Experimental Liquid Rocket Swirl Coaxial Injector Study Using Non … · American Institute of ... 4.14 MPa and propellant mass flowrates of 0.118 kg/s and 0.039 kg/s for ... it is

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/268259345

Experimental Liquid Rocket Swirl Coaxial Injector Study Using Non-Intrusive

Optical Techniques

Article · July 2005

DOI: 10.2514/6.2005-4299

CITATIONS

8READS

566

6 authors, including:

Some of the authors of this publication are also working on these related projects:

NSF-DOE DME project View project

DOE: Evaporation Sub-Model Development for Volume of Fluid (eVOF) Method Applicable to Spray-Wall Interaction Including Film Characteristics with Validation at High

Pressure-Temperature Conditions View project

Seong-Young Lee

Michigan Technological University

97 PUBLICATIONS 1,497 CITATIONS

SEE PROFILE

Shri Lal Pal

Kulwant College Dehradun


SEE PROFILE

Robert J. Santoro

Pennsylvania State University


SEE PROFILE

All content following this page was uploaded by Robert J. Santoro on 29 January 2015.

The user has requested enhancement of the downloaded file.

https://www.researchgate.net/publication/268259345_Experimental_Liquid_Rocket_Swirl_Coaxial_Injector_Study_Using_Non-Intrusive_Optical_Techniques?enrichId=rgreq-0b801f9f19fcee497db0bdb7744f5cfa-XXX&enrichSource=Y292ZXJQYWdlOzI2ODI1OTM0NTtBUzoxOTA5ODc5NDA3NTM0MDhAMTQyMjU0NjQ4MTI1Nw%3D%3D&el=1_x_2&_esc=publicationCoverPdfhttps://www.researchgate.net/publication/268259345_Experimental_Liquid_Rocket_Swirl_Coaxial_Injector_Study_Using_Non-Intrusive_Optical_Techniques?enrichId=rgreq-0b801f9f19fcee497db0bdb7744f5cfa-XXX&enrichSource=Y292ZXJQYWdlOzI2ODI1OTM0NTtBUzoxOTA5ODc5NDA3NTM0MDhAMTQyMjU0NjQ4MTI1Nw%3D%3D&el=1_x_3&_esc=publicationCoverPdfhttps://www.researchgate.net/project/NSF-DOE-DME-project?enrichId=rgreq-0b801f9f19fcee497db0bdb7744f5cfa-XXX&enrichSource=Y292ZXJQYWdlOzI2ODI1OTM0NTtBUzoxOTA5ODc5NDA3NTM0MDhAMTQyMjU0NjQ4MTI1Nw%3D%3D&el=1_x_9&_esc=publicationCoverPdfhttps://www.researchgate.net/project/DOE-Evaporation-Sub-Model-Development-for-Volume-of-Fluid-eVOF-Method-Applicable-to-Spray-Wall-Interaction-Including-Film-Characteristics-with-Validation-at-High-Pressure-Temperature-Conditions?enrichId=rgreq-0b801f9f19fcee497db0bdb7744f5cfa-XXX&enrichSource=Y292ZXJQYWdlOzI2ODI1OTM0NTtBUzoxOTA5ODc5NDA3NTM0MDhAMTQyMjU0NjQ4MTI1Nw%3D%3D&el=1_x_9&_esc=publicationCoverPdfhttps://www.researchgate.net/?enrichId=rgreq-0b801f9f19fcee497db0bdb7744f5cfa-XXX&enrichSource=Y292ZXJQYWdlOzI2ODI1OTM0NTtBUzoxOTA5ODc5NDA3NTM0MDhAMTQyMjU0NjQ4MTI1Nw%3D%3D&el=1_x_1&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Seong-Young-Lee?enrichId=rgreq-0b801f9f19fcee497db0bdb7744f5cfa-XXX&enrichSource=Y292ZXJQYWdlOzI2ODI1OTM0NTtBUzoxOTA5ODc5NDA3NTM0MDhAMTQyMjU0NjQ4MTI1Nw%3D%3D&el=1_x_4&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Seong-Young-Lee?enrichId=rgreq-0b801f9f19fcee497db0bdb7744f5cfa-XXX&enrichSource=Y292ZXJQYWdlOzI2ODI1OTM0NTtBUzoxOTA5ODc5NDA3NTM0MDhAMTQyMjU0NjQ4MTI1Nw%3D%3D&el=1_x_5&_esc=publicationCoverPdfhttps://www.researchgate.net/institution/Michigan_Technological_University?enrichId=rgreq-0b801f9f19fcee497db0bdb7744f5cfa-XXX&enrichSource=Y292ZXJQYWdlOzI2ODI1OTM0NTtBUzoxOTA5ODc5NDA3NTM0MDhAMTQyMjU0NjQ4MTI1Nw%3D%3D&el=1_x_6&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Seong-Young-Lee?enrichId=rgreq-0b801f9f19fcee497db0bdb7744f5cfa-XXX&enrichSource=Y292ZXJQYWdlOzI2ODI1OTM0NTtBUzoxOTA5ODc5NDA3NTM0MDhAMTQyMjU0NjQ4MTI1Nw%3D%3D&el=1_x_7&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Shri_Pal?enrichId=rgreq-0b801f9f19fcee497db0bdb7744f5cfa-XXX&enrichSource=Y292ZXJQYWdlOzI2ODI1OTM0NTtBUzoxOTA5ODc5NDA3NTM0MDhAMTQyMjU0NjQ4MTI1Nw%3D%3D&el=1_x_4&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Shri_Pal?enrichId=rgreq-0b801f9f19fcee497db0bdb7744f5cfa-XXX&enrichSource=Y292ZXJQYWdlOzI2ODI1OTM0NTtBUzoxOTA5ODc5NDA3NTM0MDhAMTQyMjU0NjQ4MTI1Nw%3D%3D&el=1_x_5&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Shri_Pal?enrichId=rgreq-0b801f9f19fcee497db0bdb7744f5cfa-XXX&enrichSource=Y292ZXJQYWdlOzI2ODI1OTM0NTtBUzoxOTA5ODc5NDA3NTM0MDhAMTQyMjU0NjQ4MTI1Nw%3D%3D&el=1_x_7&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Robert-Santoro?enrichId=rgreq-0b801f9f19fcee497db0bdb7744f5cfa-XXX&enrichSource=Y292ZXJQYWdlOzI2ODI1OTM0NTtBUzoxOTA5ODc5NDA3NTM0MDhAMTQyMjU0NjQ4MTI1Nw%3D%3D&el=1_x_4&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Robert-Santoro?enrichId=rgreq-0b801f9f19fcee497db0bdb7744f5cfa-XXX&enrichSource=Y292ZXJQYWdlOzI2ODI1OTM0NTtBUzoxOTA5ODc5NDA3NTM0MDhAMTQyMjU0NjQ4MTI1Nw%3D%3D&el=1_x_5&_esc=publicationCoverPdfhttps://www.researchgate.net/institution/Pennsylvania-State-University?enrichId=rgreq-0b801f9f19fcee497db0bdb7744f5cfa-XXX&enrichSource=Y292ZXJQYWdlOzI2ODI1OTM0NTtBUzoxOTA5ODc5NDA3NTM0MDhAMTQyMjU0NjQ4MTI1Nw%3D%3D&el=1_x_6&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Robert-Santoro?enrichId=rgreq-0b801f9f19fcee497db0bdb7744f5cfa-XXX&enrichSource=Y292ZXJQYWdlOzI2ODI1OTM0NTtBUzoxOTA5ODc5NDA3NTM0MDhAMTQyMjU0NjQ4MTI1Nw%3D%3D&el=1_x_7&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Robert-Santoro?enrichId=rgreq-0b801f9f19fcee497db0bdb7744f5cfa-XXX&enrichSource=Y292ZXJQYWdlOzI2ODI1OTM0NTtBUzoxOTA5ODc5NDA3NTM0MDhAMTQyMjU0NjQ4MTI1Nw%3D%3D&el=1_x_10&_esc=publicationCoverPdf

American Institute of Aeronautics and Astronautics

1

Experimental Liquid Rocket Swirl Coaxial Injector Study Using Non-Intrusive Optical Techniques

D. M. Kalitan*, D. Salgues†, A.G. Mouis‡, S. Y. Lee§, S. Pal**, and R. J. Santoro†† The Pennsylvania State University, University Park, PA, 16802

Uni-element liquid rocket experiments were completed for two different swirl coaxial injector configurations using liquid oxygen/gaseous methane propellants. Several different non-intrusive optical techniques such as OH planar laser induced fluorescence, OH* and CO2* chemiluminescence, laser light scattering and shadowgraph imaging, were employed to observe flame position and liquid core structure. All measurements were made in the near injector region of the chamber for steady state chamber pressures of approximately 4.14 MPa and propellant mass flowrates of 0.118 kg/s and 0.039 kg/s for liquid oxygen (LOX) and gaseous methane respectively. The two injectors studied varied in the size of the methane annulus, maintaining the LOX post diameter constant. The application of optical diagnostics, particularly those involving simultaneous measurements using two different techniques, provided detail information on the spray and flame structure for each injector.

Nomenclature I.D. = inner diameter O.D. = outer diameter τmix = mixing time dLOX = diameter liquid oxygen post υCH4 = velocity of gaseous methane υLOX = velocity of liquid oxygen J = momentum flux ratio ρCH4 = density of gaseous methane ρLOX = density of liquid oxygen dCH4 = thickness of the methane annulus σ = surface tension of the liquid oxygen We = Weber number ηC* = C* efficiency A = area •

m = mass flowrate γ = ratio of specific heats R = gas constant * Research Assistant, , Student Member AIAA. † Research Assistant,. ‡ Research Associate, AIAA Member. § Research Associate, AIAA Member. ** Senior Research Associate, AIAA Member. †† Professor, AIAA Member.

Copyright 2005 by Robert J. Santoro. Published by the American Institute of Aeronautics and Astronautics, Inc. with permission.

41st AIAA/ASME/SAE/ASEE Joint Propulsion Conference AIAA 2005-4299 Tucson, AZ, July 10-13, 2005


2

I. Introduction DVANCEMENT in the design, configuration, implementation, and analysis of injector performance in liquid rocket engines has not occurred on a large scale in the past decade. Without a specific mission goal as the

driving force for an in-depth injector study, it is difficult to justify the need for improvement. However, with renewed interest in missions to locations beyond the current shuttle capabilities,1 the necessity for an optimally performing injector configuration is of great importance. In addition, specific propellant combinations that are candidates for such a mission need to be investigated in order to gain a more complete understanding of injector/propellant interactions in reacting high-pressure, high-temperature conditions. The present study therefore seeks to gain needed understanding of the swirl coaxial injector mixing and atomization characteristics using LOX/methane propellants for potential use in space applications. The swirl coaxial injector, a modification to the widely used shear coaxial injector, has been implemented primarily in Russia2-5 and the United States6-11 since the 1950’s for use in liquid rocket engines (LRE’s). Although Russia continues to integrate swirl coaxial injectors into their LRE’s, they are not as widely used in the United States with the RL-10 being the most noTable exception. This is in part due to the lack of fundamental data for the swirl coaxial injector’s mixing and atomization capabilities under hot-fire conditions. Recently however, as technology has advanced, it is now possible to gain understanding of the combustion phenomena that occur inside the rocket chamber using non-intrusive optical techniques such as planar laser induced fluorescence (PLIF), chemiluminescence imaging, laser light scattering and shadowgraph imaging.12-14 Few injector studies have indeed implemented these techniques to study reacting flows under rocket engine conditions, however, because of the complexity of the experiments, they are mostly qualitative in nature.15-18 There is therefore a need for research that that provides quantitative injector characterization information. The Cryogenic Combustion Laboratory (CCL)19 at The Pennsylvania State University has been the site of focused studies pertaining to injector characterization research for over a decade. A variety of injector configurations have been studied including shear coaxial,20-23 swirl coaxial,24,25 and impinging type26,27 elements in addition to utilizing a range of propellant types such as bi-propellants (LOX/H2, LOX/GCH4, LOX/RP1, LOX/ethanol) and tri-propellants (LOX/RP1/H2). In the present study, the near injector flowfield was examined for two, uni-element, tangentially swirled, coaxial injector configurations employing LOX/GCH4 propellants. Using simultaneous non-intrusive optical techniques, it was possible to superimpose images of the liquid core and the flame zone, thereby observing how mixing and atomization are occurring within the rocket chamber.

II. Experiment All experiments were conducted at the Penn State Cryogenic Combustion Laboratory. The uni-element rocket

chamber is constructed of oxygen-free high conductivity copper and has a modular design to accommodate an injector assembly, window section, igniter section, several blank sections to allow for variable chamber length and a removable water cooled nozzle section. A cross-sectional view of the rocket assembly is shown in Figure 1 and is explained in detail in Ref. 16. The interior of the chamber is 50.8 x 50.8 mm (2 in.2). The window section contains four, 50.8 mm (2 in.) diameter quartz windows that are cooled using gaseous nitrogen to prevent failure under the harsh operating conditions. It is through these four windows that the non-intrusive optical techniques are applied in order to gain an improved understanding of the combustion phenomena occurring in the near injector region of the rocket chamber.

Two different swirl coaxial injector configurations were examined, varying only in the dimension of the methane annulus, keeping the LOX post size constant. Chamber pressure was approximately 4.14 MPa (600 psia) and propellant mass flowrates were 0.118 kg/s and 0.039 kg/s for LOX and GCH4 respectively. The oxidizer to fuel ratio (O/F) was close to 3.0 making the combustion slightly fuel-rich (stoichiometric O/F ratio is 4 for LOX/GCH4) Complete injector characteristics for the two assemblies are stated in Table 1 and the entire swirl coaxial injector cross-section is shown in Figure 2.

The swirl coaxial injector differs from the shear coaxial injector in the manner in which the liquid

A

Liquid Oxygen

GaseousMethane

Nitrogen Purge

Quartz Window

Igniter Cooling Water In

Cooling Water Out

Liquid Oxygen

GaseousMethane

Nitrogen Purge

Quartz Window

Igniter Cooling Water In

Cooling Water Out

Fig. 1. Cross-sectional view of the optically accessible rocket chamber. All dimensions are in inches.


3

travels inside the injector. For the present injectors, LOX enters the swirl nut of the injector tangentially through three slots at 120 degree intervals and subsequently, as the liquid travels the length of the injector post, a liquid sheet is formed which develops into a conical sheet as it leaves the injector and enters the rocket chamber. The co-flowing high velocity gaseous methane acts to break-up the liquid sheet into ligaments and drops, which then vaporize and burn. Perhaps the two most important non-dimensional parameters for measuring system performance with regards to liquid jet or sheet break-up and their subsequent atomization are the momentum flux ratio and Weber number4-11,16,24, and are defined respectively as follows,

2

244

LOXLOX

CHCHJυρ

υρ= (1)

( )συυρ

444

2CHLOXCHCH dWe

−= (2)

where ρCH4 and υCH4 and ρLOX and υLOX are the density and velocity of methane and LOX respectively, σ is the surface tension of the LOX (σ=0.001 kg/m/s for LOX at 120K) and dCH4 is the width of the methane annulus.

The temporal and spatial aspects of the breakup of the conical liquid sheet are known to be sensitive to the ratio of the momentum flux of the gas to the liquid. The effect of increasing the annular area for Injector 2 is to significantly reduce the momentum flux ratio and thus it would be expected to degrade the performance of Injector 2 as compared to Injector 1. Higher Weber numbers are known to indicate improved secondary atomization of liquid ligaments and drops and this effect is also expected to contribute to a better performance from Injector 1 than Injector 2. The momentum ratio for Injector 1 was approximately J = 15.3, and J = 2.6 for Injector 2. The Weber number for Injector 1 was approximately 203,860 while the Weber number for Injector 2 was approximately 67,510. These large differences for the two injectors are due to the difference in the methane velocity between Injector 1 and Injector 2, i.e., 214 m/s compared to 87 m/s and are expected to result in a significant performance difference between the two injectors.

One of the objectives of the present study is to ascertain how the optical diagnostics qualitatively and quantitatively yields insights into the expected performance differences of the two injectors. The optical diagnostics employed in the present study allow measurements of several spray characteristics and combustion processes occurring for these two injectors. In order to visualize the reacting flowfield for the two different injector configurations, four different non-intrusive optical techniques were employed, OH planar laser-induced fluorescence (OH PLIF), OH* and CO2* chemiluminescence, laser light scattering, and shadowgraph imaging. In some cases two of the above-mentioned techniques were applied simultaneously. Figure 3 illustrates the combined optical set-up for the OH PLIF, chemiluminescence and laser light scattering (LLS) measurements. OH PLIF was used to capture two-dimensional images of OH radicals in the reacting flow. Since the OH radicals are present in high temperature regions, they are a reasonable indicator of the combustion zone location. A Nd-YAG laser (532 nm) was used to pump a dye laser followed by a frequency doubler. The system was tuned to the Q1 branch of the J’’= 9 rotational level in the transition from ν’’ = 1 ← ν’ = 0 at 283.93 nm. The ICCD camera was fitted with one

Table 1. Tangential swirl coaxial injector characteristics for Injectors 1 and 2. LOX Injector 1 Injector 2No. of Rectangular Tangential Inlet Slots 3 3Rectangular Tangential Slot Width (mm) 0.762 0.762Rectangular Tangential Slot Length (mm) 2.794 2.794Post I.D. (mm) 3.429 3.429Post O.D. (mm) 4.191 4.191Tangential Velocity (m/s) 12.707 12.707Axial Velocity (m/s) 22.009 22.009Pressure Drop (kPa) 758.42 758.42

MethaneAnnulus I.D. (mm) 4.191 4.191Annulus O.D. (mm) 5.182 6.35Annulus Area (mm2) 7.290 17.874Pressure Drop (kPa) 577.200 96.06Velocity (m/s) 213.650 87.161Mach Number 0.474 0.1934

Fig. 2. Cross-sectional view of swirl injector. All dimensions are in mm.


4

UG11 filter and three WG305 filters to eliminate interferences from the elastically scattered light from the high-density liquid oxygen present in the near injector region. The laser light scattering diagnostic took advantage of the fact that the LOX conical sheet, ligaments and droplets elastically scatter the incident laser beam and thereby provide information concerning the structure of the liquid core. The LLS technique used the same laser configuration as the above-mentioned OH PLIF technique (Figure 3) however; the ICCD camera was fitted with only one UG11 filter to collect the vertically polarized scattered light. Line of sight (LOS) chemiluminescence imaging of OH* and CO2* were used to visualize the flame location. This diagnostic used the same camera locations as illustrated in Figure 3, however, the laser was not needed for this particular diagnostic. The ICCD cameras were fitted with narrowband interference filters centered at 310 nm and 430 nm for OH* and CO2* respectively. Similar to laser light scattering, the shadowgraph technique was used to image the conical liquid sheet. The shadowgraph technique however, required a different experimental configuration than the one illustrated in Figure 3. To collect the shadowgraph images, only one ICCD camera was used and it was shifted from the top window location to a side window location. A strobe light and a pane of diffuser glass were placed on the opposite side of the chamber thereby creating a backlighting effect.

III. Results and Discussion In the following sections, results for both injector configurations are presented. A comparison of each optical diagnostic for both injectors is presented as well as, superimposed results of simultaneously captured images from two different diagnostic techniques. The major combustion performance parameter determined was the C* efficiency which is based on the measured chamber pressure. The C* efficiency, ηC*, is defined as

lTheoretica

AcutalC C

C*

** =η (3)

where, C* is the characteristic exhaust velocity and is defined as,

[ ] )1()1(*

)1(2 −+•

+==

γγγγ

γRT

m

APC tchamber (4), so *Cη can be calculated as lTheoretica

AcutalC P

P=*η (5).

Laser 532 nmDye Laser

566 nm

Frequency Doubler 283.55 nm

UV Laser Beam

Data Acquisition System

Flow Direction

Mirrors ICCD CamerasLaser 532 nm

Dye Laser 566 nm

Frequency Doubler 283.55 nm

UV Laser Beam

Data Acquisition System

Flow Direction

Mirrors ICCD Cameras

Fig. 3. Optical configuration used for the PLIF and laser scattering measurements.


5

In the above equations, Pchamber is the chamber pressure, At is the area of the nozzle throat, •

m is the mass flowrate, γ is the ratio of specific heats, R is the gas constant and T is the chamber temperature. The C* efficiency is a good indicator of overall engine performance. Results for this analysis show that Injector 1 has a higher performance, operating with an average C* efficiency of 94.8% while Injector 2 had on average an efficiency of 90.9%. It is apparent from the C* efficiency measurements that Injector 1 performs better than Injector 2 as expected from Weber number and momentum flux ratio calculations. The only variable in this study is the velocity of the methane. Injector 1 and Injector 2 both use the same LOX injector post and therefore, the velocity of the LOX, as well as its momentum, are the same for both injectors. However, the outer diameter of the methane annulus for Injector 1 is 0.204 in., whereas for Injector 2 the diameter was increased to 0.25 in. Therefore, for the same mass flowrate of propellant, both the velocity and momentum of the methane for Injector 1 will be greater than Injector 2, which will significantly affect the momentum flux ratio and the Weber number.

A. Laser Light Scattering and Shadowgraph Imaging Laser light scattering (LLS) and shadowgraph imaging were used to capture characteristics of the swirled LOX

sheet structure in order to gain a more complete understanding of how a swirl injector operates under reacting conditions. Figure 4 shows typical Injector 1 (left) and Injector 2 (right) LLS images for two windows positions (upstream and downstream pictures were taken during different runs for a given injector). A drawing of the window and injector has been superposed onto the pictures for reference. A first glance at LLS images, such as the ones pictured in Figure 4, shows the appearance of a solid core. This is unexpected since a tangentially swirled injector creates a liquid sheet upon entering the combustion chamber. The solid core appearance is attributed to the internal reflections and intense light scattering from the liquid sheet and also from the large liquid drops and ligaments that are being created as a result of atomization. This renders the interpretation of LLS images difficult. However it was observed that less light was scattered at the downstream position for Injector 1 which implies that the liquid oxygen

Injector #2

Injector face

45.72 mm (1.8 in.)15.24 mm

(0.6 in.)

Injector #1

Injector face

45.72 mm (1.8 in.)15.24 mm

(0.6 in.)

Injector #2

Injector face

45.72 mm (1.8 in.)15.24 mm

(0.6 in.)

Injector #2

Injector face

45.72 mm (1.8 in.)15.24 mm

(0.6 in.)

Injector #1

Injector face

45.72 mm (1.8 in.)15.24 mm

(0.6 in.)

Injector #1

Injector face

45.72 mm (1.8 in.)15.24 mm

(0.6 in.)

Fig. 4. LOX scattering images for Injector 1 (left) and Injector 2 (right) at two window locations. Images have been each taken in separate runs. The laser sheet is directed from top to bottom.


6

sheet breaks up earlier for Injector 1 than Injector 2. The images from Figure 4 also show that the spreading angle is larger for Injector 2 than Injector 1 due to the higher gas momentum impairing the radial spread of the sheet as observed also in cold flow testing16.

The shadowgraph technique is seen as complimentary to the LLS technique because they both depict the liquid sheet. However, unlike the LLS images, shadowgraph images are not planar but are line of sight averaged images. These images illustrate density gradients within the chamber and are capable of showing mixing layers between gas (i.e., CH4) and liquid (i.e., LOX). Figures 5 and 6 show typical shadowgraph images for Injectors 1 and 2, respectively. An interesting aspect that can be seen on these images is the dark liquid sheet surrounded by a lighter grey region. This light grey region is the co-flowing, high-velocity, gaseous methane that accelerates the break up of the liquid sheet into ligaments and droplets. In both cases the CH4 quickly dissipates only a short distance from the injector face indicating rapid mixing of the methane and oxygen in the near injector region.

B. OH Planar Laser Induced Fluorescence

As stated earlier, the OH PLIF technique was used to image the reaction zone. Several parameters were measured in order to understand the flame zone better and to gain insight on the performance of each injector. Figure 7 illustrates typical single shot OH PLIF images for Injectors 1 (left) and 2 (right). Note that only the top edge of the PLIF images shows some signal due to the attenuation of the laser sheet by absorption by the OH radicals and laser light scattering from the spray as the laser sheet passed through the reactiion zone and the LOX spray.

Fig. 5. Injector 1 shadowgraph image.

Fig. 6. Injector 2 shadowgraph image.

highlow

Fig. 7. OH-PLIF images for Injector 1 (left) and Injector 2 (right) where the laser is directed from top to bottom.


7

In order to gain quantitative information from the OH PLIF images, a threshold process was applied before measurements were made in order to visibly refine the aspects of the images that were most useful for data analysis without compromising the integrity of captured signal. Subsequently, four parameters were measured for both injector configurations. First, the angle of inclination of the OH PLIF signal (i.e., half angle) to the chamber centerline was measured. The length of the signal along this inclination and an average thickness were measured. The axial location of maximum intensity in the OH PLIF image from the injector face was also measured. Average results for both Injectors 1 and 2 are presented in Table 2. The measurements confirm the visual information from Figure 7 and show that, the OH radical region for Injector 1 is typically shorter (the location of maximum intensity is closer to the injector face) and thinner than for the Injector 2. The OH region is also closer to the centerline for Injector 1. The length of the OH signal for both injectors is almost identical but the OH region starts closer to the injector face for Injector 1, in fact it appears to be anchored on the injector face. A main difference between the two injectors is that detached areas of OH radical signals (pocket burning),visible toward the top of the window in Injector 2 in Figure 7, are often observed in Injector 2 images whereas Injector 1 images do not contain such features. This zone of combustion located away from the centerline is probably due to the recirculation of unburned gases in this low performing injector. An important aspect of this test campaign was the ability to capture simultaneous images of the reacting flow using different optical techniques and then subsequently forming a single composite image to compare the spatial locations of the two measurements. The two cameras used in the experimental set-up were synchronized and it is possible to compare simultaneous OH PLIF and laser light scattering for Injectors 1 and 2.

Table 2. OH PLIF summary of results.

Measured Parameter Injector 1 Injector 2 Average Angle (Degrees) 18.56 25.68

Standard Deviation (Degrees) 3.79 6.81 Average Length (mm) 12.9 13.94

Standard Deviation (mm) 4.5 1.5 Average Thickness (mm) 1.04 1.6 Standard Deviation (mm) 0.33 0.43

Location of Maximum Intensity (mm) 8.43 14.33 Standard Deviation (mm) 1.44 0.85

Fig. 8. Injector 1 composite image (top) of OH PLIF (green color, original image bottom left) and Laser Light Scattering (red color, original image bottom right).

Laser Light ScatteringOH PLIF Laser Light ScatteringOH PLIF Fig. 9. Injector 2 composite image (top) of OH PLIF (green color, original image bottom left) and Laser Light Scattering (red color, original image bottom right).


8

The image processing was fairly simple and included scaling all images to the same size, cropping the images so that the injector face was flush with the left side of each image and the centerline of the chamber was exactly the centerline of the image. After all geometrical factors were taken into consideration, the images were overlaid and false coloring was applied to give each diagnostic a more noticeable and distinctive appearance in order to facilitate the interpretation of the data.

Figures 8 and 9 show composite images (top) followed underneath by the two original images from which the composite images were obtained for PLIF and LLS measurements for Injectors 1 and 2. The most interesting aspect in both figures is the location of the OH radical (as a flame indicator) from the OH PLIF images with respect to the liquid sheet as determined from the LLS images. The flame appears directly on the edge of the methane for both injectors and slightly separated from the liquid sheet. A correspondence can be seen between the wave-like structure in the liquid core, methane flow and OH PLIF showing that the structure of the liquid flow affect the gas flow and the flame structure.

C. OH* and CO2* Chemiluminescence OH* and CO2* chemiluminescence measurements were also used to image the reactive zone of the combustion

chamber. Monitoring a chemiluminescence signal is analogous to the OH PLIF technique because they both convey images of high-temperature reacting flows; however, kinetically their formation pathways are different and therefore, due to the turbulent reacting flowfield, one can expect that they may not occur in the same locations. Please note, it was observed in the early stages of this analysis that the OH* and CO2* chemiluminescence signals were almost identical and only OH* results are presented in a quantitative manner for brevity.

Similar to the OH PLIF images, the chemiluminescence images also underwent processing techniques. First, an intensity threshold was applied to all images. Second, the images were transferred into binary form and finally, the edges of the images were enhanced in order to show a clear demarcation between the chemiluminescence signal and the surrounding environment. Figure 10 shows a typical OH* image before processing (left) and after processing (right). It is important to note that the two inflection points (top and bottom) in the processed image are not real but are due to the window curvature. Therefore, signal from the inflection point to the right edge of the window was not used in the analysis.

Two types of chemiluminescence images are presented in this analysis. First, line of sight (LOS) chemiluminescence images are used to compare the signal edge positions for Injectors 1 and 2. Then, these images are converted to planar images using an Abel inversion transformation.28 Using the Abel images, the location of maximum intensity was found for Injectors 1 and 2 and then compared to results obtained from the OH PLIF analysis.

Results from the LOS chemiluminescence measurements for Injectors 1 and 2 are compared in Figure 11 and 12. Figure 11 shows the location of the edge of chemiluminescene images while Figure 12 show the chemiluminescence images for both injectors both at the injector face and 25.4 mm (1 inch) downstream.

Fig. 10. Typical OH* chemiluminescence image. OH* image before processing (left) and thresholded image (right) for Injector 1.


9

Figure 12 shows some qualitative but nevertheless intersting trends. The chemiluminescence zone from Injector 1 is very compact close to the injector but shows a sudden expansion toward the end of the upstream window location. The chemiluminescence zone for Injector 2 is larger and more intense close to the injector and shows a lot of variability from picture to picture. Both injector show more intense and spread out burning at the more downstream window location (this may not be obvious on the Figure because the intensity scale is different for upstream and downstream picture but the collected intensities are higher at the downstream location). The higher level of chemiluminescence close to the injector observed for Injector 2 is unexpected since Injector 2 is the poorer performing injector. However, this is a local result and cannot be linked to the overall performance of the system. Chemiluminescence measurements over the whole length of the chamber would be needed in order to link chemiluminescence to the overall injector performance (C* efficiency).

It is clear from Figure 11 that Injector 2 encompasses more of the chamber than Injector 1. This results agrees with the OH PLIF edge location results; however, it is important to remember that the chemiluminescence images are LOS (3-D image projected onto a 2-D plane) images and are not planar. Therefore, it is difficult to directly compare the OH PLIF and chemiluminescence results. As mentioned earlier, in order to more accurately compare OH PLIF with OH* chemiluminescence, an Abel inversion transformation was performed on chemiluminescence images so that both techniques can be compared on a common basis.

The Abel transformation algorithm28 used a three point method and required, for symmetry purposes, to apply the code separately to the top and bottom halves of the average image, (i.e., symmetric about the centerline of the

-20

-15

-10

-5

0

5

10

15

20

0 5 10 15 20 25 30

Axial position (mm)

Rad

ial p

ositi

on (m

m)

-20

-15

-10

-5

0

5

10

15

20

0 5 10 15 20 25 30

Axial position (mm)

Rad

ial p

ositi

on (m

m)

Flame centerline

Injector #2

Injector #1

Fig. 11. Averaged OH* chemiluminescence edge location from 12 images (80 µs exposure). The ±standard deviations are also plotted in the figure.

Injector #2Injector face

45.72 mm (1.8 in.)15.24 mm(0.6 in.)

Injector #1

15.24 mm (0.6 in.)

Injector face

45.72 mm (1.8 in.)

Injector #2Injector face

45.72 mm (1.8 in.)15.24 mm(0.6 in.)

Injector #1

15.24 mm (0.6 in.)

Injector face

45.72 mm (1.8 in.)

highlow

Fig. 12. OH* images (80 µs exposure) for Injector 1 (left) and Injector 2 (right) at two window locations. Images have been taken in separate runs. The scale of intensity is different between upstream and downstream images.


10

chamber). After the code was implemented, the top and bottom Abel inversion images were combined. Figure 13 shows an average chemluminescence image before (left) and after (right) the Abel transformation. Using the Abel inversion images, the location of maximum intensity of the chemiluminescence was calculated and then compared to the location of maximum intensity for the OH PLIF images. For this analysis, the location of maximum intensity was calculated based on both the axial and radial directions (x, y), unlike the previous OH PLIF location of maximum intensity (Table 2) which was only based on the axial direction (x-direction).

Results for this analysis are presented in Figure 14 and clearly illustrate that the OH PLIF maximum intensity lies farther from the chamber centerline than that of the OH* chemiluminescence maximum intensity. There is more scatter in the maximum intensity location further away from the injector face, which can be expected, due to turbulent mixing and the break-up of the liquid sheet. In addition, the maximum intensity results mimick the OH PLIF edge results with respect to Injector 1 and Injector 2 location. Injector 1 maximum intensity lies closer to the chamber centerline than Injector 2. This can be linked to the larger cone angle on the liquid sheet for Injector 2 than Injector 1.

Figures 15 and 16 compare chemiluminescence with laser light scattering. It can clearly be seen in both figures that the highest intensity light emission occurs just as the liquid jet has broken up and more oxygen becomes available for combustion. This links peak energy release (i.e., peak chemiluminescence intensity) breakup length and overall system performance. Close to the injector, the chemiluminescence edge is virtually undistinguishable from the LLS edge which agrees with the finding of separation between chemiluminescence and on PLIF since the

Fig. 13. Average OH* chemiluminescence image before (left) and after Abel transformation (right) for Injector 1.

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 0.5 1 1.5 2 2.5 3

Axial Distance From Injector Face (cm)

Rad

ial D

ista

nce

From

Inje

ctor

Fac

e (c

m)

OH PLIF Injector 1

OH PLIF Injector 2

OH* Abel Injector 1OH* Abel Injector 2

Fig. 14. Comparison of OH PLIF and OH* chemiluminescence maximum intensity location from de-convoluted images.


11

PLIF was observed (Figures 8 and 9) to be located close to the methane flow and thus away from the LLS edge. The LSS images showed that breakup occurs earlier for Injector 1 and this seems to correspond to the sudden pickup in chemiluminescence observed for Injector 1 in figures 12 and 15. For Injector 2, the influence of sheet breakup is not as obvious which concurs with the assumption of less efficient atomization.

IV. Summary Swirl coaxial injectors are an important subset of liquid rocket injectors and have been employed in engines in

both the United States and Russia. One reason why they are of interest is due to their good mixing and atomization capabilities, which can lead to better performance over traditional shear coaxial injectors. Although this type of injector is promising, it is not as widely used in the U.S. rocket industry. This is due in part to a lack of empirical data on the swirl coaxial injector in both cold flow and hot fire tests. Hot fire testing is difficult due to the extremely harsh environment in which the experiments are occurring. Therefore, placing instruments such as probes inside the rocket chamber for data collection is unrealistic. This series of experiments circumvented this problem by providing optical access to the chamber and then employing several non-intrusive optical techniques such as OH planar laser induced fluorescence, OH* and CO2* chemiluminescence, laser light scattering and shadowgraph imaging to reveal the nature of the reacting flow.

As the main focus of the study, two different swirl coaxial injectors were tested in the Cryogenic Combustion Laboratory at Penn State University using gaseous methane and liquid oxygen as the propellants. The LOX was swirled through the central post of the injector while the methane was expelled into the chamber from the surrounding annulus. The two injectors differed only by the size of the methane annulus (Injector 1: Douter = 0.204 in. and Injector 2: Douter = 0.25 in.) and the same propellant flowrates were maintained for both cases.

These experiments confirmed that Injector 1 was better performing than Injector 2. Injector 1 had a higher C* efficiency (approximately 4% higher). This was expected due to the significantly higher Weber number and momentum flux ratio (a factor of 3 and 5 greater, respectively) of Injector 1. The performance difference was created only by a difference in methane velocity (and methane flow area) between Injectors 1 and 2.

The concurrent optical analysis demonstrated that it was possible to capture images of the reacting flow in the near injector region of the rocket chamber. Some quantitative parameters relating to the flame position were

Laser Light ScatteringCO2* Chemiluminescence Fig. 15. Injector 1 composite Image (top) of CO2* chemiluminescence (green color, original image bottom left) and Laser Light Scattering (red color, original image bottom right).

Laser Light ScatteringOH* Chemiluminescence Laser Light ScatteringOH* Chemiluminescence Fig. 16. Injector 2 composite image (top) of OH* chemiluminescence (green color, original image bottom left) and Laser Light Scattering (red color, original image bottom right).


12

measured using OH PLIF and OH* and CO2* chemiluminescence images. In addition, a qualitative analysis was completed by examining the structure of the liquid core through the use of laser light scattering and shadowgraph imaging and also by using composite pictures of simultaneous images from two different optical techniques.

It was found that for both injectors the position of the liquid sheet directly affects the chemiluminescence zone location and shape. Close to the injector face the chemiluminescence was located very close to the liquid sheet but showed significant expansion and increase in intensity when atomization became important further downstream. This was especially noticeable in the case of the better performing injector. For both injectors a significant difference was found in the location of the maximum of the deconvoluted chemiluminescence signal and OH PLIF intensity since the OH PLIF signal was found to be located close to the methane flow.

The better performing injector was characterized by a more compact sheet geometry and combustion zones, the absence of OH PLIF signal away from the center line of the rocket, and shorter and thinner OH PLIF zones. The chemiluminescence imaging showed increased intensity and flame area when the liquid sheet starts to disintegrate. This phenomenon was more noticeable for the better performing injector which is the injector expected to have better atomization. The light scattering data showed that the liquid core of the better performing injector disappears earlier than that of the poorer performing injector confirming the superior atomization and mixing for the better performing injector.

In general, the application of optical diagnostics, particularly those involving simultaneous measurements using two different techniques, provided detail information on the spray and flame structure for each injector. Continuation of similar studies involving other injector geometries, operating conditions and diagnostic techniques appears to be justified based on the results of the present study.

Acknowledgments The authors gratefully acknowledge support of this work by NASA under the Constellation University Institute

Program. The authors would also like to acknowledge Mr. Larry Schaaf for his invaluable assistance with the experiments.

References

1Pelaccio, D., Jacobs, M., Collins, J., and Scheil, C., “Engine System Assessment Study Using Martian Propellants,” AIAA Paper 92-3446, AIAA/SAE/ASME/ASEE 28th Joint Propulsion Conference and Exhibit, Nashville, TN., July 6-8, 1992.

2Abramovich, G. N., “Theory of the Centrifugal Nozzle,” Monograph – Industrial Aerodynamics (Russian), ZAGI, 1944, pp. 18.

3Novikov, I. I., “Atomization of Liquids by Centrifugal Nozzles,” (Translated from Russian, Journal of Technical Physics, Vol. 18, No. 3, 1948, pp. 345-354), The Engineer’s Digest, Vol. 10, No. 3, 1949, pp. 72-74.

4Gill, G. S., and Nurick, W. H., Liquid Rocket Engine Injectors, NASA SP-8089, National Technical Information Service, 1976.

5Bazarov, V., “Self-Pulsations in Coaxial Injectors with Central Swirl Liquid Stage,” AIAA Paper No. 95-2358, 31st AIAA/ASME/SAE/ASEE Joint Propulsion Conference and Exhibit, San Diego, CA. July 10-12, 1995.

6Dombrowski, N., and Hasson, D., “The Flow Characteristics of Swirl (Centrifugal) Spray Pressure Nozzles with Low Viscosity Liquids,” AI Che Journal, Vol. 15, No. 4, 1969, pp. 604-611.

7Tate, R. W., and Marshall, W. R., Jr., “Atomization by Centrifugal Pressure Nozzles – Part I and II,” Chemical Engineering Progress, Vol. 49, No. 4, 1953, pp. 169-174 (Part I), and Vol. 49, No. 5, 1953, pp. 226-234 (Part II).

8Fogler, B. B., and Kleinschmidt, R. V., “Spray Drying,” Industrial and Engineering Chemistry, Vol. 30, No. 12, 1938. 9Doumas, M., and Laster, R., “Liquid-Film Properties for Centrifugal Spray Nozzles,” Chemical Engineering Progress,

Vol. 49, No. 10, 1953, pp. 518-526. 10Longwell, J. P., “Fuel Oil Atomization,” Ph.D. Dissertation, Massachusetts Institute of Technology, 1943. 11Rice, E. J., “The Effect of Selected Fluid Parameters on Spatial Drop Size Distribution,” Ph.D. Dissertation, University of

Wisconsin, 1966. 12Kychakoff, G., Howe, E. D., Hanson, R. K., and McDainel, J. C., “Quantitative Visualization of Combustion Species in a

Plane,” Applied Optics, Vol. 21, 1982, pp. 3225-3227. 13Long, M. B., Webber, B. F., and Chang, R. K., “Instantaneous Two-Dimensional Concentration Measurements in a Jet

Flow by Mie Scattering,” Applied Physics Letters, Vol. 34, 1979, pp. 22-24. 14Long, M. B., Fourguette, D. C., Escoda, M. C., and Layne, C. B., “Instantaneous Ramanography of a Turbulent Diffusion

Flame,” Optics Letters, Vol. 8, 1983, pp. 244-246. 15Moser, M. D., Pal, S., and Santoro, R. J., “Laser Light Scattering Measurements in a GO2/GH2 Uni-Element Rocket

Chamber,” AIAA 95-0137, 33rd Aerospace Sciences Meeting and Exhibit, Reno, NV, January 9-12, 1995.


13

16Rahman, S. A., “Primary Atomization Study of a Swirl Coaxial Liquid Propellant Rocket Injector,” Ph.D. Dissertation in Mechanical Engineering, The Pennsylvania State University, August 1997.

17Snyder, R., Herding, G., Rolon, J. C., and Candel, S., “Analysis of Flame Patterns in Cryogenic Propellant Combustion,” Combustion Science and Technology, Vol. 124, 1997, pp. 331-370.

18Singla, G., Scouflaire, P., Rolon, C., and Candel, S., “Transcritical Oxygen/Transcritical or Supercritical Methane Combustion,” Proceedings of the Combustion Institute, Vol. 30, 2005, pp. 2921-2927.

19Santoro, R. J., Pal, S., and Woodward, R. D., “Rocket Testing at University Facilities,” AIAA 2001-0748, 39th AIAA Aerospace Sciences Meeting and Exhibit, Reno, NV, January 8-11, 2001.

20Moser, M. D., Merenich, J. J., Pal, S., and Santoro, R. J., “OH-Radical Imaging and Velocity Field Measurements in a Gaseous Hydrogen/Oxygen Rocket,” AIAA 93-2036, AIAA/SAE/ASME/ASEE 29th Joint Propulsion Conference and Exhibit, Monterey, CA, June 28-30, 1993.

21Pal, S., Moser, M. D., Ryan, H. M., Foust, M. J., and Santoro, R. J., “Flowfield Characteristics in a Liquid Propellant Rocket,” AIAA 93-1882, AIAA/SAE/ASME/ASEE 29th Joint Propulsion Conference and Exhibit, Monterey, CA, June 28-30, 1993.

22Beisler, M. A., Pal, S., Moser, M. D., and Santoro, R. J., “Shear Coaxial Injector Atomization in a LOX/GH2 Propellant Rocket,” AIAA 94-2775, AIAA/ASME/SAE/ASEE 29th Joint Propulsion Conference and Exhibit, Indianapolis, IN, June 27-29, 1994.

23Moser, M. D., Pal, S., and Santoro, R. J., “Uni-Element Rocket Studies,” 1994 Conference on Advanced Earth-to-Orbit Propulsion Technology, NASA Marshall Space Flight Center, AL, May 17-19, 1994.

24Rahman, S. A., Pal, S., and Santoro, R. J., “Swirl Coaxial Atomization: Cold-Flow and Hot-Fire Experiments,” AIAA 95-0381, 33rd Aerospace Sciences Meeting and Exhibit, Reno, NV, January 9-12, 1995.

25Yeralan, S., Pal, S., and Santoro, R. J., “Experimental Study of Major Species and Temperature Profiles of Liquid Oxygen/Gaseous Hydrogen Rocket Combustion,” Journal of Propulsion and Power, Vol. 17, No. 4, 2001, pp. 788-793.

26Anderson, W. E., Ryan, H. M. and Santoro, R. J., Impinging Jet Atomization, Chap. 8, in Liquid Rocket Engine Combustion Instability, V. Yang and W. Anderson, eds., Progress in Astronautics and Aeronautics, Volume 169, American Institute of Aeronautics and Astronautics, Inc., Reston, VA (1995).

27Woodward, R. D., Miller, K. L., Bazarov, V. G., Guerin, G. F., Pal, S., and Santoro, R. J., “Injector Research for Shuttle OMS Upgrade Using LOX/Ethanol Propellants,” AIAA 98-3816, 34th Joint Propulsion Conference and Exhibit, Cleveland, OH, July 13-15, 1998.

28Dasch, C. J., “One-Dimensional Tomography: A comparison of Abel, Onion-Peeling, and Filtered Back Projection Methods,” Applied Optics, Vol. 31, No. 8, 1992, pp. 1146-1152.

View publication statsView publication stats
https://www.researchgate.net/publication/268259345

Documents

Experimental Liquid Rocket Swirl Coaxial Injector Study Using Non … · American Institute of ... 4.14 MPa and propellant mass flowrates of 0.118 kg/s and 0.039 kg/s for ... it is