ESS-38-90
Rotating Detonation Engines
Zbigniew Szawerdo
Cranfield University, Cranfield, Bedfordshire MK43 0AL, United Kingdom, z.szawerdo@cranfield.ac.uk
Abstract
Detonation is a process fairly well understood. However its application to propulsion is in a very early stage. Detonative
combustion results in an increase in pressure, as compared to deflagrative combustion currently widely used. This can be utilized
generally in Pulse Detonation Engines (PDE) or Rotating Detonation Engines (RDE). This paper focuses on the latter, since
rotating detonation combustion chambers have a possibility to be easily adopted in existing designs of a turbojet, ramjet or a
rocket engine. A principle of work is shown, and a brief history of the RDE’s development is given, which starts as early as 50
years ago, but only recently, with progresses of the computational powers, the research has been intensified. This has also to do
with possibilities of such engines to improve propulsive efficiency by a great deal, which is clearly in interest of aircraft engine
manufacturers in view of rising fuel prices. Different applications and configurations are presented, together with a critical
review of advantages and disadvantages of applying rotating detonation to existing designs.
Keywords: rotating detonation; RDE; detonation engine
Nomenclature 2. Deflagration and detonation
CRD Continuously Rotating Detonation The main difference between deflagration and detonation
PDE Pulse Detonation Engine is that in the former the pressure and the combustion products
RDE Rotating Detonation Engine density behind the flame front decrease and in the latter the
pressure and density increase.
The relationship comes from rearranging the mass and
1. Introduction momentum conservation equations and is presented below:
The detonation process has been described in 1881 by P2 − P1
Berthelot, Vieille, Mallard and Le Chatelier and the theory 1
= ṁ2 > 0 (1)
described in 1899 by Chapman and Jouguet [1, 2]. First at- ρ1
− ρ12
tempts to apply pulse detonation to propulsion were made
in the 1950s by Nicholls at the University of Michigan and The equation shows that only such transformations can
around ten years later first establishment of continuously exist whose pressure change has the same sign as the density
rotating detonation was demonstrated by Vojciechovski, Met- change [4].
rofanov and Topchiyan at the Institute of Hydrodynamics of This shows the first advantage of detonation over deflag-
Siberian Branch of Soviet Academy of Sciences in Novos- ration. No pressure drop in the combustion chamber. There
ibirsk [1]. is a substantial increase in pressure that improves the overall
Around twenty to thirty years ago the research has been performance.
re-initiated with the development of computational tools to Another difference lays in the speed of the combustion
help understand the detonation process better, and the need front propagation. In case of deflagration it is typically in the
to develop newer solutions to outperform ones used today range of dozens of meters per second, while for detonation
which are believed to slowly reach their limit of performance. the flame front propagates in kilometres per second. This
Both PDE and RDE propulsion is being looked closer upon. results in a possibility to shorten the combustion chamber and
This paper concentrates on the latter. manufacture a more compact and thus lighter engine. The
More and more papers are being published on the topic temperature for both is very similar, but the residence time
and it is one of the centres of interest on combustion related and combustion zone size are smaller for the detonation, and
conferences. Detonative propulsion poses a few unique ad- this results in a reduced NOx production.
vantages over traditional deflagrative one and together with Since these two processes are significantly different the
distributed propulsion is being named as the future of air performance results of application of either one are also com-
transport technology [3]. pletely different. Figure 1 presents a comparison of ideal
1
38th Engine Systems Symposium - March 2013 ESS-38-90
Z. Szawerdo / Rotating Detonation Engines
thermodynamic cycles for these detonation and deflagration. 3. Detonation in propulsion
The combustion process starts at point 2 after same com-
pression 1-2 for both cycles. The detonative pressure peak There are many ways to utilize detonative combustion in
is clearly seen, while the Brayton cycle shows the volume propulsion systems. Although this paper focuses on CRD
increase. engines it is necessary to name some other different designs.
The concept of a pulse detonation engine has been men-
tioned earlier. In more detail it is described in another paper
for this symposium. For the purpose of later comparison it is
important to list the key characteristics of the design.
Figure 2: First PDE designed by Nicholls et al. at the University of
Michigan [7].
A schematic drawing of PDE is shown on figure 2. It typic-
ally consists of a long tube, which is filled with fresh fuel–air
Figure 1: Thermodynamic Brayton-Joule and Fickett-Jacobs mixture. The flame initiated by ignition must, in a relatively
cycles [5]. short time, accelerate to the detonation velocity, such that
the transition from deflagration to detonation happens in a
relatively small distance. Detonative combustion results in
The efficiency of both cycles can be calculated using the very high pressure, which is then converted to thrust. When
following formulas: the detonation went through all of the mixture in the tube, the
products have to be exited and fresh mixture must be quickly
1 resupplied, for the cycle to be repeated [6].
ηB = 1 − k−1 (2) The main drawback of such engines is that their operate in
p k 2
p1
pulsed mode and the thrust changes in time which is usually
in range of a tens to hundreds of Hertz [6]. Also the geometry
T 00 1k is quite complicated as the engine requires fast purging and
3
1 −1
ηF−J
T
= 1 − k k−1 T2 00 (3) refilling of the tube, and the detonation has to be reinitiated
p2 k 3
−1 each time.
T2
p1
Provided that the initial compression process is the same
the highest efficiency can be clearly obtained for the Fickett-
Jacobs cycle. This is shown in table 1 where the ideal effi-
ciency is calculated for three different fuels.
Table 1:
Comparison of ideal cycle efficiencies for different fuels for an initial
compression ratio equal to 5 [5]
Figure 3: Initial idea of the Standing Detonation Wave Engine [8].
Fuel Brayton (%) Fickett-Jacobs (%)
Hydrogen - H2 36.9 59.3 Apart from the PDEs other concepts also receive research-
Methane - CH4 31.4 53.2 ers attention. These concentrate on employing a standing
detonation wave and are simply called Standing Wave Det-
Acetylene - C2 H2 36.9 61.4
onation Engines (SDWE), on figure 3, or detonation driven
ramjets (dramjets). These can be used however for a limited
The values presented clearly show why there is interest range of flight velocities so they could have very few, if any
in the detonative combustion. Such an increase in cycle applications [6].
efficiency will improve the specific fuel consumption substan-
tially. This is the main driver of the research in the field.
Lastly the design of the detonative combustion chamber
can be much simpler compared to the existing deflagrative
solutions. No recirculation is needed to hold the flame in
place, and in a lean mixture detonation the temperature of the
products is low enough so that no mixing before the turbine
is needed [6]. Figure 4: Schematic of a ram accelerator in the detonative mode [9].
2
38th Engine Systems Symposium - March 2013 ESS-38-90
Z. Szawerdo / Rotating Detonation Engines
There is also a proposed idea of a ram accelerator 1960s. They achieved stabilized rotating detonations in a
(RAMAC), to accelerate projectiles through a fuel-oxidizer cylindrical chamber for relatively long durations. These ex-
mixture filled tube - as presented on figure 4. These could periments conducted at the Institute of Hydrodynamics of
reach velocities up to 20 km/s. RAMAC could be for example the Siberian Branch of the Soviet Academy of Sciences in
used for a low cost delivery of payloads to the Earth Orbit [6]. Novosibirsk must be considered as the first fundamental step
The most promising one, however is the RDE concept and to developing the Rotating Detonation Engine [6].
is described in more detail in the following section. Further on theoretical analyses and some more pioneering
experimental research of the application of rotating detona-
4. Rotating detonation engine tions to rocket propulsion were conducted at the University
of Michigan [6].
The principle of RDE is based on a continuously propagat- After some time the research slowly stopped and the RDE
ing detonation in a ring-like combustion chamber. A rep- concept, just as the PDE was almost forgotten. This happened
resentation of such chamber is presented in figure 5. Air is as a result of low crude oil prices, and thus aircraft fuels as
supplied through a narrow slit at sonic conditions [6]. It is well. Fuel efficiency was not a primary consideration.
then mixed with fuel which enters from a number of holes on More recently, successful experimental research on RDE
the side of the cylinder (not presented). The rise in pressure has been restarted. Predominantly the research is being car-
from the detonation wave temporarily shuts off a portion of ried out at the Institute of Hydrodynamics in Novosibirsk in
the injector orifices, but refuelling begins when the pressure Russia, at the Warsaw University of Technology in Poland,
behind the wave reduces below that of the plenum cham- ENSMA and MBDA of France, and at the University of Texas
ber [10]. at Arlington and Pratt & Whitney in the US. Many groups
are supporting experimental research with numerical calcu-
lations, and researchers in institutions in some countries, for
example Japan, are relying on computational results only [6].
There is a lot of academics involved and CRD is one of the
centres of interest on many conferences worldwide.
In 2004, Tobita, Fujiwara, and Wolański applied for a
patent on the Rotating Detonation Engine and the patent was
issued in 2005 [1, 12].
4.2. Experimental research
Figure 5: Basic RDE combustion chamber representation [10].
As mentioned in the previous section a substantial amount
To further help understand the processes happening in of experimental research is being carried out. Different geo-
the combustion chamber an expanded view is presented in metries are being examined - a few example configurations
figure 6. The triangular fresh mixture layer region, the det- are shown on figure 8. A variety of methods is being used to
onation wave front, and the attached shock can be clearly investigate the rotating detonation process closer, from pres-
seen. sure transducers through smoked foil to the compensation
photographic technique.
The compensation photographic technique is based on
projection of a moving object onto a moving photographic
film. Velocity of the film has to be controlled in a way that the
photographed object is stationary in the film moving frame
coordinates [6]. The film, although only exposed through a
narrow window catches the whole image of a propagating
detonation wave. Examples of photographs recording the
rotating detonation wave structure are presented on figure7.
Figure 6: Schematic of the rotating detonation wave structure [10].
Naturally, the new fresh mixture should have sufficient
time to fill up the region where the detonation has already
passed, before the next appearance of the detonation front.
Therefore, the frequency of chamber operation will depend
on the Chapman-Jouguet detonation velocity and the size
of "ring-like combustion chamber" [11]. The direction of
detonation is not specified in any way. It can follow both
ways. The combustion is also not restricted to one detonation
wave and 2, 3 and even more can occur at the same time.
It is important to notice that the air inlet and also more
importantly the combustion products outflow is axial. Figure 7: Continuous Rotating Detonation wave structure in a cyl-
4.1. Research history indrical chamber obtained for different mixtures [13].
Vojciechovski, Metrofanov, and Topchiyan performed ex- Extensive research is being carried out at the Warsaw Uni-
periments on continuously rotating detonations in the early versity of Technology. Different geometries are being tested
3
38th Engine Systems Symposium - March 2013 ESS-38-90
Z. Szawerdo / Rotating Detonation Engines
Figure 8: Basic geometric configurations for continuously rotating detonation studies. “O” indicates locations of oxidizer injection and “F”
location of fuel injections [13].
Figure 9: Pictures of the experimental test stands for Continuous Rotating Detonation at Warsaw University of Technology, (a) assembled
cylindrical chamber, (b) cylindrical chamber with air and fuel supply lines and dump tank (c) temperature measurements for hydrogen–air
injections [6].
Figure 10: Pressure–time history for rich methane–oxygen mixture for an initial pressure of 1 bar, for different distances of pressure transducers
from the injection plane (cylindrical geometry): (a) 20 mm, (b) 40 mm, and (c) 60 mm [14].
4
38th Engine Systems Symposium - March 2013 ESS-38-90
Z. Szawerdo / Rotating Detonation Engines
with different fuels from hydrogen and basic hydrocarbons to lows this. Many papers are being presented on the topic, and
aviation kerosene. A picture of one of the test rigs and sample some of the results are being presented on figures 11 to 13.
measurements of exhaust gas temperature are presented in
figure 9. The thrust is also measured together with pressure in
multiple locations allowing to evaluate the detonation direc-
tion and even the structure of the CRD wave [6, 14]. Sample
pressure measurements are presented on figure 10. Note the
timestep on the x-axis in the order of tenths of microseconds.
The experimental research is typically made on small com-
bustion chambers, not comparable with sizes of engines used
in aviation. Recently a full size Continuous Rotating Detona-
tion combustion chamber research facility has been opened
at the Institute of Aviation in Warsaw with the aim to replace
the original deflagration mode combustion chamber of the
GTD-350 gas turbine in the near future [6].
4.3. Numerical research
Figure 13: Enlarged structure of the 3D Numerical Schlieren of
First numerical calculations of CRD were performed in
detonation propagating in a cylindrical channel [18].
the 1970s. Since that time the computational powers of com-
puters have increased more than significantly and today very
detailed calculations of the flow structure in a RDE are pos- The results apart from supplying the researchers with a
sible [6]. To the point that the academics are taking them as lot of insight, describing which is beyond the scope of this
the source of their knowledge about detonation processes in work, show general characteristics of the flow that have been
detail [15, 17]. No real means of investigating this in experi- mentioned earlier, in previous sections. The axial outflow
ments, increasing modelling complexity and generally good of reactants is visible on figure 11c, and the possibility of
agreement of computations with other experimental data al- two detonation waves occurring in the combustion chamber
Figure 11: Computational results showing (a) 2-D CRD structure, (b) computational “soot print” revealing structure of detonation front, (c)
velocity vector in laboratory coordinate system [15].
Figure 12: Detonation waves rotating in an annular chamber at 2.88 × 10−3 s (pressure contours) [16].
5
38th Engine Systems Symposium - March 2013 ESS-38-90
Z. Szawerdo / Rotating Detonation Engines
on figure 12b. Clearly the detonation front is shorter, as The values of specific impulse calculated for such design
there is just half of the combustible mixture available for the are higher than those of the conventional rocket engines [19].
detonation wave.
Figure 13 presents the level and depth of detail achiev- 4.4.2. Tutbojet / turbofan engine
able with numerical calculations which is not possible with
experimental methods. The CRD can also be applied in turbojet engines, but
Numerical calculations allow the researchers to change the research into such an application is only at the conceptual
different parameters of the flow and the combustion cham- stage [6]. The idea is to switch the traditional deflagration
ber geometry with zero cost of building new test rigs. The combustion chamber for a rotating detonation one. No re-
numerical research is developing very fast, but should be designing of the turbomachinery would be needed. Fuel
considered as a support to experimental work. Numerical system could also stay in large parts the same, as aviation
calculations of the rotating detonation waves for different kerosene is a suitable fuel for CRD chambers.
boundary conditions in different geometries offer a great tool Application of continuously rotating detonation to existing
which not only helps better understand the mechanisms of designs of turbojet and turbofan engines would result in a
detonation propagation process but also allows for optimiz- more compact burner, and so the engines would be shorter
ing the geometry and feed parameters for future propulsion and simpler, plus due to the pressure increase in the process
systems based on this concept [6]. of detonative combustion they would achieve better perform-
ance [6]. Not to mention the lower mass and simpler design.
4.4. Continuous rotating detonation applications Moreover such an engine would be more environmentally
friendly as the NOx emission is reduced.
Main concepts of applying the CRD into existing designs
are presented below.
4.4.3. Ramjet engine
4.4.1. Rocket engine
The most straightforward solution of employing the con-
tinuously rotating detonation into an existing design is the
rocket engine. The fuel and oxidizer are injected together to
the ring-like detonation chamber, and the rotating detonation
continuously propagates, as long as the fuel and oxidizer are
supplied. Because of the fact that the products from the deton-
ation chamber are flowing out with supersonic velocity, there
is no need to apply the converging-diverging nozzle and the
aerospike nozzle can be attached directly to the detonation
chamber [1, 6]. This type of engine has been most extensively
researched and picture of running tests of such designs are Figure 16: Schematic diagram of the supersonic ramjet based on
presented in figure 14. RDE [1].
Figure 14: Pictures of rotating detonation rocket engines. (a) Pratt & Whitney (b) MBDA (c) University of Texas [6].
Figure 15: Size comparison of two turbofans using conventional combustion chamber (left) and a CRD combustion chamber (right) [1].
6
38th Engine Systems Symposium - March 2013 ESS-38-90
Z. Szawerdo / Rotating Detonation Engines
Another application of the continuously rotating deton- general, but be able to withstand them for prolonged periods
ation is the supersonic ramjet engine. In such design, the of time [20].
detonation can be organized in a special sub-chamber, where
a rich fuel-air mixture can detonate (figure 16) or in a whole 7. Conclusions
cross-section behind the normal shock wave. In both cases,
the engine would be much shorter than it is in a conventional Rotating detonation combustion offers significant advant-
ramjet [1]. ages over the typical deflagrative one. The combustion occurs
The main drawback of ramjet engines is that they cannot at a very small distance, the chamber together with the whole
operate at flight speeds that are below a certain Mach number. engine can be shorter. The pressure over the combustion
A concept of a combined cycle engine basing on CRD exists chamber rises and the thermal efficiency together with it,
to deal with this problem. It combines the rocket and ramjet significantly. NOx emissions are decreased.
designs presented above into one propulsion system. Plane There are many benefits to RDE propulsion, and the possib-
would initially accelerate on rocket mode, and then gradually ility to apply it in the existing designs looks very promising.
transfer to ramjet for cruising [1]. Naturally carrying an However with the advances in research more and more prob-
oxidizer onboard for the rocket mode would be needed, but lems arise and the introduction of RDEs to the industry is
only a small amount for take-off and acceleration. During the still in a quite distant future. But the advantages of detonative
whole cruise the engines would work on ramjet mode. propulsion, most importantly the massively increased fuel
Such a way of propulsion could be considered for a super- efficiency, drive the research forward and it is only a matter
sonic aircraft of the future. of time before we will see PDEs and RDEs in majority of
aircrafts.
5. PDE versus RDE
References
Since the research of detonative combustion engines con-
centrates mainly on these two design concepts it is useful to [1] Wolański, P. (2011) Detonation Engines. Journal of KONES
compare them together. Powertrain and Transport, 18(3).
Most notably the in a pulse detonation engine the thrust [2] Chapman, D. L. (1899) On the Rate of Explosion in Gases.
varies with time. The pulses occur at low frequency compared Philosophical Magazine, 47, 90–104.
to the rotating detonation. The design is also more complic-
ated as it needs to initiate detonation for each cycle plus either [3] Sehra, A. K. and Whitlow, W. (2004) Propulsion and Power
needs sophisticated valves or a complicated geometry for a for 21st Century Aviation. Progress in Aerospace Sciences,
40(4-5), 199–235.
valve-less system. [6]
PDEs operate at stoichiometric mixture. This not only [4] Gieras, M. (2011) Spalanie - wybrane zagadnienia w zadaniach.
means higher pressure and temperature (mixing air before Oficyna Wydawnicza PW, Warsaw, Poland.
the turbine needed - which is not the case for RDE), but also [5] Kindracki, J. (2008) Badania eksperymentalne i symulacje nu-
higher NOx emission. ˛ detonacji gazowej. Ph.D.
meryczne procesu inicjacji wirujacej
Judging on these differences the RDE can be seen as the thesis (in Polish), Politechnika Warszawska, Warsaw, Poland.
better one, and researchers certainly want to look at it this [6] Wolański, P. (2013) Detonative Propulsion. Proceedings of the
way. But it is the PDE propulsion that is almost out of the Combustion Institute, 34(1), 125–158.
laboratories, ready to be applied in the industry, and the RDE
is still at its very early development stage. [7] Nicholls, J. A., Wilkinson, H. R., and Morrison, R. B. (1957)
Intermittent Detonation as a Thrust-Producing Mechanism. Jet
Propulsion, 27(5), 534–541.
6. RDE challenges and limitations
[8] Dunlap, R., Brehm, R. L., and Nicholls, J. A. (1958) A Pre-
The rotating detonation looks very well in the laboratory liminary Study of the Application of Steady-State Detonative
in the perfect laboratory conditions. Employing it for use in Combustion to a Reaction Engine. Jet Propulsion, 28(7), 451–
real engines is in the distant future and many problems have 456.
to be overcome. [9] Kailasanath, K. (2000) Review of Propulsion Applications of
The biggest challenge is associated with the pressure losses Detonation Waves. AIAA Journal, 38(9), 1698–1708.
in the feeding system otherwise the gain from detonative [10] Braun, E. M., Lu, F. K., Wilson, D. R., and Camberos, J. A.
combustion will not be sufficient to compensate the feeding (2012) Airbreathing Rotating Detonation Wave Engine Cycle
losses, so a superior solution must be found [6]. Analysis. Aerospace Science and Technology, 1, 1–8.
Another challenge is connected to the inlet conditions. As
[11] Wolański, P., Kindracki, J., Fujiwara, T., Oka, Y., and Shima-
it has been shown they do not have to be ideally matched to
uchi, K. (2005) An Experimental Study of Rotating Detona-
design, but have to be in a certain range to allow for a proper tion Engine. 20th International Colloquium on the Dynamics
rotating detonation wave [6]. Off design performance is one of Explosions and Reactive Systems Conference Proceedings,
key of the challenges of RDE. Moreover as the combustion is Montreal.
to be operated at lean mixtures any disturbance could result
in a flame-out. [12] Tobita, A., Fujiwara, T., and Wolański, P. (2005), Detona-
tion Engine and Flying Object Provided Therewith, US Patent
Laboratory tests last typically fractions of seconds to
2005/0284127.
seconds. Not much attention is given to cooling of the com-
bustion chamber which experiences very high rates of heat [13] Bykovskii, F. A., Zhdan, S. A., and Vedernikov, E. F. (2006)
transfer. Continuous Spin Detonations. Journal of Propulsion and
Plus since the detonation wave continually runs near the Power, 22(6), 1204–1216.
head-end section of the combustion chamber, the inlet micro- [14] Kindracki, J., Wolański, P., and Gut, Z. (2011) Experimental
nozzles can be subjected to intense pressures and temper- Research on the Rotating Detonation in Gaseous Fuels–Oxygen
ature and they would not only have to withstand them in Mixtures. Shock Waves, 21(2), 75–84.
7
38th Engine Systems Symposium - March 2013 ESS-38-90
Z. Szawerdo / Rotating Detonation Engines
[15] Hishida, M., Fujiwara, T., and Wolański, P. (2009) Fundament-
als of Rotating Detonations. Shock Waves, 19(1), 1–10.
[16] Yi, T.-h., Turangan, C., Lou, J., Wolanski, P., and Kindracki, J.
(2009) A Three-Dimensional Numerical Study of Rotational
Detonation in an Annular Chamber. AIAA 2009-634.
[17] Schwer, D. and Kailasanath, K. (2011) Numerical Investigation
of the Physics of Rotating-Detonation-Engines. Proceedings of
the Combustion Institute, 33(2), 2195–2202.
[18] Eude, Y. and Davidenko, D. (2011) Numerical Simulation and
Analysis of a 3D Continuous Detonation under Rocket Engine
Conditions. Detonation Wave Propulsion Workshop, Bourges,
France.
[19] Davidenko, D., Eude, Y., Goekalp, I., and Falempin, F. (2011)
Theoretical and Numerical Studies on Continuous Detona-
tion Wave Engines. Detonation Wave Propulsion Workshop,
Bourges, France.
[20] Schwer, D. and Kailasanath, K. (2013) Fluid Dynamics of
Rotating Detonation Engines with Hydrogen and Hydrocarbon
Fuels. Proceedings of the Combustion Institute, 34(2), 1991–
1998.
