Professional Documents
Culture Documents
Hypersonics I
J. J. Bertin
R. Glowinski
J. Periaux
Editors
Volume 1
Defining the Hypersonic Environment
ISBN 0-8176-3418-5
Volume 2
Computation and Measurement of Hypersonic Flows
ISBN 0-8176-3419-3
2-Volume Set
ISBN 0-8176-3420-7
Advances in Hypersonies
Defining the Hypersonic Environment
Volume 1
J. J. Bertin
J. Periaux
J. Ballmann
Editors
987654321
Contents
by
E.H.Hirschel
Messerschmitt-B6lkow-Blohm GmbH
Military Aircraft Division
P.O. Box 801160
0-8000 Mlinchen 80
SUMMARY
1. INTRODUCTICl'l
2
pects. Chapters 5, 6 and 7 are devoted to an overview and discussion of
viscous phenomena, heat transfer phenomena and heat load problems, and
finally real-gas effects. In the concluding Chapter 8 it is tried to
give a synthesizing overview in the sense of the title of the paper.
2. AERO'lBERID>YNAC DESlm OF
A'lK>SPHERIC HYPER500IC AIRPLANES
the configuration,
the structure and materials concept,
the propulsion system,
the cooling of the propulsion system,
the fuel.
3
propulSion TURBOJET I
----- -- RAMJET
'''----- --~
"'-
fuel
structure
JP
AI MIN 11M
..
R'~ii. ",
"'"'
I
LNG
",~HZ
• LNG:
•• RSR:
"-
>
l,quified natural gas
rapid sol,dif,catlOn rate materlal
material
I
TlTANIUM",-lfR.:~~~ __________
I ~ERAMICS
",SUPERALLO RSR"
CARBONiCARBON
SUPERALLOYS
LKl LK2 LK3 LK4
I I I I
2 4 6 8 10 M 12
Fig. 2.1 Technology jumps with supersonic and hypersonic aircraft at
rising Mach numbers, Ref.l
Or----~---~~------------------------_,
engine requirement
at 24 kPa U4ach 6)
dynamic pressure
.4
coolant flow /"72 kPa
fuel flow*) / r24kPa
.6
.8
surplus
at 24kPa
(Mach 6)
I.OI---+-n-a-v"- -,""l-aL\bl-e------~~~
1.2
*) stoichiometric
combustion
L4·~4-----~---~~----------~8------------~10
v Mach mnber
5
Fig. 2.2 Engine cooling requirements of hypersonic airplane, Ref.3
o AlAA - used with permission
4
For possible future hypersonic passenger airplanes the passenger
comfort degradation can be a limiting factor. Fig. 2.3, partly taken
from Ref. 2, shows that acceleration and especially deceleration limits
can affect very much a flight mission and its economy. At a Mach number
Moo = 12, which amounts roughly to a flight speed of vOl) = 4 kID/s, a 25
percent g-reduction occurs, which certainly would lead to problems in
commercial passenger transportation.
a)
2
--
acceleration ( 0.59)
cruise
~
deceleration
(-0.1259)
-0.2el
(-0.29)
(-0.29)
10 12 14 16 18 20
distance [1/1000km)
c)
b) 1.0 _ _= - - - - - - - - - - - - ,
0.5~-----+-~:----_i
5
hypersonic vehicle (present): numerical
airplane (classical) b) methods as additional design tool
there is only a very small data base and there are only few
statistics from earlier designs, because there are only very
few earlier designs, in addition much knowledge for several
reasons is not available,
the approximate methods are old and limited, partly the error
bounds are not established (validation problem),
6
- the wind-tunnel simulation has several major shortcomings,
the error bounds are not established (validation problem),
the classical methodology to test separately for the involved
parameters Mach number and Reynolds number is questionable,
other parameters are also important.
7
from Ref.6, where the specific impulse of the propulsion system is used
as an additional parameter, which can be neglected for the present dis-
cussion.
nominal: 1
300~----------------~~~
LID = 6.0
0.76
200
8
However, a reduction of the viscous drag would indirectly also
affect the heat loads. If in general the heat loads can be reduced, al-
so at the propulsion system (boundary-layer diverter, external nozzle,
etc.), the mass of the heat-protection means can be reduced, and the
ratio m.mpty~.ke-off be shifted to a smaller, less weight-critical
value. Of course, a progress in structural and material technology can
bring the same effect; in any case both effects should be combined.
9
- to study the efficiency and performance of configuration
parts (inlet, nozzle, flaps, etc.), systems and subsystems.
other very important items are, and they depend on the novelty
and the risks of the project:
3. ~C CHARACl'ERISTICS OF
NIKlSPHERIC HYPERSaUC AIRPLANES
11
H
[kmJ
.
50
SAENGER
lower stage
2(fo1;.=3)
-"'T"""- .""-~/
1(M",=2.2) 8
---
l!'~ea;;;c-::::Pt (LK)--
corde 14
v [km/sJ
o 4
Fig. 3.1 Hypersonic airplane concepts in the altitude-velocity plane,
after Ref.2
L. . ./::::;;::~~i;S~t--1---::1~015 ~turbUlent
r----~~-~-~~-,_~~~r-=_~~--_;~------~ L=l~
H
[kmJ
r /
flow
Rell =
50
1m- I ]
10'
5· 10'
o 2 4 f, v Ikm/sJ 7
Fig. 3.2 Hypersonic airplane concepts with Reynolds numbers and real-
gas total temperatures, after Ref.2
stagnation
area, sphere
.,., L.
r = 0.3 m
:0
Ol-
L.
i6'{
~~
-0
::IN
>0
o 2 4
Fig. 3.3 Hypersonic airplane concepts and effects,
after Ref.2
12
Indicated is the approximate location of the vulnerable ozone layer,
where no extended flight should take place.
Fig. 3.1 shows the nominal flight points and the Mach numbers.
It should be noted that with pr~sent-day structures dynamic pressures
will not exceed IInlch qQ) = 70 kPa. All the flight points lie at lower
values. Turbo propulsion is possible up to approximately v", = 1 krovs,
and ram propulsion up to approximately vQ)= 2.5 kmVs, see also Fig.
2.1.
Fig. 3.3. finally shows that below Veo = I krovs the air can be
considered as perfect gas. For I krovs;li Vat: 3 krovs vibration excitation
has to be taken into account, and above voo " 3 krovs oxygen dissoziation
(see Chapter 7). Probably it is sufficient for speeds below v'" = 3.;. 4
krovs (Moo < 10 to 12) to assume equilibrium real gas behaviour. An ex-
ception of course is the propulsion system, and with regard to external
aerothermodynamics the external nozzle and the base.
13
viscosity-effect dominated, which will be demonstrated in the following
chapters. In Table 3.1 a rough classification is given of hypersonic
airplanes (cruise vehicles), reentry vehicles and ascent and reentry
vehicles. Aeroassisted orbital transfer vehicles are not considered. It
is important to make this differentiation, because hypersonic flight
has different consequences of all kind for these different classes of
hypersonic vehicles.
OPPosite demands at
configuration slender blunt ascent and reentry
flight time long short short
angle of attack small large small/large
viscosity effects pressure field viscoSity effects
flow field dominated dominated dominated
thermodynamic weak (except for ex-
effects ternal nozzle area) strong strong
opposite demands at
drag small large ascent and reentry
OPPosite demands at
11 ftldrag large small ascent and reentry
thermal problem thermal household thermal loads thermal loads
propulsion large Mach number large Mach number span
speCial problems Integration span opposite demands I
4. cmFIGURATIcmL ASPECTS
14
The most important fact is that airbreathing hypersonic air-
planes are highly integrated lift- and propulsion systems. The lower or
pressure side of the vehicle produces the lift (see Sub-Chapter 4.2),
it also houses the propulsion system with the inlet, the engines, and
the nozzle (see Sub-Chapter 4.3). The lower side of the forebody thus
must be considered as part of the inlet, and the lower side of the
base/tail as part of the nozzle, which there is an external asymmetric
nozzle. Of course, the pitching moment household is strongly affected
due to these pecularities (see Sub-Chapter 4.4).
15
10
LID
\
~ hypersonic
airplanes
5 ~ ..... r--
. +====l , .........
reentry
vehicles
...
-
-""'-
~
o
o 10 20 30
Fig. 4.2 Approximate orders of magnitude of the lift to drag ratio
LjD in the flight Mach number ranges of different hypersonic
vehicles (following Ref.12)
- with Mach numbers higher than Moo '" 4 and large operation ran-
ges liquid hydrogen becomes necessary as fuel (see Fig. 2.1).
This leads to large tank volumes, which makes it necessary to
use the fuselage for storage. This and other considerations
lead to the "blended-body" concept, where slender wing and
fuselage are highly integrated, however with an airplane sur-
face as small as possible. Certain considerations then lead
to the wave-rider concept, see for instance Ref.12, which is
not considered here.
- with rising Mach number the lift is produced more and more
only on the lower side, i.e. the pressure side of the wing.
16
Fig. 4.3, following Ref.IS ,demonstloates this. The limit, New-
tonian flow, can be understood in this sense, too: the lee-
side lies in the shadow of the windward side and does not add
to the forces on the body (note that the Newton-theory gives
good estimates of forces down to Moo '" 4). For iii hypersonic
airplane the consequence is that the lower side of the fuse-
lage is part of the lifting system, i.e. the whole underside
of the airplane is the wing. Usually therefore also the whole
airplane planform surface is taken as the reference surface
for coefficients.
1.0~-----------r-.------------~----------~
O.5~-.~------4-+------------r----------~
O~----~--~~----~----~----~----~
o 4 8 M 12 co
Fig. 4.3 Production of lift moves to pressure side with increasing
Mach number (schematically), after Ref.IS
o AIM - used with permission
These three aspects - high slenderness, large volume, lift pre-
dominantly on the pressure side - lead to the typical long, slender
blended-body configurations studied today. Because of the large running
lengths, boundary layers on such configurations became very thick, and
hence viscous effects in general become very important, and partly even
dominate the aerothermodynamics of such airplanes.
- take-off/landing performance,
- transonic performance,
17
4.3 Aspects of propulsion Integration
18
A boundary-layer diverter in the classical sense will lead to
structural design problems, but also to structural weight increase due
to heat loads (no radiation cooling), if it must operate also in the
ram or scram mode. In any cas~ the diverter must be as small as possitr
Ie in order to keep the diverter drag as small as possible. It is evi-
dent therefore that an accurate prediction of the boundary-layer thick-
ness at the inlet is necessary. In addition, if possible, the thickness
should be reduced by an appropriate forebody shaping. Nose bluntness
and pressure field shaping can delay transition laminar-turbulent to a
location further downstream (Chapter 5), lateral shaping can probably
be used to reduce the boundary-layer thickness via three-dimensional
flow effects.
The two major aspects with regard to the longitudinal trim pro-
blem are
the large shift of the center of pressure, and hence the neu-
tral point, over the Mach number range (up to 10 per cent of
the airplane's length for a Mach 5 plane),
19
shows, and this is the more true for the pitching-moment household.
Trinnning by fuel management is a possibility, as well as the employment
of canards in a certain Mach-number range.
Trim and control problems therefore are potentially and also ac-
tually affected strongly by viscous effects.
net thrust
v",'"
thrust-l ift
component
g-reduction
Fig. 4.4 Schematic of mass-force polygon of hypersonic airplane at
high speed
20
4.5 Simulation Problems
22
5. ASPECTS OF VISCOOS l'III!1OtENA
a... F'ow
. _ - - - - - - - L ::: 80 m- - - - - - - - - J I I
Fig. 5.1 Aerothermodynamic effects on SAENGER configuration with
upper stage HORUS (schematically)
23
Three-dimensional effects are also weak especially on the low-
er side of the forebody. Whether a strong attachment line topology
will exist on the lower side of the forebody is not yet known. There
"leading-edge contamination", which is surmised to have been a factor
in the transition process on the Space Shuttle, Ref.19, can be an ad-
ditional transition mechanism.
24
lent wind tunnel-walls governs the transition process in a wide Mach
number range, Ref.2l. Besides the demand of Mach and Reynolds-number
similarity, the thermal state of the body surface and the temperature
gradients (radiation) might be crucial to simulate. A quiete tunnel to-
gether with total temperature and thermal wall condition duplication is
necessary, if a high ~gree of accuracy in the determination of the
transition process is demanded by design considerations.
25
general predictions of turbulence boundary layers are not too bad at
lower Mach numbers, if the flow is attached, Ref.23. How good the tur-
bulence models are at Mach numbers larger than three to four cannot be
decided, because too few experimental data of sufficiently high quality
are available. Introduction of density fluctuation terms into turbu-
lence models like in Ref.24 gives no conclusive answers, Ref.25. Final-
ly, if the transition location dominates the development of the turbu-
lent boundary layer, the question of exact turbulence models ceases to
be of primary importance.
shockjboundary-layer interactions,
- vortex/boundary-layer interactions,
shock/shock/boundary-Iayer interactions,
hypersonic viscous interaction.
26
stance in locations where the bow shock is being intersected by a con-
figuration part, Fig. 5.3. Such locations must be accurately predicted
together with the heat loads in order to introduce local thermal pro-
tection systems if necessary.
27
Fig. 5.4 Schematic of longitudinal corner flow with glancing shock
surface
28
flow must be turbulent in the proper geometrical (integral) relations.
Trim and control surface efficiency and heat loads can be measured with
a complete configuration, which, however, in any case will be small
(Reynolds number!, boundary-layer tripping). Enlarged component experi-
ments can be imagined, again the oncoming boundary layer must have the
proper geometrical relations. Inlets will be studied as components, be-
cause on a complete model no actually operating inlet is imaginable
with the present techniques. The same is true for external nozzle and
base-flow problems.
-11-----~~--~f_-------'f-----_1 0.05
AC M due to
viscous effects
(no flop deflectionl
r O.1...:----1.\II--I--:----~----~----J
10.3
SAENGER lower stage
10.2 10'·
0
lifi Moo lyReoo I 10·
Large heat loads and thermal problems are typical for high speed
flow. A substantial part of the kinetic energy of the hypersonic air-
plane is converted into heat by irreversible processes. This part of
the total drag is transferred by convectional and diffusional processes
towards the surface of the airplane. Two major aspects, see Fig. 6.1,
taken from Ref.29, can be distinguished:
29
+I radiation
sun
Mco
coo 1 i ng needs:
fuselage
engines
Fig. 6.1 Heat sources (heat loads) and heat sinks at hypersonic air-
plane (schematically), Ref.29
1. The heat loads. This is the heat which actually is carried to-
wards the airplane surface. Important is not only the heat flux as
such, which is proportional to some temperature gradient, but also the
temperature level, on which this happens. Aerothermodynamic heat loads
occur on all parts of the fuselage, and, what is very important, also
in the propulsion system (inlet, engine, nozzle). In the engine and in
the nozzle they are of course coupled with the combustion heat loads.
Other heat loads exist in the airplane (systems, payload).
30
earth-orbit reentry trajectory - emission and absorbtion of radiative
energy by the airflow past the flight vehicle can be neglected, Ref.30.
Of course the emitting surface portion must not face another emitting
surface portion. Where this is the case, for instance in the boundary-
layer diverter, in the internaj parts of inlet and nozzle, and also in
the engines themselves, no net heat can be radiated away. The allevia-
tion of heat loads can be drastic, as Fig. 6.2 demonstrates (defini-
tions see Fig.6.3). The computations were made with the approximate me-
thod described in Ref.31. In any case it holds that the hotter the
wall, the more energy is being radiated away.
radiation
radiation adiabatic
wa 11 temperature Tra
(E: = 0.85)
total temperature To
31
z
qw
Fig. 6.3 Schematic of heat fluxes at a radiating surface, z: direc-
tion normal to the wall (Ref.32)
32
number range, turbulence tripping (if the Reynolds number is too small)
in many cases are available. This holds also for the total enthalpy
(total temperature). The question then is which of the three cases dis-
cussed above (cases a, b, c) is realized. With a cold-model technique
radiation is not being taken into account. Model techniques which allow
a fast heating of the surface are still questionable because due to
short blowing times in many hypersonic wind tunnels a steady state
might not be reached. Even if the right radiation properties are there,
the question arises whether the tunnel walls will re-radiate a part of
the emitted energy. This part might be small, if they remain essential-
ly cold during the run.
•
I 4
:;Q
numerical solution
Moo = 10
Too = 220K
L = 0.075m -
£1 = £2 = 0.8
Rel =1.2·10G,Re2 =1.2· 10'
x/L perfect g~S' laminar I
a _l 1
a x
a 0.5 L
1.0
Fig. 6.4 Comparison of approximate and numerical-solution ratio of
radiation-adiabatic wall temperature at a hyperbola for two
different Reynolds numbers (Ref.32)
33
mental simulations in principle too, apart from radiation) will give a
certain heat load (temperature, heat flux) at a given location. If in
reality for instance a vortex/boundary-Iayer interaction happens at
that location, which was not taken into account in the simulation (ei-
ther because only parts of the configuration were considered, or be-
cause the vortex was not present due to, for instance, insufficient
Reynolds number similarity), heat loads possibly much higher than pre-
dicted can appear. The reason is that the vortex transports fluid with
the original total enthalpy towards this location, with the resulting
high heat loads. This mechanism is comparable to that of the vortex ge-
nerator in low-speed aerodynamics where external flow momentum is
brought on purpose into the boundary layer in order to reduce separa-
tion effects.
Apart from the engines, the internal and the external nozzle
flows, however, experience strong, even non-equilibrium, real-gas ef-
34
c
8'
f")
N "'-
C ~
00 . . . .
.....................................................................
CD
IMinar, perfect gal,
,-r~;
c-O.
turbulent, perfect gal, c-O.
f-X/l IMinar, re.1 gal, cwO.
. : :
o 1 )( turbulent, real gal, c-O.
o bainar, real gal, cwO.IS
turbulent,
". real ga., cwO.IS
\1:
~.
. . .
..... ········:··············· ... ··············1············· ..: .............. .
o
~ +------,-------r------T-----~------~
0,0 0,2 O,~ 0,6 0,8 x/L 1,0
Fig. 7.1 Influence of modelization on wall temperatures (qw = 0) in
symnetry lines of a generic forebody, Ref.33 (M.., = 6.8, ReL
= 1.22 .108 , L = 55m, 0< = 6°)
@ AIAA - used with permission
35
8. CClIlDLUDING RmfARKS
36
which must encompass the whole aerothermodynamic community at universi-
ties, research establishments and industry. If such programmes can be
started in time, if they make use of the different abilities of the
partners in an optimum fashion, the design of hypersonic airplanes will
profit from them in a decisive way.
~
The author wishes to thank his colleagues N.C.Bissinger and G.Cucinel-
Ii for several very helpful discussions, and A.Ko~ for the computation
which he performed for Fig. 6.2. He thanks G.Koppenwallner for the per-
mission to use Fig. 5.5 from Ref.28.
The author also thanks the following institutions, who granted reprint
permissions: AIAA/ICAS (Figs. 2.2, 4.1, 7.1), u.s. National Research
Council (Fig. 2.1), and AGARD (Fig. 5.5, the original version of this
figure was first published by the Advisory Group for Aerospace Research
and Development, North Atlantic Treaty Organization (AGARD/NATO) in
Conference Proceedings CP428 - "Aerodynamics of Hypersonic Lifting Ve-
hicles" in November 1987).
9. REF'ERENCES
37
9. P.Sacher: "Hypersonic Experimental Aircraft Technology Demonstra-
tor HYTEX". MBB-Flugzeuge, Ottobrunn, 1990.
24. M.Situ, J.Schetz: "New Mixing Length Model for Turbulent High-
Speed Flows". AIM-Paper 89-1821, 1989.
38
26. B.Edny: "Anomalous Heat Transfer and Pressure Distributions on
Blunt Bodies at Hypersonic Speeds in the Presence of an Impinging
Shock". FFA Report 115, 1968.
27. B.H.Anderson: "Three-Dimensional Design Methodology of Supersonic
Inlet Systems for Advanced Technology Aircraft". In: Numerical Me-
thods for Engine-Airframe Integration (S.N.B.Murthy, G.C.Paynter,
eds.), Vol. 102 of Progress in Astronautics and Aeronautics, AlAA,
New York, 1986, pp. 431-380.
28. G.Koppenwallner: "Low Reynolds-Number Influence on Aerodynamic
Performance of Hypersonic Lifting Vehicles". AGARD-CP 428, 1987,
pp. 11-1 to 11-14.
29. Th.Gottmann: "Aspekte des Hyperschallfluges und Beschreibung des
Leitkonzepts 4 der BMFT-Studie Uberschallflug". MBB/LKE127/HY-
PAC/R/5/A, 1987.
30. R.K.Hoeld: "Die Berechnung dreidimensionaler Hyperschallstromungen
mit Hilfe der Viscous-Shock-Layer-Gleichungen". Fortschrittsbe-
richte VOl, Reihe 7, Nr. 171, DUsseldorf, 1990 (Doctoral Thesis,
Universitat der Bundeswehr Mtinchen, 1989).
31. A.Ko9: "Aerodynamische Aufheizung des Demonstrators". MBB/FE112/
HYPAC/TN/0130, 1990.
32. E.H.Hirsche1, Ch.Mundt, F.Monnoyer, M.A.Schmatz: "Reynolds-Number
Dependency of Radiation-Adiabatic Wall Temperature". MBB-FE122-
AERO-MT-872, 1990.
39
CONCEPTS OF HYPERSONIC AIRCRAFT
P. PERRIER (DASSAULT AVIATION - France)
O. INTRODUCnON
and :
- aerothermal limits of the materials (with the limited
improvement possible in the program schedule),
- mass-volume and thei r balance (defi ni ng useful or necessary
payload and equipments),
- control and guidance (that make also the project critical or
not, with margins).
We have to discover that such constraints have been and will be
probably the selecting factors of the concepts.
41
AERODYNAMIC
HYPERSONIC PROBLEMS
AEROTHERMAL
Table I
On the second line there are the vehicles with radius of the nose
of about one meter and wi th 1eadi ng edges radi i of the order of a
portion of meter. Their sweep angle increases with reduction of size.
Only large size of the u.S. Orbiter allows intersection of shocks at the
leading edge of the wing; smaller sizes oblige to return progressively
to 1i fti ng body and to spheri ca 1 shapes - of pure convex topology, at
least on windside.
43
AEROTHERMAL CONSTRAINT WITH SIZE
(ReynOlds numb., effect)
FIg .... 1
Low temperature
Mean temperature
lJD - 1
~
~
Very high temperature
+ SIIIIIIiuI on some peru
• nose UD 2.5 Wave Propulsion first
·L.E. rider
~, UD 3.8
44
2 EXTREME DESIGNS
Figure 2
Star H Hermes
multiple shock on inside bow shock
air intake, corner flow without any shock interactions:
t;
FORCES
:::>
4 - Entropy building Ivs x) ~k~
-- ~ s !--=:tC ~
'[ Wave rider
~x (minimum of S production)
)(
)1.
Table 2 describes the different analysis tools that can be used
thanks to the CFD approach. It is effectively possible to go deeper in
the analysis work and understand how to change the basic concepts or
refine it, knowing where the constraints are active and where the
performance can be improved.
47
STRATEGY FOR CONCEPTS VALIDATION
Table II
RfD Do RfD numerics
numerical analysis
CFD ~~ configuration
SUbstan~ . . survey by CFD '\. ~ configuration
~ refinement
RfD ~
physical analysis
49
EXAMPLE OF BALANCE
Figure 4
5 _ _ _- 10 - MN __ Rarefied flows
C.P.
for
L~
HERMES 242 I ORBITER - Ech. 41. I HERMES 0-0 I HERMES 0-1
Figure 5
2000 HERMES
extreme~ _ ORBITER
Temp
nose 1 LIMIT
Ul
- Repudiation hypothesis
/"/////~~&1
-- -
... .
NOSE! HERMESI ORBITER
- 100 0
L.E. - 50 0 - 150 0
y/b
NOSE SALMON
On figure 6 maximum peak temperature is gi ven versus sweep
angle of the leading edge at the bow-shock impigement. The
four curves are for plane wing, wing with winglet, double
sweep wi ng homothetic to orbiter US pl anform without
consideration of shock-impingement and with the corresponding
overheating.
T
l't)
2'-0
1,'"
'T't,£,
52
VARIOUS DESIGN CONCEPTS
Figure 7
G:---jl ,,
53
2. CONSTRAINTS BY PROPULSION INTEGRATION
2.0 - Main constraints come from the location and geometry of air
intake and exit of the engines. For the research, for better flow field
in front of air intake (it needs correct design of front fuselage), or
for better integration of nozzles, a large degree of variation in design
conditi ons is needed ; 1arge vari ati on in the ai rfl ow is requi red by
engi nes with Mach number. The high pressure induced in ai r intakes as
the high temperature in nozzles gives criticity to the design of
necessary movable parts. So the concepts are the developments of the
proven .state of the art in vari ab 1e geometry for lower pressure and
temperature.
54
AIR-INTAKE INTEGRATION
Figure 8
/' ~ ~
20 3 0 axisymetric Complete 3D
historically coming from coming form axisymetric
characteristics 2 0 characteristics
calculations calculations
~/ (corner flow)
External Define by n rotations SR 71
20 + p aspirations ~-
External - internal 3 0
(corner flow +
B.L. convergence)
-.~}~-
~
main question : What on the sides • Main question : rebuild area
~
secondary: large M.N. variation 7 distribution only with translation -
Complete 3-D design can take advantage of a semi-open isentropic
internal compression as sketched in fig. 8 (reverse flow versus nozzle
exit flow) but such design concentrate boundary layer at the throat. A
spike flow works in opposite direction and mixing of both approaches as
to be examined as a compromise concept.
56
C
M
'iii
E 'iii
...
CD
.5
E
!!
.5
'iii
E
CD
C
M
X
CD
CD
~
Q.
E
o
(,)
. \
.
CD
::I
ii:
c
o
";
'7ii
c
E
I-
.~
C
M
c('oj
c
('oj
'iii
c
CD
X
w
57
o
z
-ow
~
C 0....
W Q)
:E
W
~
.-
a: u.
~
W
58
ANALYSIS TOOLS
Figure 11
fI.I!M1
1 - Flux Heating visualisation
2 - Flux history
I~
~
-' -
-----./~
VI
\0
2 - Drag-thrust distribution external/internal ~~
I""~ 'J ~
L--(~ ~
4 - Entropy building
I--~
THRU ST - DRAG ANALYSIS
(PRESSURE)
Figure 12
"C,
combustion
g
Figure 13
61
THE STAR-H SHOCK WAVE PATTERN
M ~ 6.
F"4Jur~ 14
62
BOUNDARY LAYER THICKNESS
ON THE BODY lOWER PART
63
THE EXTERNAL SEPARATION
COMPARISON OF THE TWO DESIGNS
F1gure 16
64
3. CONSTRAINTS BY CONTROL REQUIREMENTS
65
CONTROLS
Figure 17
Only to be
evaluated with
a flight simulator
~
Analysis tools :
(limited answer)
~
,- - .......
"
eN
Main problem :
- uncertainties to be covered by
(!L F.e.S. robustness
- uncertainties not to cover
inversion of efficiency
Example : lateral control on
......... Hermes/Orbiter
As a matter of fact, control constraints are essential on concepts
evaluation; then in the sketches of possible configurations of controls
in figure 18, we don't retain such finned configuration in lifting
reentry vehicles of low LID; however such fin is useful on low angle of
attack aircraft with air propul sion. Two additive concepts are to be
retained as possible configuration of controls. One with use of flaps on
fl are (as anci ent Dassault TAS desi gn), the other with lower rudder,
foldable for landing. Such rudder may have in low LID vehicles a
tri angul ar secti on i ncreasi ng the 1atera 1 pressure in zero defl ecti on
and so efficiency with deflexion.
67
SOME CONCEPTS FOR CONTROL
Figure 18
+ Low LID
~ ~..I" ~ /FSJl:l
~ -!.~..:.... <~-
II
cambered low side Lifting body non-linear efficiency
a..
00
¥~
without heating limitation
~
_ _""F)~===:::-)~)--.)--...-> -:JJ~ LZi'
~)CJJ,
~4
~'-
40 -----
_ _- -
\ c(~
, ; , ...... -- _ .......... Jii.;i
« 1.s 1 -to. ....~. 15. 40. is. JO. 15.
Calculation Euler {; = , °5 DL - 3° ON - 20°
+
+
+ ++++++ ......++++
Go AVIOM5 " WIM4HlT
'f" II ~
:ttttt\!~tlrttt:
Warping --x--
Yaw angle
Flap .
~
0--
69
We can say that such long approach, if short-circuited, will not
permi t suffi ci ent concept selection and val i dati on ; it wi 11 increase
the risks associated with the program by insufficient evaluation of the
critical problems and insufficient efforts devoted to solve or reduce
such problems. So hypersonic vehicle design remain a not mature activity
where great challenges can be out-passed or not fullfilled by an
insufficient work in the conceptual phasis of the program.
70
- REFERENCES -
71
Hypersonic Wind Tunnel Testing*
R. K. MATTHEWS * *
Introduction
In general, there are four primary reasons for wind tunnel testing:
1. understanding aerodynamic flow (research)
2. development of a database
3. parametric configuration studies and
4. validation of codes/correlations
In the sixties and seventies wind tunnel tests were often directed at
verification or validation of empirical correlations. Today the primary
objective of most wind tunnel tests is the substantiation of some form of a
Computational Fluid Dynamic (CFD) code.
The most common omission in designing a wind tunnel test program is
lack of a precisely defined test objective. Once a clearly defined test objective
is stated and understood by all concerned, it becomes much easier to design
a test approach, select the proper facility, and define the test matrix. A well-
defined test objective includes a detailed understanding of how the data will
be used. For example, if the test objective is "to substantiate the predicted
heating distribution on a space shuttle configuration at hypersonic Mach
numbers," it is important to plan every detail, even to the extent of deciding
the format of how the results will be plotted. The predictions should be made
at the specific tunnel conditions, angles of attack, and Reynolds number of
the test program and, of course, for the same configuration as that tested.
It is interesting to note that on both the Gemini and Apollo programs there
were significant differences between the configurations tested and the flight
vehicles (see Ref. 1).
The types of hypersonic whid tunnels can be classified as:
1. impulse (run time of 1 sec or less)
2. intermittent (run time of several minutes), or
3. continuous (run time of many hours)
* The research reported herein was performed by the Arnold Engineering Development Center
(AEDq, Air Force Sytems Command. Work and analysis for this research were done by personnel
of Calspan Corporation I AEDC Operations, operating contractor for the AEDC aerospace flight
dynamics facilities. Further reproduction is authorized to satisfy needs of the U. S. Government.
• * The author appreciates the material developed by the CalspanlAEDC staff with particular
thanks to W. T. Strike and F. L. Crosswy.
Corresponding examples of these three types are:
1. impulse
Naval Surface Warfare Center (NSWC) Tunnel 9 (Fig. I)
CALSPAN 96-in. shock tunnel (Fig. 2)
2. intermittent
Ames Research Center (ARC) 3.5 ft Tunnel
Langley Research Center (LRC) 8 ft Tunnel, and
3. continuous
Arnold Engineering Development Center (AEDC)
Tunnels B & C (Figs. 3 and 4)
There are many other hypersonic facilities and Ref. 2 presents a
comprehensive listing of these facilities. However, the facilities listed above
have produced the vast majority of existing hypersonic data in the U.S.
Compared to flight testing there are significant advantages of ground
testing:
• many times less costly than flight
• known, controlled environment
• relevant factors can be altered independently
• extensive, detailed instrumentation
On the other hand there are also significant shortcomings:
• incomplete Reynolds number simulation
• requires extrapolation to flight
• interference from tunnel walls and support system possible
• tunnel-to-tunnel variations
• incomplete simulation (velocity or time)
After the decision to conduct a wind tunnel test is made there are three
fundamental issues to be addressed: (I) what are the important simulation
parameters, (2) how do the flight values compare to what can be obtained
in the tunnel, and (3) can the nozzle contour and tunnel reservoir conditions
generate the desired (and known) test section values. More specifically, is
the tunnel reservoir pressure high enough to produce a free-stream Reynolds
number which will provide a naturally turbulent boundary layer on the model?
The fundamental simulation parameters are typically Mach number and
Reynolds number. It is highly desirable that the test section Mach number
be uniform (within ± O.02) in both the longitudinal and radial directions. Since
test section Mach number controls the model shock shape and flow field (and
the resulting pressure and heat-transfer distribution) any flow nonuniformities
can influence these distributions. Another concern in code validation is the
determination of the average Mach number level in the test section. In the
calibration of hypersonic tunnels it is often assumed that the flow is isentropic;
however, it has been shown (Ref.3) that this is not always a valid assumption.
The other important simulation parameter, Reynolds number, is perhaps
even more elusive. In general, it is desirable to match the flight Reynolds
number base of vehicle length with that provided in the tunnel based on
model length. In this case one would expect similar boundary layers, friction
73
drag, separation, lee side vorticity, and similar transition. Again, it has been
shown (Ref. 4) that transition in the wind tunnel is influenced by the noise
from the nozzle walls and, therefore, wind tunnels cannot be relied on to
duplicate flight transition data. Of more fundamental importance is the state
of the boundary layer i.e., laminar vs. turbulent. Boundary-layer trips (see
Fig. 5) have been successfully used for many years to effectively produce
turbulent flow on wind tunnel models. There have been many debates on
this subject, but most aerodynamicists agree that large models (30 + in. long)
are not adversely affected by small O.030-in.-diam trips located close to the
nose and that the resultant turbulent flow provides a heating distribution that
is indistinguishable from that of a naturally turbulent boundary layer.
In selecting a hypersonic test facility there are two other important
considerations in addition to those discussed above. These are:
1. facility support provided, and
2. test techniques available
Test techniques represent probably the single most important factor in
selecting a hypersonic test facility. As a result, the next section is devoted
to this topic.
Test Techniques
The operating characteristics of the chosen facility will immediately restrict
the choice of test techniques because of the tunnel operating mode. Since
the time response of the measurement technique must be compatible with
the run time of the facility, some techniques are immediately ruled out for
certain facilities. An overview of the types of tests performed in the continuous
AEOC tunnels is presented below:
1. force and moment·
• pitch-pause
• continuous sweep
• mass addition
2. pressure
3. flow-field diagnostics
• intrusive
- pitot pressure
- total temperature
- Mach number/angularity
• nonintrusive
- boundary-layer transition detector (BLTO)
- laser particle monitor (LPM)
- laser Ooppler velocimetry (LOY)
- laser-induced fluorescence (LIF)
* Since the principles and techniques for force and moment (balance) testing are not particularly
unique to hypersonics they are only briefly mentioned in this chapter but an excellent example
of this type test is presented in Ref. 5. The emphasis in this chapter is placed on flow field
measurements and aerothermal methodology.
74
4. heat-transfer
• thermal mapping
- phase change paint
- infrared scanning
- thermographic phosphor
• discrete measurements
- thin-skin
- coax gages
- Schmidt-Boelter gages
- Gardon gages
- thin-film
5. material/structures
• screening
• characterization
• component survivability
• component thermal response
A general understanding of test techniques is very important in planning
a test program. Therefore, the basic principles of these test techniques are
discussed in the remainder of this chapter.
Pressure Measurements
Experience has indicated that the model orifice diameter of tubing installed
in the model should be about 0.040 in. The tubing external to the model can
have an outside diameter of 0.093 in. with an inside diameter of at least 0.063
in. These dimensions reduce lag time, yet are still small enough so that the
pressure orifice does not disturb the measurement.
A limited number of miniature electronically scanned pressure (ESP)
sensors can be packaged inside the model or mounted at the base of the
75
model. A typical pressure sensor contains 32 transducers, an electronic
scanner, an analog amplifier, and an integral pneumatic switching valve. The
valve is used to switch the 32 transducers from either the model pressures
or to a common calibrate manifold. The pressure sensor modules (Fig. 7) are
nominally 2.5 in. long by 2.0 in. high by 1.0 in. wide with the internal volume
of each transducer being approximately 0.004 in. 3 Typical pressure ranges
are 2.5 psid full scale and 15 psid full scale. The resolution of the 2.5-psid
sensors is approximately 0.00036 psid/count, and the resolution of the 15-psid
sensors is approximately 0.002 psid/count. The accuracy of the sensors is
0.15 percent of full-scale pressure rating.
Flow-Field Diagnostics
77
number of the flow. The flow angularity sensitivity of the probe can be
expressed in terms of the pressure differential, AP13 or AP24, between two
opposite static pressure orifices. These probes have been very useful in
defining the flow field within hypersonic inlets where the flow is very complex.
80
Structures larger than those listed above would be too big for a high-
enthalpy airflow facility and typically are tested in a structures test facility
using radiant heating. A structures test facility (no-flow) is also best suited
to perform life cycle testing.
In contrast, airflow facilities can provide a better simulation of flight and
a more comprehensive evaluation of potential failure modes. An obvious
example is the case where the hot airflow leaks into the substructure and
exposes low-temperature elements to very high temperatures. In addition,
an airflow facility is needed to test components for which the primary failure
mode is related to aerodynamic heating, shear, acoustics, or vibration.
The next generation of hypersonic vehicles poses many technical challenges
in the development of structural components. Though exposed to severe
aero heating and aerodynamic loading, structural components should be
reusable, light, and maintainable. Structural features and typical aerothermal
flight issues of hypersonic vehicles are summarized in Fig. 17. Vehicle surfaces
will typically be exposed to very high heating, reaching temperatures ranging
from 2,000 to 3,5000F. At these extreme conditions, the structural design of
certain components like leading edges, control surfaces, protuberances, and
shock interference regions is complicated by the need for internal cooling. Joints
and seals must protect underlying cooling passages and storage tanks from
severe environments without degradation. Unfortunately, the ability to predict
the boundary-layer state accurately-laminar versus turbulent-continues to
elude the aerodynamicist. In the recent past, transition Reynolds numbers were
estimated from engineering correlations, and these techniques will probably
continue to be used for future hypersonic vehicles. A successful response to
the challenges of structural component design lies in the selection of materials
and the proper integration of engineering analysis and experimentation to
verify the survivability of the structure.
The key role of analysis cannot be overemphasized since the design options
are so varied. For example, aerothermal issues can be resolved by any (or
all) of the following: thermal protection systems (fPS), active cooling systems,
or passive cooling systems. The optimum design of any of these systems
requires many trade studies and verification by experimental data.
There are two fundamental phases in the development of structural
components: (1) definition of the flight thermal environment and (2)
demonstration of hardware survivability(see Fig; 18).
h = (3)
where {j comes from the solution of Eq. (1) since the left-hand side is known.
Prior to each run, the model is cleaned and cooled with alcohol and then
spray painted with Tempilaq. The model is installed on the model injection
mechanism at the desired test attitude, and the model initial temperature (Tj)
is measured. The model is then injected into the airstream for approximately
25 sec; during this time the model surface temperature rise produces isotherm
melt lines. The progression of the melt lines is photographed with 70-mm
sequenced cameras operating at two frames per second. Typical examples
82
of phase change paint photographs obtained during a run are presented in
Fig. 20.
A complete description of the phase change paint technique as applied
to a particular test situation is presented in Ref. 25.
Infrared Scanning - Thermal mapping techniques used in wind tunnel
test applications generally involve the use of heat-sensitive model surface
coatings. The major drawback to these methods has been the time required
to obtain quantitative data from photographic test results.
With an infrared (IR) scanning camera system, heat-transfer coefficient
data in the form of tabulations, plots, and surface maps have been produced
within minutes of test run completion. The typical IR camera is a scanning
optical-mechanical device that produces data in the form of an analog signal
rather than using film. The output signal is amplified, converted from analog
to digital form, and transferred to the computer. The analog signal is also
received by a video monitor that provides real-time displays of the developing
model surface temperature pattern. The digital data received by the computer
are reduced to coefficient form using the Stephan-Boltzman law to calculate
temperature and semi-infinite heating assumptions to calculate h. A more
complete description of the infrared system, the data reduction, testing
techniques, and presentation of sample test results can be found in Ref. 26.
Thennographic Phosphor Paint - Thermographic phosphor paints are
sensitive to temperature, and as temperature increases their luminosity
decreases. The luminosity of these paints is typically excited by UV or laser
light and as the temperature patterns develop, they are photographed or
recorded with a video camera. Thermographic phosphor paints have been
used both in wind tunnel testing and in turbine engine tests. In the wind tunnel
these paints are applied to models to measure surface temperature patterns
which can be used to infer heating rates. In turbine engine tests the phosphor
paints have been applied to turbine blades to infer temperature. Temperature
measurement with thermographic phosphors is practical from about 100 to
900°F.
83
Thermal radiation and heat conduction effects on the thin-skin element
are neglected in the above relationship and the skin temperature response
is assumed to be due to convective heating only. It can be shown that for
constant Tr' the following relationship is true:
to [ Tr - T·I ]
(8)
Tr - Tw
is linear with time. This linearity assumes the validity of Eq. (7), which applies
for convective heating only. Thus, if the data show anon-linearity, effects
other than convective heating are present. In most cases, the nonlinearity
will be caused by conduction effects. Machine plots of data from each
thermocouple provide the opportunity for quick visual examination of test
data with the objective of evaluating conduction effects. Once areas and/or
time frames during which significant conduction effects were present are
identified, the remaining valid data are used to calculate the heat-transfer
coefficient via Eq. (7).
A complete discussion of the thin-skin test technique as applied to a
particular test situation (including presentation and analysis of the resultant
data) is contained in Ref. 27.
Coaxial Gage - For applications where surface contour is critical or
where measurements at severe wall temperature conditions are required, the
coaxial thermocouple,gage (or coax gage) is often used. The coax gage (Fig.
22) is simply a surface thermocouple comprised of an insulated Chromel®
wire fixed concentrically within a constantan jacket. The thermocouple
junction is formed at the sensing surface by blending the two materials
together with a file. This filing process is also used to contour the gage surface
to match the model surface exactly. Because of its simple construction, the
coax gage can be made very small; gage diameters of 0.065 and 0.125 in.
are in common use. Also because of its construction, this gage has a very
large temperature operating range (- 3200 to 1,OOOOF and above) as compared
to other measurement techniques.
The data reduction equation used to obtain the heat flux from the
measured temperature of a coax gage is as follows:
84
where q(tn) = surface heat flux at time tn
Tw (t·) = surface temperature at time tj
C
The term (t n) is essentially the scale factor for each coax gage with the
thermal properties temperature dependence included.
The semi-infinite solid assumption used in the development of the data
reduction technique described above is valid for only a limited time since
the gage has a finite depth. Another factor which can limit the accuracy
and/or run time is lateral conduction effects. The combined thermophysical
properties of Chromel and constantan are quite similar to stainless steel. Thus,
if the gage is mounted in a steel model, the conduction errors usually become
negligible for run times less than 2 sec.
The time response of the coaxial gage is extremely small; typically, the
only limitation is the response and sensitivity of the signal recording
equipment. .
The coax gage technique is described for a particular test application in
Ref. 28.
Schmidt-Boelter Gage - The Schmidt-Boelter gage (Fig. 23) alleviates
many deficiencies found in the thermopile Gardon gage. Schmidt-Boelter
gages have seen considerably wider usage in recent years for heat-transfer
measurements in continuous-flow wind tunnels and flight test applications
(Ref. 29). This is primarily due to the attractive operating characteristics com-
mon to this type sensor. These include (1) excellent durability, (2) good sensi-
tivity, (3) self-generating output signal directly proportional to incident heat
flux, (4) continuous service temperature of 700oF, and (5) semicontourability.
The principle of operation of the Schmidt-Boelter gage is based on axial
heat conduction and involves measuring the temperature difference, ~T,
between two parallel planes on the top and bottom of a slab or wafer which
is backed by a heat sink, as shown in Fig. 23. This temperature difference
is generally measured with a differential thermocouple. The hot junction
temperature, TH, is on the top surface of the slab, and the cold junction
temperature, Te, is on the bottom surface. The material and thickness of the
slab can vary widely; the heat sink is usually a material with a high thermal
conductivity such as aluminum, copper, etc. Excellent sensitivity is achieved
by using a series thermocouple (thermopile) arrangement to detect the
temperature difference between the top and bottom surfaces of the slab. The
thermopile is constructed by winding 20 to 40 turns of small (0.002-in.-diam)
constantan thermocouple-grade wire around the anodized aluminum wafer.
One-half of the constantan coil is electroplated with copper, creating a multi-
element copper-constantan differential thermocouple. The steady-state output
signal of the transducer is proportional to the incident heat flux at the surface
(Co.
85
Experimental calibrations of Schmidt-Boelter gages are performed using
a radiant heat source as described in Ref. 30. These experimental procedures
enable a calibration scale factor (CSF) to be obtained for each gage, and
heating rates are calculated as follows:
q
h=--- (11)
Tr - Tw
h = q (12)
(Tr - Tw)
Equation (13) has the form of a straight line when hTr is assumed constant
(which is valid for this application).
(14)
86
Also note that h = - Al and setting q = 0 leads to the relationship for Tr,
(15)
87
The test article may range from a simple insulated material panel to a
complex structure with LH2 cooling passages. In addition, it may be necessary
to provide structural loading on the test article using hydraulic actuators
mounted under the wedge. If the test article incorporates backside cooling,
it is important to simulate both the aero heating (qin) and the heat removed
by the coolant (qout). Therefore, the mass flow, heat capacity, temperature,
and viscosity of the coolant must be duplicated in the ground test.
The use of analysis tools to design a test is illustrated in Fig. 27. An
"aerothermal response code" combines the material properties, the flight
trajectory, and other inputs with a heat conduction model of the test article
to provide a prediction of surface temperature versus time. A similar code
combines the results of the wedge calibration data (Le., qversus WA*) and
the facility flow conditions to produce the test article surface temperature
predictions during the wind tunnel test. In this manner, the wedge angle can
be adjusted until the temperatures agree as shown in the figure. For the test
article to reach the predicted temperature and temperature gradients, run
times of many minutes may be required. Consequently, impulse facilities
cannot be used for this type of testing.
In general, materials/structures testing can be grouped into the four
categories listed in Fig. 28.
Screening Test
Screening tests make relative comparisons among many candidate test
articles in a constant (or repeatable) test environment. It is desirable to
fabricate several identical test articles to investigate the repeatability of failure
modes.
Characterization Test
Characterization tests are the inverse of screening tests in that one test
article design is exposed to a variety of test conditions. For example, the test
variables may be temperature and load, and each is held constant while the
other is varied over the range of interest. The data can be developed into
an algorithm that characterizes the test article.
* WA - Wedge Angle
88
math model can be developed. The math model can be used to predict internal
component temperatures for a variety of flight profiles. The disadvantage
of this technique is that it is only valid for the specific component that is used
to generate the database. If the component design changes after the test, it
may be necessary to rerun the entire test for the new design.
Concluding Remarks
Nomenclature
90
15. Layne, T. C. and Bomar, B. W. "Discrete Fourier Transform Laser
Velocimeter Signal Processor." ICIASF '87 Record, International Congress
on Instrumentation in Aerospace Simulation Facilities, College of William
and Mary, Williamsburg, Virginia, June 22-25, 1987.
16. Crosswy, F. L. "Particle Size Distribution of Several Commonly Used
Seedling Aerosols." Wind Tunnel Seeding Systems for Laser Velocimeters,
NASA Conference Publication 2393, Hampton, Virginia, March 19-20,
1985, pp. 53-75.
17. Crosswy, F. L., Kingery, M. K. (AFSC/ AEDC) and Schaefer, H. J. Pfeifer,
H. J. (French-German Research Institute at Saint Louie, France). "Laser
Velocimeter Seed Particle Sizing by the Whisker Particle Collection and
Laser Aerosol Spectrometer Methods." AEDC-TR-89-3 (AD-A21091O), July
1989.
18. Crosswy, F. L. and Sherrouse, P. M. "Electro-Optically Slaved, Forward-
Scatter Receiver/Traverse System for Laser Velocimetry." SPIE Technical
Symposium Southeast, Orlando, Florida, May 17-22, 1987.
19. Heltsley, F. L. "Recent Experience in Seeding Transonic/Supersonic Flows
at AEDC." Wind Tunnel Seeding Systems for Laser Velocimeters, NASA
Conference Publication 2393, Hampton, Virginia, March 19-20,1985, pp
121-140.
20. Donaldson, J. C. "Laser Doppler Velocimeter Application in Supersonic
Boundary-Layer Flow." AEDC-TR-86-44 (AD-A178395), March 1987.
21. Trimmer, L. L., Matthews, R. K., and Buchanan, T. D. "Measurement of
Aerodynamic Heat Rates at the von Karman Gas Dynamics Facility."
International Congress on Instrumentation in Aerospace Simulation
Facilities, September 1973.
22. Matthews, R. K., Nutt, K. W., Wannenwetsch, G. D., Kidd, C. T., and
Boudreau, A. H., "Developments in Aerothermal Test Techniques at the
AEDC Supersonic/Hypersonic Wind Tunnels." Vol. 103, AIAA Progress
in Astronautics and Aeronautics, 1986.
23. Stallings, D. W., Matthews, R. K., and Jenke, L. M. "Recent Developments
in Aerothermodynamic Test Techniques at the AEDC von Karman Gas
Dynamics Facility." International Congress on Instrumentation in
Aerospace Simulation Facilities, September 1979.
24. Jones, Robert A. and Hunt, James L. "Use of Fusible Temperature
Indicators for Obtaining Quantitative Aerodynamic Heat-Transfer Data."
NASA-TR-R-230, February 1966.
25. Matthews, R. K. and Gilley, G. E. "Reduction of Photographic Heat-
Transfer Rate Data at AEDC." AEDC-TR-73-90 (AD-762928), June 1973.
26. Boylan, D. E., Carver, D. 8., Stallings, D. W., and Trimmer, L. L.
'Measurement and Mapping of Aerodynamic Heating Using a Remote
Infrared Scanning Camera in Continuous Flow Wind Tunnels." AIAA 10th
Aerodynamic Testing Conference, San Diego, California, April 19-21,
1978.
91
27. Carter, L. D. and Kaul, C. E. "Heat Transfer Tests on the Rockwell
International Space Shuttle Orbiter with and without Simulated
Protuberances." AEDC-TR-75-20 (AD-AOI2876), July 1975.
28. Wannenwetsch, G. D. and Martindale, W. R. "Roughness and Wall
Temperature Effects on Boundary Layer Transition on a 0.0175-Scale
Space Shuttle Orbiter Model Tested at Mach Number 8." AEDC-TR-77-19
(AD-A038895), April 1977.
29. Kidd, C. T. "A Durable, Intermediate Temperature, Direct Reading Heat
Flux Transducer for Measurements in Continuous Wind Tunnels." AEDC-
TR-81-19 (AD-AI07729), November 1981.
30. Kidd, C. T. "Determination of the Uncertainty of Experimental Heat-Flux
Calibrations." AEDC-TR-83-13 (AD-AI31918), August 1983.
31. Matthews, R. K. "A Summary Report on Store Heating Technology."
AEDC-TR-78-46 (AD-A059415), September 1978.
32. Wannenwetsch, G. D., Ticatch, L. A., Kidd, C. T., and Arterbury, R. L.
"Measurement of Wind-Leading-Edge Heating Rates on Wind Tunnel
Models Using the Thin-Film Technique." AIAA Paper 85-0972, June 1985.
33. levaIts, J. O. and Matthews, R. K. "Aerothermal Evaluation of High-
Temperature Structural Materials for Use in High Speed Missile Design."
AEDC-TR-79-38 (AD-A080023), January 1980.
34. Matthews, R. K. and Harper, D. C. "Aerothermal Tests of the Space Shuttle
External Tank Insulating Material." AEDC-TR-75-94 (AD-AOI7497),
November 1975.
35. Knox, E. O. "Thermal Response and Reusability Testing of Advanced
Flexible Reusable Surface Insulation and Ceramic RSI Samples at Surface
Temperature to 1200°F." AEDC-TR-79-62 (AD-A09771l), April 1981.
36. Wannenwetsch, G. D. and Matthews, R. K. "Prediction Techniques Used
in the M505A3E2 Fuze Auto-Ignition Investigation." AIAA-81-1041, June
1981.
92
VACUUM SPHERE
~ DIFFUSER
,......TEST CELL TRANSFER CART
TEST CELL WITH
_ MODEL SUPPORT SYSTEM
MACH 10 NOZZLE
GAS HEATER
PRESSURE CONTROL VALVES
FIGURE 1. Hypervelocity wind tunnel at NSWC.
SCHEMATIC
16 FT ++ 48.S FT
NOZZLE
23 FT---J
TEST SEmON-
RECEIVER TANK
I
96FT DlAM
HIGH
PRESSURE
STORAGE
ATMOSPHERE
VENT
I
TANK
GROUND ACCESS
flOOR flOOR
FIGURE 3. AEDC 50-in.-diam Tunnel B.
94
1.0
o
00
__
-
l
-cr-..ri - - -
TURBULENT
0 0
~ _f)
0
Oodntfo
0
000 0 - - - a
o 00
cP 0000
h
-h 0.1
ref
SYM Re oo , L TiTo
o 4.5 X 106 ",,0.53
o 6.6 X 106 ",,0.53
FLAG - BOUNDARY LAYER TRIPS "" 0.60
95
*GENERAL PURPOSE 6·COMPONENT BALANCES (12 AVAILABLE)
• NOMINAL BALANCE SIZES: LENGTH, IN. 5.1 TO 13.8
DIAMETER, IN. 0.6 TO 2.1
• NORMAL FORCE RANGES, LB: 20 TO 1,500
• SIDE FORCE RANGES, LB: 20 TO 700
• AXIAL FORCE RANGES, LB: 4 TO 300
96
1' - - - 2.2 ' - ,,, 3"
f@oooo o;oo~
' " C2 0 0 0 0 O? 0 0 c,
0000 0000 REFERENCE TUBE .OW
l.@oooo oooo@
PX-32 TUBES .040" or .063" CAl TUB .063"
0.10
0.08
1.00
0.06
0.80
0.60
0.50
0.40
"-
] 0.30
0.20
0.02
0.15
0.10
0.09
0.07
0.06 L...-_ _...L.-_L...-...l..-...I-.J..-...J........J
1.2 1.5 2.0 3.0 4.0 5.0 6.0 7.0 8.0 1 4 5 6 8 10
FREE·STREAM MACH NO. FREE· STREAM MACH NO.
Typical data for probe flow Typical data for probe Mach
angularity sensitivity number calibration
FIGURE 10. Mach-flow angularity probe.
98
BEAM SPLITTER II TRANSMITTING LENS
IMAGE OF FRINGES BOUNDARY LAYER
ON SLIT As
I~~ I
I- l------r
112 AS
BOW SHO(K-------~
DIRECTION OF -
FRINGE MOVEMENT
6 '
6o 5
!!54
:e
--'
3
-~
<1
2 f---t'-IHIrllth-rr-t-'L..!..Ilf"WllffuLiI\l
If--+-~~~~~~~
O~~~~~~~~~~~~
FLOW
WINDOW .+ WINDOW
. . ....
' .....
.', :. : CONTAMINANT PARTICLES': .... ".: '.
I • , . '. . PARTICLE
:
~ '. : , . .' '. . '. DETEaJON
PULSES
BEAM-1F7~~~~~~~~~~tE~~~till:U
BLOCK . . . . .
.... " ",
.' . ' .' ..... .
!l_!~---, TUNNEL A ~~
-'C 6. I LDV SURVEYS ===
:::::::::::=;
~
EXTERNAL FLOW
c::::::>
MI- 2.0
MISSILE
NOSE
100
PARAMETER
TECHNIQUE STATUS ADVANTAGES DISADVANTAGES CONCERNS
MEASU RED
Intrusive !
Pitot Probe FO Pitot Pressure, • Classical direct • Error induced if AOA > 20 • Pressure tubing length
Pp measurement deg (responses time)
• Many, many years of • Measurements in boundary
experience layer can be distorted if
• Very simple close to wall
• Relatively insensitive to • Small probe diameters
AOA desired (0.010 - 0.034) are I
(i.e. s 20 deg) difficult to fab and are
delicate
Total Temp Probe FO Local flow total • Direct measurement • Error induced if AOA > 20 • Radiation effects (shielded
temperature • Many years of experience deg vs unshielded)
§ T, • Relatively simple • Local flow in BL can be
• Relatively insensitive to distorted if close to wall
AOA • Small probe diam desired
(i.e. < 20 deg) are difficult to fab and are
delicate
Mach/Flow FO Inferred M p , and • Provides basic aerodynamic • Requires extensive M p, Reo> • Measurement outside
Angularity Probe local flow angle information calibration calibration boundaries
• Simple pressure • Fabrication difficult • Pressure response time
measurements • Relative large size - might • Alignment of probe
• Years of experience distort local boundary layer
Legend: FO - Fully Operational ID -In Development AOA -Angle of Attack BL - Boundary Layer
FIGURE 16. Summary of flow-field diagnostics techniques used in the AEDC hypersonic facilities.
PARAMETER
TECHNIQUE STATUS ADVANTAGES DISADVANTAGES CONCERNS
MEASURED
Nonintrusive
Boundary-Layer FO Boundary layer • Simple setup and operation • Viable only in model vertical • Laser safety
Transition Detector state inferred • Performance comparable plane I
I
(BLTD) from optical with conventional means
density • Extensive applications
fluctuation experience
Laser Particle 10 Particle flux • On-line data • Competes with schlieren, • Laser safety I
Monitor (LPM) • Simple setup and operation cameras, etc. for optical • Vibration-induced
access misalignment
Laser Doppler 10 Local velocity of Measures:
Velocimeter (LDV) particles • 1, 2, or 3 velocity • Costly primary and support • Laser safety
components with ± sense equipment • Particle dynamics
• Instantaneous samples of U • Equipment heavy and bulky • Data biased by large
from which other velocity • Requires large area optical contaminant particles
o parameters are computed access
N • Close (0.005 in.) to model • Setup time significant
surface • Gas velocity information
• Small sensing volume (S.O inferred from particle
x 10·6in.3) velocity measurements
• Experience in aerodynamic
and aeropropulsion test
environments
Laser-Induced 10 nand T of 02, • Direct molecular energy • Costly primary and support • Signal confusion with laser
Fluorescence (LIF) NOand H2 state and number density equipment radiation scattered by
measurements • Setup time significant contaminant particles
• Requires UV test facility • Limited application I
• Instanteous sampling of
number density and windows experience
temperature • Through validation of laser
• IPoint and planar capability operating characteristics
necessary
• Handling of large data
volume
• Background radiation from
luminsecent flows
-------- _.. _--- ----
INSULATION
~---~--~-~=-------_/
PROVIDES: • CODE VERIFICATION ~ EXTRAP. TO FlT.
• HEATING INPUTS (q)
• THERMAL ENVIRONMENT
(i.e., WHAT HEATING RATESITEMPERATURES ARE ENCOUNTERED IN FLIGHT?)
PHASE 2 • DEMONSTRATE HARDWARE SURVIVABILITY (STEPS 2!3!~)
• DUPLICATE LOCAL
ENVIRONMENT
~ TEST ARTIClES
(i.e. q LOCAL~ FLY)
STEPS
(2) SElECT MATERIAL
(3) TEST STRUCTURAL DESIGNS
(~) PERFORM FLIGHT HARDWARE VERIFICATION TESTS
103
ADVANTAGES DISADVANTAGES REF.
THERMAL MAPPING
• PHASE·CHANGE PAINT VIVID ILLUSTRATION OF HOT SPOTS MUST REAPPLY PAINT, DATA 25
HIGH SPATIAL RESOLUTION PRESENTATION CAN BE CONFUSING
• IR SCANNING CAMERA COMPUTER-GENERATED PLOTS AND SPATIAL RESOLUTION 26
COLOR MAPS, AND NON INTRUSIVE
• THERMOGRAPHIC PHOSPHOR COMPLETE MODEL, GOOD SPATIAL MODEL PREPARATION AND DATA 21
RESOLUTION PRESENTATION
DISCRETE MEASUREMENTS
• THIN-SKIN HIGH QUALITY DATA, DENSE SPACING EXPENSIVE MODEL FAB, 27
CONDUCTION EFFECTS
• COAX GAGE EASY TO INSTALL, CONTOURABLE, DURABLE LOW OUTPUT, SHORT TEST TIMES 28
• SCHMIDT -BOELTER GAGE HIGH OUTPUT, SLIGHTLY FAB AND CALIBRATION TIME 22
CONTOURABLE, VERY DURABLE REQUIRED
• GARDON GAGES YEARS OF EXPERIENCE, FAST RESPONSE GAGE ATIRITION RATE, NOT 31
{HI TEM~ LOW TEMP) CONTOURABLE
• HIN-FIL DENSE SPACING, FAST RESPONSE, CAN RELATIVELY DIFFICULT INSTAL- 32
BE USED ON SMALL RADII LATlON, MATERIAL CRACKING
I
._t· ....
"1\l'-hIni
:.' 0
.' U,"
EPOXY
COPPER WIRES
o ANODIZED
ALUMINUM
HEAT SINK
SECTION A-A
FIGURE 23. Section drawing of 3/16-in.-
diam Schmidt-Boelter gage.
105
r
I-f-
SENSING FOIL
0.25 IN.---1
T
O.OOI-IN.-DIAM AIR SPACE
COPPER WIRE --t..r-~
""",,-mI"l"7"rT"4............<
SOLDER --",.:a.'-oI_~
EPOXY POTTING
J--->,--- MA(HINABLE
CERAMI( SUBSTRATE
(MACOR®)
~-.-r~--(QPPER
LEAD WIRE
(HROMEL® PINS
(0.010 IN. DlAM)
ANALYSIS EXPERIMENTS
• ESTIMATE FLIGHT ENVIRONMENT • FLIGHT COMPONENT (OR LARGE
• ENGR CODE) SIZE REPLICA) EXPOSED TO
• CFD SIMULATED FLIGHT ENVIRONMENT
• THERMAL RESPONSE CODES IN TERMS OF TEMPERATURE, HEAT
(TEMPERATURE) RATE, SHEAR, PRESSURE AND LOADS
• STRUCTURAL LOADS • AEDC - APTU
(STRAIN) • AEDC AEROTHERMAL TUNNELS
• LaRC 8-FT TUNNEL
• AMES 3.S-FT TUNNEL
• ARC FACILITIES
• AMES, AEDC, MDAC, JSC
"' ; . 1
FLIGHT ENVIRONMENT
SURVIVABILITY IS AFUNCTION OF:
• WALL TEMPERATURE
• SHEAR
• THERMAL SHOCK ~ . ~/
• HEATING RATE
• ACOUSTIC ENVIRONMENT
• SPAlLiNGIEROSION
• PARTIClE IMPACT
• PRESSURE TEST INSTALLATION
SURFACE TEMPERATURE
PREDICTED FOR FLIGHT
SURFACE
TEMPERATURE
OBTAINED
DURING WIND
TUNNEL TEST
(TYP.)
o 60
108
Wind-Tunnel Based Definition of the APE
Aerothermodynamic Environment
Charles G. Miller
M.S. 408 EAB/SSD
NASA Langley Research Center
Hampton, VA 23665
W.L. Wells
M.S. 408 EAB/SSD
NASA Langley Research Center
Hampton, VA 23665
ABSTRACT
To date, NASA's primary experience with reentry flight centers about the
Apollo vehicle, which plunged through the Earth's atmosphere upon return from
the moon, and the Space Shuttle orbiter which reenters along an altitude-
velocity trajectory quite different than that for Apollo. Measurements performed
on the orbiter during reentry (Le., orbiter experiments (OEX) and definition flight
instrumentation (OFI» have provided the aerodynamic!aerothermodynamic
community with extremely valuable data that provides insight to hypersonic!
hypervelocity (high enthalpy or real-gas) flow phenomena through the rarefied
to continuum regimes. Unfortunately, relatively little flight information was
monitored on the Apollo vehicles during reentry. Because proposed ASTV's
would have different trajectories than the orbiter (higher velocities at higher
altitudes) and Apollo (similar velocities, but at different altitudes), there is a
dearth of flight information to serve as a building block for the design of an
ASTV.
no
As discussed by Walberg 1982, a number of shapes providing a rela-
tively wide spectrum of lift-to-drag ratio (UD) have been proposed for ASTV's.
Concepts range from drag brakes having near zero lift to slender, asymmetric
configurations such as a bent-nose biconic having a UD near unity. On the low
UD end of the spectrum, rigid aeroshe"s using lift modulated control,
deployable flexible aerobrakes, and inflatable devices using drag modulated
control have received attention. One common factor about these low UD ASTV
concepts is that they are quite large, varying from about 50 ft in diameter for
GEO or lunar return to over 100 ft in diameter for Martian return. Existing
hypersonic ground-based facilities are not capable of duplicating the enthalpy-
density flow environment of ASTV's and certainly cannot accommodate full-
scale models. Computational fluid dynamic (CFD) codes and coupled
inviscidlviscous codes can provide predictions of the flow environment for
continuum flow and direct simulation Monte Carlo (DSMC) or particle kinetic
techniques can predict the environment for the free-molecular regime. How-
ever, these codes are, in many cases, uncalibrated by experimental ground-
based data and/or unvalidated by flight data.
The ASTV designer, who has goals to reduce cost and risk, is con-
.fronted with relatively large uncertainties associated with the capabilities of
ground-based facilities and CFD codes to address the flow environment for
ASTV's everywhere along the flight trajectory. Because the high-velocity, low-
density flow environment cannot be duplicated or adequately simulated in
present test facilities nor accurately predicted by existing computational
techniques, a precursor sub-scale ASTV flight experiment was proposed and
successfully advocated within NASA. (The primary thrust for this proposal was
provided by the NASA Johnson Spacecraft Center (JSC).) This precursor is
referred to as the aeroassist flight experiment (AFE). The AFE will provide the
aerodynamiclaerothermodynamic information necessary to embark on the
design and construction of ASTV's with a higher level of confidence, particularly
needed for manned-rated vehicles. More specifically, the primary objectives of
the AFE are to: (1) resolve radiative heat transfer issues, particularly those
associated with thermal and chemical nonequilibrium flows; (2) determine the
effects of wall catalysis; (3) assess aerodynamics and control for high velocity-
altitude trajectories; (4) develop/demonstrate thermal protection system (TPS)
materials; (5) define wake flow characteristics; and (6) provide a benchmark
data set for the validation of .CFD codes for a highly three-dimensional
configuration over a wide range of flow conditions. To this end, 12 onboard
experiments will be performed to provide information on the aerodynamic
characteristics, aerothermodynamic environment, and material and structure
111
response to the aerodynamic loads and heating. Thus the emphasis on the
science side of the experiment is to accurately monitor the flow environment
and the state of the vehicle in terms of structures and materials; on the project
side, the emphasis is to successfully fly the vehicle, have it survive, and be
recovered. (The AFE project is managed by the NASA Marshall Space Flight
Center (MSFC).)
Systems studies for AFE performed in the mid-1980s (e.g., Roberts 1985,
Gamble et al.,1984) evolved towards a vehicle having a rigid (as opposed to
flexible), blunt forebody with a low ballistic coefficient (lift-to-drag ratio of about
0.2 to 0.3), roll controlled (Le., vehicle rolled about longitudinal axis to change
lift direction), non-ablative heatshield, and made as large as possible for
delivery to orbit by a Space Shuttle orbiter. To minimize cost, systems
developed for and used by the Shuttle orbiter would be used wherever
possible, including the thermal protection system (thus, the surface temperature
for the AFE vehicle could not exceed 3360°R.) A blunted, raked-oft elliptiC cone,
similar to configurations studied in the 1960's at the NASA Langley Research
Center (LaRC) (Bernot 1965, Mayo et aI., 1965, and Molloy 1966), was selected
for the aeroshell. A fast-paced parametric aerodynamic study was performed in
several hypersonic wind tunnels at NASA LaRC to assess the effect of nose
bluntness and radius and arc-length of the skirt. The baseline configuration
emerged from this parametric study which was directed by NASA JSC. Since
that time, over 700 wind tunnel runs have been made by NASA LaRC and
several CFD codes have been exercised for this configuration.
ll2
279,000 ft altitude, the velocity will be 31,700 ftlsec and perigee will be
achieved between 237,000 and 263,000 ft. A 30 sec quiescent or quiet period
will be provided prior to perigee in which all control rockets will be shut off.
Certain onboard experimental measurements will be made during this quiet
period. The vehicle will be rolled after traveling for some period at the minimum
altitude to begin its exit and retum to LEO. During this exit period, guidance and
control become extremely important as the dynamic pressure begins to
decrease and the uncertainty in the atmospheric density profile increases.
Once in LEO, it will rendezvous with the orbiter, be retrieved and stored back in
the cargo bay, and returned to Earth for in-depth post-flight analysis. This
retrieval and post-flight inspection is a prime requirement for the mission
because reusability without TSP refurbishment is a key feature of ASTV's. (It
should be noted that only one AFE vehicle will be constructed and this vehicle
is scheduled to be flown one time in 1994.)
The AFE vehicle will experience free molecular to continuum flow. At the
entry interface, the mean free path will be the order of the diameter of the
vehicle. Thus, a shock will not form about the vehicle until it descends to lower
altitude. At the rarefied flow condition, the contribution of shear to aerodynamic
coefficients relative to the pressure contribution will be larger than at the lower
altitude, continnum flow condition. Consequently, a Significant change in
aerodynamic characteristics may occur during the flight. The vehicle will
transition from the free-molecular to the continuum regime as the altitude
decreases, eventually encountering the period of peak heating and dynamic
loads. This period is sometimes referred to as the "golden period" by the AFE
scientific community since the onboard experiments will provide a wealth of
information during this time. As one example, for the period that non-
equilibrium flow occurs within the boundary layer, many atoms (produced by
dissociation immediately behind the shock) will reach the vehicle surface and
find a partner, recombining to release energy. If the surface enhances this
recombination process (i.e., is catalytic), the convective heating may increase
by a factor of two compared to a noncatalytic surface. It should be noted that
even at the high velocities of AFE, radiative heating for this sub-scale ASTV will
be less than convective heating. (At high velocities, the radiative heating
increases with body nose radius, whereas the convective heating decreases
with nose radius.) Although the AFE is quite blunt, the equivalent nose radius
(which is relatively small compared to proposed low UD ASTV concepts) and
range of velocity are such that convective heating will dominate everywhere
along the trajectory.
ll3
overview of the ground-based program, including the rationale and the
approach taken, description of the principle facilities (i.e., ·workhorse· facilities)
used in the study, models, measurement techniques, data reduction procedures
and samples of the results. A brief discussion of future test plans is also
presented.
SYMBOLS
d model base height in symmetry plane, inch (fig. 2) (L is also used for
base height)
114
M Mach number
p pressure, psia
T temperature, OR
U velocity, ftls
P density, Ibm/ft 3
Subscripts:
aw adiabatic wall
B balance
d diameter
115
ref reference value
RP rake-plane center
stagnation conditions
w wall conditions
00 free-stream conditions
AFE CONFIGURATION
Section Preface
116
Top view
~------14ft------~
Instrument
carrier
o ~ROCketmotor
Front view
Side view
117
Elliptic II
•
cone\/
I
I
I
Ellipsoid
I
I
:(7(- --,----
I
:=
00 /" I
Flow I
I
I
I
I Cone
I region
I
I
Rake plane
Figure 2. Development of AFE configuration from original elliptic cone. Symmetry plane shown.
Facilities
Langley 2Q-lnch Mach 6 CF4 Tunnel. The Langley 20-lnch Mach 6 CF4
Tunnel is a blowdown wind tunnel that uses tetrafluoromethane (CF4) as the
test gas. The ratio of specific heats of CF4 is approximately 20 percent lower
than air. The CF4 is heated to a maximum temperature of 15300 R by two
molten-lead-bath heat exchangers connected in parallel; maximum pressure is
2500 psia. Flow is expanded through an axisymmetric, contoured nozzle
deSigned to generate a Mach number of 6 at the 20-in.-diameter exit. This
facility has an open jet test section with two 24- by 30-in. clear tempered glass
windows on opposite sides. Run duration can be as long as 30 sec, but 10 sec
is typical for most tests. A detailed description of the CF4 tunnel and calibration
results are presented by Midden and Miller 1985. The calibration data revealed
a disturbance in pitot pressure focused on,the nozzle centerline. A 4-in.-
diameter hemisphere pressure-distribution model tested 3 in. off centerline
produced excellent agreement with theory. However, when pOSitioned on the
nozzle centerline, the sphere pressure distribution was degraded substantially.
119
Models
Force and Moment.- Aerodynamic models fabricated and tested are iden-
tical except for size; the base heights d (fig. 2) at the symmetry plane were either
3.67 in. (2.2 percent scale) or 2.50 in. (1.5 percent scale). A photograph of one
of the models is shown in figure 3 along with a sketch that provides information
pertinent to the aerodynamic tests. (Notice that the moment reference center is
at the rake-plane center. This position is convenient to locate in computer
codes. The flight vehicle center of gravity may be somewhat aft of this position.)
Each model is made in three parts: a stainless steel forebody (aerobrake), an
aluminum afterbody (instrument carrier and propulsion motor), and a stainless
steel balance holder. The forebody was machined to the design size and
shape within a tolerance of ±O.003 in. The balance-holder axis is parallel to the
Original cone axis. Although stainless steel subjects the balance to a greater
tare weight than aluminum, steel was chosen as the forebody material because
of its lower thermal conductivity and resistance to abrasion. Heat penetration
through the blunt, shallow forebody can result in thermal gradients across the
balance sensing elements, thereby compromising the calibration. The
balances were water cooled and provision was made for an air gap between
the balance forward end and the holder cavity surface to further reduce heat
transfer. The first models were fabricated with a cylindrical instrument carrier
and a simulated propulsion motor. Because of a redesign in the carrier to a
hexagonal shape to better accommodate instrumentation attachment, more
recent models were fabricated with the new carrier shape.
Two shrouds (fig. 4) were built to shield the balance from base-flow
closure. The first shroud is used when the afterbody is attached and the second
when the afterbody is removed. The shrouds attach to the sting, and clearance
is provided to avoid interference with the balance and model movement when
forces or moments are applied.
120
(a) Photograph
Balance
Flow moment
center
Z
Reference
moment
center
(Rake-plane center)
(b) Sketch with reference moment center and transfer distances X and Z identified.
Figure 3. AFE force~and-moment wind-tunnel model. a=O°.
121
For use
without afterbody
Afterbody
For use
". .. ".
. ,. . . . :. with afterbody
Figure 4. Photograph of APE wind-tunnel model (oblique aft view) with two balance shrouds.
Figure 5. AFE pressure model in wind tunnel.
123
cI>~
00
o
315~ I
o
-A-
290--............
17 Or--
T 3.67 in.
1.092 In.
I DIAMETER
250~ + _L
t::s
.j:>.
_t_
180 0
Thermal map.- Molds were made from the force and moment models and
thermal mapping models were cast with these molds using stycast. Models
were built to provide qualitative forebody heating distributions and data on a
cylinder in the near wake region.
125
.: • • ') ,. • 8 'J -=
'p,, 1 1001 '!"~
' l-! .
F \ •
N
0'1
"!
location
RTVrubber 0
",----seal U
I I I I I I I i I
3.67
it
N
00
stycast
~WireleadS
forebody 0.25 p 0.88-diam.
1.01-diam. Macor cylinder locknut
Aluminum cylinder/cone section
2.20
Figure 9. Sketch of AFE wake-heating-study model showing resistance gage inserts partially removed.
Dimensions are in inches.
(J
~----1'"
''-
:.:::~
....,......" "
.
.' -
""
I
90 ~ . :.
\", j / ' _~ A
~. ~-
N
. ., ' . ~.~,,\~\"\l\~
\0
Section A.-A
1.~~,~1.""
Figure 10. Photograph of AFE wake-heating-study model. Location of heat-transfer gage array.
Section A-A shows other possible locations.
Surface streamlines and wake-flow impingement.- The force-and-
moment models and the thermal mapping models were used with the oil-flow
technique to visualize forebody surface flow directions and impingement of
base-flow closure on surfaces in the wake region.
Instrumentation
Heat-Transfer.
130
thin-skin models, a geometric correction factor must be applied to the measured
skin thickness. This correction factor is, naturally, small for large radius of
curvature such as on the cone section of the AFE models, but can be quite large
for the skirt region (fig. 2).
131
at any given time. Estimates of the heat transfer rate can be made using the
thermophysical properties of the model and the time required to reach the
phase change temperature. This time is obtained using a camera and high
intensity electronic stroboscopic lamps to illuminate the model. These lamps
are synchronized with the camera shutter and the duration of a single flash is
only 25 J.lSec to prevent the lamps from imparting significant energy to the
model surface. The camera shutter speed generally is 30 frames per sec.
Usually, paints having melt temperatures between 5600R and 7100R are used,
although higher temperature paints may be required in some instances.
Test Conditions
Angles of attack were varied from 10° to -10° and angles of Sideslip
varied from 5° to -5°. (The AFE vehicle is expected to fly at _5° Sa S 5°.)
Force and moment. Each of the three test facilities has a dedicated
stand-alone data system. Output signals from the balances were sampled and
digitized by an analog-to-digital converter and stored and processed by a
computer. The rates at which the analog signals were sampled were 50 per
second in the CF4 and Mach 10 air tunnels, and 20 per second in the Mach 6
132
Table I. - Nominal test conditions.
Re oo Pt Tt Poo Too Moo Uoo qoo P2/Poo Re2 Pt,2 T t,2 "(2
( x \06/fl ) Obf/in 2) (OR) (I hf/in 2) (OR) . (fl/s) (Ihr/in 2) . ( xlO51fl) (lbf/in 2) (OR) -
0.25 150 1800 0.0045 97.4 9.55 4624 0.29 6.0 0.211 0.54 1800 1.34
0.54 350 1835 0.0095 96.9 9.74 4679 0.63 6.0 0.447 1.17 1835 1.34
1.09 700 1810 O.ot75 91.7 9.90 4651 1.20 6.0 0.866 2.23 1810 1.34
c:; 2.10 1450 1830 0.0324 90.5 10.05 4689 2.29 6.0 1.626 4.25 1830 1.34
w
0.63 30 845 0.023 108.3 5.84 2975 0.54 5.2 1.00 1.00 845 1.40
2.21 126 910 0.084 112.5 5.94 3095 2.10 5.2 3.52 3.86 910 1.40
0.30 970 1160 0.026 300.0 6.24 2850 0.63 11.8 0.97 1.23 1150 1.11
0.46 1500 1160 0.039 292.0 6.29 2844 0.95 11.7 1.48 1.86 1150 1.11
air tunnel. A single value of data reported herein represents an average of
values measured for 2 sec in the CF 4 and Mach 6 air tunnels, and for 0.5 sec in
the Mach 10 air tunnel. Corrections were made for model tare weights at each
angle of attack and for interactions between different elements of the balances.
Corrections were not made for base pressures. Output Signals were related to
forces and moments by a laboratory calibration that is accurate to within ±O.S
percent of the rated load range for each component. The moments about the
model rake-plane center have greater uncertainty than those measured at the
balance moment center. The pitching moment at the balance has only the ±O.S
percent of rated-load uncertainty, whereas the moment at the rake-plane center
also includes uncertainties associated with the forces included in the transfer
equation. «Pitching moment)RP = (Pitching moment)B -(X)(Normal force) -
(Z)(Axial force) where the subscripts RP and B denote the rake-plane center
and the balance moment center, respectively. The transfer distances X and Z
are illustrated in fig. 3.) Balance related uncertainties are tabulated in table II.
Heat Transfer.
Thin skin.- Values of heat transfer rate were inferred from the one-
dimensional heat conduction equation assuming radiative heating to be
negligible (Miller 1981). Thus, the heat transfer rate is directly proportional to
the density and specific heat of the thin skin material, skin thickness, and the
measured variation of inner surface temperature with time. Primary contribu-
tions to uncertainties in values of heat transfer inferred in this manner are: (1)
lateral and/or longitudinal heat conduction within the thin skin such that the
assumption of one-dimensional behavior is not valid, (2) variation of the heat
capacity with temperature, (3) measurement of skin thickness (e.g., an error of
0.001 in. corresponds to a 4 to 5 percent error in the inferred heat transfer rate),
(4) correction for effective thickness (Miller 1981) and (5) uncertainties in the
measured temperature time history. Of particular concern is the determination
of zero time; that is, the time when the outer surface first experiences
aerodynamic heating. All things considered, the uncertainty in heat transfer
inferred from this technique is believed to be ±~ 0 percent on the nose and cone
region of the AFE model and ±15 to ±20 percent on the skirt.
134
Table 11.- Balance related uncertainties in experimental longitudinal aerodynamic
coefficients.
Uncertainly
Moo = 6, air
Moo=6,CF4
136
Model at test •
stream center
·
q
137
The geometric stagnation point was used to define the origin of the coordinate
system. An optical sighting device was used to locate and record approximately
70 points along each shock, corresponding to a step size of approximately 0.06
in. on the photograph. The silhouette of the model symmetry plane was also
digitized from the schlieren photograph and recorded in the same manner as
the shock and in the correct relation to the shock. The digitized data from each
photograph were stored in an individual computer file and later plotted by a
graphics plotter. The accuracy of this process for the smallest standoff distance
(near the stagnation point) is believed to be within approximately ±5 percent
and improves with increasing standoff distance.
PREDICTION
Predictions presented herein were obtained with the HAllS (High Alpha
Inviscid Solution) computer code (Weilmuenster and Hamilton, 1986) and/or the
LAURA (Langley Aerothermodynamic Upwind Relaxation Algorithm) computer
code (Gnoffo, 1989). The HAllS code is a time-asymptotic solution of the Euler
equations, where the solution space is the volume between the body surface
and the bow shock wave that is treated as a time-dependent boundary. The
code will handle arbitrary perfect gases (constant ratio of specific heats) or real
gases in thermodynamic equilibrium. Test-stream flow conditions were used as
inputs to HAllS. For the CF4 computations, the program was modified to
include the thermodynamic properties of CF4. The wind-tunnel and numerical
model geometry were the same except for the region downstream of the aft
corner. The numerical model was modified to prevent the onset of
computational instabilities due to expansion of flow around the aft rim of the
skirt. The modification was a cylindrical extension rearward from the skirt.
138
To date, over 700 wind tunnel tests corresponding to over 23,000 data
points, have been performed in support of the AFE ground-based program.
With only a few exceptions, all tests have been performed in the Langley
Hypersonic Facilities Complex. Several tests have been performed in the Ames
Research Center ballistic range providing measurements of aerodynamic
characteristics and flow visualization of free-flight models at hyperveloc-
ity conditions. A few force and moment tests were performed in the Naval
Surface Warfare Center Tunnel 8A at Mach 18 in nitrogen with limited success.
Quite recently, detailed heat transfer distributions and shock shapes were
measured on 1.5-in.-diameter and 2.5-in.-diameter AFE models in the NASA
Hypulse facility (formerly the Langley Expansion Tube (Miller 1985)) at the
General Applied Science Laboratory (GASL). These measurements represent
the first ground-based hypersoniclhypervelocity aerothermodynamic data for
AFE and follow a calibration of Hypulse specifically for the AFE tests. Heating
rates were measured in helium, which behaved ideally (i.e., density ratio equal
to 3.7) even though the free stream velocity was 21,000 ftlsec, air at a velocity
equal to 17,000 ftlsec and corresponding density ratio of 11.5 (indicating
significant dissociation within the shock layer), and C02 at a velocity equal to
16,000 ftlsec and density ratio of 19.
Tests scheduled for the remainder of calendar year 1990 include mea-
surement of forces and moments and pressure distributions in the Wright
Research and Development Center Hypersonic (Mach 12 and 14) Wind Tunnel
and heating distributions and flow visualization at hypersoniclhypervelocity
(velocities in excess of 20,000 ftlsec) free stream conditions provided by the
Australian National University piston-driven shock tunnel T3. A second entry
into the Hypulse may be possible to test models having catalytic and
noncatalytic surfaces at conditions corresponding to chemical nonequilibrium
within the shock layer.
139
The synergistic relationship of the Mach 6 and 10 tunnels allowed the
effects of compressibility (i.e., Mach number) to be determined for a given free
stream unit Reynolds number and similar values of density ratio. Reynolds
number at Mach 6 and 10 was varied by an order of magnitude to assess
viscous effects for continuum flow. The flow over the forebody is believed to be
laminar for all tests, whereas the shear layer formed as the boundary layer
separates from the surface at the base may be transitional or turbulent.
The 20-lnch Mach 6 CF4 Tunnel provides NASA and the country with a
unique test capability. It is generally acknowledged that the effects of Mach
number on the inviscid flow field will be small for hypersonic (Moo> 5) blunt
bodies (Le., configurations for which the majority of the flow field about the
windward surface is subsonic (M2 < 1)). This is generally referred to as the
Mach number independence principle (Hayes and Probstein 1959). Similarly,
the effects of unit Reynolds number are expected to be small on the windward
flow field of hypersonic blunt bodies in continuum flow primarily because of the
corresponding thin boundary layer; this is not to imply, however, that Reynolds
number effects will be small in the near-wake region. The primary simulation
parameter that influences the inviscid flow field of blunt bodies is the normal
shock density ratio. During hypervelocity flight, molecules passing through the
shock about a blunt vehicle become vibrationally excited and for sufficiently
high enthalpy levels, dissociate to form a mixture of atoms and molecules within
the shock layer. In so doing, an appreciable amount of energy is placed in
dissociating these molecules, resulting in a decrease in the temperature as
compared to the case of no dissociation. This phenomena, generally referred to
as a real gas, does not influence the pressure behind a normal shock
substantially. Thus, the decrease in temperature behind the shock results in a
corresponding increase in density. Now, considering a small area of the bow
shock, the mass flow into this area (Le., product of free stream density and
velocity) remains constant. However, the density immediately behind this area
of the shock has increased substantially due to dissociation, whereas the post
shock velocity is relatively independent of real-gas effects. From continuity of
mass conSiderations, the shock detachment distance from the surface will
decrease significantly for a real gas (e.g., Jones and Hunt 1969, Miller 1976,
and Miller 1977). Thus, density ratio is an extremely important parameter in the
study of real gas effects for hypersonic blunt bodies. During entry into the
Earth's atmosphere, real-gas effects may result in density ratios as high as 18
which is a factor of three higher than that provided by conventional-type
hypersonic wind tunnels using air as the test gas. However, by using a gas
having a low ratio of specific heats, relatively large values of density ratio can
be achieved at hypersonic conditions and relatively low enthalpy (Jones and
Hunt 1969). This is the case for the Langley CF4 tunnel, which simulates the
large density ratio aspect of real gas by providing a ratio equal to 12 at Mach 6.
(This is not to infer that the chemistry of a real gas is duplicated.) It is for this
reason that the CF4 tunnel has been a workhorse for the AFE ground-based
program.
140
Samples of shock shape, force and moment, pressure distribution, and
heat transfer distribution measurements are presented here to illustrate the
effects of Mach number, Reynolds number, density ratio, and angle of attack.
With one exception, the results presented herein will be for forebody flow,
although a number of tests have been performed to examine near-wake flow
characteristics. The reader is referred to Mico11987, Mico11989, Wells 1987,
Wells 1988, Wells and Frank 1988, Wells 1989, and Wells 1990 (NASA TP
2954 and 2956) for a more detailed presentation of AFE tests performed to date.
Shock Shapes
Typical Schlieren photographs for tests in CF4 (P2/Poo = 11.7) and in air
(P2/Poo = 5.2) are shown in figure 12. These photographs illustrate that a factor
of 2 increase in density ratio significantly decreases the shock detachment
distance. At a = 0°, the shock detachment distance at the stagnation point in
CF4 is less than half the distance in air. In CF4, an inflection in the shock is
observed near the ellipsoid-cone juncture, which indicates a flow overexpan-
sion process. This inflection becomes more pronounced with decreasing angle
of attack and implies an overexpansion should be expected in measured
pressure distributions over the AFE fore body at these conditions. (To be
discussed subsequently.) The effect of angle of attack on shock shapes in
Mach 6 air is summarized in figure 13 and reveals the detachment distances
are greater (over most of the body) at a = 10° and decrease as a decreases to
-10°. This is expected because the body presents a more blunt cross section to
the oncoming flow as a increases and at a = 10° appears similar to a flat-faced
cylinder where the flow is subsonic everywhere over the cylinder face (sonic
corner). No effect of Reynolds number on shock characteristics were observed
for the range available in these tests.
141
~
~
i§
(a) CF4 test gas: pJpoo=I1.7. (b) Air test gas: pJpoo=5.2.
-
Figure 12. Typical schlieren photographs for APE at Mach 6 and a=O.
.8
.4
-.4
y/d
-.8
-1.2
(1, deg
10
-1.6 o
-10
-2.0 L---L.----IL--L.......J_...L........I._..L---L_L..-L._L---L.----IL--L.......J.-J
-1.6 -1.2 -.8 -.4 o .4 .8 1.2 1.6
x/d
143
Shock shape
.4 ",,-,:::.-::.-~--- - -
Flow ______ /;.~-----.
~ /f~
-.,......., " Itl
o a = o· J : -:::::
: I'
•i III'
-.4 ,! \\,
i, ,i,
y/d !, ,i,
-.8 \, \i,
\\ ,\'-
\, \'
-1.2 \, \
\, \
-1.6
\,
\'\
'\
\,
""
'" "-
"-
"
\''\
-2.0 L.--'-_'---'-_'--....a...--JL..-....L.--J'--..L..-..I_..L..-..I~..J---L_-'--....J
"
-1.6 -1.2 -.8 -.4 0 .4 .8 1.2 1.6
x/d
Figure 14, Comparison of predicted flight and wind-tunnel-measured shock shapes for a.=O°.
144
Force and Moment
145
1.5
1.4 9 --~--------~
~-,Q1--- -~ a a a
fA _-- a
C A 1.3
ltf-- a ~_------
0'--
1.2 ~----
1.1 ~~~-L~~~~~~~~~~~J-~~~~
-12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12
n, deg
P2'P., d. in.
a 52 3.67
0 11.7 3.67
/', 11.7
0.5 -,- 5.2
--x·- 11.7
0.4
CN
0.3
--
~ Uncertainly
n:¢Il
0.2
-12 -10 -8 ·6 -4 -2 0 2 4 6 8 10 12
n, deg
(b) Normal force coefficient.
Figure 15. Effect of density ratio across normal shock on
longitudinal aerodynamic coefficients at Mach 6 and J3=()9.
146
0.04
0.03
0.02
~"
r night In M~ch 31
equlhbrium Air
(HAllS)
Uncertainty
¢?~
em
0.01
0.00 ~
-0.01
-0.02
-0.03
-12 -10 -8 -6 -4 -2 o 2
n, deg
3.67 }
3.67 Experiment
0.5 2.50
HAllS } Com led
HAllS Pu
0.4
UD 0.3
0.2
0.1 ~~~-L~~~~-L~~~~~~~~~~~~
-12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12
n, deg
147
predicted flight trim angle is only slightly greater than a = 00 , which is in much
better agreement with the CF4 wind-tunnel data than with the air wind-tunnel
data.
To evaluate the effect of the afterbody, tests were conducted with the
afterbody attached and removed in CF4 with the same model, balance, and test
conditions. These tests indicated that the longitudinal coefficients and UD are
identical with the afterbody on or off. The afterbody is shielded from the flow at
hypersonic speeds and therefore does not effect the vehicle aerodynamics.
Consequently, the data presented should represent the current vehicle with
modified afterbody.
A limited number of tests were conducted with sideslip angles other than
00 and revealed the configuration is directionally stable (positive rate of change
of yawing moment coefficient with B) and laterally stable (negative change of
rolling moment coefficient with B). Sideslip angle does not have a significant
effect on the longitudinal coefficients or UD for B s 40 • The data indicated a
slight but consistent decrease in CA with increasing B for -50 Sa S 50; however,
this variation was within the range of uncertainty.
Pressure Distributions
The effect of varying angle of attack between -100 to 100 on the pressure
distribution for the symmetry plane is shown in figure 16 for Mach 6 air. Along
the cf» = 1800 ray, the pressures are well behaved and increase with increasing
angle of attack. A slight overexpansion of the flow from the ellipsoid nose to the
conical surface is observed along this ray for the lowest angle of attack (a = -
100 ), and inflections in the distribution near the nose-cone junction (S/L = 0.22)
are noted for a = _50 and 00. This overexpansion and/or inflection is due to the
surface discontinuity and influence of the cone section on the flow expansion
over the nose. For a > 00 , the pressure decreases monotonically on the cone
section in the direction of the cone-skirt junction (S/L = 0.76). The movement of
the stagnation region with angle of attack is also noted, with this region moving
farther up and around the elliptical nose as angle of attack is decreased from
zero. At the higher angles of attack, the pressure distribution approaches that of
a flat-faced cylinder with rounded corners. It should be noted that the flow over
the nose and cone section is subsonic for all values of angle of attack. (If the
flow within the shock layer expands isentropically from the stagnation region, it
will become supersonic when cplcp,ref < 0.518 for air and 0.566 for CF4.)
148
1.2
«Il = 00 «Il = 1800
1.0 ~O e~~ooo
OA. c oA.A.A.0OO
<>A.O <>c OOOAA.A~O
Co <>c OOOOAO
0.8 0 cccccc 0A.
<> ccoo
A. <> 000<><>000 ce
Cp c <>c
0.6 00 <>
Cp, ref A.
Sonic
a a
a,deg
0.4
0+10
~
0 A +5
c 0 0
0.2 0 c ·5
~ o ·10
0.0
i
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
s/L
Figure 16.- Effece of angle of aetack on pressure distributions; Mach 6;
air; Re - 1.95 x 10 5 . • - 0° and 180°.
-.L
149
Pressure distributions along the symmetry plane (~ = 0° and 180°) are
compared for Mach 6 and 10 air in figure 17(a), (b), and (c). An apparent effect
of Mach number is observed, particularly as the flow expands off the nose onto
the conical section. This apparent compressibility effect increases with
decreasing angle of attack. It should be noted that the density ratio increases
from 5.2 to 6 as the Mach number increases from 6 to 10. Whereas air behaves
thermally and calorically perfect everywhere in the flow field at Mach 6 (i.e.,
obeys the perfect gas equation of state and specific heats are constant), at
Mach 10 the flow field is thermally perfect but calorically imperfect (i.e., because
of vibrational excitation, the ratio of specific heats decreases from 1.4 to around
1.34 behind a normal shock). As will be discussed subsequently, the trends
observed in figure 17 are representative of the effect of density ratio or ratio of
specific heats. The effect of Mach number, for a given Reynolds number and
density ratio, on the pressure distribution is believed to be quite small.
The effect of angle of attack on the heat transfer coefficient along the
symmetry plane at Mach lOin air is shown in figure 19. The stagnation region,
as defined by the region of maximum heating, is observed to move up and
150
1.2
1.0
8
Co
Q
~
o 6
o 10
co
0.8 8
8
gooogggg8s
Cp
Cp, ref 0.6
000 e
Sonic
0.4 o
8
0.2
o
O.OL--L..----JL..-.---L_---L_-.L._--'-_--L...._.....J
151
liL
1.2
cf)= 0 0
cf) =180 0
o 6
1.0 o 10
c 8 Co 0
0 0 0
0.8
0
0
0088 88 8§
~
~
Cp
0.6 8
Cp, ref
Sonic
0.4
0.2
8
C
0.0 L----JI.-----1_--L_--L_--I..._--1-_-.1....._....1
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
s/L
(b) a - 00
Figure 17.- Continued.
152
0.2
0.0 L......--.l....-..-.l....-.._'----IL----IL--_'----IL--~
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
s/L
(c)" .. _ 10 0
153
Measured
Predicted Test gas Mo:> P2 1Po:>
- - - Air (HALlS code)
- - - - _. CF4 (HAllS code) o Air 5.84 5.2
- - - Modified Newtonian [J CF4 6.24 12.0
1.2
1.0
0.8
Cp
I
Cp, ref 0.6 Sonic CF4
(
Sonic Air l
0.4 I
0.2
0.0L----l.....L.---L.--'---'---L---..J_~.l...._.__I
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
s/L
(al a - -10 0
Figure 18.- Effect of normal shock density ratio on pressure
distributions.
154
Measured
Predicted Test gas Mco P2/ Pco
- - Air (HAllS code)
- - - - - - CF4 (HAllS code) o Air 5.84 5.2
- - - Modified Newtonian o CF4 6.24 12.0
1.2 cI> = 00 --;- cI> = 180 0
1.0
Cp 0.8
Cp, ref
0.6 Sonic CF4
Sonic Air
0.4
0.2
155
Measured
Predicted
Test gas M= P2/P=
- - Air (HAllS code)
- - - - - - CF4 (HAllS code) o Air 5.84 5.2
- - - Modified Newtonian o CF4 6.24 12.0
1.2 CI>= 0°
1.0
0.8
Cp
Cp, ref 0.6
Sonic CF4
Sonic Air
0.4
0.2
0.0 l..-----L.-LJJ.-..I..._.....1-_...L..-_l....----1_-"-....l!....-.......J
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
s/L
(c) Q - 100
Figure 18.- Concluded.
156
1.4 a,deg
0+10
1.2 l::.. +5
o 0
Cl -5
1.0 0-10
0.8
Ch, ref <p-
0.6
dJ
0.4 o
6.
0.2 o
4l= 00 -t-- 4l= 180 0
0.0 L - - _ . L . - - . _ . . . . L - _ . . . . I . . . . . . _ - - I - _ - . L . . _ - - ' - _ - - - I _ - - - - J
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
s/L
Figure 19.- Effect. of angle of attack on heat-transfer distributions;
Mach - 10; air; Re•• L - 1. 63 x 10 5 .
157
around the elliptical nose with decreasing angle of attack, moving to the nose-
skirt junction (siL = -0.2) at the lowest angle of attack (a = -10°). At a = 0°, the
center of the measured region of maximum heating does not coincide with the
geometric stagnation point. The magnitude of the heating in the flow stagnation
region (as opposed to the geometric stagnation region) increases significantly
with decreasing angle of attack, corresponding to a decrease in the effective
nose radius. Heating over the skirt section is well behaved; that is, the heating
increases systemically with increasing angle of attack (heating increases
approximately 50 percent as a increases from -10° to 10°). As the flow
approaches the skirt, heating increases and reaches a peak near sIL = 0.73.
This increase in heating in the vicinity of the skirt is the result of the rapid
acceleration of the subsonic flow to a supersonic condition on the skirt. At the
highest angle of attack (a = 10°), heating in the region of the cone-skirt
junction is only about 25 percent less than heating in the stagnation region. At
the lowest angle of attack, heating in the cone-skirt region is about half of that in
the stagnation region.
158
1.4
o Measured data
1.2
- - LAURA code
1.0
0.8
Ch, ref
0.6
0.4
0.2
o.o~--~--~--~--~--~--~--~--~
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
s/L
159
1.4
1.2
o Measured data
1.0 - - LAURA code
Ch 0.8
Ch, ref
0.6
0.4
0.2
0.0~_.l...-_...l.-.._...l.-.._....I.-._...l..-_.....L...._.....l-_.....J
160
1.4
0.8
Ch, ref
0.6
0.4
0.2
0.0 L...-_L...-----.JI....----1_----'-_--L_---L._---L.._-I
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
s/L
(c) .. _ 50
161
n, deg Heating Oil flow
+5 4 A
••
~l
0 <>
-5 c I~e=oo
-10 • 0
x 103 ,
60 \
i
50 0
\
~
!
\
\
40 \
\
\
\
30 \
\
\
\
\
20 I
I
Ae 2, rl \
\
0
I
\ I
I I
I I
10 I I
I I
I I
I I
I I
Q A
I
I I
162
surface. The magnitude of heating on the top of the cylinder (~ = 00), particularly
at the maximum heating point, is significant relative to the computed forebody
stagnation point heating. For example, for angles of attack from _50 to 50, the
corresponding range of heating rate varied from 45 percent to 28 percent of the
forebody stagnation point value. Comparisons of heating-rate distributions
predicted with a Navier-Stokes computer code (LAURA) are generally in good
agreement with the measurements as illustrated in figure 22.
RESUME
An overview of the ground-based hypersonic testing program to establish
an aerodynamiclaerothermodynamic data base for the baseline aeroassist flight
experiment (AFE) configuration is presented. Primary objectives of this program
are: to provide a benchmark data base over a wide range of hypersonic simula-
tion parameters for the calibration (as opposed to validation) of computational
fluid dynamic (CFD) computer codes; to assess the effects of compressibility,
viscosity, real-gas and angles of attack and sideslip on the aerodynamic
characteristics for development of guidance, navigation and control algorithms;
determine aerothermodynamic characteristics for establishment of limits for
thermal protection system of aeroshell and for aft mounted instrumentation
carrier; and to enhance the understanding of the fluid dynamic environment
about the vehicle (forebody flow and near-wake region) for optimization of
various onboard experiments.
163
.6
o Experiment
- LAURA, refined grid
.5 - - LAU RA, base grid
.4
qJqref Data
.3 "'~I ±8% accuracy
o
.2
~
.1
0 1.0 1.5
lId
Appendix A
Austin, R. E., Cruz, M. I., French, Jr. R. 1982. System DeSign Concepts and
Requirements for Aeroassisted Orbital Transfer Vehicles. AIAA Pap. No.
82-1379.
Bird, G. A. 1986. Direct Simulation of Typical AOTV Entry Flows. AIAA Pap. No.
86-1310.
165
Blanchard, R. C. and Hinson, E. W. 1989. Free-Molecule-Flow Force and
Moment Coefficients of the Aeroassist Flight Experiment Vehicle. NASA
TM 101600.
Carlson, Leland A., Bobskill, Glen J., Greendyke, Robert B. 1988. Comparisons
of Vibration Dissociation Coupling and Radiative Heat Transfer Models
for AOTV/AFE Flowfields. AIAA 88-2673.
Carlson, Leland A., Gaily, T. A. 1989. The Effect of Electron Temperature and
Impact Ionization on Martian Return AOTV Flowfields. AIAA 89-1729.
Celenligil, M. C., Moss, J. N., and Blanchard, R. C. 1989. Three-
Dimensional Flow Simulation about the AFE Vehicle in the Transitional
Regime. AIAA 89-0245.
Celenligil, M. C., Moss, J. N., and Bird, Graeme A. 1989. Direct Simulation of
Three-Dimensional Flow about the AFE Vehicle at High Altitudes. AIM
Progress in Astronaytics and Aeronautics, Vol. 118.
166
Cheatwood, F. M., DeJarnette, F. R., and Hamilton, H. H. 1987. An Interactive
Approach to Surface-Fitting Complex Geometries for Flowfield
Applications. AIAA-87-1476.
Davy, W. C., Park, C., and Arnold J. 1985. Radiometer Experiment for the AFE.
AIAA-85-0967.
Dogra, Virendra K., Moss, Jim N., and Simmonds, Ann L. 1987. Direct
Simulation of Aerothermal Loads for AFE Vehicle. AIAA-87-1546.
Gamble, J., Spratlin, K., Skalecki, L. 1984. Lateral Directional Requirements for
a Low UD Aeromaneuvering Orbital Transfer Vehicle. AIAA Pap. No. 84-
2123.
Gibson, Lorelei S., Siemers, Paul M., III, and Kern, F. A. 1989. Pressure
Distribution and Air Data System for the AFE. ISA Paper 89-0048.
Gupta, Roop N. 1988. Stagnation Flowfield Ionization for an AFE Vehicle. AIAA
Pap. No. 88-2613.
167
Gupta, Roop N. 1987. Navier-Stokes and Viscous Shock-Layer Solutions for
Radiating Hypersonic Flows. AIM 87-1576.
Jones, J. J. 1987. The Rationale for an Aeroassist Flight Experiment. AIM 87-
1508.
Li, C. P., and Wey, T. C. 1988. Numerical Simulation of Hypersonic How over an AFE
Vehicle. AIAA Pap. No. 88-2675.
168
Moss, J. N., and Price, J. M. 1988. Direct Simulation of AFE Forebody and
Wake Flow with Thermal Radiation. NASA TM 100673.
Rochelle, W. C., Ting, P. C., Mueller, S. R., Colovin, J. E., Bouslog, S. A., Curry,
D. M., and Scott, C. D. 1989. Aerobrake Heating Rate Sensitivity Study
for the Aeroassist Flight Experiment (AFE). AIM 89-1733.
Scott, C. D. Ried, R. C., Maraia, R. J., Li, C, P., Derry, S. M. 1984. An AOTV
Aeroheating and Thermal Protection Study. AIM Pap. No. 84-1710.
Scott, C. D., Roberts, B. B., Nagy, K., Taylor, P., Gamble, J. D., Cerimele, C. J.,
Kroll, K. R., Li, C. P., and Ried, R. C. 1988. Design Study of an Integrated
Aerobraking Orbital Transfer Vehicle. NASA TM 58264.
Scott, Carl D., Ried, R. C., Maraia, R. J., L, Chien-P, and Derry S. M. The
Aerodynamics and Thermal Protection System Challenges for an
Aerobraking Orbital Transfer Vehicle.
169
Shinn, J. L., Jones, J. J. 1983. Chemical Nonequilibrium Effects on Flowfields
for Aeroassist Orbital Transfer Vehicles. AIM Pap. No. 83-0214.
Stewart, David A., and Kolodziej, Paul. 1986. Heating Distribution Comparison
Between Asymmetric and Symmetric Blunt Cones. AIM 86-1307.
Stewart, David A., and Kolodziej, Paul. 1988. Wall Catalysis Experiment on
AFE. AIM 88-2674.
Striepe, Scott A., and Suit, William T. 1988. AFE Guidance "Quiet Time". NASA
TM 100556.
Walberg, G. D., Siemers, P. M., III, Calloway, R. L., and Jones, J. J. 1987. The
Aeroassist Flight Experiment. IAF-87-197.
170
Wells, William L 1990. Surface Aow and Heating Distributions on a Cylinder in
Near Wake of Aeroassist Flight Experiment (AFE) Configuration at
Incidence in Mach 10 Air. NASA TP 2954.
APPENDIXB
The need for an AFE is driven primarily by the fact that the flight regime
for ASTV's has distinct differences from previous flight experience. ASTV's will
fly at higher velocities than the Shuttle orbiter and at higher altitudes than
Apollo (i.e., higher than the altitude of major heating and loads for Apollo). A
large diameter (e.g., in excess of 50 ft), very blunt aeroshell will produce a
thicker shock layer than experienced previously in flight at very high velocities.
Although this thicker shock layer may serve to lower convective heating, it is
expected to increase radiative heating. The magnitude of the increase in
radiative heating is not well determined. Because the aeroassist pass is made
at high altitudes, the low-density effects of viscosity and flow nonequilibrium are
expected to have a greater impact on ASTV design than would be the case for
an entry vehicle. The Apollo heatshield was capable of withstanding higher
heating rates than expected for an ASTV, but it was an ablative thermal
protection system (TPS) and thus not reusable. The shuttle heatshield is
reusable and demonstrated the ability to maintain a low wall catalysis, but
operated in a lower dissociationlionization regime because of its lower velOCity.
The successful development of aeroassist must, therefore, be based on a
171
technology advancement program which will provide accurate data on the
ASTV environment so that a minimum weight, highly reliable aerobrake system
can be designed.
At altitudes below about 200,000 feet, the density is suffiCiently high that
the chemical reactions occur quickly compared to the transit time of a particle
through the shock layer. In this case, the chemical eqUilibrium state is
approached and accurate flow field analysis is possible using the assumption of
eqUilibrium flow. At altitudes above 250,000 feet, the flow chemistry is not in
eqUilibrium since the comparatively lower density (and resulting lower collision
rate) has increased the time required to relax the flow; thus, the flow chemistry
everywhere deviates significantly from local equilibrium. Immediately behind
the bow shock, the intemal energy modes such as vibration, dissociation and
electronic excitation have not yet absorbed their share of the energy. The
excess appears as thermal energy and thus the local translational temperature
greatly exceeds the eqUilibrium value. For air, this overshoot in temperature is
accompanied by an overshoot in local gas radiation, because some of the most
important radiating species are too hot and are overpopulated compared to
eqUilibrium values. Because of the nonequilibrium condition, the gas near the
body surface is still dissociated so atoms, as well as molecules, are colliding
with the wall. A significant portion of the flow energy may reside in the heat of
dissociation, so the heat transfer to the wall depends on what happens when
the atoms strike the wall. A highly catalytic wall will promote the rapid
recombination of all dissociated particles, releasing the attendant heat of
dissociation to the wall itself; a wall with a low recombination rate coefficient will
inhibit this process and the accompanying heat release. Thus, chemical
nonequilibrium in the flow may have important effects on both radiative and
convective heat transfer. As noted by Jones 1987, significant uncertainties
exist in the ability to predict the magnitude of the effects.
172
Experience with wall catalysis is largely limited to the shuttle TPS
materials at shuttle entry conditions. The glassy coating on the shuttle tile had
low recombination coefficients and was consequently able to reduce the heat
transfer rate to roughly 60 percent of the calculated value for a fully catalytic
wall. At lower altitudes where the collision frequency has increased sufficiently
to approach equilibrium flow, the discrepancy disappears. Shuttle flight data
also indicates a possible tendency for the tile coating to become more catalytic
with continued use. A 20 percent increase in heating was noted for flight 5
compared to flight 2, indicating changes in either the emittance or the catalysis
of the tile surface. Since the ASTV atmospheric pass occurs at higher velocity
than the shuttle entry, the shock layer gases will be more highly dissociated. In
particular, nitrogen dissociation is expected to playa more significant role than
for the shuttle case, for which oxygen dissociation is the dominant mode. Thus,
the potential exists to achieve reductions in convective heating rate if suitably
low recombination coefficients for both oxygen and nitrogen can be maintained
on the TPS surface. However, a major shortcoming is lack of flight-verified data
for other materials than the coated shuttle tile. Current technology, as well as
advanced technology, candidate ASTV materials and coatings need to be tried
in the correct (higher velocity) environment.
APPENDIXC
173
x
~ x
~
6 HEATSHIELD PERFORMANCE
E'.J WALL CATALYSIS
PRESSURE DISTRIBUTION
ATPM
~ •
III AIR DATA SYSTEM
TOTAL RADIOMETER
~ X THERMOCOUPLE
'-: HIGH RESOLUTION SPECTROMETER fQ) WCE PRESSURE
Flow in the base region will be monitored using sensors mounted on the
carrjer vehicle and instrumented booms; these sensors include thermocouples,
radiometers, and pressure transducers mounted on the carrjer vehicle,
Langmuir probes, and video cameras. The Langmuir probe system (LPS) will
consist of six probes mounted on a short boom to determine electron
concentration and temperature; instrumented booms will be in the field of view
of the cameras. This experjment will be a broad study of base flow phenomena
throughout the aeroassist pass. It will provide data to which predictions from
CFD codes may be compared and allow an assessment of afterbody thermal
loads.
FQrebody Aerothermal Characterization Experjment (FACE):
176
Heating Experiment (RHE), the convective heat transfer rate may be extracted
from the measured total heating. Again, these flight data will provide a bench-
mark to which ground-based tests and CFD code predictions may be compared.
177
previously, will allow convective heat transfer rates to be extracted from the
measured total heating (FACE). The RHE data will also be valuable in
validating CFD flow field calculations.
This experiment will consist of high resolution triaxial linear and angular
accelerometers and rate gyros aligned to the vehicle axes which will
supplement the baseline accelerometers and gyros to provide data in the
transition and free-molecule flow regimes. These data will be used to
accurately determine vehicle aerodynamic coefficients in these low-density
viscous flight regimes, and provide a benchmark. data set to which predictions
from Direct Simulation Monte Carlo (DSMC) codes may be compared.
In summary:
178
ACKNOWLEDGMENT
REFERENCES
Cook, W. J., and Felderman, E. J. 1966. Reduction of Data from Thin-Film Heat
Transfer Gages: A Concise Numerical Technique," AIAA Joyrnal, Vol. 4,
pp. 561-562.
Fay, J. A., and Riddell F. R. 1958. Theory of Stagnation Point Heat Transfer in
Dissociated Air. Journal of Aeronautics Science, Vol. 25, No.2, pp. 73-
85.
Jones, J. J. 1987. The Rationale for an Aeroassist Flight Experiment. AIM Pap.
No. 87-1508.
Jones, Robert A., and Hunt, James L. 1969. (appendix A by James L. Hunt,
Kathryn A. Smith, and Robert B. Reynolds and appendix B by James L.
Hunt and Lillian R. Boney), Use of Tetrafluoromethane to Simulate Real-
Gas Effects on the Hypersonic Aerodynamics of Blunt Vehicles. NASA
TR R-312.
Jones, R. A., and Hunt, J. L. 1966. Use of Fusible Temperature Indicator for
Obtaining Quantitative Aerodynamic Heat-Transfer Data, NASA TR R-
239.
179
Micol, J. R. 1987. Simulation of Real-Gas Effects on Pressure Distribution for a
Proposed Aeroassist Flight Experiment Vehicle and Comparison to
Prediction. AIM Pap. No. 87-2368.
Midden, Raymond E., and Miller, Charles G., m. 1985. Description and
Calibration of the Langley Hypersonic CF4 Tunnel--A Facility for
Simulating Low @ Flow as Occurs for a Real Gas. NASA TP-2384.
Miller, Charles G., III. 1975. Shock Shapes on Blunt Bodies in Hypersonic-
Hypervelocity Helium, Air, and C02 Flows, and Calibration Results in
Langley 6-lnch Expansion Tube. NASA TN 0-7800.
Miller, Charles G., III. 1975. A Comparison of Measured and Predicted Sphere
Shock Shapes in Hypersonic Flows with Density Ratios from 4 to 19.
NASA TN 0-8076.
Miller, Charles G., III, and Gnoffo, Peter A. 1981. Pressure Distributions and
Shock Shapes for 12.840 170 On-Axis and Bent-Nose Biconics in Air at
Mach 6. NASA TM-83222.
Miller, C. G., III. 1984. Experimental and Predicted Heating Distributions for
Biconics at Incidence in Air at Mach 10. NASA TP 2334.
Miller, Charles G., III, Micol, John R., and Gnoffo, Peter A. 1985. Laminar Heat-
Transfer Distributions on Biconics at Incidence in Hypersonic-
Hypervelocity Flows. NASA TP-2213.
180
Walberg, G. D. 1982. A Review of Aeroassisted Orbit Transfer. AIM Pap. No.
82-1378.
Walberg, G. D., Siemers, P. M., III, Calloway, R. L., and Jones, J. J. 1987. The
Aeroassist Flight Experiment. IAF-87-197.
Wells, William L., and Franks, A. M. 1988. Measured and Predicted Shock
Shapes for AFE Configuration at Mach 6 in Air and in CF4. NASA TM
100660.
181
High-Enthalpy Testing in Hypersonic Shock Tunnels
B. Esser, H. Gronig, H. Olivier
Shock Wave Laboratory, Technical University Aachen, Germany
Abstract
After an introduction into different types of high-enthalpy hypersonic ground
test facilities the performance of the shock tunnel is described in detail with spe-
cial consideration given to the Aachen tunnel. The basic shock tube perform-
ance is presented with numerical calculations using an equilibrium air model.
The influence of boundary layer effects and van der Waals driver gas effects
on the shock tube flow is included. Numerical calculations of equilibrium and
frozen air flow in the nozzle are shown. The relaxation zone behind a bow
shock is calculated using the air model together with 34 elementary reactions.
The attainable simulation parameters are discussed. The experimental meth-
ods cover the measurements of pressures, forces and moments and heat flux;
time resolved flow visualization is possible. The experimental results given
include nozzle calibrations and a study of the nozzle starting process. A ta-
ble is included giving properties behind incident and reflected shocks at initial
pressures of 0.01, 0.1 and 1 bar for equilibrium air.
1. Introduction
For the development of the European space shuttle HERMES and of future
projects, for example SANGER II or HOTOL, the study of the hypersonic flow
phase during re-entry or steady flight is of essential importance. In ground test
facilities the really existing hypersonic flow conditions should be simulated as
closely as possible. This includes not only the effects of Mach and Reynolds
number as in classical aerodynamics but also real gas effects due to the high
stagnation temperatures. They influence, as the shuttle flights have shown,
such phenomena as boundary layer heat transfer, shock layer vortex shedding,
shock detachment, center of pressure location and control surface effectiveness.
The real gas effects include molecular vibration and at higher speeds essentially
dissociation of first oxygen and then nitrogen. To simulate these real gas effects
in ground test facilities a duplication of the vehicle velocity is necessary. This
implies to duplicate also the stagnation enthalpies corresponding to several
thousand degree Kelvin.
In conventional wind tunnels high enthalpies are not attainable. In shock-
heated tunnels, however, such high values of velocity and enthalpy may be
achieved unfortunately only for a test time of a few milliseconds or less.
Shock heating avoids the necessity of transferring energy from outside which
would not be feasible at the required stagnation temperatures. The dynamics
of shock motion determines uniquely all the conditions in the test gas [lJ.
The following Chapter 2 is devoted to a brief description of high-enthalpy
hypersonic ground test facilities as shock tunnels, free piston shock tunnels and
gun tunnels. Also special arc heated devices will be mentioned. In Chapter 3
the shock tunnel is closer considered, and especially the performance of the
Aachen tunnel is described including some shock tube data for equilibrium
air. Finally simulation of flight conditions will be dealt with and after a short
description of experimental techniques test data are present yd.
183
of methods have been proposed in the past to heat the driver gas including con-
ventional ones, as electrical energy discharge in the driver, internal or external
heating (by combustion or electrical heating), and piston compression [9]. Be-
sides these conventional techniques there have been proposed and tested various
detonation devices to compress the driver gas and also magnetohydrodynamic
processes to accelerate the shock heated driven gas [9]. An informative col-
lection of different non-conventional driver techniques is presented in the Pro-
ceedings of the Seventh International Shock Tube Symposium [10]. The test
times generally decrease with increasing flow velocity.
2.2 Free Piston Shock Tunnel
In the free piston shock tunnel it is utilized that the high pressure and temper-
ature conditions of the driver gas need to exist only for a short period. Thus
Stalker [11] proposed a free piston technique to heat the shock tube driver gas.
The free piston shock tunnel is illustrated schematically in Fig. 3 [12]. Initially
helium of about 2 bar is filled in the compression tube with the piston at its far
upstream end. The reservoir pressure of about 200 bar (usually air) accelerates
the piston after release; the piston accumulates kinetic energy until the pressure
of the compressed driver helium ahead of the piston exceeds the pressure be-
hind it. This will occur when the piston is not far from the downstream end of
the compression tube [13]. The considerable kinetic energy of the piston must
be transferred to the helium as driver gas over a short distance, causing high
pressure and temperature to be developed as the piston approaches the end of
the tube. Under the quoted conditions the values of pressure and temperature
in the driver before diaphragm bursting will be 2000 bar and 4500 K. With
air in the driven section of 1 bar and 300 K a shock is produced with a Mach
number of about 10 leading to nozzle stagnation conditions of Ps = 1000 bar
and Ts = 8500 K [12]. Free stream velocities of up to 10 km/s are achievable.
2.3 Gun Tunnel
Figure 4 explains the principal arrangement of the gun tunnel and its wave
diagram. The driver and driven gases are separated by a piston. The piston
is usually accelerated to supersonic speeds. Due to the strong acceleration of
the piston a shock wave is generated in front of it. Thus the nozzle reservoir
gas is non-isentropic ally heated and compressed by successive reflections of the
primary shock between the nozzle entrance and the piston face [14].
The following discussion in this section rests essentially on parts of the paper
of Stollery and Stalker [12]. In the beginning of the studies of gun tunnel
performance the mass of the piston evoked considerable interest. In Figs. 5a
and b the two extreme situations of piston mass are considered; one extreme
is represented by the "mass-less" piston (Fig. 5a) which is realized by the
non-isentropic compression in a reflected type shock tube. The driver-driven
interface acts as a piston without mass, which is, however, transparent for the
leading expansion wave. The other extreme is represented by a "very heavy"
piston, which moves only slowly giving almost isentropic conditions. [For given
184
P4b the important parameter is mg/(p4 - pt)Aj. In between these extremes
there is a variety of finite piston masses which produce a range of stagnation
conditions by a mixture of shock waves and isentropic compression fans.
A comparison between these two extreme cases regarding the ratios of stag-
nation pressure and temperature is shown in Figs. 6a and b. The pertaining
relations are given in Appendix A. Figures 6a and b show a clear superiority of
the isentropic compression in generating high stagnation pressures except for
low pressure ratios (P41 < 20). Conversely, shock compression is better for gen-
erating stagnation temperatures until P41 exceeds about 900. Piston strength
problems usually prevent operation above these levels.
A further comparison between the gun tunnel and the shock tunnel reveals two
other differences to be seen in Figs. 3 and 4. In the gun tunnel the piston
damps many of the reflected waves so that all the gas in the driven section is
used for test purposes. The piston also seals off the driver gas from the dump
tank at the end of the test run. As a result much of the original driver pressure
P4 can be recovered, so giving economic operation.
The running time of the reflected shock tunnel is usually limited by the head
of the expansion wave (Fig. 2) and is therefore one or two orders of magnitude
less than that of a gun tunnel. Nevertheless the compression process by a
single reflected shock in a shock tunnel is very clean, the stagnation conditions
produced are very uniform and the brief test time ensures that any reservoir
temperature decay is modest [so Sect. 3.2j.
Table I lists the performance of the CALSPAN and AACHEN shock tunnels and
the free piston shock tunnel (Stalker tunnel) T4 of the University of Queens-
land, Australia. For comparison the gun tunnel of the Imperial College and
the Longshot gun tunnel of VKI at Rhode St. Genese near Brussels are in-
cluded. The latter makes use of the heavy piston pressure "trapped" behind
check valves and accepts a decay of stagnation conditions during the test time.
Figure 7 shows the schematic and operation cycle of Longshot tunnel. The
last line in Table I shows the performance data of the projected High Enthalpy
tunnel of DFVLR at .Qottingen (HEG); its data are generally a little higher
than the T4 tunnel of the University of Queensland.
2.4 Tunnels with Electrical Discharge
The disadvantage of shock tunnels lies in the fact, that with increasing nozzle
stagnation pressure the stagnation temperature falls; one may also say that the
highest stagnation temperature may be obtained in reflected shock tunnels at
lower pressures. This suggests to heat the test gas of intermediate pressure in
a stagnation chamber immediately by an electrical discharge in order to obtain
high temperatures and pressures in a relatively simple way. Such a device as
stagnation chamber attached to a hypersonic nozzle forms a so-called hotshot-
tunnel. A number of hotshots existed in the USA during the years 1958 and
1964. In Fig. 8 a comparison is shown between the stagnation conditions of a
185
reflected shock tunnel with 1000 atm driver pressure and the AEDC Hotshot I
with 1 MJ heating energy. It clearly shows that the stagnation temperature
decreases for the shock tunnel with increasing stagnation pressure, but vice
versa for the hotshot (s. also Figs. 16a and b).
The environmental conditions, however, in the arc chamber and in the vicinity
of the nozzle are severe. Using air as test gas the oxygen fraction is reduced by
oxydation and severe erosion occurs. Thus subsequently most of the hotshot-
tunnels were used with nitrogen as test gas instead of air. They cannot therefore
simulate chemical effects under hypersonic flow conditions. These tunnels, how-
ever, are very useful for measuring aerodynamic forces, moments and pressure
distributions in a Mach-number range of 15 to 20. Heat transfer measurements
have also been obtained (in the absence of realistic real gas behaviour) [14].
Arc-heated wind tunnels are steady or intermit tend facilities for high-enthalpy
testing. But they are also not capable of providing the same purity of the
gas. They cannot usually reproduce the Mach and Reynolds number range
because of lacking sufficiently high stagnation pressures. Besides of their high
enthalpy they yield a longer test time, and they are applicable to many phases
of heat-transfer research and aerodynamic force studies.
186
helium of 493 K and 1500 bar as driver and room temperature air of 1 bar as
driven gas one obtains a shock Mach number of Ms = 6.6. The corresponding
nozzle stagnation conditions are Ts = 4130 K and Ps = 415 bar, resp. Varying
the initial pressure PI leads to higher or lower Ms depending whether PI is
decreased or increased. To obtain generally higher test flow Reynolds numbers
the initial pressure has to b~ increased, which implies generally lower stagna-
tion temperatures. A detailed description of the shock tube and nozzle flow
will be given in the next sections.
3.2 Theoretical Description of Shock Tube Flow
The elementary shock tube performance is well documented in a number of
text books [17-20]. It will thus be sufficient to bring into attention the most
important relations and the usual notations of the different flow fields and
kinds of reflections. The initial conditions are shown in Figs. 2 and 10: in
the beginning driver and driven section are separated by a diaphragm. Typical
driver pressures are between 100 and 1500 bar and driven gas pressures between
0.1 and 10 bar.
By a simple mechanism the diaphragm between driver and driven section is
allowed to burst. In Figs. 2 and 10 ideal diaphragm opening and ideal wave
behaviour (no damping, no boundary layer influence) are assumed. Then a
shock wave forms in the test gas (1) and compresses it to a pressure P2 with a
temperature T2 in the region (2). Behind this shock wave the contact surface
follows which separates the driver and driven gases. The driver gas acts like a
piston which compresses the test gas rapidly. The density and sound velocity
on both sides of the interface are different, but pressure and flow velocity are
equal in (2) and (3). This pressure balance is produced by an expansion wave
which moves upstream into the driver section reducing the pressure from its
initial value P4 to P2.
A few milliseconds after the bursting of the diaphragm the incident shock wave
arrives at the end wall of the driven section. The shock tunnel at Aachen
operates in the reflected mode, i.e. the incident shock wave is reflected at the
end wall and propagates upstream. During this reflection process the second
diaphragm between driven section and nozzle bursts and the nozzle flow starts.
The flow velocity behind the reflected shock is zero to a first approximation
if the outflow from the nozzle is neglected. The complete kinetic energy of
the shock heated gas in region (2) is thus converted to high temperature and
pressure in region (5) behind the reflected shock. The compressed test gas hav-
ing a temperature of a few thousand degrees centigrade expands subsequently
through the nozzle. The high stagnation enthalpy is thus converted to a high
free stream velocity in the test section.
The reflected shock interacts after a short time with the contact surface. During
this interaction usually new waves are generated which are propagated again
downstream towards the end wall while the reflected shock moves with changed
187
velocity further upstream. Between the contact surface and the end wall a re-
gion of multiple reflections develops which lead to an equilibrium pressure as
will be shown in Sect. 6.3. Different test times may be seen from Fig. 10 be-
tween the arrival of the incident shock wave and various reflections. At B the
re-reflected shock from the interface terminates the homogeneous region (5).
In case C the reflected shock interacts with the expansion wave and compres-
sion waves are running downstream terminating the equilibrium conditions. In
many shock tunnels, however, the driver length is smaller than indicated in
Fig. 10, thus the arrival of the head of the reflected expansion wave, indicated
by D, occurs earlier. In the case of a special combination of the initial pa-
rameters the reflected shock penetrates the contact surface without generating
waves of finite amplitudes, the so-called tailored interface case.
Figure 11 shows the tailored and both the undertailored and overtailored case;
in the undertailored case the incident shock Mach number is lower than for
the tailored case and it is higher in the overtailored case [4]. For given driver
and driven gases at given temperatures and pressures there is only one Mach
number of the primary shock for which tailoring occurs.
The basic shock tube equation relating shock Mach number and initial pressure
ratio P41 = P4/Pl has the following dependence
(3.2-1)
the exact equation may be found in App. B for perfect gases. Thus varying
P41 different Mach numbers Ms can be obtained. The condition of tailoring
requires Us = U7 = Us = 0 and Ps = P7 = ps (between (5) and (7) only a sonic
wave exists in this case). This leads to a relation between the sound velocity
ratio a32 and the pressure ratio PS2 of the form (see App. B)
Thus only one Mach number is attainable for given initial conditions.
Real Gas Calculation (Equilibrium)
For the shock tube and the nozzle detailed calculations of the flow conditions
were performed; here are especially regarded the properties behind the incident
(2) and reflected (5) shock, the interaction between the reflected shock and
the interface, the conditions of the test flow at the nozzle exit (00) and the
stagnation conditions (s) on the model.
The calculation of the entire flow inside a shock tunnel requires a numerical
method for solving the Euler equations, as e.g. van Leers MUSCL scheme [21].
The important quantities, however, can be obtained using an exact Riemann
solver which takes into account high temperature real gas effects [22].
188
Neglecting dissipative effects such as viscosity, heat transfer and radiation, high
temperature flow problems are described by the Euler equations. Written in
conservation form they are for one dimension
au + of(U) = 0 (3.2-2)
at ax
with U = (e, eu, eet)T, F(U) = (eu, eu2 + p, euhdT
et = e + u 2 /2 and h t = et + pi e
e, u, p and e denote the density, flow velocity, pressure and internal energy.
The Riemann problem is a special initial value problem to (3.2-2) with piecewise
constant initial data
forx>O
(3.2-3)
forx<O
separated by a jump at x = O. The shock tube problem is a special Riemann
problem with Ul = U4 = O.
The set of equations is completed by the equations of state: the thermal equa-
tion of state
p = pee, T, Ni), (3.2-4)
which relates the pressure to density, temperature T and chemical composition
of the mixture, and the caloric equation of state
e = e(T,Ni ), (3.2-5)
and (02p)
oe /0
2
189
where s is the entropy which can be detennined from the equations of state by
the second law of thermodynamics. The first condition is necessary to make
sure that the Euler equations are hyperbolic. Due to the second condition the
equations of state remain convex. For air these conditions have been found to
be valid in a wide range of pressure and temperature.
The second assumption made is that the gas is always in equilibrium. Accord-
ingly the chemical composition of the gas may be detennined definitely by two
thermodynamic state variables such as pressure and temperature. Numerically
this evaluation, however, is the most difficult step in the solution of a Riemann
problem. Details are described in Appendix C.
As an example, Fig. 12 shows the equilibrium composition of air versus tem-
perature at a constant pressure of 0.01 bar. The air is assumed to be composed
of N2 , O2 , N, 0, N+, 0+, NO, NO+ and e-. The mass fractions of the compo-
nents strongly depend on the temperature and there is also a weak dependence
on the pressure.
In general the solution of the Riemann problem consists of four constant states
Vu"" V 4 adjacent two being separated by a wave [27] (Fig. 13). The states
VI and V 4 are known from the initial data (3.2-3). The unknown states V 2
and V 3 are detennined by intersecting the shock and expansion curves in the
pressure-velocity-plane using a secant iteration method.
The shock curves are detennined from the Rankine-Hugoniot conditions
(3.2-6a)
(3.2-6b)
(3.2-6c)
where W is the mass flux across the shock wave. For given V I and P2 first
the temperature T2 is evaluated from (3.2-6a). This is again done by a secant
iteration, since (3.2-6a) cannot be solved for T2 explicitly. At each iteration step
the chemical composition and (h must be determined. Knowing T2 the mass
flux W follows from (3.2-6b) and afterwards u 2 is evaluated from (3.2-6c). In
(3.2-6b) the positive square root must be chosen for left-facing and the negative
one for right-facing shock waves.
The expansion curves arise from the constancy of Riemann invariants. They
can be determined from
(3.2-7a)
(3.2-7b)
190
where a denotes the sound velocity. Here T3 and U 3 are determined indepen-
dently from U 4 and P3. T3 is calculated from (3.2-7a) using a secant iteration
method and u 3 follows from (3.2-7b). The integral is evaluated numerically
using a Romberg method. The positive sign is valid for right-facing and the
negative sign for left-facing expansion waves.
Figure 14 shows as an example the solution of a shock tube problem in air.
The air is composed as described above. The results are compared to the ideal
gas case. There are evident deviations in the temperature, density and velocity
profiles, whereas the pressure profile and the shock velocity are almost the same
in both cases.
Using this Riemann solver shock tube tables for equilibrium air have been cal-
culated (s. Table II). The Mach number ranges from 1 to 15; data are calculated
for the incident and reflected shock including the ratios of the specific heats '"(2
and '"(s. For the reflected shock additionally the ratio of hs/a~ is listed. The
data have been calculated for the initial pressures PI of 0.01 bar, 0.1 bar and
1 bar. For comparison shock tube data for an ideal gas with '"( = 1.4 are listed
at the end of Table II, they are calculated with the same number of digits as
the Ames Tables [28]. The most significant differences between equilibrium
air and ideal gas appear at the lowest initial pressure and the highest Mach
number; comparison between these two entries reveal the importance of taking
into account real gas effects.
To illustrate the influence of the driver/driven gas combinations in Fig. 15
the shock Mach number is plotted which may be obtained by a certain initial
pressure ratio PH. The effectiveness of low driver gas molecular mass and high
temperature is evident.
The pressure Ps behind the reflected shock is also an important property for
the performance of the shock tunnel: increasing PI and thus Ps the Reynolds
number of the test section free stream is also increased. High values of Ps
also lessen the possibility that dissociated oxygen freezes during the nozzle
expansion. Figure 16a illustrates the influence of increasing PIon the behaviour
of Ps for a constant driver gas pressure of P4 = 1500 bar. Helium/air and
hydrogen/air combinations are taken into account with the stated values of
TI and T4 • The curves show a broad maximum at initial pressures of about
10-20 bar which lead to values of Ps higher than the driver pressure. Figure 16b
shows the temperature Ts behind the reflected shock as a function of the driven
pressure for the same gas combinations. This confirms the results already found
in Fig. 8 that with increasing pressure Ps the temperature Ts decreases in shock
tunnels.
Figures 17a and b show the dependence of the shock Mach number on the
driver gas temperature T4 for tailored conditions. The gas combinations are
H2 /air, and He/air. In the first case both hydrogen and air are considered as
real gases; the hydrogen composition includes the following components: H2 ,
H, Ht, H+, H-and e-. For comparison the curve (5) for hydrogen and air as
191
ideal gases is plotted. Helium is considered as ideal gas; in combination with
equilibrium air (curve (6)) the influence of the initial pressures between 0.01
and 1 bar is very small (at 4000 K the deviation amounts to 2.6 %) so that
only the curve for initial pressure of 0.01 bar is plotted. Figure 17a shows the
typical temperature range for conventional shock drivers, and Fig. 17b is rather
applicable to free piston drivers.
Influence of Boundarv Layer
So far any boundary layer effects have been neglected in Figs. 2 and 5. It is
well known, however, that the boundary layer behind the incident and reflected
shocks influences both, characteristic times and flow conditions [4,29,30J. In
shock tunnel applications a serious limitation results from premature driver gas
contamination of the test gas. Explanations of this effect were given by Davies
and Wilson [31J on the basis that bifurcation ofthe reflected shock wave occurs
during its interaction with the boundary layer.
Figure 18 shows schematically the structure of the reflected shock and its inter-
action with the boundary layer. According to an investigation of Mark [32J the
reflected shock before it is transmitted through the contact surface is observed
as a bifurcated shock. He found that the stagnation pressure in the boundary
layer (assumed without internal structure) behind the reflected shock is higher
than the pressure ahead of it. It is then assumed that by boundary layer sep-
aration a bifurcation of the reflected shock is generated. It has been shown by
Mark that this happens within the range Ms = 1.8 to 16 for air (-y = 1.4) [4J.
Davies and Wilson [31J applied the model of Mark also to the situation when
the reflected shock is transmitted through the contact surfaces shown in the
actual situation of Fig. 18. While in this case the fluid in the boundary layer
also comes to rest under the bifurcated shock the adjacent gas layer between the
foreward limb and the tripel point passes through both oblique shocks. This gas
thereby suffers a smaller change in velocity than the gas which passes through
the normal part of the reflected shock and comes to rest in the region (8). Thus
the gas which passes through the bifurcation region has a velocity towards the
contact surface which causes it to penetrate the contact surface as a jet along
the wall [31J.
Figure 19 shows a series of schlieren photographs of the flow pattern of this
interaction between the reflected shock wave and the boundary layer [33J. The
photographs taken at 20 J.tS and 120 I-IS after shock reflection illustrate clearly
the development of the bifurcation due to boundary layer interaction. At 120 I-IS
the contact surface is seen to approach from the left hand 'side; it is not visible
as a sharp front like a shock wave but rather as a turbulent region. The
photographs at 145 J.tS and later show that the bifurcation not only persists
but also increases as the reflected shock is transmitted into the driver gas.
From the schlieren photographs at later times one notes that gas is flowing
close to the walls. Thus the test gas can be contaminated by cold driver gas.
192
Davies and Wilson [31] find that early contamination should occur at shock
Mach numbers in the overtailored case for He/N 2 combinations (Ms > 3.5).
For high-enthalpy reflected shock tunnels Stalker extended their theory to take
into account experimental results in which early contamination persisted at
shock Mach numbers down to 60 % of the tailored interface value which in this
case was close to Ms = 22 [34].
The driver gas contamination leads to an accumulation of cold gas at the end
wall of the shock tube. Experimental results of Lapworth and Townsend [35]
show that the duration of pressure near the end wall increases with increasing
shock Mach number, while the duration of experimental high temperature is
decreasing drastically in the same Mach number range. The experimental in-
vestigation of Bull and Edwards lends further support to the theory based on
the bifurcation model in preference to the contact surface instability interpre-
tation; it also discounts explanations based on non-ideal diaphragm bursting
phenomena [36].
Decreasing initial pressure PI enhances usually these effects due to an increase
of the boundary layer thickness. Thus the use of large diameter shock tubes
and high initial driven pressures PI relieves the situation for not too high Mach
numbers.
Compressibility of the Driver Gas
Using high pressures of the order of 1500 bar in the driver section bulk prop-
erties of the driver gases may effect the shock tunnel flow. Davies et al. [37]
studied the effects of the influence of the compressibility on the test time and
the gas properties behind the reflected shock for elevated initial pressures PI
up to about 30 bar. Though the sound speed of helium was found to increase
with pressure up to more than a factor of 1.4 at 1000 bar compared to nor-
mal pressure, the influence on the shocked properties is usually negligible [38].
The test times, however, are reduced when the reflected head of the expansion
wave terminates the tunnel flow [37]. In Figs. 20a and b the performance of a
shock tube is illustrated using hydrogen/air as driver/driven gas; two cases are
considered: hydrogen as van der Waals or ideal gas. The air is in both cases
considered as equilibrium air of the high temperature model. Figure 20a shows
the dependence on P4I of the shock Mach number and b of the temperature
ratio for the incident shock. For high P4I small deviations to a lower shock
Mach number and temperature ratio may be seen from the figures. The same
calculations and plots for He/equilibrium air show deviations only in line width
of the curves [38].
3.3 Nozzle Flow
The main function of a nozzle is to convert the high enthalpy of the reservoir
into kinetic energy or velocity. A hypervelocity nozzle is defined by Burke and
Bird [39] as one in which the Mach number is sufficiently high (Moo ;::: 8), so
that the expanded gas has nearly approached the lilnit velocity corresponding
193
to the reservoir enthalpy. On the other hand such a nozzle usually operates
from a stagnation temperature sufficiently high (T5 > 1500 K), so that the air
does not follow perfect gas behaviour.
Convergent-divergent Laval nozzles contoured to provide uniform parallel flow
in the test section are used in most supersonic (as distinguished from hyper-
sonic) wind tunnels. The nozzle contours are developed with precision by the
method of characteristics. A small correction is added to the inviscid con-
tour to account for the boundary layer. At Mach numbers above about five,
these conditions which make the usage of contoured nozzles attractive at the
lower Mach numbers are no longer present. For example, the boundary layer
is no longer thin so that the effective inviscid contour of the nozzle changes
significantly with tunnel reservoir conditions [39].
One of the characteristics of hypersonic nozzles designed to produce uniform
flow is their extreme length. Although it is possible to expand the flow along
the centerline of a contoured nozzle to the desired test section Mach number
in a short distance, the rate, at which the uniform flow region widens with
nozzle length, is determined by the Mach angle arcsin(1/M). At hypersonic
conditions this length constitutes a major portion of the nozzle. At Mach
number 20 for example, if the final radius of the test core is to be 0.6 m, just
this portion of the nozzle will be 12 m long [39]. Methods of shortening these
nozzles have been reported [40]. Frequently contoured hypersonic nozzles are
so-called source flow configurations; they consist of a conical section in which
a source flow is assumed to exist followed by a transition section to the final
contoured region.
For these reasons most operational hypervelocity nozzles are simple cones which
circumvent these difficulties. The area ratio, and therefore the Mach number,
may be conveniently adjusted by the use of replaceable throats of different size.
The nozzles can be kept reasonably short by use of relatively large cone angles
which also minimize the boundary layer buildup. Conical nozzles do have a
defect, however, in that they produce axisymmetrical flow. As a consequence,
any test model will be immersed in a stream which is diverging and which has
axial gradients in Mach number, static pressure and the related flow properties.
Methods have been developed to take into account the effects of the free stream
nonuniformities for various configurations [39, 41].
For pure source flow one obtains from the mass conservation equation and
the relation between area and Mach number (see Appendix D) for the Mach
number gradient at the distance x from the source
dM x
- - =,-1
dx M
194
and for the gradients of local density and static pressure
d(l:' =-2
dx (l
dpx
and --
dxp
= -2,
In Sect. 6.2 the measured Mach number gradients exhibit the influence of the
three stagnation conditions of T5 = 1500 K, 3470 K and 5180 K on the ratio
of the specific heats.
To estimate the inviscid flow properties in the test section a one-dimensional
numerical method has been developed to compute the nozzle flow. This method
uses the same nine component model for air as mentioned above. The aim of
this calculation is to be capable to assess the simulation conditions of the shock
tunnel. Mach number, Reynolds number and specific heat ratio are dimension-
less variables of major significance in flight simulation and are commonly used
to form an opinion about the wind tunnel performance. However, many phe-
nomena ocurring in hypersonic flight require virtually complete duplication.
According to Hertzberg et al. [6] duplication of flight conditions requires that
the flight velocity and the ambient free-stream conditions of pressure, temper-
ature, density and gas composition be identically matched and that the model
and flight vehicle be of identical geometry and size. Simulation refers to testing
wherein not all of the flight conditions are duplicated. Duplication is needed,
e.g. when studying equilibrium and non-equilibrium real-gas effects, radiation
phenomena and the interaction of electromagnetic radiation with an ionized
gas.
The difficulties in duplicating ambient flow conditions at hypersonic speeds
are clearly seen in Fig. 21. This figure represents an altitude-velocity map in
which wind tunnel stagnation pressures and temperatures are indicated. For
shock tunnels these conditons occur behind the reflected shock. The stagnation
conditions are necessary for duplicating flight conditions assuming adiabatic
expansion of real equilibrium air. For duplication of hypersonic velocities both,
high stagnation enthalpies and pressures are required.
The assumptions to calculate the nozzle flow are an inviscid, adiabatic high-
temperature gas streaming through an convergent-divergent nozzle. The reser-
voir conditions are produced by the incident and reflected shock wave either as
tailored interface or equilibrium interface conditions. For equilibrium air the
stationary isentropic quasi-one-dimensional equations have been numerically
solved. The solution method is described by Vincenty and Kruger [Ref. 23,
Sect. VI.3]
Due to the high stagnation temperatures ahead of the nozzle real gas effects
in the sence of vibrational excitation, oxygen dissociation and at extreme con-
ditions also nitrogen dissociation occur combined with chemical reactions of
producing NO, NO+ and electrons. When this mixture of components flows
195
through the nozzle the rapid expansion can lead to such low values of the den-
sity that the necessary collisions for the energy exchange between the different
and differently excited species are lacking. Thus deviations from an equilib-
rium nozzle flow will occur. In equilibrium flow the characteristic chemical
times are small compared to the characteristic flow times, due to a sufficient
number of particle collisions. In the case of chemical times being very large
compared to flow times the chemical composition freezes to constant values
from some cross section in the nozzle downstream. In the extreme case one
may assume that freezing occurs already at the composition of the stagnation
condition. High number densities and therefore high collision rates obviously
alleviate these freezing problems. In the calculations both cases, equilibrium
and frozen composition, have been taken into account.
Some results of this numerical calculation are given in Figs. 22a-d [42]. The
stagnation conditions in this case are Po = 680 bar and To = 5150 K. The
diagrams show the free stream values of static temperature T, density {!, area
ratio A/A" and Reynolds number per meter as functions of the Mach number
M which varies in these calculations from 0 to 25. In each diagram the solid line
shows the behaviour of equilibrium air expansion while the dotted line gives the
dependence of frozen chemical composition at the stagnation conditions. The
deviation between equilibrium and frozen conditions are not too large because
of the relatively high stagnation pressure. Decreasing the stagnation pressure
leads to an increase of these deviations. The 500 mm nozzle has an area ratio of
approximately 175 leading to a free stream Mach number of about 6 for these
conditions.
Estimations of the boundary layer thickness have been included using the re-
sults of Lee [43] given in Ref. [6] for the present air model. In Fig. 23 the relative
displacement thickness at the nozzle exit is shown as function of the stagna-
tion pressure for three free-stream temperatures of 50' K, 100 K and 300 K
and two nozzle exit diameters and constant stagnation temperature. It can
be seen that high stagnation pressures lead to a decrease of the displacement
thickness. With decreasing stagnation temperatures the curves are shifted to
lower values on the left hand side. Nozzle boundary layer data are difficult
to apply in general since the particular configuration (for example contouring,
nozzle back pressure etc.) of each facility has a significant effect [6]. Figure 23
may be taken as an indication that the high pressures required for duplication
are also necessary to obtain large cores of uniform flow in the test section.
As already stated high pressure operation of a shock tunnel will also decrease
some problems of interface mixing, reflected-shock boundary layer interaction
and shock wave attenuation [6].
Non-Equilibrium Effects
Non-equilibrium effects occur not only in the nozzle but also behind the bow
shock ahead of the model in the test section. While in the nozzle, however, de-
excitations prevail as the predominant reactions, behind the bow shock wave
excitations of the molecules are the main reactions.
196
In the normal part of the bow shock a non-equilibrium region is established in
which the flow properties relax to the equilibrium values far downstream. If the
shock stand-off distance is too small, equilibration of the flow properties will not
be achieved. The usual behaviour within the relaxation zone behind the shock
wave is an increase of density and pressure, and a decrease of temperature and of
the flow velocity relative to their values immediately behind the shock. Because
a sufficient number of particle collisions is needed especially for dissociation
reactions the gas is in a non-equilibrium state for some distance behind the
shock. An indication of this distance is given in Fig. 24 which shows the
density ratio behind the shock relative to the density immediately behind the
shock versus the distance x behind the shock in form of the binary reaction
variable X = px/u, where p is the pressure and u the velocity both at a distance
x behind the shock wave. A similar plot has been given by Stalker [44). Pairs
of curves are shown for the same flight velocity at two different altitudes of 30
and 75 km, corresponding to 1.117.10- 2 bar and 226.7 K and 2.14.10- 5 bar
and 196.7 K. These pairs of curves only separate after a considerable distance
behind the shock wave.
The reason for this is that this part of the non-equilibrium process is dominated
by the foreward chemical reactions which increase the dissociation levels. All
these are elementary reactions which involve only two components [Ref. 23,
Sects. VII. 7-9). Therefore the rates at which they proceed are proportional
to the density or, at the same temperature and dissociation level, to the pres-
sure [44). For the calculation of the curves in Fig. 24 the nine species air model
was used together with the chemical reaction rates given by Park [45); details
are given in Appendix E. The data of Fig. 24 compare very well with the graphs
given by Stalker [44), though he assumes constant pressure p behind the shock.
Also the use of the binary reaction variable X given above instead of the integral
J(p/u)dx has not a great influence. It may be seen from Fig. 24 that, when the
gas begins to approach an equilibrium state the curves, which were identical,
increasingly diverge. This is because the backward, recombination, reactions
play an increasingly important role and since these involve ternery collisions
the direct proportionality between the reaction rate and density no longer ap-
plies. The binary scaling variable X then ceases to be effective in correlating
flows at different densities. (The undershoot in the curves for 9.15 km/s is a
numerical artefact).
Some Special Problems of the Nozzle Flow
Erosion
Due to the high temperature and high pressure conditions ahead of the nozzle
throat erosion has been observed. Therefore the throat pieces are replaceable.
The nozzle is made from stainless steel and exhibits no signs of erosion. Model
erosion sometimes occurs from diaphragm particles which mainly originate from
the second diaphragm between low pressure section and nozzle.
Very severe erosion was once observed using a conical aluminum nozzle also
equipped with replaceable aluminum throats. The initial conditions were
P4 = 500 bar He and pure oxygen in the driven section. Due to an error the
driven pressure was 1 bar instead of about 40 mbar. The wrong initial con-
ditions gave a shock Mach number of Ms = 4.8. The conditions behind the
reflected shock obtained from Bernstein's tables [46] yielded a temperature
of 2600 K and a pressure of 193 bar. Figure 25a shows the just machined
aluminum throat piece with the double-notched clamping ring for the second
diaphragm and Fig. 25b the eroded throat piece (here with a four-notched
clamping ring) and part of the aluminum nozzle forming the end wall of the
driven section. The aluminum was molten and by the oxygen flow evaporated.
The fine dispersed aluminum vapour condensed subsequently on the surfaces
of the model and the test section [47].
Effect of Water Vapour
Recently Boudreau and Adams, Jr. [48] presented a critical view on the char-
acterization of hypersonic wind tunnel flow fields. The conclusions from their
studies are that most hypersonic wind tunnels operating at or above Mach
number 8 appear to suffer a loss of free-stream Mach number because of non-
isentropic processes occurring in the expansion nozzles. They hypothesize that
the principal non-isentropic process consists of a rapid non-equilibrium vibra-
tional relaxation which raises free-stream temperature and pressure. This re-
laxation phenomenon is associated with impurities, such as water vapour, in
the free-stream flow which act as third bodies. Theoretical calculations sup-
port their hypothesis and suggest that as little as one percent of the reservoir
enthalpy needs to be frozen and subsequently released to produce the effects
observed. They conclude that conventional methods of determining the free-
stream Mach number (i.e. Pitot pressure measurements) are insensitive to such
non-isentropic processes and hence are poor indicators of true Mach number.
Blunt, low-angle cones, on the other hand, have been shown to be excellent
indicators of free- stream Mach number [48].
In the high-enthalpy shock tunnel freezing is also a problem at low stagnation
pressures. Thus the non-isentropic processes may also play a significant role
as shown by Hall & Treanor [49]. The water-vapour content of the test air has
been completely circumvented in the shock tunnel in Aachen because synthetic
air (20 % O2 , rest N2 , hydrocarbon-free) is used throughout.
Angle of Conical Portion
The choice of the half angle of the conical portion of the nozzle has also to
be a compromise between two adverse effects. In case of a small angle the
gradients of the Mach number and other related flow properties decrease; as a
consequence to obtain a constant area ratio of the nozzle its length increases
and also the thickness of the boundary layer. In case of a large angle the nozzle
gets short with a small boundary layer thickness; the gradients, however, of
198
the Mach number and other flow parameters increase and even flow separation
may occur.
= P2. + e20u~.
Poo + eoou~
and = h2• + u~./2
hoo + u~/2
where equilibrium flow is assumed and 020 denotes the equilibrium conditions
behind the shock wave. For an estimate the properties behind the normal shock
should be the same as those behind the bow wave of a re-entry body. We then
have
199
The first relation then gives together with the Mach number independence
principle for the pressure at the stagnation point
Thus for the simulation of the correct pressure distribution the duplication of
the momentum flux eoou~ or the kinetic energy (eoo/2)u~ is required. A
similar argument leads to the heat flux in the stagnation region as of the order
Some of these properties are compared in the following section together with
the re-entry path of Hermes.
4.2 Comparison with Hermes Re-Entry-Path
The Figs. 26 to 30 show the results of a quasi one-dimensional numerical cal-
culation of the flow properties of equilibrium air in the test section taking into
account the air model mentioned earlier. As parameters the driver pressure
P4' the driver temperature T4, the driven pressure PI and the nozzle area ratio
A/A* are used. When simulating real gas effects the free-stream velocity U oc
is of essential importance [53]. Therefore the interesting simulation parame-
ters are plotted mostly with respect to the free-stream velocity. In addition a
typical re-entry path of a space transport with wings is drawn as simulation
aim. In all cases shown the driver pressure is P4 = 1500 bar and the driver
temperature T4 = 493 K. As driver gas helium is used.
Figure 26 shows the test section Reynolds number as a function of the free-
stream velocity for different nozzle area ratios. The pressure PI in the driven
section was varied between 10 bar and 0.05 bar. The lower limit of PI is
determined by the nozzle back pressure. The lower PI the lower is also the
nozzle exit pressure Poc' The initial pressure in the dump tank should be lower
than Poc ; but this depends on the pumping capacity and the leak rate of the
tank. For the calculation of the Reynolds number a model length of 250 mm
200
was assumed. From Fig. 26 it is to be seen that a simulation of the correct
Reynolds numbers is possible up to a velocity of about 2.6 km/s. This requires
driven pressures PI from 10 to 5 bar varying the area ratios A/A * between 177
and 500.
A much larger simulation region is obtained if Reynolds and Mach number have
to be duplicated (s. Fig. 27). In this case the initial pressures PI are nearly the
same. A comparison of Figs. 26 and 27, however, shows that a simultaneous
duplication of the velocity within the whole Mach-Reynolds-number region is
not possible.
Figure 28 shows the simulation region of the binary scaling factor {!ooL. Here
also a model length of 250 mm was assumed while the actual length of the
reentry vehicle is 16.56 m. A duplication is also possible up to velocities of
about 2.6 km/s with pressures between PI ~ 1.5 and 10 bar.
The hatched area in Figs. 26 to 28 denotes a region of simulation in which it is
possible to nearly duplicate simultaneously the parameters Re, Moo, {!ooL and
U oo in one experiment, i.e. with the same nozzle geometry and pressure Pl' A
comparison, for example, of the operation point at PI = 5 bar for the nozzle
with A/A * = 177 yields, that this point in all these figures coincides reasonably
with the reentry path.
Figure 29 represents the simulation region of the free- stream density {!oo versus
velocity U oo ' The driver conditions are the same as in Figs. 26 to 28. It is
obvious that under these conditions the simulated densities are too large. An
arbitrary shift of the simulation region to lower densities, however, is possible
by applying lower driver pressures; then also the velocities are reduced. In
Fig. 29 the simulation region for the driver conditions of P4 = 500 bar and
T4 = 293 K for He/air is shown, too. The simulation of the real conditions
for {!oo' U oo during re-entry is possible by choosing the proper values of initial
pressure and area ratio. Thus when {!oo and U oo are duplicated the parameters
({!00/2)u;, and ({!00/2)u~ are duplicated simultaneously.
4.3 Further Pertinent Parameters
At high altitudes hypersonic flight is usually associated with low density, so
that the effect of gas rarefaction plays an important role. The characteristic
parameter for rarefaction effects is the Knudsen number defined as the ratio of
the mean free path A to a characteristic body length L. The usual definition
of the mean free path according to
M
Knoo = A/I = 1.26 v'7 Re 00
00
201
If one takes the boundary layer thickness 6 as a characteristic length, then
Knoo '" Moo/ ..jReoo ' since 6/1 '" 1/ ..jReoo . However, one would not expect
the results of ordinary boundary layer theory, which assumes that the Reynolds
number is high, to be applicable to a rarefied gas flow in which Re is usually
small [Ref. 51, pp. 210]. Though this particular definition of Kn is not very
appropriate, and a careful discussion of the problem leads to another criterion
for free molecular flow [54]. The rarefaction parameter
given above may be taken as applicable for not too high altitudes. As a
crude division one could expect continuum flow for V < 0.01, slip flow for
0.01 < V < 10 and free molecular kinetics for V > 10 [53].
Low density also leads to large boundary layers, so that the concept of the
boundary layer may no longer be tolerable. The inviscid flow and the shock
wave may be influenced by the viscous layer leading to the so-called viscous
interaction [Ref. 52, pp. 337]. The pertinent parameter X is defined as
_ M!,JC:
x= ..jRe", 00
with Re = tloouooxoo
"'00 Poo
- 1/2
and C ( T ) 1 + 5 1 /Too (4.3 - 1)
00 = Too (T/Too) + 5 1 /Too
Coo represents the Chapman-Rubesin factor [55] which is a constant for a single
flight condition depending on Tw ' Too and Moo' T is an average temperature in
the boundary layer and 51 the reference temperature in Sutherland's viscosity
relation. For X > 4.0 strong interaction occurs, below that value the weak
interaction prevails. In the strong interaction case the boundary layer for a
flat plate grows like X 3 / 4 and not like X l / 2 as in the weak interaction. From
different authors different definitions of the viscous interaction parameter are
introduced [56].
Figure 30 shows the Reynolds-Mach number plot including the lines of con-
stant viscous interaction parameter (solid) and lines of constant rarefaction
parameter (dashed). The hatched area is the same as given in Fig. 27 indicat-
ing the performance of the shock tunnel according to details shown in Fig. 27.
The regimes of continuum and slip flow and also of strong and weak viscous
interaction are indicated. The points with the numbers 1 to 4 represent four
test conditions with driver conditions of P4 = 1500 bar and T4 = 493 K in
He/air with PI = 10 bar and area ratios A/A· = 175; 500; 1300 and 15000
(s. Fig. 27). The values of the viscous interaction parameter for these test
conditions produced by the tunnel have also been given in Fig. 30 as XT' It
is seen, that the deviation of the flight viscous interaction parameters X and
202
the values produced by the tunnel XT are not too large. Computing the flight
viscous interaction parameter the Chapman-Rubesin factor has to be evaluated
according Eq. (4.3-1). But to do this an assumption about the specific heat
ratio has to be used. To show the influence of the real gas behaviour on the
value of X, for the line of X = 1.0 and Moo = 16 the value of the effective 'Y for
that condition is varied by ±0.1. The result is only a small shift of the X = 1
line in both directions.
In conclusion to this section it is seen that a general simultaneous duplication
of all interesting parameters is not possible at these extreme flow conditions.
In a narrow region the simultaneous duplication of Re, Moo, (!ooL and U oo is
indeed possible. Thus a study of the influence of real gases on the classical
flow phenomena (compressibility, boundary layer, friction etc.) can be done.
The duplication of density, stagnation pressure and heat flux requires different
initial conditions of the shock tunnel. Here it is possible to decrease the driven
gas pressure PI and thus the density (!I j this leads to higher stagnation tem-
peratures ahead of the nozzle and thus to higher free stream velocities. But,
as we have seen earlier, low pressures PI are responsible for thicker shock tube
boundary layers which may adversely influence the flow quality and test times
available. Low pressures PI also tend to increase non-equilibrium effects or
even to freeze the chemical composition in parts of the nozzle which also lessen
the desired flow quality.
5. Experimental Methods
5.1 Introduction
The experimental methods for high-enthalpy shock tunnel testing include the
measurement of pressures, forces and moments, and heat transfer and flow visu-
alization by means of schlieren, shadow and interferometer techniques. These
methods are complemented by special techniques which are usually applica-
ble in special test gases or gas property ranges. These methods include e.g.
the electron beam technique to measure species concentration and/or temper-
atures, mass spectrometry and spectroscopy, laser applications for different
fluorescence techniques, microwave methods and Langmuir probes for studying
ionized test gases. Because the test times of high-enthalpy ground-test facilities
are generally short, all experimental methods mentioned have been developed
under the aspects of short-time resolution. This concerns not only the various
transducers but also the needed amplifiers, impedance converters and recording
systems.
5.2 Pressures
In shock tunnel work pressure measurements have to be performed in the driven
section, near or at the end wall, in the nozzle, as Pitot pressures in the test
section and at various locations on the surface of the model. The pressures
measured in the driven section serve to obtain information of the incident
203
shock, i.e. its strength and velocity, and of the reflected shock and thus of the
stagnation pressure ahead of the nozzle. For these purposes piezo-electric pres-
sure gauges are available with resonance frequencies of about 200 kHz. Since
the pressures can be of the order of 1 to 2 kbar, the requirements for sensitivity
is not very high. The pressure on the model surface is measured by piezoelec-
tric or piezoresistive gauges where the latter allow a recalibration in the model
by pressurizing or evacuating the test section. In this case the pressure of a
manometer is then compared with the voltage output of the gauges. Usually
this calibration procedure takes a few minutes and it is not possible with piezo-
electric gauges which loose their charge within a short time. Using piezoelectric
foiles with thicknesses down to 6 pm improvements of piezoelectric gauges are
possible with respect to a higher frequency response [57).
5.3 Forces and Moments
The time resolution is one of the greatest problems in the development of a
force balance for short-duration shock tunnels. In spite of this problem force
balances have been described earlier [58,59). Based on previous force balance
work at Aachen a new piezoelectric six component balance was manufactured
and tested in the shock tunnel. As piezoelectric elements PVDF foils were
used. The individual force measuring elements were slotted cubes similar to
the CALSPAN balance [59), however, of half their size. The prototype of this
balance used six of these force transducers and six additional accelerometers.
Figure 31 shows the complete balance with six cables for force transducers
(FA' FB , Fe, Fv, FE' FF) and six cables for accelerometers three of them
mounted to the ground plate (Gr x' Gr y' Gr %) and three other ones mounted
on the lid (De x , Dey, De%) to which the model is screwed on. The side view
in Fig. 32 shows the positions of the accelerometers inside the balance. Three
accelerometers are mounted together to measure the acceleration in the three
coordinate directions; they may be distinguished from the force transducers
by their circular shape in Fig. 32. The first few test runs encouraged our
engineers to build another six component balance which is more rigid, has
a higher frequency response and works with semi-conductor strain gauges to
improve the steady calibration procedure.
5.4 Heat-Flux
Thin film resistance thermometers measure the surface film temperature and
obtain from the temperature history the heat flux to the model. This technique
is well described and developed especially regarding short test times [60,61).
A sort of a standard method to measure the surface temperature is represented
by the platinum thin film resistance thermometer which is deposited on an
insulating substrate. The platinum film is deposited either by high vacuum
evaporating or sputtering or by a special platinum paint with subsequently fir-
ing. As substrates certain kinds of pyrex (Corning), quartz or macor (Corning)
are in use where the latter consists of a glass ceramic which is machinable by
common tools. Compared with quartz and pyrex macor has the drawback to
204
be softer and probably less durable at extreme test conditions. In Fig. 33 thin
film thermometers are shown which are manufactured by painting technique
on differently shaped pyrex. The film gauges have a response time of a few
microseconds. They are connected to a constant current source. For not too
high surface temperature changes the measured change of the resistance is pro-
portional to the temperature. The evaluation is done either by analog circuit
or by the method of Cook and Felderman [62].
For severest conditions special thin film thermometers or thermocouples have
to be applied the sensitivity of the latter being only 1/60 of the usual thin film
gauges. Recently also copper calorimeter gauges l have been applied for the
severest test runs.
5.5 Flow Visualization
The classical methods of visualization are also used in connection with short-
duration high-enthalpy facilities. These classical methods are the shadowgraph,
the schlieren system and interferometric techniques [63]. For short-duration
facilities short single sparks or multiple sparks are used as light sources in
connection with a still or a time resolving camera. The shadowgraph is sensitive
to the second derivative of the index of refraction or the density, the schlieren
system to the first derivative of the density and interferometers usually to
the density itself. The shadowgraph is most appropriate for visualizing the
geometry of shock and sonic wave patterns. The schlieren system is obtained
by only slight modifications from a shadowgraph, however, it is much more
sensitive to density changes.
The shock tunnel in Aachen is equipped with a shadowgraph and a schlieren
system. As light sources a N anospark unit is available and a ruby laser strobo-
scope. The N anospark has a duration of 15 to 30 ns depending on the size
of the discharge capacitor [64]; this light source can be fired once per tunnel
test giving one photograph at a predetermined time on a medium-size camera.
The ruby laser stroboscope produces a series of light flashes with a usable
frequency between 90000 - 400000 flashes per second. In combination with a
rotating mirror camera and a stationary film one obtains a fast movie of the
time history of the test section flow. This technique has been applied to study
the starting process of the shock tunnel (s. Sect. 6.3).
A double pulse ruby laser system is also available. It is used either as a single
pulse light source in combination with the shadowgraph or schlieren system in
case, when the luminosity from the shock tunnel test section is too high, or to
produce holographic interferograms.
205
5.6 Special Techniques
An electron beam fluorescence technique was developed and applied for the
shock tunnel work in the seventies to measure the rotational temperature of
N2 in the nozzle flow [65). The widely used Nt -system on 391.4 nm was em-
ployed to measure the rotational temperature from the intensities of two spec-
tral ranges . Two optical channels were set up using both, interference filters
and photomultiplier for high sensitivity and reasonable time response. The
ratio of the intensities can be interpreted as a function of the rotational tem-
peratures. The measured temperature range extended from 78 to 320 K. The
local resolution was 3 mm 3 and the time response about 20 I's.
It was possible to measure the static temperature in the free stream nozzle flow
by the electron beam fluorescence technique only in a region of shock tunnel
parameters where the luminosity of the test gas itself was not too intensive. It
was additionally possible to measure the temperature in the stagnation region
of a blunt body by observing the Nt -system from the self-luminosity of the
radiant gas layer behind the shock wave. In this case no electron beam exitation
was needed [65).
A number of experimental techniques has been developed for shock tube stud-
ies; they may also be successfully applied to shock tunnel work. Microwave
techniques were used to measure electron densities and temperatures [7,66).
Time-of-flight mass spectrometry was used to obtain a complete mass spec-
trum of the shock reflection process every 20 I's [67). In a similar system
the electron velocity distribution was studied by an electron energy analyzer
attached to the end wall of the shock tube [68).
6. Experimental Results
6.1 General Instrumentation and Recording
At the moment the shock tunnel is equipped with 32 fast measuring channels
as transient memories with 16 kByte memory for each channel. The resolution
is 8 bit. The transient memories are connected to a computer and fast plotter.
During the test times of about 1 to 10 ms various properties are recorded at
different locations in the shock tunnel.
1. The pressure in the driven section is measured at three positions 2.94 m,
1.39 m and 0.14 m from the end wall.
2. Light emission is measured by means of fiber optics and photodiode 0.22 m
from the end wall.
3. In a cross section 0.295 m from the end wall two opposite measuring ports
exist for simultaneous optical or pressure measurements.
4. The Pitot pressure is generally recorded 0.33 m from the nozzle exit.
5. For the time being 20 pressure or heat transfer gauges can be monitored.
206
6. Light emission from the test section and from the end wall of the dump
tank can be measured.
7. The steady temperature inside and outside of the driver section is sepa-
rately recorded.
6.2 Nozzle Calibration
To determine the pressure distribution in the test section a rack housing nine
Pitot pressure gauges was mounted in the test section. All Pitot tubes were
located in the same plane normal to the nozzle axis. By changing the axial
position the pressure distribution of the test core was determined.
Three nozzle stagnation conditions (Po. To) were generally utilized:
The conical nozzle has a throat diameter of 43.3 mm and an exit diameter of
572 mm. Using the equilibrium air calculation described in Sect. 3.3 one obtains
from the stagnation conditions and Pi tot pressure the free stream Mach num-
ber. Vibrational freezing effects in the nozzle according to Boudreau et al. [48]
have not yet been taken into account.
In Fig. 34 the experimentally obtained Mach number distribution on the axis
of the test section is plotted for the three given stagnation conditions. For the
stagnation temperature To = 1500 K the free stream Mach number amounts
to Moo = 7.7 in a short distance behind the nozzle exit. Because of the conical
nozzle geometry the test section air expands further after leaving the nozzle
exit. This leads to an increase of the Mach number of 10.8 % downstream.
With the nozzle area ratio AlA * = 175 one would obtain a Mach number
Moo = 7.85 for an ideal gas with 1 = 1.4 [28]. The experimentally determined
Mach number is a little less due to two effects. The nozzle boundary layer causes
a reduction of the effective area ratio leading to a decrease in the Mach number.
(A displacement thickness 6* :::; 15 mm at the nozzle exit which is estimated
on the basis of the results of [43] accounts almost for the difference between
measured and calculated free stream Mach number). The second effect is due
to the excitation of internal degrees of freedom which leads also to a decrease of
the Mach number. This second effect is more pronounced at the higher nozzle
stagnation temperatures.
2m
In Fig. 34b the Mach number distribution for To = 3420 K is plotted. The
relative increase of the Mach number amounts to 8.4 %. The Mach numbers
in this case are lower compared to the 1500 K runs because of the increased
excitation of internal degrees of freedom.
At To = 5180 K a relative Mach number increase of 5.5 % resulted from the
calibration runs. The Mach number is Moo = 6.18 in this case.
With increasing stagnation temperature the Mach number gradient thus re-
duces. This confirms the estimation of Sect. 3.3 that for a conical nozzle (with
free-stream Mach number tending to infinity) the relative Mach number gra-
dient behaves as (dM/dx)/(M/x) = 1-1. For the three given conditions
this value was determined to be 0.65; 0.55 and 0.37 with increasing stagna-
tion temperature. Although these absolute values are too high the tendency is
obvious.
6.3 Testing Time
The advantage of a shock tunnel compared to continuous wind tunnels lies
essentially in the fact that extreme free flight conditions can be generated with
stagnation temperatures of several thousand degrees. This advantage is partly
neutralized by the shortcoming of relatively short testing times. Thus generally
the testing time amounts to several milliseconds in shock tunnels. As testing
time the period is considered when all flow conditions have reached steady
values.
After the reflection of the incident shock at the end wall the diaphragm between
driven section and nozzle bursts. The flow in the nozzle starts consisting of a
system of two shock waves and an expansion [69]. This starting flow is highly
unsteady. The period until at a certain location a steady flow establishes cor-
responds to the starting time which may be estimated according to Smith [69]
l
by the following relation
Xa du
tB~ --. (6.3-1)
x H u-a
This relation describes the period which an upstream running disturbance needs
to travel from the throat x H to the nozzle exit x A' For a conical nozzle and a
perfect gas with I = 1.4 Eq. (6.3-1) can explicitly be solved [69] to
208
about 3 ms the pressure behind the reflected shock, i.e. the nozzle stagnation
pressure, is constant. Under these conditions the shock tunnel operates with a
tailored-interface. The pressure remains constant until the reflected expansion
wave arrives at the end wall.
The numerical integration of Eq. (6.3-1) along the nozzle axis yielded a starting
time of 1.2 IDS also in agreement with that of the explicit expression and the
measured time. The delay time llt D in Fig. 35a is due to both, the period for
the opening of the diaphragm and the time which the flow needs to get from
the stagnation region to the pitot probe behind the nozzle exit.
The situation is more complicated with the test conditons 2 and 3 at higher
stagnation temperatures when the tunnel operates in the overtailored mode
(s. Sect. 3.2). As shown in Fig. 10 the reflected shock is again reflected at the
contact surface. This process repeats several times with an increasing pressure
after each reflection at the end wall. Because of the boundary layer and the
starting nozzle flow the reflected waves become weaker with each reflection. The
nozzle stagnation pressure thus reaches an equilibrium value shown in Fig. 36b.
For these test conditions it is approximately constant for about 4.8 IDS. The
pressure after the first shock reflection is equal to Ps = 143 bar. This pressure
is nearly constant until the first reflection from the contact surface arrives at
the end wall. In the Pitot pressure history two more or less distinct rises
may be recognized. The first one occurs at the stagnation pressure Ps and
attains approximately 0.9 bar. The numerical integration of Eq. (6.3-1) yields
a rise time of t.l = 0.7 ms which agrees reasonably well with the experimental
value (s. Fig. 36a). In the subsequent rise the corresponding nozzle stagnation
pressure increases from Ps to the equilibrium pressure Po' This is a complicated
unsteady process because the nozzle and the stagnation conditions vary with
time. A simple estimation of the nozzle starting time using Eq. (6.3-1) is no
longer possible in this case.
After about 2.5 ms a steady Pitot pressure is reached (s. Fig. 36a) which does
not change within the phase of constant nozzle stagnation pressure Po'
Apparently the testing time in the overtailored case is longer than in the tailored
one. This would be correct if the stagnation pressure were the only criterion for
the evaluation of the testing time. For the nozzle flow, however, the duration
of the constant stagnation temperature is of essential importance. Because the
driver gas is highly cooled during the expansion from state 4 to 3 the test gas
may be cooled by driver gas contamination (s. Sect. 3.2). Because of high driven
section pressures and not too excessive enthalpies [34] ahead of the nozzle this
driver gas contamination has not yet been observed as a terminating effect for
the testing time.
One has to ascertain, however, that during the testing time the contact surface
must not arrive at the nozzle entrance. By a simple mass balance of the test
209
gas and the numerically calculated mass flux through the nozzle the testing
time may be approximately calculated from
(6.3-3)
with I?I' I?* initial and throat density, AI' A * cross section of driven section
and throat, L length of driven section, a* sound velocity at the throat, D.tCB
period of time until the contact surface arrives at the nozzle entrance. For
the conditions of Fig. 36 Eq. (6.3-3) yields a value of D.tCB = 7.7 ms which is
indicated in Fig. 36a. A final testing time of 4.6 ms remains.
A similar behaviour is also found for the test conditions 3 of Po = 580 bar and
To = 5180 K. Because of the higher shock Mach number the time intervals
are shorter between the arrival of the different waves at the end wall. Fig-
ure 37 shows the Pitot and stagnation pressure history for these conditions.
The p5-plateau after the first shock reflection is shorter in this case. As a con-
sequence the two rise times in the nozzle flow cannot be as clearly seen as in
Fig. 36a. After 1.2 ms a steady Pitot pressure is attained with a duration of
4.8 ms. Because of the higher shock Mach number the tailored conditions are
farther away than with conditions 2. The pressure plateau in Fig. 37b is not
as clear as in Fig. 36b. During the indicated interval of 3.5 ms the stagnation
pressure increases about 20 %. The time until the contact surface arrives at the
nozzle entrance was also estimated using Eq. (6.3-3) yielding D.tcs = 4.5 IllS.
Thus a final testing time of 2 ms remains.
In this estimate losses were not taken into account. Heat transfer and radiation
causing both, a decrease in nozzle stagnation temperature and a reduction of
the usable testing time.
Furthermore it has not been taken into account that a certain time is needed
until the steady flow around the model is developed. This depends on both,
the size and shape of the model and the local flow conditions. Different flow
properties develop differently during the starting process. Thus, for example,
the pressure needs only half the time to attain a steady value compared to the
heat flux [70].
Figure 38 shows the time elapsed between the arrival of the incident shock wave
and that of the contact surface at the end wall. Here multiple reflections of the
reflected shock have been taken into account. The time intervals are shown for
three area ratios. For comparison also the arrival of the head of the reflected
expansion wave is plotted.
The starting of the nozzle occurs simultaneously with the establishment of the
flow around the model. In Fig. 39 schlieren photographs of the starting process
are shown. There is a time interval from the first to the last photograph of
about 500 /LS. The angle of attack is zero, Mach and Reynolds number are
Moo = 8 and Re oo = 1 . 105. From the series of the schlieren photographs one
can see the double shock system passing the model. In Fig. 39a the primary
210
shock of the starting process has just reached the canopy of the model. Behind
the shock a turbulent region follows flowing from left to right. In Fig. 39c the
second shock is visible as a dark line. On the model in Figs. 39c and d a bow
shock develops indicating that the flow is accelerated to supersonic velocities.
The turbulent region ends with the second shock; when this has passed the
model the flow becomes steady and smooth with a corresponding stationary
shock configuration (s. Fig. 39f).
7. Future Activities
A number of problems in high-enthalpy testing needs further investigations.
This concerns the various interactions occurring simultaneously at high veloc-
ities. These are e.g. effects of chemical non-equilibrium together with rarefied
gas problems. There is a need for detailed measurements of chemical non-
equilibrium surface slip flow. Catalytic effects are very important but the sur-
faces are usually not very well defined regarding their cleanness even in space
conditions. Non-equilibrium turbulent boundary layers are difficult to realize
in ground facilities.
Viscous interaction between the laminar boundary layer and the inviscid flow
is also important regarding control flaps, jet interaction and hypersonic engine
intakes. Interaction of several planar shock waves or the focusing of non-planar
shock waves lead to much higher values of flow properties as known for plane
shocks.
These problems require test facilities of high enthalpy and high pressures to
attain Mach numbers in the range of about 15 and larger and Reynolds numbers
of order 3 . 10 6 /m. Thus there is a general interest in high-enthalpy shock
tunnels by utilizing, for example, Stalker drivers of considerable lengths and
piston masses [16]. Suggestions are also made to increase the nozzle stagnation
conditions by the use of new detonation drivers. Also the attempts to accelerate
the ionized gas behind a sufficiently strong initial shock wave by magneto-
hydrodynamic forces are resumed after some years elapsed. The problem of
short testing times and influences of viscous effects is being tried to lessen by
influencing the boundary layers in the shock tunnel and the nozzle.
Careful experimental investigations of non-equilibrium flow and of complicated
interaction problems can only be solved with a simultaneous development of so-
phisticated measuring methods. Here optical methods seem to be very promis-
ing regarding their non-intrusive applications. The experience made so far,
however, that with increasing enthalpy the testing time decreases means a fur-
ther challenge for developing new optical methods. Laser induced fluorescence
(LIF) or laser induced predissociated fluorescence (LIPF) are basically poten-
tial techniques also for short-duration testing. The experimental investigations
have more than ever to be accompanied by computational fluid dynamic for
mutual benefit.
211
Acknowledgement
Members of the Shock Wave Laboratory assisted in many ways. Our special
thanks go to Mrs. von Hoegen for typing the manuscript and Mr. Klose for
preparing the figures.
References
[1) Charwat, A.F.: A survey of hypersonic problems and the characteristics
of shock-heated tunnels, ZfW, ~, 125- 134 (1960).
[2) Wittliff, Ch.E.: Short duration facilities. Short Course in Hypersonics.
State University of New York at Buffalo, Calspan-UB Research Center.
August 1986.
[3) Russo, A.L.; Hertzberg, A.: Modifications of the basic shock tube to im-
prove its performance. Cornell Aeron. Lab., AD 162251 (1958).
[4) Davies, L.: The interaction of the reflected shock with the boundary layer
in a shock tube and its influence on the duration of hot flow in the reflected-
shock tunnel. Part I. Aeron. Res. Council C.P. No. 880 (1967).
[5) Gronig, H.; Schumacher, H.: Experimentelle Untersuchungen an Raum-
flugkorpern im Hyperschallbereich. Teil II: Entwicklung und Aufbau eines
StoBwellenkanals mit erweitertem Leistungsbereich. StoBwellenlabor,
RWTH Aachen, BMFT-FB W 75-22 (1975).
[6) Hertzberg, A.; Wittliff, Ch.E.; Hall, J.G.: Summary of shock tunnel devel-
opment and application to hypersonic research. Cornell Aeron. Lab. Rep.
No. AD-1052-A-12 (1961).
[7) Dunn, M.G.: Experimental shock-tube investigation of conditions behind
incident and reflected shocks in air for shock Mach numbers between be-
tween 8.5 and 16.5. Fifth Hypervelocity Techniques Symposium, March
1967, Denver, Colorado.
[8) Wittliff, Ch.E.; Wilson, M.R.; Hertzberg, A.: The tailored-interface hy-
personic shock tunnel. J. Aerospace Sci. 26, 219-228 (1959).
[9) Warren, W.R.; Harris, C.J.: A critique of high performance shock tube
driving techniques. In: 1.1. Glass, Ed.: Shock Tubes. Proceedings of
the Seventh International Shock Tube Symposium. University of Toronto
Press, Toronto and Buffalo, 1970, pp. 143-176.
[10) Glass, 1.1. Ed.: Shock Tubes. Proceedings of the Seventh International
Shock Tube Symposium. University of Toronto Press, Toronto and Buffalo,
1970, pp. 177- 538.
[11) Stalker, R.J.: Isentropic compression of shock tube driver gas. ARS J. 30,
564 (1960).
[12) Stollery, J.L.; Stalker, R.J.: The development and use offree piston wind
tunnels. In: R.D. Archer & B.E. Milton (Eds.): Shock Tube and Waves.
New South Wales University Press, Sydney, 1983, pp. 41-50.
[13) Stalker, R.J.: Development of a hypervelocity wind tunnel. The Aeronau-
tical Journal 76, 374-384 (1972).
212
[14] Enkenhus, KR.: Intermittend Facilities. VKI Short Course on Advanced
Shock Tube Techniques, Febr. 1969.
[15] Wendt, J.F.: European hypersonic wind tunnels. AGARD Conference
Proceedings No. 428: Aerodynamics of Hypersonic lifting Vehicles.
AGARD-CP-428, Nov. 1987, paper 2.
[16] Biitefisch, K-A.; Hefer, G.; Hornung, H.; Koppenwallner, G.; Stanewsky,
E.: Pflichtenheft fur eine Konzeptstudie des geplanten Hochenthalpiewind-
kanals HEG. DFVLR IB 222-87 A 09, 9. Mai 1987, Gottingen.
[17] Oertel, H.: StoBrohre, Springer-Verlag, Wien-New York, 1966 p. 670.
[18] Thompson, Ph.A.: Compressible-Fluid Dynamics, McGraw- Hill Book
Company, New York, 1972, Section 8.9.
[19] Liepmann, H.W.; Roshko, A.: Elements of Gasdynamics, John Wiley &
Sons, New York, 1960, Section 3.2.
[20] Glass, 1.1.; Hall, F.G.: Shock Tubes: UTIAS Review No. 12, Part I + II,
May 1958.
[21] Van Leer, B.: Towards the ultimate conservative difference scheme. V. A
second-order sequel to Godunov's method. J. Compo Phys. 32, 101-136
(1979).
[22] Esser, B.; Gronig, H.: Equilibrium shock tube flow of real gases. In: H.
Gronig, Ed.: Shock Tubes and Waves. Proceedings of the 16th Int. Symp.
on Shock Tubes and Waves. VCH-Verlagsgesellschaft, Weinheim 1988,
663-669.
[23] Vincenti, W.G.; Kruger, Ch.H.: Introduction to Physical Gas Dynamics,
Wiley & Sons, New York, 1967, Chapter IV, Sect. 10.
[24] Horton, T.E.: The Computation of Partition Functions and Thermochem-
istry Data for Atomic, Ionic, Diatomic and Polyatomic Species, JPL Tech-
nical Report 32-1425 (1970).
[25] Horton, T.E.; Menard, W.A.: A Program for Computing Shock-Tube Gas-
dynamics Properties, JPL Technical Report 32-1350 (1969).
[26] Huber, KP.; Herzberg, G.: Molecular Spectra and Molecular Structure,
IV, Constants of Diatomic Molecules, van Nostrand Reinhold, New York,
1979.
[27] Smoller, J.: Shock Waves and Reaction-Diffusion Equations, Springer,
New York, 1983.
[28] Ames Research Staff: Equations, tables and charts for compressible flow.
Ames Aeronautical Laboratory, Rep. 1135 (1952).
[29] Mirels, H.: Shock tube testing time limitation due to turbulent-wall bound-
ary layer. AIAA J. 2, 84-93 (1964).
(3~] Vyska, K: Akustische Theorie zur Ermittlung des Einflusses des Rei-
bungsdruckabfalles auf den Gaszustand hinter einer reflektierten StoBwelle
in einem StoBwellenrohr. PhD Thesis, RWTH Aachen, 1969.
(31] Davies, L.; Wilson, J.L.: Influence of reflected shock and boundary layer
interaction on shock-tube flows. Phys. Fluids 12, Supp. I, 1-37-1-43 (1969).
[32] Mark, H.: The interaction of a reflected shock wave with the boundary
layer in a shock tube. NACA TM 1418, March 1958.
213
[33] Holder, D.W.; Schultz, D.L.: The duration and properties of the flow in a
hypersonic shock tunnel. In: F.R. Riddell, Ed: Hypersonic Flow Research,
Academic Press, New York, 1962 pp. 513-546.
[34] Stalker, R.J.; Crane, K.C.A.: Driver gas contamination in a high-enthalpy
reflected shock tunnel. AIAA J. 16,277-279 (1978).
[35] Lapworth, K.C., Townsend, J.E.G.: Temperature and pressure studies in
the reservoir of a reflected-shock hypersonic tunnel. Nat. Phys. Lab., NPL
AERO Rep. 1127, Dec. 1964.
[36] Bull, D.C.; Edwards, D.H.: An investigation of the reflected shock inter-
action process in a shock tube. AIAA J. Q, 1549-1555 (1968).
[37] Davies, L.; Pennelegion, L.; Gough, P.; Dolman, K.: The effects of high
pressure on the flow in the reflected shock tunnel. National Physical Labor-
atory, NPL AERO REPORT 1072, ARC 25071, Hyp. 349, Sept. 1963.
[38] Kurth, G.: Effects of van der Waals gas driver on shock tube flow. Diploma
Thesis, Shock Wave Laboratory, RWTH Aachen, 1988.
[39] Burke, A.F.; Bird, K.D.: The use of conical and contoured expansion
nozzles in hypervelocity facilities. CAL Report No. 112, July 1962.
[40] Geiger, R.E.: Short hypersonic contoured nozzles. ARS J. 30, 368-369
(1960).
[41] Baradell, D.L.; Bertram, M.H.: The blunt plate in hypersonic flow. NASA
TN D-408, October 1960.
[42] Ritzerfeld, E.: Work in progress in Shock Wave Laboratory, RWTH
Aachen (1988).
[43] Lee, J.D.: Axisymmetric nozzles for hypersonic flows. OSU Rept. No. TN
(ALOSU) 459-1, WADC TN 59-228, June 1959.
[44} Stalker, R.J.: Shock tunnels for real gas hypersonics. In: Aerodynamics
of Hypersonic Lifting Vehicles. AGARD Conference Proceedings No. 428,
1987.
[45} Park, Ch.: Convergence of computation of chemical reacting flows. In:
J.N. Moss & C.D. Scott, Eds.: Thermophysical Aspects of Re-entry Flow.
AIAA Series Prog. in Astron. and Aeron. Vol. 103, pp. 478-513, (1986).
[46] Bernstein, 1.: Tabulated solutions of the equilibrium gas properties behind
the incident and reflected normal shock-wave in a shock-tube. I-Nitrogen,
II-Oxygen. Ministry of Aviation, Aeronautical Res. Council C.P. No. 626,
1963.
[47] Finke, K.H.: Die zentrierte zweidimensionale Nichtgleichgewichts-Expan-
sionsstromung im Hyperschallbereich. Dissertation Aachen, 1970.
[48} Boudreau, A.H.; Adams, J.C. Jr.: Characterization of hypersonic wind
tunnel flow fields. AIAA 15th Aerodynamic Testing Conference, May 18-
20,1988.
[49] Hall, T.G.; Treanor, Ch.E.: Nonequilibrium effects in supersonic-nozzle
flows. Technical Report, CAL No. 163, March 1968.
[50] Gibson, W.E.; Marrone, P.V.: A similitude for non-equilibrium phenom-
ena in hypersonic flight. AGARDograph 68: High Temperature Aspects
of Hypersonic Flow. Pergamon Press, Oxford, 1963, pp. 105-129.
214
[51] Cox, R.N.; Crabtree, L.F.: Elements of Hypersonic Aerodynamics. The
English Universities Press Ltd., London 1965, p. 40.
[52] Hayes, W.D.; Probstein, R.F.: Hypersonic Flow Theory. Academic Press,
New York 1959, p. 25.
[53] Koppenwallner, G.: Aerothermodynamik - Ein Schliissel zu neuen Trans-
portgeriiten der Luft- und Raumfahrt. Z. Flugwiss. Weltraumforsch. 12,
6-18 (1988).
[54] Probstein, R.F.: Shock wave and flow field development in hypersonic
re-entry. ARS J. 31, 185-194 (1961).
[55] Chapman, D.R.; Rubesin, M.W.: Temperature and velocity profiles in the
compressible laminar boundary layer with arbitrary distribution of surface
temperature. J. Aeron. Sci. 16,547-565 (1949).
[56] Chevallier, J.P.: Retrospective sur les moyens d'essais et de mesure en
hypersonique. 22eme Colloque d' Aerodynamique Applique, Lille 1985.
[57] Miiller, H.M; Gronig, H.: Experimental investigations on shock focusing
in water. Proc. 12th Int. Congr. on Acoustics Vol. III, H3-3, Toronto 1986.
[58] Bernstein, L.: Force measurement in short-duration hypersonic facilities.
AGARDograph No. 214, Nov. 1975, London.
[59] Anonymous: Hypersonic shock tunnel. Description and capabilities.
ARVIN/CALSPAN, Dec. 1984.
[60] Schultz, D.L.; Jones, T.V.: Heat transfer measurements in short-duration
hypersonic facilities. AGARDograph No. 165, Febr. 1973.
[61] Vidal, R.J.: Model instrumentation techniques for heat transfer and force
measurements in a hypersonic shock tunnel. Corn. Aeron. Lab. Rep. AD-
917-A1, 1956.
[62] Cook, W.J.; Felderman, E.J.: Reduction of data from thin film and trans-
fer gauges. A concise numerical technique AIAA J. 1, 561-562 (1966).
[63] Merzkirch, W.: Techniques of flow visualization. AGARD AG 302, Dec.
1987.
[64] Miyashiro, S.; Gronig, H.: Low-jitter reliable nanosecond spark source for
optical short-duration measurements. Exp. in Fluids il, 71-75 (1985).
[65] Van der Meer, J.: Kurzzeitmessung der Rotationstemperatur von Stick-
stoff mit Hilfe der Elektronenstrahlfluoreszenz. Dissertation, RWTH
Aachen, 1977.
[66] Schneider, K.-P.; Park, Ch.: Shock tube study of ionization rates of NaCI-
contaminated argon. Phys. Fluids 18, 969-981 (1975).
[67] Fritz, K.R.: Zeitaufgeloste massenspektrometrische Untersuchungen der
Kinetik schneller homogener Reaktionen hinter reflektierten StoBwellen.
Dissertation RWTH Aachen, 1976.
[68] Thor, H.J.; Gronig, H.: Dynamic measurement of the electron velocity
distribution in rarefied weakly ionized argon. Experiments in Fluids ~,
42-46 (1984).
[69] Smith, C.E.: The starting process in a hypersonic nozzle. J. Fluid Mech.
24, 625-640 (1966).
215
[70] Holden, M.S.: Establishment time of laminar separated flows. AIAA J. ~,
2296-2298 (1978).
1 1 + (1'1 - 1 )P21
T51 = 1'1 [(31'1 -1)P21 - hI -1)] (1'1 + I)P21 + (1'1 -1) (A-2)
with P21 = P2/Pl representing the shock strength of the incident shock. The
initial pressure ratio P41 over the diaphragm is related to the shock-strength
by the following implicit relation
(A-3)
For the isentropic compression by means of a heavy piston the gas (1'1) in the
volume VI and pressure PI is compressed to the volume v by the gas (1'4) in V4
under pressure P4 (see Fig. 5). Using isentropic relations the following recursion
formula determines the pressure ratio PSI
To51 -- ph,
51
-1)/'"1,
. (A-5)
For 1'4 = 1'1 Eq. (A-4) turns over to the corresponding relation given in [12]
1/'"1 + 1 ]1'
_ V41 P41
PSI - [ V + 1 (A-6)
41
216
or as function of the shock strength as given by Eq. (A-3).
The requirements of equal pressure and velocity on both sides of the contact
surface (i.e. P3 = P2' u 3 = u 2 and Ps = Ps, Us = Us = 0) lead to
for PS2 -1 ~ 1.
Eq. (B-2) may be reduced to a relation which only contains the ratio a 41 of the
initial sound velocities of driver and driven gas and the shock strength P21 •
a23 in Eq. (B-2) is rewritten as
Shock relations are then used for a 21 and PS2 as functions of P21 :
(B-3)
_ (1 + 2(I)P21 - (1
(B-4)
PS2 -
1 + (IP21
The ratio a 43 of the sound velocities is given by the isentropic relation
(B-5)
Inserting Eqs. (B-3), (B-4), (B-5) and (B-1) into Eq. (B-2) the relation between
a41 and P21 for tailored interface conditions follows:
217
Appendix C: Calculation of the Chemical Composition
We consider a gas mixture consisting of n species Ai (i = 1, ... , n). The
species AI' ... , Am are called basic species. These are all chemical elements in
the mixture including electrons if existing. The remaining n - m species are
called composed species. Each of them is thought to be formed from the basic
species by a reaction
for each composed species and the thermal equation of state (3.2-4). The func-
tions Ii on the right hand side of the laws of mass action can be derived from
statistical mechanics in terms of the partition functions [231. The mass bal-
ances are linear in the unknown quantities Ni and e, but all other equations
are nonlinear and therefore the solution must be achieved iteratively. Unfor-
tunately, the equations cannot be solved conveniently for the Ni and e. The
main reason for this is that the Ni may only take positive values and that the
solutions are often so near to zero that they cannot be found by an iterative
method. Hence, new variables
N·
lJi = In eVL
are introduced. Using the lJi the mass balances become nonlinear equations
n
em + L ajie'l; = const, (i=l, ... ,m) (C-l)
j=m+l
218
Appendix D: Gradients in a Conical Nozzle
The flow in a hypervelocity conical nozzle is approximated by a pure source
flow for which with u = const
with x the axial distance from the source. From Eq. (D-1) follows
2dx dA
(D-2)
x A
with the area A = 41l"x 2 • Thus from Eq. (D-2) follows
d{!:. = -2 (D-3)
dr (J
Finally from the Mach number - area relation [s.e.g. Ref. 18, Eq. (65)] for
M->oo
1 dM ,-1 1 dA
----
Mdx 2 Adx
and with Eqs. (D-2)
dM x
dxM='-l. (D-5)
(E-4)
219
_ 1 ei
Here x·--- 1~i ~ k (E-5)
• Mi e
is the mole mass fraction with Mi the mole mass and ei the partial density of
species i. k is the number of the species and cl)i the chemical production term
of the i-th reaction
r
(E-7)
Here are
r: number of reactions
vii,lIij: stochiometric coefficients of the educts and products
kfj: rate constant of the forward reaction
kj = kfj/k b{ equilibrium constant of reaction j from statistical mechanics.
p(x = 0) =p.
e(x = 0) = e.
u(x=O)=u.
xi(x = 0) = Xi.
with x = 0 indicating the shock front.
The thermal and caloric equations of state are
T = p/(eR L: x;)
h = e(T, Xi) + p/ e.
The calculation has been performed for k = 34 reaction equations.
220
Table I: The perlormance of several short-duration hypersonic shock and gun tunnels
Place Type po/bar To/K Moo Rev l Nozzle diameter Test time Ref.
millions D/m ms
Aachen Shock tunnel 1500 2 8000 2 6 - 20 2.5 0.5 0.5 - 10 -
Germany 1.0
2.0
CALSPAN Shock tunnel 1300 2 8000 2 6.5 - 20 0.02 - 250 1.2 0.7 - 20 [2]
USA
--
Univ. of Queensland Stalker tunnel 2000 85993 6 2 0.25 0.4 [12]
Australia
~
Imperial College Gun tunnel 550 1070 9 25 0.45 5 [15]
GB
222
Table II: Shock 'lUbe Data (continued)
(Equilibrium Air: PI = 0.01 bar; Tl =293 K; f!1 = 0.011843 kg/m3 ; Cl = 343.66 m/s)
Ms P21 T21 f!21 C21 M2 '"'(2 P51 T51 f!51 C51 h5/C~ '"'(5
8.1 80.92 10.19 7.747 2.958 2.385 1.185 742.6 15.53 41.86 3.931 4.592 1.252
8.2 83.05 10.32 7.838 2.977 2.403 1.185 771.0 15.81 42.49 3.983 4.586 1.255
8.3 85.20 10.43 7.929 2.996 2.421 1.184 800.2 16.10 43.09 4.035 4.580 1.257
8.4 87.38 10.55 8.020 3.015 2.438 1.184 830.2 16.40 43.68 4.087 4.575 1.259
8.5 89.60 10.67 8.112 3.035 2.456 1.185 861.1 16.71 44.26 4.139 4.571 1.260
8.6 91.83 10.78 8.203 3.054 2.473 1.185 892.7 17.03 44.82 4.190 4.570 1.260
8.7 94.10 10.89 8.294 3.074 2.489 1.186 925.0 17.36 45.38 4.238 4.574 1.258
8.8 96.39 11.00 8.384 3.094 2.505 1.186 958.1 17.70 45.93 4.285 4.582 1.256
8.9 98.71 11.11 8.474 3.114 2.521 1.187 991.8 18.03 46.49 4.328 4.596 1.253
9.0 101.1 11.22 8.563 3.134 2.537 1.189 1026 18.36 47.05 4.370 4.614 1.248
9.1 103.4 11.33 8.651 3.154 2.552 1.190 1061 18.69 47.62 4.409 4.636 1.244
9.2 105.8 11.44 8.739 3.175 2.566 1.191 1097 19.01 48.21 4.446 4.661 1.239
9.3 108.3 11.55 8.825 3.196 2.580 1.193 1133 19.32 48.82 4.482 4.689 1.235
9.4 110.7 11.66 8.910 3.217 2.594 1.194 1169 19.63 49.43 4.517 4.719 1.231
9.5 113.2 11.76 8.994 3.238 2.608 1.196 1206 19.92 50.06 4.550 4.750 1.227
9.6 115.7 11.87 9.076 3.260 2.620 1.198 1244 20.21 50.70 4.583 4.781 1.224
9.7 118.2 11.98 9.157 3.282 2.633 1.199 1282 20.48 51.35 4.615 4.813 1.221
9.8 120.8 12.09 9.236 3.305 2.645 1.201 1320 20.75 52.01 4.647 4.845 1.218
9.9 123.4 12.20 9.314 3.327 2.656 1.203 1359 21.01 52.68 4.679 4.877 1.216
10.0 126.0 12.31 9.389 3.351 2.667 1.205 1398 21.26 53.34 4.710 4.909 1.215
10.1 128.6 12.42 9.463 3.374 2.677 1.207 1438 21.50 54.02 4.741 4.941 1.214
10.2 131.3 12.54 9.534 3.399 2.686 1.209 1478 21.74 54.69 4.772 4.973 1.213
10.3 133.9 12.65 9.604 3.423 2.695 1.212 1518 21.97 55.36 4.802 5.005 1.212
10.4 136.6 12.77 9.670 3.449 2.704 1.214 1558 22.19 56.02 4.833 5.036 1.212
10.5 139.4 12.89 9.734 3.474 2.712 1.216 1598 22.41 56.69 4.863 5.068 1.212
10.6 142.1 13.01 9.796 3.501 2.719 1.218 1639 22.62 57.34 4.893 5.098 1.213
10.7 144.9 13.13 9.854 3.528 2.725 1.221 1680 22.82 57.98 4.923 5.129 1.213
10.8 147.7 13.26 9.910 3.556 2.731 1.223 1720 23.03 58.61 4.953 5.159 1.214
10.9 150.5 13.39 9.962 3.585 2.735 1.226 1761 23.22 59.23 4.983 5.189 1.215
11.0 153.3 13.52 10.01 3.614 2.739 1.228 1802 23.41 59.84 5.012 5.219 1.215
11.1 156.2 13.66 10.06 3.645 2.743 1.231 1842 23.60 60.42 5.042 5.249 1.217
11.2 159.1 13.80 10.10 3.676 2.745 1.234 1883 23.79 60.99 5.071 5.278 1.218
11.3 162.0 13.95 10.14 3.708 2.746 1.237 1923 23.97 61.54 5.100 5.307 1.219
11.4 164.9 14.10 10.17 3.742 2.747 1.240 1962 24.15 62.07 5.130 5.335 1.221
11.5 167.8 14.26 10.20 3.776 2.747 1.242 2001 24.32 62.57 5.159 5.363 1.222
11.6 170.8 14.43 10.22 3.810 2.747 1.245 2040 24.49 63.05 5.188 5.391 1.223
11.7 173.8 14.60 10.25 3.846 2.745 1.248 2079 24.66 63.50 5.217 5.419 1.225
11.8 176.8 14.78 10.26 3.882 2.744 1.251 2116 24.82 63.93 5.246 5.446 1.227
11.9 179.8 14.96 10.28 3.918 2.742 1.253 2154 24.98 64.34 5.275 5.473 1.228
12.0 182.8 15.16 10.29 3.954 2.740 1.255 2191 25.14 64.72 5.303 5.500 1.230
12.1 185.9 15.35 10.30 3.989 2.739 1.256 2227 25.30 65.09 5.332 5.526 1.232
12.2 189.0 15.56 10.30 4.024 2.738 1.256 2264 25.45 65.44 5.361 5.552 1.234
12.3 192.1 15.76 10.30 4.057 2.738 1.256 2300 25.61 65.78 5.390 5.578 1.235
12.4 195.2 15.97 10.31 4.089 2.738 1.255 2337 25.76 66.11 5.419 5.603 1.237
12.5 198.3 16.19 10.31 4.119 2.740 1.253 2373 25.91 66.44 5.447 5.628 1.239
12.6 201.5 16.40 10.31 4.148 2.743 1.250 2410 26.06 66.77 5.476 5.652 1.241
12.7 204.7 16.62 10.32 4.175 2.747 1.247 2448 26.20 67.11 5.506 5.677 1.243
12.8 208.0 16.83 10.33 4.200 2.752 1.243 2487 26.35 67.45 5.535 5.701 1.244
12.9 211.2 17.04 10.34 4.224 2.758 1.239 2527 26.50 67.81 5.564 5.724 1.246
13.0 214.5 17.24 10.35 4.247 2.765 1.234 2568 26.65 68.19 5.594 5.747 1.248
13.1 217.9 17.45 10.36 4.269 2.772 1.230 2610 26.79 68.58 5.624 5.770 1.250
13.2 221.2 17.65 10.38 4.291 2.780 1.226 2653 26.94 68.99 5.654 5.792 1.252
13.3 224.6 17.84 10.40 4.311 2.788 1.222 2697 27.09 69.42 5.684 5.814 1.254
13.4 228.1 18.03 10.42 4.332 2.797 1.218 2743 27.24 69.86 5.715 5.835 1.255
13.5 231.5 18.21 10.45 4.351 2.806 1.215 2790 27.39 70.32 5.746 5.856 1.257
13.6 235.0 18.39 10.48 4.371 2.814 1.212 2839 27.54 70.79 5.777 5.876 1.259
13.7 238.5 18.57 10.51 4.391 2.823 1.209 2889 27.68 71.28 5.808 5.896 1.261
13.8 242.1 18.74 10.55 4.410 2.833 1.206 2940 27.83 71.78 5.840 5.916 1.262
13.9 245.7 18.90 10.58 4.429 2.842 1.204 2992 27.98 72.30 5.872 5.935 1.264
14.0 249.3 19.07 10.62 4.448 2.851 1.202 3046 28.13 72.82 5.904 5.953 1.266
14.1 253.0 19.22 10.66 4.467 2.860 1.200 3101 28.29 73.35 5.936 5.971 1.268
14.2 256.7 19.38 10.70 4.486 2.870 1.199 3157 28.44 73.90 5.969 5.989 1.269
14.3 260.4 19.52 10.74 4.505 2.879 1.198 3214 28.59 74.45 6.002 6.006 1.271
14.4 264.1 19.67 10.79 4.524 2.888 1.197 3273 28.74 75.01 6.035 6.022 1.272
14.5 267.9 19.81 10.83 4.543 2.897 1.196 3332 28.90 75.57 6.068 6.038 1.274
14.6 271.7 19.95 10.88 4.562 2.906 1.196 3393 29.05 76.14 6.102 6.054 1.275
14.7 275.6 20.09 10.93 4.581 2.916 1.195 3455 29.20 76.71 6.135 6.069 1.277
14.8 279.5 20.22 10.98 4.599 2.925 1.195 3518 29.36 77.29 6.170 6.083 1.278
14.9 283.4 20.35 11.03 4.618 2.934 1.195 3582 29.52 77.86 6.204 6.097 1.279
15.0 287.3 20.48 11.08 4.637 2.943 1.195 3647 29.67 78.44 6.239 6.110 1.280
223
Table II: Shock Tube Data (continued)
(Equilibrium Air: PI = 0.1 bar; T1 = 293 K; III = 0.11843 kg/m 3 ; C1 = 343.66 m/s)
Ms P21 T21 1121 C21 M2 '"(2 P51 T51 1151 C51 h5/C~ '"(5
1.0 1.000 1.000 1.000 1.000 0.0000 1.399 1.000 1.000 1.000 1.000 2.503 1.399
1.1 1.245 1.065 1.169 1.032 0.1544 1.398 1.540 1.131 1.361 1.063 2.507 1.397
1.2 1.514 1.128 1.342 1.061 0.2883 1.397 2.238 1.262 1.773 1.122 2.511 1.396
1.3 1.806 1.190 1.517 1.090 0.4065 1.397 3.110 1.395 2.230 1.179 2.518 1.393
1.4 2.121 1.254 1.692 1.119 0.5119 1.396 4.169 1.530 2.724 1.233 2.526 1.390
1.5 2.460 1.319 1.866 1.147 0.6070 1.395 5.428 1.670 3.250 1.287 2.537 1.386
1.6 2.823 1.386 2.037 1.175 0.6932 1.393 6.897 1.815 3.800 1.339 2.550 1.382
1.7 3.209 1.455 2.205 1.203 0.7720 1.392 8.584 1.965 4.369 1.391 2.565 1.377
1.8 3.619 1.527 2.369 1.232 0.8443 1.390 10.50 2.120 4.951 1.442 2.582 1.372
1.9 4.053 1.602 2.529 1.261 0.9109 1.388 12.64 2.280 5.543 1.493 2.601 1.367
2.0 4.510 1.680 2.684 1.290 0.9725 1.386 15.01 2.445 6.140 1.543 2.621 1.361
2.1 4.992 1.761 2.834 1.320 1.030 1.384 17.63 2.616 6.740 1.592 2.642 1.356
2.2 5.497 1.845 2.979 1.350 1.083 1.381 20.48 2.791 7.339 1.642 2.664 1.350
2.3 6.027 1.932 3.120 1.380 1.133 1.378 23.58 2.972 7.936 1.691 2.687 1.345
2.4 6.581 2.022 3.255 1.410 1.179 1.375 26.93 3.158 8.529 1.740 2.710 1.340
2.5 7.159 2.115 3.385 1.440 1.223 1.372 30.53 3.348 9.117 1.788 2.733 1.336
2.6 7.762 2.211 3.511 1.471 1.264 1.369 34.38 3.545 9.699 1.837 2.756 1.331
2.7 8.389 2.309 3.633 1.502 1.303 1.366 38.48 3.746 10.27 1.885 2.780 1.327
2.8 9.041 2.411 3.750 1.532 1.340 1.362 42.84 3.952 10.84 1.933 2.804 1.323
2.9 9.718 2.516 3.863 1.563 1.375 1.359 47.46 4.163 11.40 1.981 2.828 1.319
3.0 10.42 2.623 3.972 1.594 1.408 1.356 52.35 4.379 11.95 2.029 2.852 1.315
3.1 11.14 2.733 4.078 1.626 1.439 1.352 57.49 4.600 12.50 2.077 2.877 1.311
3.2 11.90 2.846 4.180 1.657 1.469 1.349 62.90 4.826 13.03 2.124 2.901 1.308
3.3 12.67 2.962 4.279 1.688 1.498 1.346 68.58 5.057 13.56 2.171 2.926 1.304
3.4 13.47 3.080 4.374 1.719 1.525 1.342 74.52 5.292 14.08 2.218 2.952 1.300
3.5 14.30 3.201 4.467 1.751 1.552 1.339 80.74 5.531 14.60 2.264 2.978 1.296
3.6 15.15 3.324 4.557 1.782 1.577 1.336 87.22 5.775 15.10 2.310 3.004 1.293
3.7 16.02 3.451 4.644 1.814 1.601 1.333 93.97 6.022 15.60 2.356 3.031 1.289
3.8 16.93 3.579 4.729 1.845 1.624 1.331 101.0 6.274 16.10 2.401 3.058 1.285
3.9 17.85 3.711 4.811 1.877 1.646 1.328 108.3 6.529 16.59 2.445 3.086 1.281
4.0 18.80 3.845 4.890 1.909 1.667 1.325 115.9 6.787 17.07 2.489 3.114 1.277
4.1 19.78 3.981 4.968 1.940 1.688 1.322 123.7 7.048 17.55 2.533 3.143 1.273
4.2 20.78 4.120 5.044 1.972 1.708 1.320 131.8 7.312 18.03 2.576 3.173 1.269
4.3 21.81 4.262 5.117 2.003 1.727 1.317 140.2 7.578 18.50 2.618 3.204 1.265
4.4 22.86 4.406 5.189 2.035 1.745 1.315 148.9 7.846 18.97 2.659 3.236 1.261
4.5 23.94 4.552 5.259 2.067 1.763 1.312 157.8 8.115 19.44 2.700 3.269 1.256
4.6 25.04 4.701 5.327 2.098 1.781 1.310 167.1 8.385 19.91 2.740 3.303 1.252
4.7 26.17 4.852 5.394 2.129 1.798 1.307 176.5 8.655 20.38 2.779 3.338 1.247
4.8 27.32 5.005 5.459 2.161 1.815 1.305 186.3 8.925 20.85 2.817 3.374 1.243
4.9 28.50 5.161 5.523 2.192 1.831 1.302 196.4 9.194 21.32 2.854 3.412 1.238
5.0 29.71 5.319 5.585 2.223 1.846 1.300 206.7 9.461 21.80 2.891 3.450 1.234
5.1 30.94 5.479 5.647 2.254 1.862 1.297 217.3 9.726 22.27 2.927 3.490 1.230
5.2 32.19 5.641 5.707 2.285 1.877 1.295 228.2 9.988 22.76 2.962 3.530 1.226
5.3 33.48 5.805 5.767 2.316 1.892 1.292 239.4 10.25 23.24 2.996 3.570 1.222
5.4 34.78 5.971 5.825 2.346 1.907 1.289 250.8 10.50 23.73 3.031 3.611 1.219
5.5 36.12 6.139 5.883 2.377 1.921 1.287 262.6 10.76 24.23 3.065 3.652 1.216
5.6 37.48 6.309 5.940 2.407 1.935 1.284 274.7 11.01 24.73 3.098 3.692 1.214
5.7 38.86 6.481 5.997 2.437 1.949 1.281 287.0 11.25 25.23 3.132 3.733 1.212
5.8 40.28 6.654 6.052 2.466 1.963 1.279 299.7 11.50 25.74 3.165 3.772 1.211
5.9 41.72 6.829 6.108 2.496 1.977 1.276 312.8 11.74 26.25 3.199 3.812 1.210
6.0 43.18 7.006 6.163 2.525 1.991 1.273 326.1 11.97 26.77 3.232 3.850 1.209
6.1 44.67 7.183 6.218 2.554 2.005 1.270 339.8 12.21 27.28 3.266 3.887 1.209
6.2 46.19 7.362 6.273 2.582 2.018 1.267 353.9 12.44 27.81 3.300 3.924 1.209
6.3 47.74 7.542 6.328 2.610 2.032 1.263 368.4 12.67 28.33 3.334 3.959 1.210
6.4 49.31 7.723 6.383 2.638 2.046 1.260 383.2 12.90 28.86 3.369 3.993 1.211
6.5 50.92 7.904 6.438 2.665 2.060 1.256 398.5 13.13 29.39 3.404 4.026 1.212
6.6 52.54 8.085 6.494 2.692 2.074 1.253 414.2 13.36 29.92 3.439 4.057 1.214
6.7 54.20 8.267 6.550 2.718 2.089 1.249 430.4 13.59 30.46 3.476 4.087 1.216
6.8 55.89 8.448 6.608 2.743 2.104 1.245 447.0 13.82 31.00 3.512 4.116 1.218
6.9 57.60 8.628 6.665 2.769 2.118 1.242 464.2 14.05 31.54 3.549 4.143 1.221
7.0 59.34 8.808 6.724 2.793 2.133 1.238 481.9 14.28 32.08 3.587 4.168 1.223
7.1 61.11 8.987 6.784 2.817 2.149 1.234 500.1 14.52 32.62 3.626 4.192 1.226
7.2 62.91 9.164 6.845 2.841 2.164 1.230 518.9 14.76 33.17 3.665 4.213 1.229
7.3 64.74 9.340 6.907 2.865 2.179 1.227 538.3 14.99 33.71 3.705 4.233 1.232
7.4 66.60 9.514 6.971 2.888 2.195 1.223 558.3 15.24 34.26 3.746 4.252 1.235
7.5 68.49 9.686 7.035 2.910 2.211 1.220 579.0 15.48 34.80 3.788 4.268 1.239
7.6 70.41 9.855 7.101 2.933 2.227 1.217 600.3 15.73 35.35 3.831 4.282 1.242
7.7 72.36 10.02 7.168 2.955 2.242 1.214 622.3 15.99 35.89 3.875 4.295 1.245
7.8 74.33 10.19 7.235 2.977 2.258 1.211 644.9 16.25 36.43 3.919 4.305 1.249
7.9 76.34 10.35 7.304 2.999 2.274 1.209 668.3 16.52 36.97 3.965 4.314 1.252
8.0 78.38 10.51 7.374 3.020 2.289 1.207 692.3 16.79 37.50 4.011 4.322 1.255
224
Table II: Shock Tube Data (continued)
(Equilibrium Air: PI = 0.1 bar; Tl = 293 K; (!1 = 0.11843 kg/m 3 ; Cl = 343.66 m/s)
Ms P21 T21 (!21 C21 M2 12 PSI T51 (!SI CSI h5/C~ 15
8.1 80.44 10.67 7.445 3.042 2.305 1.205 717.0 17.07 38.02 4.058 4.327 1.258
8.2 82.53 10.83 7.516 3.064 2.320 1.204 742.5 17.36 38.54 4.106 4.332 1.261
8.3 84.65 10.98 7.587 3.086 2.335 1.203 768.6 17.65 39.05 4.155 4.336 1.264
8.4 86.81 11.13 7.660 3.107 2.350 1.202 795.5 17.95 39.55 4.203 4.339 1.266
8.5 88.98 11.28 7.732 3.129 2.365 1.201 823.0 18.26 40.05 4.252 4.342 1.267
8.6 91.19 11.43 7.805 3.151 2.380 1.201 851.3 18.58 40.53 4.301 4.345 1.269
8.7 93.43 11.58 7.878 3.173 2.394 1.201 880.2 18.90 41.01 4.350 4.349 1.270
8.8 95.69 11.72 7.950 3.195 2.408 1.201 909.8 19.23 41.49 4.398 4.354 1.270
8.9 97.98 11.87 8.023 3.217 2.422 1.201 940.1 19.56 41.96 4.446 4.361 1.269
9.0 100.3 12.01 8.095 3.240 2.435 1.201 970.9 19.90 42.43 4.492 4.370 1.268
9.1 102.6 12.15 8.167 3.262 2.448 1.202 1002 20.24 42.90 4.537 4.380 1.267
9.2 105.0 12.29 8.239 3.285 2.461 1.203 1035 20.57 43.37 4.581 4.393 1.265
9.3 107.4 12.43 8.309 3.308 2.473 1.204 1067 20.91 43.84 4.624 4.407 1.263
9.4 109.8 12.57 8.380 3.331 2.485 1.205 1100 21.25 44.31 4.666 4.424 1.261
9.5 112.3 12.71 8.449 3.354 2.497 1.207 1134 21.58 44.79 4.706 4.442 1.258
9.6 114.8 12.85 8.517 3.378 2.509 1.208 1168 21.91 45.27 4.746 4.461 1.255
9.7 117.3 12.99 8.585 3.401 2.520 1.210 1203 22.24 45.75 4.784 4.482 1.253
9.8 119.8 13.13 8.651 3.425 2.530 1.211 1238 22.56 46.24 4.822 4.504 1.251
9.9 122.4 13.26 8.716 3.450 2.541 1.213 1273 22.87 46.73 4.859 4.526 1.248
10.0 124.9 13.40 8.780 3.474 2.550 1.215 1309 23.18 47.23 4.896 4.549 1.246
10.1 127.5 13.54 8.843 3.499 2.560 1.217 1346 23.48 47.73 4.932 4.573 1.245
10.2 130.2 13.68 8.904 3.525 2.569 1.219 1382 23.77 48.23 4.968 4.596 1.243
10.3 132.8 13.82 8.964 3.550 2.578 1.221 1419 24.06 48.73 5.003 4.620 1.242
10.4 135.5 13.96 9.022 3.576 2.586 1.223 1457 24.35 49.23 5.038 4.644 1.241
10.5 138.2 14.10 9.078 3.603 2.593 1.226 1494 24.62 49.73 5.072 4.668 1.240
10.6 140.9 14.25 9.133 3.629 2.601 1.228 1532 24.90 50.23 5.107 4.692 1.240
10.7 143.7 14.39 9.186 3.657 2.608 1.230 1570 25.16 50.72 5.141 4.716 1.239
10.8 146.5 14.54 9.237 3.684 2.614 1.232 1608 25.43 51.21 5.175 4.740 1.239
10.9 149.3 14.68 9.286 3.712 2.620 1.235 1647 25.68 51.69 5.209 4.763 1.239
11.0 152.1 14.83 9.333 3.741 2.625 1.237 1686 25.93 52.17 5.242 4.787 1.240
11.1 155.0 14.99 9.378 3.770 2.630 1.239 1724 26.18 52.64 5.276 4.810 1.240
11.2 157.8 15.14 9.420 3.799 2.635 1.242 1763 26.42 53.11 5.309 4.833 1.241
11.3 160.7 15.30 9.461 3.829 2.639 1.244 1802 26.66 53.56 5.343 4.856 1.242
11.4 163.6 15.46 9.499 3.860 2.643 1.246 1841 26.90 54.00 5.376 4.878 1.242
11.5 166.6 15.62 9.535 3.891 2.646 1.249 1880 27.13 54.44 5.409 4.900 1.243
11.6 169.5 15.79 9.568 3.922 2.648 1.251 1919 27.35 54.86 5.442 4.922 1.245
11.7 172.5 15.96 9.600 3.954 2.651 1.253 1958 27.58 55.27 5.475 4.944 1.246
11.8 175.5 16.14 9.629 3.986 2.653 1.255 1996 27.80 55.67 5.508 4.966 1.247
11.9 178.6 16.31 9.655 4.019 2.655 1.257 2035 28.01 56.06 5.541 4.987 1.248
12.0 181.6 16.50 9.680 4.051 2.656 1.259 2074 28.23 56.44 5.574 5.008 1.250
12.1 184.7 16.68 9.702 4.084 2.657 1.260 2112 28.44 56.80 5.607 5.029 1.251
12.2 187.8 16.87 9.722 4.117 2.658 1.261 2151 28.65 57.15 5.640 5.049 1.253
12.3 190.9 17.07 9.741 4.150 2.660 1.262 2189 28.85 57.49 5.672 5.069 1.254
12.4 194.0 17.26 9.758 4.183 2.661 1.263 2228 29.06 57.82 5.705 5.089 1.256
12.5 197.2 17.46 9.773 4.215 2.662 1.263 2266 29.26 58.15 5.738 5.109 1.258
12.6 200.4 17.67 9.788 4.247 2.664 1.263 2305 29.46 58.46 5.771 5.128 1.259
12.7 203.6 17.87 9.801 4.278 2.666 1.263 2343 29.66 58.77 5.804 5.147 1.261
12.8 206.8 18.08 9.814 4.309 2.668 1.262 2382 29.85 59.07 5.836 5.166 1.263
12.9 210.1 18.29 9.827 4.339 2.670 1.261 2421 30.05 59.37 5.869 5.184 1.265
13.0 213.4 18.51 9.839 4.368 2.674 1.259 2461 30.24 59.67 5.903 5.203 1.267
13.1 216.7 18.72 9.852 4.397 2.677 1.257 2501 30.44 59.97 5.936 5.220 1.268
13.2 220.0 18.93 9.866 4.424 2.681 1.255 2542 30.63 60.28 5.969 5.238 1.270
13.3 223.4 19.14 9.880 4.451 2.686 1.253 2583 30.82 60.58 6.003 5.255 1.272
13.4 226.8 19.35 9.895 4.477 2.691 1.250 2625 31.01 60.89 6.036 5.271 1.274
13.5 230.2 19.56 9.912 4.502 2.696 1.248 2667 31.20 61.20 6.070 5.288 1.276
13.6 233.6 19.77 9.930 4.527 2.702 1.245 2711 31.39 61.52 6.104 5.304 1.278
13.7 237.1 19.98 9.949 4.551 2.708 1.243 2755 31.58 61.85 6.138 5.320 1.279
13.8 240.6 20.18 9.970 4.575 2.714 1.240 2801 31.77 62.18 6.173 5.335 1.281
13.9 244.2 20.38 9.992 4.598 2.720 1.238 2847 31.96 62.52 6.207 5.350 1.283
14.0 247.8 20.58 10.02 4.621 2.727 1.235 2894 32.15 62.87 6.242 5.364 1.285
14.1 251.4 20.77 10.04 4.644 2.734 1.233 2942 32.34 63.22 6.277 5.378 1.287
14.2 255.0 20.96 10.07 4.666 2.741 1.231 2991 32.54 63.58 6.313 5.392 1.289
14.3 258.7 21.15 10.09 4.688 2.748 1.229 3041 32.73 63.95 6.348 5.406 1.290
14.4 262.4 21.33 10.12 4.710 2.755 1.227 3093 32.92 64.32 6.384 5.418 1.292
14.5 266.1 21.51 10.15 4.732 2.763 1.226 3145 33.11 64.70 6.420 5.431 1.294
14.6 269.9 21.69 10.19 4.753 2.770 1.225 3198 33.31 65.09 6.456 5.443 1.296
14.7 273.7 21.87 10.22 4.775 2.777 1.223 3252 33.50 65.47 6.493 5.455 1.297
14.8 277.5 22.04 10.25 4.796 2.785 1.222 3307 33.70 65.86 6.530 5.466 1.299
14.9 281.3 22.21 10.29 4.818 2.792 1.222 3363 33.89 66.26 6.567 5.477 1.300
15.0 285.2 22.38 10.32 4.839 2.799 1.221 3420 34.09 66.66 6.604 5.487 1.302
225
Table II: Shock Tube Data (continued)
(Equilibrium Air: PI = 1 bar; Tl = 293 K; (!1 = 1.1843 kg/m 3 ; Cl = 343.66 m/s)
Ms 1121 T21 (!21 C21 M2 72 P51 T51 (!51 C51 h5/C~ 75
1.0 1.000 1.000 1.000 1.000 0.0000 1.399 1.000 1.000 1.000 1.000 2.503 1.399
1.1 1.245 1.065 1.169 1.032 0.1544 1.398 1.540 1.131 1.361 1.063 2.507 1.397
1.2 1.514 1.128 1.342 1.061 0.2883 1.397 2.238 1.262 1.773 1.122 2.511 1.396
1.3 1.806 1.190 1.517 1.090 0.4065 1.397 3.110 1.395 2.230 1.179 2.518 1.393
1.4 2.121 1.254 1.692 1.119 0.5119 1.396 4.169 1.530 2.724 1.233 2.526 1.390
1.5 2.460 1.319 1.866 1.147 0.6070 1.395 5.428 1.670 3.250 1.287 2.537 1.386
1.6 2.823 1.386 2.037 1.175 0.6932 1.393 6.897 1.815 3.800 1.339 2.550 1.382
1.7 3.209 1.455 2.205 1.203 0.7720 1.392 8.584 1.965 4.369 1.391 2.565 1.377
1.8 3.619 1.527 2.369 1.232 0.8443 1.390 10.50 2.120 4.951 1.442 2.582 1.372
1.9 4.053 1.602 2.529 1.261 0.9109 1.388 12.64 2.280 5.543 1.493 2.601 1.367
2.0 4.510 1.680 2.684 1.290 0.9725 1.386 15.01 2.445 6.140 1.543 2.621 1.361
2.1 4.992 1.761 2.834 1.320 1.030 1.384 17.63 2.616 6.740 1.592 2.642 1.356
2.2 5.497 1.845 2.979 1.350 1.083 1.381 20.48 2.791 7.339 1.642 2.664 1.350
2.3 6.027 1.932 3.120 1.380 1.133 1.378 23.58 2.972 7.936 1.691 2.687 1.345
2.4 6.581 2.022 3.255 1.410 1.179 1.375 26.93 3.158 8.529 1.740 2.710 1.340
2.5 7.159 2.115 3.385 1.440 1.223 1.372 30.53 3.348 9.117 1.788 2.733 1.336
2.6 7.762 2.211 3.511 1.471 1.264 1.369 34.38 3.545 9.699 1.837 2.756 1.331
2.7 8.389 2.309 3.633 1.502 1.303 1.366 38.48 3.746 10.27 1.885 2.780 1.327
2.8 9.041 2.411 3.750 1.532 1.340 1.362 42.84 3.952 10.84 1.933 2.804 1.323
2.9 9.718 2.516 3.863 1.563 1.375 1.359 47.46 4.163 11.40 1.981 2.828 1.319
3.0 10.42 2.623 3.972 1.594 1.408 1.356 52.35 4.379 11.95 2.029 2.852 1.315
3.1 11.14 2.733 4.078 1.626 1.439 1.352 57.49 4.600 12.50 2.077 2.877 1.311
3.2 11.90 2.846 4.180 1.657 1.469 1.349 62.90 4.826 13.03 2.124 2.901 1.308
3.3 12.67 2.962 4.279 1.688 1.498 1.346 68.58 5.057 13.56 2.171 2.926 1.304
3.4 13.47 3.080 4.374 1.719 1.525 1.342 74.52 5.292 14.08 2.218 2.952 1.300
3.5 14.30 3.201 4.467 1.751 1.552 1.339 80.74 5.531 14.60 2.264 2.978 1.296
3.6 15.15 3.324 4.557 1.782 1.577 1.336 87.22 5.775 15.10 2.310 3.004 1.293
3.7 16.02 3.451 4.644 1.814 1.601 1.333 93.97 6.023 15.60 2.356 3.030 1.289
3.8 16.93 3.579 4.729 1.845 1.624 1.331 101.0 6.274 16.10 2.401 3.057 1.285
3.9 17.85 3.711 4.811 1.877 1.646 1.328 108.3 6.529 16.59 2.446 3.084 1.281
4.0 18.80 3.845 4.890 1.909 1.667 1.325 115.9 6.788 17.07 2.490 3.112 1.278
4.1 19.78 3.981 4.968 1.940 1.688 1.322 123.7 7.050 17.55 2.534 3.140 1.274
4.2 20.78 4.120 5.044 1.972 1.708 1.320 131.8 7.316 18.02 2.578 3.168 1.270
4.3 21.81 4.262 5.117 2.003 1.727 1.317 140.2 7.584 18.49 2.621 3.197 1.267
4.4 22.86 4.406 5.189 2.035 1.745 1.315 148.9 7.856 18.96 2.664 3.225 1.263
4.5 23.94 4.552 5.259 2.067 1.763 1.312 157.9 8.130 19.42 2.706 3.254 1.260
4.6 25.04 4.701 5.327 2.098 1.781 1.310 167.1 8.406 19.88 2.748 3.284 1.257
4.7 26.17 4.852 5.394 2.129 1.798 1.307 176.7 8.685 20.33 2.790 3.313 1.253
4.8 27.32 5.005 5.459 2.161 1.815 1.305 186.5 8.967 20.79 2.831 3.343 1.250
4.9 28.50 5.161 5.523 2.192 1.831 1.302 196.6 9.250 21.24 2.872 3.373 1.247
5.0 29.71 5.319 5.585 2.223 1.846 1.300 207.0 9.534 21.69 2.912 3.404 1.244
5.1 30.94 5.479 5.647 2.254 1.862 1.297 217.7 9.821 22.14 2.953 3.434 1.241
5.2 32.19 5.641 5.707 2.285 1.877 1.295 228.7 10.11 22.59 2.992 3.465 1.238
5.3 33.48 5.805 5.767 2.316 1.892 1.292 240.0 10.40 23.04 3.032 3.495 1.236
5.4 34.78 5.971 5.825 2.346 1.906 1.290 251.5 10.68 23.49 3.071 3.526 1.233
5.5 36.12 6.140 5.883 2.377 1.921 1.287 263.5 10.97 23.94 3.110 3.557 1.231
5.6 37.48 6.310 5.939 2.407 1.935 1.285 275.7 11.26 24.39 3.149 3.587 1.229
5.7 38.86 6.482 5.996 2.437 1.948 1.282 288.2 11.55 24.84 3.187 3.617 1.227
5.8 40.28 6.656 6.051 2.467 1.962 1.279 301.1 11.84 25.29 3.226 3.647 1.226
5.9 41.71 6.832 6.106 2.497 1.976 1.277 314.3 12.13 25.74 3.265 3.677 1.225
6.0 43.18 7.009 6.160 2.527 1.989 1.274 327.8 12.42 26.19 3.303 3.706 1.224
6.1 44.67 7.188 6.214 2.557 2.002 1.272 341.6 12.71 26.64 3.342 3.734 1.223
6.2 46.19 7.369 6.268 2.586 2.015 1.269 355.9 13.00 27.09 3.380 3.762 1.223
6.3 47.73 7.551 6.321 2.615 2.028 1.267 370.4 13.29 27.54 3.419 3.789 1.223
6.4 49.30 7.734 6.373 2.644 2.041 1.264 385.4 13.58 27.99 3.458 3.815 1.223
6.5 50.90 7.919 6.426 2.673 2.054 1.261 400.7 13.86 28.44 3.497 3.840 1.224
6.6 52.52 8.105 6.478 2.701 2.066 1.259 416.3 14.15 28.89 3.537 3.864 1.225
6.7 54.17 8.293 6.531 2.729 2.079 1.256 432.4 14.44 29.34 3.577 3.887 1.226
6.8 55.85 8.481 6.583 2.757 2.091 1.254 448.9 14.73 29.79 3.617 3.909 1.227
6.9 57.55 8.670 6.635 2.785 2.104 1.251 465.8 15.03 30.24 3.658 3.930 1.229
7.0 59.29 8.860 6.687 2.813 2.116 1.249 483.1 15.32 30.68 3.699 3.950 1.231
7.1 61.04 9.051 6.739 2.840 2.129 1.246 500.9 15.61 31.12 3.741 3.969 1.233
7.2 62.83 9.242 6.791 2.867 2.141 1.243 519.1 15.91 31.56 3.783 3.986 1.235
7.3 64.64 9.433 6.844 2.894 2.154 1.241 537.8 16.21 32.00 3.826 4.002 1.238
7.4 66.49 9.625 6.897 2.920 2.166 1.238 556.9 16.51 32.44 3.869 4.018 1.240
7.5 68.35 9.817 6.950 2.947 2.179 1.236 576.6 16.81 32.87 3.913 4.032 1.243
7.6 70.25 10.01 7.003 2.973 2.191 1.234 596.7 17.11 33.30 3.957 4.044 1.246
7.7 72.18 10.20 7.057 2.999 2.204 1.231 617.4 17.42 33.72 4.002 4.056 1.249
7.8 74.13 10.39 7.111 3.025 2.216 1.229 638.5 17.74 34.15 4.047 4.067 1.251
7.9 76.11 10.58 7.165 3.050 2.229 1.227 660.3 18.05 34.57 4.093 4.076 1.254
8.0 78.12 10.77 7.220 3.075 2.241 1.225 682.5 18.37 34.98 4.139 4.085 1.257
226
Table II: Shock Tube Data (continued)
(Equilibrium Air: PI = 1 bar; Tl = 293 K; (II = 1.1843 kg/m 3 ; Cl = 343.66 m/s)
Ms P21 T21 {l21 C21 M2 12 P51 T51 {l51 C51 h5/C~ 75
8.1 80.15 10.96 7.275 3.101 2.253 1.224 705.3 18.70 35.39 4.186 4.093 1.260
8.2 82.22 11.15 7.330 3.126 2.265 1.222 728.7 19.03 35.80 4.233 4.101 1.263
8.3 84.31 11.34 7.386 3.151 2.277 1.221 752.7 19.36 36.20 4.281 4.107 1.265
8.4 86.43 11.52 7.442 3.176 2.289 1.219 777.2 19.70 36.60 4.329 4.114 1.268
8.5 88.58 11.71 7.498 3.201 2.301 1.218 802.2 20.04 36.99 4.377 4.120 1.270
8.6 90.75 11.89 7.554 3.226 2.313 1.217 827.9 20.38 37.38 4.425 4.125 1.272
8.7 92.95 12.08 7.610 3.251 2.325 1.217 854.1 20.74 37.77 4.473 4.131 1.274
8.8 95.18 12.26 7.666 3.276 2.336 1.216 880.9 21.09 38.15 4.521 4.138 1.275
8.9 97.44 12.44 7.722 3.301 2.347 1.216 908.2 21.45 38.53 4.569 4.144 1.276
9.0 99.73 12.62 7.778 3.326 2.358 1.216 936.1 21.81 38.91 4.616 4.151 1.277
9.1 102.0 12.80 7.834 3.351 2.369 1.216 964.6 22.17 39.28 4.663 4.159 1.277
9.2 104.4 12.98 7.890 3.377 2.379 1.216 993.5 22.53 39.65 4.710 4.167 1.278
9.3 106.7 13.16 7.945 3.402 2.390 1.216 1023 22.90 40.02 4.756 4.177 1.278
9.4 109.1 13.34 8.000 3.427 2.400 1.217 1053 23.27 40.39 4.802 4.187 1.277
9.5 111.6 13.52 8.054 3.453 2.410 1.218 1084 23.64 40.76 4.846 4.198 1.277
9.6 114.0 13.69 8.108 3.479 2.419 1.219 1114 24.00 41.13 4.891 4.210 1.276
9.7 116.5 13.87 8.161 3.505 2.428 1.220 1146 24.37 41.50 4.934 4.222 1.275
9.8 119.0 14.05 8.214 3.531 2.437 1.221 1178 24.73 41.87 4.977 4.236 1.274
9.9 121.5 14.22 8.266 3.558 2.446 1.222 1210 25.09 42.23 5.019 4.250 1.273
10.0 124.0 14.40 8.317 3.584 2.455 1.224 1243 25.45 42.60 5.061 4.265 1.272
10.1 126.6 14.58 8.367 3.611 2.463 1.225 1276 25.81 42.97 5.102 4.281 1.271
10.2 129.2 14.75 8.417 3.638 2.471 1.227 1309 26.16 43.34 5.143 4.297 1.270
10.3 131.9 14.93 8.465 3.665 2.478 1.228 1343 26.51 43.71 5.183 4.313 1.269
10.4 134.5 15.11 8.512 3.693 2.485 1.230 1378 26.85 44.08 5.223 4.330 1.268
10.5 137.2 15.28 8.559 3.720 2.492 1.232 1412 27.19 44.44 5.263 4.346 1.267
10.6 139.9 15.46 8.604 3.748 2.499 1.234 1447 27.53 44.81 5.302 4.364 1.267
10.7 142.6 15.64 8.648 3.777 2.506 1.236 1482 27.86 45.18 5.341 4.381 1.266
10.8 145.4 15.82 8.691 3.805 2.512 1.238 1518 28.19 45.54 5.379 4.398 1.266
10.9 148.1 16.00 8.733 3.834 2.517 1.240 1553 28.52 45.90 5.417 4.416 1.266
11.0 150.9 16.18 8.773 3.863 2.523 1.242 1589 28.84 46.26 5.456 4.433 1.266
11.1 153.8 16.36 8.812 3.893 2.528 1.245 1625 29.15 46.61 5.493 4.450 1.266
11.2 156.6 16.55 8.850 3.922 2.533 1.247 1662 29.47 46.96 5.531 4.468 1.266
11.3 159.5 16.73 8.886 3.952 2.538 1.249 1699 29.77 47.31 5.569 4.485 1.266
11.4 162.4 16.92 8.921 3.982 2.542 1.251 1735 30.08 47.65 5.606 4.502 1.267
11.5 165.3 17.11 8.955 4.013 2.546 1.253 1772 30.38 47.99 5.643 4.519 1.267
11.6 168.3 17.30 8.987 4.043 2.550 1.255 1810 30.67 48.32 5.681 4.536 1.268
11.7 171.2 17.49 9.018 4.074 2.553 1.257 1847 30.97 48.65 5.718 4.553 1.269
11.8 174.2 17.68 9.048 4.105 2.557 1.259 1884 31.26 48.97 5.755 4.569 1.269
11.9 177.2 17.88 9.076 4.137 2.560 1.261 1922 31.54 49.29 5.792 4.586 1.270
12.0 180.3 18.08 9.103 4.168 2.563 1.263 1960 31.83 49.59 5.829 4.602 1.271
12.1 183.3 18.28 9.129 4.200 2.566 1.265 1998 32.11 49.90 5.866 4.618 1.273
12.2 186.4 18.48 9.153 4.231 2.568 1.266 2036 32.39 50.20 5.902 4.633 1.274
12.3 189.5 18.69 9.177 4.263 2.571 1.268 2074 32.66 50.49 5.939 4.649 1.275
12.4 192.7 18.89 9.199 4.295 2.573 1.269 2112 32.93 50.77 5.976 4.664 1.276
12.5 195.8 19.10 9.220 4.327 2.576 1.270 2150 33.20 51.05 6.013 4.679 1.278
12.6 199.0 19.31 9.240 4.359 2.578 1.271 2189 33.47 51.33 6.050 4.694 1.279
12.7 202.2 19.52 9.259 4.390 2.580 1.272 2228 33.74 51.60 6.087 4.709 1.281
12.8 205.5 19.74 9.278 4.422 2.583 1.273 2267 34.00 51.86 6.124 4.723 1.282
12.9 208.7 19.96 9.295 4.453 2.585 1.273 2306 34.26 52.12 6.160 4.737 1.284
13.0 212.0 20.17 9.312 4.484 2.588 1.273 2345 34.52 52.38 6.197 4.751 1.286
13.1 215.3 20.39 9.329 4.515 2.591 1.273 2385 34.78 52.63 6.235 4.764 1.287
13.2 218.6 20.61 9.345 4.545 2.593 1.273 2425 35.04 52.88 6.272 4.778 1.289
13.3 222.0 20.84 9.362 4.575 2.596 1.272 2465 35.29 53.13 6.309 4.790 1.291
13.4 225.4 21.06 9.378 4.605 2.600 1.272 2506 35.55 53.38 6.346 4.803 1.293
13.5 228.8 21.28 9.394 4.634 2.603 1.271 2547 35.80 53.62 6.384 4.816 1.294
13.6 232.2 21.50 9.410 4.663 2.606 1.270 2588 36.05 53.87 6.422 4.828 1.296
13.7 235.7 21.73 9.426 4.692 2.610 1.269 2630 36.31 54.11 6.459 4.839 1.298
13.8 239.2 21.95 9.443 4.720 2.614 1.268 2672 36.56 54.36 6.497 4.851 1.300
13.9 242.7 22.17 9.460 4.748 2.618 1.267 2716 36.81 54.60 6.535 4.862 1.302
14.0 246.2 22.40 9.477 4.775 2.623 1.265 2759 37.06 54.85 6.574 4.873 1.304
14.1 249.8 22.62 9.495 4.802 2.627 1.264 2803 37.31 55.09 6.612 4.884 1.306
14.2 253.4 22.84 9.514 4.829 2.632 1.263 2848 37.56 55.34 6.651 4.894 1.308
14.3 257.0 23.05 9.533 4.855 2.637 1.261 2894 37.81 55.59 6.690 4.904 1.310
14.4 260.7 23.27 9.553 4.881 2.641 1.260 2940 38.06 55.84 6.729 4.913 1.311
14.5 264.4 23.49 9.573 4.907 2.647 1.258 2987 38.31 56.10 6.768 4.923 1.313
14.6 268.1 23.70 9.594 4.932 2.652 1.257 3034 38.56 56.35 6.807 4.932 1.315
14.7 271.8 23.91 9.616 4.957 2.657 1.256 3083 38.81 56.61 6.847 4.940 1.317
14.8 275.6 24.12 9.639 4.982 2.662 1.255 3132 39.06 56.87 6.887 4.948 1.319
14.9 279.4 24.33 9.662 5.007 2.668 1.254 3181 39.32 57.13 6.927 4.956 1.321
15.0 283.2 24.54 9.686 5.032 2.673 1.253 3232 39.57 57.40 6.968 4.964 1.323
2T1
Table II: Shock Tube Data (continued)
(Equilibrium Air: PI = 10 bar; Tl = 293 K; &11 = 11.843 kg/m 3 ; Cl = 343.66 m/s)
Ms 1'21 T21 &121 c21 M2 "(2 P51 T51 &151 c51 h5/C~ "(5
1.0 1.000 1.000 1.000 1.000 0.0000 1.399 1.000 1.000 1.000 1.000 2.503 1.399
1.1 1.245 1.065 1.169 1.032 0.1544 1.398 1.540 1.131 1.361 1.063 2.507 1.397
1.2 1.514 1.128 1.342 1.061 0.2883 1.397 2.238 1.262 1.773 1.122 2.511 1.396
1.3 1.806 1.190 1.517 1.090 0.4065 1.397 3.110 1.395 2.230 1.179 2.518 1.393
1.4 2.121 1.254 1.692 1.119 0.5119 1.396 4.169 1.530 2.724 1.233 2.526 1.390
1.5 2.460 1.319 1.866 1.147 0.6070 1.395 5.428 1.670 3.250 1.287 2.537 1.386
1.6 2.823 1.386 2.037 1.175 0.6932 1.393 6.897 1.815 3.800 1.339 2.550 1.382
1.7 3.209 1.455 2.205 1.203 0.7720 1.392 8.584 1.965 4.369 1.391 2.565 1.377
1.8 3.619 1.527 2.369 1.232 0.8443 1.390 10.50 2.120 4.951 1.442 2.582 1.372
1.9 4.053 1.602 2.529 1.261 0.9109 1.388 12.64 2.280 5.543 1.493 2.601 1.367
2.0 4.510 1.680 2.684 1.290 0.9725 1.386 15.01 2.445 6.140 1.543 2.621 1.361
2.1 4.992 1.761 2.834 1.320 1.030 1.384 17.63 2.616 6.740 1.592 2.642 1.356
2.2 5.497 1.845 2.979 1.350 1.083 1.381 20.48 2.791 7.339 1.642 2.664 1.350
2.3 6.027 1.932 3.120 1.380 1.133 1.378 23.58 2.972 7.936 1.691 2.687 1.345
2.4 6.581 2.022 3.255 1.410 1.179 1.375 26.93 3.158 8.529 1.740 2.710 1.340
2.5 7.159 2.115 3.385 1.440 1.223 1.372 30.53 3.348 9.117 1.788 2.733 1.336
2.6 7.762 2.211 3.511 1.471 1.264 1.369 34.38 3.545 9.699 1.837 2.756 1.331
2.7 8.389 2.309 3.633 1.502 1.303 1.366 38.48 3.746 10.27 1.885 2.780 1.327
2.8 9.041 2.411 3.750 1.532 1.340 1.362 42.84 3.952 10.84 1.933 2.804 1.323
2.9 9.718 2.516 3.863 1.563 1.375 1.359 47.46 4.163 11.40 1.981 2.828 1.319
3.0 10.42 2.623 3.972 1.594 1.408 1.356 52.35 4.379 11.95 2.029 2.852 1.315
3.1 11.14 2.733 4.078 1.626 1.439 1.352 57.49 4.600 12.50 2.077 2.877 1.311
3.2 11.90 2.846 4.180 1.657 1.469 1.349 62.90 4.826 13.03 2.124 2.901 1.308
3.3 12.67 2.962 4.279 1.688 1.498 1.346 68.58 5.057 13.56 2.171 2.926 1.304
3.4 13.47 3.080 4.374 1.719 1.525 1.342 74.52 5.292 14.08 2.218 2.952 1.300
3.5 14.30 3.201 4.467 1.751 1.552 1.339 80.74 5.531 14.60 2.264 2.977 1.296
3.6 15.15 3.324 4.557 1.782 1.577 1.336 87.22 5.775 15.10 2.310 3.004 1.293
3.7 16.02 3.451 4.644 1.814 1.601 1.333 93.97 6.023 15.60 2.356 3.030 1.289
3.8 16.93 3.579 4.729 1.845 1.624 1.331 101.0 6.274 16.10 2.401 3.057 1.285
3.9 17.85 3.711 4.811 1.877 1.646 1.328 108.3 6.530 16.59 2.446 3.084 1.282
4.0 18.80 3.845 4.890 1.909 1.667 1.325 115.9 6.789 17.07 2.491 3.111 1.278
4.1 19.78 3.981 4.968 1.940 1.688 1.322 123.7 7.051 17.55 2.535 3.139 1.274
4.2 20.78 4.120 5.044 1.972 1.708 1.320 131.9 7.317 18.02 2.579 3.166 1.271
4.3 21.81 4.262 5.117 2.003 1.727 1.317 140.3 7.586 18.49 2.622 3.194 1.268
4.4 22.86 4.406 5.189 2.035 1.745 1.315 148.9 7.859 18.95 2.665 3.222 1.264
4.5 23.94 4.552 5.259 2.067 1.763 1.312 157.9 8.135 19.41 2.708 3.250 1.261
4.6 25.04 4.701 5.327 2.098 1.781 1.310 167.2 8.414 19.87 2.751 3.277 1.258
4.7 26.17 4.852 5.394 2.129 1.798 1.307 176.7 8.696 20.32 2.794 3.305 1.255
4.8 27.32 5.005 5.459 2.161 1.815 1.305 186.5 8.981 20.77 2.836 3.332 1.253
4.9 28.50 5.161 5.523 2.192 1.831 1.302 196.7 9.269 21.21 2.878 3.360 1.250
5.0 29.71 5.319 5.585 2.223 1.846 1.300 207.1 9.561 21.65 2.921 3.386 1.248
5.1 30.94 5.479 5.647 2.254 1.862 1.297 217.8 9.855 22.09 2.963 3.413 1.245
5.2 32.19 5.641 5.707 2.285 1.877 1.295 228.8 10.15 22.53 3.004 3.439 1.243
5.3 33.48 5.805 5.767 2.316 1.892 1.292 240.2 10.45 22.96 3.046 3.465 1.241
5.4 34.78 5.971 5.825 2.347 1.906 1.290 251.8 10.76 23.39 3.088 3.491 1.240
5.5 36.12 6.140 5.883 2.377 1.920 1.287 263.8 11.06 23.82 3.130 3.516 1.238
5.6 37.48 6.310 5.939 2.407 1.934 1.285 276.1 11.37 24.25 3.172 3.540 1.237
5.7 38.86 6.482 5.995 2.438 1.948 1.282 288.7 11.69 24.67 3.214 3.564 1.235
5.8 40.27 6.656 6.051 2.468 1.962 1.280 301.7 12.00 25.09 3.256 3.588 1.234
5.9 41.71 6.832 6.105 2.498 1.975 1.277 315.0 12.32 25.51 3.298 3.611 1.233
6.0 43.18 7.010 6.159 2.528 1.988 1.275 328.6 12.64 25.92 3.340 3.633 1.233
6.1 44.67 7.189 6.213 2.557 2.001 1.272 342.6 12.96 26.34 3.382 3.655 1.232
6.2 46.18 7.371 6.266 2.587 2.014 1.270 356.9 13.29 26.75 3.425 3.675 1.232
6.3 47.73 7.553 6.318 2.617 2.027 1.268 371.6 13.62 27.15 3.468 3.696 1.232
6.4 49.30 7.738 6.370 2.646 2.039 1.265 386.6 13.95 27.56 3.511 3.715 1.232
6.5 50.89 7.924 6.422 2.675 2.051 1.263 402.1 14.29 27.96 3.554 3,734 1.232
6.6 52.51 8.112 6.473 2.704 2.063 1.261 417.8 14.63 28.36 3.597 3.752 1.233
6.7 54.16 8.301 6.524 2.733 2.075 1.259 434.0 14.97 28.75 3.641 3.769 1.233
6.8 55.84 8.492 6.574 2.762 2.087 1.257 450.5 15.31 29.15 3.685 3.785 1.234
6.9 57.54 8.684 6.624 2.791 2.099 1.255 467.5 15.66 29.54 3.729 3.801 1.235
7.0 59.27 8.878 6.674 2.820 2.110 1.253 484.8 16.01 29.92 3.773 3.816 1.236
7.1 61.02 9.073 6.723 2.849 2.122 1.251 502.5 16.36 30.30 3.818 3.830 1.238
7.2 62.80 9.270 6.772 2.877 2.133 1.249 520.6 16.72 30.68 3.864 3.843 1.239
7.3 64.61 9.468 6.821 2.906 2.144 1.247 539.1 17.07 31.05 3.909 3.855 1.241
7.4 66.44 9.667 6.869 2.934 2.155 1.245 558.0 17.44 31.42 3.955 3.866 1.242
7.5 68.30 9.867 6.918 2.962 2.166 1.243 577.4 17.80 31.78 4.001 3.877 1.244
7.6 70.19 10.07 6.966 2.991 2.176 1.242 597.1 18.17 32.14 4.048 3.887 1.246
7.7 72.10 10.27 7.013 3.019 2.187 1.240 617.3 18.55 32.49 4.095 3.896 1.248
7.8 74.04 10.48 7.060 3.047 2.197 1.239 637.9 18.92 32.84 4.142 3.905 1.250
7.9 76.01 10.68 7.108 3.075 2.208 1.237 659.0 19.30 33.19 4.190 3.913 1.252
8.0 78.00 10.89 7.154 3.103 2.218 1.236 680.5 19.69 33.53 4.238 3.920 1.255
228
Table II: Shock 'fube Data (continued)
(Equilibrium Air: PI = 10 bar; Tl = 293 K; Ul = 11.843 kg/m 3 ; Cl = 343.66 m/s)
Ms P21 T21 U21 C21 M2 1'2 PSI T51 U51 C51 h5/C~ 1'5
8.1 80.02 11.09 7.201 3.131 2.228 1.235 702.4 20.07 33.86 4.286 3.927 1.257
8.2 82.07 11.30 7.247 3.159 2.238 1.234 724.8 20.46 34.19 4.334 3.934 1.259
8.3 84.14 11.51 7.294 3.186 2.248 1.232 747.6 20.86 34.52 4.383 3.940 1.261
8.4 86.24 11.72 7.339 3.214 2.257 1.231 770.9 21.26 34.84 4.432 3.946 1.263
8.5 88.37 11.93 7.385 3.242 2.267 1.231 794.6 21.66 35.15 4.480 3.952 1.265
8.6 90.52 12.14 7.430 3.270 2.276 1.230 818.8 22.06 35.46 4.529 3.958 1.267
8.7 92.70 12.35 7.475 3.298 2.285 1.229 843.4 22.47 35.77 4.578 3.964 1.268
8.8 94.91 12.56 7.520 3.325 2.294 1.229 868.5 22.88 36.08 4.627 3.970 1.270
8.9 97.14 12.77 7.565 3.353 2.303 1.228 894.0 23.29 36.38 4.675 3.976 1.271
9.0 99.40 12.98 7.609 3.381 2.312 1.228 920.0 23.71 36.68 4.724 3.982 1.273
9.1 101.7 13.19 7.653 3.409 2.320 1.228 946.4 24.13 36.97 4.772 3.989 1.274
9.2 104.0 13.41 7.696 3.437 2.329 1.228 973.2 24.55 37.26 4.820 3.995 1.275
9.3 106.3 13.62 7.739 3.465 2.337 1.228 1000 24.97 37.55 4.868 4.003 1.275
9.4 108.7 13.84 7.781 3.493 2.345 1.228 1028 25.39 37.84 4.915 4.010 1.276
9.5 111.1 14.05 7.824 3.522 2.353 1.228 1056 25.81 38.12 4.962 4.018 1.277
9.6 113.5 14.26 7.865 3.550 2.360 1.228 1085 26.24 38.41 5.009 4.027 1.277
9.7 116.0 14.48 7.906 3.579 2.368 1.229 1114 26.66 38.69 5.055 4.036 1.277
9.8 118.4 14.69 7.947 3.607 2.375 1.229 1143 27.08 38.97 5.101 4.046 1.277
9.9 120.9 14.91 7.987 3.636 2.382 1.230 1173 27.50 39.25 5.146 4.056 1.277
10.0 123.4 15.13 8.026 3.665 2.389 1.231 1203 27.93 39.52 5.191 4.066 1.277
10.1 126.0 15.34 8.065 3.694 2.395 1.232 1233 28.35 39.80 5.236 4.077 1.277
10.2 128.6 15.56 8.103 3.723 2.402 1.233 1264 28.76 40.08 5.280 4.088 1.277
10.3 131.2 15.78 8.140 3.752 2.408 1.234 1295 29.18 40.35 5.324 4.100 1.277
10.4 133.8 16.00 8.177 3.782 2.414 1.235 1327 29.59 40.62 5.367 4.112 1.277
10.5 136.4 16.22 8.213 3.811 2.420 1.237 1359 30.01 40.89 5.410 4.124 1.277
10.6 139.1 16.44 8.248 3.841 2.425 1.238 1391 30.42 41.17 5.453 4.137 1.276
10.7 141.8 16.66 8.283 3.871 2.430 1.239 1423 30.82 41.43 5.495 4.149 1.276
10.8 144.5 16.88 8.316 3.901 2.436 1.241 1456 31.23 41.70 5.537 4.162 1.276
10.9 147.3 17.10 8.349 3.931 2.441 1.242 1489 31.63 41.97 5.579 4.175 1.276
11.0 150.0 17.33 8.381 3.962 2.445 1.244 1523 32.03 42.23 5.620 4.189 1.276
11.1 152.8 17.55 8.412 3.992 2.450 1.246 1556 32.42 42.50 5.661 4.202 1.276
11.2 155.7 17.78 8.443 4.023 2.454 1.247 1590 32.81 42.76 5.703 4.215 1.276
11.3 158.5 18.00 8.472 4.054 2.458 1.249 1624 33.20 43.02 5.7"3 4.228 1.276
11.4 161.4 18.23 8.501 4.085 2.462 1.251 1659 33.59 43.27 5.784 4.242 1.277
11.5 164.3 18.46 8.529 4.116 2.466 1.253 1693 33.97 43.53 5.825 4.255 1.277
11.6 167.2 18.69 8.556 4.148 2.470 1.255 1728 34.35 43.78 5.865 4.268 1.277
11.7 170.2 18.92 8.582 4.179 2.473 1.256 1763 34.73 44.02 5.905 4.282 1.278
11.8 173.1 19.15 8.607 4.211 2.477 1.258 1798 35.10 44.27 5.946 4.295 1.278
11.9 176.1 19.38 8.632 4.242 2.480 1.260 1834 35.47 44.51 5.986 4.308 1.279
12.0 179.1 19.62 8.655 4.274 2.483 1.262 1870 35.84 44.75 6.026 4.321 1.280
12.1 182.2 19.85 8.678 4.306 2.486 1.263 1906 36.21 44.99 6.066 4.334 1.281
12.2 185.2 20.09 8.700 4.338 2.489 1.265 1942 36.57 45.22 6.105 4.346 1.282
12.3 188.3 20.33 8.722 4.370 2.492 1.266 1978 36.93 45.45 6.145 4.359 1.282
12.4 191.5 20.57 8.742 4.402 2.495 1.268 2015 37.29 45.68 6.185 4.371 1.283
12.5 194.6 20.81 8.762 4.434 2.497 1.269 2052 37.65 45.90 6.225 4.383 1.285
12.6 197.8 21.05 8.782 4.466 2.500 1.271 2089 38.00 46.12 6.265 4.395 1.286
12.7 201.0 21.29 8.800 4.498 2.502 1.272 2126 38.36 46.34 6.305 4.407 1.287
12.8 204.2 21.54 8.819 4.530 2.505 1.273 2164 38.71 46.55 6.345 4.418 1.288
12.9 207.4 21.78 8.836 4.562 2.508 1.274 2202 39.06 46.76 6.385 4.430 1.290
13.0 210.7 22.03 8.853 4.594 2.510 1.275 2240 39.40 46.97 6.425 4.441 1.291
13.1 214.0 22.28 8.870 4.626 2.513 1.276 2278 39.75 47.18 6.465 4.452 1.292
13.2 217.3 22.52 8.887 4.657 2.515 1.277 2317 40.09 47.38 6.505 4.462 1.294
13.3 220.6 22.77 8.903 4.689 2.518 1.277 2356 40.43 47.58 6.545 4.473 1.295
13.4 224.0 23.02 8.919 4.720 2.521 1.278 2395 40.78 47.78 6.585 4.483 1.297
13.5 227.4 23.27 8.934 4.751 2.523 1.278 2435 41.12 47.97 6.626 4.493 1.299
13.6 230.8 23.52 8.950 4.782 2.526 1.278 2474 41.46 48.17 6.666 4.503 1.300
13.7 234.2 23.78 8.965 4.813 2.529 1.279 2515 41.79 48.36 6.707 4.512 1.302
13.8 237.7 24.03 8.980 4.844 2.532 1.279 2555 42.13 48.55 6.748 4.521 1.304
13.9 241.2 24.28 8.995 4.874 2.535 1.279 2596 42.47 48.74 6.789 4.530 1.306
14.0 244.7 24.53 9.011 4.904 2.538 1.279 2638 42.80 48.93 6.830 4.539 1.307
14.1 248.3 24.78 9.026 4.934 2.541 1.279 2680 43.14 49.12 6.871 4.547 1.309
14.2 251.8 25.03 9.041 4.964 2.544 1.278 2722 43.48 49.31 6.912 4.555 1.311
14.3 255.4 25.29 9.057 4.993 2.548 1.278 2764 43.81 49.49 6.954 4.563 1.313
14.4 259.1 25.54 9.073 5.022 2.551 1.278 2808 44.15 49.68 6.996 4.570 1.315
14.5 262.7 25.79 9.088 5.051 2.555 1.277 2851 44.48 49.87 7.038 4.578 1.317
14.6 266.4 26.04 9.105 5.080 2.558 1.277 2896 44.82 50.05 7.080 4.584 1.319
14.7 270.1 26.28 9.121 5.109 2.562 1.277 2940 45.15 50.23 7.123 4.591 1.321
14.8 273.8 26.53 9.138 5.137 2.566 1.276 2985 45.49 50.42 7.165 4.597 1.323
14.9 277.6 26.78 9.154 5.165 2.570 1.276 3031 45.83 50.60 7.208 4.603 1.325
15.0 281.4 27.03 9.172 5.193 2.573 1.275 3078 46.16 50.79 7.251 4.609 1.327
229
Table II: Shock Tube Data (continued)
(Ideal Gas: "1 = 1.4)
Ms 1'21 T21 £'21 C21 M2 "12 P51 T51 £'51 C51 h5/C~ "15
1.0 1.000 1.000 1.000 1.000 0.0000 1.400 1.000 1.000 1.000 1.000 2.500 1.400
1.1 1.245 1.065 1.169 1.032 0.1542 1.400 1.540 1.132 1.360 1.064 2.500 1.400
1.2 1.513 1.128 1.342 1.062 0.2877 1.400 2.237 1.263 1.771 1.124 2.500 1.400
1.3 1.805 1.191 1.516 1.091 0.4053 1.400 3.108 1.397 2.224 1.182 2.500 1.400
1.4 2.120 1.255 1.690 1.120 0.5101 1.400 4.167 1.536 2.714 1.239 2.500 1.400
1.5 2.458 1.320 1.862 1.149 0.6044 1.400 5.425 1.679 3.231 1.296 2.500 1.400
1.6 2.820 1.388 2.032 1.178 0.6897 1.400 6.893 1.829 3.769 1.352 2.500 1.400
1.7 3.205 1.458 2.198 1.208 0.7672 1.400 8.579 1.985 4.321 1.409 2.500 1.400
1.8 3.613 1.532 2.359 1.238 0.8380 1.400 10.49 2.149 4.881 1.466 2.500 1.400
1.9 4.045 1.608 2.516 1.268 0.9028 1.400 12.63 2.321 5.442 1.523 2.500 1.400
2.0 4.500 1.688 2.667 1.299 0.9623 1.400 15.00 2.500 6.000 1.581 2.500 1.400
2.1 4.978 1.770 2.812 1.331 1.017 1.400 17.61 2.687 6.552 1.639 2.500 1.400
2.2 5.480 1.857 2.951 1.363 1.067 1.400 20.45 2.883 7.093 1.698 2.500 1.400
2.3 6.005 1.947 3.085 1.395 1.114 1.400 23.53 3.087 7.623 1.757 2.500 1.400
2.4 6.553 2.040 3.212 1.428 1.157 1.400 26.85 3.299 8.137 1.816 2.500 1.400
2.5 7.125 2.138 3.333 1.462 1.197 1.400 30.40 3.520 8.636 1.876 2.500 1.400
2.6 7.720 2.238 3.449 1.496 1.234 1.400 34.19 3.749 9.119 1.936 2.500 1.400
2.7 8.338 2.343 3.559 1.531 1.268 1.400 38.21 3.987 9.583 1.997 2.500 1.400
2.8 8.980 2.451 3.664 1.566 1.300 1.400 42.47 4.234 10.03 2.058 2.500 1.400
2.9 9.645 2.563 3.763 1.601 1.330 1.400 46.95 4.489 10.46 2.119 2.500 1.400
3.0 10.33 2.679 3.857 1.637 1.358 1.400 51.67 4.753 10.87 2.180 2.500 1.400
3.1 11.04 2.799 3.947 1.673 1.384 1.400 56.61 5.026 11.26 2.242 2.500 1.400
3.2 11.78 2.922 4.031 1.709 1.408 1.400 61.78 5.307 11.64 2.304 2.500 1.400
3.3 12.54 3.049 4.112 1.746 1.430 1.400 67.17 5.597 12.00 2.366 2.500 1.400
3.4 13.32 3.180 4.188 1.783 1.451 1.400 72.78 5.896 12.34 2.428 2.500 1.400
3.5 14.13 3.315 4.261 1.821 1.471 1.400 78.61 6.204 12.67 2.491 2.500 1.400
3.6 14.95 3.454 4.330 1.858 1.490 1.400 84.66 6.521 12.98 2.554 2.500 1.400
3.7 15.80 3.596 4.395 1.896 1.507 1.400 90.92 6.846 13.28 2.616 2.500 1.400
3.8 16.68 3.743 4.457 1.935 1.524 1.400 97.40 7.180 13.57 2.680 2.500 1.400
3.9 17.58 3.893 4.516 1.973 1.539 1.400 104.1 7.523 13.84 2.743 2.500 1.400
4.0 18.50 4.047 4.571 2.012 1.553 1.400 111.0 7.875 14.10 2.806 2.500 1.400
4.1 19.45 4.205 4.624 2.051 1.567 1.400 118.1 8.236 14.34 2.870 2.500 1.400
4.2 20.41 4.367 4.675 2.090 1.580 1.400 125.4 8.605 14.58 2.933 2.500 1.400
4.3 21.41 4.532 4.723 2.129 1.592 1.400 133.0 8.984 14.80 2.997 2.500 1.400
4.4 22.42 4.702 4.768 2.168 1.604 1.400 140.7 9.371 15.02 3.061 2.500 1.400
4.5 23.46 4.875 4.812 2.208 1.615 1.400 148.6 9.767 15.22 3.125 2.500 1.400
4.6 24.52 5.052 4.853 2.248 1.625 1.400 156.8 10.17 15.41 3.189 2.500 1.400
4.7 25.61 5.233 4.893 2.288 1.635 1.400 165.1 10.59 15.60 3.254 2.500 1.400
4.8 26.71 5.418 4.930 2.328 1.644 1.400 173.7 11.01 15.78 3.318 2.500 1.400
4.9 27.84 5.607 4.966 2.368 1.653 1.400 182.4 11.44 15.95 3.382 2.500 1.400
5.0 29.00 5.800 5.000 2.408 1.661 1.400 191.4 11.88 16.11 3.447 2.500 1.400
5.1 30.18 5.997 5.033 2.449 1.669 1.400 200.6 12.33 16.27 3.511 2.500 1.400
5.2 31.38 6.197 5.064 2.489 1.676 1.400 209.9 12.79 16.42 3.576 2.500 1.400
5.3 32.60 6.401 5.093 2.530 1.683 1.400 219.5 13.25 16.56 3.641 2.500 1.400
5.4 33.85 6.610 5.122 2.571 1.690 1.400 229.2 13.73 16.69 3.705 2.500 1.400
5.5 35.13 6.822 5.149 2.612 1.697 1.400 239.1 14.21 16.82 3.770 2.500 1.400
5.6 36.42 7.038 5.175 2.653 1.703 1.400 249.3 14.71 16.95 3.835 2.500 1.400
5.7 37.74 7.258 5.200 2.694 1.709 1.400 259.6 15.21 17.07 3.900 2.500 1.400
5.8 39.08 7.481 5.224 2.735 1.715 1.400 270.2 15.72 17.18 3.965 2.500 1.400
5.9 40.44 7.709 5.246 2.777 1.720 1.400 280.9 16.24 17.29 4.030 2.500 1.400
6.0 41.83 7.941 5.268 2.818 1.725 1.400 291.8 16.77 17.40 4.095 2.500 1.400
6.1 43.24 8.176 5.289 2.859 1.730 1.400 302.9 17.31 17.50 4.160 2.500 1.400
6.2 44.68 8.415 5.309 2.901 1.735 1.400 314.2 17.86 17.60 4.226 2.500 1.400
6.3 46.14 8.658 5.329 2.943 1.739 1.400 325.7 18.41 17.69 4.291 2.500 1.400
6.4 47.62 8.905 5.347 2.984 1.744 1.400 337.4 18.98 17.78 4.356 2.500 1.400
6.5 49.12 9.156 5.365 3.026 1.748 1.400 349.3 19.55 17.87 4.422 2.500 1.400
6.6 50.65 9.411 5.382 3.068 1.752 1.400 361.4 20.13 17.95 4.487 2.500 1.400
6.7 52.20 9.670 5.399 3.110 1.755 1.400 373.7 20.72 18.03 4.552 2.500 1.400
6.8 53.78 9.933 5.415 3.152 1.759 1.400 386.2 21.32 18.11 4.618 2.500 1.400
6.9 55.38 10.20 5.430 3.194 1.763 1.400 398.8 21.93 18.18 4.683 2.500 1.400
7.0 57.00 10.47 5.444 3.236 1.766 1.400 411.7 22.55 18.25 4.749 2.500 1.400
7.1 58.64 10.74 5.459 3.278 1.769 1.400 424.7 23.18 18.32 4.814 2.500 1.400
7.2 60.31 11.02 5.472 3.320 1.772 1.400 437.9 23.81 18.39 4.880 2.500 1.400
7.3 62.00 11.30 5.485 3.362 1.775 1.400 451.4 24.46 18.45 4.946 2.500 1.400
7.4 63.72 11.59 5.498 3.404 1.778 1.400 465.0 25.11 18.52 5.011 2.500 1.400
7.5 65.46 11.88 5.510 3.447 1.781 1.400 478.8 25.77 18.58 5.077 2.500 1.400
7.6 67.22 12.17 5.522 3.489 1.784 1.400 492.8 26.45 18.63 5.142 2.500 1.400
7.7 69.00 12.47 5.533 3.531 1.786 1.400 507.0 27.13 18.69 5.208 2.500 1.400
7.8 70.81 12.77 5.544 3.574 1.789 1.400 521.3 27.81 18.74 5.274 2.500 1.400
7.9 72.64 13.08 5.555 3.616 1.791 1.400 535.9 28.51 18.80 5.340 2.500 1.400
8.0 74.50 13.39 5.565 3.659 1.794 1.400 550.7 29.22 18.85 5.405 2.500 1.400
230
Table II: Shock Tube Data (continued)
(Ideal Gas: I = 1.4)
Ms P21 T21 /121 C21 M2 12 PSI T51 /151 C51 h5lc~ 15
8.1 76.38 13.70 5.575 3.701 1.796 1.400 565.6 29.93 18.89 5.471 2.500 1.400
8.2 78.28 14.02 5.585 3.744 1.798 1.400 580.7 30.66 18.94 5.537 2.500 1.400
8.3 80.20 14.34 5.594 3.787 1.800 1.400 596.1 31.39 18.99 5.603 2.500 1.400
8.4 82.15 14.66 5.603 3.829 1.802 1.400 611.6 32.13 19.03 5.669 2.500 1.400
8.5 84.12 14.99 5.612 3.872 1.804 1.400 627.3 32.89 19.07 5.735 2.500 1.400
8.6 86.12 15.32 5.620 3.915 1.806 1.400 643.2 33.65 19.12 5.801 2.500 1.400
8.7 88.14 15.66 5.628 3.957 1.808 1.400 659.2 34.41 19.16 5.866 2.500 1.400
8.8 90.18 16.00 5.636 4.000 1.810 1.400 675.5 35.19 19.19 5.932 2.500 1.400
8.9 92.24 16.34 5.644 4.043 1.811 1.400 692.0 35.98 19.23 5.998 2.500 1.400
9.0 94.33 16.69 5.651 4.086 1.813 1.400 708.6 36.78 19.27 6.064 2.500 1.400
9.1 96.44 17.04 5.658 4.129 1.815 1.400 725.4 37.58 19.30 6.130 2.500 1.400
9.2 98.58 17.40 5.665 4.171 1.816 1.400 742.5 38.39 19.34 6.196 2.500 1.400
9.3 100.7 17.76 5.672 4.214 1.818 1.400 759.7 39.22 19.37 6.262 2.500 1.400
9.4 102.9 18.12 5.679 4.257 1.819 1.400 777.1 40.05 19.40 6.328 2.500 1.400
9.5 105.1 18.49 5.685 4.300 1.821 1.400 794.6 40.89 19.44 6.394 2.500 1.400
9.6 107.4 18.86 5.691 4.343 1.822 1.400 812.4 41.74 19.47 6.460 2.500 1.400
9.7 109.6 19.24 5.697 4.386 1.823 1.400 830.4 42.59 19.50 6.526 2.500 1.400
9.8 111.9 19.62 5.703 4.429 1.825 1.400 848.5 43.46 19.52 6.592 2.500 1.400
9.9 114.2 20.00 5.709 4.472 1.826 1.400 866.9 44.34 19.55 6.658 2.500 1.400
10.0 116.5 20.39 5.714 4.515 1.827 1.400 885.4 45.22 19.58 6.725 2.500 1.400
10.1 118.8 20.78 5.720 4.558 1.828 1.400 904.1 46.11 19.61 6.791 2.500 1.400
10.2 121.2 21.17 5.725 4.601 1.829 1.400 923.0 47.02 19.63 6.857 2.500 1.400
10.3 123.6 21.57 5.730 4.645 1.831 1.400 942.1 47.93 19.66 6.923 2.500 1.400
10.4 126.0 21.97 5.735 4.688 1.832 1.400 961.4 48.85 19.68 6.989 2.500 1.400
10.5 128.5 22.38 5.740 4.731 1.833 1.400 980.9 49.78 19.71 7.055 2.500 1.400
10.6 130.9 22.79 5.744 4.774 1.834 1.400 1001 50.71 19.73 7.121 2.500 1.400
10.7 133.4 23.21 5.749 4.817 1.835 1.400 1020 51.66 19.75 7.188 2.500 1.400
10.8 135.9 23.62 5.753 4.860 1.836 1.400 1040 52.62 19.77 7.254 2.500 1.400
10.9 138.4 24.05 5.758 4.904 1.837 1.400 1061 53.58 19.79 7.320 2.500 1.400
11.0 141.0 24.47 5.762 4.947 1.838 1.400 1081 54.55 19.82 7.386 2.500 1.400
11.1 143.6 24.90 5.766 4.990 1.839 1.400 1102 55.54 19.84 7.452 2.500 1.400
11.2 146.2 25.33 5.770 5.033 1.840 1.400 1122 56.53 19.86 7.518 2.500 1.400
11.3 148.8 25.77 5.774 5.077 1.840 1.400 1143 57.53 19.87 7.585 2.500 1.400
11.4 151.5 26.21 5.778 5.120 1.841 1.400 1164 58.54 19.89 7.651 2.500 1.400
11.5 154.1 26.66 5.781 5.163 1.842 1.400 1186 59.55 19.91 7.717 2.500 1.400
11.6 156.8 27.11 5.785 5.207 1.843 1.400 1207 60.58 19.93 7.783 2.500 1.400
11.7 159.5 27.56 5.789 5.250 1.844 1.400 1229 61.62 19.95 7.850 2.500 1.400
11.8 162.3 28.02 5.792 5.293 1.844 1.400 1251 62.66 19.96 7.916 2.500 1.400
11.9 165.0 28.48 5.795 5.337 1.845 1.400 1273 63.71 19.98 7.982 2.500 1.400
12.0 167.8 28.94 5.799 5.380 1.846 1.400 1295 64.78 20.00 8.048 2.500 1.400
12.1 170.6 29.41 5.802 5.423 1.847 1.400 1318 65.85 20.01 8.115 2.500 1.400
12.2 173.5 29.88 5.805 5.467 1.847 1.400 1340 66.93 20.03 8.181 2.500 1.400
12.3 176.3 30.36 5.808 5.510 1.848 1.400 1363 68.02 20.04 8.247 2.500 1.400
12.4 179.2 30.84 5.811 5.553 1.849 1.400 1386 69.11 20.06 8.313 2.500 1.400
12.5 182.1 31.33 5.814 5.597 1.849 1.400 1410 70.22 20.07 8.380 2.500 1.400
12.6 185.1 31.81 5.817 5.640 1.850 1.400 1433 71.34 20.09 8.446 2.500 1.400
12.7 188.0 32.31 5.820 5.684 1.850 1.400 1457 72.46 20.10 8.512 2.500 1.400
12.8 191.0 32.80 5.822 5.727 1.851 1.400 1480 73.59 20.11 8.579 2.500 1.400
12.9 194.0 33.30 5.825 5.771 1.852 1.400 1504 74.74 20.13 8.645 2.500 1.400
13.0 197.0 33.80 5.828 5.814 1.852 1.400 1528 75.89 20.14 8.711 2.500 1.400
13.1 200.0 34.31 5.830 5.858 1.853 1.400 1553 77.05 20.15 8.778 2.500 1.400
13.2 203.1 34.82 5.833 5.901 1.853 1.400 1577 78.22 20.17 8.844 2.500 1.400
13.3 206.2 35.34 5.835 5.945 1.854 1.400 1602 79.39 20.18 8.910 2.500 1.400
13.4 209.3 35.86 5.837 5.988 1.854 1.400 1627 80.58 20.19 8.977 2.500 1.400
13.5 212.5 36.38 5.840 6.032 1.855 1.400 1652 81.78 20.20 9.043 2.500 1.400
13.6 215.6 36.91 5.842 6.075 1.855 1.400 1677 82.98 20.21 9.109 2.500 1.400
13.7 218.8 37.44 5.844 6.119 1.856 1.400 1703 84.19 20.22 9.176 2.500 1.400
13.8 222.0 37.97 5.847 6.162 1.856 1.400 1728 85.42 20.23 9.242 2.500 1.400
13.9 225.2 38.51 5.849 6.206 1.857 1.400 1754 86.65 20.25 9.308 2.500 1.400
14.0 228.5 39.05 5.851 6.249 1.857 1.400 1780 87.89 20.26 9.375 2.500 1.400
14.1 231.8 39.60 5.853 6.293 1.858 1.400 1806 89.14 20.27 9.441 2.500 1.400
14.2 235.1 40.15 5.855 6.337 1.858 1.400 1833 90.39 20.28 9.508 2.500 1.400
14.3 238.4 40.71 5.857 6.380 1.859 1.400 1859 91.66 20.29 9.574 2.500 1.400
14.4 241.8 41.26 5.859 6.424 1.859 1.400 1886 92.94 20.30 9.640 2.500 1.400
14.5 245.1 41.83 5.861 6.467 1.859 1.400 1913 94.22 20.31 9.707 2.500 1.400
14.6 248.5 42.39 5.862 6.511 1.860 1.400 1940 95.51 20.31 9.773 2.500 1.400
14.7 251.9 42.96 5.864 6.554 1.860 1.400 1968 96.82 20.32 9.840 2.500 1.400
14.8 255.4 43.53 5.866 6.598 1.861 1.400 1995 98.13 20.33 9.906 2.500 1.400
14.9 258.8 44.11 5.868 6.642 1.861 1.400 2023 99.45 20.34 9.972 2.500 1.400
15.0 262.3 44.69 5.870 6.685 1.861 1.400 2051 100.8 20.35 10.04 2.500 1.400
231
Figures
Fig. 1: Driven section, nozzle, test-section and receiver tank of the Aachen
Hypersonic Shock Tunnel.
t
Time
Te~tin9
Time
1
Air
Stotion Dis.placemeont ~
t
Double Diaphragm
tlDul,.
L-_____D_ri_._'r__
s'_c_t_io_n____ ~)L)__________________A_i_r_S_._ct_i_on__________________-J~oot
L..~:~L--~~:ar-:_TU_be____: (
Fig. 3: Free piston shock tunnel: Schematic arrangement and wave diagram
of shock tube flow [12].
L-..J)-----~
t-
V!
w
1
x
Fig. 4: Schematic arrangement of the gun tunnel and operating cycle [10].
233
P, P,
/'
.,/'
I,-
t /'" ~
~ .... 41
5 \I) E
/ 41·-
/.,/' 10'-1-
<~
""~,
~
4'~ 1
x
Initial
PA
1 m P, ICondition
VA V,
Final
Ps fa Ps I Condition
VA + V,-v v
/
/ 5
/
\/
/
)
/
I'"~
"-
4 ~.
'"
x
234
10000 Isentropic
t
Compression.
Q.'"
V'1 .1
0 1000
~
..5
.In
In
ct
100
Ref lected - Shock
-- -- - Compression.
c: A'1 =1
Q
"0
c: 10
'"0
Vi
1~----~----~----~----~-'
1 10 100 1000 10000
Drive Pressure Ratio - P'1 _
11 Isentropic
Compression
V,,=1
t
,.;9
,
o Reflecled - Shock
&7
~ //
/'Compression
A,,=1
e. "
~
/
CI.
E
\I
,!
c
2
oc
'"
a
Vi
1~~~~--~----~----~--.
1
Drive Pressure Ratio - P,., --+
235
If--..--27m·-~·1
....---tP4
.
~: _=-----~a;~
Piston moves 48 check ~
Hydrogen at at 600 m/sec. valves open.
1000 atmospheres
-.
Po ~ 2000 atm.
To ,., 2000 0 K
Piston
rebounds
Upstream processes
unimportant.
8000
Q
...
o~
~ 6000
REFLECTED SHOCK
TUNNEL, 1000 ATM
w
..... DRIVER
Z
0
;::
-«
Z
0 4000 AEDC HOTSHOT I
-«
.....
V'l (106 JOULES)
2000
I 10
STAGNATION PRESSURE (ATM)
236
Fig. 9: Side view to scale of the Aachen shock tunnel with "nominal" exit
diameters of 500 mm and 1000 mm.
Distance ____
237
(a)
Undertailored can
Transmitted shock
.-
.- .-
~
~ (b)
~
/ Ovcrtailored case
Contact ~.-
surface -..,,,,
.- ~
/
.- ~"
Transmitted shock
(c)
Tailored case
238
II
~1F.r======~~~==~~
N2
10-2
10-4
10~
o
T (lOOOK)
p-O.OI br
cr.lcel CCIIPOIIt Ian or equlllbrllll air
Fig. 12: Equilibrium composition of air at a constant pressure of 0.01 bar [22].
x
Fig. 13: Solution of a Riemann problem in a space-time diagram [22].
239
! r----------,
- - reel gee
~ ,..------------,
1000
i-,..------------, i~ ,..------------,
- - reel gee
800
- . - ldeel gee
1.5
800
400
200
0.5
- - reel gee
_ . - ldeel gee
o
240
P41 -CD
20
p,"0.6bar
-- -- -- - - -
.- ----
10
-'-
.........
./'
/./'
./
~~~~------------------------
5 ./
--~------
~'---
P41
Fig. 15: Basic shock tube performance, air as equilibrium gas, driver gas ideal.
~104,----------r--------~r---------~--------~----------,-,
L
~ 5
~
III
0. 2
-2
10
P 1 [bar]
Fig. 16a: The influence of the driven gas pressure PIon the pressure Ps behind
the reflected shock with P4 = 1500 bar. Driver and driven gas
combination with their initial temperatures are indicated.
241
~ 10r---------~--------~~--------~--------~----------~
o
o --- H2 (493 KI / Air (293 KI
o
~ - -- - H2 (293 KI / Air (293 KI
If) 8 -.-.- He (493 KI / Air (293 K)
f- ------ He (293 KI / Air (293 KI
p. - 1500 bar
-2 -1
10 10 103
PI [bar]
Fig. 16b: The influence of the driven gas pressure PIon the temperature Ts
behind the reflected shock with P. = 1500 bar. Driver and driven
gas combination with their initial temperatures are indicated.
~ 10r-----~----~-----r----_r----~----~----~r_~~~~__,
....m
'-
!! 9
...c
o 8
Q)
'-
o
::: 7
.,m
ui 6
L
5
oL--~--~--~--~--~--~-~--~--~
300 350 400 450 500 550 600 650 700
T4 [K]
Fig. 17a: Tailored shock Mach number versus driver gas temperature T4 for
various driver/driven gas combinations and typical shock tunnel
driver gas temperatures.
242
Q)
~ 35r------r------r------r------r------r------~----~----__,
....
t.
Q)
+' 1: H2 / Air (0.01 bar)
c:
.... 30 2: H2 /Air(0.lbar)
u
Q) 3: H2 / Air (1 bar)
t.
o 4: H2 (ideal gas) / Air (0.01 bar)
:;:: 25
5: H2 (ideal gas) / Air (ideal gas)
'"
+'
6: He / Air (O.Olbar)
ui
2: 20 7: He / Air (ideal gas)
15 T,-293K
10
oL-----~----~------~----~----~------~----~----~
T 4 [K]
Fig. 17b: Tailored shock Mach number versus driver gas temperature T4 for
various driver/driven gas combinations and typical Stalker tunnel
driver gas temperatures.
End wall
I
Transmitted shock .
3 8 \ Contact
surface
Streamline
Boundary layer
243
'-20,Jol UC: t _ t9Sp SIte
,-'<4S IJ ue
Fig. 19: Interaction of reflected shock with boundary layer and interface.
Schlieren photographs of flow pattern for M. = 6, PI = 9.3 mbar,
152.4 mm x 88.9 mm shock tube; the times quoted are measured
after the reflection of the shock from the end wall (Taken from [33]).
10r---~----~---r---'-----'----~--~-----r--~----~-'
Ul
L 9 - - - - - H2 (van der Waals gas) / Air
- - - - H2 (ideal gas) / Air
8
PI • 0.5 bar
7 TI "293K
T4"293K
6
P41
Fig. 20a: Shock tube performance for van der Waals driver gas; Shock Mach
number as function of diaphragm pressure ratio
244
"'C\I 13
f-
12 H2 (van der Waals gas) / Air
- - - H2 (ideal gas) / Air ,/;
11
h
10 ~
PI = 0.5 bar
TI =293K
#'
9
T4 -293K
8
3
2
0
2 5 101 2 5 102 2 5 103 2
P41
Fig. 20b: Shock tube performance for van der Waals driver gas; Temperature
ratio T21 as function of diaphragm pressure ratio.
IE 100r------.--~r_--_,----n_--~r_~--r7--~--~~--~------,
~
.c
80
60
40
20
2 4 6 8 10
v"" [km/s]
Fig. 21: Shock tunnel stagnation pressures and stagnation temperatures re-
quired for flight duplication.
245
~
...
'.\
10' \
'"
a)
..•..
....'-.
",r-...
.~
~
"-
~
10' ~r-.-
0.2 0,4 0,6 0,8 1,0 1,2 ',4 t.b 1,8 2.0 2,2 2.' • 10'
M
~
~10'
\.
100 \"
\.
'\'\,
\,
I,\, b)
~
~
~
~
~ t--....
" ....... r--.
t----
0,2 0,4 0.& O,B I,D 1.2 1,4 1,6 1,8 2.0 2.2 2.4
• 10'
M·
Fig. 22: Steady nozzle flow from stagnation conditons of Po = 680 bar and
To = 5150 K. Solid line: equilibrium air; dotted line: frozen chemical
composition at stagnation conditions. Free stream conditions versus
Mach number
a) temperature, b) density
246
/v
/' V
/' V
/ ' ::,....-
10' /V
~
V
10' ~
.'
c)
1/"
...-
7'
If.'
i
10' I
P
1/
10" 1\
0.2 0,4 0,' 0,8 1,0 1,2 1,4 1,6 1,8 2,0 2.2 2,4 * 10'
H
"'" ~...
~\
, '.
\.
10' IY·...
\'" ...
1\:\ d)
10' ~
'" ~
K
'" '-... ~ I-
r---.... .......
0.2 0,4 0.' 0.8 1,0 1.2 1,4 1,6 l,B 2,0 2,2 2.4
• 10'
H
Fig. 22: Steady nozzle flow from stagnation conditons of Po = 680 bar and
To = 5150 K. Solid line: equilibrium air; dotted line: frozen chemical
composition at stagnation conditions. Free stream conditions versus
Mach number
c) area ratio and d) Reynolds number.
247
- - Tm= 50 K. 0=0.572 m
II
~ - - - Tm= 50 K. 0=1.073 m
1.0 _._.- Too -100 K. 0=0.572 m
0.8 - - - - - Tm=IOO K. 0=1.073 m
------- Too =300 K. 0=0.572 m
.............. Tm =300 K. 0=1.073 m
0.6
0.4
0.2
o 1
10 2 5 2 5 103 2
P5[bar]
2.4
2.1
2.0
I.B
I.-
75~m
• 1,0
X (Ns m- 2)
Fig. 24: Density relation behind a normal shock versus binary reaction vari-
able X = p(x)· x/u(x)
248
Fig. 25a: Just machined aluminum throat piece.
Fig. 25b: Aluminum nozzle throat and end wall eroded by oxygen of 2600 K
and 193 bar.
249
Re Re
, P4 =1500bar \
,A/A'=177
P4 =1500bar
'\ T4 = 493K
107
\
, T4 =493K 107 \
He/ Air
He/Air
"- "-
....
_-
....... ........
.... .... ~
1r1> '-1~OOO
'" ~e-entry path
,, l1=loba~
,,
,, '\
'\reentry path
\
\
105 10S
\
1rf 104
7 u~{km/sl 12 16 20 24 28 M~
~~'L 10·
~ {kg
l'.. P4 =1S00bar
P4 =1500bar T4 =493K
T4 ·493K He / Air
He/ Air 10-1
10-1
10-
"-
"-
"-
'\ '\ reentry path
P4 =500 bar
,
'\
T4 = 293K '\
\
\
\
\
7 um[kmls 7 u~{kmlsl
Fig. 28: Fig. 29:
Simulation region of the binary Simulation of (loo - U oo region.
scaling factor.
250
Re
fj1
V=O,01
h4,O
17=0,02
1rf.
" Strong viscous
" interaction
"-
'\
\
\
\ V:,OJ.-
-.,,- - - ,\
................. \
.... \
, /' R..ntry path
104
1, XT=0,07 V:l:1.se..... Rarefaction paramlter
YRe.;
2, ir= 0,2
3 XT= 0,49
X _K;~ VISCOUS Interaction parof!teter
- YRE;;;
4, XT=4,5
4 8 12 16 20 24 28 Ma>
Fig. 30: Reynolds-Mach-number plot with rarefaction and viscous interaction
parameter.
251
Fig. 32: Cutted side view of six component balance.
252
Moo ..._ =7.85
"..,or:
To =1500 K
oj
70
1-----370
To =3420 K
.-L....I.~_ _ bJ
--------~~-90---- 370
Mco=6.1B 1 ~=5.5%
l
-----ll-
To: 5180 K
-- cJ
90 -
370
253
~~
----+-----_.-
I
0.5
Pp
0.4 ~
0.3
-"-
"' 0.2
.:2
cu
c... 0.1
:::J
en
en
cu 0.0
c...
C1.
~tD ts=\2 IDS 0)
40.0
30.0 Ps
- 20.0 3 ms
-
'-
IC
.c constant
stagnation pressure
cu 10.0
c...
:::J
en
en
cu
c...
0.0
C1.
=
Fig. 35: Test conditions: Po 33 bar; To = 1500 K.
a) Pitot pressure in the test section.
b) Stagnation pressure ahead of the nozzle.
254
3.0
2.0
'-
ttl
e 1.0
(1J
1 - - - - - 4.6 ms--""';
t- test time
:::J
VI 0.0
VI
r-- 2,Sms-,
(1J
t-
o.
r
r a)
100.0
(1J
t-
:::J
VI 50.0
VI
(1J
a. 0.0
t-
b
0.0 2.0 4.0 6.0 8.0 10.
time [ms]
255
B.O
6.0
4.0
'-
ra
.:S
Q.I 2.0
'-
::::J
III
III
Q.I 0.0
'-
0..
al
600.0
·po=S80 bar
I
_ 400.0 3,Sms--""i
c-
constant
a
/O
Ps...= 263bar stagnation pressure
i:00 . 0
-1
Q.I I
'- !-2 ms
::::J
III
III
test time
~ 0.0
0.. b)
256
-
:i!l
~
1
--- contact surface, Al/"*-90
"CO
c: - - - contact surface. "1/l-30
surface, ,,/l-10
CO>
5 _._.- contact
.....
CO>
.=
........ ----- head of expansion wive
.-
U
0=
.=
en
2 "1 : A*
.....
--...
CO>
.=
0=
Air
.;= 10- 1 Tt -293K
...
c...
c...
Pl- 0.6 bar
-.............
c...
CO>
5
......
!
10-2
3 4 5 6 7 8
Fig. 38: Testing times: Arrival of contact surface at the end wall compared
with arrival of the head of the expansion wave.
257
Fig. 39: The starting process of nozzle and model flow: High-frequency
sequence of schlieren photographs.
258
Low Density Facilities
Georg Koppenwallner
Hyperschall Technologie Gottingen HTG
Labor Lindau
Max Planck Str. 1
3411 Lindau/Harz
Gottingen, Germany
List of Symbols
A reference area
cp pressure coefficient
D diameter
D drag
E energy flux
h specific enthalpy
hd specific enthalpy for dissociation
Kn Knudsen number
kr rate constant for specific reaction
1 body length
1 char characteristic flow field length
1r reaction length
L nozzle length
M molecular mass (kg/K mol)
Ma Mach number
IiI mass flow
n number density of molecules
p pressure
q dynamic pressure
q heat flux
Q heat flow
r recovery factor
r radius of curvature
Re Reynolds number
R
s
similarity parameter for a chemical reaction
S molecular speed ratio
259
T temperature
V magnitude of celocity vector
V volume flow (equal suction speed S)
u,v,w, velocity components
X,y,z local rectangular coordinates
Greek Symbols
Indices
c continuum
FM free molecular
i incident molecules
n mormal component to surface
r reflected molecules
s stagnation point on a body
t total conditions with isentropic compression
w wall
o stagnation chamber conditions
1 test section conditions, free stream
2 condition behind a normal shock wave
free stream
* distance from nozzle throat
260
1. Introduction
At very high altitudes where the molecular mean free path iI. is
much larger than the whole body length free molecular flow
exists. The condition is Kn 1= iI. /1 ~ 10.
261
Summarizing we have the following flow regimes which are also
explained in Fig.2
Kn 100= 1.26 . ~
Ma
Re
I'"
The isolated simulation of rarefaction effects in a wind tunnel
is therefore achieved if Mach- und Reynolds number of the flight
are duplicated in the facility.
i v <",
2 ;0 hd (requirement for dissociation).
In order to distinguish between frozen, nonequilibrium and
equilibrium chemistry we adopt the concept of a specific reaction
length 1. which is similar to the mean free path concept [4],[5]
as shown below
262
c
A
v
1r =
V
or 1 =
v (binary reaction).
V Z r p k
r
1 = V K·
IT Ma . A and with Ma » 1
r
8 z' z'
1 » A.
r
263
The second point concerns the combined simulation of aerodynamics
and chemical kinetics. This simulation can in principle only be
achieved in the nonquilibrium regime, where the recombination is
completely frozen and only dissociation occurs. An explanation is
given in Table 1 which explains the simulation requirements.
2M
It is evident that this kind of simulation would require a low
density wind tunnel with extremely high .gas velocities.
In a flow regime with shock waves and boundary layer the high
kinetic energy is provided by the high temperatures in the
boundary layer close to the surface.
In free molecular flow the high kinetic energy results directly
from the high free stream mass velocity , with which the particles
impinge on the surface.
It therefore results that gas surface interaction including
chemical surface reactions are of importance in the whole low
density hypersonic flow regime.
265
dynamic similarity laws for wind tunnel testing.
Gas surface interaction testing needs therefore specialized
facilities, like molecular beam systems or arc jet tunnels.
266
The highest Reynolds numbers necessary result from the low
altitude limit at Ma = 20, where the viscous parameter has a
value of Ma/~ =
.02
K 1. 5 . Ma < 10
nl ~
which gives
Rei min '" 0.15 . Ma
'" 4 for Ma 28.
We note that the following relation holds for the mass flow:
m= _Tl_ • Il (T) . D . Re - ~ . D . Re
4 1 lD 1 11
267
p
o
268
operation of low density tunneiS 1S therefore possible. Power
supply may however become extremely high for tunnels operating close
continuum flow limit and simulating the true flight
velocity. Fig. 8 demostrates this by showing the test section
kinetic energy flux , which would in windtunnels be necessary for
Hermes flight simulation.
269
4. The Different Elements of Low Density Tunnels
270
however the problem of gas dissociation in the stagnation chamber
and monequilibrium nozzle expansion, which results especially at
low densities in freezing of the dissoziation during the nozzle
expansion. Arc heated tunnels therefore are the testing tasks
defined in section 2.5 and will therfore be treated in section 5
as special type of facility.
4.2.1 The nozzle boundary layer thickness and the test section
core flow size
In the following a simple analysis based on experimental results
in various tunnels is given. From Pitot pressure and static
pressure measurements the nozzle boundary layer thickness was
determined. The evaluation of these data shown in Fig. 10 gives a
simple correlation formula between boundary layer thickness ~,
nozzle length L and a Reynolds number Re LO ' namely
k
with
..; Re LO
k = 6.25; Re LO :
. D
.Ma
we obtain
Ci
D/2
k . (
K -
2
1
1/4
) /2-. tg e
Reo ~
2
tg e
k2
/ K - 1
2
Ma
272
or A
D
< / 2K
K - 1
. tg e .
Re D min
3970
1.27 . 10~ .
4.2.3 The free jet expansion for highly rarefied flow conditions
For the free jet expansion only a sonic nozzle or even a thin
orifice is used. A wall boundary layer ist therefore only in the
subsonic flow possible. The flow expands there after in the
273
supersonic parts without guidance by walls into the tunnel test
section. The flow field of free jets from sonic nozzles has been
the subject of many investigations [18,19,20] and is extremely
well known. As standard reference the work of Ashkenas and
Sherman [18] is world wide adopted. The striking features of the
free jet flow field are the extreme independence of streamlines
and Mach number from the stagnation conditions. Fig. 13 shows the
structure of a free jet, with straight radial streamlines in the
farfield and a low density flow region, which is shielded from
the surrounding higher pressure gas by the barrel shock and the
Mach disc. Mach disc position and therefore the size of the free
jet is determined by the pressure ratio po/p~. Fig. 14 shows the
Mach number distribution along the jet axis, which depends only
on the normalized distance X/D. Some important formulas for the
calculation of the flow properties are summarized in Fig. 15. For
free jets there exist also a lower useful operating limit, which
is given when viscous effects influence the sonic nozzle flOW,
this is the case when the nozzle throat Reynolds number becomes
smaller than Re oo ~ 100.
In the widely used conical nozzles and in the free jet expansion
no uniform test section flow can be obtained.
The test section flow has usually radial and angular gradients of
density and pressure. The streamlines are divergent and seem to
emanate from a virtual source point. For an inviscid conical flow
and for the free jet expansion this source points is with a
good approximation located at the nozzle throat. At hypersonic
conditions the following relation for the gradients exists:
1<
~ = -2 . q (x )
1<
dx X
1<
d Ma = (Ie -1) . Ma (x )
1<
dx X
274
The normalized gradients however as shown in Fig. 17 remain
constant.
Fig. 18 shows the Mach number along the axis of the SR 3 tunnel
at various stagnation conditions [21]. Included are the actual
and the normalized gradients. The normalized gradients are
somewhat smaller than inviscid theory would predict, which is
caused by the contouring effect of the nozzle boundary layer. The
next Figure 19 shows the complete flow flield at an even smaller
operating Reynolds number of the tunnel. Here the axial gradient
becomes even smaller due to the better contouring of the thicker
boundary layer, which however reduces the useful test core size.
/2 .e .
tg
Ma
Re D
275
shows for an ideal dissociating gas [24] the degree of dissociation a
as function of T and p. We note, that the enthalpy for a
dissociating gas is given by
h *~ RT + hd a
Fig. 24 shows as result of Bray [26] the area ratio in the nozzle
at which the dissociation will become frozen. Nozzle size is
introduced by the parameter ~.
We can generally conclude that in nozzles of low density facilities
freezing of vibration or if the temperatures are high enough
freezing of dissoziation will occur. This freezing will change
the test section flow conditions to a large extent.
277
The open test section.
"
1
" + 1
" - 1
. (_.:.:,,:..-.:.+---=1=--_ " - 1
" - 1 2 • " • Ma 2
278
1 K+1
=( 1
K
K - 1
. ( /(. +
2
1
K-1
279
steam ejectors (Princeton)
air ejectors (AEDC)
vapour booster pumps (CNRS)
roots-blowers (DLR, Imp. College)
diffusion pumps (DLR)
cryogenic pumps (DLR, Univ. of Southern Cal.)
steam ejectors are usually able to work at pressures between .05
and 1000 mbar and can handle ext rem large volume and massflow
ranges. In order to obtain the compression to atmospheric pressure
4-5 stages with intermediate condensors are necessary.
Fig. 28 shows the principal set up of such a system.
At the time between 1960 and 1975, when most of the present
available low density facilities have been installed, many of
these tunnels where equiped with arc heaters. Arc heaters [31]
working principle and operational bahaviour is explained below:
2W
Working Principle: An electric arc column heats the gas which
flows through the heater.
Within the arc column the gas is electric
conducting and ionized.
The arc column is maintained between two
electrodes.
Many different ways for the arc column path
have are in use.
Main problems Heating of the electrodes, therefore
Strong cooling necessary
Rotating arc principle .
Hollow electrode heaters.
Heating of the whole arc chamber at high
pressures. Above p ~ 100 bar the arc becomes
optical thick, radiative heat transfer to the
wall becomes extremely high.
Strong water cooling and well placed test gas
injection necessary.
Arc stabilization, therefore vortex tube gas
injection.
Rotating arc heaters.
Rotating arc heaters.
Fig. 34 shows a 1 foot hypersonic tunnel, located at NASA Langley
[32], whose heater works on the rotating arc principle.
Arc rotation is achieved by a magnet coils surrounding the arc.
The Lorenz forces of the magnetic field rotate the arc at high
speed, thus reducing the local electrode heating.
Hollow electrode heaters.
Hollow electrode type heaters with a DC-power-supply are used at
the DLR tunnels at Cologne [14 ]. Fig. 35 shows the heater, whose
hollow electrodes are water cooled and therefore the flow
contamination due to erosion of the electrodes is reduced. Due to
the reduced contamination test gases like air may be used in this
facility, which might be of interest for the study of atomic
oxygen reactions on the model surface.
Coaxial electrode heaters.
As example the heater of the IRS Stuttgart [33] is shown in
Fig.36. From the cathode the arc discharge is maintained through
the nozzle throat to the anode, which forms part of the
supersonic nozzle. As the gas will be heated during nozzle
expansion it seems quite difficult to define equivalent
stagnation chamber conditions for the expansion.
281
As these heaters are operated at· temperatures above the
dissociation limit of the test gases TO > 3000 K the dissociated
molecules will in general not recombine in the low pressure
nozzle expansion. Therefore the freestream gas in the test
section will be in a strong chemical nonequlibrium state, which
is of main disadvantage for aerodynamic testing in these type of
facilities. This flow of dissoziated gas my however be used for
testing material behaviour and gas- surface reactions . For this
task the arc jet shall however simulate the actual free flight
flow close to the vehicle wall. This means a simulation of the
free flight flow at the edge or inside the the vehicle boundary
layer. To establish general similarity conditions for this type
of testing is still a task, which must be attacked.
Another important limitation of arc heaters is given by the upper
limit for the operating pressure of Po * 100 bar, which does not
allow the simulation of flow conditions close to continuum flow.
Due to limitation of the operating pressure of the Arc heater
the simulation capability for aerodynamic testing (Re- Ma-
simulation) or for chemical reactions in a vehicie flow field
does not at all reach the necessary similarity parameters of
the free flight values .Due to the pressure and power
requirements it seems unlikely that this can be achieved with a
contiously running arc jet.
An arc jet facility offers however the possibility to simulate
the actual in flight values of aerodynamic heat flow and pressure
on a vehicle. Fig.37 and 38 show as a demonstrative example the
performance of the DLR facility in terms of heat flow
duplication. The following table therefore summarizes the main
tasks for testing and research in low density arc jet facility.
282
6. Molecular Beam Facilities.
In the present context the molecular beam will be used for the
study of reactive and nonreactive gas surface reactions.
Fig. 39 explains the general question of gas surface interaction.
For such a study it will be necessary simulate the following
conditions:
a. Use beam gases , which represent the individual
components of the in flight gas impinging on surfaces.
b. Duplicate velocity, which means energy of the impinging
particles.
Fig 40 shows in an altitude velocity chart the necessary kinetic
energy Ekin of beam particles ,given in electron Volt eV, which
is a standard measure in molecular beam systems [35].
It is clear that there arise for the production of high energy
molecular beams similar problems as for production of high energy
wind tunnel flows.
A molecular beam is usually produced by expansion of the
test gas from a source into vacuum . There exist however also
283
other methods for high energy beam production. Therefore a short
summary of various principles for molecular beam production is
given in the following.
v = -c / 8
n
284
method was first proposed by Kantrowitz [36] and later optimized
by many investigators like Bossel [37], Campargue [38]
As also shown in Fig. 41 the test gas is first expaned in a continuum
jet to high Mach numbers in order to obtain a mass velocity V,
which is determined by the·source temperature. This requires the
addition of a gasdynamic expansion chamber with a nozzle.
From the jet a molecular beam is extracted with a supersonic
skimmer. The skimmer is placed at a jet position, where the mean
free path equals about the skimmer entrance diameter.
Skimmer outside and inside angles have be selected in such a way
, that no shock waves are introduced. The beam extracted by the
skimmer runs then through a collimation chamber, which has to be
kept at very low vacuum in order to avoid again attenuation of
the beam by collisions with background molecules.
The beam velocity equals the maximum velocity obtained with
gasdynamic expansion:
Ekin = .5 m V 2 =
285
~ Electrical resistance heaters.
With this type of heaters stagnation temperatures up to
2500 • K can be achieved. Heater materials are Kanthal,
Tungsten or Tantal.
~ Arc heaters and induction plasma heaters [39],[40].
With this type of heaters higher temperatures up to 10000 K
can be achieved.
There exist however also other methods for producing high energy
molecular beams.
2~
energy gain is obtained. The experimental energy gain factor was
8.4 compared to a theoretical maximum gain of 11.
In a molecular beam system the main gasdynamic expansion is after
the skimmer continued by a pure collision free expansion.
During this expansion the light carrier gas of the seeded beam
will due to its higher thermal speed spread much more than the
havier test gas. The beam core therefore gets continously
enriched with the havier test gas.As result of this different
thermal spreading the beam will behind a collimator chamber
mainly consist of the havy high energy component.
The seeded beam technique therefore offers a clean way to produce
high energy molecular beams of heavy particles.
2'67
6.7 Large Molecular Beams as Free Molecular Windtunnel.
A large molecular beam , with a diameter d much larger than the
mean free path, could be used as wind tunnel. In order to obtain
such a beam with reasonable pumping capacity Legge [47] proposed
the beam skimmming to be performed at continnum flow conditions.
The first skimmer has to be designed and placed into the jet in
such a way that no gasdynamic disturbances run from the skimmer
leading edge into the jet centerline. Alomg the jet centerline
expansion shall then continue in the same rate as without
skimmer. A concept of this system, which is presently under
development is shown in Fig. 46.
7. Conclusions
288
8. References
289
[15] Maslach, G.J., Sherman, F.S., Design and Testing of an Axi-
symmetric Nozzle for a Low Density Wind Tunnel,
WADS-TR-56-34-341, University of California, Report 150-134,
1956.
[16] Vas, I.E., Allegre, J., The N-4 Hypersonic Low Density
Facility and Some Premilinary Results on a Sharp Flat Plate.
Rarefied Gas Dynamics, Vol. II, 1967, pp. 1015-1030. Proc.
5th Int. Symposium. New York: Academic Press, 1967.
[17] Hefer, G., Kienappel, K., Erprobung einer mit Stickstoff
gektihlten Dtise des Hypersonischen Vakuumwindkanals. DLR-FB
70-41, 1970.
[18] Ashkenas, H., Sherman, F.S., The Structure and Ultilization
of Supersonic Free Jets on Low-Density Wind Tunnels.
Proceedings of the 4th Int. Symposium on Rarefied Gas
Dynamics, Toronto, 1964 (ed. J.H. de Leeuw), Vol. II, pp.
84-105, New York: Academic Press, 1966.
[19] Bisch, Ch., Etude de jets libres et de jets emis an culot
d'un cylindre place dans un ecoulement hypersonique.
Aeronautique et l'Astronautia ve No. 63, pp. 31-41, 1977-2.
[20] Christ, S., Sherman, D.M., Glas, D.R., Study of the Highly
Underexpanded Sonic Jet. AIAA Journal, Vol. 4, 1966,
pp. 68-71.
[21] Allegre, J., Private Communication.
[24] Lighthill, M.J., Dynamics of a dissociating gas, Part I
Equilibrium Flow .,J. Fluid Mech. Vol.2 1957, pp 1-32
[25] Stollery ,J.L.,Park,C., Computer solutions to the problem
of vibrational relaxation in hypersonic nozzle flow.
J. Fluid Mech. Vol.19, 1964, pp 113-123
[26] Bray K.N.C , Atomic recombination in a hypersonic wind
tunnel nozzle., J. Fluid Mech. Vol.6, 1959, pp 1-32
[27] Koppenwallner G., Review of Flow Quality obtained in Conical
and contoured Nozzles of Hypersonic High Enthalpy Wind
Tunnels.( Vibrational Relaxation).
DFVLR IB 222-88 A 33 , 1988
[28] Anonym, Handbook of Supersonic Aerodynamics, Section 17,
Ducts, Nozzles and Diffusors. NAVWEPS-Report 1488, Vol 6,
U.S. Government Printing, 1964.
[29] Allegre, J., Raffin, M., Etude experimentelle d'un diffuseur
en ecoulement hypersonic de gaz rarefie. Lab. d'Aerothermique,
Paris, Rep. 68-4, 1968.
[30] Koppenwallner, G., Der hypersonische Vakuumwindkanal der
Aerodynamischen Versuchsanstalt GOttingen. Betriebsverhalten
und erste Ergebnisse tiber reale Gaseffekte in Dtisenstromungen.
DLR-FB 66-62, 1966.
290
[31] Anonym, ARC heaters and MHD accelerators for aerodynamic
purposes. AGARDograph 84 , 1964
[32] Boatright, W.B. et alea ,Summary of some of the Arc Heated
Hypersonic Windtunnel Development Effort underway at the
Langley Research Center. AGARDograph 84 ,1964,pp 353-378
[33] Auweter-Kurz et alea, Steady State MPD Devices for Reentry
Simulation., DGLR/AIAA/JSASS International Propulsion
Conference, Garmisch Partenkirchen ,W.Germany 1988
[34] French, J.B., Molecular Beams for Rarefied Gas Dynamic
Research. AGAR-Dograph 112, 1966.
[35] Bossel U. Erzeugung intensiver Molekularstrahlen sub-
orbitaler Energien fur Streuexperimente. DLR- FB 72-52, 1972
[36] Kantrowitz, A., A High Intensity Source for the Molecular
Beam. Rev. of Scientific Instruments, Vol. 22, May 1955,
pp. 328-332.
[37] Bossel U. Molecular beam Extraction from Equilibrium gas
Flow., AIAA J. Vol.9 1971,pp 2060-2062
[38] Campargue R., Lebehot A., High intensity supersonic
molecular beams with Extremely narrow energy spread.
Rarefied Gas Dynamics. Proc. 9th Int. Symposium,
DFVLR Press 1974
[39] Knuth,E.L. Kiuluva, N.M. Performance of an Arc- Heated
Supersonic Molecular Beam and its Application to Molecule-
Molecule Collisions. AGARD CP 12 (1967) pp 277-338
[40] Stark J.P.W , Kinnerslay M.A., Development of a Low Power,
High Velocity Atomic Oxygen Source. in Rarefied Gas
Dynamics ,Edited byE.P. Muntz et alea. Vol. 116 Progress in
Astronautics and Aeronautics 1989, pp 156-170.
[41] Skinner ,G.T., On the design of experiments with a Shock tube
Driven Molecular Beam. AGARD CP-12 pp 423--441
[42] Jones T.V., Experiments on the Formatiuon of a 2 eV Argon
Beam from a Shock Tube Source, Rarefied Gas Dynamics, Proc.
5th Symposium, Academic Press 1967 ,pp 377-409. 09.
[43] Cross,J.B., Blais N.C., High -Energy/Intensity CW Atomic
Oxygen Beam Source., in Rarefied Gas Dynamics, Edited
by E.P. Muntz et alea. Vol. 116 Progress in Astronautics and
Aeronautics 1989, pp 143-154 ..
[44] Abuaf, N.,Anderson ,J.B., Andres, R.P., Molecular beams with
Energies above one Electrn Volt. SCience, 155 (1967), 997-999
[45] Abuaf, N.,Anderson ,J.B., Andres, R.P.,
Studies of Low Density Supersonic Jets, in Rarefied Gas
Dynamics, Proc. 5th Symposium, Academic Press 1967 ,pp
1317-1336
291
[46] Sjolander G.W., Froechtenicht J.F., Laboratory Results for
5-ev Oxygen Atoms on Selected Spacecraft Materials.
in Rarefied Gas Dynamics, Edited by E.P. Muntz et alea.
Vol. 116 Progress in Astronautics and Aeronautics; 1989,
pp 180-179.
[47] Moser H.O. Investigation of the influence of low density gas
atmospheres on spacecraft by means of accelerated cluster ion
beam, Z. Flugwiss. Weltraumforsch. ,11 (1987) pp 291-294
[48] Dankert, C., Legge, H., High Intensity Molecular Beams
Skimmed in Continuum Flow. Rarefied Gas Dynamics, Vol, 74,
pp. 882-894. Progress in Astronautics and Aeronautics, 1981;
ed. S. Fisher.
292
Table 1 Simulation of reentry aerodynamics
Simulation in wind tunnel requires the duplication of the relevant similarity parame-
ters.
These are:
I. For compressibility the Mach number
V
Ma=~
aoo
RS = 2 = ( p ~kr ) / L = p\ • l
with k, the reaction constant.
In case the velocity is duplicated k, will be the same as in the free night.
Therefore (p. Llsoo'• = (p • L)ru,ht must be duplicated in order to achieve a
simulation of reaction kinetic.
293
300, III geost. II 3001r-~'-----------~--~
Shuille transf.
km Orbiter Orbit km
free molecular
flow
200·-
12001-
, Kn: ~ :10
...- ...-
.c I .c near tree
rn rn
.c ~ AOTV I .c molecular flow
~ --4------------- Kn:.1..: 0.1
100·- \ J 100 flow I
'-/ .
I amlnar Ma/YRe;'=0.1
10 6
Rel:
(l:10m)
OL--_-L_ _ . L - - - I . - . L . ._ _" ' - -_ _
OLI--~~--~~--~~--~~~~~~
o 2 4 6 8 km/s o 2 4 6 km/s
fl ight veloci ty V ~ flight velocityV ..
80
Continuum flow
60
h
l.O Chemical effects Chemical nonequilibrium _
unexis!ent _ _. _ . _ . - .
or frozen
Chemical equilibrium
2 3 5 6 7 km/s 9
u-
0.2
( )
model
scale
0.1
0.06
1 O.Ol.
/j
V1G
Ma (1/160 I
{rfe 0.02 I
Sl.Ma
(1/l.0)
he'"/~.ec'
,o~
~~"5C.\,e
0.01 Y,e '\
R3 CHI
-:>'(
,~ S'<;-
(1/100) / e<:-\jS
0 5 20 25 30
295
o
-m
heater nozzle test section diffusor cooler vacuum
stagn. o diameter pumping
chamber I model lenght system
1000
bar
1
800
...
QI
.0
E
0
.t:
0
c
.2
"0 To=2000K
c
01
0
Vi ·~IMa-6]!
I
Po
model length I
t = rIJ 12.5
vehicle len th L =30 m
0
10 15 20
flight Mach number Ma
25
..
Tunnel stagnation chamber pressure Po for
Moch- Reynolds number simulation of reentry
Fig. 6
296
K r I I 1000
100 Ma .
\ - Re - simu Ia I'Ion
3000 50 MW[- with
condens. 10 V and p·L duplication
at equil.
vapour ~
pressure
I
·w
l
,g 100
polbarJ ,..
2000 ~
e
....
.!.!
TOmlO
Q;
....MrW:th 20K .!:
-"
co~dens. ~
~10050
~ ,.join" ./..... 10 e
~ ''[: ,,,' "'''00 '.'m
model lenglh t,p.25m ~
u
OJ
Ul
iii
2 -CI [_E.t·Vl+Dl]~ ~ T
11 I
0 1 I I
1[, 10 15 20 25
8 10 12 16 IS 20
flight Mach number Ma
Ma
N
\0
00
GRAPHITE
WATER-IN.OUT
a RADIATION
POWER
NOZZLE SECTION t
0.15 ------;---------+--+-------1
1 t
6 Windtunnel 140
r
-----+---+-0--- ... Univ. von Toronto 2
0.10
... ~
• .. Serke(ey 6
• Princeton Univ. 24
[J AVA -V1G =21
(;) AVA - V 1 G 7
0,05 ---+--7F-
6 = boundary layer
6,= boundary layer
displacemen t thick ness
°
° 0,01 0,02
1
0,03
lRe .'---
Lo'
299
8 ~.
~
; ,I.
._._._.,?O .-.-.-.-.-.-tk~ '&
N2-paSSage
....
~
/'
j. 452.5 ~-4!
Fig. 11: The liquid nitrogen (LN2) cooled nozzle for VIG.
Q25 ---_.•. _.
Ma
c 2 Toronto}
4 Toronto
"0 6 8erkely
uncooled
aI5~-------~---~-r+-------;--------~
0.; ·Y
t:;.
22 20 300 cooled
• 22 30
90 } LN2
A 22 20 90
90 cooled
0.05 • 22 10
o
0.005 0.015 0.025 0.045
301
Jel boundary
... 1
Iota· A(-0-
x•x )
10
91--+-1--+-1-~-~~-~-r-r-~
Ashkenas - ..... V
8.1--+-1--+-1 Sherman the?
7~~~-~-+_+--r~~1~L-r--r~~
6
I I ~ ./.~ 0
5
-Oll~
- . -x /~,.4::.-~-+--E.....Jx'-p-er..J.im-.-O...L.A-N-K.l-ER-T-i
U'Ma
~
~o
Ma
I.J---.---r---r..r--t--t--l 0 = 10 mm
3:I---I--I-..Q/~-1--t--i Po = 0.5 ala
o To = 290 K
21-4--.:J 4---I--+--t----l 0 s tat ic pr e ssure
o
l . impact pressure
1 0
o ~fl
-1 0 2 3 I. 5 6 7 8 9 10
x/D - - -
302
Free jet formulas accordi ng Ashkenas I
Sherman
01 Condilions along axis x (x/D 21
Mochnumb-er Mo
x- X x-I
! (x+l)
2 x-I
Ha = A'(T) X- X x-I
A'(T)
Pitot pressure p 2
-- t,
x 1
~ 1t-l 1
Pt,2 = ('K+l) .(X+l) • ~~_ __
Po x-I 2lt ...L.
It-l X - Xo
2
A • (-n-)
position of Mach-disc
bl Angular profiles
p(r.El) 2 Tt El
= cos (2' q;)
p(r.9=0)
fla (r , Ell
Ha(r,8=0)
= ( e
1
Tt
COS(2' Cli)
r-
.
1
Pig. 15
303
Conical nozzle
nozzle type flow regime
conventional -.......
conical nozzles continuum
Ma
slip flow I
q
Reo:: 35.~
tg <3
L ./
liquid nitrogen cooled
Conditions at distance L from throat:
short nozzles slip flow
merged layer
MaiL) ; d Ma
d x
~ Reo ;c 3.5. Ma
tg e q III ; ~
d x
cooling
Normalized gradien Is:
sonic nozzle with
free jet expansion slip to d MajdX = lI.-1
Ma III L
free mol. flow
~/ dx =-2
. q III L
ReDO;:;
~ . 100
dq 2 d Ma
-q-= - lI.-1 . MO
Fig. 16: The various flow expansion methods
in rarefied gas dynamics. Fig. 17: The axial gradients in hypersonic conical
nozzles.
nozzle dMa [_1] dMa x·
exit dx mm Cii('Ma
0.01 0.328
Ma Relem
20.3 7L87
0.009 0.329
1
Me
20
18.2 7770 0.009 0.368
16.3 5115 0008 0.350
15 lL.8 L675
x": 667
-
10
0 100 mm 200
x
o ~--~--~--~--~--r---r-~~---d~r~.la-------
~--
Moehnumber
~S~~~§§~~~~~~=
::19
gradients
d:~a : 0.0058811Jmm)
-5
_ _ _.1..1_ _ _ _ _ _- ' - -_ _ _ _- - 1 . _ dMo .~: 0.205
y 30 35 em LO dx Me
x---
x
305
Tunnel O[cm] e Ma
o V1G 25 10° 7 - 22
1-40 dMo
0. V2G l.0 15° 10 • y.-
R
23 10° 15-20 tD
• SR3
• SR3 30 12.5° 15 -20
1
6
0/2 =
........ 0. t
Y 2 Ma ' 6
l..1 8 . tg8. Reo 0/2
empirical
boundary layer
O.S
thickness
formula (x:1.l.1
..-=;=j"":J " I , .! I 0
0.1
2 Ma ..
Ig 8Re o
0.2
1
0.15
ev
h 0.1
0.05
a 1000 2000 K
T ..
Fig. 21: Ratio of vibrational energy to total enthalpy of
ideal nitrogen.
a.(p.Tl
0.8
0.6
O.L.
0.2
O~~~--~~~~--~----~--~--~
2000 L. 000 8000 K
Temperature T
Fig. 22: Dissociation degree of oxygen and nitrogen as
function of temperature and pressure.
3m
102~------------------------------~
m·bar
To=1.2 ·T *
po=1.9·p*
10 '
CAL SPAN
r*·p*
tg ~ 0.6
Calibrated 0.8
TFITO in V lG
( Danker t )
1.0
10- 2
1000 2000
T* ..
3000
308
15~----------~--------~--~----~~~~
--1
To- 0'10
0'12
15
11~-+--~--~~~----------~------------~
2 Wood
(1956)
90~----------~1~~----------~2~~----------~3~
Fig. 24: The rate parameter ~ and area ratio for sudden
freezing of dissoziation according to Bray.
309
closed lesl seclion diffusor
I simple lubl! diffusor
O.Sr.- - - - - - - - - - - - - - - . . . . . ,
Mo.22 flow air inlet
____ -
----- - -EE:-.-
. 00
.~ ....... > " , . , . ....... -...,.
t;;;;-r'r.... ~'-:::.:*. -+3-
boundary 'posItion of shock 0.6
open lest section layer undetermined
with nozzle exponsion diffusor
conslricled lube diffusor ( Po)
./ 1'" ....
maximum
I
J_~~
,
1__ :;a~:E§ -E} ..!'L
I" recovery safety
'compression shocks Pu ~orgin. ~
'!- slart in boundary layer
w 0.2 operating
centerbody diffusor cond i t ion m;.
~ !(~Po.;)Ea. \
o open lesl seclion
wi" " .. ~ iol .hoI., ,,' 3.10' 4.10' Reo ~
1 as diffusor
r:ff~8- I I I
compression shock 20 40 60 bar 80
°°I
in inviscid core Po
--r::: -
Fiq. 25: The different test Fiq. 26: The different Fiq. 27: Diffusor performance
section types. hypersonic diffusor in the V1G-Mach 20
principles. test section.
iF 1-4 : Dampfstrahlpumpen 1.-4. Siule
S • Ob .. lllichenhaupikondenulor
6 • OberflOchenl",ischenkondonsolor
7 • Absporrvenlile
354 mbar 8 • Drossolvenlile
9 • Kondonsahbscheider mil Schmuhliinger
13 -C
10 • Kondonsalbehiiller loffen'
\4 mbar
A • Vakuumanschlun
B • Enlluflung
C • Troibdamphnochlun
n • Treibdamplreinkondensol
6 E • Verunreinigl .. Kondensol
F • Kuhlmillel 15010. Brunnen-. Shdl-.
-IAmbar Kuhllurm- odor Dberlliichen",ass .. '
i
7k 7
W
=::
(PI\ I loIul.Ju~ul4 I i I
C rtl
9
J-·I~I
8 1010 mbar
I I
To 'a
10 Fig_ 28:
4-stage steam ejector
ET --- with condensators.
10 6 = 0 ; : 1 ; - -'. - .~
t=~N.1-111rn;;i - ~ :: - t$: - - ~:lI~·tfllll~~·"t1HII~fUJ'
m3 ~~ViJtll~ - ~ ~ri' !&~ . . ~ -:-t-'t" ~I SlUl~~
h ~ ~~9f 107~b. I"-'~ , ~~,. ~ max. moss rlow
lOSi~ ~lJnJIII'-: r" [" _' . ' " " ' " , , ~ (450 kg) ~~
Ruvac 87 • ~. ,,; I h ~" .1'-
~ ~- J ~~ .... !~~. ~~. I~r--. ~ .f'. I~P
I~ " . I.~~ ,....., , . ; r,,<; I"~ . '" , ., ~~ " : ~ .:
" ' ,I: , "', I" ", '. I'\: 1':-- . 1'-':" .: t~~ ..;~, ,~
10' ,""" ,', I'.'. ",. ~~ ~ ~ ...
~-- ~'- . . ~"- ~
11'1
7 ~ ~. ~ - ~
"C ' 5.
~
~ ,Buvac 8,~. r~·, t kg
Q) ~ f~ .. , '"\" ".: ' I".: I~, , ~, " , .; x r-; " .-
Q)
a. K'" Iv[,\"!,-' " '- I",' ~" h
11'1 '" '. ~ If~ I " , " , :. . I" tS~"~n~ .1500
w 10 3II
C - r=' - 'j;:E 1200
N I&': ' ~ V .>.>c. ~ ~ 1... ~a -900
.~
u
~ ~~ "''''., '. I~tt:.".:-. ~ , 600
s,
-:J
III
~ :~~~ ~ ~ ~~ 1-': I~~ I~'~ '~n~ ~';' -300
If"; , ' ' 3x 5800 j"t:. 1'\ I~, "\ " ~: as ,:' ~ 1\ ~ :
10 -150
~> ~~ ~'k ~I' ~~- ~ -- ~~_ ~-.
t~b~;,' r~~ ~'~ ~~ ~~; ~", ~',
10 I~I~~' ~'\ ~ ,
. -~ 'f- ,\Rf' "I~ , '\r-,:, 1":,
10- 4 10- 3 10- 2 10-1 10° 10' Torr 10l
Suction pressure p
oY'OOOOo
107
Ruvac: 87
RV'I'CX S7
v> Ruvoc 46
c;
w
\ ~--t-r-Il
~
high frequency
plasma generator
-~+
gate valve to main lest section diffUSion
root s blowers with doors pump
V= 30000 1/5 and windows II = 75000 lIs
m150 bar 1/5 rh 1 O.O~ bar lis
315
INNER ELECTRODE
WATER-COOLED MAGNET
'\ ~ ~,=c
(COPPER) ~ COIL
b
AIR INLET
(SWIRLED)
~~ER
c= _=oJ ELECTRODE
(WATER-COOLED COPPER)
ARC HEATER CONFIGURATION
i[Q :'Fb
TO
,.---,
ct::::J::=:=!It
r--- ,
'- ___ J
t 5-STAGE
STEAM
EJECTOR
ARC TEST
HEATER SECTION
coolin
---------
Neutral Segments
]1'iq. 36:
The IRS coaxial
electrode arc heater.
---~ Anode
317
y
1.0
MWrm
m2 0.5
q VR:RH 0.3
0.1
0.05
0.03
bar
H km
1.0
MW ... r=
m2 Ym 0.5 80
q yrr:.3
0.1
0.05
0.03
2 10 20 :so 40
---- Hermes
Fig. 38 Heat flow simulation capability of DLR arc heater
q versus total enthalpy
318
molecular beam
densi ty n
velocity -
~i
questions:
• how is beam reflected?
-dependence on
n
surface state?
/ molecular state?
surface
-
Fig. 39: Main question in gas-surface interaction.
150 EKIN
SATELLITEN------.....
EROUMLAUF
AI'O LL
RUCKK.
9 [e V)
Z(km)
V. MONO
15
so
6 a U[km/sl 10 12
319
0 1 2
L
"'0> ds
To
no
molecular effusion !
PUMP
1
CLASSICAL MOLECULAR BEAM
1 2 3
CHAMBER
SOURCE
COLLIMATION TEST CHAMBER
CHAMBER
-
CHAMB ER
mol. beam
mol. beam
SKIMMER dc
COLLIMATOR
continuum
f re e jet PUMP
1 1
expansion
Fig. 41
320
MATERIALS DEGRADATION ST.UD1ES
CO 2 Laser Beam
L =ZnSe Lens
(25.4 mm Diameter)
SamPI.
.. I!5cm
lZScm
Slmpl,
• MlnlpUI110t .3
~
toIonipulitor 12
+ MoMcuSar 8.am
"'-'Scm AppalllUI Wall
SlralGht
1010_... 0- SamPIt
Btlm
APPlI"I1UI WI'
ZOO ... Mlnlpulllor"
/J':.
T..... TurtJoZ
22001/1 5OO1.Js
-
Tumo Turbo.
500lJa 1500111 Roc_ItOuodNpOlt
M... FIIt.. WId\~
Pumping 51011" On
+
2. Atomic oxygen exposure facility.
321
eV
MIXTURE
KINETIC ENERGY 5 [
of ARGON 1 HOL % A
99MOL % He
ATOMS 4
3
ABUAF et 01.
2 KIN. ENERGY GAIN
w
~ 100 % A
EXPJ.tNSION
ol7 1000
~ 2000 3000
OK
-To
Goals Status
1 Beam energy - 5 eV 5 ± 0.8 eV
2 Flux - 1015 atoms cm- 2 -s- 1 5 x 10 14 (! 5 eV
0.2-in. diam
3 Fluence - 10 18 atom cm- 2 over > 10 19 (! 5 eV
a h time 8 hr run, 0.2 in.
diam
4 Neutral AO beam Neutral plasma
5 Uniform exposure - 0.5 in. 0.5 inch diam with
reduced flux density
skimmer_ skimmer_
stagnation
chamber
7mm
- 320 mm -~""'--I
t
Fig. 45: The free jet core-molecular beam system as
proposed by H. Legge.
323
HYPERSONIC BOUNDARY-LAYER TRANSITION
KENNETH F. STETSON
Foreword
325
much valuable information that remains relevant to
current hy~ersonic transition problems. Surveys by
Reshotko,2, Arnal,4 and Morkovin and Reshotko 5 are also
recommended reading.
326
grow, their growth rates, their return to a stable
condition, the particular disturbance frequency that will
obtain the maximum disturbance amplitude, and the effect
of various parameters (e.g., Mach number, pressure
gradient, wall temperature, etc.) Stability theory can
provide much valuable information about boundary-layer
disturbances, but it cannot predict transition. This is
an important point. There is no transition theory. All
transition prediction methods are empirical. Transition
prediction methods based upon stability theory (e. g., the
eN method) must relate transition to some empirically
determined condition.
The introduction of linear boundary-layer stability
theory by Tollmien and Schlichting met with strong
opposition. This was primarily because the wind tunnel
experiments of that time could find no evidence of the
instability waves predicted by the theory, and there
seemed to be no connection between linear stability
theory and transition .15 The classic experiments of
Schubauer and Skramstad16 completely changed the
opinions. Wind tunnels in use at that time had high
freest ream turbulence levels that completely obscured the
existence of small boundary-layer disturbances. The low
turbulence wind tunnel of Schubauer and Skramstad
provided the first demonstration of the existence of
instability waves in a laminar boundary layer, their
connection with transition, and the quantitative
description of their behavior by the theory of Tollmien
and Schlichting. These experiments, as well as
subsequent experiments, provided verification that when
the freest ream disturbance amplitudes are small, linear
stability theory adequately described the onset of small
disturbance growth in a subsonic boundary layer and the
growth characteristics of the disturbances through their
major growth history, up close to the transition
location. Subsequently, linear stability theory found
wide applications in the description of instability
parameters and in the prediction of transition for
subsonic flows.
327
turbulence. Other disturbances may be unstable and
experience growth, but they do not grow enough to cause
transition. These events can be conveniently illustrated
by means of a stability diagram such as schematically
shown in Fig. 1 (from Ref. 17). The bottom of this
figure illustrates a standard "thumb curve" stability
diagram, which graphically shows the boundary between
stable and unstable regions in terms of disturbance
frequency and Reynolds number. The solid lines (I and
II) are the neutral boundaries that separate the stable
and unstable regions. If one follows a specific
frequency with increasing distance (increasing Reynolds
number), disturbances at that frequency are initially
stable and experience no growth. As they reach the
Reynolds number that corresponds to the crossing of
neutral branch I they become unstable and start to grow.
The initial disturbance amplitude at the crossing of
neutral branch I (Ao) is an important parameter since it
directly influences the amount of growth required to
obtain the critical breakdown amplitude (Ac ). The
initial disturbance amplitude depends upon the
characteristics of the disturbances to which the boundary
layer is exposed, the receptivity of the boundary to
these disturbances, and the extent of the initial stable
region. As the disturbance waves proceed downstream they
become better "tuned" to the boundary-layer thickness and
they amplify at increasing rates. They reach a point of
optimal tuning (the maximum amplification rate) and then
gradually detune as they approach neutral branch II. The
amplification rate decreases to zero at the Reynolds
number that corresponds to the crossing of neutral branch
II and the disturbances have obtained their maximum
amplitude. Plots of amplitude vs Reynolds Number (such
as shown in the top portion of Fig. 1) are inflected
curves with a zero slope at branches I and II. In the
example illustrated in Fig. 1, disturbances of frequency
Fl and F2 obtain their maximum growth and then attenuate
before boundary-layer transition occurs. These
disturbances are presumed to have no influence on
transition. Note that the onset of disturbance growth
for the F2 disturbances (the crossing of neutral branch
I) occurs at a larger Reynolds number than the Fl
disturbances; however, the F2 dist~rbances have a longer
period of growth and obtain a larger amplitude. If
boundary-layer transition occurs at R.r, then the F3
disturbances are the dominant disturbances since they are
the first disturbances to grow to the amplitude required
for breakdown. These disturbances presumably cause
328
transition. F4 disturbances have the potential of
obtaining even larger amplitudes, but they do not get the
opportunity since the boundary layer becomes transitional
first. It is generally assumed that the growth rate of
the disturbances is not influenced by changes in the
freest ream turbulence levels (as long as the turbulence
levels are not large enough to force boundary-layer
disturbance growth by some mechanism other than boundary-
layer instability). Therefore, the effect of the
freest ream turbulence levels is felt through its
influence of Ao' Increasing Ao for all frequencies would
shift all of the growth curves upward, such that some
higher frequency, such as F 2 , would first obtain the
critical amplitude. Reducing Ao (as in a quiet tunnel)
would lower the curves and some lower frequency, such as
F 4 , would then be the first disturbance to obtain the
critical amplitude.
329
layers and, physically, they represent new instabilities
that can influence hypersonic transition. Of the many
contributions that Mack has made toward the understanding
of hypersonic boundary-layer stability, the discovery of
the higher modes is probably the most significant (the
higher modes are sometimes called "Mack modes" to honor
the importance of Mack's contribution). Thus, subsonic
and low supersonic boundary layers contain relatively low
frequency, vorticity disturbances called first mode
disturbances (Tollmien-Schlichting waves) and hypersonic
boundary layers contain both first mode and Mack mode
disturbances.
330
results illustrate the characteristics previously
described.
331
the stability calculations. Therefore, it is not just a
matter of having a valid theory, but also one of having
valid mean boundary-layer profiles to provide data to
input to the stability eguations. The success of linear
stability theory for subsonic and low supersonic boundary
layers does not guarantee its success with a hypersonic
boundary layer. There are so many different features of
hypersonic boundary-layer stability that an independent
verification is required. Section l.d will discuss the
first attempt at verification of hypersonic linear
stability theory.
332
layer thickness. Second mode disturbances are not
related to a specific frequency range, but can occur
anywhere from relatively low frequencies (for "thick"
boundary layers) to very high frequencies (for "thin"
boundary layers). Situations that correspond to a change
in boundary-layer thickness change the frequency of the
second mode disturbances. For example, going to higher
altitudes thickens the boundary layer and lowers the
second mode disturbance frequencies. The normal growth
of the boundary layer along a vehicle surface results in
a steady lowering of the most amplified disturbance
frequencies. Second mode disturbances grow much faster
than first mode disturbances and rapidly become the
dominant disturbances. It can also be observed in Fig.
3 that disturbance growth is occurring at frequencies
higher than the ridge of second mode disturbances. These
disturbances are believed to be a harmonic of the second
mode and are not explained by a linear theory. All of
the previously mentioned hypersonic stability experiments
have observed the high frequency nonlinear disturbances.
Even though the boundary-layer disturbances had grown to
a relatively large amplitude by the end of the model,
the boundary layer still had the mean flow
characteristics of a laminar boundary layer.
333
thickness), and disturbances extended well beyond the
defined boundary-layer edge.
334
transition Reynolds number than if the second mode
disturbances were present.
335
the figure that the disturbance amplitudes are getting
smaller in this region of the cone frustum. This stable
region extended to an S/RN of approximately 121. This
corresponded to a location on the cone frustum where most
of the entropy layer generated by the nosetip had been
swallowed by the boundary layer. Thus, for this case,
the region of the cone frustum where the entropy layer
was being swallowed by the boundary layer was a stable
region. The sharp cone, at corresponding local Reynolds
numbers, showed a steady growth of disturbances. In
fact, at a local length Reynolds number of 5.1 x 10 6 , the
boundary-layer disturbances on the sharp cone had growth
to sufficient amplitudes to initiate second mode wave
breakdown (presumably, an early stage of transition) .
336
present experiments. The disturbances amplified at a
nearly constant rate and transition occurred at Reynolds
numbers several times the value of the expected critical
Reynolds number. With small nosetip bluntness the
critical Reynolds numbers were extremely large and the
disturbances amplified rapidly once the critical Reynolds
number was exceeded. Transition information was not
obtained; however, it would be expected that transition
Reynolds numbers would not be a great deal larger than
the critical Reynolds numbers.
337
swallowed, these disturbances entered the boundary layer
and experienced rapid growth. The source of the inviscid
disturbances and why they were unstable within the
boundary layer at Reynolds numbers which were stable for
small bluntness, is unknown. A possible explanation for
the boundary-layer disturbance growth is that the
situation is analogous to the forcing concept described
by Kendal1 8 and Mack. 6 Kendall found that when the
boundary layer was subjected to a strong external
disturbance environment, disturbances were found to grow
before the predicted location of instability. It may be
that, as the nosetip radius is increased, the entropy
layer disturbances experience more growth, until they
become large enough to drive the boundary-layer
disturbances. Additional details of these stability
experiments can be found in Ref. 11.
Hot-wire boundary-layer stability data were also
obtained on the sharp cone at angle-of-attack. Data were
obtained on the windward meridian at 2 and 4 degrees
angle-of-attack and on the leeward meridian at 2 degrees
angle-of-attack. Fig. 10 compares these data with the
zero angle-of-attack data of Ref. 10. It was found that
the growth rates of the boundary-layer disturbance were
not greatly affected by angle-of-attack; however, the
onset of disturbance growth was significantly affected.
The onset of disturbance growth was delayed on the
windward meridian and occurred earlier on the leeward
meridian, as compared with the zero angle-of-attack data.
These stability trends are compatible with the observed
movement of transition location with angle-of-attack.
Details of these stability results can be found in Ref.
12.
The theory of Mack 6,7 indicated that second mode
disturbances would be destabilized by lowering the
surface temperature. The hypersonic boundary-layer
stability experiments of Demetriades 9 confirmed that
cooling the surface increased the growth rates of second
mode disturbances and that the transition Reynolds number
was reduced by a corresponding amount.
338
vs. Reynolds number are shown. Al is the disturbance
amplitude at the first measuring station. It can be seen
that disturbances in the boundary layer on the cold wall
grew much faster than those in the boundary layer of the
hot wall. The initial amplitudes (A l ) are most likely
different for the two cases shown. Therefore the
significance of the difference between amplitude ratios
at transition is not known.
Fig. 12 compares maximum growth rates for second
mode disturbances for a cooled and uncooled cone (from
Ref. 14). As observed in the previous figure, the second
mode disturbances grew much faster in a cold wall
boundary layer. Thus, these stability experiments
confirm the prediction of stability theory that cooling
the wall is destabilizing for second mode disturbances.
The increased growth rates of the second mode
disturbances for the cooler wall condition would be
expected to result in a reduction in the transition
Reynolds number. Boundary-layer transition data obtained
by Demetriades 9 and Stetson et al lO - 13 indicated that
transition Reynolds numbers were changing in a
corresponding manner.
(l.d) COMPARISON OF THEORY AND EXPERIMENT
The early experiments of Kendall a verified the
existence and dominance of second mode disturbances in a
hypersonic boundary layer. A more extensive comparison
between theory and experiment was not made until Mack l9 ,20
obtained numerical solutions for the conditions of the
experiments of Ref. 10. The comparison was for a sharp
cone at zero angle-of-attack in a perfect gas, Mach 8
flow (cone half angle 7 deg, Re./ FT = 1 x 10 6 ,
Me = 6.8, To = 1310·R, T..," 1100"R) . The results of this
comparison pointed out some discrepancies which presently
cannot be explained and need to be resolved.
339
dimensional, second mode disturbances. Since the second
mode disturbances are the major disturbances that
presumably initiate transition, they become the most
important comparison. The most unstable frequency (the
peak) is in close agreement; however, the maximum growth
rate and the location of the upper neutral branch (where
the amplification rate goes to zero) are significantly
different.
340
PART 2: COMMENTS ON PARAMETRIC TRENDS
(2.a) INTRODUCTION
341
in wind tunnels. The decrease in transition Reynolds
number with Mach number in the supersonic range is most
likely the result of the disturbances in the freest ream
of the wind tunnels. Flight experiments on a 5-deg half
angle cone supported this contention by demonstrating
that transition Reynolds number increased with Mach
number up to M = 2 (the maximum Mach number of the
experiment). Fig. 16 shows some of the flight data and
compares flight transition data with wind tunnel
transition data. All data were obtained with the same
model and same instrumentation (Fig. 16 is from Ref. 30).
Wind tunnel results at hypersonic Mach numbers have
consistently shown a large increase in transition
Reynolds number with increasing Mach number.
Unfortunately it has not been possible to separate out
the wind tunnel effects and the Mach number effects.
Most experimenters have speculated that the Mach number
effect in the hypersonic regime is one of increasing
transition Reynolds number with increasing Mach number.
This conclusion is further supported by theory. The
stability theory of Mack 6 ,7 has shown that, at hypersonic
Mach numbers, the maximum amplification rates decrease as
the Mach number increases. A decrease in the maximum
amplification rate would be expected to result in larger
transition Reynolds numbers. The Mach number effect may
not be as pronounced in flight transition data as in wind
tunnel transition data since in a wind tunnel the
environment effect varies with the Mach number. Fig. 17
(from Ref. 31) includes additional data to illustrate
Mach number effects on transition and includes both wind
tunnel and flight results. The flight data has
variations due to nosetip bluntness, angle-of-attack,
wall temperature differences; and, at the higher Mach
numbers, ablation and high temperature flow field
effects. So many effects are simultaneously influencing
flight transition data that comparisons with wind tunnel
data can be misleading.
Available data suggests that high transition
Reynolds numbers are to be expected on cones with small
nosetip bluntness and small angles-of-attack when the
local Mach number is about 10 or above. There is
uncertainty as to the magnitude or the functional
relationship between transition Reynolds number and Mach
number. The correlation, Reo/Me = constant, requires a
judgment as to this functional relationship. This topic
will be discussed in more detail under Part 4.
342
(2.c) EFFECT OF NOSETIP BLUNTNESS
343
agreement. The two major effects associated with the
entropy layer are changes in the transition Reynolds
number and reductions in the local Reynolds number. The
reduction of the local Reynolds number is an extremely
important piece of information in the interpretation of
nosetip bluntness effects on frustum transition; however,
this is not the major issue since this information is
readily obtainable, with uncertainties being related only
to the accuracy and limitations of the flow field program
being utilized. The major problem area is associated
with understanding how nosetip bluntness affects the
transition Reynolds number. Limitations in the Reynolds
number capability of wind tunnels have limited wind
tunnel results to Mach numbers less than 10. These
results are useful to illustrate trends; however, the
effects of higher Mach numbers and the magnitude of
transition Reynolds numbers expected in free flight are
not well known. Fig. 21 (from Ref. 25) contains the
results form a large amount of nosetip bluntness data
obtained in a Mach 6 wind tunnel. The movement of
transition location is shown, along with changes in
transition Reynolds number and the Reynolds number
reduction that contributed to the changes in transition
location. Note that when the entropy layer was nearly
swallowed at the transition location (XT/X sw close to 1),
the transition Reynolds numbers were significantly larger
than sharp cone transition Reynolds numbers and the
Reynolds number reduction was small. The change in
transition location in this region was primarily a
function of the change in transition Reynolds number.
The maximum change in transition location occurred in
regions of the entropy layer where the transition
Reynolds numbers were less than the sharp cone values and
the Reynolds number reduction was the major effect. For
maximum transition displacement, the local Reynolds
number was reduced by a factor of 7.3 and the transition
Reynolds number was 58% of the sharp cone value, with the
displacement being represented by the product of the two
effects, or 4.2 times the sharp cone transition location.
The Reentry F flight experiment 35 ,36 is probably the
best source of data for the effect of nosetip bluntness
on slender cone transition in hypersonic free flight.
The lack of information regarding the nosetip changes
during reentry as a result of ablation, along with small
angles of attack, produce some uncertainties in the
interpretation of the results.
344
There is another nosetip consideration that should
be included - the very low transition Reynolds numbers
associated with transition on the nosetip and the region
of the frustum just downstream of the nosetip. Nosetip
transition Reynolds numbers can be as much as two orders
of magnitude less than cone frustum transition Reynolds
numbers. This situation requires that a separate
transition criterion be applied to this portion of a
configuration. The potential of transition first
occurring in this region, and producing a turbulent
boundary layer over the entire portion of the
configuration influenced by the tip, must be considered.
It is well documented that blunt nosetips have low
transition Reynolds numbers, even at hypersonic
freest ream Mach numbers (e.g., Refs. 37-39). Boundary-
layer transition has been related to the local boundary
properties at the sonic point and the surface roughness.
The low transition Reynolds numbers associated with the
region of the frustum just downstream of the nosetip has
only recently been identified25 and the transition
criterion for this region is not as well understood as
that of the nosetip. It appears that transition in this
region is dominated by the nosetip and may be related to
nosetip conditions, analogous to nosetip transition
criteria. Fig. 22 (from Ref. 25) provides an example of
transition criteria for transition on the nosetip and
also those conditions which produced early frustum
transition for Mach 5.9 wind tunnel experiments.
345
value, transition occurred regardless of the two-
dimensional Reynolds number. For example, Owen and
Randall's43 subsonic experiments with a swept wing
observed an instantaneous jump of transition from the
trailing edge to near the leading edge when a critical
crossflow Reynolds number was exceeded. This critical
crossflow Reynolds number was approximately 175, based
upon the maximum crossflow velocity, a thickness defined
as nine-tenths of the boundary-layer thickness, and the
density and viscosity at the edge of the boundary layer
(ReeF = Pe wmax .9&) . Pate' s44 results indicated that
Ile
this criterion could be extended to supersonic Mach
numbers. However, high values of critical crossflow
Reynolds number have been obtained for incompressible
flows and there is uncertainty as to the generality to
this criterion. The appropriate value for hypersonic
flows in unknown and must be estimated on the basis of
lower speed transition experiments.
346
are not. Fig. 23 (from Ref. 45) illustrates the
transition movement on the windward and leeward rays of
sharp and blunt 8-deg. half angle cones at ~ = 5.9. The
transition distance (X T ) is normalized by the transition
distance on the sharp cone at ex = 0 deg [( (X TS ) a,=o)
varies with unit Reynolds number]. Fig. 24 (from Ref.
45) is a sample of the transition patterns obtained for
a sharp cone. ~ = 0 deg is the windward meridian and~
= 180 deg is the leeward meridian. The shaded area
represents the transition region, with curve B indicating
the beginning of transition and curve E the end of
transition. The beginning and end of transition at
ex = 0 deg is shown for reference. Fig. 25 (from Ref.
45) presents a summary of the sharp cone angle-of-attack
results, in a nondimensional format. Figures 26 and 27
(from Ref. 45) present similar results for a cone with
10% nosetip bluntness (Rn = 0.2 in).
347
disturbances provide the stimulus for exciting boundary-
layer disturbance growth and are responsible for the
initial boundary-layer disturbance amplitudes. If, by
some mechanism, the initial amplitude of the most
unstable boundary-layer disturbances could be increased
or decreased, the transition Reynolds number would
correspondingly be increased or decreased (this will be
discussed under the next topic, environmental effects).
The unit Reynolds number, in effect, provides a possible
mechanism. The frequencies of the most unstable
boundary-layer disturbances are directly related to the
unit Reynolds number (by the effect of unit Reynolds
number on boundary-layer thickness, as discussed in Part
1). Thus, increasing the unit Reynolds number increases
the frequency of the most unstable boundary-layer
disturbances, which means that the most important
environmental disturbances will, very likely, have a
smaller amplitude and, in some situations, a suitable
environmental stimulus may be lacking for some
frequencies. Intui ti vely, it would be expected that unit
Reynolds number, through its control of the frequency of
the most unstable boundary-layer disturbances, would
influence transition. Morkovin has commented many times
that unit Reynolds number probably influences transition
in several ways, thus other unit Reynolds number effects
should be considered likely.
348
that transition occurs when some boundary-layer
disturbances have obtained the critical amplitude
required for breakdown of the laminar flow, a change in
the initial amplitude of the dominant disturbances
changes the required period of growth to obtain the
critical amplitude. Thus, a change of the environment
will most likely change the transition Reynolds number.
This critical element of the transition problems is often
overlooked. When one or several sets of data are used to
make a transition prediction if a new situation, a
similarity is implied for not only the geometric and flow
parameters, but also the environment. It is assumed that
the case in question has the same environment as the data
base. Environmental differences provide a reasonable
explanation for most of the differences in transition
Reynolds numbers obtained in wind tunnels and those
obtained in flight. In supersonic and hypersonic wind
tunnels the strong acoustical disturbances in the
freest ream that are generated by the turbulent boundary
layer on the wall of the nozzle generally produce
transition Reynolds numbers lower than found in flight.
Differences in wind tunnel environments can result in
significant differences among wind tunnel transition
Reynolds numbers, thus presenting problems in correlating
only wind tunnel transition data. The data of Schubauer
and Skramstad 49 and Wells 50 provide an interesting
example. The classical experiments of Schubauer and
Skramstad were carried out on a sharp, flat plate in a
low turbulence, low speed wind tunnel. Turbulence levels
in the freestream could be controlled by varying the
number of damping screens. Transition Reynolds numbers
were found to be directly related to the freest ream
turbulence level, with transition Reynolds number
increasing as the turbulence level decreased. At low
tunnel turbulence levels, the transition Reynolds number
obtain a maximum value of 2.8 x 10 6 and remained at this
level with still further reductions in turbulence levels.
Wells repeated this experiment in a different wind
tunnel. In the Schubauer and Skramstad experiment,
control over the damping screens provided control over
the velocity fluctuations in the freest ream of their wind
tunnel but the screens had little effect on the
acoustical disturbances which were present. In the Wells
experiment, the tunnel was designed so as the minimize
the acoustical disturbances as well as to provide control
over the velocity fluctuations. Wells found the same
trends as obtained by Schubauer and Skramstad, but his
maximum transition Reynolds number was approximately 5 x
349
10 6 . Both experiments were dealing with the same
boundary layer phenomena. What was different was the
environment. Fig. 29 (from Ref. 50) contains these
results. Wells indicated that most of the freest ream
energy in his experiment occurred at frequencies below
150 cps with acoustic content less than 10% of the total
energy. The tests of Schubauer and Skramstad involved
significant energy levels out to 400 cps, and, in
addition, the spectrum exhibited large acoustic energy
peaks at 60 and 95 cps which accounted for approximately
90% of the total disturbance energy that was measured for
intensities less than about 0.05%. Spangler and Wells 51
continued the study by systematically investigating the
effects of acoustic noise fields of discrete frequencies.
Large effects were found when the acoustic frequencies
(or a strong harmonic) fell in the range where Tollmien-
Schlichting waves were unstable. It is significant to
note that transition prediction methods cannot account
for these large differences in transition Reynolds number
unless the differences in the freest ream environment are
somehow taken into account.
350
critical amplitude that produced nonlinear effects and
the eventual breakdown of the laminar flow. If
transition must wait for disturbances with a smaller
growth rate to obtain the critical amplitude, then a
delay in transition would be expected. There are many
hypersonic flow situations, both in ground test
facilities and in flight, where the potentially most
unstable boundary-layer disturbances may not be excited.
Thus, some transition delay, due to a lack of
environmental stimulus of the potentially most unstable
disturbances, may be a common hypersonic occurrence.
Stetson 12 has pointed out that for a sharp, 7-deg half
angle cone in a Mach number 8 wind tunnel at a freest ream
unit Reynolds number of 20 million, the most unstable
boundary-layer disturbances would have frequencies
greater than a megahertz. Available instrumentation
cannot measure disturbances in this frequency range;
however, it seems unlikely that there would be much
freestream disturbance energy at such high frequencies to
stimulate boundary-layer disturbance growth. Transition
under this situation would be expected to be the result
of disturbances which were not the theoretically most
unstable. This should provide larger transition Reynolds
numbers. The Reentry F flight experiment 35 reported
transition Reynolds numbers as high as 60 million. As
estimation of the frequency of the most unstable
boundary-layer disturbances indicated they were greater
than 500 kHz. There is a possibility that these high
transition Reynolds numbers were obtained because the
theoretically most unstable disturbances were not
present.
351
we would still have to somehow prescribe the external
disturbances. The freest ream disturbances are a very
important initial condition of any boundary-layer
transition problem and, unfortunately, they are generally
not well known. The uncertainty of the disturbance
environment in flight puts an additional uncertainty into
any transition prediction.
(2.g) EFFECT OF WALL TEMPERATURE
352
at hypersonic Mach numbers. The results of Sanator et
al 59 (not shown in Fig. 31 because the value of (ReXT)AD
was not known) at Me = 8.8 found no significant change of
transition location on a sharp cone with changes of Tw/To
from 0.08 to 0.4. Some additional data (not shown in
Fig. 31) of Stetson and Rushton 33 at ~ = 5.5 and Mateer 60
at ~ = 7.4 report a reduction in transition Reynolds
number with a reduction in the temperature ratio.
353
(2.h) EFFECT OF SURFACE ROOGHNESS
354
thickness. Any effect which influences boundary layer
thickness can affect the influence of roughness.
Therefore, body location, unit Reynolds number, wall
temperature, Mach number, and mass addition or removal
can all influence the effect of roughness. Wind tunnel
experiments have shown there is a strong effect of Mach
number on roughness effects. The roughness size required
to trip the boundary layer increases rapidly with
increasing Mach number and even at low hypersonic Mach
numbers the roughness heights reguired are of the same
order as the boundary layer thickness (e.g., see Ref.
64) . Part of the problem in trying to understand
roughness effects is associated with the many roughness
parameters involved. Roughness is usually characterized
by its heights, but other parameters, such as,
configuration and spacing are very important. Also
important are whether the roughness elements are two-
dimensional or three-dimensional, individual elements or
distributed (e.g., sand grain) type. The nosetip of a
hypersonic vehicle, where the Mach number is subsonic and
the boundary layer is very thin, can be very sensitive to
roughness. The frustum of a hypersonic vehicle, where
the local Mach number is hypersonic and the boundary
layer is relatively thick, is expected to be insensitive
to small or moderate roughness.
355
complicates the understanding of transition associated
with pressure gradients. It is then necessary to
consider the two competing effects on transition -- the
effect of the first and second mode disturbances and the
effect of the Gertler vortices. When there exists a
concave curvature of the streamlines (not necessarily a
concave surface) the associated centrifugal forces
result in the formation of pairs of counterrotating
vortices called Gertler vortices, the axes of which are
parallel to the principal flow direction (see Fig. 32).
The growth of Gertler vortices can be calculated from a
linear stability theory (e.g., see the papers of Flor~an
and Saric,65 El-Hady and Verma, 66 and Spall and Malik ).
Experimentally, surface visualization techniques, such as
oil flow, are believed to show the existence of Gertler
vortices. Also, Ginoux 68 noted that the vortices produce
large peaks in the heat transfer rate in the lateral
direction. An interesting case has been found in the
study of transition on wind tunnel nozzles. Transition
was found to occur on Mach 3.5 and 5 nozzle walls earlier
than expected. 69, 70 Oil flow studies showed streaks that
were believed to result from Gertler vortices. Stability
calculations,71 for the M = 3.5 nozzle, indicated that
the strong favorable pressure gradient damped the first
mode disturbances and the Gertler vortices were the major
disturbances.
356
with other effects; for example, its effect on roughness
and surface cooling.
357
most unstable boundary-layer frequencies, that they would
be of little consequence.
PART 3: SOME ADDITIONAL GENERAL COMMENTS
358
Crossflow: An inflectional instability of the
crossflow velocity profile. Little is known about the
characteristics of these disturbances. Experimental
transition data imply these disturbances can have rapid
growth rates and they may be the dominant disturbances in
three-dimensional flow fields.
359
edge to the predicted location of instability; i.e., in
a region where linear stability theory indicated the
boundary layer should be stable for all disturbance
frequencies. This early growth of disturbances was
attributed to the strong sound field generated by the
turbulent boundary layer on the nozzle wall.
360
might result if by-pass transition occurs.
361
plate transition Reynolds numbers were found to vary from
about 0.8 for low-noise freest ream conditions to about
1.2 for higher noise conditions. These new quiet-tunnel
experimental results support the implications of the
analytical results obtained using linear stability theory
and indicate that the transition data of Fig. 33 was not
a general result, but was dominated by wind tunnel
freest ream noise.
362
appears evident that the results of Fig. 33 should not be
used for flight applications.
363
For most situations the transition Reynolds numbers
obtained in wind tunnels are lower than corresponding
flight transition Reynolds numbers. It should be
remembered that the differences between wind tunnel and
flight transition Reynolds numbers are not the same
throughout the Mach number range. The largest
differences are generally at supersonic Mach numbers and
the smallest differences are at subsonic and large
hypersonic Mach numbers. Figs. 16 and 17 illustrate
these differences. Also the specific configuration is a
factor. In some cases, a transition parameter may be
dominant enough to overshadow the difference in the
freest ream environment (e.g., bluntness or surface
roughness). The wind tunnel transition Reynolds numbers
obtained on the shuttle configuration were not much less
than found in flight.
364
flow should relate to other situations.
365
detecting transition onset. The location of transition
onset has been found to vary depending upon the method of
detection whereas the end of transition was essentially
independent of the method used. For example, transition
onset detected optically is consistently further
downstream than onset detected by heat transfer rate or
surface total pressure. These findings prompted Pate to
make his correlations based upon the end of transition,
rather than onset. Harvey and Bobbitt 76 have reported
that in low noise wind tunnels and flight the transition
region can be much shorter than the laminar region, with
(Re XT ) B / (ReXT) E varying from about 0.5 to 0.9. Most
flight experiments have added uncertainties due to the
inability to control the flow conditions and vehicle
altitude, coupled with more restrictions on vehicle
instrumentation. An exception was the carefully
controlled flight experiments of Dougherty and Fisher.3~
A 5-deg. half angle cone, which has been extensively
tested in transonic and supersonic wind tunnels, was
mounted on the nose boom of an F-15 aircraft and flight
tested. The same instrumentation, primarily a surface
Pitot probe, detected transition both in flight and in
the wind tunnels. The flight experiments, up to a Mach
number of 2.0, measured a very short transition region,
with (ReXT)B /(ReXT)E being between 0.8 and 0.9. Mach 6
wind tunnel experiments 45 (see Figures 24 and 26), on a
8-deg half angle cone with both a sharp tip and small
nosetip bluntness, found XTB/X TE to be approximately
0.75. With larger nosetip bluntness, which produced
early frustum transition, there was typically a very long
transition region. Usually the transition region
extended to the end of the model so that the end of
transition could not be measured, with the transition
length being several times as long as the laminar length.
The Reentry F flight test data showed large variations in
the length of the transition region. At 84,000 feet,
(Re XT ) B / (Re XT ) E = 0.64 and at 60,000 feet, the value
reduced to 0.19. These results very likely reflect the
coupling of several effects and are difficult to
interpret.
366
(3.f) SOME PROBLEMS WITH FLOW-FIELD CALCULATIONS
367
The PNS calculations were repeated, doubling the grid
points to 120, and the PNS profile became much closer to
the boundary-layer code profile, but there were still
differences in the outer part of the boundary layer. A
third PNS calculation was made, still with 120 grid
points, but increasing the density of grid points in the
boundary layer. These results (not shown) were only
slightly different, moving the profile in the direction
of closer agreement with the boundary-layer code results
and the experimental data. There seem to be several
messages from the information on this figure which
warrant further investigation to see if they represent an
isolated or the general case.
(4.a) INTRODUCTION
368
possible to predict the outcome. Even if one were
successful in identifying the major parameters, it would
not be possible to account for their individual effects
in a transition correlation technique. Usually an
empirical correlation is based upon a dominant parameter
and the others are neglected. Many effects become hidden
in the empirical relationship. As long as the transition
correlation is being applied to a configuration and flow
condition similar to those of the data base used to
establish the correlation, the hidden effect may not be
greatly dissimilar. A problem exists, however, when one
wants to apply a transition correlation to a
configuration or flow condition unlike those of the data
base. A change in the outcome of the competition of the
various factors, or a change in the contribution of the
various hidden effects, can greatly reduce the accuracy
of the transition prediction.
369
to establish a meaningful empirical relationship. Fig.
35 (from Ref. 79) illustrates the problem. It becomes
essential to be selective in the data used and to include
only those data that most nearly correspond to the
problem in question. The decision of what data to use in
the establishment of a empirical relationship and the
transition criteria is always a difficult choice since it
can have a large effect on the resulting transition
predictions. Such a procedure then limits the generality
of the prediction method. The trend seems to be that
improvements to the prediction method are made only at
the expense of greater limitations on the application of
the method. It is clear that one should always know what
data were used to establish the transition prediction
method being considered.
When it becomes necessary to predict transition on
a new configuration or at new flow conditions empirical
prediction methods have problems. The data base can only
be used as a guide and any transition prediction for such
a situation will have a large uncertainty associated with
it.
370
reasonable representation of the data. For a sharp cone
and Me > 8, a slope of about 100 seems to be reasonable.
Note that for subsonic Mach numbers the constant can
exceed 1000. Therefore, for Mach numbers up to 8, the
constant is varying by a factor of 10. When
consideration is given to entropy layer effects generated
by a nosetip, there is no region where a constant slope
has any credibility. The best that can be done is to use
some average slope. The fact that Space Shuttle flight
transition data gave a slope in the range of 200-400 atMs -2
is of no value in predicting transition on a hypersonic
vehicle with large local Mach numbers.
100 36.9 x 10 6
200 148 x 10 6
300 332 x 10 6
400 590 x 10 6
Considering that the Reentry F flight data indicated a
sharp cone transition Reynolds number of approximately 40
x 10 6 , which corresponds to an Rea/Ms just over 100,
there seems to be no rationale for using large values ofRea/Ms
for this case.
371
(~·20) cases and selected those which met the following
criteria:
372
effects are only partially included, but as long as only
slender reentry vehicles with small nosetip bluntness are
considered, bluntness effects are nearly similar. That
is, using Rotta's34 similarity approach for highly cooled
sphere-cones, the boundary layer properties within the
entropy layer resulting from the nosetip are a function
S/RN
of where the constant K is primarily a
K(Reoo / FT, R N ) 173
function of cone angle and Mach number and can be
obtained from Fig. 20. Thus, for situations where
K(Re.1 FT, RN ) 1/3 does not vary significantly, S/RN, by
itself, adequately accounts for the variation of boundary
layer properties within the entropy layer. Note, also
that it is the product of these terms that is important,
not their individual values. Thus, if the freestream
unit Reynolds number is decreased an order of magnitude
(increasing altitude by approximately 50K feet) and the
nosetip radius in increased an order of magnitude, the
entropy layer, in terms of S/RN is unchanged.
(4 . d) eN METHOD
373
Van Ingen 82 . Linear stability theory is utilized to
calculate amplitude ratios. Transition is presumed to
occur with the earliest attainment of some preassigned
amplitude ratio, usually expressed as eN. The solutions
of the linear stability equations yield the disturbance
growth rate (-u i ) which can be integrated to compute the
exponent N: N = In (A/Ao) = f:o -uids. So is the
location of the onset of instability (at Re c ) and Ao is
the disturbance amplitude at Re c . This method is often
criticized as having no theoretical justification for
predicting transition since all it does is compute an
amplitude ratio (A/~). It ignores the environment (Ao)
and the actual transition process. The value of N must
be input, based upon available experimental data, and
transition is predicted to occur when N reaches the
preassigned value. In spite of such criticisms, it is
presently the best analytical transition prediction
method in general use and Bushnell and his associates at
NASA/Langley Research Center have reported rather
remarkable results for subsonic, supersonic, and low
hypersonic situations. Their results account for first
mode, second mode, Gortler, and cross-flow disturbances
and have been applied to cones, flat plates, airfoils,
bodies of revolution, swept wings, swept cylinders, a
rotating disk, and a wind tunnel nozzle wall. Ref. 83
contains a list of the references describing these
results. Some recent results of Malik 22 contain some
hypersonic results. His computations for sharp cones,
using a N-factor of 10, showed that first mode
disturbances were responsible for transition at adiabatic
wall conditions for freestream Mach numbers up to 7. For
cold walls, second mode disturbances dominated the
transition process at lower hypersonic Mach numbers due
to the destabilizing effect of cooling on the second
mode. Malik's results also show that a favorable
pressure gradient and suction are stabilizing for second
mode disturbances.
Verification of the eN method for hypersonic, three-
dimensional, high temperature flows with entropy layers
will be an extremely difficult task. Of course,
verification of other aspects of such flows will face
similar difficulties. Obtaining valid mean profiles to
input the stability calculations and obtaining reliable
flight transition data to determine the proper N-factors
are seen as particularly difficult tasks.
374
PART 5: COMMENTS ON PREDICTION METHODOLOGY
(S.a) NOSETIP
375
(5.b) EARLY FROSTUM
(5.c) CROSSFLOWS
There is little guidance available for estimating
the effects of crossflow on hypersonic transition.
Experimental data are available for the leeward side of
cones at angle-of-attack (samples are shown in Figures
23-27) and indicate low transition Reynolds numbers in
this region. If the cone configuration is relevant to
376
the problem at hand, transition estimates may be based
upon the cone data. A more general method would be to
base cross flow influenced transition upon a crossflow
Reynolds number. The laminar boundary-layer profile in
a three-dimensional, viscous flow has a twisted profile
that can be resolved into tangential (u) and crossflow
(w) velocity components. The crossflow component of the
velocity is used for the computation of cross flow
Reynolds number. Owen and Randall 43 performed subsonic
experiments with swept wings and found there was a
critical crossflow Reynolds number that caused transition
to make an instantaneous jump from the trailing edge to
near the leading edge. Using a length dimension of nine-
tenths of the boundary-layer thickness, the critical
crossflow Reynolds number was 150 to 175. Higher
subsonic laminar crossflow Reynolds numbers have been
reported, so the generality of the Owen and Randall data
is not known. Pate 44 , 74 has indicated that the Owen and
Randall results appear to be valid for supersonic flows.
If the boundary-layer thickness is used as the length
dimension, a value of 200 seems like a reasonable
conservative guess for hypersonic flows. The procedure
would be to make calculations of crossflow Reynolds
numbers and see if any condition resulted in a number
which exceeded 200. For those conditions where the
crossflow Reynolds number exceeded 200, it could be
expected that crossflow instabilities would dominate and
cause transition. The crossflow Reynolds number is
defined as:
P.. wmax~
Re CF =
11 ..
377
boundary layer can be laminar, transitional, or
turbulent, depending upon the values of the pertinent
parameters. However, the boundary layer on the
attachment line of an infinite swept cylinder is unique
in that it is invariant with position on the cylinder.
Thus, in the absence of any parameter variations, the
state of the attachment line boundary layer (e.g.,
laminar or transitional) is invariant with position on
the cylinder. (A swept wing with a constant leading edge
radius can be considered analogous to a cylinder.)
378
cylinder in terms of momentum thickness Reynolds number,
boundary-layer edge Mach number, and wall-to-recovery
temperature. For Reynolds numbers less than the critical
value, turbulence contamination in the attachment line
boundary layer is damped and the boundary layer remains
laminar. Remember that since the attachment line
boundary layer on a cylinder is not growing, it remains
laminar regardless of the length of the cylinder. When
the critical Reynolds number is exceeded, the
disturbances grow and cause transition on the attachment
line.
In the check for dominant mechanisms, first make a
judgement as to whether or not the leading edge boundary
layer will be contaminated with turbulence from an
adjoining surface. If the leading edge boundary layer is
contaminated, calculate the attachment line momentum
thickness Reynolds number to see if it is greater than or
less than the critical value given in Fig. 39. I f the
Reynolds number exceeds the critical value, transition
can be expected on the leading edge.
379
(such as Fig. 17) or by a correlation technique such as
discussed in Part 4. Remember that flight data such as
contained in Fig. 17 already contain effects such as
small nosetip bluntness, small angles-of-attack, and some
surface temperature variations. Some of the flow and
geometric parameters which influence the instabilities
mentioned in Sa through 5e also influence the growth of
second mode disturbances (e. g., nosetip bluntness and
surface curvature). Therefore, it is necessary to
consider how second mode disturbance growth is modified
by parametric effects. The parametric trends such as
discussed in Part 2 can be used as a guide. Most of the
parametric trends come from wind tunnel data. The
influence of the wind tunnel noise may present some
uncertainty in the trends; however, they are generally
thought to be correct. A possible exception is the
effects of unit Reynolds number. The wind tunnel
freestream environment can produce a unit Reynolds number
effect not expected in flight. However, it is speculated
that unit Reynolds number effects boundary-layer
transition in several ways, therefore a unit Reynolds
number effect in flight should not be ruled out. Until
the situation is clarified, it is suggested that a unit
Reynolds number effect be applied to flight data, if such
an adjustment would be a more conservative estimate.
380
there are presently no other data for such comparisons.
Until further information becomes available, it is
suggested that small nosetip bluntness be assumed to
increase the transition Reynolds number by a factor less
than two, as shown in Fig. 40. An order of magnitude
increase, as predicted by Ericsson 91 , would suggest that
optimum nosetip bluntness for the Reentry F vehicle
should produce transition Reynolds numbers of about 400
x 10 6, an unrealistic prediction.
381
which dominate the transition process. For the
particular case in point, compare all the possible
disturbance mechanisms and make a judgement as to which
one will dominate. Having decided upon the dominant
disturbance mechanism, use what you consider to be the
best available transition method and criterion for that
instability to estimate the location of boundary-layer
transition.
382
appropriate: "It is of utmost importance that our
continuing work proceed with open eyes and open mind;
that new knowledge be subject to the tests of the U.S.
Transition Study Group (Reshotko 2 ), especially the
generalized guideline number four: 'Experiments (and
computations) where possible should involve more than one
facility. Tests should have ranges of overlapping
parameters, and where possible, redundancy in transition
measurements.' Only in this way will our efforts avoid
inferences based on insufficient evidence and yield a
furthering of our understanding of laminar-turbulent
transition."
REFERENCES
383
8. Kendall, J.M., "Wind Tunnel Experiments Relating to
Supersonic and Hypersonic Boundary Layer Transition,"
AIAA Journal, Vol. 13, No.3, pp 290-299, March 1975.
9. Demetriades, A., "New Experiments on Hypersonic
Boundary Layer Stability Including Wall Temperature
Effects," Proceedings of the Heat Transfer and Fluid
Mechanics Institute, pp. 39-54, 1978.
384
1947.
385
Conditions," Philco-Ford Corp., Mechanical Engineering,
TN 110, October 1966.
386
39. Demetriades, A., "Nosetip Transition Experimentation
Program, Final Report, Vol. II," SAMSO-TR-76-120, July
1977.
387
51. Spangler, J.G., and Wells, C.S., Jr., "Effects of
Free Stream Disturbances on Boundary Layer Transition,"
AIAA Journal, Vol. 6, No.3, pp.543-545, March 1968.
388
Roughness," Phys. Fluids, Vol. 29, No. 10, pp. 3199-3213,
1986.
389
Boundary Layers," Journal Aero. Sci., Vol. 19, November
1952.
390
84. Batt, R.G., and Legner, H.H., "A Review of Roughness
Induced Nosetip Transition," AlAA Paper No. 81-1223, June
1981.
391
NOMENCLATURE
h Enthalpy
k Roughness height
kHZ Kilohertz
P Pressure (psia)
R Radius (inches), also (Re x )1/2
Re Reynolds number
Rc Radius of curvature
U Velocity
392
u Tangential velocity component
A wavelength of disturbance
I.L Viscosity
P Density
AD Adiabatic
B Beginning or blunt
D Diameter
393
E End
N Nose
0 Reservoir or initial
S Sharp
ST Model stagnation point
T Transition, total
W Wall
00 Freestream
394
AC
(- ex i )MAX
3
10
1.2
F1
F 0.8
0.4
0
0 2 3 4 5 6 7 8 9 10
Me
I
RT
R
FIG. 1 A Schematic of a Stability Diagram FIG. 2 Maximum First and Second Mode Spatial
Amplification Rates at R s 1500
Moo = 8
a c = 7°
SHARP TIP
x(IN.) Rex
- 36 -4.3 X10'
-34
- 32
- 30 - 3.6 X10'
- 28
- 26 - 3.1 X101
- 24
l~~iiiililill"~
- 20
- 18 - 2.2 X 10'
- 16 -22-2.6Xl01
- 14 - 1.7 X10·
- 12
- 10 - 1.2 X 10'
o 1DO 200 300 400 500 600
FREOUENCY (KHz)
Moo = 8
ec = 7°
SHARP TIP
396
SHARP CONE
Mao = 8
8c = 7°
Re x = 4.4 X106
YUN.)
- .14
SHARP CONE
M.., = 8
8c = 7°
Rex = 4.4 X108
Y(JN.)
-~
- .22
A
o 100 200 300 400
FREQUENCY (KHz)
FIG. 5b Fluctuation Spectra, Normal to the Surface.
From the Surface, Looking Out.
3CJ7
3r----r----r---~--~~--~--~~--~--~----~--~----~--~
o
o o
Re./FT IINIIN.)
0 1>cl(Jf1 0
1:1. 2)(1(Jf1 0
0 2.5)(10' 0.15
2 3 4 5 6 7 8 9 10 11 12
Re,ll0- 1)
SHARP CONE
M.=8
Me=6.8 6
Re~ =3.28xlO
IJ DISTURBANCES
4 AMPLIFIED
FOO')
:~
Moo =8
Oc 7"=
=
RN 0.1SIN.
-10.S x106
-9.3 X 108
-8.3 X 106
-7.1 X 108
-S.9 X 106
~.7 X 108
-3.S X 106
f(kHz)
RN (IN.) Re~/FT
• 0 1 X 10'
• 0 1.3 X 10'
.005 • .1S 2.5 X 10' •
•004 •
•••••
• ..~. • •
-a,.003 +++
.002
TRANSmON •
• .+
••
.001
0
SHARP CONE ~ ••
Re~/FT= 1 X 10' P-I:++ • •
399
SECOND MODE
·004
.003
-al MAX
.002
.DD1
Ov 8 12 18 18 20 22 24
10 14
R(10-2)
28
f = 115 kHz. TW/T. -.42
24
20
! 16 f - 90kHz
A1 TWflu =.82
12
•
4
~ RexT(~- .82)
0
2
Rex (10-&)
3 4 " 5
400
TwIr• He./FT
.42 .12
5 0.511111
0
•••••••••• • a
0
1... 1111
1.311111
4 • •••• A 2.0.,111
0
60
AO
•
0
06
o 8 10 12 14 18 18 20 22 24
R1111'2)
FIG. 12 Maximum Growth Rates for Second Mode Disturbances
4.5
4.0
3.5
3.0
0
0
2.5
- cc i 0
(10 3) 2.0
0
1.5 51'
1.0
401
SHARP CONE
M~ =8
Me ,",6.8 UNEAR STABIUTY
THEORY (MACK2~
Reag 1m z 3.28x106
(- ex i )
3
(10 3 ) •••
•••••
...................
MAX
2
WIND TUNNEL DATA
10 12 14 16 18 20 22 24
R(10- 2)
14 X 106
12
10
Re XT 8
0
0 2 4 6 8 10
M
FIG. i5 Effect of Mach Number on Transition
402
10.0
8. 0
7.0
D
6.0
o
...... 5.0
Ii>
a: Sym ~
a NASA/Ames 12 PT
4. 0 • NASAlLangley 16 IT
••
A NASAlLangley 16 TOTo
NA SAlLangley 8 TPT
NSR&DC 7 x 10 T
10 \l RAE Bedford 8 x 8 SWT
o NASA/Ames 9 x 7 SWT
I:> NASAlLangley 4 SPT
o NASAllangley 4 SUPWT ITS No. II
2.0 c1 NASAlLangley 4 SUPWT ITS No. 21
o RAE Bedford 3 x 4 HS ST
D AEDC VKF Tunnel A
6 AEDC Tunnel 4T (Walls with Tape or Screens)··
1.0 ~ NASA/Ames 11 TWT (Walls Tapedl··
403
10· FLIGHT DATA: 77 POINTS
~
<>~ Il. <>
~~L
Il.
<:I>
-
Il. b. b.
Rex b.
T 107 Il.
Il. Il.
Il.
Il.
~
~~~~~;;~~._Y.~~~BOUNDARY
LAYER ..
BOW
SHOCK
*FLUID AT EDGE OF BOUNDARY
LAYER PASSED THROUGH A
NEARLY CONICAL SHOCK
404
--r
SHARP
M _C~j:
5 .10 20 X 106
4 .08
p/P ST
M Re/FT
3 .06
lOX 106
2 .04
.02
XswlROTTAI
OL---~--~----~--~----L----L--~~ 0 0
o 2 4 6 8 10 12 14
X (IN)
FIG. 19 Calculations of Local Flow Properties on an 8-Deg.
Half Angle Cone with 2% Bluntness at M~ = 5.9
16
14
12
10
8
-------7°
6
-8°
4 _ _ _ _ 10°
2
--------215°°
4 8 12 16 20
Moo
FIG. 20 Entropy-Layer-Swallowing
Distance Parameter
405
1.1
<D
I"
I
-~"
ii
1A
0
0
.
1.2
RWIIII CIP~
0 .01
D
1.G 0 .G3
(Rox"', '"
.Il
.IN
.All
(AoII't1e•
..
(AoII't)aa
lRox"'. 0
<I
rI
.10
.11
.ao
IS .25
6 .ao
. tu.wl
T.
--.II
64.~4fI~
TO
A 6
406
1000~--------------~--------------~-------------'
_f
-
DEMETRIADES 39
. ----.....
TRANSITION
[Re, ] SONIC
POINT
-
100 -~
......
"NORMAL" FRUSTUM/.' NOSETIP
TRANSITION '"tAMINAR
,
40% PANT/DEMETRIADES
10
0.1 1.0 10 100
[(1-)( ~: )] SONIC
POINT
.-- t-t.\
--- .-.-.---.....\:'\ ••
---
•...
RN/Re
o
Reo,'FT
9.7X10 6
... .
.02 19.4Xl0 6
.
19.4X10 8
4 ~ .05
.10 19.4Xl0 6
.15 24.9Xl0 6
~.--- ~\~
\\'* "-
.
3
~.
--- ---
III
...
:---1--.-
...
....
". t. __
... ___ ... _ ... ~.\
.-.- ~
;.:------._'----
'\
'- ------
..'I...."..., -- ---.. - *•
..... _----
o ~I____~I~--~I~--~~I~--~I----~I~--~I~--~~I~--~I
1.0 .75 .50 .25 0 .28 .50 .75 1.0
WINDWARD LEEWARD
407
14 SHARP CONE
0( =2°
12
Reo
1200
10
-=__~====~==~~~ E
E
600
-------------------------------------------- 8 400
o .
~ ----~----~ ____ ____ ____ ____..______. ____
~. ~ ~. ~----~.
1.4
•
o·
SHARP CONE
M .. - 5 .9
1.2 Se· 8 °
30·
60·
90·
--------------120·
----------150·
-----------160·
.2
Ot"
o .1 .2 .4 .S
408
10% BLUNTNESS
14
oC= 2·
12 E
= o·
10
B
8
X E
(IN .)
6
500
B
4
400
0 I I I I I I I I I I
0 20 40 60 80 100 120 140 160 180
~ (DEG.)
•
1.4
1.2
o·
30·
1.0
60·
90·
~~::---120·
150°,180°
.2
o<~----~----~~--~I~--~'~--~'
o .1 .2 .3 .4 .5
c(t8.
409
108 SYM Me 8" ...
0 4.3 10
0 4.3 10
4.3 10
••..•
IlJ
2.1 10
2.1 10
2.1 10
2.0 4
o
o
o o
•
• 0
o ~
• • 0
107 f/l • • CD e~ 0
fIlt •• 0615'
.. f/l " 0
• •• 0~ ..10 0
• ~~o~
• .~:~!'
o #)
B
• •
2 X 106
2 X 105 106 107
R~
IN,
SCHUBAUER AND
SKRAMSTAD
410
a
I,
, AVCO SHOCK TUNNEL
,
\ 8° HALF ANGLE CONE
Cl ••
<t
I-
0.4 • WIND TUNNEL (NACA )54
Moo = 3.1
?; CONE-CYL g030'TOTAL
I- ANGLE
c:C Oc:2
•
N.O.L 55(RANGE)
0.2
a a 10° TOTAL ANGLE
a:J aa o Moo = 3
0-
0- 0- a Moo = 5
- ALL LAMINAR
0
• SHOCK TUNNEL
1.0
Moo=5.5
0.8 Be = 8°
Cl
<t TRIP SIZE
I- 0.004" 0,0.0005"
0.6
?;
0.001"
I-
0.4
- - REF 56
SHARP Mc.o =2.7
0.2 CONE Be = 5°
56 )
a
0 2 4 6 8 10 12 xlo 6
Re xT
411
4
- - CONE
___ PlATE
3.5
M. = 3.54
3
Re XT
2.5
(Re X ) n
T AD 8.2 "
2 --....., "
/'
, / \ 6.8
'4 \ ,
1.5 6.0,
4.9 ,.......
\
;a. . . -" . . . . . _
- -'-....:..... .... ------"'::...'---
11.i~ ... _
----~~--
- - - __ _
4.1
--
0.5 L...--iL...---iI....---I_--I_--'_--'_....j,,_....j,,_....j,,_.....
o 0.2 0.4 0.6 0.8 1.0
~
(Tw) AD
2V4---U(y)
w'
u'
412
3
2 c- tB ~ 0i d.
l~ i
0
(Rex T) CONE
(Rex T) PLANAR
~ Q
• •(f
(f
-
0 I I I I 1 I
3 4 5 6 1 8
MACH NUMBER
.005
o
o 1000 2000 3000 4000
U (FT ISEC)
413
2 x 10 7
0
10 7 ~
i 4l.
~) SHARP DElTA WIN6S
0
~
1}_m...
~
e
*
[7
LI Ll j LI
~
C>
0
4J.
C> <#
•
a C>
LID
~ 0 Ii ~ 0
0
SIWIP CONE
SUIiItllYBUIIIT COIlE
10 8 ~ l ¢9'~~o 0 0 i~
o ~ ~}::
, .t:Ji,. oo~ ~o ~ ~ COIRURATIIIIS
R8 XT 411 0 <J~ &> q 0 0 SIIARP PYRAMID
~)
~ . . I. 2 ~C> ~Q 0 0
8I.UN1 PYRAMID
APCIUlI
1 I. ,. c> <>~ ~ !} FJI£E.fU6IfT DATA
"'III ~
l}=-~
I.
1. ~ ~ OPEN SYMBOLS 10-s "'. s 45-
10 5
/"'111 .~ ~.~ CLOSED SYMBOLS
HALF·OPEN SYMBOLS
48·5 0 .
eo· s CI.
S
s 70·
59-
~
5x104
0 2 4 6
Me
Res/Me
1000
I
I
I
I
I
I
I
I
I
I
TRANSITION TREND
FOR SHARP CONES
o 2 4 6 8 10 12 14 16
Me
FIG. 36 An Illustration of Ree/Me variations
414
Re e/ Me
10 9
SHARP CONES 400
300 10'
FRUSTUM MATERIAlS:
200 o TEFlON A CARBON PHENOLIC
C METAlLIC o QUARTZ PHENOLIC
10 8 o GRAPHITIC o 3· 0 QUARTZ PHENOLIC
100
10 7
$~§1,08
Rex TRANSITION TREND Re eT oetl~
10' I-
FOR SHARP CONES
10 6
'h~~~D
~ ~ 00 :1a
10 5 A
10' I
10 4 ,
10· 10' 10' 10' 10·
2 4
O~~~L.-..I...-...I
6 8 10 12 14 16
Me XT/R N
240
o 2 3 4 5 6
Me
FIG. 39 Transition Onset at the Attachment Line
of a Swept Cylinder
1.4
1.2
1.0
(Re x )
T BLUNT
,
.8
(Re x )
-1
T SHARP
I
.6
.4 ,y., ,",,,'" , ,
(M. ) SHARP ~ 12 ,,',f}
.2
(Me)
SHARP
Z 15 rf
0
1.0 10
X TS
Xs;-
FIG. 40 Transition Reynolds Number Variations Within the Entropy Layer
416
S I AN 8 e = 5°
1.5" FOR: _ 3 CI = 0
Re- , A)
AN" .75" 10"
200
K (- 00 1/3 COLD WALL
N
FT AL T(KFT) Moo
180 200 20
150 16
160 100 10
80 6
60 4
140 40 2.5
120
ALT
(KFT)
100
80
60
40
o
o~o 40 60 80 100 120 140 160 180 200 220 240 260
(FT)
417
SOME VISCOUS INTERACTIONS AFFECTING
THE DESIGN OF HYPERSONIC INTAKES AND NOZZLES •
by
J.L.Stollery
The College of Aeronautics
Cranfield Institute of Technology
Cranfield, Bedford MK43 OAL
England
SUMMARY
1. INTRODUCTION
2. 'TWO-DIMENSIONAL' INTERACTION
419
relation M a
00 I
'" 80 i 112 where a
I
is measured in degrees. Although
intended for Moo>5 this equation seems to give a rough guide even at
supersonic Mach numbers. What is obvious is thlt super~onic laminar
layers are easily separated e.g. M..= 3., Re L = 10, a l '" 4. Even at
hypersonic Mach numbers the turnin~ angle is only modest before separation
begins, e.g. M = 6, Re = 106 , a = 6. (In both cases above the value of C
00 L I
has been taken as unity). Fortunately in most flight conditions the
Reynolds numbers will be high enough for the boundary layer to be
turbulent and hence much more resistant to separation.
All the mathematical modelling has assumed steady flow and, provided
the boundary layer is entirely laminar, the experiments confirm this
assumption even when the flow is well separated. This is not always the
case in turbulent flow as described in the section 2.3.
420
2.3 Turbulent Flow
The corresponding pressure and heat transfer distributions for
turbulent flow over a compression corner are shown in Fig. 5. In contrast
to laminar flow the heat transfer in the separated region increases and
the close similarity in form between the two distributions is obvious.
This has led to a number of simple methods of calculating the heat
transfer once the pressure distribution is known (see for example Ref. 3
by Coleman and Stolleryl.
421
3. GLANCING INTERACTION
422
calculations confirmed the vortical nature of the separated zone and gave
good agreement with the measured pressure distributions on the side wall.
A more serious problem occurs on the leading edge of the strut. The
bow shock causes boundary layer separation which in turn generates an
oblique shock springing from the separation line. This oblique shock wave
intersects the bow shock and modifies its shape (Fig. 9). Immediately
below the shock intersection point in the plane of symmetry, the flow is
now processed by two (weaker) oblique shock waves instead of by the strong
bow shock wave surrounding the rest of the strut. The result is a
supersonic stream of high total pressure which impinges on the strut
leading edge to give a very localised region of high pressure and very
high heat transfer rate (region B, Fig. 9). Measurements with turbulent
boundary layers at supersonic Mach numbers have indicated local heat
transfer rates around three times the stagnation point value. Laminar
boundary layers are much more prone to separation so the interaction
region for a given strut leading edge diameter is far greater. Laminar
flow measurements at Mao= 14 along the leading edge of a circular nosed
strut reached a peak q of 10 times the stagnation point value due to the
shock/shock interaction described above. This type of flow is also very
important near intake lips and will be discussed in greater detail later.
423
The mathematical modelling of these complex blunt-strut/
glancing-interaction flows is improving rapidly. Pictures taken from the
work of Hung and Buning (Ref. 10) clearly show the vortex formation and
match the experimental surface oil flow patterns (Fig. 10).
Two ways of reducing the interaction problems are (j) to make the
strut leading edge radius as small as possible, (ii) to sweep back the
leading edge. If the leading edge diameter CD) is less than the
boundary-layer thickness then the interaction is significantly weakened.
However since q is proportional to l/~there is clearly a limit to the
reduction possible. The effects of sweep back are very powerful and
mostly beneficial. As the blunt edge is swept backwards so the size of
the interaction region collapses (Fig. 11), the strengths of the vortices
are weakened and the pressure and heat transfer rate peaks in the surface
distributions are greatly reduced. More importantly the shock/shock
interaction pattern is changed and the peak heat transfer rate along the
strut leading edge is significantly lowered (Fig. 12, taken from Ref. 12).
The only adverse effect may be due to leading edge contamination. The
flow at the leading edge of an unswept strut will be laminar and the
maximum heat transfer rate will be the laminar stagnation point value. If
the strut is swept back and is mounted from a surface over which the
boundary layer is turbulent then the turbulent flow in the junction may
contaminate the complete leading edge attachment line, so significantly
increasing the heat transfer rate values. Contamination depends on the
radius of the leading edge and the sweep angle, as well as the Mach number
and Reynolds number. Poll (Ref. 13) has established the conditions under
which contamination occurs.
If a straight swept strut spans two surfaces then one junction will
have a swept-forward configuration. Intuitively such a junction looks
unattractive but there are currently few experimental data on which to
base a judgement. However some recent pictures taken in a gun tunnel at
Cranfield at Moo = 8.2 with a blunt strut (a circular cylinder) protruding
from a flat plate, show that sweepforward increases the size of the
separated flow region.
424
4. SHOCK-SHOCK INTERACfION
Measurements are difficult because the peak heat transfer rates are
very localised and the flow may again be unsteady. The practical
importance and the severity of shock/shock interaction was demonstrated as
early as 1967 when a strut carrying a ram-jet model below the X-IS burned
through during a flight at M = 6.7 and the model tore loose. In Holden's
tests (Ref. 18), specifically designed to examine the intake lip problem,
F values for a single oblique shock configuration ranged from around 5 for
laminar flow and reached more than 20 for turbulent jet conditions.
Of course an intake may have two (or more) wedges to help compress
the flow. In Holden's experiments a single wedge of semi-angle 12.5 0 was
replaced by a double wedge system turning the Mach 8 flow through 7.5 0 and
then So. If the two oblique shocks coalesced just before interacting with
the bow shock ahead of the lip, then both the peak pressure and peak
heating rate were increased (F rose from around 20 to nearly 30). However
by suitably positioning the lip so that the two oblique shocks were still
separated when reaching the lip-bow-shock, the interaction could be spread
over a larger area. The peak pressure was reduced by a factor of 4 below
the single wedge case, whilst F was reduced from around 20 to 7.
425
comparisons with experiment were possible at that time. Further numerical
results were sparse until the papers by Klopfer and Yee (Ref. 19a)
(1988), Stewart et al (Ref. 19b) (1988) and Singh et al (Ref. 19c) (1989).
In the latter report some interesting comparisons with experiment were
presented. The calculated results agreed very well when an adaptive grid
technique was used.
5. VISCOUS INTERACTION
The last example shows that (as mentioned earlier) the effects of
viscous interaction with a turbulent boundary layer are less dramatic but
still need consideration. Fig. 16 plots the pressure distribution over a
15° flap at M = 9.22. There is little upstream influence ahead of the
corner but '"
downstream the effect of viscous interaction is again to
426
relieve the abrupt pressure jump forecast by an inviscid calculation.
Using the theory outlined above gives a reasonable estimate of the
experimental pressure distribution. (Fig. 16).
Tests already completed show how complex the flows are surrounding
glancing interactions. With a sharp-edged strut the incipient separation
angle is far lower than for the corresponding '20' flow. Once separation
has occurred there are doubts concerning the steadiness of the flow. For
a blunt strut, separation is more likely and the degree of flow
unsteadiness will increase. The benefits of sweep back are clear but the
effects of sweep forward are harmful. The utility of fillets at junctions
does not seem to have attracted much attention and would be worthy of
further study.
427
7. REFERENCES
428
16. Edney, B. Anomolous heat transfer and pressure distributions on
blunt bodies at hypersonic speeds in the presence of an impinging
shock. FFA Report 115, Aero. Res. Inst. of Sweden (1968).
17. Keyes, J. W. & Hains, F.D. Analytical & experimental studies of shock
interference heating in hypersonic flows. NASA TN D-7139 (1973).
18. Holden, M.S. Shock-shock boundary layer interactions. AGARD Report
No.764 (1990). See also Weiting, A.R. and Holden, M.S. AIAA J.
vol. 27 , no. 11, (1989).
19. Tannehill, J.C., Holst, T.L., Rakich, J.F. and Keyes, J.W.
Comparison of a two-dimensional shock impingement computation with
experiment. AIAA J. vol 14, no 4, (1976)
19a. Klopfer, G.H. and Yee,. H.C. AIAA Paper 88-0233, (1988).
19b. Stewart, J.R., Thareja, R.R., Wieting, A.R. & Morgan, K. AIAA
Paper 88-0368, (1988).
19c. Singh, D.J., Kumar A. & Tiwari, S.N. Influence of shock-shock
interactions on the blunt body flow field at hypersonic flight
speeds. AIAA 7th Applied Aerodynamics Conference Seattle (1989).
20. Stollery, J.L. and Bates, L. Turbulent hypersonic viscous
interaction. J.Fluid Mech. Vol 63, part 1, (1974).
21. Kumar, A. Numerical analysis of flow through scramjet engine inlets.
AGARD Conference Proceedings 428 "Aerodynamics of hypersonic lifting
vehicles", Bristol, England (1987).
22. Kubota, H. and Stollery, J.L. An experimental study of the
interaction between a glancing shock wave and a turbulent boundary
layer. J.Fluid Mech., vol. 116, (1982).
23. Price, A.E. and Stallings, R.L. Investigation of turbulent separated
flows in the vicinity of fin type protuberances at supersonic Mach
numbers. NASA TN-D-3804, (1967).
24. Stollery, J.L. and Beyer, U. Hypersonic viscous interaction
revisited. Proceedings of the International Conference on hypersonic
Aerodynamics, Manchester, U.K. Royal Aero. Soc. London, (1989).
25. Stollery, J.L. Hypersonic viscous interaction an experimental
investigation of the flow over flat plates at incidence and around an
expansion corner. Aerospace Res. Labs., Wright-Patterson Air Force
Base, ARL 70-0125, (1970).
8. NOTATION
Other symbols are defined in the text or in the figures.
C constant in the law Il = CT ~ boundary layer thickness
Cr skin friction coefficient ~L value of ~ at x = L
D diameter of leading edge Il viscosity
f function of, see section 5 P density
L distance to compression corner (see figs)
p pressure Suffices
q heat transfer rate e value at edge of
ReO Reynolds number:; P",u", Of Il", boundary layer
St Stanton number :; q f P.o 1{. (h. - hw) f.p flat plate
T temperature 0 stagnation point
u velocity (stag. also used)
x,y,z Cartesian co-ordinate system (see figs) r recovery
ex. wedge angle or compression corner angle w wall
ex. 1 value of ex. for incipient separation free-stream
'"
429
M~»l
ShoCk I Shock
Int~ractlon
_ ~-, _,
57
Glenclng IOtifcctlon on roof and floor
caused by fud InJcctlon rut
Turbulent viscous
interaction
If
\ laminar viscous
- --
\. interaction
, . ~r-/ /-'------
~
/
Transition ZtiQUC shock boundary layer Glancing interaction
int(rQctions and turbulent on side wailS.
viscous interaction.
Cylindrical leading
Bow shock
edge of cowL
Incident shock
from intake surfaclZ
Sup~rsonic jet
impinges on surface
430
SOlid symbolS. ol1Qchcd flow
~
Ref. 1.
t--L: L ,I
,..
Symbol oc..
7.•
10.0
1'.0
15.3
x (inchn) •
JC(inc:hU)
(a) Detailed distributions near the corner. (b) Overall distributions, attached flow .
..
Pip", -
~; V7 ~
r,;
f.p.
J.... ex = 7.6'
7 •
x (inches)
, .
)( (inches)
431
For Key to Dato Poinh Sec Ref 1
l00~,-~~~~~~-,__,
V ,... = 1.4 0 M.= 16.0
o 9.7 6.2
<> 14.5. 3.0
14.6 6.0
• turbulent
10 l-4-----~r==~c;,;,4 ...............j
.
'-' ,
Symbol
(a) Pressure distribution near the corner. (b) Heat tronsfllf rotlZS "czar the co~ner.
00'
M. =: 9.22 , RC L = 26 II 10 6
TwITo = 0.28
"
~
d
F, 1
60
PIP.
" ClI._IS·
(Ww-'d';';;;lu()
432
30
40
.,..',
20
30
a. i..
20 R06L
(b) Predicted values of ~L from ref. 5.
40
10
--..,
10~-----'0~S'------~'~06'--------:,0'
4
0
........
.......... T --
M.st, .••T~~
~r·' 'I}T•.O.lly
r::.~.: I.
M.: '.3
R06L
(a) Data collection, '00 rot. 5.
(c) Predicted values of ~L from ref.5.
Fig.6. INCIPIENT SEPARATION AT A COMPRESSION CORNER - TURBULENT
FLOW.
Appt"".lI'Oundar)'·I.,.rr
"Wclr.nn,
Con"""",rlillt.
433
20 ~~I 0'p cn symbOlS - attochC'd flOW} fot dote
0'... ,I...., DetaCh.d F,,,.d <,mbols - scpo'oted "ow <0 " " sec "'S ~a ~ n
lO / ,-. ShoCk [ U
.
/1
. ,....
11 ...... '--.,--T-.. . __
I LaW('975l\
Korltcg i·s c r i t u i o ; - . . . . r - - - - -
Hold,n(l98<) - -,"
O,~--~-~------~-----~-~-~--~-~,7,--~-~Il·
ond heot
Sf:PQ(olion
lin ...
'5r----------------.e>.o~w~~---
snoc'll
'Ie
- l l) -0.5 o
'/0
(,a ) j)~rt l c l c p~thS. I n tne .plane of symm1ur)'
434
8
X,em
32
VI$> ~Y A:45"
-::JL. "0·
•
\
q----o--<l-_-o- __ o-
o - -
y.cm y, em Y, em
435
IS~
IP
DS EJ::pansiCWI
I",
Type Yl.
J\ET
as
".~<I _SL IP
IS
M<I
Type IV.
-
Supersonic jd impingement
M~
m
IS - in(id£nl snock
BS - Bow shock
TS - Tronsft,Jltcd sl"lock
Sl - Shear IQy~r
I P - Impingement point
as'Z~IP
M~~~SL
Type III.
Shear loycr ottachment
IS .... ~
Type II.
Shock impingement
T .... I
Shock IMpingement
160
140
Olstance ,metres J
0:- 120
.~
::["--
100
~
0
, l '
~ 1
(
,. --.,10· '0
12 ~
t.
.
1.0
~ o 8 16 24 32 '0
0' Distance, metres
~ 0.6
~ 0.'
c
~ 0.2
~
16
" 32
Distance, mctrl:s
'0
436
60 MIlD =IL.8
Re.,-"O.S.,0S ~~
Tw/To =0.57
C;: 2.2S H
LO
Experiment (A_f. IS'
Theory (Ret l"
lnviscid
0 --;o;--:o::--::-------
i~ ________________
0
o_____ ~ __o___ _
0 0
'_IT. s 0.Z8
12;- _ _ _
!
10 I 10
I
I
8,
I
6 I o Experiment
Theor),
'nviscid
-2 6 -6 -4 -2 4 6 8
(0) z.-L(cm) (b) ",-L(em)
437
PERMISSIONS
Figs. 3, 5b, 7. Reprinted with permission of New South Wales University Press from
Stollery JL, Stalker RJ (1983): The development and use of free piston wind tunnels.
]n: Shock Tubes and Waves, Archer RD, Milton BE, eds. Sydney: New South Wales
University Press.
Fig. 19. Reprinted with permission of Academic Press from Holder OW, Schultz DL
(1962): The duration and properties of the flow in a hypersonic shock tunnel. In:
Hypersonic Flow Research, Riddell FR, ed. New York: Academic Press.
Chapter 8, J. L. Stollery:
Fig. 6a-<:. Reprinted with permission of Cambridge University Press from Elfstrom
GM (1972): Turbulent hypersonic flow at a wedge compression corner. J Fluid Mech,
Vol. 53.
Fig. 7a,b. Reprinted with permission of Cambridge University Press from Kubota H,
Stollery JL (1982): An experimental study of the interaction between a glancing shock
wave and a turbulent boundary layer. J Fluid Mech, Vol. 116.
Fig. 10. Reprinted with permission of Cambridge University Press from Hung CM,
Buning PG (1985): Simulation of blunt-fin-induced shock-wave and turbulent
boundary-layer interaction. J Fluid Mech, Vol. 154.
Fig. 14. Reprinted with permission of U.K. Royal Aero. Soc. from Stol1ery JL, Beyer
U (1989): Hypersonic viscous interaction revisited. In: Proceedings of the International
Conference on Hypersonic Aerodynamics.
Fig. 16. Reprinted with permission of Cambridge University Press from Stollery JL,
Bates L (1974): Turbulent hypersonic viscous interaction. J Fluid Mech, Vol. 63.