A Review Towards the Design Optimization

of High Performance Additively

Manufactured Rotating Detonation Rocket
Engine Injectors
Thomas W. Teasley1, Christopher S. Protz2, Andrew P. Larkey3,
Benjamin B. Williams1, and Paul R. Gradl4.
NASA Marshall Space Flight Center, Huntsville, AL, 35812, Madison
Abstract
Rotating Detonation Rocket Engines (RDRE) have been marketed primarily for their higher
specific impulse potential over constant pressure (CP) liquid rocket engines. However, several other
performance advantages exist with RDREs over CP engines such as heat transfer advantages for gas
expander cycle, increased completeness of combustion at low chamber L*, compact engine design,
reduced coolant channel pressure drop potential, and improved injector C* performance. NASA has
paved the way for liquid engine system performance enhancement since the Apollo program and
continues to do so with metal additive manufacturing (AM), super-alloy materials, and advanced
propulsion concepts. A team of propulsion development engineers at NASA are in the process of
developing high-performance 7K lbf class RDRE hardware for their potential use in lander, upper stage,
and even launch vehicle applications. Clear advantages have been demonstrated with AM including
program cost and schedule reductions of up to 50%. It is well known that injector performance is integrally
linked to the global performance of a combustion device. This is especially the case for RDREs since
detonation stability is heavily dependent on the mixedness of propellants. A major program goal is to
rapidly produce ultra-high-performance AM injectors. This paper reviews the available literature on liquid
rocket injector design optimization as well as the experimental work conducted to date on injectors tested
in RDREs. Major lessons learned are document and suggestions given towards the design of high-
performance liquid RDRE injectors. In addition, the integration of metal AM into the design of liquid RDRE
injector schemes is discussed. Finally, several candidate AM RDRE injector elements were produced to
obtain their diodicity and cold flow characteristics.

I. Nomenclature
AM = Additive Manufacturing
CWN = Channel Wall Nozzle
DED = Directed Energy Deposition
EB = Electron Beam Welding
GH2 = Gaseous hydrogen
GRCop-84 = NASA GRC Copper-alloy (Cu-8 at.% Cr-4 at.% Nb)
H2 = Hydrogen
ID = Internal Diameter
K-lbf = thousand pounds of force (thrust)
LCF = Low Cycle Fatigue
LOX = Liquid Oxygen

1
Combustion Devices Engineer, Component Technology Branch
2
Team Lead Combustion Devices Engineer, Component Technology Branch
3
Pathways Intern, Combustion Devices Engineer, Component Technology Branch
4
Senior Combustion Devices Engineer, Component Technology Branch, Associate Fellow, AIAA

1
LWDC = Laser Wire Direct Closeout
MCC = Main Combustion Chamber
MIG = Metal Inert Gas Arc-welding
MSFC = Marshall Space Flight Center
MTD = Manufacturing Technology Demonstrator
PBF = Powder-bed Fusion
psi = Pounds Per Square Inch
RDRE = Rotating Detonation Rocket Engine
RDE = Rotating Detonation Engine
RDC = Rotating Detonation Combustor
SLM = Selective Laser Melting
TCA = Thrust Chamber Assembly
TIG = Tungsten Inert Gas
WAAM = Wire Arc Additive Manufacturing

II. Introduction
Detonations have been substantially researched in ducts, area contractions, and various combustor
geometries. Their underlying physics are well understood with classical models that have been long since
established. The rapid increase in pressure across a detonation front minimizes enthalpy generation
causing thermal efficiency to reach a maximum when compared with the Brayton or Atkins cycles for
constant pressure combustion. Harnessing this efficiency gain for practical use in modern combustion
technology is another challenge altogether. The common solution for maintaining a stable detonation is by
making use of an annular combustion chamber with axial or radial injection of propellants. Numerous
experimental works have been conducted to date in an effort to understand the effects of geometry, scaling,
and operability of this type of detonative combustor configuration. Not to mention the sheer magnitude of
modeling efforts conducted to date. Regardless, these works have substantially researched the efficacy of
using detonations to increase engine efficiency and are currently at a crossroads in terms of their
development.
It is commonly cited that the peak benefit to detonative combustion devices is the promised ~15% boost
in specific impulse (Isp) compared to the maximum theoretical Isp in an equivalent constant pressure (CP)
combustor. However, other major performance characteristics are advantages with these engines rather
than quantities such as Isp and thrust. For one, these engines are often much more compact allowing for
reduced hardware mass a geometry. The completion of combustion has been shown to occur much faster
than CP combustors and thus drastically reduces L* requirements as well as the potential to maximize
characteristic velocity (C*) efficiency. Other works have suggested that these combustion devices could
produce lower overall heat fluxes and lower total heat load to active cooling geometries. In addition, lessons
learned from the literature indicates that the heat flux curve in a detonative combustor may decrease as the
flow gets closer to the chamber exit. This would allow for the design of ultra-low pressure drop integrated
coolant channels, which is another major performance advantage for this engine type. Each of these
performance criteria make a strong case for investment into detonative combustion device technologies.
With that said, no single literature source summarizes the design optimization procedure for a detonative
combustor or detonative liquid engine system. It is the goal of this paper to review, discuss, and summarize
major findings in the available literature for;
1. The optimization of an annular type rotating detonation rocket engine (RDRE) utilizing cryogenic
liquid/gas injectors.
2. Discuss the desirable performance criteria for thermal steady state RDRE liquid injector
operation.
3. Characterize the feasibility and advantages of using additive manufacturing techniques for use in
a high thrust application RDRE injector.

The term rotating detonation combustor (RDC) or even rotating detonation engine (RDE) are general
terms used interchangeably in the literature that refers to any combustion device utilizing a continuous
detonation front to achieve combustion. This device could use air or a stored oxidizer while a rotating
detonation rocket engine (RDRE) is restricted to utilizing a stored oxidizer that is limited in supply to
achieve continuous rotating detonations. The majority of the experimental work conducted to date on

2
RDC’s utilizes short duration heat sink type continuous detonation. Several examples of which are the
works of [1]–[8] and many others. In addition, most of these efforts utilize air as the oxidizer including [9]–
[12] and many others. The focus of this work is to produce hardware design schemes that successfully
demonstrate flight practical hot fire burns while maintaining operable hardware lifecycle performances.
The means by which this will be achieved is the incorporation of radially running square coolant channels
directly integrated into a laser- powder bed fusion built (L-PBF) GRCop-alloy annular chamber. NASA’s
Marshall Space Flight Center (MSFC) has pioneered the use of regeneratively cooled metal additive
manufactured (AM) thrust chamber hardware for over a decade and has successfully demonstrated this
hardware at stoichiometric conditions for cumulatively tens of thousands of seconds and thousands of
starts [13], [14]. Several images of these hot fire tests, in most cases using fully AM thrust chamber
assemblies (TCA), are shown in the figure below.

Figure 1. Hot fire testing of metal additively manufactured thrust chamber assemblies including additive
injectors, chambers, igniters, and nozzles in various propellant combinations and thrust classes.

Several of these metal AM components are operated in extreme environments often at

stoichiometric and even lean combustion for hundreds of seconds at a time. The highly conductive
material GRCop-84/42, a NASA invented super-alloy, has been the basis for this success. Other
materials that are easily used in additive manufacturing include Inconel 625/718, JBK-75, Monel K500,
NASA HR-1, and numerous other refractory and super-alloy materials. A list of metal additive materials
commonly used in industry and grouped by their base material are shown in the figure below.

3
 Figure 2. Metal additive materials commonly used in industry.

Of particular interest for use in RDRE injector design are the high-temperature nickel alloy
material Inconel and iron alloy material NASA HR-1 / JBK-75. These materials have been used in
additively manufactured injectors and nozzle hardware under numerous test programs at MSFC including
but not limited to PK058, PI051, PJ030, PJ141, PK020, and PJ129. Several super alloy metal additive
injectors are shown in the figure below.

Figure 3. Examples of various thrust class and material metal additively manufactured liquid rocket injectors
hot fire tested at NASA MSFC.

All tested injectors shown above demonstrated exceptional life cycle performances with
cumulatively thousands of starts and tens of thousands of seconds of mainstage duration. For this reason
alone, these materials would be an ideal candidate for use in RDRE injectors. The reader is referred to
[15] which gives a comprehensive summary of AM injectors hot fire tested at NASA MSFC. Since
reasonable life cycle performance is desired during a long duration burn of an RDRE, these materials
would stand the best chance for demonstrating hardware survivability. Several images of example nickel-
alloy injector designs that have been tested at NASA MSFC and demonstrate exceptional life cycle
performances are shown in the figures below.

4
 Figure 4. JBK-75 additively manufactured shear coaxial injector. (Left) Prior to testing under PJ141 and
PK020. (Middle) Post-testing with More than 3130 seconds of cumulative duration and 21 starts over 2 test
series in LOX/GH2 at a nominal Pc of 1000 psi and mixture ratio of close to stoichiometric. Slight face
erosion shown in the red circle but not deleterious to injector operation. (Right) Hot fire test demonstration
with NASA carbon composite nozzle.

Data from this injector scheme was used to design a co-flow type injector which is discussed later on in this work.
The injector shown above used a JBK-75 additive material produced using L-PBF with a 32-
element shear coaxial heritage design. This injector demonstrated great life cycle performance with no
element erosion and minimal face erosion circled in red. This injector produced a very high C* performance
of nearly 100% theoretical in some test cases. This is however, typical of LOX/GH2 injectors when coupled
with high L* chambers. This was also a lower thrust class injector producing thrusts upwards of 2500 lbf.
An example of a 7K lbf class lander engine Inconel 625 pentad impingement injector is shown in the figure
below.

Figure 5. Inconel 625 L-PBF 7K lbf class pentad impinging injector face. (Left) pre-hot fire testing with
NiCrAlY thermal barrier. (Middle) More than 720 seconds of cumulative hot fire duration and greater than
32 starts in LOX/GCH4 at a nominal Pc of 750 psi and mixture ratios up to 3.6. (Right) Hot fire test
demonstration of AM impinging injector, L-PBF GRCop-42 chamber, and LP-DED NASA HR-1
regenerative nozzle.

This injector design produced high C* performances upwards of 99% theoretical at throttled
chamber pressures between 100 psi and 800 psi. Alteration of the hot wall surface finish later showed a C*
performance improvement of around 2% at the same operating conditions. Results yet to be published.
This in and of itself lays the groundwork for tailor made performance improvement for metal AM components
in liquid rocket engines. Experimental data from this injector type was used as a guide towards the design
of a similar “close triplet” scheme described later in this work.
These two previously tested injector configurations show reasonably high C* performance and
lifecycle performances that can be achieved with additive materials under extreme hot fire conditions. An
RDRE variant AM injector will need to have similar mixing, atomization, and life cycle performances to
successfully demonstrate flight practical hot fire test durations as well as achieving the ultimate goal of
demonstrating equivalent or higher engine specific impulse.
It is not well understood or even completely known if modern combustion device materials and
cooling schemes will be able to handle the extreme environments that an RDRE will produce. This is

5
especially the case for flight practical operating durations needed during space missions and launch
vehicles. For this reason, a NASA sponsored program denoted the Advanced Rotating Detonation Variant
Rocket Combustor (ARDVARC) program was established under the Space Technology Mission
Directorate (STMD) announcement for collaborative opportunity (ACO). This effort will characterize the
life cycle performances of 7K lbf class AM injectors, chambers, and nozzles to the extreme environment
of an RDRE over long duration hot fire tests. There are a substantial number of considerations for
hardware development of high thrust regeneratively cooled and additively manufactured hardware as
illustrated in the considerations process diagram figure below.

Figure 6. Consideration diagram for development of an optimized performance RDRE.

This work primarily focuses on RDRE injector hardware development in preparation for the
experimental program. Several topics and considerations are discussed in reference to optimizing
performance for the detonative combustion cycle. Design recommendations are made and incorporated
into the design schemes presented.

III. Background and Literature Review

Prior to the design process of an optimized RDRE injector, the available literature was reviewed
on annular detonation injectors. Second, an in-depth review on the available constant pressure (CP)
combustion device injector performance was also conducted to identify attributes of high C* efficiency
injector elements. These reviews are presented below.

A. RDC/RDE/RDRE Injector Research to Date

A limited amount of work has been conducted on detonation combustion device injection. Several
articles are in the available literature that describe how injection properties, injector design, and injection
mixing effects RDC operation. It was found that air breathing injection schemes dominate the experimental
literature in terms of most tests conducted. The typical scheme used is the annular slot injector or annular

6
cross flow orifice injector. It is also very common to see variations in radial and axial injection schemes.
Two examples of injection feed systems are shown in the figure below from [1], [16].

Figure 7. Injection feed systems tested and reported in (left) [17] and (right) [1]

The propellant phase commonly employed is the gas/gas bi-propellant combination with
hydrogen/air and methane or natural gas (NG) /air is typical. Other works have utilized methane, NG,
hydrogen, propane, and even refined petroleum (RP) in conjunction with gaseous oxygen.

Injector scheme aside, several articles were available that looked at variable flow area at constant
mass flow rate. Bennewitz et al. 2018 [18], conducted an experimental test series on an uncooled rotating
detonation engine at AFRL. The injector flow area and thus pressure drop was varied as well as the mass
flow rate across a broad range of mixture ratios using GCH4/GOX as the propellant. Key results noted were
a maximum of thrust and specific impulse at or near stoichiometric mixture ratios and a linear increase in
thrust with increasing injection mass flow rate. Another interesting feature to note, though not explicitly
stated, the number of wave modes present in the combustor for a given test was dependent upon the
injection mass flow rate or unburned injected flow velocity.
It was apparent that the number of wave modes present balanced with the propagation velocity of
the detonation wave and the annular geometry of the combustor. This lends credibility to the notion that if
detonation is achieved in an RDC, sustainability of the wave is not so sensitive to injection properties and
channel geometry that the wave will simply be quenched. Instead, the wave will divide or merge the
balancing number of wave modes to become anchored in the annulus. Further observations of the
experimental data include the following; pressure tap data suggests anchoring of the detonation wave
occurred very close to the injector face, specific impulse is maximized when wave speed is the highest and
thus when the fewest number of waves are present, a larger injection flow area seems to yield higher wave
speeds and higher specific impulse but may be predominantly due to constant mass flow rate error. Finally,
the authors note that the engine tended to operate in a specific frequency range which may be caused by
a coupling between the RDC natural acoustic modes and the detonation modes.
Anand and Gutmark 2018 [19], summarized the RDC work conducted at the University of
Cincinnati. This paper focuses on the experimental results of both an annular RDC and a hollow RDC. It
was found that lower injection areas and higher-pressure ratios contribute to an optimal RDC injector
design. Both of these variables contribute to a net higher Injector fluidic impedance thus reducing the
backflow potential commonly encountered in these combustion devices. Furthermore, a high number of
injector ports with small diameters compared to a low number of injector ports with larger flow area, thus
holding total injection area constant, was better for sustaining detonation. The cause of this is primarily due
to refresh rates and mixing efficiencies. A higher number of ports increases the mixing while simultaneously
reducing backflow potential thus improving fresh propellant injection timing. Interestingly, the authors also
found that the wave mode operation of detonations to be dependent on the pressure ratio across the injector
area. This could be due to a change in mixedness of the propellants which has been previously found to
cause wave mode shifts [16].

7
 Full propagation of rotating detonation was observed for pressure ratios that were greater than the
chocked criteria pressure ratio. At lower pressure ratios, azimuthal pulsed detonations were observed. It
was determined that this was caused by a “chugging” type phenomena where combustion products back
flowed through the injector and out again. Finally, this work summarized the wave mode operation space
of an RDC. In general, stable wave mode operation occurs at equivalence ratios between 0.8-1.75 and
from mass flow rates of approximately 0.2 kg/s and up. These values are likely specific to a certain
propellant combination and RDC geometry but nevertheless demonstrates that a minimum mass flow rate
and a bounded equivalence ratio band allows for steady and stable detonation to process in an annular
RDC.
Since it has been inferred that propellant mixing is of such high importance to wave mode operation,
[20] conducted an experimental study on the effects of three different injector configurations in an RDRE.
These element configurations from [20] are shown in the figure below.

Figure 8. Impinging injector element schemes a) aligned 72 element pairs ~0.131” spacing, b) 36 element pairs
~0.262” spacing, c) misaligned 36 element pairs ~0.131” spacing [20].

The authors found that poor mixing sustained fewer waves in the annulus and causes a breakdown
of detonation mode structure, less well-defined waves, and counter propagating waves. In addition, both
thrust and wave speed were reduced significantly with lower mixing quality. As the mass flow rate and thus
mass flux increased through the annulus, the number of waves present in the annulus increased linearly to
seven waves at a mass flow of ~1.3 lbm/s. It is not discussed if the injectors were chocked or not at any
given mass flow rate. Some inferences can be made in the data presented in this work to that end. The
wave mode operation does increase linearly until about 0.75 lbm/s for the 72-element injector. The wave
mode operation appeared to remain constant at five waves until the mass flow rate increases to ~1.3 lbm/s.
The tests conducted at this mass flow shows operation at both six waves and seven waves for two different
tests. From these observations, it could be that the injector was chocked at and above the 0.75 lbm/s mass
flow rate. Since the injection area is not known, this is not certain. The authors did give measured CTAP
pressure data for increasing mass flow rates. As expected, the averaged chamber pressure did increase
linearly with increasing mass flow rate. Since a single RDRE geometry is presented in this work and as
discussed above the wave mode operation ceases to increase linearly at about 0.75 lbm/s, the wave mode
operation may plateau and then vary between six and seven waves for the last two cases. At the mass flow
rate of 0.75 lbm/s, the measured CTAP pressure is above ~30 psig, which is greater than ~2X atmospheric
pressure. This would likely mean that annulus exit was chocked and thus the injector outlet would also be
chocked since the injection area is very likely less than the annulus exit area. It is theorized from these
results that wave mode operation in an RDC is not as strongly influenced by the chamber pressure as it is
by the mixing efficiency, injection velocity, and exit chocked condition.
The last interesting observation made by the authors was that degraded mixing resulted in higher
average chamber pressure measurements due to shifting of the reaction zone closer to the measurement
location. It can be concluded from this work that injector design for high mixing efficiency is of extreme
importance for RDC operation.
Walters et al. 2019 [11], conducted an experimental investigation using natural gas and pre-heated
air as well as two different injector configurations. The first was an axial slot injector and the second was
an dubbed a sting injector. Both configurations are shown in the image below from [11].

8
 Figure 9. a) Axial slot injector and b) axial sting injector from [11].

It was found that injector configuration (b) produced somewhat higher performance metrics than configuration (a).
The thrust efficiency was higher for this configuration and was approximately the same for larger cell sizes. It is
unclear to the authors why this was the case. Similarly, injector (b) produced higher C* efficiency band than injector
(a).
Fotia et al. 2016 [1], examined the performance gains of axial fed injection versus radially fed
injection. Typical injection systems of modern rocket engines employ axial feed systems that inject directly
along the fluid flow path through the main combustor. To date, very little work has been done to quantify
the effects of axial feed injection in RDREs. This study compared several configurations with JP-8 fuel and
hydrogen fuel. It was determined that an increase of ~15% in the corrected thrust was attained with axial
fed propellants in similar channel gap configurations when compared to radial feed systems.
Mizener and Lu 2016 [21], built an analytical model of an RDC to determine performance trends
with a parametric analysis. The model was designed based on integrated conservation equations and
varied several key annular channel designs and operational parameters. It was found that for optimal
performance the geometry should have a high injection pressure, low propellant temperature, and positive
injection swirl. It was chosen by the authors to model injection swirl likely due to the increased mixing
effects. Positive injection swirl likely increased the residence time and thus the mixing efficiency, though it
is uncertain if this level of detail was captured in the presented model.
Li et al. 2018 [22], investigated injection strategies for liquid-fuel RDEs. A premixed and non-
premixed combination of liquid jet A-1 fuel and air was injected into the chamber via a radial slot inlet
scheme. It was found that a rotating mode detonation formed in the startup and transitioned to an axial
pulsating mode using premixed propellants. In the non-premixed cases, similar results were observed. In
all cases, the detonation front was fairly weak with a low wave pressure gradient.
Stechmann et al. 2017 [16], conducted a high-pressure experimental investigation of an RDRE
using O2/methane, O2/NG, and O2/H2. The facility and test article utilized an oxygen pre-burner setup and
annular slot with various upstream and downstream fuel orifices in a crossflow arrangement to the oxidizer
slot. An image of the injection element arrangement is shown in the figure below from [16].

9
Figure 10. Arrangement of injector and channel schemes experimentally investigated in (Stechmann, et al.,
2017). On top left is V1.0 configuration, top right is the V1.1 configuration, and on the bottom are various
V1.2 configurations [16].

This work is the most applicable to the injector operating conditions expected during this program.
The injector orifice and slot sizing was altered as shown in the figure above to gauge the detonation stability
and engine performances. The element scheme shown on the top left (V1.0) is more similar to CP liquid
rocket engine elements. Detonation in the V1.0 and V1.1 configurations only occurred near shutdown when
the injector plenum pressures were low. It must be noted that a significant analysis was given on flame
holding and parasitic deflagration through short ignition delay. This was likely a problem experienced due
to the hot oxygen from the pre-burner greatly increasing the chance for parasitic deflagration. Other works
such as [23], were able to achieve rotating detonation in O2/H2 using much lower injection temperature
thus increasing the ignition delay. The V1.2 injector did successfully yield rotating detonation in all
configurations. It was concluded that the detonation behavior was highly sensitive to the injector and throat
configuration. The work of [16] and other works by this author are further reviewed in a later section.
Sosa et al. 2020 [23], conducted an experimental investigation with gaseous O2/H2 in an uncooled
annular RDRE. The injector configuration incorporated 72 aligned impingement ox and fuel jets with orifice

10
sizes of 0.9 mm and 1.1 mm in diameter. Consequently, this element scheme is identical to a configuration
described in [24]. An image of the test article and injection scheme is shown in the figure below.

Figure 11. RDRE test article hot fire tested in [23].

This injector configuration was tested up to just under a total mass flow rate of 1.2 lbm/s and produced 5
rotating detonation waves in all cases shown. CTAP pressures measured reached up to 153 psia. All cases
tested utilized some percentage of methane as a tracer to visualize the rotating detonation wave front.
As mentioned above, a similar experimental investigation by Hargus et al. 2018 [24], was conducted at
the Air Force Research Laboratory (AFRL) using an uncooled annular combustor with typical hot fire tests
lasting 1.25 seconds. In addition to the injector geometry presented in Sosa et al. 2020, another misaligned
geometry was considered with a 2.5 degree spacing which translates to roughly 0.065 inches. These two
element schemes are shown in the figure below from [24].

Figure 12. a) Aligned impingement injector element scheme, b) non-impinging and misaligned injector
element scheme [24].
The misaligned element scheme solely relies either on wall impingement and/or secondary detonation front mixing
while the aligned injector scheme relied on jet impingement in addition to secondary mixing processes. This
investigation was also capable of altering the element flow areas from 1, 1.5, 2, and 2.5 times the smallest area. A
suite of tests were conducted across equivalence ratios and mass flow rates. The authors note that the performance of
the engine varied only slightly with altered orifice area and equivalence ratio. It was also noted that substantial drops
in plenum pressure did not appear to be detrimental to engine performance. A closer look at the data suggests that the
2X area elements performed the highest in the range of typical liquid rocket engine equivalence ratios but this may
not be statistically significant. This may be due to some optimization between element spacing causing increased
mixedness as the elements encompassed a large amount of the injector faces “dead space”. Once the orifice area
became large enough, atomization may have been adversely affected. As for the injector element type, impingement
or misaligned, there is a notable drop in engine performances with the misaligned elements. As tests moved away
from stoichiometric conditions, the performance of both element schemes showed similar performances. The
performance metrics mainly assessed included thrust and specific impulse. The authors also note that mixedness may

11
compete with equivalence ratio for direct impacts on the detonation cell width. An image during hot fire test of this
RDRE configuration is shown in the figure below from [24].

Figure 13. AFRL GHKN RDRE during hot fire test [24].
Goto et al. 2019 [25], conducted an experimental investigation of an RDE with different nozzle and
injector configuration to asses performance. Two injector types were used including a triplet and doublet
element scheme. These schemes from [25] are shown in the figure below.

Figure 14. Injector element configurations tested under the experimental investigation in [25].

The doublet scheme used 120 sets of 1mm diameter fuel and oxidizer orifices. The triplet scheme used 72 sets of 1mm
fuel orifices and 1.4mm oxidizer orifices. A slot injector configuration was also included in the study, likely as a
control. Both the triplet and doublet injector schemes showed clear performance advantages over the slot injector.
Both schemes trended closely with the optimally calculated specific impulse. As the pressure ratio increased, the
specific impulse increased as expected but with the injector and chamber configuration producing nearly 100%
theoretical specific impulse in some cases. Not surprisingly, cases that employed a throat constriction performed very
well. It was interesting to note that a single case of the triplet injector with open throat produced higher specific
impulse than the highly constricted throat configuration at nearly the same operating conditions. These results were,
however, within measurement error. The authors go on to note that characteristic velocity was nearly identical for all
cases of the triplet and doublet schemes at elevated conditions. The slot injector on the other hand was consistently in
the range of 15-40% lower in normalized chamber pressure at the same conditions. Finally, the authors note that the
doublet configuration cause damage to the combustor wall likely due oxidizer rich environment near the chamber wall
causing substantial erosion. The authors thus recommend a triplet scheme due to improved wall compatibility.
The study above as well as the work of Ishihara et al. 2017 [26] utilized the same triplet injector configuration in
addition to a slot injector shown in the figure below from [26].

12
 Figure 15. Injector element schemes tested under [26] (top) and region of anchored detonation causing
damage to the chamber carbon composite wall using the triplet configuration.

Consequently, the figure below also reveals the location of where the detonation was anchored confirmed by an
indentation from where the detonation resided causing damage to the combustor wall. This piece of information is
invaluable as it suggests that the detonation can be stood off from the injector face by means of moving the
impingement location similarly to altering the location of the anchored flame in a CP combustion device. This work
also reveals the length of the detonation front which is not typically reported in the experimental literature. In this
case, the detonation “groove” in the chamber wall appears to be about 0.4 inches in length. Of course, this must be
confirmed prior to making significant assumptions.
Lim 2019 [27], conducted an exhaustive study (as is typical with a PhD dissertation) on the responds of liquid
injector orifices to transverse detonation waves at elevated pressures and hot fire tested an RP-2/gaseous oxygen RDE.
Several performance parameters of the RDE were quantified including heat flux, multiple injector element schemes,
thrust and Isp, and visual confirmation of detonation stability. Tests were conducted up to 7.7 lbm/s with static
chamber pressures up to 258 psia. A significant study on liquid injector response to a passing transverse detonation
wave was given. Summarizing the findings, the author found a correlation of backflow potential and orifice geometry.
Several different injectors were hot fire tested and their performances compared. The thrust was found to be
consistently between 85% and 95% that of a CP equivalent engine operating at 100% efficiency.

Several articles have been reviewed in an attempt to understand how to optimally design an injector for
operation in an RDRE. Several lessons learned were noted and design schemes that appeared to yield
higher overall performances. These lessons learned are documented below.
• Higher thrust and stronger wave fronts are observed for well mixed propellants.
• High element density with a large number of elements in a close spacing produced higher overall
engine performance and detonation stability.
• Small inlet orifices yield higher fluidic impedances reducing the likelihood of back flow and
increasing the refresh rate for identical back pressures.
• Axial fed propellants produce higher overall engine performances than radial fed propellants.
• Low feed pressures don’t appear to be detrimental to engine performance.

13
 • A chocked annulus exit and chocked flow at the injection orifices may isolate the detonation fronts
and increase detonation stability.
• High mixedness tends to yield fewer but stronger detonation waves.

The literature presented above appears to suggest that RDRE injectors are not so dissimilar from
CP combustion devices. They do also rely on highly mixed and well atomized propellants to improve overall
combustion performance as well as sustain the detonation.

B. Review of Constant Pressure Combustion Device Injector Design

In an effort to better understand how to optimize the performance of an RDRE injector, and since
RDREs are not so dissimilar from CP combustion devices, the available literature was reviewed on the
design optimization of traditional liquid rocket injector elements. Several of these traditional element
schemes are shown in the figure below.

Figure 16. Traditionally manufactured and most common liquid bi-propellant injector elements schemes.
Blue represents liquid phase propellants and red represents gas phase propellants.

A particular focus was placed on the review of co-flow or co-axial elements as they were the most
prevalent in the literature and in many articles, the highest performing. Impingement type elements were
also reviewed from the available literature but to a lesser extent.
Coaxial Type Injector Elements
The coaxial injector configuration is a common type of non-impinging injector element, well known
for its use on the Space Shuttle Main Engine. It is noted for excellent mixing and atomization, low
pressure drop, and proven dependability [28]. The element typically consists of a central low velocity
liquid oxidizer surrounded by a high velocity gaseous fuel. Mixing, atomization, and mass distribution rely
upon shearing or swirled cross-flow of the liquid with the gaseous exterior. Maintaining a high velocity
gradient between the gaseous fuel and liquid oxidizer is crucial to ensuring that proper breakup and
atomization occurs [29]. Mixing plays a significant role in determining the performance and wave
propagation characteristics of RDREs [20]. While research on RDRE coaxial injectors is sparse, coaxial
injectors for constant pressure combustors have been studied in great depths. An image representation of
a typical coaxial injector element is shown in the figure below from [30].

14
 Figure 17 Coaxial injector element and atomization spray field [30].
The state of available co-flow injector literature was reviewed for ways to optimize performance,
specifically C* efficiency which is a the confluence of atomization and mixing efficiency [31]. Several
articles reviewed provided a fundamental background for coaxial injectors [28], [29], [32]–[36], with
articles [20], [36], [37] specifically detailing the state of injectors in RDRE’s. Additive manufacturing
methods and applications were reviewed in articles [15], [37]–[42]. Aspects of atomization were
investigated in articles [20], [30], [31], [43]–[58], while flame dynamics were reviewed in [59]–[68] . Ignition
dynamics literature was found to be less prevalent but reported in [69], [70] as well as injector face heat
flux in [71]. Finally, supercritical propellant conditions are examined in [35], [48], [51], [65], [66], [72]–[75].
Atomization describes the ability of the injector to break down the propellants into small droplets,
where the droplets mix in the spray field, vaporize, and begin to combust. The atomization process
strongly depends upon injection conditions, propellant properties, and the local flow of the spray field. The
ability for a coaxial injector to atomize liquid propellants is essential to high-performance. The complexity
of the atomization process means that coaxial injector design must be based on both basic principles and
empirical rules, leading to numerous experimental and theoretical investigations.
Maintaining a high velocity ratio between the gaseous fuel and liquid oxidizer is critical for overall
performance by helping to properly atomize the propellants [28]. LOX and fuel velocities can be
influenced by several factors. These include injector geometry such as the LOX or fuel post diameters, or
system parameters such as propellant density and mass flow rate. Adding a tapered angle to the inside of
the LOX post has consistently been found to increase the mixing quality. Mixing can be further improved
by recessing the distance of the exit of the LOX post from the exit of the fuel post. Ideal performance
occurs when the recess length is approximately the same distance as the LOX post diameter [52]. Other
studies have been conducted to determine a specific model for predicting the diameter of atomized
droplets. It was found that droplet diameters of 1000 microns or less should be required, with an ideal
injector creating droplets smaller than 10 microns [46]. A test campaign completed by R.J. Burick of the
Rocketdyne Corporation found, for example, that a reduction in droplet diameter by a factor of 2 would
increase efficiency enough to allow a reduction of L* by a factor of 4 [31]. It should be noted that there are
many other non-dimensional numbers or ratios besides the velocity ratio that are commonly used to
evaluate injector performance in coaxial injector studies. The momentum flux ratio and Weber number
can play an important part in characterizing an injector [55].
Investigations of reactive sprays have shown a notable interaction of combustion and
atomization. This means that aerodynamic forces between the propellants are not the only thing
controlling atomization. The release of heat and combustion reaction products change the conditions for
atomization and droplet vaporization [61]. Flame dynamics and the type of fuel can have a further effect
on atomization. Different atomization regimes can occur depending on whether the flame is attached or
detached from the LOX post. Flame blowout and liftoff can initiate by a high velocity ratio and small LOX

15
post diameter. A tapered LOX post has been shown to improve atomization by inducing an oscillation in
the liquid jet core [31]. Consequently, this also reduces the vorticity region in the gap between the LOX
and fuel annulus where a flame would typically anchor [59]. Another study found that the high-pressure
flame emission spectra of LOX/CH4 is similar to that of LOX/H2. They found that as soon as the oxidizer
was supplied, the flame anchored in the wake of the LOX post at all observable operating conditions. The
flame shape itself was found to be influenced by the injection velocity ratio and momentum flux. Sonic
injection of both Ch4 and H2, can result in low velocity ratios and produce stretched flames, while a larger
momentum flux can cause a more confined spreading angle and constricted flame [60]. Another
experiment comparing the flames of H2 to CH4, found that under the same chamber pressure, the liquid
intact core length was shorter for CH4 than H2. This is attributed to the higher density of CH4. Higher
momentum flux ratio tends to promote jet instability and result in an earlier onset of jet disintegration. Key
findings in this work showed that as the weber number decreased, the flame lift off distance and the intact
core length decreased [61]. Flame-acoustic interaction was found to be most sensitive to changes in fuel-
oxidizer density ratio [63]. In a test of supercritical hydrogen-oxygen with a chamber pressure of 80-bar,
dynamic mode decomposition results showed that the flame dynamics were strongly influenced by the
LOX injector acoustics, whereas no flame response to injector acoustics was observed for a 50-bar
chamber pressure [66].
As for the optimized design of a coaxial element, a few articles stand out. The work of Oschwald
et al. 2008 [61], describes a number of advantageous design characteristics for coaxial injectors. First,
the atomization process is highly complex and can be described by the momentum flux ratio. For large
values, the liquid jet tends to result in earlier onset of disintegration due to imparted instability at the
surface of the jet. Similarly, the Weber number represents the relative disturbance of the gaseous co-flow
to the liquid jet surface. It has been observed that as Weber number increases, ligament and droplet
dimensions become smaller and secondary atomization is encouraged. Droplets have also been
observed being transported to a larger radial distance. Key findings in this work showed that as the weber
number decreased, the flame lift off distance decreased and the intact core length also decreased.
Finally, it was found that the intact LOX core length decreased when the LOX post thickness was
decreased. This also resulted in a thinner anchored flame at the post and in the stagnation region
downstream.
Next, the work of Burick 1972 [31], gave a comprehensive experimental review of co-flow type
injector design optimization. Several key takeaways can be gathered from this work. First, as the gas
density increased, so did the mixing efficiency and overall C* performance. A LOX post recess of
approximately one LOX post diameter resulted in higher overall efficiencies as described previously.
Element spacing plays a key role in the mixing process and generally speaking, a well distributed field of
elements produces higher efficiencies. Smaller LOX jet diameters produce smaller droplet diameters. As
the liquid jet velocity increases, the average droplet diameter decreases. High gas injection velocity
allows for improved mixing and secondary atomization of the liquid jet. Finally, small droplet diameters are
essential for high efficiency combustion.
Finally, several secondary observations have been made from the available literature. Mixing
quality increased with an increase in LOX post inner taper and injection gas density [52]. Nondimensional
numbers characterizing fluid-dynamic interactions are not sufficient to scale coaxial injector performance
in hot fire tests from one fuel to another [55]. If the liquid jet has a much higher velocity than the flame
speed, flame holding can only occur when the post thickness is greater than or on the order of the flame
thickness [59]. Under the same chamber pressure, the liquid intact core length was shorter for CH 4 than
H2 which is attributed to the higher density of CH4 [61]. The combustion efficiency of the subcritical oxygen
case had the highest performance with a slightly increasing efficiency as the injection velocity ratio
increased [72]. Higher chamber pressures correlate to higher efficiencies at the same droplet sizes [52].
In summary, there are two types of control that can be used to alter the performance of coaxial
injectors; geometry and flow parameters. These parameters are listed below.
1. A beveled internal LOX post seems to allow the liquid jet to disperse and impinge on the high
velocity gas sooner, thus reducing the in-tact core length.
2. A recessed LOX post of 1X post diameter seems to be optimal in imparting an oscillation in
the jet which allows for the jet to fan out and break up sooner.

16
 3. High velocity gas and low relative velocity liquid allows for quicker breakup of the liquid core.
4. A higher velocity liquid core produces better atomization but reduces the mixedness between
injector elements.
5. Swirling of the liquid core increases the mixedness and atomization and thus drastically
improves combustion performance.
In the case of the swirl coaxial element, put simply, swirling allows for the effective dispersion or
“thinning” of the liquid core into an annular sheet which is then effectively atomized in the crossflow of the
high velocity gaseous annulus. Consequently, this swirling also rapidly disperses the propellants so that
mixedness between elements is maximized. An image of the swirl coaxial AMDE injector from [15]
demonstrating this high degree of atomization and mixedness in cold flow testing is shown in the bottom
right image of Figure 20.
Impingement Type Injector Elements

Figure 18. Process flow diagram for considerations in combustion performance of an injector
assembly.

C. Integration of Additive Manufacturing into Injector Design

Several different coaxial type injectors have been tested at NASA MSFC outlined in [15]. Several
different thrust class, bi-propellant, and material AM co-flow injectors from [15] are shown in the figure
below.

17
 Figure 19. Various L-PBF Injectors manufactured and tested at NASA MSFC.
Several works have found clear advantages with using AM to produce injector hardware including reduced number
of parts, rapid manufacturing, equivalent performance to traditionally manufactured injectors, and reduced overall
costs of manufacturing [15], [76]. In addition, AM allows for the production of components with complex internal
structures not previously possible with single piece traditional manufacturing [76]. Even large scale AM injectors can
be produced where the work of [77] designed and hot fire tested a 35K lbf class upper stage AM injector.
Water flow testing to characterize the mixing and atomization has been a benchmark for injector design. This is
particularly the case for AM produced injectors since validation of free and clear channels is vital to their operation.
Often powder can remain in the elements, thus water flow testing can be used to confirm their flow uniformity. Several
images of AM injector water flow testing is shown below.

Figure 20. Water flow testing of various thrust class AM injectors.

Water flow testing can also give an engine designer a qualitative idea of its effective atomization and
mixing since a high degree of which would translate to high performance. The common notion for injector
design would be to simply print thousands of extremely small orifices that effectively atomize and
disperse propellants into the chamber to optimize combustion performance. However, there are still

18
limitations with metal AM that the design must be aware of. First, there is a lower limit of orifice size that
can be effectively printed. Laser powder bed fusion (L-PBF) would yield the lowest orifice size possible of
all AM build methodologies. This limit is further compounded by the requirement of the orifice being
cleared of powder in post processing. All powder must be cleared from the orifice prior to heat treatments.
In the case of materials that require stress relieving and solutioning heat treatments, the designer must
also consider the build orientation. For Inconel components, residual stresses must be relieved prior to
removal from the build plate. Thus the engineer must be able to remove powder from all orifices while the
part remains on the build plate. An example image of an Inconel 718 L-PBF produced NASA
impingement pentad injector on the build plate and off the build plate is shown below.

Figure 21. NASA 7K lbf LOX/methane pentad impingement Inconel 718 injector.

Both NASA and industry have been developing AM injector elements for the last several years.
Several works are available that describe AM injector performances under cold flow and hot fire
conditions, their development process, and empirical models for design iteration.
Sims and Hulka 2019 [78], developed AM liquid propellant swirl injector flow elements. This work
documented their design, modeling, cold flow testing, hot fire testing, and lessons learned. It was found
that wall powder buildup after the L-PBF process caused non-uniformities in the spray cone when the slot
was very small. It was also found that traditional models could not accurately predict the spray cone
evolution to full flow but a Reynolds number like nondimensional value may exist to predict such
evolution. This information was then used to develop a multi-element injector which yielded high
performances at 50% power level.

D. Traditional and Potential Injector Element Flow Properties

From the start, it was apparent that the element design trade space could be significantly broadened by
integrating metal AM technology with the design of an injector. First, traditionally manufactured injectors
typically rely on circular ports to transport propellants into the combustion chamber. The flow vector of
traditional element schemes often incorporates axial injection with the exception of the like or unlike
doublet element scheme. A representation of element orifice flow vectors relative to the axial chamber
geometry are shown below.

Figure 22. Traditionally manufactured injector element orifice flow vector relative to the axial chamber
coordinate system.

19
These element schemes often rely on high gaseous fuel velocity or rapid chemical kinetic rates to effectively breakup
the axially injected liquid oxidizer jets. It has been observed during hot fire testing at NASA MSFC that low injector
pressure drop conditions on the gas fuel side can produce low relative C* performance due to ineffective breakup of
the liquid jet. This has generally been observed to be the case for all element types that employ axial liquid injection.
With slower chemical kinetic propellants, these performance detriments are only further amplified. The performance
of axial injected gas elements is affected to a lesser extent and typically produce slightly higher on average
performances with similar peak performances. Some, if not all, traditional injector types can produce higher
performances with low engine throttling. This has been observed to be caused by slower axial injection of liquid
propellants which increases their residence time in the chamber and decreases the jet stiffness allowing for more
effective breakup by secondary mechanisms such as attached flame breakup. A plot that depicts these observations is
shown in the figure below.

Figure 23. General performance trends of traditionally manufactured injector elements.

These trends do not describe all injector types but rather are a representation of those observed during hot fire testing
at NASA MSFC. Another important observation is that these the optimum injector performance may not align with
the design performance of the engine. This is due to a number of challenges including but not limited to off nominal
injector flow area and reduced effectiveness of secondary mixing processes.

AM has the unique capability of producing new element schemes that were not previously possible with traditional
manufacturing techniques. This includes slots, ovals, mesh screens, and micro-elements. These geometries have the
potential to produce better atomization and mixing, increased residence time, and improved rate of vaporization all
with equivalent injector pressure drop performances to traditionally manufacture element schemes.

The desirable attributes for AM detonative injectors are further highlighted in the following section.

IV. Desirable Attributes for an Additively Manufactured RDRE Injector

Several attributes that may solve significant challenges present in RDREs are discussed in this
section with realizations from the available literature. Backflow, mixedness, fill height, system losses, and
hardware robustness are all discussed.

E. Element Diodicity

20
 Diodicity is typically utilized as a measure of performance for fluid diodes such as tesla valves. This is
formulated as the ratio of the pressure drop in the reverse direction of flow to the pressure drop in the
forward direction of flow.

𝐷𝑖 = 𝛥𝑃𝑟 /𝛥𝑃𝑓
Several articles were reviewed for information on the design and integration of a simple fluidic diode
structures [79]–[86]. Experimental values of 𝐷𝑖 for optimized fluid diodes can be as high as 10 for low
Reynolds numbers (~100) [79]. It is theoretically possible to achieve 𝐷𝑖 values significantly higher for
systems where the Reynolds number is significantly higher. Similarly, the impulsive diodicity may be
substantially higher still. The impulsive diodicity is a transient measure of the diodicity where the reverse
flow is achieved from transient pressure gradient rather than steady state pressure. An example of a fluid
diode is shown in the figure below.

Figure 24. Tesla type fluid diode with high resistance in the reverse direction and low resistance in the
forward direction. Individual junctions or cavities also shown.

It may be more applicable to incorporate the impulsive diodicity rather than the steady state diodicity
for applications in RDEs. For an RDRE injector element, the Diodicity of a element may not fully characterize
the elements performance potential to negate back flow when a detonation front passes. In this case, the
impulsive diodicity is more a function of the fluids inertia, duration of the transient event, and the original
back pressure of the element. Modeling of this transient system can really only be fully realized through
computational fluid dynamic modeling or experimental testing. So for the purposes of this effort, a general
fluid diode design, informed by the literature and limitations on AM build processes, will be incorporated
into the AM injector elements to reduce the potential to back flow. Several AM informed fluidic diode
schemes under consideration are shown in the figure below.

Figure 25. Potential fluidic diode structures integrated into an injector orifice and control case for
comparison.

Not surprisingly, to achieve the mixedness and atomization desired in an RDRE for stable
detonation propagation small element orifices are practically a must for a uniform and homogeneous

21
mixture. As a general rule of thumb for maximizing mixedness the smaller the element size and more
uniformly distribute the elements the better [31], [52]. This design practice for injector elements has also
yielded higher on average C* efficiency for constant pressure thrust chamber assemblies. It stands to
reason that this would similarly hold true for detonative combustors. This is however, only one side of the
coin for consideration with injector design. The other consideration is high atomization of the mixture which
can be achieved by optimizing the injector elements geometry and comparable pressure drop across both
the fuel and oxidizer orifices [31]. Pressure drop is really a stand in for injection velocity and injection
momentum of the working fluid.
The phase state of the propellant, typically gas or liquid, is also an important consideration for
negating the potential of the fluid to backflow. Liquids can theoretical reduce the bulk back flow potential
due to their inherent incompressibility but with the drawback of reducing their effective fill height due to
low injection velocity compared to gaseous phase propellants. Gasses, on the other hand, can recover
much quicker but at the detriment being forced further back into the plenum than liquids. Implementation
of fluidic diodes could theoretically accentuate the positive attributes of propellant injection and hinder the
negative attributes. For example, an incompressible liquid would naturally act as a rigid column when
experiencing a sharp pressure gradient. In a fluid diode structured element, the entire column would resist
backflow and allow for quicker refresh once the detonation front passes. Gaseous propellants will not
behave this way due to their natural compressibility. In this case, the gases would likely compress with
the sharp pressure gradient like a spring would if encountering a sudden force. Regardless, a fluid diode
could in theory hinder back flow but would need to act on time scales similar to that of the rate of
detonation passage. In the image of a fluid diodes presented in Figure 25, the first fluid junction would
play the most important role in hindering back flow since it would experience the sharp pressure gradient
the earliest. These short time scales would also require very short junctions that can delay, purge, and
refresh the unburned propellants into the chamber. With the advent of additive manufacturing, these
geometries could be feasibly printed directly into injector elements. Their lifecycle performance, however,
would be unknown until after hot fire testing.

F. Considerations for Flame Holding Mechanics and Combustion Delay

Parasitic deflagration is a major concern for RDCs in general, but it is not clear that it would be a
concern for a liquid/liquid or even a gas/liquid RDRE. The vaporization time, ignition delay, and on average
droplet size would need to be considered [87]. The vaporization time alone would be on the order of
magnitude of milli-seconds which puts the droplets complete vaporization on the same order as typical
rotating detonation cycle time [87], [88]. If the on average droplet size was approaching that of the lower
limit of what is capable of being generated for droplet spray fields in combustion devices this would place
the droplet size around 50 micro-meters in diameter. At this diameter of droplet size, the vaporization
efficiency and subsequent C* efficiency of a typical combustion device would be close to 100% [52]. Also
if this were the case, the droplets lifetime would be very short relative to the detonation cycle time. Some
simple mathematics for droplet lifetime are laid out by [89]–[91]. If a simple example of a 50 micron diameter
on average droplet is assumed with the theoretical combustion environment for a liquid/gas RDRE then the
chamber length required for complete vaporization of the droplet would be in the range of 30 to 50 inches.
This of course neglects secondary mixing from the detonation front. It is assumed that the detonation would
likely cause the remaining mixing of the droplet and subsequent rapid combustion at super-critical
conditions just behind the wave front.
As stated previously, parasitic deflagration and resultant flame holding is a major concern with
RDC’s particularly when considering realistic thrust chamber operating pressures. A study conducted by
[87], discussed the role of ignition delay on RDE performance and operability. Several important factors are
discussed in detail including the challenge of overcoming back-flow through the injector. The injector type
effectively modeled and discussed in this article was of an annular slot type. It was determined that if the
manifold to annulus pressure ratio was large enough, 1.5-2 times the annulus pressure, meeting the critical
pressure criteria and choking the injector orifices then back-flow is unlikely to happen. This is in part due to
the large pressure pulse meeting the momentum of the injection flow. While this is the case for compressible
gas/gas detonations, it is likely that the dynamics of liquid/gas and liquid/liquid detonations would respond
differently. The authors then utilize chemical kinetics along with other modeling methodologies to determine
that flame holding for methane oxygen environments would be unlikely to dominate chamber behavior
unless the chamber pressure was at or a much greater than 100 atm. The authors then go on to give two

22
possible means of flame holding or parasitic deflagration. The first is auto ignition phenomena causing
multiple waves to form until a limit cycle donation process is reached and standard deflagration holds. The
second is the wave mode operation dependency on vortical recirculation time scales. If this were the case
then the number of waves at any given operation condition would not change significantly as the pressure
increased. However the amount of combustion or preignition of propellants would increase. It is then noted
that the above two possibilities are for gas/gas propellants. If liquids were to be used, the delay due to
vaporization time would significantly change RDC operability. This is supported by fundamental detonation
work in spray-fields such as the work of [92].

G. Detonation Propagation in Liquid Droplet Spray Fields

To further expand on the concept of using liquid propellants, Borman and Raglands Combustion
Engineering 2011 [93] text gives two sections on detonations. One in purely gaseous mixtures and one in
liquid-gaseous mixtures. Of particular interest is the detonation of liquid fuel sprays as well as spray
detonations. A clear correlation is laid out for the required on-average droplet size in a spray field for a
detonation wave to achieve its full CJ-velocity. The text references the work of [92] that conducted
detonation tube experiments of liquid fuel spray fields dispersed in gaseous oxygen. It was found that the
detonation structure was similar to that of a gaseous mixture detonations but with a much thicker reaction
zone behind the detonation front. The authors point out that this is due to deformation, stripping,
vaporization, and diffusion of the liquid fuel by the detonation. The breakup-combustion process for the
droplets is further described as being synergistic for the wave front sustainability. Primary and secondary
explosions were observed and described as a powerful means of accelerating the detonation front.
Furthermore, this section for which the work of Ragland et al. 1968 [92] is referenced, also
discusses the effects of detonations in spray fields of various droplet sizes. To summarize, it was found that
detonations occurred in a fuel-oxygen combination, controlled for minimizing initial fuel vapor in the system,
in a wide range of on average droplet size spray fields of 2 μm – 2600 μm in diameter. It was found that the
propagation velocity was significantly less than the theoretical CJ-velocity in a gaseous equivalent in larger
droplets size spray fields but only above ~1000 μm in diameter. It was also found through modeling efforts
that for droplets below 10 μm, the mechanism of vaporization was enough to sustain detonation. Droplets
from 10 μm to 1000 μm would need an additional mechanism for creating micro sprays from the detonation
front breaking up the droplets first and then vaporizations. Regardless, as long as the average droplet size
is below 1000 μm on average, the detonation wave propagates with approximately a 2% detriment to the
theoretical CJ-velocity of the gaseous equivalent mixture. It must also be pointed out that this work was
conducted at ambient initial temperatures and pressures. It is not certain how these trends would change,
if at all, to the spray field of cryogenic propellants or even a spray field of oxidizer rather than fuel. Since
the droplet lifetime of cryogenic propellants is significantly shorter than ambient propellants, it is feasible to
imagine that the detonation limits for propagation would shift to higher on average initial droplet sizes.
The implications of the works presented above is multi-faceted when it comes to the application of
RDC technology. First, high-pressure detonations can be achieved without the need to account for parasitic
deflagration. An injector can be designed so that it accounts for ignition and vaporization delays alone.
Furthermore, lower operating injection temperatures can be utilized with cryogenics to enhance detonation
properties. All experimental RDC literature reviewed to date has utilized gaseous propellant combinations.
An RDC that operates with a liquid fuel or oxidizer is a next step for design advancement in the technology.
This is currently being pursued by several academic sources such as the University of Alabama at Huntsville
[94].
According to the review of the available literature above, liquid propellant injection may hold the key to
rotating detonation combustor performance at realistic rocket chamber pressures. Further results suggest
that more delays in parasitic deflagration and flame holding can be achieved by use of cryogenics by means
of the latent heat of vaporization. Ultimately, a major conclusion of this work was that flame holding
operating modes of these combustors will be pushed out to higher pressures if the correct propellant
combinations are chosen. Injector design is also highly important for prioritizing mixing and to optimize total
propellant Injection height. It has been successfully demonstrated that rotating detonation modes are
possible with liquid/gas injection with recent works [27]. There is a single inherent concern for liquid
injection, that is the limitation on fill height. This was covered somewhat in the previous section by assisting
the liquid oxidizers injection velocity with a gaseous fuel. It is not yet known if this would work in practical
applications.

23
 V. Lessons Learned
Lessons learned from the available literature have been summarized and discussed towards their
integration into high performance AM RDRE injectors. This section combines all topics discussed in
previous sections as well as the hot fire test experience from the authors and experimental investigation of
AM injectors. Numerous desirable attributes for a high-performance detonative injector have been outlined.
Several of these design features are only possible with the use of additive manufacturing. Otherwise, they
would require labor intensive machining, EDM, and post-processing capabilities that do not currently exist
using traditional manufacturing techniques. In addition, the use of AM would allow for the rapid production
of these hardware and minimal post-processing to optimize the components material, flow area, and heat
load performance.

VI. Summary and Conclusions

Provide a concise summary of the work, as well as conclusions and recommendations drawn from the
effort. Do not introduce any new material

Appendix
An appendix, if needed, should appear before the acknowledgments.

Acknowledgments
An Acknowledgments section, if used, immediately precedes the References. Sponsorship information and
funding data are included here. The preferred spelling of the word “acknowledgment” in American English is without
the “e” after the “g.” Avoid expressions such as “One of us (S.B.A.) would like to thank…” Instead, write “F. A.
Author thanks…”

24
 References

[1] M. L. Fotia, F. Schauer, T. Kaemming, and J. Hoke, “Experimental study of the performance of a rotating detonation
engine with nozzle,” J. Propuls. Power, vol. 32, no. 3, pp. 674–681, 2016.
[2] S. Theuerkauf, P. King, F. Schauer, and J. Hoke, “Thermal management for a modular rotating detonation engine,” in 51st
AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition, 2013, p. 1176.
[3] A. Roy et al., “Experimental study of rotating detonation combustor performance under preheat and back pressure
operation,” in 55th AIAA aerospace sciences meeting, 2017, p. 1065.
[4] C. M. Brophy and J. Codoni, “Experimental performance characterization of an RDE using Equivalent Available
Pressure,” in AIAA Propulsion and Energy 2019 Forum, 2019, p. 4212.
[5] F. A. Bykovskii, S. A. Zhdan, and E. F. Vedernikov, “Continuous spin detonations,” J. Propuls. power, vol. 22, no. 6, pp.
1204–1216, 2006.
[6] A. Okninski, J. Kindracki, and P. Wolanski, “Rocket rotating detonation engine flight demonstrator,” Aircr. Eng. Aerosp.
Technol. An Int. J., 2016.
[7] J. A. NlCHOLLS, R. E. CULLEN, and K. W. Ragland, “Feasibility studies of a rotating detonation wave rocket motor.,”
J. Spacecr. Rockets, vol. 3, no. 6, pp. 893–898, 1966.
[8] S. Zhou, H. Ma, C. Liu, C. Zhou, and D. Liu, “Experimental investigation on the temperature and heat-transfer
characteristics of rotating-detonation-combustor outer wall,” Int. J. Hydrogen Energy, vol. 43, no. 45, pp. 21079–21089,
2018.
[9] S.-J. Liu, Z.-Y. Lin, W.-D. Liu, W. Lin, and F.-C. Zhuang, “Experimental realization of H2/air continuous rotating
detonation in a cylindrical combustor,” Combust. Sci. Technol., vol. 184, no. 9, pp. 1302–1317, 2012.
[10] B. A. Rankin, D. R. Richardson, A. W. Caswell, A. G. Naples, J. L. Hoke, and F. R. Schauer, “Chemiluminescence
imaging of an optically accessible non-premixed rotating detonation engine,” Combust. Flame, vol. 176, pp. 12–22, 2017.
[11] I. V Walters, C. Journell, A. I. Lemcherfi, R. Gejji, S. D. Heister, and C. D. Slabaugh, “Performance Characterization of
a Natural Gas-Air Rotating Detonation Engine at Elevated Pressure,” in AIAA Propulsion and Energy 2019 Forum, 2019,
p. 4214.
[12] F. A. Bykovskii, S. A. Zhdan, E. F. Vedernikov, and A. N. Samsonov, “Scaling factor in continuous spin detonation of
syngas–air mixtures,” Combust. Explos. Shock Waves, vol. 53, no. 2, pp. 187–198, 2017.
[13] P. R. Gradl, C. S. Protz, and T. Wammen, “Additive Manufacturing and Hot-fire Testing of Liquid Rocket Channel Wall
Nozzles Using Blown Powder Directed Energy Deposition Inconel 625 and JBK-75 Alloys,” in AIAA Propulsion and
Energy 2019 Forum, 2019, p. 4362.
[14] P. Gradl, T. Teasley, C. Garcia, C. Protz, and S. Greene, “Hot-Fire Performance of Additively Manufactured GRCop-42
and GRCop-84 Channel-Cooled Combustion Chambers,” 2019.
[15] P. R. Gradl et al., “Additive Manufacturing of Liquid Rocket Engine Combustion Devices: A Summary of Process
Developments and Hot-Fire Testing Results,” in 2018 Joint Propulsion Conference, 2018, p. 4625.
[16] D. P. Stechmann, “Experimental study of high-pressure rotating detonation combustion in rocket environments.” Purdue
University, 2017.
[17] D. P. Stechmann, S. D. Heister, and S. V Sardeshmukh, “High-pressure rotating detonation engine testing and
flameholding analysis with hydrogen and natural gas,” in 55th AIAA Aerospace Sciences Meeting, 2017, p. 1931.
[18] J. W. Bennewitz, B. R. Bigler, W. A. Hargus, S. A. Danczyk, and R. D. Smith, “Characterization of detonation wave
propagation in a rotating detonation rocket engine using direct high-speed imaging,” in 2018 Joint Propulsion Conference,
2018, p. 4688.
[19] V. Anand and E. Gutmark, “Rotating detonation combustors and their similarities to rocket instabilities,” Prog. Energy
Combust. Sci., vol. 73, pp. 182–234, 2019.
[20] B. Bigler, J. W. Bennewitz, S. A. Danczyk, and W. A. Hargus, “Injector Mixing Effects in Rotating Detonation Rocket
Engines,” in AIAA Propulsion and Energy 2019 Forum, 2019, p. 3869.
[21] A. R. Mizener and F. K. Lu, “Preliminary parametric analysis of a rotating detonation engine by analytical methods,” in
52nd AIAA/SAE/ASEE Joint Propulsion Conference, 2016, p. 4876.
[22] J.-M. Li et al., “Application of Graphene Oxide in Jet A-1 in Air to Enhance Combustion Process,” in 2018 AIAA
Aerospace Sciences Meeting, 2018, p. 133.
[23] J. Sosa et al., “Experimental evidence of H2/O2 propellants powered rotating detonation waves,” Combust. Flame, vol.
214, pp. 136–138, 2020.
[24] W. A. Hargus, S. A. Schumaker, and E. J. Paulson, “Air Force Research Laboratory Rotating Detonation Rocket Engine
Development,” in 2018 Joint Propulsion Conference, 2018, p. 4876.
[25] K. Goto et al., “Propulsive performance and heating environment of rotating detonation engine with various nozzles,” J.
Propuls. Power, vol. 35, no. 1, pp. 213–223, 2019.
[26] K. Ishihara et al., “Study on a long-time operation towards rotating detonation rocket engine flight demonstration,” in 55th
AIAA Aerospace Sciences Meeting, 2017, p. 1062.
[27] D. Lim, “Experimental Studies of Liquid Injector Response and Wall Heat Flux in a Rotating Detonation Rocket Engine.”
Purdue University Graduate School, 2019.
[28] G. S. Gill and W. H. Nurick, “Liquid Rocket Engine Injectors.,” NASA Spec Publ SP-8089. 1976.
[29] D. H. Huang and D. K. Huzel, Modern engineering for design of liquid-propellant rocket engines. American Institute of

25
 Aeronautics and Astronautics, 1992.
[30] M. Glogowski and M. M. Micci, “Shear coaxial injector spray characterization near the LOX post tip region,” 31st Jt.
Propuls. Conf. Exhib., 1995, doi: 10.2514/6.1995-2552.
[31] R. J. Burick, “Space storable propellant performance program coaxial injector characterization,” 1972.
[32] R. D. Sutton, M. D. Schuman, and W. D. Chadwick, “Operating manual for coaxial injection combustion model,” 1974.
[33] S. A. Schumaker, S. A. Danczyk, M. D. A. Lightfoot, and A. L. Kastengren, “Interpretation of core length in shear coaxial
rocket injectors from X-ray radiography measurements,” 2014, doi: 10.2514/6.2014-3790.
[34] J. Hulka and D. Makel, “Liquid oxygen/hydrogen testing of a single swirl coaxial injector element in a windowed
combustion chamber,” 1993, doi: 10.2514/6.1993-1954.
[35] R. K. Cohn, P. A. Strakey, R. . Bates, D. G. Talley, J. A. Muss, and C. . Johnson, “Swirl Coaxial Injector Development,”
2003.
[36] B. R. Bigler, J. W. Bennewitz, S. A. Danczyk, and W. A. Hargus, “Rotating Detonation Rocket Engine Operability Under
Varied Pressure Drop Injection,” J. Spacecr. Rockets, pp. 1–10, 2020, doi: 10.2514/1.a34763.
[37] M. C. Waters, “Analysis of Additively Manufactured Injectors for Rotating Detonation Engines,” 2018.
[38] S. Soller et al., “Design and Testing of Liquid Propellant Injectors for Additive Manufacturing,” EUCASS, no. Eucass,
2017.
[39] S. Soller et al., “Selective laser melting of Inconel 718 and stainless steel injectors for liquid rocket engines,” Sp. Propuls.
2016 Proc., no. May, 2016.
[40] D. S. Morrow, A. P. Nair, and R. M. Spearrin, “Minimizing hydraulic losses in additively manufactured swirl coaxial
injectors,” AIAA Propuls. Energy Forum Expo. 2019, 2019, doi: 10.2514/6.2019-4310.
[41] G. Barnett and D. B. Bullard, “Summary of Liquid Oxygen/Hydrogen, Direct Metal Laser Sintering Injector Testing and
Evaluation Effort at Marshall Space Flight Center,” 2015.
[42] K. Melle, S. Rappillard, and V. Demaison, “Development and Test Firing of a One-Piece Additively Manufactured Liquid
Rocket Engine,” Ecole Supérieure des Techniques Aéronautiques et de la Construction Automobile, 2019.
[43] P. A. Strakey, D. G. Talley, and J. J. Hutt, “Mixing characteristics of coaxial injectors at high gas/liquid momentum ratios,”
Journal of Propulsion and Power, vol. 17, no. 2. pp. 402–410, 2001, doi: 10.2514/2.5756.
[44] R. J. Santoro, “Swirl effects on coaxial injection atomization.” The Pennsylvania State University, 1996.
[45] S. A. Rahman, S. Pal, and R. . Santoro, “Similitude in pressure-swirl atomizer sprays.pdf,” Phys. Fluids, 1996.
[46] S. A. Rahman and R. J. Santoro, “A review of coaxial gas/liquid spray experiments and correlations,” 1994, doi:
10.2514/6.1994-2772.
[47] G. E. Lorenzetto and A. H. Lefebvre, “Measurements of drop size on a plain-jet airblast atomizer,” AIAA J., vol. 15, no.
7, pp. 1006–1010, 1977, doi: 10.2514/3.60742.
[48] F. C. Zhuang and J. G. Sun, “Effects of swirl coaxial injector parameters on LOX/GH2 engine combustion performance,”
38th AIAA/ASME/SAE/ASEE Jt. Propuls. Conf. Exhib., no. July, 2002, doi: 10.2514/6.2002-3697.
[49] W. H. Nurick, “Analysis of sprays from rocket engine injectors,” J. Spacecr. Rockets, vol. 8, no. 7, pp. 796–798, 1971,
doi: 10.2514/3.30321.
[50] J. H. Rupe, “The liquid-phase mixing of a pair of impinging streams,” 1953.
[51] N. Zong and V. Yang, “A numerical study of high-pressure oxygen/methane mixing and combustion of a shear coaxial
injector,” 43rd AIAA Aerosp. Sci. Meet. Exhib. - Meet. Pap., 2005, doi: 10.2514/6.2005-152.
[52] R. J. Burick, “Atomization and mixing characteristics of gas/liquid coaxial injector elements,” J. Spacecr. Rockets, vol. 9,
no. 5, pp. 326–331, 1972.
[53] L. J. Yang, M. H. Ge, M. Z. Zhang, Q. F. Fu, and G. B. Cai, “Spray characteristics of a recessed gas-liquid coaxial swirl
injector,” J. Propuls. Power, vol. 24, no. 6, pp. 1332–1339, 2008, doi: 10.2514/1.23977.
[54] T. Inamura, H. Tamura, and H. Sakamoto, “Characteristics of liquid film and spray injected from swirl coaxial injector,”
J. Propuls. Power, vol. 19, no. 4, pp. 632–639, 2003, doi: 10.2514/2.6151.
[55] B. Yang, F. Cuoco, and M. Oschwald, “Atomization and flames in LOX/H2-and LOX/CH4-spray combustion,” J.
Propuls. Power, vol. 23, no. 4, pp. 763–771, 2007, doi: 10.2514/1.26538.
[56] L. E. Sanchez, R. Ellis, V. Dorado, and A. Choudhuri, “An experimental study of the effects of plate geometry on the
spray atomization of LOX/CH4 in uni-element shear coaxial injectors,” 50th AIAA/ASME/SAE/ASEE Jt. Propuls. Conf.
2014, pp. 1–10, 2014, doi: 10.2514/6.2014-3788.
[57] S. Pal, M. D. Moser, H. M. Ryan, M. J. Foust, and R. J. Santoro, “Shear Coaxial Injector Atomization Phenomena For
Combusting and Non-combusting Conditions,” 2020.
[58] R. D. Woodward, S. Pal, S. Farhangi, G. E. Jensen, and R. J. Santoro, “LOX/GH2 shear coaxial injector atomization
studies: Effect of recess and non-concentricity,” Collect. Tech. Pap. - 45th AIAA Aerosp. Sci. Meet., vol. 10, no. January,
pp. 6925–6946, 2007, doi: 10.2514/6.2007-571.
[59] C. Nicoli, P. Haldenwang, and B. Denet, “Flame holding downstream from a co-flow injector,” Comptes Rendus - Mec.,
vol. 334, no. 7, pp. 408–413, 2006, doi: 10.1016/j.crme.2006.06.002.
[60] J. Lux and O. Haidn, “Flame stabilization in high-pressure liquid oxygen/methane rocket engine combustion,” J. Propuls.
Power, vol. 25, no. 1, pp. 15–23, 2009, doi: 10.2514/1.36852.
[61] M. Oschwald, S. Naclerio, B. Yang, and F. Cuoco, “Experimental investigation of flame stabilization in coaxial LOX/H2
and LOX/CH4 spray flames,” Int. Conf. Liq. At. Spray Syst., no. January, pp. 8–10, 2008.
[62] N. Guézennec, T. Dawson, P. Sierra, and S. Menon, “Flame holding dynamics during combustion instability in a shear-

26
 coaxial injector combustor,” Int. Symp. Turbul. Shear Flow Phenomena, TSFP 2013, vol. 1, pp. 1–6, 2013.
[63] A. Ghosh, Q. Diao, and K. H. Yu, “Effect of flow parameters on the acoustic susceptibility of shear-coaxial flames,” J.
Propuls. Power, vol. 32, no. 2, pp. 295–304, 2016, doi: 10.2514/1.B35420.
[64] Y. Wang and W. Song, “Experimental investigation of influence factors on flame holding in a supersonic combustor,”
Aerosp. Sci. Technol., vol. 85, pp. 180–186, 2019, doi: 10.1016/j.ast.2018.12.002.
[65] N. Zong and V. Yang, “Near-field flow and flame dynamics of LOX/methane shear-coaxial injector under supercritical
conditions,” Proc. Combust. Inst., vol. 31 II, pp. 2309–2317, 2007, doi: 10.1016/j.proci.2006.08.106.
[66] W. Armbruster, J. S. Hardi, D. Suslov, and M. Oschwald, “Injector-driven flame dynamics in a high-pressure multi-
element oxygen - Hydrogen rocket thrust chamber,” J. Propuls. Power, vol. 35, no. 3, pp. 632–644, 2019, doi:
10.2514/1.B37406.
[67] V. P. Zhukov and M. Feil, “Numerical Simulations of the Flame of a Single Coaxial Injector,” Int. J. Aerosp. Eng., vol.
2017, 2017, doi: 10.1155/2017/5147606.
[68] M. Roa and D. G. Talley, “Wave dynamic mechanisms in coaxial hydrogen/liquid-oxygen jet flames,” J. Propuls. Power,
vol. 35, no. 2, pp. 369–381, 2019, doi: 10.2514/1.B37079.
[69] C. Pauly, J. Sender, and M. Oschwald, “Ignition of a gaseous methane/oxygen coaxial jet,” pp. 155–170, 2009, doi:
10.1051/eucass/200901155.
[70] V. Schmidt, U. Wepler, O. J. Haidn, and M. Oschwald, “Characterization of the primary ignition process of a coaxial
GH2/LOX spray,” AIAA Pap., no. January, pp. 7592–7601, 2004, doi: 10.2514/6.2004-1167.
[71] G. Jones, C. Protz, B. Bullard, and J. Hulka, “Local Heat Flux Measurements with Single Element Coaxial Injectors,”
2006.
[72] J. J. Smith, M. Bechle, D. Suslov, M. Oschwald, O. Haidn, and G. Schneider, “Steady-State High Pressure LOX/H2
Rocket Engine Combustion.”
[73] W. O. H. Mayer et al., “Atomization and breakup of cryogenic propellants under high-pressure subcritical and supercritical
conditions,” J. Propuls. Power, vol. 14, no. 5, pp. 835–842, 1998, doi: 10.2514/2.5348.
[74] H. Tani, S. Teramoto, and K. Okamoto, “Effects of injector geometry on cryogenic shear coaxial jets at supercritical
pressures,” J. Propuls. Power, vol. 31, no. 3, pp. 883–888, 2015, doi: 10.2514/1.B35530.
[75] W. Mayer and H. Tamura, “Propellant injection in a liquid oxygen/gaseous hydrogen rocket engine,” J. Propuls. Power,
vol. 12, no. 6, pp. 1137–1147, 1996, doi: 10.2514/3.24154.
[76] G. L. Barnett and D. B. Bullard, “Summary of Liquid Oxygen/Hydrogen, Direct Metal Laser Sintering Injector Testing
and Evaluation Effort at Marshall Space Flight Center,” 62nd JANNAF Propuls. Meet. Jt. Meet. 10th Model. Simul.
Subcomm. / 8th Liq. Propuls. Subcomm. / 7th Spacecr. Propuls. Subcomm., vol. Nasheville, p. June 1-5, 2015., 2015.
[77] D. D. Bullard et al., “Design, Fabrication and Hot-Fire Test of an Additively Manufactured Liquid Oxygen/Hydrogen 35-
klbf Swirl Coaxial Element Injector,” Jt. Meet. Program. Ind. Base / 11th Model. Simul. Subcomm. / 9th Liq. Propuls.
Subcomm. / 8th Spacecr. Propuls. Subcomm., p. Phoenic, AZ.
[78] J. Sims and J. R. Hulka, “Development of Swirl Elements for Additively Manufactured Liquid Propellant Rocket Engine
Injectors,” JANNAF 13th Model. Simulation, 11th Liq. Propulsion, 10th Spacecr. Propuls.
[79] S. Lin, L. Zhao, J. K. Guest, T. P. Weihs, and Z. Liu, “Topology optimization of fixed-geometry fluid diodes,” J. Mech.
Des., vol. 137, no. 8, 2015.
[80] S. Okuhara, M. Takao, H. Sato, A. Takami, and T. Setoguchi, “A Twin Unidirectional Impulse Turbine for Wave Energy
Conversion—Effect of Fluidic Diode on the Performance,” Open J. Fluid Dyn., vol. 4, no. 05, p. 433, 2014.
[81] K. Matsumoto, M. Takao, S. Okuhara, M. M. A. Alam, and Y. Kinoue, “A Study on Fluidic Diode for Wave Energy
Conversion-Effect of Bypass Geometry on the Turbine Performance,” Open J. Fluid Dyn., vol. 10, no. 03, p. 270, 2020.
[82] S. Okuhara, M. M. A. Alam, M. Takao, and Y. Kinoue, “Performance of fluidic diode for a twin unidirectional impulse
turbine,” in IOP Conference Series: Earth and Environmental Science, 2019, vol. 240, no. 5, p. 52011.
[83] H. Sato, M. Takao, S. Okuhura, M. M. Ahsraful Alam, and T. Setoguchi, “A Twin Unidirectional Impulse Turbine With
Fluidic Diode for Wave Energy Conversion,” in Fluids Engineering Division Summer Meeting, 2015, vol. 57213, p.
V01AT22A004.
[84] M. Fripp, L. Zhao, and B. Least, “The theory of a fluidic diode autonomous inflow control device,” 2013.
[85] J. Kordík and Z. Trávníček, “Novel fluidic diode for hybrid synthetic jet actuator,” J. Fluids Eng., vol. 135, no. 10, 2013.
[86] T. W. Bermel, “Fluidic diode.” Google Patents, Nov. 25, 1969.
[87] D. P. Stechmann, S. Sardeshmukh, S. D. Heister, and K. Mikoshiba, “Role of ignition delay in rotating detonation engine
performance and operability,” J. Propuls. Power, vol. 35, no. 1, pp. 125–140, 2019.
[88] V. Yang, “Modeling of supercritical vaporization, mixing, and combustion processes in liquid-fueled propulsion systems,”
Proc. Combust. Inst., vol. 28, no. 1, pp. 925–942, 2000.
[89] K. W. Ragland and K. M. Bryden, Combustion engineering. CRC press, 2011.
[90] J. Buckmaster, T. L. Jackson, and A. Kumar, Combustion in high-speed flows, vol. 1. Springer Science & Business Media,
2012.
[91] K. K. Kuo, Principles of combustion, no. TJ254. 5 K85 2005. 2005.
[92] K. W. Ragland, E. K. Dabora, and J. A. Nicholls, “Observed structure of spray detonations,” Phys. Fluids, vol. 11, no. 11,
pp. 2377–2388, 1968.
[93] G. L. Borman and K. W. Ragland, “Chemical kinetics of combustion,” Combust. Eng., pp. 107–141, 1998.
[94] E. C. Unruh, M. Spaulding, D. M. Lineberry, K. G. Xu, and R. A. Frederick, “Development of an Optically Accessible

27
Racetrack-Type Rotating Detonation Rocket Engine,” in AIAA Propulsion and Energy 2020 Forum, 2020, p. 3868.

28