Spike-Nosed Bodies and Forward Injected Jets in Supersonic Flow

12 Pages·2002·4.1 MB·English

by  Gilinsky, M.

AIAA 02-3918
Spike-Nosed Bodies and Forward
Injected Jets in Supersonic Flow
M. Gilinsky, C. Washington
Hampton University, Hampton, VA 23668
I.M. Blankson
NASA Glenn Research Center, Cleveland, OH 441325
and
A.I. Shvets
Institute of Mechanics of Moscow State University
Moscow, 117192, Russia
38&A IAA/ASME/SAE/ASEE Joint Propulsion
Conference and Exhibit
7-10 July 2002/Indianapolis, Indiana
1
Spike-Nosed Bodies and Forward Injected Jets in Supersonic Flow
M. Gilinsky*,C . Washington**
Hampton University, USA
LM. Blankson***
NASA Glenn Research Center, USA
and
A.I. Shvets****
Institute of MechanicsMoscow State University, Russia
Abstract
The paper contains new numerical simulation and experimental test results of blunt body drag reduction using thin
spikes mounted m hnt of a body and one- or two-phase jets injected against a supersonic flow. Numezical
ix. simulations utiking the NASA CFUD code were conducted at the Hampton University Fluid Mechanics and
AcousticS Labomtory @ M U )a nd e cal tests were conducted using the facilities of the IM/MSU
Aermne~hania~n~d Gas Dynamics Laboratory. Previous resulfs were presented at the 37* AIAAIAsMEIsAE/
ASEE Joint Propulsion Conference [l]. Those red@w ere based on some experimental and numerical simulation
tests for super~onicfl ow a ~Spdike- n~sedOT Shell-nosed bodies, and Illrmerical ~irmllatiom e re conducted only
for a single spike-nosed or shell-nosed body at zero attack angle, M0I.n this paper, exp Itcstdtsof
'
gas, liquid and solid particle jet injection against a supersonic flow a ~pere sented. In addition, numerical simulation
results for supersonic flow around a mPlltiple spike-nosed body with non-zero attack angles and with a gas and solid
particle forward jet injection are iucluded. Aerodynamic coeBcients: drag, CB iirt, CL and longitudinal
momentum, Mz, obtained by numerical sixnuhion and experimental tests are compared and show good agmment
Z. Preview. A thin spike mounted in front of specified Mach and Reynolds numbers of the
a blunt body can be used to decrease the drag oncoming flow which include the regime of
and the heat transfer in a high oncoming flow stationary flow with drag reduction and the
velocity. By varying the length of the regimes of unsteady pulsating and fluctuating
protruding spike, the aerodynamic flows. A short review of different
characteristics of the apparatus can be experimental and numerical simulation
controlled; a thin straight spike is a approaches for the analysis of this problem
convenient and simple means of control. A was presented in the paper [l]. A detailed
summary of the early experimental and review of this area is in preparation at the
theoretical papers where flow past a body present time and will published as a NASA
with a spike was studied is contained in the GRC report soon.
well-lmown book by P.K. Chang (1970) and
in the Russian book [2]. 2. Experimental Test Results. Spike-Nosed
Six regimes of supersonic flow past a Blunt Bodies
cylinder with conical nose and a spike Experimental test results [l] were obtained
protruding h m it are distinguished for using an axisymmetic cylindrical model with
a flat forward part, i.e. butt-end, of diameter,
* Research Prof-, Senior Member AIAA D=80mm. Two types of models were tested.
** Assistanthf~r
a) with 1.needle (n=l) and b) with 5 needles
*** Senior Scientist, AIkA Associate Fellow
**** MSU Professol, IM/Msu Lab. Chieg Russia (n=5). Several such designs are shown in
2
Figure 1. In both cases, one central needle needle for two different instants of the
was mounted at the center of the circular pulsation flow regime.
kont flat part of radius, R4Omm
Several additional experimental tests were
For the second series of models, another 4 conducted to determine the influence of
needles were placed symmetrically around multi-needle ( ~ 5 a)nd shell applications to
the centd needle. They were placed at the the force coefficients. The needle numbers
angular interval of 90" and at the radial were: n=53 and n=103. These needles are
distance, d . 5 €2 Non-dimensional needle mounted Symmetrically along coincident
lengths, UD, were chosen based on circles located from the circle center on the
constructive opportunities of application, and radial distance^: R4.5-3511~1.l. R d th at
on hown refmced data in this field. These the butt-end circle diameter is D=8Omm. A
data show that maximum drag reduction drag coefficient comparison for 'all
using needles at supersonic speeds can be considered cases, n=l; 5; 53 and 103
achieved in the range of needle lengths, illustrates a preferable application for drag
UD=1.5-2.0, so that three needle lengths reduction of single needles.
were man&tured and test& D 1 . 5 , 1.0,
and 0.5. For this needle length range, the end The principal conclusions of these test d t s
of the needle at one limit, L/D =1.5, is for spikes were: multiple spikheedle
located farther h m t he butt-end than the application leads to decrease of drag
detached bow shock wave which occurred in reduction benefits by comparison with the
the case of flow without a needle at the body case of one central mounted needle at the
fbnt. For the other limit, uD= 0.5, the fbnt of a blunt body, as well as in
needle is located completely inside the longitudinal moment, but an increase in lift
compression shock layer, i.e. closer to the benefits. Dependence of the butt-end drag
butt-end than the corresponding detached coefficient, CD, of relative needle length,
bow shock wave. As a result of this analysis, L/D, and relative needle diameter, r/D, is
the following needle geometric parameters shown in Figure 2. Here D is the butt-end
were chosen: D/L = 0.05; 0.81; 0.1, i.e. for diameter; fiee stream Mach number, M=3,
the given model size, the needle diameters and angle of attack, a=Oo.
are: d=4; 6.5; and 8mm.
3. Numerid Simulation Test Resdts:
. .
Optical methods of flow vhahato~n were Spike-Nosed Blunt Bodies
employed for better understanding of this Numerical simulation tests for the 1-needle
phenomenon and for explanation of the model were conducted using the IM/MsU
aerodynamic charactdc changes. They Russian code, and the main results were also
include a traditional schlieren method and presented at the previous propulsion
high resolution video fitminp of the unsteady confmce [l]. This code is based on the
pulsation regimes. In Figure 1, the main Navier-Stokes equations for modeling of gas-
models tested in supersonic flow with Mach dynamic and chemical processes taking place
number, M=3, at zero angle of attack are in internal as well as external gas flows.
shown. From the left to the right, the first These results were obtained using a laminar
row: 1)-butt-end without needles; 2)-4) - 5 viscous gas model as well as the k-a
thin needles model of thiclmess, d/D=O.05, turbulence model. It was observed that the
and different needle lengths, uD-0.5,l.O and flow regime around the spike-nosed blht
1.5. In the second row: 5>7) -with a single body obtained in numerical simulation is
3
very sensitive to changes of the numerical Mach contours in the SymmetrvXZ-plane; b)
scheme parameters or applied models. For Mach contours in the cross section X=const
example, for the same conditions in inviscid at the butt-end h n tw all: c) a portion of grid
flow (Euler approximation) or for a laminar employed in this cross section; d) 3D grid for
viscous flow, numerical simulation results the domain between two symxnetry.phes.
show periodic oscillatory variation of the This grid contains 15 blocks. In this case, slip
flow. Such observations were found in [1 1 for boundary conditions were applied. Note that
the case: fiee stream Mach number, &=3, numerical simulation results are in good
UD=l, pressure, P, 4.14, temperature, T, agreement with experimental test results
=102"K, and Reynolds number, Re, obtainedinthisresearch.
=1.8x106. However, application of the k-a
turbulence model has led to stabilization of 4. Liquid Jet Injection from the Blunt
this process for the same outer boundary BO~FYr ont against a Supersonic FI~W
conditions in agreement with experimental
test results known h m o ur own and several 4.1 Technique of Experiments. The
other published papers. structure of interaction of liquid jets
exhausted towards to subsonic or supersonic
Numerical simulation tests of a 2D airflow has been considered. Two types of
supersonic flow for the 2-step configuration interaction: were analyzed: injection fiom a
have been conducted for different h est ream conical divergent ledge with an exposure
Mach numbers, M=1.5-4.0 and different angle, a=12Oo, installed at the blunt body
lengths of forward directed small height h n t and injection directly h m t he flat part
steps. This 2D problem is analogous to the of butt-end. In the latter case, the nozzle was
problem of supersonic flow around a 2- installed at the end of the cylindrical
needle model described above in the previous divergent tube of diameter, 0.1 D (8 mm),
section. Accurate solution of this problem and wall thickness varied from 0.2 to 0.5mm.
allows resolution of many obstacles This tube was strengthened at the center of
connected with the more complicated 3D- the flat part butt-end. The tested models had
problem solution for the flow in the 5- the following geometric characteristics: butt-
needles model. end diameter, D =80 111111, and nozzle exit
diameter was varied discretely d = 0,34;
, ,... New numerical simulation results were 0,64;1 ,48; 2,8; 3; 4; and 5 mm. It is known
+..;-:., ,7.-, :
.. .-. obtained for the 3D problem using the NASA that liquid jet injection from a nozzle can be
CFL3d code [3] for the case of zero attack accompanied by CaVitatiOn. To avoid
angle, a=O, as well as non-zero attack angle cavitation, the nozzle had a smooth
in the range of: -15'--<+15'. Some longitudinal shape of the input portion
infomation for supersonic flow numerical extended by a cylindrical tube of various
.....~.... ., ... -, . simulation with h e s tream Mach number, lengths fiom 80 to 150 nozzle diameters.
. ..
,..,. . 1<. ... & around a 5-needle nosed buttend at zero
.'.I.. .%(.
'.; .., :. angle of attack was presented at the Experimental tests were conducted in the
.- . conference [4]. The NASA CFL3D code was wind tunnel, A-7, and in the special jet stand,
.:1' .:
:. employed, and Navier-Stokes Equation both located at the Institute of Mechanics of
:.
i.
. . (NSE) based simulations were conducted Moscow State University. The test chamber
with slip and no-slip boundary conditions at of the wind tuxmel has square cross section,
the solid walls. An example of such results is 0.6 x 0.6m. Free stream Mach number, My
illustrated in Figure 3. Presented thm are: a) varies from M = 0.3 to M = 3, and Reynolds
4
number, Re, varies h m 1x106 to 3x10’ droplets and bubbles. At the blunt jet portion,
calculated using fkee stream parametem and the bow shock wave, 7, is fomed. Also, a
the characteristic length, l=lm During separation zone 3 is formed between the
experimental tests, schlieren photos of the conical and cylindrical liquid jet portions. In
flow field were made and the main fiee the separation zone, static pressure was
stream and injected liquid jet parameters measured using the special pressure
were measured. Also, the pressure transducer, 2.
distriiution along the butt-end radial
direction was measured as well as at the The structural picture of liquid jet interaction
nozzle exit and inside of the separation zone. with the main air stream depends on the
The average velocity value of the liguid jet nozzle shape and relative jet diameter. If the
exhausted was calculated ftom the liquid jet penetration depth is much more than the
mass flow rate. The latter was determined nozzle diameter, the liquid jet interacthg
using pressure difference at two sizes of the with the air stream is close to stationary.
special measured washers. The relative root- Photos of such interaction are shown above
mean-square error of velocity measurements in Figure 4a In the presented experimental
of a liquid was estimated as f 0.05. tests, fluctuations of jet penetration depth
were observed for some regimes. Therefore
4.2 Experimental Test Results: the visualized shock waves are a little bit dim
Geometrical Characteristics. In Figure because of the relatively large exposure
4a,b, a shlieren picture of a liquid jet -0.01 employed.
interacting with a supersonic airflow is
shown for two cases. In a) the jet exhausted It was observed that in the case of liquid jet
h m a conic body against the flow with zero injection a stationary regime of interaction
angle of attack b) the jet exhausted h m a takes place only for relatively small mass
flat forward part of the butt-end with zero flow rate of the injected jet. h this case, the
angle of attack. Dependence of butt-end drag jet does not penetrate the bow shock wave
coefficient, CD,v s the relative specific mass formed at the blunt body without injection
flow rate of a liquid jet, G=pjUj /paym and (Figure 4a). Othexwise, an unsteady
the relative specific impulse KTjUj /p.JJm2 oscillatory regime takes place (Figure 4b).
are shown in Figure 4c. The same effect was observed in the case of
solid needle application noted in [l].
Four potential schemes of supersonic airflow However, for liquid jets, this effect is more
interaction with liquid jet exhausted fkom a sensitive to the boundary conditions
conical nozzle and butt-end were observed in producing strong interaction of two
the experimental test series. These schemes deformation media-a gas and liquid. At large
are illustrated in Figure 5. For moderate liquid mass flow rates, there are intensive
speeds of the liquid jet exhausted fi-om the fluctuations of the bow shock wave. Shock
conical nozzle (Figure 5a,b), the jet has wave fluctuations are a cause of liquid mass
almost a cylindrical-shaped portion, 4, and flow rate fluctuations at the body h n t etc.
then the jet sharply transforms to almost The schematic picture of such interaction is
spherical shape, 6, as shown in Figure 5a or illustrated in Figure 4d. Fo~mation of
almost mushroom-shape as shown in Figure protuberances emitted through an e x t dj et
5b. Further downstream, these jets break up confirxns that the main jet portion
to multiple droplets. This area of the flow experienced intensive cross fluctuations
represents a complicated mixture of liquid connected with fluctuations in the jet mass
5
flow rate. In the non-stationary interaction angle, cp and external diameter D, of the
regimes, static pressure along the butt-end returnable jet in the plane of the nozzle exit.
-
face was essentially reduced (on 70 % 80 Results of measurements are shown Figure 7,
%, relative to pressure behind the n o d where dependence on angle, cp, and relative
portion of the bow shock wave, a distinctive returnable jet diameter, & =DcId , are shown.
analogue with a hydraulic jump. Practically, these values do not depend on
fiee'stream Mach number in the region: M =
Geometric characteristics of a liquid jet 0, 3 - 1, 2; 2; 3. They are dehed by the
exhausted fiom a conical nozzle and butt-end
parameter as well as by the relative jet
obtained in experimental tests are presented
penetration depth, E. In this figure, points 1-3
in Figure 6. In this figure, dependence of
are for &*E), d=3,4,5mm, and 4-6 are for
relative jet penetration depth, E, vs root of a cp= q~)d,= 3,4,5mm. Experimental data are
&
relative jet impulse is shown. H~XEar e
approximated Using the dependence cp = 5 / 3 ~
t;- l/r, where 1 is jet penetration depth and r = / E + 4, d c= 3.6 E In
nozzle radius. Points 1-3 correspond to the
jets with diameters: d = 3,4,5 mm, flowing
Based on the conducted experiments, some
fiom a conical nozzle into the airflow with
iic dependence between the diameter of the
Mach number, M=3. The point 4 corresponds
return jet and diameter of the butt-end was
to the jet flowing fiom butt-end hole with
observed. So, for example, it is possible to
d=3mm. One can see that, in the case of a
predict the transition h m s tatio- to non-
stationary regime (KC I) when the jet does
stationary regimes of interaction. For
not interaction with a shock wave, the jet
relatively mall jet penetration depths, which
penetration depth surpasses the approPriate
do not surpass the shock layer thickness
depth for a fi-ee jet. In contrary case (K>1),
(without a jet) at the butt-end, isolated jet
when interaction between shock waves takes
stationary regimes can be predicted. With
place, jet penetration depth becomes less than
increase of jet speed accompanying jet
the corresponding value for the b e j et
penetration, whole shock layer, non-
flowing to a still space. However, there is
stationary modes of interaction can be
monotone behavior of jet penetration depth
predicted. The experimental data show that
vs K even with change of the interaction
non-stationary regimes take place until the
regime. Note that, in regimes of non- characteristic size satisfies the inequality, &
stationary interaction, the fastest destruction 4, where 4 =DJd. A further increase of
of the liquid jet takes place in the vicinity of
speed leads to the stationary interaction
the blunt body.
, regime, in which the returnable jet does not
reach the h n t face platform of the blunt
Liquid jets injected h m t he conical forward
body.
part of the body interact with a contrary gas
stream by formation of a forward thickening
4.3 Blunt body drag coefficient variation
drop-shaped structure. This occurs for a wide
with liquid jet injection. VariouS methods
range of characteristic parameters with ring-
were employed for the purpose of blunted
shaped returnable gas-fluid jet downstream at
bodies drag reduction. These methods hvor
the body periphery. The external boundaq of
the formation of separation zones at the blunt
the returnable ring jet can be described as a
body h e . As a result, a detached shock
conical surface with good accuracy. Thus,
wave transforms to a wave similar to the
the form of the returnable jet can be
shock wave arising at the conical or ogival-
characterized by two parameters: conical
shaped body in a contrary supersonic flow.
6
With the purpose of formation of such shown that the value of reduction of
separation zones, gas or liquid jet injection - aerodynamic drag of the model is defined not
may be applied as well as firm particle only by the mass flow rate of the liquid jet,
emission alongside with widely investigated but also by jet diameter. For each nozzle size,
methods of solid needle application there is a size relative to the mass flow rate at
which a drag coefficient, CD is minimal. A
A special construction was mazlllfactured for further increase of the jet mass flow rate
more accurate drag measurements. The leads to growth in the drag coefficient, CD.
model was a cylindrical body, with a flat
forward part. A flexible hose was attached to For solid particle injection, a nozzle of
.-
the body with the device for measuring the diameter, DJD=O.l, was manufactured.
reduction of force. Calibration of the Appropriate schlieren photos of tfie flows
tension-balance was carried out at various have shown that injected particles smash and
pressures of the flow system. On the forward destroy the bow shock waves detached at the
part of the cylinder, a rifled hole was made butt-end model. It is observed that particle
for installation of nozzles of various injection leads to butt-end drag reduction,
diameters. then the drag coefficient increases.
5 x
Comparison of these experimental test results
Experimental tests were carried out with a with numerical simulation results will be
cylindrical model of small len@ & = D, D = presented in the next section.
80 mm.) with both flat and hemispherical
head parts. The liquid flowed from a 5. Numerical simulation test results:
cylindrical nozzle. The internal nozzle Solid Particles
diameters were 4 and 2 mm and the nozzle An existing discrete-trajectory approach
length was equal to 8 mm. The first tests using the two-phase version of the order
have shown that water jet injection fiom the Godunov scheme for numerical simulation of
nozzle towards a supersonic stream with unsteady and steady 2D and axisymtnebk
Mach number, M = 3.0, leads to hezing of inviscid flows [5] was improved. The
the liquid and transferring to ice at some improvement lay in two directions: 1) use of
distance h m the nozzle. Therefore in these “almost adaptive” ds for gas phase and 2)
experiments, an anti-fieezing liquid was application of a 28rder numerical scheme
used, represented by the mixture of ethylene- for gas phase and calculation of particle
glycol (30-50 %) and water (70 - 50 %). trajectories and their parameters along these
Density of this liquid, y = 0.8 - 1.05. trajectories. By such improvements, more
exact numerical results were obtained for the
The efficiency of influence of liquid jet problem of supersonic flow at the blunt body
injection was examined for kze stream Mach with solid particle injection against a flow.
number, M=3. In Figure 8, the dependence of Several numerical simulations were
the drag coefficient, CD,o f a cylindrical body conducted for the cases tested exphentally
with flat and spherical forward parts vs in the IM/MSU aerodynamic wind tunnel A-
& 7. At the present time, we are developing the
parameter of the liquid jet is illustrated.
discrete-trajectory method for the NASA
In both cases, this jet was exhausted fkom a
CFUd code for numerical simulation of two-
nozzle with the same relative diameter, d/D.
phase flows based on the NSE approXimation
Figure 6 illustrates that jet injection leads to
for gas phase.
monotonic body drag reduction with increase
&.
of parameter These results have also
7
This method has been used to determine retum flow and are moved as a thin layer by
sensitivity of the wave drag of a cylinder and the incident flow. This layer seems to be
face (with the radius r* to several parameters. describable by the term “sheet” [6].
In particular, a) the mass flow rate E of the
particles in the injected mixture, b) the Figure 10 contains the hction C, (E) for
particle size a,,, and c) the particle slip ratio the case with particles of size a,,=l plan
Ip= up,iu, (with respect to the carrier gas (Enel). The line 6 represents the distance xp
velocity) at the injector exit section (of radius (E) attained by the paaicle in their back-
penetration along the symmetry axis. Here,
Yw).
this distance is roughly the same as the
The flow conditions are as follows: maxjmum separation of the shock wave and
k, = k, =1.4:M, =3; M, =1.05; as the length of the circulation return flow-
region. The rate of the decrease in Cxoi s the
y, /re = 0.l;p , /p, = 0.67; ,T /T, = 1;
greatest, as great 12%, at ~=0.05, 23%.
Tp /T, = 1; /3 = 0.014; y = 4.8.10”;
Thus, we conclude that even a minor amount
Re. = 1.48.10’ ( ap = I-). of particles leads to a noticeable reduction in
the wave drag. This same figure contains the
Parameters of the particles varied in the
following ranges: 4 ukm, < 1, num&cdy obtained function Go (E) for a
0% O<E
body with a needle of length lqp( dot-md-
O<L,$l, where L=UJLL. Used as
dash h e 4, for an “effective” needle).
reference values are he-stream parameters
Injecting a heterogeneous jet is seen in curve
for the dagmtion point and the cylinder
5 for a non-monotonous function C,(+)
radius. In the present analysis, two relative
which attains its maximum at a 4 . 5 pkm
values are essential, the jet gas mass flow
(E,, = 0.3). The dashed line 3 is for
rate G, = pwu, /p,v, = 2.84; and the
experiments involving bodies with needles of
momentum Kw= pu, 2 ,/p ,v2- =1.49; length 1%. As long as + < 1.5 mm,
! Non-dimensional parameters, J3 and y convergence ofresults is satisf&ory. When
characterize force interaction and heat wl.5 pkm, the practical back-penetration
transfer between single particle and carrying distance (curve 2) and the circulation return
gas. Reynolds number, Re+, is calculated flow region length monotonously increase in
using gas critical (sound) parameters and ap, and GO continues to decrease. However,
I
characteristic len& 1 p . the experiment perfomed for needle lengths
reveals an abrupt increase in Cxod ue to
Figure 9 demonstrates location of a bow the flow separation onset being shifted h m
shock wave depending on the particle size, the needle tip toward the body (the so-called
ap, and the mass flow rate of particles, E. It is delayed separation). Computation does not
readily seen that particles introduced in the predict this effect.
jet do enhance the distance between the
shock wave and the body, the influence of The numerical result leads to the conclusion
the particle size being more notable than that that large particles (ap> ap+) are prefdle to
of the particle mass flow rate. Particles small ones. However, this is not so in the
represented by the shaded area in the figure general case. The realiv is that particles at
(at 64.3, +=l mm) move along the the nozzle exit section are slower than the
symmetry axis to a Surface where horizontal carrier gas. The minimum value of GO on the
components of their velocities become zero. curve 1 in fig.12 is at L=l when the
Mer being turned, they are in the circulation particles fly the maximum distance ( m e2 )
8
and decelerate least near the nozzle. The to thin spikes mounted in h n t o f a body and
dashed line 3 corresponds to the “effdve one or two-phase jets injected into a
needle”, l=xp When L, <OS, the coefficient supersonic flow have been presented. The
C,, is greater than Cox,, so that a pure gas jet experimental and numerical results are in
is more advantageous. With due account for good agreement and show that sigzllficant
the reaction force, the effective drag , h g r eduction can be acbieved by proper
coefficient choice of spike and flow parameters.
C, = F =cxo+{ $)cw(l+ELw) Beneficial parameter ranges are described.
?n;zp~v21D1 2
7. Acknowledgements
We would like to acknowledge the NASA
is in notable excess over the C, only if the
Glenn Research Center, especially Mi.
particle mass flow rate is great enough. The
Robert C. Hendricks for hitiation, interest
functions &= C, (E) and C,= C, ( E ) are
and support of our research. This research
plotted as curves 4 and 5, respectively.
was partially conducted under the NASA
grant NAG-3-2422 and under the supporting
Calculation shows that the extent of decrease
CRDF grant RE1-2068. We would like to
in drag depends on the distance fiom the
thanks Dr. Jay C. Hardin for his attention,
body along the symmetry axis to the point
interest in our research and usefid
where the particles stop (i.e. depends on the
suggestions.
“tW~-phase”c one apex angle). One of the
8. References
likely options for shifting this point is to
1. Gilinsky, M., Bhkson, LM., Sakharov V.I.,
supply the heterogeneous jet through a
Shvets, kL, 2001, Shock Waves Mitigation at
needle nozzle of length 1. How C, depends
Bluut Bodies using Needles and Shells against a
of &=l/r* is represented in Figure 11 by
supersonic Flow, 37& AIAA/AsmsAE/ASEE
curves 6 and 7 for E=O and E 4.3, Joint Propulsion Conference, 8-11 July, 2001,
+
respectively. The ‘heedle jef‘ combination Salt Lake City, UT, USA, 12p.
definitely causes C, to further decrease. For 2. Shvets AI., Aerodynamics of Supersonic
example, if 1,=2, the drag is reduced by a Shapes, 1987, Moscow State UniverSity Press,
factor of approximately two. No oscillation Moscow.
in a flow with the mentioned parameter is 3. Krist, S.L., Biedron, ET., and Rumsey, C.L.,
predicted. 1996, CFL3D User‘s Manual (Version 5.0),
NASA Langley Research Center, 3 1l p.
4. Oilinsky, M., Patel, K., Alexander, C., at al.,
Finally, in Figure 12, a comparison of
2001, Numerical simulation of One- and Two-
numerical simulation and experimental data
Phase Flows m Propulsion Systems,
is shown for Werent sizes of mono-
HBCUdOMUs Research Conference Agenda
dispersion quartz particle jets interacting and Abstracts, OAI[MASA GRC, Apnl 17-18,
With the air flow with Mach number, M=3. 2001, p.23.
The experimental tests were conducted at the 5. Gdmslq, M.M., et al., 1994, Supersonic
IM/MSU supmonk wind tunnel, A-7. This Gasdispu-sional Jets: Models and Applications,
comparison demonstrates a satisfactory AIAA Paper 940135,32nd Aerospace Sciences
agreement between the theory and Meeting and Exhibit, Jan. 10-13, 1994Ren0,
N v .
experiment.
6. Krayko, A.N., 1982, To two-liquid model of a
gas flow with dispersion particles, Applied
6. Conclusion
Mathematics and Mechanics, V.6, No 1, pp. 96-
New numerical simulation and experimental
106
test results on blunt body drag reduction due
9
I I
'
OD I I
on as ID
Hg. 1 Interaction of supersonic air flow with spike-nosed Fig3 Drag coefficient, CD, v s relative needle length,
blunt body having 1 and 5 needles. Experimental test results. WD, where D-butt-end diameter, for single needle, s l,
Shlieren photos. Free Stream Mach number, M=3, angle of and 5 needles, n=5. Two experimental test results are
attack, GO". shown for Camparison: l-obtained m Bestall experiments,
and &obtained m W S Utr ansonic wind tunnel, A-8.
pig3 Mach 3 supersonic flow numerical simulation results around 5-needles nosed butted at zero attack angle. NASA
CFL3D code. Navier-Stokes based simulation with slip boundary conditions. a) Mach contours m the symmetry XZplane; b)
Mach contours in the cross d o nX =const at the butt-end front wall; k) portion of employed grid in this cross section; d) 3D