Shock Waves (2002) 11: 289 295
Self-ignition and ignition of aluminum powders in shock waves
V.M. Boiko, S.V. Poplavski
Institute of Theoretical and Applied Mechanics, Russian Academy of Sciences, Institutskaya 4/1, 630090 Novosibirsk, Russia
Received 4 December 2000 / Accepted 30 May 2001
Abstract. Ignition of fine aluminum powders in reflected shock waves has been studied. Two ignition
regimes are found: self-ignition observed at temperatures higher than 1800 K and  low-temperature igni-
tion at temperatures of 1000 1800 K. The possibility of initiating the ignition of aluminum powders in air
using combustible liquids has been studied too.
Key words: Aluminum particle-gas mixture, Reflected shock wave, Ignition
1 Introduction ings, alloying, or amalgamation to decrease or eliminate
protective features of the oxide shell (Zolotko et al. 1980).
The second approach implies thermal influence on the gas
Ignition of aluminum powders includes a wide spectrum
medium with a suspension of powder hard to ignite. This
of problems which have been attracting the attention of
may be reached using mixtures of powders of hard-to-
researchers for a long time both at the level of aluminum
oxidation kinetics and from the viewpoint of external char- ignite and combustible metals (Yagodnikov and Voronet-
acteristics of the process such as critical conditions, ig- ski 1997) or mixtures including combustible liquids (CL)
and metal powders (Boiko and Poplavski 1998).
nition delays, and effective energies of activation. Many
The most part of data on Al ignition delays tign is ob-
of these questions have been studied quite extensively,
but there are also many unsolved problems. In connec- tained under steady conditions - in heated quartz tubes,
in burners, etc. These researches allow one to reveal the
tion with the expanding use of aluminum powders, this
main kinetic regularities of oxidation and to find the crit-
topic has not lost its importance up to the present time.
Namely, aluminum is used as an admixture for rocket pro- ical temperatures of ignition. The most reliable data on
"
the critical temperature T are derived for large, easily
pellants and in technological processes and as a reagent in
controlled single particles. In the work of Derevyaga et al.
high-temperature synthesis of new materials. In addition
1977, upon inductive heating of particles with d = 3 4 mm
to individual features of aluminum dust, there is a global
and the heating rate of 380 700 K/sec, it is established
problem connected with the explosion hazard of gas-dust
mixtures of metal and organic powder materials in gen- that T = 2050 K. In the work of Ermacov et al. 1982,
upon heating of spherical particles with d = 0.4 1.2 mm
eral. In this aspect, the safety of aluminum-based powder
technologies demands the research of external character- in a CO2-laser beam with the heating rate <" 104 K/sec,
T = 2070 Ä… 50 K was obtained. The microfilming shows
istics and ignition mechanisms under conditions typical of
that ignition is related not to melting of the oxide shell
explosion processes.
(2318 K) but to the rupture of its integritydue to thermal-
The peculiarities of high-temperature oxidation of Al
mechanical tensions during heating.
are caused by the presence of an oxide shell Al2O3 whose
Concerning the small particles, a detailed control of
diffusional resistance is several orders higher than that of
their initial state is difficult; the sample is characterized
the gas. It is the initial characteristics of Al2O3 and their
statistically with a number of undetermined parameters.
change in the process of particle heating that determine
the period of induction and the critical ignition temper- In this connection, a wide scatter of the studied parame-
ters is observed. Thus, in the experiments with spherical
ature. Large periods of ignition induction are typical of
particles (d = 15 65 µm) carried out on an installation
Al powders, which does not allow full realization of their
with a gas burner (Friedman and Macek 1962), it was
energetical potential.
"
Great attention is paid to intensification of the alu- found that T is close to the melting point of the Al2O3
minum ignition processes. For example, it is material pul- oxide and practically independent of d. According to the
verization to fine and ultra-fine fractions. Various meth- data Gurevich et al. 1970 and Gurevich et al. 1978 ob-
tained for particles with d d" 50 µm on an argon  arc
ods of influence on powders or the ambient gas medium
are also used. The first method is the use of different coat- burner, T depends on the particle size and has a mini-
mum <" 1000 1300 K at d <" 6 µm, which is considerably
Correspondence to: V.M. Boiko
lower than the melting point of Al2O3. It is the opinion
(e-mail: [e-mail])
290 V.M. Boiko, S.V. Poplavski: Self-ignition and ignition of aluminum powders in shock waves
of Gurevich et al. 1970 that the determining role in this powder on a thin substrate at a level of the channel axis
process belongs to crystal-structure changes in the oxide 10 40 mm from the reflecting wall. The powder spraying
coating. As a result, the concentration of defects in the and gas-dust mixture formation occurred behind the in-
protective layer and its diffusion penetrability increase. cident SW; the ignition took place behind the reflected
In two-phase systems like suspensions of fine particles SW. The mixture ignition was registered with the help of
in a gaseous oxidizer, the ignition processes behind the a high-speed camera in the regime of multi-frame shadow
shock wave (SW) are more complicated because, along laser visualization and in the regime of photochronograph
with chemical reactions of oxidation, there is a complex of flame radiation (streak camera). It allowed registration
of physical-mechanical processes responsible for prepara- of both time and space characteristics of the process. The
tion of a combustible mixture to ignition and supporting radiation of burning particles was registered through a
subsequent burning. Relatively few works deal with the glass window 4 × 100 mm. A laser stroboscope was used
study of ignition characteristics of Al dust-gas mixtures for synchronization of the process with the moment of
"
in SW. For instance, it is established that T does not SW reflection from the wall and also for introduction of
exceed 1300 K for Al particles (d = 15 20 µm) in oxygen the time scale. The images of light marks are seen in the
behind a reflected SW (Borisov et al. 1984), which does upper part of the pictures made with the photochrono-
not contradict the data under steadyconditions (Gurevich graph (Fig. 1 is a typical streak picture). The intervals
et al. 1970) but does not agree with the data for large par- between the light pulses were "t = 100 Ä… 0.1 µs. The first
ticles. This is a typical example of studying the external pulse was formed 100 µs after SW reflection from the wall.
characteristics of the process without controlling the state The ignition delay was determined as the residence time
and behavior of separate particles. The main purpose is to the ignition site in the high-temperature area behind the
establish a tendency of heterogeneous mixtures to detona- reflected SW before luminescence appearance.
tion. It is determined from a comparison of the ignition
delay tign in the latter with tign for gas mixtures taken as
reference data. Though such tests are important for explo-
3 Self-ignition of aluminum
sion safety of powder technologies, they permit to study
dust-gas mixtures in shock waves
neither the determining parameters, nor the mechanism
of low-temperature ignition of small Al particles.
Self-ignition is understood as regimes studied previously
In the present work, results of an experimental re-
by other authors under static conditions but reproduced
search of ignition and combustion of Al powders behind
byus in shock waves for simulation of explosion processes.
a reflected SW at the temperature T = 1000 2000 K and
Figure 1 shows streak pictures of Al powder ignition with
pressure P = 1 3 MPa obtained with the help of highly
different masses of the sample. In the late stage of burning,
informative methods of photographic record are described.
separate tracks of large particles are clearly seen. In any
case, the volume glow without visible tracks corresponds
to ignition beginning. This means that the self-ignition
2 Experimental facility
of a sprayed powder begins with the smallest particles
but even for a comparatively narrow fraction composition
The experiments were carried out in a shock tube
of the sample, large particles are identified in the streak
equipped with devices for the optical visualization of the
picture and considerably affect the total burning period.
shock wave processes in two-phase media, for measuring
When m rises from 0.25 to 1 mg, the total burning period
the pressure profiles and the shock wave velocity. Driver
increases significantly, which is associated with ignition of
and driven sections were 2.4 and 5 m long, respectively;
progressively large particles of this fraction. The latter is
the channel cross section was 52 × 52 mm2. The pressure
obviously caused by the igniting affect of small particles.
of the driver gas (helium) and the driven gas (oxygen,
The further increase of m from 1 to 5 mg does not change
air) were, respectively, 2.5 10 MPa and 0.01 0.1 MPa. The
the type of glow intensity distribution in time. These re-
SW Mach range was M = 2.3 4.5. The gas parameters
sults show also that the ignition delayof a polydispersional
behind the SW were determined through the measured
sample does not depend on its mass.
Mach number M taking into account the temperature de-
The data for self-ignition delays of Al powders ver-
pendence of the specific heat ratio (Lapworth 1970).
sus the oxygen temperature behind the reflected SW are
A manufactured aluminum powder with the form of
given in Fig. 2 (1). Curve (3) is calculated as the period
particles close to spherical was used. The powder compo-
of particle heating (the average diameter of the particle is
sition was 99.2% Al, 0.2% Fe, 0.2% Si, and 0.02% H2O.
"
d =4 µm) up to some critical temperature T <" 1800 K
The initial powder was divided into narrow fractions. The
(Friedman and Macek 1962):
boundaries of separation were determined through parti-
cle sizes with equal probability of their location both in

Ád2 T - T H
large and small fractions. The fractions with particle sizes 0
tign = c ln + ,
"
of 3 5, 10 14 and 14 20 µm were used. 12 T - T T - Tm
The samples of powders with the mass m varied from
0.2 to 20 mg were introduced into the channel either with where Á =2.7 × 103 kg/m3 is the particles density,  =
the help of an electromagnetic striker placed at a distance 0(T/T0)0.75 is the coefficient of thermal conductivity of
of 10 mm from the reflecting wall, or through placing the the gas, 0 =2.4 × 10-2 J/(m·sec·K), T0 = 290 K is the
V.M. Boiko, S.V. Poplavski: Self-ignition and ignition of aluminum powders in shock waves 291
Fig. 1a c. Ignition and combustion of Al powder in oxygen for different sample masses. T = 1900 K, P =1.1MPa; a  5 mg,
b  1 mg, c  0.25 mg
4 Ignition of aluminum powders in the SW
Low-temperature ignition means a regime that do not
agree with the data of other authors obtained under static
conditions but occur under conditions of shock-wave ex-
periments. The point is that the effect of ignition of single
Al particles at medium temperatures considerably lower
"
than the critical T <" 1800 K was found in the described
experiments (see Fig. 2, points 2). Figure 3 shows typical
streak pictures of this process. Let us note its character-
istic features:
 a small part of ignited particles from the total sample
mass, and from experiment to experiment this part
is not equal (for example, from the sample 5 7 mg
(Fig. 3a) only a few dozen of particles are ignited and
their estimated mass is <" 10-5 mg);
 simultaneous ignition of particles of various diameters
including relatively large ones;
 poor recurrence of the results, namely, a wide scatter of
ignition delays under identical experimental conditions
(0.4 2 msec);
 separate ignited particles may result in the ignition of
the whole gas-dust cloud (Fig. 3b).
It is found that one of the conditions of the igniting
effect of abnormal particles on the entire sample is a high
Fig. 2. Ignition delays of Al powders in oxygen versus the
concentration of the gas-dust suspension, but the nature of
oxygen temperature: (1)  self-ignition; (2)  ignition; (3) 
abnormal particles is not known. Moreover, a few mecha-
calculation
nisms of abnormal ignition can exist. In this case, not the
averaged parameters of the system but attendant facts
initial particle temperature, Tm = 933 K is the melting may become important, such as defects of the particle
point, c =1.01 kJ/(kg·K) is the average heat capacity in surface and shape, the presence of a small share of alien
"
the temperature range from T0 to T , and "H = 4 × particles, local overheating of the medium, etc.
105 J/kg is the melting heat. A comparison of data for Al ignition delays obtained by
The data presented above for self-ignition of fine parti- Borisov et al. 1984 with the results presented here shows
cles in the SW are in excellent agreement with the results that they also may be explained through the regime of
of static tests for large particles.  abnormal ignition. In our opinion, this is the main rea-
292 V.M. Boiko, S.V. Poplavski: Self-ignition and ignition of aluminum powders in shock waves
Fig. 3a,b. Streak pictures of low temperature ignition of Al powder for different sample masses: T = 1370 K; a  m = 7 mg, b
 m =10 mg
son for the difference between the critical ignition tem-
peratures of large particles (Derevyaga et al. 1977; Erma-
cov et al. 1982) and fine fractions (Gurevich et al. 1970;
Borisov et al. 1984).
The effect of abnormal  low-temperature ignition of
separate particles might be interesting from the academic
point of view if, under certain conditions, it did not initiate
ignition of the whole sample. This circumstance simply
expands the temperature range of aluminum dust ignition;
as for danger, it levels off the phenomena of abnormal
Fig. 4a,b. Streak pictures of ignition of the Al + CL mixture;
ignition with regular regimes.
T = 1200 K; a  in oxygen, b  in air
5 Intensification of the processes
of aluminum powder ignition
The possibility of initiating  low-temperature ignition of
aluminum powders by small additives of combustible liq-
uids (CL) is studied as one of the ignition peculiarities
in the SW. It is possible to assume that, due to wide di-
versity, combustible liquids will allow one to affect the
characteristics of aluminum powders ignition in wider lim-
its. Tridecane (n - C13H28), isopropylnitrate (C3H7NO2-
IPN), and a mixture of diesel fuel (DF) with promoting
admixtures (Boiko and Poplavski 1999) were used in the Fig. 5a c. Streak pictures of ignition of an Al powder layer
present work. The influence of CL on pre-ignition heating (fraction 10 14 µm) behind the reflected SW in oxygen; P =
3.4MPa, T = 1100 K; a  pure Al powder; b, c  Al +
of metal particles in the reflected SW was studied stage
by stage beginning from concentrations typical for sus- isopropyl-nitrate for the distance between the drop and the
reflecting wall L = 5 mm (b) and L =50 mm(c)
pensions with a further decrease in the liquid share in the
system down to Õ <10-2 when only a small part of the
sample is wetted with the CL.  the ignition delays of the suspensions are determined
A drop of the suspension with a constant CL mass of by the liquid, and the metal particles are ignited later;
<" 2 mg was put in a cup fixed on a fine tungsten wire at the  the duration of suspension burning is longer than the
level of the axis of the shock tube channel at a distance of corresponding period for a drop of a pure liquid and
70 mm from the reflecting wall. Mixture spraying occurred are determined by dispersion and amount of powder
in the incident SW, as it happened in the case with pure in the drop.
powders, and ignition was observed in the reflected SW.
A more detailed description of the ignition processes
Figure 4 shows typical streak pictures of this process. The
of metal powders, pure CL and promoting admixtures, as
experiments have shown the following:
well as hybrid fuels can be found in the papers Boiko et al.
 with CL present, Al is ignited at lower temperatures 1989, Boiko et al. 1991, Boiko and Poplavski 1998, Boiko
than the critical one (1800 K) both in oxygen (Fig. 4a) and Poplavski 1999. To summarize the facts presented, let
and in air (Fig. 4b); us note the following:
V.M. Boiko, S.V. Poplavski: Self-ignition and ignition of aluminum powders in shock waves 293
Fig. 6a c. Streak pictures of ignition and combustion of Al powder (3 5 µm) in a mixture with CL for various ratios of the
solid and liquid components Õ = mAl/ml; ml = 7 mg. T = 1500 K; a  Õ =0; b  Õ =1.5 for mAl = 10 mg; c  Õ = 10 for
mAl = 100 mg
 in spite of the fact that the masses of the liquid and micro-spray both as drops of a pure liquid and as a film
metal powders are comparable, the influence of metals on particles, but the latter does not affect liquid phase
on CL ignition delays is not observed; ignition. After CL ignition, particle heating follows the
 upon transition from oxygen to air, the suspension ig- heat transfer mechanism, which is limited by the thermal
nition delay increases (by <" 3 times) and the type of conductivity of combustion products.
burning changes from explosion (in oxygen) to defla-
The correctness of this approach is confirmed by spe-
gration (in air); the same behavior was noted previ-
cial experiments on ignition of solid and liquid com-
ously for pure CL (Boiko et al. 1991).
ponents, which were not mixed beforehand (Boiko and
Poplavski 1998). In this case, an Al powder sample was
The insensitivityof CL ignition parameters to the pres-
located as described before, and there was a drop of an
ence of a considerable mass of the solid component is likely
initiating liquid 5 mm upstream. Other methods of sepa-
to be connected to the peculiarity of suspension drop de-
rate disposition of the components were also used. These
struction in the incident SW. In fact, a three-phase four-
experiments show that the wetting of the powder with a
component system is formed at the moment of CL igni-
liquid fuel as well as the method of separate disposition
tion. The burning in it is developed as follows:
of these components does not influence considerably the
 CL burning noticeably increases the gas temperature,
ignition delay and the burning type in the system.
which favors the heating of solid particles;
In the present work, a series of experiments on ignition
 after CL ignition, the temperature of some share of
of low-temperature Al powders was also performed with
metal particles (the smallest ones) rises up to the crit-
powders arranged as a thin layer on the bottom wall of
"
ical value T and their ignition occurs;
the channel. A small share of CL (<" 1 mg) was put on the
 due to burning of small particles, the medium temper-
powder layer at various distances L from the reflecting
ature increases, which results in further heating of the
end. The liquid wetted a small share of the dust sample,
"
total mass of metal particles (up to the same T ) and
and the rest sample remained dry.
their ignition.
Figure 5 shows typical streak pictures of ignition of
The CL insensitivity to the presence of the solid com- the Al powder layer (fraction 10 14 µm) in oxygen be-
ponent in the course of suspension ignition means that the hind the reflected SW for M = 2.6, P = 3.4 MPa, and
mixture formation (drop destruction, micro-spray evapo- T = 1100 K. In Fig. 5a, ignition of pure Al powder is reg-
ration) and also the ignition of the suspension micro-spray istered; Fig. 5b,c illustrates the process initiated by IPN
follows the same mechanism as pure liquids. The above at different location of the drop. Apart from the fact of Al
scheme of the process implies that CL is present in the ignition and stable burning, these streak pictures allow us
294 V.M. Boiko, S.V. Poplavski: Self-ignition and ignition of aluminum powders in shock waves
to state that the point of ignition is in the initial location
of the initiating admixture.
As was shown above, the method of disposition of the
components (different location of the drop and the pow-
der, the powder/liquid mixture put on a flat substrate or
in a cup) does not affect noticeably the ignition delays
and burning type in the system. An exception could be
the case of the powder/liquid mixture with a great excess
of the solid phase (Al particles wetted by CL). With a
small diameter of the particles and a small thickness of
the liquid film, such a complex may happen to be very
stable to breakage. The mechanisms of liquid fuel vapors
entering the gas phase will be limited byevaporation from
some fixed surface without drop destruction with its typ-
ical increase in the effective liquid surface.
A quantitative estimation of the ratio of the masses
of the solid and liquid components of a hybrid fuel is pre-
sented for the initial state of the mixture when the powder
has a packet density and the CL fills the pores between
the particles. Thus, we consider that the liquid excess may
occur behind the SW as a microspray, if the liquid volume
exceeds the total volume of the pores. It is known that
the relative volume of the pore space is <" 0.3 for the
most dense packing of a monodisperse powder; the poly-
dispersity may both increase and decrease this value. The
average mass concentration of the liquid in the pores is
assumed to be <" 0.3(ÁAl/Ál). Since we have ÁAl/Ál <" 3
Fig. 7. Ignition delays of Al powder mixtures with CL versus
for Al and liquid hydrocarbons, the mass concentration
the air temperature for various ratios of the masses of the solid
of the liquid is <" 0.1 and free liquid will be obviously
and liquid components. Õ =0.3 for mAl = 10 mg (points 1);
present in the microspray if the components have com- Õ =1.5 for mAl = 10 mg (points 2); Õ = 10 for mAl = 100 mg
parable masses. It is clear from the aforesaid why hybrid (points 3); curves 4 and 5 are approximation for the data of (3)
and (1,2), respectively; curve 6 refers to pure CL (70% diesel
mixtures are studied mainly with comparable masses of
fuel +30% CnH2n+1ONO2: n = 6 11)
the components though some experiments were carried out
with powder excess.
With an increase in the solid component concentration
According to the results of liquid fuel researches (Boiko
in a hybrid fuel, physical conditions of formation of a
and Poplavski 1999), the following mixture was chosen
combustible mixture change and the energy Ea increases.
for low-temperature ignition of Al powders in air: 70%
Thus, for Õ = 10 the ignition delays are approximated by
of diesel fuel + 30% of alcohol nitrate. Ignition delays
the equation tign =(1.2 × 10-9) exp(17.3 × 104/RT ) [sec].
of this mixture are well approximated by the dependence
The value of Ea approaches the value typical for hydrocar-
tign =(6 × 10-6) exp(3.6 × 104/RT )[sec].
bon fuel vapors such as a mixture of methane and oxygen
Figure 6 shows streak pictures of ignition and burning
diluted by nitrogen (187 kJ/mole) (Zellner et al. 1983).
of Al powder (3 5 µm) with this CL mixture for various
ratios of the solid and liquid components Õ = mAl/ml.
In these experiments, we had ml = 7 mg. A considerable
6 Conclusion
increase in the burning duration testifies to Al ignition
(see Fig. 6b). The streak picture in Fig. 6c was obtained
for Õ = 14, which correlates with the  film-like presence Ignition of fine aluminum powders in reflected shock waves
of the CL. has been studied. Two ignition regimes are found: self-
ignition observed at temperatures higher than 1800 K
Figure 7 shows the ignition delays of Al powder/CL
and  low-temperature ignition at temperatures of 1000
mixtures as functions of the air temperature behind the
1800 K. A high concentration of the gas-dust mixture is
reflected SW for various ratios of the masses of the solid
one of the conditions of  low-temperature regime.
and liquid components. It is seen that points 1 and 2
are grouped around line 5 corresponding to the equation The possibility of initiating the ignition of aluminum
tign =(1.4 × 105) exp(3.6 × 104/RT ) [sec]. For these mix- powders in air using combustible liquids with promoting
tures, we have Ea = 36 kJ/mole, which is close to the admixtures has been studied. The mixture of diesel fuel
same value for the liquid component (line 6). This sug- and alcohol nitrate is chosen with the indices  less than
gests that the conditions of generation of a combustible the least either in the effective activation energy or in
mixture behind the incident SW are practically identical ignition delays in the range T = 1000 2000 K. It is shown
for hybrid and pure liquid fuels, at least up to Õ d" 1.5. that, if the share of powder is increased up to comparable
V.M. Boiko, S.V. Poplavski: Self-ignition and ignition of aluminum powders in shock waves 295
masses of the solid and liquid components, the ignition Derevyaga ME, Stesik LN, Fedorin EA (1977) Ignition and
delays are determined by the liquid fraction with the ef- combustion of aluminum and zinc in air. Combustion, Ex-
fective activation energy typical of a pure gas-drop mix- plosion, and Shock Waves 6: 722 726
Ermakov VA, Razdobreev AA, Skorik AI et al. (1982) Temper-
ture behind the SW, and the burning period is determined
ature of aluminum particles at the time of ignition. Com-
mainly by the disperse composition of the solid phase.
bustion, Explosion, and Shock Waves 2: 256 257
Friedman R, Macek A (1962) Ignition and combustion of alu-
Acknowledgements. The work was supported by the Russian
minum particles in hot ambient gases. Combustion and
Foundation for Basic Research (Grant No. 99-01-00587) and
Flame. V.6, 1: 9 19
by INTAS (Grant No. 97-2027).
Gurevich MA, Lapkina KI, Ozerov ES (1970) Ignition limits of
aluminum particles. Fizika Goreniya i Vzryva 2: 172 176
(In Russian)
Gurevich MA, Ozerov ES, Urinov AA (1978) Effect of an oxide
References
film on the ignition characteristics of aluminum. Combus-
tion, Explosion, and Shock Waves 4: 448 451
Boiko VM, Lotov VV, Papyrin AN (1989) Ignition of gas sus-
Lapworth KC (1970) Normal shock wave tables for Air, Argon,
pension of metal powders in reflected shock waves. Com-
Oxygen, Carbon Dioxide, Carbon Monoxide, Hydrogen, Ni-
bustion, Explosion and Shock Waves 2: 193 199
trogen, Nitrous Oxide and Oxigen. London: Aeronautical
Boiko VM, Lotov VV, Papyrin AN (1991) Ignition of liquid fuel
Research Council, C. P. No. 1101
drops in shock waves. Dynamics of Deflagrations and Re-
Yagodnikov DA, Voronetskii AV (1997) Experimental and the-
active Systems: Heterogeneous Combustion, Prog. Astron.
oretical study of the ignition and combustion of an aerozol
Aeron. V.132: 205 219
of encapsulated aluminum particles. Combustion, Explo-
Boiko VM, Poplavski S V (1998) Influence of liquid hydro-
sion, and Shock Waves 1: 49 55
carbons on the ignition of metal powders in shock waves.
Zellner K, Niemitz KJ, Warnatz J, et al. (1983) Hydrocarbon-
Combustion, Explosion and Shock Waves 2: 215 219
induced acceleration of methane-air ignition. Flames,
Boiko VM, Poplavski SV (1999) Ignition of aluminum powders
Lasers, and Reactive Systems. Prog. Astron. Aeron. AIAA.
mixed with liquid hydrocarbon fuels in air. Combustion,
V.88: 252 272
Explosion, and Shock Waves 1: 31 35
Zolotko AN, Matsko AM, Polishchuk DI et al. (1980) Igni-
Borisov AA, Gelfand BE, Timofeev EI et al. (1984) Ignition
tion of a two-component gas suspension of metal particles.
of dust suspension behind shock waves. Dynamics of Shock
Combustion, Explosion, and Shock Waves 1: 20 23
Waves, Explosions and Detonations. Prog. Astron. Aeron.
V.94, AIAA: 332 339