Академический Документы
Профессиональный Документы
Культура Документы
This paper presents numerical simulation results of the primary atomization of a turbulent liquid jet injected into a gaseous crossflow. Simulations are performed using the
balanced force refined level set grid method. The phase interface during the initial
breakup phase is tracked by a level set method on a separate refined grid. A balanced
force finite volume algorithm together with an interface projected curvature evaluation is
used to ensure the stable and accurate treatment of surface tension forces even on small
scales. Broken off, small scale nearly spherical drops are transferred into a Lagrangian
point particle description allowing for full two-way coupling and continued secondary
atomization. The numerical method is applied to the simulation of the primary atomization region of a turbulent liquid jet q 6.6, We 330, Re 14, 000 injected into a
gaseous crossflow Re 570, 000, analyzed experimentally by Brown and McDonell
(2006, Near Field Behavior of a Liquid Jet in a Crossflow, ILASS Americas, 19th
Annual Conference on Liquid Atomization and Spray Systems). The simulations take the
actual geometry of the injector into account. Grid converged simulation results of the jet
penetration agree well with experimentally obtained correlations. Both column/bag
breakup and shear/ligament breakup modes can be observed on the liquid jet. A grid
refinement study shows that on the finest employed grids (flow solver 64 points per
injector diameter, level set solver 128 points per injector diameter), grid converged drop
sizes are achieved for drops as small as one-hundredth the size of the injector diameter.
DOI: 10.1115/1.4000148
Introduction
The atomization of turbulent liquid jets injected into fast moving subsonic gaseous crossflows is an important application for
example in gas turbines, ramjets, and augmentors. It is a highly
complex process that has been extensively studied experimentally
over the past decades. Early studies of the atomization of nonturbulent liquid jets in crossflows have recently been reviewed in
Ref. 1, whereas newer studies of this case include the work
reported in Refs. 18. Most experimental work has focused on
jet penetration, including both the column trajectory and the resulting spray penetration. Some recent jet penetration correlations
can be found in Refs. 4,8. Unlike turbulent liquid jets atomized
either by injection into still air or by a coflowing fast moving gas
stream, experimental access to the primary atomization region of
liquid jets injected into crossflows is relatively straightforward.
This is due to the fact that at least the windward side of the liquid
jet is typically not surrounded by many drops that could otherwise
conceal the jet phase interface geometry. This is not, however,
typically the case for the leeward side. Still, experimental studies
of nonturbulent liquid jets injected into subsonic crossflows have
concluded that depending on the momentum flux ratio and the
crossflow Weber number, essentially two different breakup modes
can be observed 8. For low momentum flux ratios and crossflow
Weber numbers, the liquid jet breaks up as a whole some distance
downstream of the injector. For high momentum flux ratios and
crossflow Weber numbers, on the other hand, surface breakup or
stripping at the sides of the liquid jets is observed prior to column
breakup 8. Unlike nonturbulent jets, turbulent liquid jets do not
exhibit this clear separation of breakup modes 9. Stripping at the
sides of the liquid column occurs even at low Weber numbers,
Manuscript received May 5, 2009; final manuscript received May 6, 2009; published online March 30, 2010. Review conducted by Dilip R. Ballal.
ghost cells
boundary cells
active
G-grid cells
hG
Governing Equations
The equations governing the motion of an unsteady incompressible immiscible two-fluid system are the NavierStokes
equations
u
1
1
1
+ u u = p + u + Tu + T
t
G
+ u G = 0
t
x = HG1 + 1 HG2
x = HG1 + 1 HG2
x x f = GG
G
,
G
=n
Numerical Methods
In the RLSG method, all level set-related equations are evaluated on a separate equidistant Cartesian grid using a dual-narrow
band methodology for efficiency. This so-called G-grid is overlaid
onto the flow solver grid on which the NavierStokes equations
are solved and can be independently refined, providing high resolution of the tracked phase interface geometry, see Fig. 1. Details
of the method, i.e., narrow band generation, level set transport,
re-initialization, curvature evaluation, and its performance compared with other interface tracking methods in generic advection
test cases, can be found in Ref. 18.
In the current simulations, refinement of the G-grid is limited to
a factor two in each spatial direction as compared with the flow
solver grid. While higher refinements are feasible from an efficiency standpoint, they would require the use of a subflow solver
grid model to correctly capture the otherwise nonresolved phase
interface dynamics on the G-grid scale 19. Instead, here, the
refined G-grid simply serves to increase the accuracy of the level
set-based phase interface tracking scheme.
In the NavierStokes equations, the position of the phase interface influences two different terms. The first term is due to Eqs.
4 and 5, since HG is a function of the position of the phase
interface. For finite volume formulations, the volume fraction cv
of a control volume is defined as
cv = 1/Vcv
HGdV
Vcv
with Vcv the volume of the control volume cv. In the RLSG
method, the above integral is evaluated using the high-resolution
G-grid, see Refs. 18,20 for a detailed description. Then both
control volume density and viscosity are simply
cv = cv1 + 1 cv2
10
cv = cv1 + 1 cv2
11
12
f =
cvcv + nbrnbr
cv + nbr
13
where nbr is the control volume sharing the face with cv, and
cv =
1 : 0 cv 1
0 : otherwise
14
15
3
VD
4
1/3
16
Employed Solvers
In this work we use the flow solver CDP/CHARLES that solves the
incompressible two-phase NavierStokes equations on unstrucJournal of Engineering for Gas Turbines and Power
Experiment
Simulation
1.3
1.225
1000
120.4
10.83
1.82 105
1.0 103
0.07
6.6
330
2178
5.7 105
14,079
1.3
1.225
12.25
120.4
97.84
1.82 105
1.11 104
0.07
6.6
330
2178
5.7 105
14,079
tured grids using the finite volume balanced force algorithm 18.
In the single phase regions, the employed scheme conserves the
kinetic energy discretely. Turbulence in the single phase regions
of the flow is modeled using a dynamic Smagorinsky large eddy
simulations LES model, however, none of the terms arising from
filtering the phase interface are modeled. The approach instead
relies on resolving all relevant scales at the phase interface, thus,
ideally reverting to a direct numerical simulation DNS at the
phase interface. As such, the current simulation approach is a
combination of LES in the single phase regions and ideally DNS
at the phase interface. A Lagrangian particle/parcel technique is
employed to model the small scale liquid drops of the atomizing
liquid spray 26.
The liquid/gas phase interface during primary atomization is
tracked by the interface tracking software LIT, using the RLSG
method 18. The solver uses a fifth-order WENO scheme 27 in
conjunction with a third order TVD RungeKutta time discretization 28. The phase interfaces curvature on the level set grid is
evaluated using a second-order accurate interface projection
method 18.
The flow solver CDP/CHARLES and the interface tracking software LIT are coupled using the parallel multicode coupling library
CHIMPS 18,29. In order to couple the level set equation to the
NavierStokes equation, u in Eq. 3 is calculated from the flow
solver velocity by trilinear interpolation. To achieve overall
second-order accuracy in time, the level set equation is solved
staggered in time with respect to the NavierStokes equations.
y+: slip
wall
x-: channel
inflow
z+: sl
ip
inje
cto
y
z
z-: slip
wall
wall
x+: channel
outflow
y-: w
all
Fig. 2 Computational domain and boundary conditions left and mesh detail near the
injector right
Fig. 4 Instantaneous injector exit plane velocity distributions. From left to right: crossflow direction, axial direction, and
transverse direction.
c01
c12
c23
D / 16
D / 32
8.7
3.5
D / 32
D / 64
21
13
D / 64
D / 128
110
66
x
y
= 1.37 q
D
D
1/2
17
x
y
= 2.63q0.442
D
D
0.39
We0.088
l
H2O
0.027
18
50
50
50
40
40
40
30
30
30
20
20
20
10
10
10
0
0
10
20
30
40
50
60
0
0
10
time
20
30
40
50
60
10
time
20
30
40
50
60
time
Fig. 5 Temporal evolution of tracked liquid mass for different grid resolutions: case c01 left, c12 center, and c23 right
Fig. 7 Side view snapshots of jet in crossflow atomization at t = 5, 10, 15, 20, 25, and 28.1 time units top to
bottom; case c01 left, c12 center, and c23 right
Fig. 8 Front view snapshots of jet in crossflow atomization at t = 5, 10, 15, 20, 25, and 28.1 time units top to
bottom; case c01 left, c12 center, and c23 right
tion. Note that these drops can continue to atomize via secondary
atomization mechanisms further downstream; however, within the
computational domain, only a limited number of these secondary
atomization events occur.
The drop size distributions in Fig. 10 were generated by binJUNE 2010, Vol. 132 / 061506-7
Fig. 9 Top view snapshots of jet in crossflow atomization at t = 5, 10, 15, 20, 25, and 28.1 time units top to
bottom; case c01 left, c12 center, and c23 right
ning the data into 20 bins of equal size in terms of logD. The
approximate total number of drops used to calculate each distribution was 40,000 c01, 176,000 c12, and 193,000 c23. To be
able to compare the three probability density functions pdfs, the
061506-8 / Vol. 132, JUNE 2010
coarser grid pdfs were normalized using a single bin of the finest
grid pdf. Also shown in Fig. 10 is a log-normal fit to the larger
drop sizes of the finest grid results. All three cases show a similar
behavior. Larger drop sizes collapse well to the log-normal fit,
Transactions of the ASME
c23
c12
c01
d
Fig. 10 Grid convergence study of pfd of drop diameters d
generated directly by primary atomization: case c01 open
circles, c12 open triangles, and c23 solid squares. Solid
line is log-normal fit to large drop sizes on finest grid.
however, for each grid there is a distinct departure point from the
fit from which point on smaller drop sizes do not collapse and fail
to match the log-normal fit. This departure point d p decreases in
value proportional to the employed grid size, with d p = 1.6xG for
the coarse and medium grids and d p = 1.3xG for the fine grid. The
key result of Fig. 10 is the fact that larger drops, i.e., those resolved by at least 1.6 G-grid cells, collapse to a single pdf, in this
case a log-normal distribution independent of the employed grid
resolution. These drop sizes can thus be considered grid independent. Drops smaller than about d = 1.6xG, however, show a
strong grid dependency. The reason for this behavior is the following: fixed grid methods used to track interfaces, like the level
set method used here, have an inherent topology change length
scale that is proportional to the local grid size. As soon as two
front segments enter the same grid cell, a topology change event is
automatically triggered. The exact moment of breakup is thus always a function of the employed grid size. However, for larger
drops, the error introduced by missing the exact moment of
breakup due to the inherent grid size dependency is small compared with the generated drop size. Thus larger drop sizes are grid
independent, as demonstrated in Fig. 10. For smaller drops, the
exact moment of breakup is however the dominant source of error; thus, smaller drops are dominated by the employed grid resolution. The cut-off appears to occur at drops resolved by about
1.6 G-grid cells. This number is surprisingly low and appears to
decrease with increasing grid resolution. This seems to indicate
that the numerically induced topology change starts to mimic the
physical breakup mechanism on small scales.
10
Conclusion
Detailed simulation results of the primary atomization of a turbulent liquid jet injected into a subsonic gaseous crossflow previously analyzed experimentally by Brown and McDonell 2 have
been presented. The simulations match all key nondimensional
numbers, except for the density ratio, which is artificially reduced
to study its impact on jet penetration and atomization.
A preliminary qualitative analysis of the simulation results indicates that breakup of the liquid jet occurs via two main simultaneous atomization mechanisms. In the first, instability waves on
the liquid column, predominantly visible on the windward side,
generate role-ups and continue to grow along the jet axis until
they form baglike structures that rupture resulting in a broad range
of drop sizes, not unlike the column-breakup mode. It is speculated that this instability mode is due to a KelvinHelmholtz instability. In the second mode, corrugations on the liquid jet surface
are stretched out into ligaments at the sides of the liquid jet near
Journal of Engineering for Gas Turbines and Power
the injector exit. The stretched out ligaments then break up and
form a range of drop sizes. At the current stage, the above description of the breakup mechanisms is speculative and still requires a
detailed, yet to be performed, quantitative study of the generated
time dependent simulation data.
Grid independent predictions of the jet penetration in agreement
with experimental observations can be achieved on relatively
coarse grids D / x = 16. Moreover, the large scale instability
mode conjectured to be responsible for the first atomization
mechanism is sufficiently resolved even on these coarse grids.
Analysis of the temporal evolution of the pre-atomization tracked
mass indicates that it is important to resolve turbulent eddies at
least up to the maximum flow solver resolution considered
D / x = 64, since these turbulent eddies can generate significant
surface corrugations to initiate the second atomization mechanism. The grid independent prediction of atomized drop sizes requires significant computational resources. Grid independent drop
sizes on the finest grids have been achieved for drops as small as
13 m, or one-hundredth of the injector exit diameter.
While the reduced density ratio analyzed in this paper appears
to have no significant impact on the jet penetration determined
from averaged side views of the jet geometry, it is conceivable
that the atomization mechanism and resulting drop sizes are a
function of the chosen density ratio. Currently ongoing simulations employing higher density ratios will focus on this question.
Acknowledgment
This work was supported in part by CASCADE Technologies
Inc. under the NavAir SBIR Project No. N07-046. The author
would like to thank S. Hajiloo and Y. Khalighi for valuable help in
generating the fluid solver meshes, and F. Ham for many valuable
discussions concerning the flow solver CDP/CHARLES.
Nomenclature
d
dp
D
G
H
n
p
q
r
Re
u
t
T
V
VD
We
xf
x
x
xG
y
Greek Symbols
Subscripts
drop diameter
minimum grid independent drop diameter
injector diameter
level set scalar
Heaviside function
phase interface normal vector
pressure
momentum flux ratio
radius
Reynolds number
velocity
time
surface tension force
volume
drop volume
Weber number
location of phase interface
crossflow direction
flow solver grid size
level set grid size
jet direction
Dirac delta function
phase interface curvature
dynamic viscosity
volume fraction
density
surface tension coefficient
cv control volume
f control volume face
j jet
References
1 Aalburg, C., van Leer, B., Faeth, G. M., and Sallam, K. A., 2005, Properties
of Nonturbulent Round Liquid Jets in Uniform Gaseous Cross Flows, Atomization Sprays, 153, pp. 271294.
2 Brown, C. T., and McDonell, V. G., 2006, Near Field Behavior of a Liquid Jet
in a Crossflow, ILASS Americas 19th Annual Conference on Liquid Atomization and Spray Systems.
3 Brown, C. T., Mondragon, U. M., and McDonell, V. G., 2007, Investigation
of the Effect of Injector Discharge Coefficient on Penetration of a Plain Liquid
Jet Into a Subsonic Crossflow, ILASS Americas 20th Annual Conference on
Liquid Atomization and Spray Systems, ILASS Americas, Chicago, IL.
4 Stenzler, J. N., Lee, J. G., Santavicca, D. A., and Lee, W., 2006, Penetration
of Liquid Jets in a Cross-Flow, Atomization Sprays, 16, pp. 887906.
5 Mazallon, J., Dai, Z., and Faeth, G. M., 1999, Primary Breakup of Nonturbulent Round Liquid Jets in Gas Crossflows, Atomization Sprays, 93, pp.
291312.
6 Sallam, K. A., Aalburg, C., and Faeth, G. M., 2004, Breakup of Round
Nonturbulent Liquid Jets in Gaseous Crossflow, AIAA J., 4212, pp. 2529
2540.
7 Fuller, R. P., Wu, P.-K., Kirkendall, K. A., and Nejad, A. S., 2000, Effects of
Injection Angle on Atomization of Liquid Jets in Transverse Ai, AIAA J.,
381, pp. 6472.
8 Wu, P. K., Kirkendall, K. A., Fuller, R. P., and Nejad, A. S., 1997, Breakup
Processes of Liquid Jets in Subsonic Crossflows, J. Propul. Power, 131, pp.
6473.
9 Lee, K., Aalburg, C., Diez, F. J., Faeth, G. M., and Sallam, K. A., 2007,
Primary Breakup of Turbulent Round Liquid Jets in Uniform Crossflows,
AIAA J., 458, pp. 19071916.
10 Madabhushi, R. K., 2003, A Model for Numerical Simulation of Breakup of
a Liquid Jet in Crossflow, Atomization Sprays, 134, pp. 413424.
11 Arienti, M., Madabhusi, R. K., Slooten, P. R. V., and Soteriou, M. C., 2005,
Numerical Simulation of Liquid Jet Characteristics in a Gaseous Crossflow,
ILASS Americas 18th Annual Conference on Liquid Atomization and Spray
Systems.
12 de Villiers, E., Gosman, A. D., and Weller, H. G., 2004, Large Eddy Simulation of Primary Diesel Spray Atomization, SAE Technical Report No. 200401-0100.
13 Menard, T., Beau, P. A., Tanguy, S., Demoulin, F. X., and Berlemont, A.,
2006, Primary Break Up Modeling, Part A: Dns, a Tool to Explore Primary
Break Up, ICLASS-2006, pp. 17, Paper No. ICLASS06-034.
14 Menard, T., Tanguy, S., and Berlemont, A., 2007, Coupling Level Set/Vof/
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Ghost Fluid Methods: Validation and Application to 3d Simulation of the Primary Break-Up of a Liquid Jet, Int. J. Multiph. Flow, 335, pp. 510524.
Bianchi, G. M., Pelloni, P., Toninel, S., Scardovelli, R., Leboissetier, A., and
Zaleski, S., 2007, 3d Large Scale Simulation of the High-Speed Liquid Jet
Atomization, SAE Technical Paper No. 2007-01-0244.
Desjardins, O., Moureau, V., and Pitsch, H., 2008, An Accurate Conservative
Level Set/Ghost Fluid Method for Simulating Turbulent Atomization, J. Comput. Phys., 22718, pp. 83958416.
Gorokhovski, M., and Herrmann, M., 2008, Modeling Primary Atomization,
Annu. Rev. Fluid Mech., 401, pp. 343366.
Herrmann, M., 2008, A Balanced Force Refined Level Set Grid Method for
Two-Phase Flows on Unstructured Flow Solver Grids, J. Comput. Phys.,
2274, pp. 26742706.
Herrmann, M., and Gorokhovski, M., 2008, An Outline of a LES Subgrid
Model for Liquid/Gas Phase Interface Dynamics, Proceedings of the 2008
CTR Summer Program, Center for Turbulence Research, Stanford University,
pp. 171181.
Herrmann, M., 2007, On Simulating Primary Atomization Using the Refined
Level Set Grid Method, ILASS Americas 20th Annual Conference on Liquid
Atomization and Spray Systems, ILASS Americas, Chicago, IL.
Francois, M. M., Cummins, S. J., Dendy, E. D., Kothe, D. B., Sicilian, J. M.,
and Williams, M. W., 2006, A Balanced-Force Algorithm for Continuous and
Sharp Interfacial Surface Tension Models Within a Volume Tracking Framework, J. Comput. Phys., 213, pp. 141173.
Brackbill, J. U., Kothe, D. B., and Zemach, C., 1992, A Continuum Method
for Modeling Surface Tension, J. Comput. Phys., 100, pp. 335354.
Herrmann, M., 2006, Simulating Two-Phase Flows Using the Refined Level
Set Grid Method, ILASS Americas 19th Annual Conference on Liquid Atomization and Spray Systems, ILASS Americas, Toronto, Canada.
Kim, D., Desjardins, O., Herrmann, M., and Moin, P., 2007, The Primary
Breakup of a Round Liquid Jet by a Coaxial Flow of Gas, ILASS Americas
20th Annual Conference on Liquid Atomization and Spray Systems.
Apte, S. V., Gorokhovski, M., and Moin, P., 2003, Les of Atomizing Spray
With Stochastic Modeling of Secondary Breakup, Int. J. Multiphase Flow,
299, pp. 15031522.
Moin, P., and Apte, S. V., 2006, Large-Eddy Simulation of Realistic Gas
Turbine Combustors, AIAA J., 444, pp. 698708.
Jiang, G.-S., and Peng, D., 2000, Weighted ENO Schemes for HamiltonJacobi Equations, SIAM J. Sci. Comput. USA, 216, pp. 21262143.
Shu, C. W., 1988, Total-Variation-Diminishing Time Discretization, SIAM
Soc. Ind. Appl. Math. J. Sci. Stat. Comput., 96, pp. 10731084.
Alonso, J. J., Hahn, S., Ham, F., Herrmann, M., Iaccarino, G., Kalitzin, G.,
LeGresley, P., Mattsson, K., Medic, G., Moin, P., Pitsch, H., Schluter, J.,
Svard, M., der Weide, E. V., You, D., and Wu, X., 2006, CHIMPS: A HighPerformance Scalable Module for Multi-Physics Simulation, 42nd AIAA/
ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, AIAA Paper No.
2006-5274.