41st AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit AIAA 2005-3747

10 - 13 July 2005, Tucson, Arizona

Spray Angle Variation of Liquid-Liquid Swirl Coaxial

Injectors

A. R. Ramezani1 and A. Ghafourian2

Department of Aerospace Engineering, Sharif University of Technology

Tehran, Iran, 1458889649

Spray angle behavior of a liquid-liquid swirl coaxial injector is experimentally

investigated through cold tests. This type of injector produces two coaxial liquid sprays
which interact to form a combine spray. A theoretical model based on momentum balance is
developed to predict the combine spray behavior. The spray angles for different operating
conditions are obtained from analysis of high resolution digital images. The experimental
observations indicate that the two inner and outer sprays are pulled together and interact to
form a combine spray. Their interaction results in an overall performance that is different
from the summation of each individual spray characteristic. Their behavior is not linear.
The combine spray angles are between the inner and outer spray angles. The spray angles
are calculated by the theoretical relation developed. Comparison of theoretical results with
experimental measurements shows that momentum balance can be used to predict the
observed behavior.

Nomenclature
A = discharge orifice area
Cd = discharge coefficient
m& = mass flow rate
V = velocity
P = pressure drop
= density of fluid
= half angle of inner cone spray
= half angle of outer cone spray
= half angle of combine cone spray

I. Introduction

S pray behavior in combustion chamber determines the location and pattern of energy release in chamber cavity.
Spray angle is one of the important factors in determining spray behavior. This parameter can directly influence
combustion efficiency, blow out limit, amount of unburned fuel and smoke generation.1 Knowledge of spray angles
helps in determining the interaction between adjacent injector sprays and mass distribution pattern in the chamber.2
Spray characteristics depend on many parameters including the injector type which has produced the spray.
Different types of injectors are developed over the years. Typical injectors used in liquid fueled rocket engines
are coaxial, impinging jets and pintle types.3 Coaxial injectors have a relatively high thrust to injector ratio and can
be throttled in a wide range of fuel and oxidizer flow rates. Therefore, they are normally used for main engine
applications as well as for maneuvering systems.4-5 Coaxial injectors are made of two coaxial tube passages. These
kinds of injector are divided into different types by both the geometry and fluid state. By geometry, they are divided
into three categories: shear-shear, swirl-shear and swirl-swirl. By fluid state, they are divided into two categories:
gas-liquid and liquid-liquid.
1
Research Assistance, Department of Aerospace Engineering, E-mail: Ramezani@mehr.sharif.edu.
2
Associate Professor, Department of Aerospace Engineering, E-mail: Ghafourian@sharif.edu.

1
American Institute of Aeronautics and Astronautics

Copyright © 2005 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.
 Spray angles of a particular liquid-liquid swirl coaxial injector are experimentally measured and theoretically
modeled. This type of injector produces two coaxial liquid sprays. The experimental observations made in this study
indicate that these two sprays are pulled toward each other and establish a combine spray. A theoretical model based
on momentum balance of two sprays is developed to predict the overall spray angle behavior. Some of the relevant
previous research works are reported in the next section. Theoretical modeling, experimental approach, results and
discussion are presented in the following sections.

II. Background
Even though, Liquid-liquid swirl coaxial injectors are used in eastern block propulsion systems, very limited
research work is reported on this subject.5 Many researchers have focused their activity on gas-liquid coaxial
injectors. Their results on spray pattern and swirl flow effects are useful in liquid-liquid swirl coaxial injectors study.
Ghafourian et al. have presented an extensive literature review on atomization characteristics of different types
of injectors including gas-liquid injectors.6 Important non-dimensional numbers and geometrical parameters
governing the performance of injectors are reported in their work. The shape and disintegration characteristics of
swirled annular liquid sheets formed in coaxial injectors are investigated by Ramamurthi and Tharakan.7 They
determined the effects of injection condition on the tulip shape and transition condition for the formation of conical
sheet.
A high gas-liquid momentum ratio coaxial injector with swirl in liquid phase, operating at elevated pressure is
studied by Strakey et al..8 They observed that the radial spreading of the swirl coaxial spray is less than what is
previously reported when the injector is operated at low pressure. They proposed a model to predict spray angle in
gas-liquid coaxial swirl injectors based on momentum ratio. Inamura et al. theoretically modeled the thickness of
liquid sheet formed on the inner wall of center post, spray angle and breakup length in a gas-liquid coaxial injector.9
They measured these parameters experimentally too. The theoretically calculated results compare well with the
experimentally measured values.
Amagai and Arai are among the very limited researchers who performed experimental research on liquid-liquid
shear coaxial injectors.10 They found that relative velocity between the inner and outer sprays has significant effect
on drop size distribution produced by the injector. Sivakumar and Raghunandan studied the interaction effect of two
liquid sheets formed by a liquid-liquid swirl coaxial injector.11 Their injector is helical type. They made visual
measurements of sheet angle and emphasized on the importance of possible interaction liquid sheets can have on
each other.
Swirl coaxial injectors have the potential of been used as injectors for tripropellant engines. Ramamurthi and
Madhavan-Nair experimentally studied the inner and outer spray angle variation with injection pressure of a swirl
coaxial injector.4 Liquid oxygen is injected from the inner core and liquid hydrogen mixed with gaseous nitrogen is
injected from the outer orifice. These fluids are used to simulate tripropellant injection condition. They observed that
with increasing pressure drop, inner spray angle increases. As the gas pressure increases, the outer cone angle
slightly decreases.
Merging and separating the two inner and outer liquid sheets drastically influence drop size distribution of
liquid-liquid swirl coaxial injectors as studied by Sivakumar and Raghunandan.12 They observed that for a fixed
inner flow rate by increasing the outer flow rate, the mean drop size initially increases to reach a maximum value
and then decreases. The inner sheet has influence on drop size only at low flow rates of outer sheet. At high outer
orifice flow rate, by increasing the swirl and having a positive offset of the inner jet, the quality of spray improves.
These results have practical application on cases where throttling has to be performed. In the next section a simple
theory is developed to predict spray angle of liquid-liquid swirl coaxial injectors. Experimental results are compared
with the results of this theory.

III. Theoretical Modeling

The importance of momentum ratio of two stream flows in both gas-liquid coaxial and impinging jet injectors on
the overall spray angle has been determined by previous researchers.8-13 This idea is further developed here for the
liquid-liquid swirl coaxial injector. A relation is proposed between momentums of the two streams exiting from the
inner and outer orifices and their combination which forms the final spray. By assuming steady state, inviscid
uniform pressure and velocity for the liquid exiting the orifices and neglecting body forces, momentum balance
equations in axial, z, and radial, r, directions become:

2
American Institute of Aeronautics and Astronautics
 m& iViz + m& oVoz = (m& i + m& o )V f z (1)
m& iVir + m& oVor = (m& i + m& o )V f r (2)
where m & and V are the mass flow rate and velocity. Subscripts i, o and f indicate inner, outer and combine sprays
respectively.
A resisting force is required to maintain the tangential momentum in the flow. Wall reaction provides this force
inside the injector. Tangential momentum vanishes outside the injector, due to absence of the required resisting
force. Therefore, tangential momentum balance is not considered in spray flow.
Half spray angle of a swirl injector is defined as:14
V fr
= Arc tan (3)
V fz
Substituting equations (1) and (2) into equation (3) gives
m& oVor + m& ir Vir
= Arc tan (4)
m& oVoz + m& iViz
This relation is further extended to
m& oVo Sin( ) + m& iVi Sin( )
= Arc tan (5)
m& oVo Cos ( ) + m& iVi Cos ( )
, and are the half angles of the inner, outer and combine sprays respectively. Equation (5) is used to predict
spray angle of the liquid-liquid swirl coaxial injector. Prediction results are plotted and compared with experimental
results which are presented in Results and Discussion section.

IV. Experimental Detail

A test facility including injector holder, liquid supply lines and control panel is designed and fabricated for this
investigation. The actual operating range of the experimental condition is determined through a series of tests. The
details of experimental facility and test conditions are presented in the following section.

A. Experimental setup
The liquid-liquid swirl coaxial injector
investigated is schematically shown in Figure 1.
This injector is manufactured and tested for
integrity and correctness of dimensions. Two and
four tangential ports supply the liquid to the inner
and outer simplex atomizers respectively. The
major dimensions of the injector are shown in
Table 1. The injector is kept in place by an Figure 1. Geometrical structure of the liquid-liquid swirl
injector holder. This holder provides the proper coaxial injector.
connections between the supply lines and the
injector ports.
A schematic diagram of the experimental Table 1. Major dimensions of the swirl coaxial injector.
setup showing the supply lines is given in Figure d o [mm] d i [mm] Ds [mm]
2. Two separate supply lines of liquid are used to Discharge Inlet Vortex
feed the liquid-liquid swirl coaxial injector. Water orifice orifice chamber
is used to perform the cold tests. The injection diameter diameter Diameter
environment is at atmospheric pressure.
Two pressurized tanks are used in order to
supply liquid to the injector during tests. Inner
Simplex 2 1.2 6
Regulated high pressure nitrogen gas is used to
pressurize the supply tanks in order to have a Atomizer
relatively uniform supply of liquids to the Outer
injector. The supply lines pressures prior to Simplex 5.8 1.5 9
injector are controlled with a combination of a Atomizer

3
American Institute of Aeronautics and Astronautics
regulator and a needle valve. The two supply lines
are equipped with calibrated rotameters and Control Panel
Bourdon pressure gages for measurements of flow
rates and pressure. The controllable parameters Pressure Gage
N2 Gas
are flow rates and pressure of each supply line.
The measurable parameters are the supply line
pressures prior to entrance to the injector and flow N2 Gas Flow Meter

rates.
The visual characteristics of the sprays are Water
Tank
Water
Tank
Manual Valve
determined by studying the video images. Images
Injector
of sprays are obtained by a Sony digital camera
DSC F717 with a maximum image resolution of 5
Mpix and a Sony digital video camera Model
TRV330. Images were taken with a back light Figure 2. Diagram of the test facility for investigation of a
environment supplied by a strobe flash light. Liquid-liquid swirl coaxial injector.

B. Experimental conditions
The minimum operating range limits of the injector are determined by minimum flow rates required for the
formation of sprays. The maximum flow rates are determined by the pressurization limit of the liquid supply tanks.
The test condition flow rates are designed based on the visualization study.
To determine the actual flow rates from the inner and outer discharge orifices, a series of tests are performed to
determine their discharge coefficients. Flow rate ranges for spray angle measurement of the inner and outer sprays
are 40 to 130 lit/hr and 110 to 430 lit/hr respectively. Table 2 shows the flow rate conditions for measurements of
combine spray angle. Repeatability of test results has been checked. The results used for analysis are the average
results in repeated tests.

Table 2. Flow rate conditions for measurements of combine spray angle.

Outer Mass Inner Mass Flow Rate
Flow Rate 70 90 110 130
200 70-200 90-200 110-200 130-200
250 70-250 90-250 110-250 130-250
300 70-300 90-300 110-300 130-300
350 70-350 90-350 110-360 130-350
400 70-400 90-400 110-400 130-400

V. Results and Discussion

Shape and angle of individual sprays formed by each orifice and the combine spray of the liquid-liquid swirl
coaxial injector are measured. Interaction of the two sprays on each other is investigated. Calculated results from
theoretical predictions are compared with experimental results. In the following sections, results and related
discussions are presented.

A. Visual Observation of Spray Formation

The sprays formed by the inner and outer discharge orifices are visually studied. The spray formed by the inner
orifice has the expected behavior of going through onion, tulip and cone sheet with increasing flow rate.14 The spray
formed by the outer orifice does not have the onion behavior and starts with tulip mode and moves to cone range
with increasing flow rate. When the inner simplex atomizer is removed from the injector and the outer simplex
atomizer is tested alone, the onion behavior is observed. Presence of the inner simplex provides a passage from the
middle of spray to outside which prevents the formation of onion mode.
Observations indicate that the inner and outer cone sheets are attracted to each other. The inner cone sheet with a
given fixed flow rate moves outward as flow rate from the outer orifice increases when both cone sheets are present.
The outer cone sheet showed the same behavior toward the inner cone sheet. The presence of both cones can be seen
in Figure 3. The inner cone sheet rapidly merges with the outer sheet. It is difficult to distinguish between the two
sprays after they merge. They remained indistinguishable, even when different color liquids were feed thought the

4
American Institute of Aeronautics and Astronautics
inner and outer orifices. If the flow rates are high
enough such that no cone sheets is present and spray
is formed, then these movements can not be
visualized. Special equipment like radial collector
can be used to quantitatively measure these
movements of sprays.
For both spray, cone sheet perforation, rim and
wave breakup modes are observed. These breakup
modes are described by other researchers.14-15
Different forces such as aerodynamic and surface
tension dominate in different breakup regions.
As flow rates increase, the sprays move to fully
atomize region with no unbroken liquid body such
as a cone sheet. Fully atomized spray is established
due to existence of complex phenomena such as
cavitations and turbulent structures.16 Figure 3. Two coaxial sprays' structure of a swirl coaxial
injector.
B. Discharge coefficient
The discharge coefficient, Cd, is determined by
using the following relation:
m&
Cd = (6)
A 2 P
Area, , and density, , are known values and
mass flow, m & , and pressure drop, P, are measured
values. For both the inner and outer discharge
orifices the entire areas are used as A. Usage of
equivalent hydraulic diameter for the outer
discharge orifice gives unacceptable results. This is
a very important point which has not been
considered by other researchers.
An injector is build with the same outer simplex
atomizer and without the inner simplex atomizer.
This single outlet injector is tested and its discharge
coefficient is measured to be 0.12. It is known that Figure 4. Inner discharge coefficient variation with flow
when the inner orifice is in its place, a lower rate.
discharge coefficient than 0.12 should be obtained.
This is because of the resistance it produces toward
flow, since the inner simplex atomizer body is larger
than the possible air core that might exist in the
swirl flow. If hydraulic diameter is used, the
measured discharge coefficient for the outer
discharge orifice becomes 0.17, which is not
acceptable, since it is large than 0.12. Usage of the
actual diameter gives a discharge coefficient of
0.10, which is acceptable, since it less than 0.12.
The results of measured discharge coefficients
are presented in Figures 4 and 5 verses mass flow
rate. The patterns show a decrease of discharge
coefficient value with increasing flow rate. The
values of discharge coefficient reach a steady value
after flow reaches a critical value. The critical flow
rate values for the inner and outer orifices are 90
and 300 liter/hour respectively. Figure 5. Outer discharge coefficient variation with flow
In general, swirl flow atomizers have lower rate.

5
American Institute of Aeronautics and Astronautics
discharge coefficients than the shear flow atomizers because of the existence of an air core in swirl flow. Many
researchers have worked on calculations of the discharge coefficient. They have found that discharge coefficient
becomes independent of Reynolds number at large values of Reynolds number.14

C. Operating Range
Reynolds number, Re, and Weber number, We are the two non-dimensional numbers that has been identified by
other researchers to best represent operating condition of injectors. If experimental results are to be used in industrial
applications, the values of Reynolds number and Weber number at experimental conditions have to be close to
actual values at industrial conditions. Exiting liquid from injector is fully atomized immediately after the exit plane
of the injector in industrial condition.
Experimental conditions achieved for measurements of liquid-liquid swirl injector behavior in this research are
given in Table 3. Due to swirl flow in outer orifice, actual exit area is used in discharge coefficient calculation
instead of equivalent hydraulic area. The Reynolds numbers of another researcher from the limited publications
available are also presented in Table 3. Comparison of these conditions indicates that the lowest value of Reynolds
number in this study is greater than the highest value of the other research work.
Behavior of a spray formed from interaction of the two sprays formed by the liquid-liquid swirl coaxial injector
at the above mentioned conditions are presented in the following sections.
Table 3. Experimental condition achieved in this study (Water, 300 K, 1 atm.)
Min Max
m& P Re Re m& P Re Re
[lit/hr] [bar] [Ref. 12] [lit/hr] [bar] [Ref. 12]
Inner
70 4 25000 10000 130 15.7 55000 25000
atomizer
Outer
200 2 10000 5000 400 8.4 25000 10000
atomizer

D. Spray Angle measurement from Images

Analysis of images obtained from sprays, with
controlled and repeatable lighting; indicate that with
increasing flow rate, both inner and outer spray
angles increase. Spray angle increasing rates
decrease as the flow rate reaches it upper limits as
shown in Figures 6 and 7. Inner and outer spray
angles become almost constant for flow rates above
90 liter/hr and 300 liter/hr respectively. The same
behavior is observed for discharge coefficients
variations with flow rate, except that its values
decrease with increasing flow rate. As spray angles
increase with flow rate, discharge coefficients
decrease and research their limiting values at flow
rates where spray angles become almost constant
too. The same observations have been reported by
other researchers. Figure 6. Inner spray angle variation with mass flow rate.
When both sprays are present, they merge and establish a single spray for which its angle variation for different
test conditions is shown in Figure 8. For fixed outer orifice flow rate, the inner orifice flow rate is varied from zero
to 130 liter/hr. The established spray angles for each fixed values of outer flow rate are presented by a curve on the
Figure 8. The curve for inner spray alone is also presented in the same Figure. Condition of no outer flow rate on
each curve represents the inner spray alone.
The combine spray angle decreases with increasing inner orifice flow rate. As outer orifice flow rate increases
the combine spray maintains its integrity as a single spray and its angle increases. The measured angles of combine
spray for given values of inner and outer flow rates are between the spray angles measured for inner and outer
sprays alone for the same flow rates. This behavior can be due to fact that the out coming stream from one orifice
induces a flow field in its vicinity which reduces local pressure in that region. The second stream from the other

6
American Institute of Aeronautics and Astronautics
orifice is exposed from one side to this induced low
pressure field and from the other side to ambient
environment. Pressure difference between the two
exposed sides of second stream causes a deflection
towards low pressure side. The same effect is
induced by second stream on first stream. Overall
effect is observed as an attraction effect between the
two sprays.
From geometry of each spray trajectory, the
impingement point should be much lower than the
point where the two sprays actually impinge. This
up shift in impingement location is due to the above
mentioned attraction effect between the two sprays.
Momentum balance between the two streams can be
the determining factor for combine spray angle after
the impingement point. In Figure 9, results obtained
from equation (5) based on momentum balance is
Figure 7. Outer spray angle variation with mass flow rate.
plotted and compared with experimental results.
Comparison indicates that momentum balance is an
appropriate approach for prediction of combine
spray angle. Accuracy of spray angle measurements
from images is within ±2 degrees. The sources of
inaccuracy are small fluctuations in spray surface
and image processing software ability to determine
outer spray surface limits.

VI. Conclusion
Knowledge of spray angle and mixing behavior
of injectors is essential for design of new
combustion chambers and analysis of existing
chamber performance. A liquid-liquid swirl coaxial
injector is experimentally tested and its performance
is modeled based on a momentum balance theory. Figure 8. Effect of inner and outer mass flux on combine
Each of two sprays exiting from a liquid-liquid spray angle.
swirl coaxial injector pass through different breakup
regions by increasing flow rate. They show a
behavior similar to sprays formed by simplex
atomizers. Spray angle increases with flow rate and
reaches a limit. Spray angle variations are measured
at fully atomized region. Measurements indicate that
the two sprays exiting from inner and outer orifices
induce an attraction force on each other. They
impinge and form a combine spray. Combine spray
angle is between angles of individual inner and
outer sprays. Momentum balance between two
sprays can be used to predict the angle of combine
spray. Momentum ration of two sprays is an
important parameter in determining the nonlinear
behavior of combine spray.

Figure 9. Comparison of measured spray cone angle with

theoretical results.

7
American Institute of Aeronautics and Astronautics
 Acknowledgment
This work is supported by Combustion and Fuel Lab, Department of Aerospace Engineering, Sharif University
of Technology and Optic Lab, Department of Mechanical Engineering, Tehran University. Authors thank and
acknowledge the contributions of Mr. Sarkar in fabricating the experimental facility and Mr. Morad and Mr.
Mesbahi in collecting data and for their useful suggestions.

Reference
1
Lefebvre, A. H., Gas Turbine Combustion, Taylor & Francis, 1999
2
Harrje, C.T., Reardon, F.H., and eds., Liquid Propellant Rocket Combustion Instability, SP-194, NASA, 1972
3
Anon, Liquid Rocket Engine Injectors, SP-8089, NASA, 1976
4
Ramamurthi, K., and Madhavan Nair, G., “Coaxial Swirl Injection Element for a Tripropellant Engine,” AIAA 98-3514,
1998
5
Woodward, R.D., Miller, K.L., Bazarov, V.G., Guerin, G.F., Pal, S., and Santora, R.J., “Injector Research for Shuttle OMS
Upgrade Using LOX/ETHANOL Propellant,” AIAA 98-3816, 1998
6
Ghafourian, A., Mahalingam, S., Dindi, H., and Daily, J. W., “A Review of Atomization in Liquid Rocket Engines,” AIAA
91-0283, 1991
7
Ramamurthi, K., and Tharakan, T. J., “Experimental Study of Liquid Sheets Formed in Coaxial Swirl Injectors,” Journal of
Propulsion and Power, Vol. 11, No. 6, November-December 1995
8
Strakey, P.A., Talley, D.G., and Hutt, J.J., “Mixing Characteristics of Coaxial Injectors at High Gas/Liquid Momentum
Ratios,” Journal of Propulsion and Power, Vol. 17, No. 2, March-April 2001
9
Inamura, T., Tamura, H., and Sakamoto, K., “Characteristics of Liquid Film and Spray Injected from Swirl Coaxial
Injector,” J. of Propulsion and Power, Vol. 19, No. 4, pp. 632-639, July- August 2003
10
Amagai, K., and Arai, M., “Disintegration Process of a Liquid Jet Injected from a Coaxial Dual Nozzle,” ICLASS, 94, pp.
71-78, July 1994
11
Sivakumar, D., and Raghunandan, B.N., “Jet Interaction in Liquid-Liquid Coaxial Injector,” Journal of Fluid Engineering,
Vol.118, pp. 329-334, 1996
12
Sivakumar, D., and Raghunandan, B.N., “Role of Geometric Parameters on the Drop Size Characteristics of Liquid-Liquid
Coaxial Swirl Atomizers,” Journal of Atomization and Sprays, Vol. 8, pp. 547-563, 1998
13
Sutton, G. P., Rocket Propulsion Elements, John Wiley & Sons, 2001
14
Lefebvre, A. H., Atomization and Sprays, Hemisphere Corporation, 1989
15
Bayvel, L., Orzechowski, Z., Liquid Atomization, Taylor & Francis, 1993
16
Liu, Humin, Science and Engineering of Droplets: Fundamentals and Application, Noyes Publications/William Andrew
Publishing, 2000

8
American Institute of Aeronautics and Astronautics