the conditions being examined. This analysis shows that the IPS will have
particles covering the entire regime of particle behavior.
The dust particles are of irregular shapes, evident in the picture shown in Fig. 6(b). The irregular
shape will eect the drag of each particle dierently
and will also eect how the particle bounces. It is
impossible to generalize these eects in the Stokes
Number calculation, but it is important to note its
existence.

St

Downloaded by UNIVERSITY OF MINNESOTA on October 10, 2013 | http://arc.aiaa.org | DOI: 10.2514/6.2013-1080

StD

0.1

0.01

0.001

10
d, m

100

Figure 7. Test dust domain Stokes Number

6 of 10
American Institute of Aeronautics and Astronautics

III.

Flow Visualization

Oil-streak ow visualization was completed for each OSG at a scavenge split, , of 0.16; the results are
shown in Fig. 8. The direction of the ow is from top to bottom and each OSG has been annotated with
a dashed line depicting the start of the recirculation zone and an image of the geometry to the left. The
position of the recirculation zone is indicated by a drastic change in the streaklines. The well dened inlet
streaks suddenly stop at the beginning of the zone where the oil-streak mixture is not signicantly moved
by the ow. The oil-streak results show that the OSG-2 recirculation zone pushes the furthest upstream,
followed by the the OSG-3 zone, and nally the OSG-1 zone starts furthest downstream.

OSG-2

OSG-3

Figure 8. Oil streak images of each OSG at = 0.16 comparing recirculation zone size

IV.

Eciency Results

Results for all three OSG ow paths using A4 Coarse Test Dust are shown in Fig. 9(a). The core ow
was held constant at 1.7 lbm/sec and three target scavenge splits of = 0.14, 0.16, and 0.20 were used as
primary test conditions. Additional tests for OSG-1 were conducted at lower scavenge splits to explore the
low end operability of the IPS. The measured eciencies fall with the expected range of values given by
Filippone and Bojdo 3 . It is evident that the eciency has a dependence on the scavenge split for all three
OSGs. As is increased, the eciency of the system is also increased; this implies that there are signicant
uid eects on the particles within the IPS.
1

0.9

0.9

0.8

0.8

0.7

0.7

0.6

0.6

0.5

0.5

Downloaded by UNIVERSITY OF MINNESOTA on October 10, 2013 | http://arc.aiaa.org | DOI: 10.2514/6.2013-1080

OSG-1

0.4

0.4

0.3

0.3

0.2
0.1
0
0.05

0.2

A4 Coarse Test Dust

Size Range: 0200 m
Mean Diameter: 36.8 m

0.1

0.15

OSG1
OSG2
OSG3

0.2

0.25

0.1
0
0.05

0.3

(a) Raw data

OSG1 Average
OSG2 Average
OSG3 Average
McAnnaly14
A4 Coarse Test Dust
Size Range: 0200 m
Mean Diameter: 36.8 m

0.1

0.15

Duffy15
Breitman16
Vittal17

0.2

0.25

(b) Averaged data comparison

Figure 9. IPS Eciency for A4 Coarse Test Dust

7 of 10
American Institute of Aeronautics and Astronautics

0.3

By averaging the data at each condition, the trends can be seen more easily in Fig. 9(b). The eciency
data are in good agreement with previous IPS studies that were done, specically Vittal et al. 17 and Duy 15 ,
This data validates the use of a rectangular geometry to study the multiphase interactions within the IPS
and that the conditions are similar to a typical IPS system.
Relating this data to the oil-streak data, it is evident that there is a correlation between recirculation
zone position and separation eciency. The recirculation zone present for OSG-1 is positioned furthest
downstream and has highest eciency, while the recirculation zone in OSG-2 is positioned furthest upstream
and has the lowest eciency.

Downloaded by UNIVERSITY OF MINNESOTA on October 10, 2013 | http://arc.aiaa.org | DOI: 10.2514/6.2013-1080

V.

Conclusions

There is a signicant correlation between the position of the recirculation zone and the eciency of
the IPS geometry. This nding supports the premise that the recirculation itself has a signicant impact
on separation eciency. The separation eciency of an IPS is also dependent on the scavenge ow split;
increasing the amount of scavenge ow results in an improvement in eciency and as the scavenge ow split
for OSG-1 approaches 0.10 there is a signicant increase in separation performance sensitivity to . Further
work is planned to study the inuence of the recirculation zone, including the use of mono-disperse spherical
particles to eliminate variables concerning particle size and shape, and particle image velocimetry to study
the ow eld.

References
1

D. Barone, E. Loth, and P. Snyder, A 2-d inertial particle separator research facility, in 28th AIAA
Aerodynamic Measurement Technology, Ground Testing, and Flight Testing Conference, 2012.

P. Stallard, Helicopter engine protection. Perfusion, vol. 12, no. 4, pp. 263 267, 1997.

A. Filippone and N. Bojdo, Turboshaft engine air particle separation, Progress in Aerospace Sciences,
vol. 46, pp. 224245, 2010.

A. Hamed, W. Tabako, and R. Wenglarz, Erosion and deposition in turbomachinery, Journal of

Propulsion and Power, vol. 22, no. 2, pp. 350360, March-April 2006.

F. Saeed and A. Z. Al-Garni, Analysis method for inertial particle separator, Journal of Aircraft, vol. 44,
no. 4, pp. 11501158, July-August 2007.

P. Snyder, Particle separator for a gas turbine engine, US Patent 6,698,180, March 2003.

A. Hamed, Y. D. Jun, and J. J. Yeuan, Particle dynamics simulations in inlet separator with an experimentally based bounce model, Journal of Propulsion and Power, vol. 11, no. 2, pp. 230235, March-April
1995.

P. de la Calzada, R. Vazquez, F. Fern

andez, and M. S. Segundo, Particle dynamics simulation for
aeroengine intake design, Industria de Turbo Propulsores, Tech. Rep. ISABE 99-7280, 1998.

A. Hamed, K. Das, and D. Basu, Numerical simulations of ice droplet trajectories and collection eciency
on aero-engine rotating machinery, in 43rd AIAA Aerospace Sciences Meeting and Exhibit, no. 2005-1248.
AIAA, 2005.

10

M. E. Taslim, A. Khanicheh, and S. Spring, A numerical study of sand separation applicable to engine
inlet particle separator systems, Journal of the American Helicopter Society, vol. 54, no. 4, 2009.

11

G. O. Musgrove, M. D. Barringer, K. A. Thole, E. Grover, and J. Barker, Computational design of a

louver particle separator for gas turbine engines, in Proceedings of ASME Turbo Expo, no. GT2009-60199.
ASME, 2009.

12

M. E. Taslim and S. Spring, Numerical study of sand particle distribution, density, and shape eects
on the scavenge eciency of engine inlet particle separator systems, Journal of the American Helicopter
Society, vol. 55, no. 2, 2010.
8 of 10
American Institute of Aeronautics and Astronautics

Downloaded by UNIVERSITY OF MINNESOTA on October 10, 2013 | http://arc.aiaa.org | DOI: 10.2514/6.2013-1080

13

L. Y. Jiang, M. Benner, and J. Bird, Assessment of scavenge eciency for a helicopter particle separation
system, Journal of the American Helicopter Society, vol. 57, 2012.

14

W. J. McAnally, Investigation of feasibility of integral gas turbine engine solid particle inlet separators, in Proceedings of the 10th Annual National Conference on Environmental Eects on Aircraft and
Propulsion Systems, 1971.

15

R. J. Duy, Integral engine inlet particle separator volume ii. design guide, General Electric Company,
Tech. Rep. AD-A015 064, August 1975.

16

D. Breitman, E. Dueck, and W. Habashi, Analysis of a split-ow inertial particle separator by nite
elements, Journal of Aircraft, vol. 22, no. 2, pp. 135140, February 1985.

17

B. Vittal, D. Tipton, and W. Bennet, Development of an advanced vanesless inlet particle separator for
helicopter engines, Journal of Propulsion, vol. 2, no. 5, pp. 438444, 1986.

18

R. J. Goldstein, Fluid Mechanics Measurements, 2nd ed.

Washington D.C.: Taylor & Francis, 1996.

9 of 10
American Institute of Aeronautics and Astronautics

Downloaded by UNIVERSITY OF MINNESOTA on October 10, 2013 | http://arc.aiaa.org | DOI: 10.2514/6.2013-1080

Table 1. Test Section Coordinates used in Fig. 5.

yOSG-1

yOSG-2

yOSG-3

yhub

0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
4.50
5.00
5.50
6.00
6.50
7.00
7.50
7.75
7.80
7.85
7.90
7.95
8.00
8.05
8.10
8.15
8.20
8.25
8.30
8.35
8.40
8.45
8.50
9.00
9.50
10.00
10.50
11.00
11.50
12.00

2.06
2.09
2.13
2.24
2.37
2.57
2.75
2.93
3.07
3.20
3.31
3.38
3.42
3.39
3.31
3.17
3.07
3.05
3.02
2.99
2.96
2.92
2.89
2.85
2.81
2.78
2.73
2.68
2.64
2.59
2.55
2.50
2.12
2.00
2.00
2.00
2.00
2.00
2.00

2.06
2.09
2.13
2.24
2.37
2.57
2.75
2.93
3.07
3.20
3.31
3.38
3.42
3.43
3.42
3.39
3.36
3.35
3.34
3.34
3.33
3.32
3.31
3.30
3.29
3.29
3.27
3.26
3.24
3.23
3.21
3.19
3.00
2.70
2.33
2.06
2.00
2.00
2.00

2.06
2.09
2.13
2.24
2.37
2.57
2.75
2.93
3.07
3.20
3.31
3.38
3.42
3.39
3.31
3.18
3.10
3.08
3.06
3.04
3.02
3.00
2.98
2.96
2.94
2.92
2.90
2.88
2.86
2.84
2.82
2.79
2.59
2.44
2.28
2.05
2.00
2.00
2.00

0.43
0.44
0.47
0.53
0.65
0.87
1.15
1.42
1.71
1.99
2.24
2.40
2.43
2.15
1.40
0.25

ysplitter,scav

ysplitter,core

1.78
1.90
1.92
1.91
1.89
1.87
1.81
1.76
1.71
1.67
1.63
1.59
1.55
1.51
1.48
1.45
1.32
1.32
1.32
1.32
1.32
1.32
1.32

1.77
1.65
1.55
1.44
1.35
1.25
1.16
1.06
0.96
0.86
0.76
0.67
0.58
0.48
0.38

Note: The test section width is 8 inches.

10 of 10
American Institute of Aeronautics and Astronautics