You are on page 1of 12

PHYSICS OF FLUIDS 17, 036102 共2005兲

Film cooling effectiveness on a large angle blunt cone flying


at hypersonic speed
Niranjan Sahoo
Department of Mechanical Engineering, Indian Institute of Technology, Guwahati—781039, India
Vinayak Kulkarni, S. Saravanan, G. Jagadeesh, and K. P. J. Reddya兲
Department of Aerospace Engineering, Indian Institute of Science, Bangalore—560012, India
共Received 28 July 2004; accepted 7 December 2004; published online 23 February 2005兲
Effectiveness of film cooling technique to reduce convective heating rates for a large angle blunt
cone flying at hypersonic Mach number and its effect on the aerodynamic characteristics is
investigated experimentally by measuring surface heat-transfer rates and aerodynamic drag
coefficient simultaneously. The test model is a 60° apex-angle blunt cone with an internally mounted
accelerometer balance system for measuring aerodynamic drag and an array of surface mounted
platinum thin film gauges for measuring heat-transfer rates. The coolant gas 共air, carbon dioxide,
and/or helium兲 is injected into the hypersonic flow at the nose of the test model. The experiments
are performed at a flow free stream Mach number of 5.75 and 0° angle of attack for stagnation
enthalpies of 1.16 MJ/ kg and 1.6 MJ/ kg with and without gas injection. About 30%–45% overall
reduction in heat-transfer rates is observed with helium as coolant gas except at stagnation regions.
With all other coolants, the reduction in surface heat-transfer rate is between 10%–25%. The
aerodynamic drag coefficient is found to increase by 12% with helium injection whereas with other
gases this increase is about 27%. © 2005 American Institute of Physics. 关DOI: 10.1063/1.1862261兴

I. INTRODUCTION aerodynamic as well as structural problems. Hence, alternate


form of cooling systems must be adopted for hypersonic
Management of aerodynamic forces and aerodynamic flights in order to have a feasible, cost effective, and efficient
heating is the key issue driving the design of vehicles trav- TPS in comparison with other conventional cooling tech-
eling at hypersonic Mach numbers. To some extent this is niques.
achieved by adopting appropriate body configurations for hy- Over the years, there have been many investigations4–7
personic vehicles. Use of large angle blunt cone configura- on the feasibility of using forward-facing structural spike
tions 共semiapex angle 30°–70°兲 brings down the aerody- 共extending from nose of the blunt bodies兲 to reduce the drag
namic heating within reasonable level with marginal and surface heating rates at hypersonic speeds. However, the
enhancement of drag force, which could be advantageous for spike becomes hot and ablates as a result of large stagnation
the purpose of aerobraking for reentry vehicles.1–3 Ultimately temperature and hence requires frequent replacement or ac-
the safety of the hypersonic vehicle is ensured by providing tive cooling. Many other studies have investigated focused
appropriate thermal protection system 共TPS兲. energy deposition and plasma/hot gas injection in to the ex-
In general, the design of TPS for a particular body con- ternal flow mainly to reduce the wave drag.8–11 This deposi-
figuration is based on the principle of either the absorption or tion of energy upstream of the flow essentially modifies the
rejection of the thermal energy that comes through the aerodynamic features of the body and as a consequence re-
boundary layer of the body. Accordingly, the TPSs are clas- duces peak heat fluxes on the body surfaces.10,11 At the
sified as “radiative-shield” system, “insulated” system, and
present time, a problem of great concern to the safe atmo-
“ablative” cooling systems. Carbon fibers, ceramic tiles, and
spheric reentry of space vehicles suggests another applica-
refractory materials 共such as Molybdenum and Zirconium兲
tion of the forward-facing gas injection. Continuous data
are typically used in most of these cooling systems. In prac-
transmission and communication contact of reentering ve-
tice, the strength of these materials falls off rapidly at high
hicle with the ground during its descent is lost because the
temperatures and thereby limiting the application of these
strong bow shock that forms around the body raises the air
methods for TPS. Although these cooling techniques have
temperature enough to ionize the gas which prevents trans-
been effectively used in the past, their usage in practical
mission of the radio signal. A coolant gas injected at the nose
hypersonic transport system is doubtful. For instance, the
of the body has the ability to some extent to cool the hot
ablative cooling becomes extremely costly when reusability
is the major concern. Also, the shape change that occurs plasma and to induce the recombination. In addition, in some
inevitably in the process of ablation causes various other cases, it greatly alters the shape and strength of the main-
stream shock and thereby helps in reducing the aerodynamic
drag.
Author to whom correspondence should be addressed. Telephone: ⫹91-80-
a兲

22932758, 22933021. Fax: ⫹91-80-23606223, 23600134. Electronic mail: The idea of injecting a gas into a free stream flow was
laser@aero.iisc.ernet.in explored when Lopatoff12 conducted the experimental inves-

1070-6631/2005/17共3兲/036102/11/$22.50 17, 036102-1 © 2005 American Institute of Physics


036102-2 Sahoo et al. Phys. Fluids 17, 036102 共2005兲

FIG. 1. Schematic diagram of mass transfer cooling systems.


FIG. 2. Schematic diagram of IISc hypersonic shock tunnel, HST2.

tigation to see the effect of a forward-facing sonic jet on the


tions, the aerodynamic heating and generation of aerody-
drag of a blunt body in transonic flow. Other
namic drag are complimentary phenomena at hypersonic
investigations13–15 of forward-facing jet considered the effect
Mach numbers. Hence, it is mandatory that both aerody-
of jet on the aerodynamic characteristics of the body from
namic heating and drag be addressed simultaneously. But in
which they issued. In these investigations, the controlling
most of the previous investigations, the drag and heat-
factor was the ratio of jet total pressure to free stream total
transfer reduction by a forward facing gas injection have
pressure. Ultimately, this controlling factor dictates the prac-
been attempted independently by controlling the pressure ra-
tical feasibility of the concept of heat protection or drag re-
tio of the gaseous jets.
duction devices for high-speed free stream flows during gas
Recently, a research program has been initiated in the
injection. More recently, Benjamin et al.16 carried out two
High Enthalpy Aerodynamics Laboratory of Indian Institute
numerical studies to see the effects of forward-facing jet lo-
of Science 共IISc兲 to investigate the effectiveness of film cool-
cated at the stagnation point of a blunt body on wave drag
ing by injecting coolant gases into a hypersonic free stream
and surface heat-transfer rates for Mach 6.5 flow. It was
flow. A 60° apex-angle blunt model has been used as fore
shown that upstream jet injection significantly modifies the
body and various coolant gases are injected through a hole of
flow fields and if the jet conditions are chosen properly, sub-
2 mm diameter at the nose of the test model. An array of
stantial reduction in surface heating and aerodynamic drag
platinum thin film gauges deposited on thermally insulating
can be achieved.
backing material flush mounted on the model surface are
When the effectiveness of forward facing gas injection is
used for convective surface heating measurements while an
studied for heat protection devices, it is necessary to ensure
internally mounted three-component accelerometer balance
suitable pressure ratios that should lead to cooling. Such
system measures the aerodynamic drag simultaneously. Air,
types of cooling systems are called “mass transfer cooling”
carbon dioxide, or helium gas is injected as coolant gas at
that have been found more feasible and cost effective in
standard atmospheric conditions. All the experiments are
comparison with other conventional cooling techniques for
conducted in the hypersonic shock tunnel HST2 at 0° angle
high-speed applications. Essentially, in this type of cooling
of attack with nominal flow Mach number of 5.75. In order
techniques, a cold fluid is injected into the boundary layer
to complement the experiments, CFD studies using a com-
through the surface. The fluid may be injected through a
mercial package CFX-TASC flow have also been carried out.
porous media or through small holes. When the fluid is in-
Prior to actual shock tunnel measurements, some preliminary
jected through a porous media over the entire surface, the
theoretical feasibility studies on film cooling in HST2 in-
coolant comes out as a continuous mass. Such a cooling
volving physical flow patterns are carried out that will be
system is called as “transpiration cooling” technique. When
discussed in the subsequent sections. During the experi-
the fluid is injected through discrete slots, the system is
ments, the critical pressure ratio of the injectant and mass
called as “film cooling system.” In either case, the coolant
flow rates are controlled suitably in order to achieve required
absorbs the incoming heat through its rise in enthalpy and
cooling effect over the test model surface.
thus modifies boundary layer characteristics in such a way
that the heat flow rate to the surface is less. Figures 1共a兲 and
II. EXPERIMENTAL FACILITY
1共b兲 show the schematic of transpiration and film cooling
techniques, respectively. Realization of hypersonic vehicle The HST2 hypersonic shock tunnel is schematically
will remain a distant dream unless an efficient and economic shown in Fig. 2. The shock-tube portion of HST2 consists of
cooling system is developed. Most of the space vehicles a 51 mm 共inner diameter兲 stainless-steel driver and driven
make use of blunt body shapes 共semiapex angle 30°–70°兲 sections, separated by a metal diaphragm. Platinum thin-film
with a large nose radius to facilitate aerobraking maneuver. sensors located towards the end of the driven section monitor
Aerodynamic heating is again a fundamental issue for these the shock-wave velocity. The pressure jump across the shock
body shapes particularly at the nose. In such cases, film- wave is measured using a pressure transducer 共PCB, Pi-
cooling system is one of the popular methods that can be ezotronics兲 located at the end of the driven section. The
incorporated at the desired locations particularly at the nose wind-tunnel portion of HST2 comprises a truncated conical
of the blunt cone models. But for such types of configura- nozzle terminating into 30⫻ 30 cm size test section. The
036102-3 Film cooling effectiveness on a large angle Phys. Fluids 17, 036102 共2005兲

TABLE I. Nominal test conditions in IISc shock tunnel HST2. The numbers inside the brackets indicate the
percentage in uncertainties.

Driver gas Helium Helium

Primary diaphragm thickness 共mm兲 1 1.5


Shock Mach number M s 共±2.5% 兲 3.1 3.6
Stagnation pressure P0 共kPa兲 共±8 % 兲 540 990
Stagnation enthalpy H0 共MJ/kg兲 共±10% 兲 1.16 1.6
Free stream Mach number M ⬁ 共±2 % 兲 5.75 5.75
Free stream static pressure P⬁ 共kPa兲 共±5 % 兲 0.45 0.85
Free stream static temperature T⬁ 共K兲 共±9 % 兲 140 210
Free stream unit Reynolds number Re⬁ 共m−1兲 共±16% 兲 1.4⫻ 106 1.6⫻ 106

typical test flow conditions along with the relative uncertain- field of jet injection. First, the body shape and the size are
ties are given in Table I. These test flow conditions are specified by the fineness ␭ and nondimensional body diam-
achieved by using helium as driver gases and metal 共alumi- eter D and are defined as
num兲 diaphragms of various thicknesses. The HST2 tunnel is
a db
capable of producing a reservoir enthalpy of up to 5 MJ/ kg ␭= , D= , 共1兲
and has an effective test time of about 800 ␮s. A transient b dj
PC based data acquisition system with requisite software is where a, b are the semiaxes of the elliptical section parallel
used for recording and processing of the data. and normal to the free stream, db共=2b兲 is the nose-shape
diameter, and d j is the jet diameter.
III. PHYSICAL PHENOMENA OF GAS INJECTION The properties of the free stream and the jet are specified
by their specific-heat ratios ␥ j for the jet and ␥⬁ for the free
When a jet issuing from a body interacts with
stream. It is also necessary to specify the jet total-pressure
supersonic/hypersonic incoming airstream, the body shock
ratio P and total-temperature ratio T, defined by
stands away from the body surface and takes a form appro-
priate to a new body consisting of the original body with a P0j T0j
protrusion due to the jet flow. The boundary of the protrusion P= , T= , 共2兲
P0f T0
is defined by the interface, i.e., the stream surface between
the jet flow and the mainstream flow. A model of jet and the where P0j is the jet total pressure, P0f is the free stream Pitot
interface flow is sketched in Fig. 3. The assumptions made in pressure, T0j is the jet total temperature, and T0 is the free
the analysis are 共a兲 the mainstream and jet-gas flow consist stream total temperature. In addition, the Mach number of
of perfect gases, 共b兲 the jet flow is axially symmetrical, uni- the free stream and that of the jet are taken as M ⬁ and M j,
form, and parallel to the exit plane, and 共c兲 the variation of respectively.
free stream Reynolds number due to the jet injection may be In a steady flow, both the free stream flow and the gas-
neglected and no account is taken for the heat transfer to/ eous jet come to rest on the axis at stagnation point, where
from the body. the pressure is the free stream Pitot pressure P0f . The jet total
The independent variables are then those, which specify pressure P0j must be reduced to P0f by some shock system in
the size and shape of the body, jet hole diameter, the aero- the jet. For any outflow of the jet, P0j must be greater than
dynamic properties of the jet, and the free stream. Let us P0f . As the jet total-pressure ratio P increases from its mini-
define the nondimensional parameters that control the flow mum value of 1, the bow shock moves forward. In general,
this process is continuous, except for a small critical range of
P where the trend is reversed. Below the critical range of P,
the flow is observed to have more than one jet cell and is
generally unsteady. Above the critical range, the flow is usu-
ally steady with a single jet cell terminated by a normal
shock. The change from one type of flow to the other is
intermittent in character. The estimated value of P midway
between the interception of transition and its completion is
referred to as the critical value Pcrit and is primarily a func-
tion of the jet size and the nose shape.14 Hence, the effec-
tiveness of gas injection as drag reduction device 共P 艌 Pcrit兲
or heat protection device 共1 ⬍ P ⬍ Pcrit兲 is essentially dictated
by the critical pressure ratio. At hypersonic Mach numbers,
the pressure on the blunt body is given closely by Newtonian
FIG. 3. Physical features of the flow during gas injection and corresponding theory.17 The local surface pressures are then primarily de-
independent variables. pendent on P0f and the local surface inclination. Thus, the
036102-4 Sahoo et al. Phys. Fluids 17, 036102 共2005兲

FIG. 5. Variation of the critical total-pressure ratio Pcrit for models of vari-
ous shapes with relative size D.

FIG. 4. 共a兲 Diagram and 共b兲 photograph of a 60° apex-angle blunt cone system as well as sufficiently thick Macor inserts. The plati-
model fitted with an accelerometer balance system and platinum thin film
gauges for simultaneous drag and heat-transfer measurements during film
num thin film sensors deposited on the Macor insert mea-
cooling. sures convective surface heating rates whereas aerodynamic
drag over the model is measured using the three-component
accelerometer balance system. Because of having a hole of
principal independent parameters of dynamically similar 2 mm diameter at the nose of the model, it is almost impos-
flows are thus the nondimensional body diameter D, the jet sible to mount the heat-transfer gauge exactly at the stagna-
total-pressure ratio P, the jet exit plane Mach number M j, tion point. However, the heat transfer gauge mounted closest
and the jet-gas specific heat ratio ␥ j. to the stagnation point is at the location of s / Rn as 0.3333
where s is the arc length on the cone surface from geometric
stagnation point of zero incidence and Rn is the nose radius.
IV. TEST MODEL CONFIGURATION AND
INSTRUMENTATION
V. SHOCK TUNNEL EXPERIMENTS
In the present investigation, a 60° apex-angle blunt cone
model with base diameter db of 70 mm and bluntness ratio Prior to the shock tunnel testing in HST2, some prelimi-
共defined by the ratio of nose diameter and base diameter兲 of nary theoretical investigations have been carried out in order
0.857 is chosen to study the effectiveness of film cooling by to estimate the controlling parameters of the gas injection in
gas injection. During this study, the convective surface heat- order to achieve the film cooling. In order to have cooling
transfer rates over the model surface and aerodynamic drag effects over the surface, it is mandatory to maintain adequate
on the model are simultaneously measured. The schematic pressure ratio P such that the gaseous jet will not have suf-
diagram for the test model is shown in Fig. 4共a兲 whereas the ficient strength to push bow shock away from the body and
photograph of the model is shown in Fig. 4共b兲. The coolant rather it expands over the surface. So, it is necessary to main-
gas is injected into the hypersonic flow through a hole of tain the jet-total pressure ratio, less than that of critical pres-
2 mm diameter d j at the nose of the model shown in Fig. sure ratio 共1 ⬍ P ⬍ Pcrit兲. Finley14 carried out extensive inves-
4共a兲. A flexible pipe connected to this hole passes internally tigations to study the effects of relative body size and
along the hollow sting of the accelerometer balance system fineness ratio on critical pressure ratio. The variation of criti-
and then is connected to the external gas supply system cal pressure ratio as a function of relative body size and
through a vacuum-sealed port. The external gas supply sys- fineness ratio is shown in Fig. 5. For the present model with
tem contains coolant gases at high pressure in a cylinder and D = 35 and ␭ = 1, the critical pressure ratio is found to be 6.5.
the outlet pressure from the gas cylinder is controlled by The free stream Pitot pressure is measured by mounting a
pressure regulator. Just before the onset of the shock tunnel Pitot probe in the test section for different test conditions in
operation, the coolant gases are injected from the gas cylin- the shock tunnel. However, in the present experimental
ders under standard atmospheric conditions into the flow setup, the test section sizing does not allow mounting the
through the flexible pipe. The test model has the provision to model and permanent Pitot probes simultaneously. So prior
accommodate an internally mountable accelerometer balance to actual testing of the model in the tunnel, sufficient Pitot
036102-5 Film cooling effectiveness on a large angle Phys. Fluids 17, 036102 共2005兲

TABLE III. Choked mass flow rates for various coolant gas injection.

P0j = 1 bar, T0j = 300 K, D j = 2 mm, M j = 1


Coolant gases ␥j R j 共J/kg K兲 ṁ 共kg/s兲

Carbon dioxide 1.29 189 0.000 890


Air 1.40 287 0.000 743
Helium 1.67 2080 0.000 293

thickness. The nominal free stream and stagnation conditions


along with the relative uncertainties are presented in Table I.
Three to four experiments at each condition are done to
check the repeatability of the signals. Sufficient care has
been taken to verify the repeatability aspect during the mea-
surements.

A. Heat-transfer measurements
FIG. 6. Typical Pitot pressure signals in the test section of HST2 shock
tunnel for different enthalpies. The convective surface heat-transfer rates over the test
model are measured by platinum thin film gauges. These thin
film sensors are deposited on the Macor strips using platinum
probe measurements 共ten numbers for each set of enthalpies兲
05-X metallo-organic ink 共M/s Englehard-Clal, UK兲. A slot
have been performed in the shock tunnel while maintaining
cut in the metallic blunt cone fore-body houses the Macor
almost identical test conditions in the driver and driven sec-
strips that have the provision to take out the electrical leads
tion of the shock tube for two ranges of enthalpies under
from the thin film sensors. Two Macor strips of 18 mm width
study. Typical Pitot pressure histories for two different en-
and 6 mm thick are fixed flush with the metallic model sur-
thalpy conditions are shown in Fig. 6. The jet-total pressure
face into the slots cut along the rays shown in Fig. 4. These
ratio is then computed from the atmospheric jet pressure and
sensors are powered with constant current 共⬃20 mA兲
free stream Pitot pressure and is given in Table II. At this
sources and connected to the PC-based data-acquisition sys-
operating pressure, a choked flow 共M j = 1兲 is achieved at the
tem. The initial resistances of these passive gauges are main-
exit plane of the jet for the two different test conditions. The
tained around 50 ⍀. The change in voltage across the gauge
corresponding choked mass flow rate ṁ of the jet is ex-
with respect to time gives the temperature-time history at the
pressed by the jet total temperature T0j, jet total pressure P0j,
gauge location on the model surface. These temperature sig-
diameter of jet at the exit plane d j, and properties of the
nals are then numerically integrated18 in order to get the
gases ␥ j and R j 共characteristics gas constants兲 and is given by
convective surface heat-transfer rates. The surface heat-
the following equation:

冉 冑 冊冉 冊冋冉 冊冉 冊 册
transfer rates are measured with and without coolant gas in-
␥+1 −1 0.5
j /␥ j
P0j ␲ 2 ␥j 2 jection. When the model is tested with injection of coolant
ṁ = d . 共3兲 gas, the thin film gauge located at the stagnation point is
T0j 4 j Rj ␥j + 1
replaced by a hole of 2 mm diameter. Two typical
The computed values for choked mass flow rate ṁ are given
in Table III.
The experiments are carried out with air, carbon dioxide,
and helium as coolant gases. The operating pressure ratio is
maintained such that the film cooling is ensured and at this
pressure ratio the choked flow is achieved at the nozzle exit
plane. The experiments are conducted out in the hypersonic
shock tunnel 共HST2兲 for two different stagnation enthalpies
at 0° angle of attack. For all shots, air is used as the test gas
and helium as the driver gas and the stagnation enthalpy
variation is achieved by varying the primary diaphragm

TABLE II. Controlling parameters for gas injection 共refer Fig. 2兲.

M ⬁ = 5.75, P0j = 1 bar, M j = 1, D = 35,


Pcrit = 6.5, P 共choked flow兲 = 1.893
H0 共MJ/kg兲 P0f 共kPa兲 P Remarks

1.16 19.6 5.17 P ⬍ Pcrit


1.60 36.4 2.78 P ⬍ Pcrit FIG. 7. Typical temperature-time history from platinum thin film sensors
with and without coolant gas injection.
036102-6 Sahoo et al. Phys. Fluids 17, 036102 共2005兲

temperature-time histories recorded from a platinum sensor Ct

冉 冊冉 冊
with and without coolant gas injection are shown in Fig. 7. Cd = . 共5兲
␥⬁ ␲ 2
For the case of without gas injection, the flow is attached and P⬁M ⬁2 d
2 4 b
has attained steady state over the model surface, which is
exhibited by the parabolic shape of the temperature-time his-
tory of the signal. On the other hand, the temperature-time C. Measurement uncertainties
history signal with gas injection is not of a perfect parabolic
The uncertainties in the test flow conditions, drag levels
shape, which is a clear indication of flow being unsteady in
from the accelerometer balance and surface heat-transfer
that region. Similar trends have been observed for all the
measurements have been estimated using sequential pertur-
runs for which data are presented here.
bation techniques.20 The uncertainties associated with test
flow conditions 共Table I兲 include contributions from uncer-
B. Drag measurements on the blunt cone
tainties in shock tube filling conditions, shock speeds 共in-
In order to understand the aerodynamic characteristics of ferred from platinum thin film gauges兲, and measured out-
gaseous jets during film cooling, it is also important to pre- puts from the data acquisition system. Uncertainties in the
dict the effects of gas injection on aerodynamic drag. So accelerometer sensitivity, output of the data acquisition sys-
while conducting the film cooling experiments for surface tems and restraint offered by the rubber bushes to the free
heat-transfer studies on the model, the aerodynamic drag is flight of the model during test time are some of the factors
also measured simultaneously with and without gas injec- that contribute to the uncertainties in Cd共±9 % 兲. Based on the
tion. The drag force on the model is measured using a modi- uncertainties associated with the gauge characteristics, data-
fied three-component accelerometer balance system19 where acquisition system, data reduction techniques, and calibra-
only the axial force is measured with an accelerometer tion, the measured Stanton numbers st are believed to be
screwed onto the inner surface of the blunt cone along the accurate to ±12%. All uncertainties are 95% confidence in-
axis of the model as shown in Fig. 4共a兲. The balance system terval levels.
with two stainless-steel rings is mounted in the hallow region
of the model’s rear skirt. The two metallic rings get fastened
to the wall of the model skirt by means of threaded screws. VI. RESULTS AND DISCUSSIONS
Each ring is internally fixed with a flexible rubber bushes A. Convective surface heating rates
having a central hole. These bushes are adhesively bonded
Convective fore-body surface heat-transfer rates qt mea-
through their surfaces to the internal surfaces of the metallic
sured at 0° angle of incidence at different locations on the
rings. A sting that supports the model in the test section is
60° apex-angle blunt cone model are expressed in terms of
glued to these bushes through their central hole. These two
ratio of arc length on the cone surface s measured from geo-
rubber bushes in conjunction with central fastening sting acts
metric stagnation point and nose radius Rn. These measured
as suspension system. Hence, selection of rubber bushes with
convective surface heating rates at all the gauge locations are
appropriate material properties such as the Young’s modulus
normalized with free stream velocity and are expressed in
and the stiffness determine the extent to which a sufficiently
terms of Stanton number st. The Stanton number can be ex-
soft suspension system can be assumed. Suitable rubber
pressed as
bushes that justify the free-floating condition are chosen after
carrying out detailed parametric studies and using finite ele- qt
ment modeling 共FEM兲 with a realistic model-balance system. st = , 共6兲
兵␳⬁u⬁关C p共T0 − Tw兲兴其
Some of the salient features of the FEM modeling of the
accelerometer force balance system carried out recently are where ␳⬁ and u⬁ are the free stream density and free stream
reported before.19 During the short test times encountered in velocity, C p is the specific heat of air at constant pressure,
the shock tunnel, the model along with the accelerometer and Tw is the wall temperature. The variation of convective
moves as a free body and the resulting acceleration is mea- heat-transfer rates expressed in terms of Stanton number
sured from the output of the accelerometer. The PCB- along the surface of the model with and without coolant gas
piezotronics accelerometer with sensitivity and frequency injection are shown in Figs. 8共a兲–8共c兲 at two different stag-
range of 10 mV/ g and 10 kHz, respectively, has been used nation enthalpies of 1.16 MJ/ kg and 1.6 MJ/ kg. Figure 8共a兲
in the balance system. The drag force Ct acting on the model shows the variation of Stanton number along the surface of
due to the hypersonic flow is calculated using Newton’s sec- the test model with carbon dioxide as coolant gas whereas
ond law of motion, Figs. 8共b兲 and 8共c兲 show the corresponding numbers with air
and helium as coolant gas. In general, for the blunt cone
model without any injection, the maximum heating occurs at
Ct = M ␰ , 共4兲 the stagnation point21 and it decreases gradually along the
conical portion of the model. In the present experimental
model, it is almost impossible to mount the heat-transfer
where M is the mass of the model 共=0.33 kg兲 and ␰ is the gauge at the geometric stagnation point. However, the heat-
measured acceleration from the drag accelerometer. The transfer rates near the nose of the blunt cone reduce substan-
aerodynamic drag coefficient Cd is obtained from this mea- tially with the injection of coolant gases as shown in Figs.
sured axial force using the relation 8共a兲–8共c兲 and Table III. Essentially, when a coolant gas is
036102-7 Film cooling effectiveness on a large angle Phys. Fluids 17, 036102 共2005兲

surface heating rates over the model surface. In all the Figs.
8共a兲–8共c兲, the pattern of heating rate distribution is almost
similar, i.e., st decreases along the surface of the model both
with and without injection of coolant gas. This is a clear
indication of laminar hypersonic flow over the model sur-
face.
The main effect of the introduction of a coolant is to
reduce the driving temperature difference and this in turn
reduces the heating rate. So sometimes it is necessary to
quantify the amount of reduction in convective heating rates
due to the injection of various coolant gases. This in fact
helps in designing the TPS for such type of vehicles flying at
hypersonic Mach number. So the noninjection data are em-
ployed as the reference parameters in order to quantify the
amount of reduction in heat-transfer rates at various loca-
tions of the heat-transfer gauges on surface. The percentage
of reduction in heat-transfer rates in terms of Stanton number
can be expressed as

共st兲without − 共st兲with
Dst = ⫻ 100. 共7兲
共st兲without

Figures 9共a兲–9共c兲 show the percentage reduction in st along


the model surface with various coolant gas injections at two
different stagnation enthalpies. It is observed that at H0
= 1.6 MJ/ kg, the percentage reduction of st in the vicinity of
stagnation regions 共兵s / Rn其 ⬍ 0.5兲 is higher in comparison to
other parts of the model surface for carbon dioxide and air
injection 关Figs. 9共a兲 and 9共b兲兴. This is not seen in case of
helium injection as shown in Fig. 9共c兲. Since helium is a
lighter gas, it expands faster in comparison with the other
two gases. This aspect has been discussed in the subsequent
section. At 共s / Rn兲 = 0.5, around 35%–40% reduction in heat-
ing rates is felt whereas in the other parts of the model sur-
faces 共兵s / Rn其 ⬎ 0.5兲, the percentage reduction in st at both
enthalpy conditions is on par with each other. Generally,
when a gas is injected into the hypersonic stream, it is likely
to expand at the exit plane and the operating pressure ratio P
controls the characteristics of the jet. With increase in stag-
nation enthalpy, P drops down 共Table II兲. So the flow loses
the characteristics of jet and expands nearer to the exit plane.
During this process of expansion, surface heat-transfer rates
on the model surface decreases. However, the reduction pat-
tern in heating rate distribution for all the locations is unaf-
fected by stagnation enthalpies. It is worthwhile to note from
Fig. 9共c兲 that the overall cooling performance with helium
beyond the stagnation region 共兵s / Rn其 ⬎ 0.5兲 is much better in
comparison with carbon dioxide or air. In these regions, al-
most 32%–42% reduction in heating rate is observed with
helium as coolant gas. So the cooling performance increases
FIG. 8. Variation of Stanton number over the surface of the 60° apex-angle with lighter gas injection and vice versa. This is the prime
blunt cone model at 0° angle of attack at two different enthalpies 共a兲 with function of molecular weight of the coolant gases, which is
and without carbon dioxide injection, 共b兲 with and without air injection, and
共c兲 with and without helium injection.
discussed in the subsequent section.

B. Effect of molecular weight of coolant gases on


surface heating rates
injected from the nose of the body, by flowing back over its
surface the gas forms a “heat buffer” between the free stream It is generally observed that gas injection into the
flow and the body thus providing an effective “blanketing” supersonic/hypersonic stream introduces the destabilizing
on the model surface. This ensures substantial reduction in disturbance that changes the boundary layer characteristics
036102-8 Sahoo et al. Phys. Fluids 17, 036102 共2005兲

TABLE IV. Transport properties of the coolant gases 共Ref. 17兲.

P j = 0.528 bar, T j = 250 K, and M j = 1

Coolant gases
Molecular
weight 共MA兲
Molecular
diameter 共dA兲
共A0 unit兲
共 兲 ␧A
k1
共K兲 ⍀AB
DAB
共cm2 s−1兲

Carbon dioxide 44.01 3.996 190 1.273 0.1746


Air 28.97 3.617 97 0.9878 0.2830
Helium 4.003 2.576 10.2 0.6436 1.2200

ing rates, it is necessary to define a suitable correlation pa-


rameter, which is essentially a function of the molecular
weights. This can be expressed in terms of binary diffusion
coefficient DAB and is defined by Fick’s law.17 It states that if
a species of molecular weight MA is mixed with another
species of molecular weight MB at certain pressure P̄ and
temperature T̄, then DAB can be defined as

冑冉T̄3
1
+
1
MA MB

DAB = 0.001 858 3 , 共8兲
2
P̄dAB ⍀AB

where DAB is in cm2 s−1, P̄ and T̄ are the pressure and tem-
perature of the species A in atmosphere and Kelvin, ⍀AB is
the collision integral parameter and is a function of k1T̄ / ␧AB
with k1 being Boltzmann constant, and ␧AB = 冑␧A␧B with ␧A
and ␧B the characteristic energy interaction parameters for
the species, dAB = 共1 / 2兲共dA + dB兲 with dA and dB as the mo-
lecular diameters for the species in Angstrom unit. All these
parameters are computed for all the coolant gases injected at
the nose of the model and are given in Table IV. In the
present case, the test gas 共air兲 is taken as the species B and
the coolant gases are injected at static pressure and tempera-
tures of 0.528 bar and 250 K, respectively.
The convective heating rates expressed in terms of st at
various locations are plotted against DAB in Figs. 10共a兲 and
10共b兲. In these figures, it has been observed that with in-
crease in DAB the reduction in heating rates follow an in-
creasing trend for all the gauge locations except in the vicin-
ity of stagnation region. For instance, with stagnation
enthalpy of 1.16 MJ/ kg 关Fig. 10共a兲兴, about 20% reduction is
observed at s / Rn of 0.33 with air injection and it drops down
to 8% with helium injection for the same location. At the
same location and with stagnation enthalpy of 1.6 MJ/ kg,
the corresponding values are 35% and 25% with air and he-
lium injection 关Fig. 10共b兲兴. At extreme gauge location with
s / Rn = 1.4 and at stagnation enthalpy of 1.16 MJ/ kg, about
40% reduction in heating rate is felt with helium injection
FIG. 9. Percentage reduction in Stanton number over the surface of the 60°
and it drops down to 15% with carbon dioxide/air injection
apex-angle blunt cone model at 0° angle of attack at two different stagnation
enthalpies with 共a兲 carbon dioxide injection, 共b兲 air injection, and 共c兲 helium 关Fig. 10共a兲兴. At stagnation enthalpy of 1.6 MJ/ kg, around
injection. 35% reduction in surface heating rates is achieved for helium
injection and corresponding values change to 18% with car-
bon dioxide injection at the same gauge location 关Fig. 10共b兲兴.
and produces an effect similar to cooling. The cooling effect Moreover these reductions in surface heating rates is
is essentially produced due to transport phenomena of non- achieved with the choked mass flow rate for helium injection
reacting flows and is a function of molecular weight. So in which is much lower than the corresponding choked mass
order to study the effect of coolant gases on convective heat- flow rates for air/carbon dioxide 共Table III兲. Generally, when
036102-9 Film cooling effectiveness on a large angle Phys. Fluids 17, 036102 共2005兲

TABLE V. Comparison of drag force coefficient for various coolant gas


injections.

Increase in Cd 共%兲

Coolant gases H0 = 1.16 共MJ/kg兲 H0 = 1.60 共MJ/kg兲

Carbon dioxide 27 26
Air 24 22
Helium 12 15

injection. About 12%–25% increase in the aerodynamic drag


coefficient is observed at all enthalpy conditions. As dis-
cussed earlier, the pressure difference is not sufficient to push
the body shock, and hence leads to increase in the gas den-
sity due to gas injection in the stagnation region. This in turn
accelerates the oncoming free stream flow that passes over
the model. However, injection of heavier gases can lead to
higher density in comparison to lighter gases. Hence, with
carbon dioxide injection, about 25% increase in Cd is felt
whereas 12% increase in Cd is observed with helium injec-
tion.

D. Numerical study
In order to complement the experiments, illustrative nu-
merical simulations are carried out using commercial CFD
code CFX-TASC flow.20 Axisymmetric Navier–Stokes equa-
tions have been solved with appropriate boundary conditions
based on the experimental test flow conditions for the test
case with H0 = 1.16 MJ/ kg. The typical boundary conditions
of the simulation for this case are given below.
Inlet. At this boundary of the computational domain the
velocity, static pressure, and static temperature of the flow
FIG. 10. Percentage reduction in Stanton number at various gauge locations have been specified. Typical values used here are velocity
as a function of binary diffusion coefficient with different coolant gas injec-
tion at stagnation enthalpy of 共a兲 1.16 MJ/ kg and 共b兲 1.6 MJ/ kg. = 1412 m / s, pressure= 450 N / m2, and temperature= 140 K.
These are the flow properties obtained at the inlet of the test
section 共free stream兲 in the shock tunnel for an enthalpy level
of 1.16 MJ/ kg 共Table I兲.
a lighter injected gas comes out of the exit plane, it expands Inlet-1. At the nose of the model, air is injected through
much faster along the surface of the body. So the cooling a hole of 2 mm diameter. The velocity and static temperature
effect is only felt at the later part of the surface. For heavier of the sonic air jet are 317 m / s and 250 K, respectively.
gas, in particular, carbon dioxide, the diffusion takes place Outlet. At the outlet of the computational domain, all
much faster in the vicinity of stagnation region. Hence it is variables are extrapolated from the interior domain.
concluded that the cooling performance with heavier gas in- Wall. The wall boundary condition is used at the blunt
jection 共with low DAB兲 is better in the vicinity of stagnation cone model surfaces and the fluid at these surfaces is as-
region in comparison with lighter gas injection 共with high sumed to have a no-slip condition. A constant temperature of
DAB兲. 300 K is specified at the walls.
Symmetry. Other bounds of the computational domain,
C. Aerodynamic drag measurements during gas
excluding inlet and outlet, are specified as symmetry planes.
injection
The fluid velocity is assumed to be tangential at these sym-
Since generation of surface heating and aerodynamic metry planes.
drag are complementary phenomena for blunt bodies flying A typical grid used for the computations of flow fields
at hypersonic Mach number, aerodynamic drag is also mea- around the blunt cone with gas injection is shown in Fig. 11.
sured simultaneously during the film cooling studies. The The multiblock, body fitted grid used for the computations
aerodynamic drag values are then computed by averaging the has a total number of around 25 836 nodes 共grid points兲. The
results during the steady flow duration and the drag coeffi- grid cell height near the wall is around 8 ⫻ 10−5 m. An up-
cient Cd is computed by using Eq. 共5兲. The computed values wind differencing scheme has been used for the computa-
of Cd are given in Table V for the case with and without gas tions with a local time step of 0.9. The target residuals to
036102-10 Sahoo et al. Phys. Fluids 17, 036102 共2005兲

FIG. 11. The grid used in the numerical simulation of flow fields around a
60° blunt cone with air injection. FIG. 13. Numerically computed surface heat transfer distribution 共W / m2兲
for a 60° apex-angle blunt cone model flying at Mach 5.75 with air injection.

terminate the simulation have been set at 1.0⫻ 10−4. About


the blunt cone surface with and without air injection, the
1500 iterations 共time steps兲 have been used for the conver-
corresponding values are plotted in Fig. 14. It is seen that the
gence. Approximately three hours of CPU time was required
numerically estimated data from the CFD computations match
for the simulation on an Intel Pentium 4, 1.4 GHz processor.
very well with experimentally measured data. However, the
The program was run on a Windows NT platform.
computed drag coefficient seems to be unaffected with air
The objective of this computational study has been to
injection. This deviation can be attributed to some unsteadi-
understand the flow field features around the blunt cone
ness in the flow field around the blunt cone with gas injec-
model flying at Mach 5.75 and gain insight into the experi-
tion, which could not be taken into account during the com-
mental results of drag and heating rates measured on this
putations while integrating pressure over the entire surface.
body reported in the previous sections. The simulations are
carried out only for 0° angle of incidence for the cases with
and without air injection. The numerically simulated surface VII. CONCLUSIONS
heat transfer rates over the blunt cone model with and with- The effectiveness of film cooling and its effect on the
out air injection are shown in Figs. 12 and 13. The Stanton aerodynamic drag is investigated experimentally in hyper-
number is computed from the heat-transfer data for both the sonic shock tunnel HST2 by injecting various coolant gases
cases 共with and without air injection兲 are compared with the
experimentally measured values of Stanton number. In order
to have an exact estimation of reduction in heating rates over

FIG. 14. Comparison of numerically computed reduction in surface heating


rates with the experimentally measured data over the surface of the 60°
FIG. 12. Numerically computed surface heat-transfer distribution 共W / m2兲 apex-angle blunt cone model at 0° angle of attack for H0 = 1.16 MJ/ kg with
for a 60° apex-angle blunt cone model flying at Mach 5.75. air injection.
036102-11 Film cooling effectiveness on a large angle Phys. Fluids 17, 036102 共2005兲

at the nose of the blunt cone model at free stream Mach spike-tipped bodies,” AIAA Pap. 2001–2464, 2001.
7
M. Viren, S. Saravanan, and K. P. J. Reddy, “Shock tunnel study of spiked
number of 5.75. Carbon dioxide, air, and helium are used as
aerodynamic bodies flying at hypersonic Mach number,” Shock Waves
coolant gases at 0° angle of attack. This study has shown that 12, 197 共2002兲.
8
lighter gases are more effective than heavier gases in reduc- D. Riggins, H. F. Nelson, and E. Johnson, “Blunt-body wave drag reduc-
ing the heat-transfer rates beyond the regions of stagnation tion using focused energy deposition,” AIAA J. 37, 460 共1999兲.
9
Y. C. Ganiev, V. P. Gordeev, A. V. Krasilnikov, V. I. Lagutin, V. N. Ot-
point. Also it is found that the cooling performance of mennikov, and A. V. Panasenko, “Aerodynamic drag reduction by plasma
heavier gases in the vicinity of the stagnation region is much and hot-gas injection,” J. Thermophys. Heat Transfer 14, 10 共2000兲.
10
better than that of lighter gases. Out of all the coolant gases R. Takaki and M. S. Liou, “Parametric study of heat release preceding a
used in the present study, helium is found to be the most blunt body in hypersonic flow,” AIAA J. 40, 501 共2002兲.
11
M. N. Kogan and M. A. Starodubtsev, “Reduction of peak heat fluxes by
effective coolant gas because of its overall performance in supplying heat to the free stream,” Fluid Dyn. 38, 118 共2003兲.
reducing surface heating rates at all gauge locations except in 12
M. Lopatoff, “Wing flow study of pressure drag reduction at transonic
the vicinity of stagnation point region. The effect of gas in- speed of projecting a jet of air from the nose of a prolate spheroid of
jection on aerodynamic characteristics, in particular, drag has fineness ratio 6,” NACA Report No. RM L51E09, 1951.
13
C. H. E. Warren, “An experimental investigation of the effect of ejecting a
also been studied. About 12%–25% increase in overall drag coolant gas at the nose of a bluff body,” J. Fluid Mech. 8, 400 共1960兲.
coefficient is observed for all the injected gases. 14
P. J. Finley, “The flow of a jet from a body opposing a supersonic free
stream,” J. Fluid Mech. 26, 337 共1966兲.
15
ACKNOWLEDGMENTS J. W. Keyes and J. N. Hefner, “Effect of forward-facing jets on aerody-
namic characteristics of blunt configurations at Mach 6,” J. Spacecr. Rock-
The support rendered by K. Sateesh, J. Ratan, K. Na- ets 4, 533 共1967兲.
16
B. Meyer, H. F. Nelson, and D. W. Riggins, “Hypersonic drag and heat-
gashetty, Gangadhar, Rajgopal, and Kiran Suryavamshi dur-
transfer reduction using a forward-facing jet,” J. Aircr. 38, 680 共2001兲.
ing the course of this work is sincerely acknowledged. 17
J. D. Anderson, Jr., Hypersonic and High Temperature Gas Dynamics,
Aeronautical and Aerospace Engineering 共McGraw-Hill, New York,
1
F. I. Peter and B. K. Donn, “High-speed aerodynamics of several blunt- 1989兲.
cone configurations,” J. Spacecr. Rockets 24, 127 共1987兲. 18
W. J. Cook and E. J. Felderman, “Reduction of data from thin film heat
2
D. A. Stewart and Y. K. Chen, “Hypersonic convective heat transfer over transfer gauge: A concise numerical technique,” AIAA J. 4, 561 共1966兲.
140° blunt cones in different gases,” J. Spacecr. Rockets 31, 735 共1994兲. 19
N. Sahoo, D. R. Mahapatra, G. Jagadeesh, S. Gopalkrishnan, and K. P. J.
3
R. H. Brian and N. P. John, “High enthalpy aero thermodynamics of a Reddy, “An accelerometer balance system for measurement of aerody-
Mars entry vehicle Part 1: Experimental results,” J. Spacecr. Rockets 34, namic force coefficients over blunt bodies in a hypersonic shock tunnel,”
449 共1997兲. Meas. Sci. Technol. 14, 260 共2003兲.
4 20
L. D. Huebner, A. M. Mitchell, and E. J. Boudreaux, “Experimental re- N. Sahoo, “Simultaneous measurement of aerodynamic forces and convec-
sults on the feasibility of an aerospike for hypersonic missiles,” AIAA tive surface heat transfer rates for large angle blunt cones in hypersonic
Pap. 95–0737, 1995. shock tunnel,” Ph.D. thesis, Indian Institute of Science, Bangalore, India,
5
M. Yamauchi, K. Fujii, and F. Higashino, “Numerical investigation of 2003.
21
supersonic flows around a spiked blunt body,” J. Spacecr. Rockets 32, 32 G. Jagadeesh, N. M. Reddy, K. Nagashetty, and K. P. J. Reddy, “Fore
共1995兲. body convective hypersonic heat transfer measurements over large angle
6
P. Gnemmi, J. Srulijes, K. Roussel, and K. Runne, “Flow field around blunt cones,” J. Spacecr. Rockets 37, 137 共2000兲.

You might also like