Pulsed Detonation Engine Experimental and Theoretical Review
Pulsed Detonation Engine Experimental and Theoretical Review
Pulsed Detonation Engine Experimental and Theoretical Review
I,
AIANSAEfASMWASEE
28th Joint Propulsion
L
W
assumption of a constant volume cycle would be 8:l. discovered that detonation-tube firing occurred at
Also, this early engine was limited by h e available lower frequencies than the spark-plug energizing
cycle frequency, which in turn is limitcd by ihe Irequencies. indicating that the injection flow rate and
reaction rate. A simple calculation8 showed that if ignition were out of phase. Wartime events prevented
the combustion time could be reduced in this engine further work by Hoffmann and his co-workers.
from 0.006 sec to 0.003 sec, the thrust per pound of
fuel-air mixture would increase 100%. Thus, a The Work of Nicholls and Co-Worker%. A
propulsion device based on an explosion-cycle has substantial effort in intermittent detonation research
two main disadvantages: was made by a group headed by J. A. Nicholls11-12
of the University of Michigan beginning in the early
Constrained volume combustion (as 50's. The most relevant work concerns a set of
distinguished from constant volume experiments carried out in a six foot long detonation
combustion) does not take full advantage of the tube.ll The detonation tube was consnucted from a
pressure rise characteristic of the constant one inch internal diameter stainless steel tube. The
volume combustion process. fuel and oxidizer were injected under pressure from the
(closed) left end of the tube and ignited at some
* The frequency of the explosion cycle is limitcd distance down sueam. The tube was mounted on a
by the reaction rate, which is only slightly pendulum platform, suspended by support wires.
higher than the deflagrative combustion rate. Thrust for single detonations was measured by
detecting tube (platform) movement relative to a
The main advantage of the constant pressure stationary pointer. For multi-cycle detonations,
cycle is that it leads to engine configurations with thrust measurement was achieved by mounting the
steady state processes of fuel and oxidizer injection, thrust end of the tube to the free end of a cantilever
combustion, and expansion of the combustion beam. In addition to direct thrust measurements. the
W products. These stages can be easily identified and the temperature on the inner wall of the detonation tube
engine designer can optimize them on the basis of was measured. Fuel mixtures of hydrogedoxygen,
relatively simple steady state considerations. hydrogenlair, acetyleneloxygen and acetylenelair
mixtures were used. The gaseous oxidizer and fuel
were continuously injected at the closed end of the
There have been numerous attempts in the past detonation tube and three fixed flow rates were
to take advantage of detonative combustion'for engine investigated. Under these conditions, the only
applications. The following is a brief description of parameters that could be varied were the fuel/oxidizer
some of the most relevant past experimental and ratio and frequency of ignition. A maximum gross
analytical studies of pulsed detonation. -
thrust of 3.21b was measured in the hydrogenlair
-
mixture at the frequency of 30 detonations per
The Work of N. Hoffmann. The first reponed second. The most promising results were
work on intermittent detonation is attributed to demonstrated for the Hdair mixture, where a fuel
HoffmannlO in 1940. Hoffmann's experiments on specific impulse of Isp = 2100 sec was reached. The
intermittent detonation were carried out in a long, maximum frequency of detonations obtained in all
narrow tube mounted on a test stand using acetylene- experiments was 35 Hz. The temperature
oxygen and benzine-oxygen fuel mixtures. Water measurements on the inner wall showed that for the
vapor was added to prevent the highly sensitive highest frequency of detonations the temperature did
acetylene-oxygen mixture from premature detonation. not exceed 800° F. This temperature is
Hoffmann pointed out the importance of the approximately the mean between the temperature of
detonation initiation (spark plug) location in reference the injected gasses and the detonation wave
to tube length and diffuser length. It was found that a temperature averaged over the cycle frequency.
continuous injection of the combustible mixture leads
to only a narrow range of ignition frequencies that In their later w 0 r k , ~ ~ - lthe
5 University of
./ will produce an intermittent detonation cycle. These
frequencies are governed by the time required for the
Michigan group concenuated on development of the
Rotating Detonation Wave Rocket Motor. No further
v mixture to reach the igniter, time of transition from work on the pulsed detonation cycle was pursued.
deflagration to detonation, and time of expansion of
the detonation products. Hoffmann attempted to find The Work of L. J. Krzvcki , In a setup very
the optimum cycle frequency experimentally. It was similar to Nicholl's, L. J. Krzycki16 performed an
3
W
experimental investigation of intermittent detonations experimenlal work related to the pulsed detonation
with frequencies up to 60 cps. An attempt was also engine concept stopped at this time.
made to analyze the basic phenomena using unsteady
gas dynamic theory. Krzycki's attempt to analyze the Russian Work on Pulse Detonation Devices. A
basic phenomena relied on wave diagrams to uace review of the Russian literature has not uncovered
characteristics, assumptions of isentropic flow for work concerning applications of pulsed detonation
detonation and expansion, and incompressible flow devices to propulsion. However, there are numerous
for mixture injection processes. The most repons of applications of such devices for other
convincing data from the experiments are the purposes such as for producing niuogen oxide1* (an
measurement of thrust for a range of initiation old Zeldovich idea to bind nitrogen directly from air
frequencies and fuel mixture flow rates. to produce fertilizers) and as rock crushing devices.19
Unfortunately no direct pressure measurement in the
device is reported, so there is only indirect evidence of Korovin et a1.18 provide a most interesting
the nature of the process observed. account of the operation of a commercial detonation
reactor. The main objective of this study was to
The basic test stand used by Krzycki is very examine the efficiency of thermal oxidation of
similar to that used by Nicholls and his co-workers. nitrogen in an intermittent detonative process as well
The length of the detonation tube and the internal as an assessment of such technological issues as the
diameter were exactly the same as those in Nicholl's fatigue of the reactor pans exposed to the intermittent
experiments. Figure 3 presents a schematic of the detonation waves over a prolonged time. The reactor
experimental apparatus containing common, generic consisted of a tube with an inner diameter of 16 mm
elements of the Hoffmann-Nicholls-Krzycki and length 1.3 m joincd by a conical diffuser to a
experiments. A propanehir mixture was second tube with an inner diameter of 70 mm and
continuously injected through a reversed-flow diffuser length 3 m. The entire detonation reactor was
for better mixing, and was ignited at the same submerged in running water. The detonation mixture
W
distance as in the Nicholls' experiments from the was inuoduced at the end wall of the small tube.
injection point by an automobile spark plug. The CHq.02 and N2 comprised the mixture composition
spark frequency was varied from 1 to 60 cps. The and the mixture ratios were varied during the
spark plug power output was varied inversely with continuous operation of the reactor. The detonation
the initiation frequency. and at the frequency of 60 wave velocity was measured directly by piezoelectric
cps was only 0.65 Joule. This value is too low for sensors placed in the small and large tubes. The
direct initiation of a detonation wave by the spark, detonation initiation frequency in the reactor was 2-16
and consequently all of the experiments must have Hz. It is reponed that the apparatus operated without
been based on transition from deflagration to significant maintenance for 2000 hours.
detonation. According to experimental data and
the01y.l~direct initiation of a mixture of propandair Smimov and Boichenko19 studied intermittent
at the detonability limifs requires an energy release on detonations of gasoline-air mixtures in a 3 m long
the order of lo6 Joules. Thus, we conclude that the and 22 mm inner diameter tube operating in the 6-8
required deflagration-detonation transition region Hz ignition frequency range. The main motivation
length in Kryzcki's experiments would have been for this work was to improve the efficiency of a
prohibitively large for the propane/& mixture. It commercial rock crushing apparatus based on
follows that in all of the experiments a substantial intermittent detonations of the gasoline/air
pan of the process was deflagrative. This resulted in mixtures.20 The authors investigated the dependence
low efficiency and negligible thrusr Knycki repeated of the transitional region length from deflagration to
Nicholls' experiments using basically the same rates detonation on the initial temperature of the mixture.
of injection of the detonable mixtures. Krzycki's
experimental results are very well documented, As a result of the information contained in the
allowing us to deduce a clear picture the physical Russian repons, we conclude that reliable commercial
processes occurring in the tube. The author anived at devices based on intermittent detonations have been
I the conclusion that thrust was possible from such a consuucted and operated.
device but practical applications did not appear
promising. It is unfortunate that, possibly based on Pulsed Solid Exnlosion Stud ies at JPL. Work
Krzycki's extensive but misleading results, all at the Jet Pro ulsion Laboratory (JPL) by Back, Varsi
and others27-24 concerned an experimental and
4
U
theoretical study of the feasibility of a rocket thruster wave is about two orders of magnitude higher than
based on intermittent detonations of solid explosive the speed of a typical deflagration wave. Because of
for propulsion in dense or high-pressure aunosphcrcs this, very high power densities can be created in the
of cenain solar system planets. The JPL work was detonation chamber. Each detonation can be initiated
directed at very specific applications: however, these independently and. depending on the chamber
studies also addressed more general key issues geomeuy and external flow characteristics pertaining
concerning intermittent propulsion devices such as to a particular device, a wide range of frequencies is
propulsion efficiency. In this work, a Deta sheet type possible. There is no theoretical restriction on the
C explosive was detonated inside a small detonation range of operating frequencies: they are uncoupled
chamber attached to nozzles of various length and from any acoustical chamber resonance. The
geometry. The nozzles, complete with firing plug, independence of detonation cycle frequency is the
were mounted in a containment vessel that could be feature that most differentiates the PDE concept from
pressurized with mixtures of various inert gases from the pulsejet. I t is also the feature that leads
vacuum to 70 a m . The apparatus directly measured theoretically to scalability of PDE configurations for
the thrust generated by single detonations of a small a wide range of flight applications. A key physical
amount of solid explosive charge expanding into restriction on the range of allowable detonation
conical or straight nozzles. Thrust and specific frequencies arises from the rate at which the fresh
impulse were measured by a pendulum balance fuellair mixture can be introduced into the detonation
system. chamber. Obviously the detonation products must be
discharged from the chamber before fresh charge is
The results obtained from the JPL experimental injected.
study of an explosively driven rocket led to the
5
v -
formed in the chamber and propagated with the Mach 1. Initiation and propagation of the detonation
number appropriate for the fuel-air mixture. The fucl wave inside the chamber,
used in the NPS experiments was ethylene and the
maximum detonation frequency obtained was 25 Hz. 2 . Expansion of the detonation products from
limited only by the mechanical nature of the solcnoid the chamber into the air slrcam around the chamber at
valve used for fuel injection control. Figures 5 and 6 flight Mach numbers;
are two frames from a videotape of the early NPS
experiments. Figure 5 shows the experimental 3. Fresh air inlake from the surrounding air
apparatus and Figure 6 shows the apparatus during into the chamber:
repetitive detonation. The figures also show the fuel
injector ring between the two concenuic detonation 4. The flow pattern inside the chamber during
chamber cylinders. It was determined that the post-exhaust pressure buildup, which determines the
duration of a single cycle was less than 7 msec. This strategy for mixing the next detonation charge:
means that the N P S device could have potentially
operated at frequencies up to 150 Hz in the static or 5. Strong mutual interaction between the flow
no flow (M = 0) case. At the time of the N P S inside the chamber and the external flow surrounding
experiments, performance extrapolations included the engine.
thrust levels up to 40 Ibs at 100 Hz. As described
later, SAIC simulations of static operation show All of these processes are interdependent, and
higher thrust levels at these frequencies due to new interaction and timing are crucial to engine efficiency.
ideas and improvements in the PDE concept. These Thus, unlike simulations of steady state engines. the
new ideas are incorporated in the generic PDE phenomena described above cannot be evaluated
concept independenuy. It is a challenging computational
problem to resolve the flow regime inside the
The Generic PDE Device chamber to account for nozzles, air inlets. etc., and at
W In this section, we refer to the generic PDE the same time resolve the flow outside and
device, which is represented as a small engine in surrounding the engine, where the flow regime varies
Figure 7. The figure shows a schematic of the basic from high subsonic, locally transonic and supersonic.
detonation chamber aaached to the aft end of a generic
aerodynamic vehicle. A combustible gas mixture is The single most important issue is to determine
injected at the closed end of the detonation chamber the timing of the air intake for the fresh charge that
and a detonation wave is shown propagating through leads to repetitive detonations. It is sufficient to
the mixture. Also shown are air injection inlets and assume inviscid flow for the purpose of simulating
an important part of the device that we have termed the expansion of the detonation products and fresh air
the thrust wall. The schematic suggests a small- intake. The assumption of inviscid flow makes the
payload, aerodynamic vehicle: however, as we task of numerically simulating the PDE flow
describe later, the concept can be extended to larger phenomena somewhat easier than if a fully viscous
payloads simply by scaling up the size of the flow model were employed. The effects of viscous
detonation chamber and possibly combining a number boundary layers are negligible for the size of the
of chambers into one larger engine. generic device studied in this work, with the
exception of possible boundary layer effects on the
The geomeay of the main detonation chamber, valve and inlet gmmemes discussed subsequently.
which determines the propulsion efficiency and the
duration of the cycle (frequency of detonations), is a SAIC has performed an extensive study of the
key issue for the PDE concept. Since the fresh generic PDE over a wide range of operating
charge for the generic engine is supplied from the conditions for a wide range of device
external flow field, the efficiency of the engine configurations?6-30 Numerical simulations of the
depends on the interaction of the surrounding flow unsteady flow and detonation processes, in addition to
with the internal flow dynamics. Following is a theoretical analysis. have resulted in an understanding
, partial list of the broad range of physical processes and an approach to analyzing and evaluating PDE
requiring simulation in order to model thc complex propulsion pcrformance. Although the basic concept
v
flow phenomena associated with the detonation rcmains the same, there are subtle differences in the
engine performance: PDE manifestation for particular applications. These
will be described subsequently. Details of the
6
d -
numerical simulations (including assumptions used thousands of Ibs.) depending on the engine cross
for detonation wave physics and chemistry, use of sectional area, length and detonation frequency.
adaptive unsmctured grids and Godunov methods for
the Euler Kasdynamic equations) are given Subsonic-TransonicOneration
elsewherez6-J1 The following section is a summary PDE operation in the subsonic-transonic regime
of the results from numerical and theoretical studies differs from the static casc in that the self aspiration
of various applications and operating regimes Tor the effcct decreases with increasing Mach number. This
generic PDE. is due to the formation of a rear stagnation point
behind the exhaust plane above certain Mach numbers
e Reeimes for given geometries. The stagnation region prevents
complete detonation product exhaust and subsequent
In this section we summarize the results of fresh charge injection. For example, over the Mach
several applications and operating regimes identified number range. O+ c M < 0.5, full to partial self
in the course of our studies of the PDE concept. aspiration occurs: the effectdecreases rapidly for Mach
numbers above 0.5, resulting in the need for some
M = 0 static 00erauog' type of air inlet or air intake valve configuration.
Under static conditions, M = 0, the PDE is Simulations of various detonation chamber and air
completely self-aspirating. Such was the c3se for the inlet geomerries26,28 have shown that, depending on
early NPS PDE studies. Without an external the free-stream Mach number, appropriate shaping of
airstream, the PDE must obtain fresh air charge as a the air inlet geomevy and total inlet area leads to
result of the detonation chamber overexpansion propulsion engines that are attractive for certain
immediately following exhaust of air-fuel detonation applications. We present here a summary of
products. To the lowest approximation, the available studies28 carried out in an attempt to find a
time for chamber refill due to this overexpansion satisfactory PDE configuration for a small missile
process is, for a given chamber geometry and fuel-air engine (the final configuration was not optimum, by
combination, directly proportional to its length. For any means, since all variables were not parametrically
M = 0 operation, we assume that the PDE Varied).
configuration does not contain any air inlets other
than the aft end of the device or. if inlets are present. A PENAID-type missile with associated mission
they are closed. Simulationsz6 of M = 0 PDE requirements such as range, speed, system weight,
operation show that the time required for fresh air total thrust, and specific fuel consumption was used
refill for a device with dimensions equivalent to the for the study. The detonation chamber dimensions
NPS experimental apparatus is on the order of 6-7 were 6 cm diameter and 9 cm length with a cylindrical
msec. This agrees with the NPS results and means cross-section. A schematic of PDE integration into
that a maximum frequency of 150 Hz should be such a missile configuration is shown in Figure 8.
possible. Simulated thrust levels were higher than The simulations showed that, for practically all cases
those estimated from scaling the NF'S results. This involving simple inlets (circumferential slits around
is due to a new operating scenario that was uncovered the cylindrical cross-section), the thrust data were
by the simulations: detonation initiation from the aft independent of whether the inlets open intermittently
end resulfs in the kinetic energy of the shock wave (valved) or remain open during operation. This is due
being transferred to the thrust wall. The amount of partially to the very short time that detonation
extra thrust obtained from this mode of operation is products have to escape from the inlets thereby adding
considerably larger than that expected from to negative thrust: this negative thrust, determined in
gasdynamic expansion following detonation initiation the simulations, is negligible compared to the total
at the thrust wall. The physical reason for this is integrated thrust. The thrust data do indicate a sEong
found in the shock wave energetics. dependence on external flow conditions. e.g., Mach
number. The Mach number plays a role in the wave
The importance of M = 0 PDE performance is drag: the details of valve and inlet configuration
associated with applications of the concept for full geomeuy figure prominently in the total wave drag.
scale aircraft propulsion, including rollout and These studies answered an i m p o m t question: can an
takeoff. Simple scaling laws derived from the air inlet be configured such that the inlet remains
numerical simulation results and described later, show open over the full flight regime and operating
that M = 0 thrust levels can be large (tens of conditions? The answer is "yes." Thus, at least for
this regime, the PDE offcrs the possibility of a no-
7
-
-
moving-parts propulsion device. For the PENAID Thrust................................................ 200 Ib.
missile under discussion here, a configuration was Fuel flow rate................................ 0.05 Ibhec.
found that operates between 0.2 < M < 0.9 with open Fuel weight for 12 min........................... 36 Ib.
air inlets. Oxygen weight..................................... 3.6 Ib.
Fuel for detonation tube......................... 1.2 Ib.
The foUowing performance data were obtained for Total oxygen and fuel weight................ 40.8 Ib.
the PENAID missile configuration. For M=0.8 at Total engine weight............................. 54.2 Ib.
sea level altitude and a detonation frequency, f-100 Specific fuel consumption........ LO9; Ib./(lb.*hr.)
Hz,the PDE characteristics arc:
The examples of the PDE device performance
Thrust.................................................. 79 Ib. given above are based on point design conditions
Fuel flow rate............................. 0.025 Ib./sec. arising from the simulations reported earlierz6 They
Fuel weight for 12 min........................... 18 Ib. cannot be extrapolated with any degree of reliability
Oxygen weight...................................... 1.8 Ib to other conditions or configurations. We conclude,
Fuel for detonation tube......................... 0.6 Ib. however, that the performance computed for the
Total oxygen and fuel weight................. 20.4 Ib. indicated device is encouraging from the point of view
Total engine weight............................. 30.2 Ib. of thrust, thrust control, simplicity of the device (no
Specific fuel consumption........ 1.14 Ib./(lb.*hr.) moving pans), and specific fuel consumption (SFC).
The specific fuel consumption computed above is
Assuming the PDE device geometry is kept competitive with present day small turbojet engines.
fixed. a higher detonation frequency will result in a The SFC for a PDE could be significantly lower than
linear increase in thrust and fuel flow rate at the same for small turbojets (SFC's for small turbojets are in
specific fuel consumption. For example, if the the range of 1.8-2.0 Ib./(lb.*hr)). Thus, for a given
detonation frequency is increased to 200 Hz. the mission and vehicle, a PDE propulsion unit may be
performance data are: more fuel efficient, resulting in increased range.
W
Moreover, if the expected thrust control in PDEs is
T h s t ............................................... 157 Ib. realizable, it may be possible to produce propulsion
Fuel flow rate................................ 0.05 Ib/sec. units that can slow down, loiter and maneuver, and
Fuel weight for 12 min........................... 36 Ib. finally accelerate to full thrust again rapidly.
Oxygen weight .................................... 3.6 Ib. Depending on the detonation frequency, which
Fuel for detonation tube......................... 1.2 lb. determines the thrust for all other conditions fixed,
Total oxygen and fuel weight ................ 40.8 lb. the thrust-to-weight ratio for the PDE can be as high
Total engine weight ............................ 54.4 Ib. as 201. This value is certainly competitive with
Specific fuel consumption......... 1.14 Ib./(lb.*hr.) other propulsion concepts.
At lower Mach numbers, M=0.5. the maximum The results of the scaling studies at subsonic-
operating frequencies for constant thrust will be lower transonic speeds lead to scaling laws that can be used
since the external dynamic pressure responsible for to predict the performance of PDE's over some range
supplying fresh air to the chamber is also lower. For of parameters, assuming that other parameters are held
the device under consideration here, the maximum fixed. For example, holding the external Mach
frequency is 250 Hz. For a frequency of 100 Hz: number and basic chamber and inlet geomeuy fixed
suggests that the thrust at constant specific fuel
Thrust ............................................... 100 Ib. consumption produced by the PDE scales as:
Fuel flow rate.............................. 0.025 Ib/sec.
Fuel weight for 12 min........................... 18 Ib.
Oxygen weight .................................... 1.8 Ib.
Fuel for detonation tube......................... 0.6 Ib. where Ti. (v/vl) and (f/fl) are the thrust computed for
Total oxygen and fuel weight................. 20.4 Ib. a chamber of volume v i operating at frequency fl. the
Total engine weight............................. 30.2 Ib. ratio of a new volume to v i and the ratio of the new
Specific fuel consumption..........0.9 Ib./(lb.*hr.) frequency to f l , respectively. Thus, thrust should
W scale linearly with the parameter (v/vl) * (f/fl) over
Again, if the frequency is increased the thrust some range of this parameter. Depanure from this
will increase linearly: operation at 200 Hz yields: linear variation may occur due to the following
argument First, since volume is proponional to the
u -
.
-
product of cross-sectional area and length, v r21, (r - a function of air inlet opening. To find an optimum
detonation chamber radius, I - chamber length) configuration that satisfies given performance over a
physical limits will be placed on r and 1; if r is too wide flight regime requires a more extensive
small (less than 1 cm), a detonation will not be simulation study. It was mentioned earlier that the
sustainable and if 1 is too small (less than 10 cm), it simulations presented here were carried out under the
may be difficult to mix fuel and air effectively. assumption of inviscid flow: boundary layer effects
Using the thrust relation established above, we make were not included. Boundary layers are only
the following observations. For a PDE device significant for the air inlets and valves.
producing 100 pounds thrust at 100 Hz, doubling the
frequency and increasing the volume by a factor of 5 There is an important feature of PDE operation
yields a thrust level of 1000 pounds. Assuming that for missiles such as the one considered here: if the
the aspect ratio of the chamber (chamber Iengh to expected thrust control is attainable, then the
radius) is fixed, this would require an engine only detonation frequency can be varied to produce constant
25.5 cm in diameter and 25.5 cm in length. Of thrust over a given flight envelope. or the frequency
course, the relation between thrust and (v/vl) * (f/fl) can be varied to make the missile slow down, loiter
cannot be believed over too wide a range of and maneuver, and finally ramp back to full thrust
parameters; but, it does serve to point out the more or less instantaneously. Since each detonation
flexibility permitted by the PDE concepr is controlled separately, this capability should depend
only on on-board electronics and power.
The subsonic-transonic simulations showed that
the timing of the fresh air refilling required to Suoersonic-HvnersonicOoeration
recharge the chamber for subsequent detonations is a Numerical simulations have been carried out for
strong function of the details of the valve and inlet PDE operation in the supersonic and hypersonic
geometry, the expansion of the combustion products, flight regimes.29 The results of these simulations
the resulting over-expansion of the chamber flow, and show that there are differences when compared with
v
the external flow regime and interaction of the the lower speed regimes. The main difference, with
external flow with the internal flow. For subsonic respect to operating characteristics, is the air intake
flight, Mach 0.2-0.9, the fresh air entering the inlet must be more carefully considered. For
chamber comes from two separate principal flow supersonic and hypersonic flow air scoops may be
processes: one comes from the flow through any required, adding to wave drag. For PDEs enclosed in
valve or inlet and the other comes from the self- a duct connected to upstream air inlets, pressure
aspiration or reverse flow from the aft end of the recovery from free-stream to duct inlet and finaly to
chamber due to strong over-expansion. All these PDE inlet must be accounted for. To date, several
processes are interdependent and, in order to search for detailed studies have been carried out for the higher
a given performance in a given device, require speed regimes: a supersonic, M = 2 PENAID missile
variation of many parameters. The simulation results engine simulation and a sizing analysis for a large
obtained to date provide an understanding of the engine operating in the supersonic to hypersonic
effects caused by variation of the above-mentioned flight regime.
parameters. With the information available, we
conclude that a PDE propulsion unit can be optimized umrsonic M = 2 P D E The M = 2 PENAID
(although no optimization studies were carried out) miszle study has been reported earlierz9 and,
for a given flight regime. The decrease in thrust with representative simulation results are shown on the
increasing Mach number has been described earlier to cover of this review paper. It was found that a fixed
result from increased wave drag produced by the inlet air inlet geometry could be conceprualized to operate
geometry. Optimization of the inlet geometry could over the Mach number range, 0.5 < M < 2. By this
help to eliminate a large part of the wave drag. The is meant the timing for fresh air charge allowed a
simulation data can be used to determine the detonation frequency of 200 H z at M = 2 and this, in
detonation frequency at a given Mach number yielding turn, means that any lower frequency is allowable at
constant thrust. For example, for a constant thrust any other Mach number below M = 2. Detonation
level of 90 pounds, the required detonation frequency frequency control may result in enhanced control over
W varies from 84 Hz at M-0.0 to 140 Hz to M=0.8. In missile flight trajectory since a constant thrust, a
a similar fashion, we can obtain parametric variations cruise-dash-loiter-cruise or any other tailored thrust
of other important aspects of PDE performance, such profile can be realized. We conclude that supersonic
as minimum time for refill at given Mach number as PDE operation appears possible for missile
v
applications. and there may also be advantages for
(5
7
longer range air-to-air missiles due to enhanced
propulsion energy management capability. pz= 1 . 2 c p 0 ~ -
+$ (4)
M 3 = 2 = f L q & - . (8)
The simplest condition to assume for the PDE air
inlet is choked flow. Although this is not valid over u3
much of the required regime, certainly not for Since the total pressure in the chamber equals the
subsonic external flow, it will result in a pessimistic total pressure at the PDE inlet, the static pressure in
bound on the sizing m e t e r s . Using well known the chamber as a function of chamber Mach number,
gasdynamic analysisg2 the static and tool pressures given in Eq. (8), can be related to the free-stream
and density at the PDE inlet can be found as: static pressure as follows:
7
( $( @cy5$
P,=CPo 1 + - - 1+ I-- (9)
10
shows the effect of frequency variation on A2/A3.
Recall the design point was at f = 100 Hz. Figures
10-13 contain the answers to the questions asked
Another relation between P3 and Po as a function of during this sizing analysis; reasonable physical sizes
M3 can be given as: for PDEs developing high thrust levels are predicted.
7 -7
A more rigorous analysis is required to validate these
(
P3=CPo 1 + 0-$)T(l+$)T (IO)
predictions.
11
U .
issues as detonation initiation, air inlet design Explosion-Cycle Combustion Chamber of a
including boundary layers, fuel/air injection and Jet Propulsion Engine." NACA TN-1702,
mixing can be studied and understood. In addition, September 1948.
thrust measurements, both static and in an external 9. S h u I t z - G r u no w , F . , " G a s - D y n am ic
flow are required to validate the numerical and Investigation of the Pulse-Jet Tube," NACA
theoretical predictions. Plans for such an TM-1131, February 1947.
experimental program are presently under 10. Hoffmann, N., "Reaction Propulsion by
consideration. Intermittent Detonative Combustion,''
Ministry of Supply, Volkenrode Translation.
1940.
Acknowledme ne 11. Nicholls, J.A., Wilkinson, H.R. and
The work reviewed in this paper received partial Morrison, R.B., "Intermittent Detonation as a
support from the Defense Advanced Research Thrust-Producing Mechanism," Jet Propulsion,
Projects Agency under the following contracts: 27, 534-541, 1957.
N66001-88-D-0088 (NOSC) and F49620-89-C-0087 12. Dunlap, R., Brehm, R.L. and Nicholls, J.A.,
(AFOSR). The authors would like to acknowledge "A Prcliminary Study of the Application of
Robert Pegg and Lany Hunt of NASA LaRC, John Steady State Detonative Combustion of a
Leingang of the Wright Aeronautical Laboratory, and Reaction Engine," ARS J., 28, 451-456,
Major David Neyland of DARPA for their interest and 1958.
for stimulating discussions concerning PDE 13. Nicholls, J.A., Gullen, R.E. and Ragland
applications. Over the course of these studies. SAIC K.W., "Feasibility Studies of a Rotating
colleagues Ed Rogan, Adam Drobot, Isaac Lottati and Detonation Wave Rocket Motor," Journal of
Ahron Freidman provided support and advice on many Spacecrafts and Rockets, 3,893-898, 1966.
of the technical areas required to carry out this work. 14. Adamson, T.C. and Olsson, G.R..
"Performance Analysis of a Rotating
W
Note Added in h o o f The authors would like to thank Detonation Wave Rocket Engine,"
MI Robert Pegg of NASA's Langley Research Center AsUonautica Acta, 13.405415, 1967.
for bringing their attention to early documents and 15. Shen, P.I. and Adamson, T.C., "Theoretical
reports describing pulsed detonation engine research. Analysis of a Rotating Two-Phase Detonation
in Liquid Rocket Motors,'' Asmonautica Acta,
Referenm 17,715-728, 1972.
16. Krzycki, L.J., Performance Characteristics of
1. Scott, William B.. "Renewed Interest in Pulsed an Intermittent Detonation Deviee. Navweps
Engines May be Linked to Black Aircraft," Report 7655, U S . Naval Ordnance Test
Aviation Week, 28 October 1991 (68-69). Station, China Lake, California 1962.
2. Scott. William B., "New Evidence Bolsters 17. Matsui, H. and Lee, J.H.. "On the Measure of
Reports of Secret, High-speed Aircraft,'' the Relative Detonation Hazards of Gaseous
Aviation Week, M a y 11,1992 (62-63). Fuel-Oxygen and Air Mixtures," Seventeenth
3. Wolfe, M.O.W., Luck, G.A., "Pressure Symposium (International) on Combustion.
Measurements on the F.Z.G. 76 Flying Bomb 1269-1280, 1978.
Motor,'' Technical Note No. EA237/1, Royal 18. Korovin, L.N., Losev, A., S.G. Ruban and
Aircraft Establishment, Farmborough, 1944. Smekhov, G.D. "Combustion of Natural Gas
4. Tharratt, CE., "The Propulsive Duct," Aircraft in a Commercial Detonation Reactor," Fiz.
Engineering. November 1965. (327-337). Gor. Vzryva, Vol. 17, No. 3, p. 86, 1981.
5. Tharratt, C.E., ibid, December 1965, (359- 19. Smirnov, N.N. and Boichenko, A?.,
371). "Transition from Deflagration to Detonation in
6. Tharratt, C.E., ibid, February 1966, (23-25). Gasoline-Air Mixtures," Fiz. Gor. Vzryva,
7. Kentfield, J.A.C., "Valveless Pulsejets and 22, No. 2, 65-67, 1986.
Allied Devices for Low Thrust, Subsonic, 20. Lobanov, D.P., Fonbershtein, E.G. and
Propulsion Applications," AGARD Conf - Ekomasov, S.P., "Detonation of Gasoline-&
W Proc. No. 307, Ramjets and Rockets for Mixtures in Small Diameter Tubes," Fiz. Gor.
Military Applications, March (1982). Vuyva, 12, No. 3,446, 1976.
8. Zipkin, M.A. and Lewis G.W., "Analytical
and Experimental Performance of an
12
v
21. Back, L.H., "Application of Blast Wave
Theory to Explosive Propulsion." Acta
Asuonautica, 2, No. 5/6. 391-407, 1975.
22. Varsi, G.. Back, L.H.and Kim, K.. "Blast
Wave i n a Nozzle for Propulsion
Applications," Acta Astronautica, 3, 141-156,
1976.
23. Kim, K., Varsi, G. and Back, L.H., "Blast
Wave Analysis fa Detonation Propulsion,''
AIAA Journal, Vol. 10, October 1977.
24. Back, L.H., Dowler, W.L. and Varsi, G.,
"Detonation Propulsion Experiments and
Thcory," AIAA Journal Vol. 21, October
1983.
25. Helman. D., Shreeve, R.P. and Eidelman, S.,
"Detonation Pulse Engine," AIAA-86-1683,
24th Joint Propulsion Conference, Huntsville,
1986.
26. Eidelman, S., Grossmann, W. and Lotrati, I..
"Propulsion Applications of the Pulsed
Detonation Engine Concept," SAIC Report
Number 89/1684, December 31,1989.
27. Eidelman, S . , Grossmann, W. and Lottati. I.,
"A Review of Propulsion Applications of Ihe
Pulsed Detonation Engine Concept," J.
v Propulsion and Power, Vol. 7, No. 6,
November-December 1991 (857-865).
28. Eidelman, S. and Grossmann, W.,
"Computational Analysis of Pulsed Detonation
Engines and Applications," AIAA-90-0460,
January 8-11,199Q/Reno, Nevada.
29. Eidelman, S., Grossmann, W. and Lottati, I.,
"Air-Breathing Pulsed Detonation Engine
Concepc A Numerical Study." AIAA-90-2420,
July 16-18, 1990/0rlando. Florida.
30. Eidelman, S.,Lottati,I. and Grossmann, W.,
'"A Parametric Study of the Air-Breathing
Pulsed Detonation Engine," AIAA-92-0392,
January 6-9,1992/Reno, Nevada.
31. Lottati, I.. Eidelman, S. and Drobot, A,, '"A
Fast Unsrmctured Grid Second Order Godunov
Solver (FUGGS)," AIAA-90-06l9, January 8-
11,1990/Reno, Nevada.
32. Anderson, J. D., essible Flow,
McGraw-Hill,w- 2 York.1982.
13
AI
FY.1-
AI
----
‘I
-detonation
a m wave 0.72
0.66
Figure 7. Schematic of the generic PDE.
0.64
-
A3 m*2
i . ;o ~~
0.65
aa3 0.6
0.55
Figure 8. Schematic of PDEPENAID missile
integration. 0.45
Mad number
W
h - -
10.000 m. A3-1.2 Prec = 1.0.
M - 2 . Vol 2 m*3 0.84
0.78 i 1
frequency s$ 0.72
z
80 100 120 140 160 0 .-c 0.66
YI
.2 0.6
P
;
m
0.54
0.48
I: 0 . 4 2
0.6 0.36
0.3
0.5 1 1.1671.333 1.5 1.6671.833 2
Engine Lenglh in Meters
Figure 13. Results for AdA3 as a function of Figme 15 PDE engine radius (cylindricai cmss-
detonation frequency. section) versus engine length.
w
Thrust vs Volume lor 1-1OOcpo
M
n
G 32.8
2
29.6
26.4
23.2
20 '
1 1.167 1.5
1.333 1 . 6 6 7 1.833
I
2
PDE Chamber Volume In Cubic Melers