JOURNAL OF PROPULSION AND POWER

Vol. 21, No. 1, January–February 2005

Plasma Ignition and Combustion of JA2 Propellant

Jianquan Li,∗ Thomas A. Litzinger,† and Stefan T. Thynell†

Pennsylvania State University, University Park, Pennsylvania 16802

Experiments were performed to investigate the effects of various parameters on plasma-driven ignition and
combustion of a double-base propellant under closed-chamber conditions. The parameters varied include input
electrical energy, nozzle length and inner diameter, nozzle exit to propellant distance, as well as propellant sample
thickness. Chamber pressure was measured to determine the ignition delay and to deduce the regression rate.
High-speed images of the plasma jets and combustion event were also recorded. At low plasma energies, rapid,
plasma-driven burning occurred, but self-sustained burning was not achieved. With moderate plasma energies,
combustion of the propellant exhibited a two-stage burning behavior: one stage of plasma-driven rapid burning
that occurred during the plasma pulse and a second stage of slower self-sustained burning that occurred with a
clear delay after the first stage. When plasma energy was increased further, the two-stage behavior became less
Downloaded by UNIVERSITY OF QUEENSLAND on June 16, 2014 | http://arc.aiaa.org | DOI: 10.2514/1.5866

distinct and eventually disappeared, leaving only one stage of burning. Nozzle length and diameter affected the
ignition and combustion characteristics as a result of energy losses from the plasma as it flows through the nozzle.
The propellant burning behavior is also affected by both nozzle–sample distance and sample thickness. High-speed
images revealed vigorous motion of gases in the closed chamber, which was induced by the plasma jet. Also, the
images showed what appeared to be JA2 fragments; this observation was confirmed by recovery of fragmented
propellant after some of the tests.

Nomenclature in terms of shorter and more reproducible ignition delay,8 miti-

c = specific heat gation of gun performance sensitivity to ambient temperature,9,10
dc = capillary diameter better controlled propellant mass generation rates,11 and the ability
dN = nozzle inner diameter to ignite reliably low-vulnerability ammunition propellants12 and
E = electrical energy stored in plasma generator high loading density propellant charges.13,14 However, optimization
E vap = energy for vaporization at standard state of the ETC igniter design relies on a clear and complete understand-
h vap = heat of vaporization at normal boiling point ing of the underlying fundamentals that account for these observed
I1 = electrical current flowing through capillary plasma benefits.
I2 = electrical current flowing parallel to I1 The ignition source of ETC guns is in the form of an electrother-
LN = length of nozzle (cathode) mal plasma that is produced through a rapid capacitive discharge
L c–s = distance between capillary exit to propellant sample process, where electrical energy is converted into various other
L n–p = distance between nozzle exit to stagnation plate forms, including internal, kinetic, and chemical energies. The sub-
L n–s = distance between nozzle exit to propellant sample millisecond duration of the electrical discharge is achieved through
ms = mass of JA2 propellant sample selection of the electrical circuit components, generally consisting
p = pressure of the capacitors, the inductors, and the plasma (the resistor15 ) gen-
R = electrical resistance erated by exploding a thin metallic wire in a capillary and sustained
t = time by the ablation of surrounding capillary wall material of either the
ts = thickness of propellant sample propellant16 or a hydrocarbon polymer.17−21 The electrical connec-
Vc = volume of test chamber tion between the electrodes that is initially provided by the wire is

p = peak pressure rise caused by propellant then continued by the plasma. Using the trigger wire in the capillary
combustion only allows a highly efficient initiation and establishment of the plasma.22

τ = time duration of the second stage of burning The behavior of the electrical circuit, the trigger wire explosion phe-
nomenon, as well as the fluid mechanics of the plasma propagation
Introduction into a stagnant atmosphere are fairly well understood.23−29 However,
many other aspects are much less understood, such as the underly-
T HE electrothermal–chemical (ETC) propulsion concept has
been the subject of substantial research activities as reviewed
in Refs. 1–7. This propulsion concept, which utilizes electrically
ing causes for the measured shorter ignition delay, enhanced burn
rate during the electrical discharge, and differences in the extent of
generated plasma for ignition, has potential application in both the interaction between the plasma and various propellants.
medium- and large-caliber weapon systems. Results from labora- Numerous studies have been conducted to provide basic or phe-
tory experiments and actual firing tests have revealed that the use nomenological information about the plasma–propellant interaction
of ETC plasma ignition can offer several attractive benefits over process. Early studies revealed the ability to minimize the effect of
conventional chemical powder ignition, such as precision ignition the propellant’s initial temperature on the burn rate.30 This was ac-
complished by adjusting the plasma injection process and power
levels, to achieve breech pressures at their maximum levels for a
Received 17 October 2003; revision received 7 April 2004; accepted for broad range of initial temperatures.
publication 29 April 2004. Copyright  c 2004 by the American Institute of Several works have been performed trying to characterize the ETC
Aeronautics and Astronautics, Inc. All rights reserved. Copies of this paper plasma impact on the ignition and combustion of JA2 propellant dur-
may be made for personal or internal use, on condition that the copier pay ing and following the plasma pulse. Koleczko et al.16 have studied
the $10.00 per-copy fee to the Copyright Clearance Center, Inc., 222 Rose- burn-rate differences between graphite-containing and graphite-free
wood Drive, Danvers, MA 01923; include the code 0748-4658/05 $10.00 in
correspondence with the CCC.
JA2. The burn-rate enhancement for transparent JA2 was signifi-
∗ Research Assistant, Department of Mechanical and Nuclear Engineering. cant and lasted well beyond the duration of the plasma pulse. For
Student Member AIAA. graphite-containing JA2, burn-rate enhancement was much less, but
† Professor, Department of Mechanical and Nuclear Engineering. Senior much larger compared to the use of a conventional igniter. The re-
Member AIAA. sults were explained by 1) in-depth absorption of radiant energy that
44
 LI, LITZINGER, AND THYNELL 45

produced decomposition of energetic molecules, 2) formation of in- studies that will contribute to the formulation of a complete under-
ternal voids (porous structure) and fragmentation of propellant by standing of the various chemical and physical processes occurring
decomposition gases, and 3) increased surface area for conversion in plasma–propellant interactions.
of propellant into final products. Birk et al.17 studied JA2 response
to plasma in both interrupted and noninterrupted test configurations Experimental Apparatus and Approach
and found burn rate of JA2 was enhanced by 40–120% during the
plasma pulse but not after it. Pesce-Rodriguez et al.31 and Beyer Plasma Generator
and Pesce-Rodriguez32 used desorption-gas chromatography–mass The capillary plasma-generation system is schematically shown
spectroscopy on recovered JA2 samples from interrupted tests. They in Fig. 1. The pulse-forming network (PFN) is based on a resistance–
concluded that subsurface reactions occurred in plasma-exposed inductance–capacitor circuit, which is mainly composed of an en-
samples based on the evidence of in-depth (0.75 mm) denitration ergy storage component that consists of two high-voltage fast-
of JA2 possibly due to radiation absorption, although in-depth pro- discharge capacitors connected in parallel to yield a total capacitance
file was not consistent with radiation absorption by Beer’s law. No of 192 µF, pulse-shaping components including a 20-µH inductor
conclusions were made whether denitration was caused by thermal and a crowbar diode, and a floating high-voltage mercury switch
or photochemical effects. In addition, microscopic investigations (ignitron) as the trigger unit. The capacitors can be charged up to
revealed the formation of pits and wormholes that produced an in- 10 kV, yielding maximum energy storage of 9.6 kJ. The plasma
creased surface area and possible burn-rate augmentation. Katulka chamber consists of a capillary liner, a fine metallic wire, elec-
et al.33 used a configuration where the propellant samples were trodes, and other conducting or nonconducting housing hardware.
either fully exposed to plasma normal impingement, or covered The capillary liner is typically made of either high-density polyethy-
Downloaded by UNIVERSITY OF QUEENSLAND on June 16, 2014 | http://arc.aiaa.org | DOI: 10.2514/1.5866

with Mylar® or aluminum films to separate conductive and convec- lene (PE, [C2 H4 ]n ) or polycarbonate (Lexan, [C16 H14 O3 ]n ), which
tion effects from plasma radiation effect. The Mylar film is opaque is machined to have a bore length of 26 mm and a typical diameter
for wavelengths below 300 nm. In general, Mylar-covered samples of 3.2 mm. Located at either end of the capillary are two electrodes
showed little or no evidence to the effects of radiation, because ig- made of erosion-resistant material, elkonite, a copper–tungsten al-
nition did not occur. Hence, the UV radiation component appears loy (30% Cu, 70% W). Inside the capillary is a fine copper filament
to be important. Bourham et al.34 studied the effects of plasma im- that is connected between the electrodes and serves as the discharge
pinging angle on the ignition and combustion characteristics of JA2. initiator. After being formed immediately on triggering of the igni-
Results showed evidence of an enhanced burn rate at pressures of tron, the plasma flows through a nozzle that has typical dimensions
55–90 MPa over 400-µs pulse length, and 20–40% enhancement of 3.2 mm (inner diameter) and 26 mm (length) into open air or a
in burn rate was achieved for parallel injection, whereas a factor closed chamber.
of three increases in burn rate was observed with normal injection.
Taylor35 reported that results from small-scale closed vessel tests Test Chambers
and large-caliber 155 mm tests suggested radiative energy transfer Two closed chambers with different configurations were used in
from plasma to propellant is negligible for ignition, but metallic va- the experiments. The first one, as shown in Fig. 1, has dimensions
por deposition is important to efficient ignition and with a shorter of 3.5 cm in length and 2.1 cm in diameter, corresponding to a net
ignition delay. Also, in Taylor’s work,36 it was found that wire mass, volume of 15.8 cm3 . An insert was used to allow variation in the
capillary dimensions, and propellant composition and initial temper- chamber volume, and for tests conducted with this chamber, the
ature made no measurable difference in the ignition delay. However, actual volume was 6.5 cm3 .
the amount of electrical discharge energy had a distinctive influence The second one is an optically accessible windowed chamber
on ignition delay. (shown in Fig. 2), which was fabricated to allow photography of pro-
These research efforts have provided many insights to the vari- pellant ignition and combustion processes. The windowed chamber
ous mechanisms and, hence, a greatly improved understanding of measures 3.5 cm long with a 4.3 cm diameter, yielding a net vol-
plasma ignition mechanisms. However, a complete understanding is ume about 48 cm3 . The chamber is equipped with a 2.0-cm circular
critical to the development of the ETC technology; therefore, con- viewing port, and the optical window can withstand pressures up to
tinuing research is needed. This paper reports results from recent 20 MPa.
experimental work on characterizing the ignition and combustion of In general, a sample holder, with one end screwed into the end
a double-base propellant, JA2, driven by the ETC capillary plasma. plate of the closed chamber, brings the propellant sample close to
An effort of visualizing the plasma–propellant interaction during the nozzle exit to enable normal exposure to the plasma jet. For tests
ignition and combustion was made, which is believed to be very with no propellant, an inert substitute having the same size with the
helpful to aid in the interpretation of many other experimental data. propellant sample was used to maintain the same chamber volume.
The overall objective of this work is to gain insights from parametric Two pressure transducers (P.T.) were mounted in different locations

Fig. 1 Schematic of test setup for plasma ignition.

 46 LI, LITZINGER, AND THYNELL

Table 1 Effect of plasma energy on discharge and burning behavior of JA2 with ms = 500 mg
and Vc = 6.5 cm3

E, I1 peak , τpulse-length , pplasma-peak , ppeak-in-pulse ,
p,a τsecond-stage , τend-combustion ,
τ,b

kJ kA µs MPa MPa MPa ms ms ms
0.38 3.15 179 4.8 5.4 0.6 143 191 48
0.60 4.76 180 7.1 8.5 1.4 6.5 49 43
0.86 6.32 191 11.0 12.4 1.4 2.2 33 31
1.56 9.38 262 16.1 23.0 6.9 0.39 7.9 7.5
2.40 12.70 310 21.3 42.1 20.8 0.31 1.38 1.1
a

p = ppeak-in-pulse − pplasma-peak . b

τ = τend-combustion − τsecond-stage .

Results and Discussion

Results from pressure measurements under various test conditions
will be presented first, followed by discussion of the photographic
results of the plasma jets and the burning propellant. The propellant
used in all tests was a double-base gun propellant JA2. The test
samples were cut from JA2 cylindrical sticks, which have a diameter
Downloaded by UNIVERSITY OF QUEENSLAND on June 16, 2014 | http://arc.aiaa.org | DOI: 10.2514/1.5866

of 10.9 mm and seven 0.5-mm perforations.

Pressure Measurements
Two sets of experiments were performed to investigate the effects
of plasma parameters and test conditions on ignition and combustion
characteristics of JA2. The first set of tests was conducted in the
smaller chamber (6.5 cm3 ) to achieve a higher propellant loading
density. During the tests, the input electrical energy was varied from
0.38 to 2.40 kJ. The second set of tests was conducted to investigate
the effects of nozzle length and inner diameter, the distance from
nozzle exit to propellant sample (hereinafter referred to as nozzle–
Fig. 2 Sectional view of the windowed chamber.
sample distance) L n–s , and propellant sample thickness ts . These
experiments were performed in the 48-cm3 windowed chamber to
allow imaging as well as pressure measurements to be carried out.
to allow simultaneous measurement of pressure time history in the Both PE and Lexan capillaries of 3.2 mm diameter were used in these
chamber. A venting hole was connected to a solenoid valve for tests; however, the type of capillary material did not result in any
depressurization of the chamber as needed. observable differences in ignition or combustion of the propellant.

Diagnostics Plasma Energy Effect

Diagnostics included measurements of electrical currents and It is expected that more electrical energy input to the plasma will
chamber pressures resulting from plasma jet expansion and propel- ensure more effective ignition of the propellant charge. In actual
lant combustion and capture of images of evolving plasma jets and ETC weapon system applications, however, minimizing the require-
burning propellant. Pearson coils were used to measure the currents ment of input electrical energy for ignition is important for overall
including I1 , which flows through plasma inside the capillary, and system design. Therefore, tests were conducted to determine the
I2 , which flows parallel with I1 but though a non-inductive resistor minimum electrical energy needed to ignite the propellant charge
bank (denoted as R2 in Fig. 1). The instantaneous voltage across the effectively and reliably. Typical results of tests with differing in-
plasma can then be deduced from I2 and the resistance of R2 , which put electrical energies are summarized in Table 1 and plotted in
is connected in parallel with the flow of I1 . Figs. 3–6. Each test used the same amount of JA2, approximately
Two types of Kistler pressure transducers were used, with Model 500 mg, to give a loading density about 0.08 g/cm3 , and the nozzle–
211B3 having a measuring range of 0–500 psig (0–3.45 MPa) and a sample distance was held at 5 mm for all tests.
rise time of 2 µs for lower pressure measurement and Model 211B1 In Table 1, E refers to electrical energy stored in the pulsed power
of a much higher measuring range of 0–10,000 psig (0–68.95 MPa) source; typically about 95% of the stored energy is deposited into the
and a rise time of 1 µs for higher chamber pressure measurement. generated plasma through ohmic heating. The duration of discharge
To minimize thermal and electromagnetic effects of plasma on the pulse, τpulse-length , is measured from time zero to the time when the
pressure transducers in closed-chamber testing, recess mounting discharge current has decreased to a value on the order of 10−3 kA.
was adopted to allow the use of a thin layer (∼1 mm) of vacuum Here pplasma-peak is the peak pressure caused by the plasma jet alone,
grease between the sensor diaphragm and plasma. This treatment with no propellant present; it was obtained by firing the plasma into
proved to be effective in protecting the pressure transducers from the chamber with an inert substitute in place of the propellant sample.
overheating and from direct exposure to the charged particles in the The pressure rise,
p, is the difference between pplasma-peak and
plasma. A check with compressed air showed no evidence that this ppeak-in-pulse ; the latter is the in-pulse peak pressure resulting from a
grease layer affected the rise time of the pressure transducer. combined effect of plasma jet expansion and propellant combustion
Images were taken using an intensified, multiple charge-coupled during the plasma pulse. The start of the second stage, τsecond-stage ,
device (CCD) imaging system, Cordin 222-B, which consists of is defined as the time when pressure starts to rise for a second time
8 CCDs and is capable of acquiring 16 images, each with a pixel after the pulse. The time for completion of burning, τend-combustion , is
resolution of 1300 by 1030, a 10-bit dynamic range, and exposure taken as when pressure trace reaches a second peak.
gates down to 10 ns. The first eight images can be acquired at a Figure 3 presents the chamber pressure as a function of time on
framing speed of up to 108 frames/s, followed by a delay of 1 µs, different timescales, 0–5 and 0–300 ms; the shorter timescale shows
and then a second set of eight images at the same framing rate is details of the plasma-driven burning. Figure 4 presents a comparison
acquired. of discharge power for five plasma energy levels, which illustrates
A Nicolet 120 MultiPro high-speed data acquisition system along variation in peak power and pulse length. Figure 5 shows variations
with its ProView software was used for pressure and current mea- in pressure rise due to hot gaseous products from ablation and/or
surements. burning of JA2 samples during the pulse.
 LI, LITZINGER, AND THYNELL 47

a)
Fig. 5 Plasma energy effect on peak pressures and in-pulse pressure
Downloaded by UNIVERSITY OF QUEENSLAND on June 16, 2014 | http://arc.aiaa.org | DOI: 10.2514/1.5866

rise.

b)
Fig. 3 Effect of plasma energy on ignition and combustion of JA2:
Vc = 6.5 cm3 , ms = 500 mg, and Ln–s = 5 mm.
Fig. 6 Plasma energy effect on the second stage of burning.

plasma energy is used, the pressure rise due to burning during the
plasma pulse decreases (Fig. 5), and the delay time between the two
stages lengthens. Figure 6 presents a summary of the timescales of
the second-stage burning as a function of input energy. The delay
time and length of the second stage vary strongly with input energy.
These results indicate that the minimum electrical energy re-
quired to achieve sustained burning without significant two-stage
burning behavior is approximately 1.56 kJ for a loading density of
0.08 kg/cm3 under the current test conditions. With this input en-
ergy, however, other changes in test conditions can still bring about
significant changes in JA2 burning behavior, as will be discussed in
the following sections.

Nozzle Geometry Effect

As the plasma flows through the nozzle (cathode), it experiences
viscous effects and heat transfer. It is, therefore, likely that noz-
Fig. 4 Discharge power at varied electrical energies. zle dimensions would affect the plasma flow and the actual energy
available to ignite the propellant. Therefore, tests were conducted
to determine the effects of the nozzle length and inner diameter on
As expected based on work of past researchers,30 input energy the plasma and its interaction with JA2.
has a significant influence on the ignition and entire burning of the Although the ETC plasma is sustained by ablation of capillary
propellant. With the energy at 1.56 kJ, the pressure trace in Fig. 3a wall material, for a plasma generator having a configuration shown
clearly indicates a two-stage burning behavior: a fast combustion in Fig. 1, the total mass in the plasma jet comprises contributions
within the plasma pulse that appears to extinguish before the pulse from several sources, including the end electrode (anode), the trigger
ends at about 0.26 ms, which is followed by a second ignition after wire, the nozzle (cathode), the capillary, and the air that is initially
which the propellant sample burns to completion over a relatively in the capillary. Generally, because the nozzle undergoes intense
long period of time (7.5 ms). When plasma energy is increased radiative and convective heating by the hot plasma flowing through
(2.40 kJ data), the two-stage behavior becomes much less distinct; it at high velocities, ablation of the nozzle surface material provides
burning is sustained from plasma-driven ignition to completion and a major fraction to the total plasma mass. As a result, the mass
occurs over a much shorter period of time about 1.1 ms. When less of metallic particles dominates the total mass of the plasma jet.
 48 LI, LITZINGER, AND THYNELL

Table 2 Effect of nozzle dimensions on plasma mass: E = 1.56 kJ, dN = 3.2 mm, Lexan

Plasma mass sources

Test Nozzle i.d., mm Nozzle length, mm Capillary, mg Nozzle, mg End (anode), mg Wire, mg Air,a mg Total mass, mg
0291 4.7 13 5.1 1.6 1.4 2.7 0.2 11.0
0293 3.2 13 6.5 5.1 1.6 2.7 0.2 16.1
0295 3.2 28 6.1 14.8 1.6 2.7 0.2 25.4
a
Air initially trapped in the capillary.

Table 3 Properties of electrode materiala

Material g/mol c kJ/kg/K MP, K BP, K h vap , kJ/mol E vap , kJ/g

Cu 63.546 0.38 1356.6 2843 300.3 5.69
W 183.85 0.13 3683.2 5773 824.0 5.19
a
Copper and tungsten.

Table 4 Plasma energy loss due to nozzle ablation, E = 1.56 kJ

Downloaded by UNIVERSITY OF QUEENSLAND on June 16, 2014 | http://arc.aiaa.org | DOI: 10.2514/1.5866

Nozzle Nozzle Ablation of Energy for Fraction

Test i.d., mm length, mm nozzle, mg ablation, kJ of E, %
0291 4.7 13 1.6 0.009 0.58
0293 3.2 13 5.1 0.027 1.84
0295 3.2 28 14.8 0.079 5.33

Fig. 8 Nozzle length effect on JA2 burning behavior: Vc = 48 cm3 ,

E = 1.56 kJ, ms = 255 mg, and dN = 3.2 mm.

Fig. 7 Effect of nozzle dimensions on plasma chamber pressure:

Vc = 48 cm3 and E = 1.56 kJ.

However, in terms of mole fraction, it is the species ablated from

the hydrocarbon capillary that have dominance (about 90%) over
all other species.37
Table 2 lists typical results that compare the influence of nozzle a)
length and inner diameter on the contributions of each source to
the total mass. Distinctive differences in nozzle ablation mass were
found for the three nozzles. The energy loss due to ablation of the
Cu/W alloy used in the nozzle was estimated under the assumption
that ablation was caused by vaporization only, which is reasonable
given the high plasma temperature [>10,000 K (Ref. 19)] well above
the boiling points of copper and tungsten. The data used in this
calculation, including molecular weight (MW), melting point (MP)
and boiling point (BP), are presented in Table 3.
Table 4 summarizes the energy loss associated with ablation and
shows that it accounts for as much as 5.33% for the longer nozzle,
but only 0.58% for the shorter nozzle of a 3.2 mm i.d. This difference
is large enough that it should lead to differences in plasma behav-
ior. Figure 7 presents the results of chamber pressure when plasma
entered the closed chamber through these nozzles. As expected, the
discrepancies in pressure traces are correlated reasonably well to
those found in ablation energy losses.
Not surprisingly, the use of these nozzles would produce differ-
ence in the ignition of JA2 by the plasma. Representative results b)
of using these nozzles are presented in Figs. 8 and 9. For illustra- Fig. 9 Nozzle i.d. effect on JA2 burning behavior and discharge cur-
tion purposes, also included in Fig. 8 are the variations in nozzle rent: E = 1.56 kJ, ms = 255 mg, LN = 13.5 mm, and Ln–s = 5 mm.
 LI, LITZINGER, AND THYNELL 49

configuration corresponding to the plots labeled as a, b, and c. Three

comparisons can be made based on the results in Fig. 8: a vs b, a
vs c, and b vs c. Comparison of a and b illustrates the effect of
the nozzle–sample distance, which will be discussed in detail later;
thus, discussion here will be focused on the last two cases. As indi-
cated, the nozzle–sample distances in a and c were kept at 5 mm, but
test c used a longer nozzle (28 mm), resulting in a longer ignition
delay and a slower burn rate and, hence, a lower peak pressure due
to greater energy loss to the chamber walls during combustion. It
appears that the longer nozzle caused more energy loss from the
plasma, resulting in longer ignition delay times. Another possible
reason is that although the nozzle–sample distances were the same,
the distances from the capillary exit to the sample, L c–s , were dif-
ferent, so that, with the longer nozzle, the plasma has to travel a
longer path to reach the propellant. In addition to enabling more
heat transfer from the plasma to the nozzle wall, the longer noz-
zle also enables more momentum loss due to the viscous boundary
layer that develops along the nozzle inner surface, which then re- Fig. 10 Influence of nozzle–sample distance on burning behavior of
duces the mass flow rate of the plasma at the nozzle exit, resulting JA2: ms = 260 mg, Vc = 48 cm3 , and E = 1.56 kJ.
Downloaded by UNIVERSITY OF QUEENSLAND on June 16, 2014 | http://arc.aiaa.org | DOI: 10.2514/1.5866

in a reduction in the heating rate of the propellant by the plasma.

This boundary-layer effect may be further evidenced to some extent
by comparing test c with b, where the propellant sample was kept at
the same location relative to the capillary exit, but a shorter nozzle
was used to reduce both nozzle ablation effect and boundary-layer This effect is essentially attributed to the dependence of heat trans-
effect. The resultant difference in pressure traces can be seen but fer rates on the nozzle–sample distance, which affects the heating
is much less pronounced compared to that of a vs c, which may of the propellant in several ways. First, as L n–s increases, the plasma
suggest that L c–s is more critical than L n–s to propellant ignition. mass flux that comes to a contact with the propellant becomes
However, it seems that the differing behaviors in JA2 ignition smaller, resulting in a lowered heat transfer rate from the plasma
and burning cannot be explained only by those effects already men- to the propellant. Second, higher values of L n–s cause more reduc-
tioned, as shown in Fig. 9a, which presents typical results from tests tion in the recovery of temperature due to radiation loss and possibly
that used nozzles of the same length (13 mm) but different i.d. As also because of the formation of stagnation bubbles in the region
shown in Tables 2 and 4, the ablation mass and ablation-associated between the propellant surface and the Mach disk in the plasma jet.
energy loss of the 3.2-mm i.d. nozzle are three times that of the The stagnation bubble is formed as a result of interactions of plasma
4.7-mm i.d. nozzle, and based on the foregoing analysis, longer ig- jet shocks with the plate shock that is generated when a freejet is
nition delay and lower burn rate should be seen with the smaller impinging on a plate.40,41 Consequently, the convective heat trans-
i.d. nozzle, but the result shows the opposite trend. Accordingly, fer rates will slow down with longer L n–s . Third, from a radiation
some other factors related to nozzle i.d. must exist that exert a major heat transfer point of view, because the solid angle projected on the
influence on the ignition process. sample surface from any point in the jet is inversely proportional
In the formation of the capillary plasma, the electrical energy is to the square of the distance from this source point to the sample,
converted into thermal energy of the plasma through ohmic heating, radiative heating of the propellant by the plasma will be attenuated
and thus, the temperature and pressure of the plasma are closely with longer L n–s . Fourth, as will be seen in the photographic study,
related to the discharge current. Figure 9b shows a comparison of at L n–s = 5 mm, the mushroom-shaped bright core, representing the
the discharge current when using the two different nozzles. It is hottest part of the plasma jet, maintains contact with the propellant
likely that the differences in the peak current and the discharge surface during most of the plasma event. Within this hottest mush-
pulse length are responsible for the resultant behaviors in the ignition room region, although the local density is much higher than that
and combustion of the propellant. This current effect has also been at L n–s = 15 mm, recombination reactions may not prevail yet be-
evidenced in the work of Kappen and Beyer,38 who examined the cause the temperature is still extremely high, and therefore, plasma
response of a transparent JA2 to plasma radiation and found that species are present mainly as small ions and radicals, which would
with almost the same energy a higher peak current caused more actively participate in the chemical reactions that lead to ignition of
structural damage of the propellant. the propellant.

Effect of Nozzle–Sample Distance Effect of Sample Thickness

Another factor that has an influence on the ignition and combus- Heat flux measurement by Williams and White42 found an ef-
tion of the JA2 propellant is the nozzle–sample distance. It has been fect of sample thickness on the bulk temperature rise of JA2 after
found in a previous work that a shorter L n–s enables more ablation exposure to the plasma, which showed that the temperature rise
of the propellant.39 In the present work, experiments with different of 1-mm-thick JA2 was twice that of a 2-mm-thick JA2 sample.
values of L n–s were carried out to study how this parameter would Hence, it is expected that thinner JA2 samples should ignite eas-
affect the burning behavior of the propellant. A typical result is in ier with shorter ignition delay than thicker ones. Experiments were
Fig. 10. For each test, a JA2 sample of 260 mg (2.3 mm thick) was performed to explore this effect. The results are shown in Fig. 11,
used, and the electrical energy was kept at 1.56 kJ. Nozzles having where three sample thicknesses were considered including 1.5, 2.8,
the same dimensions (d N = 3.2 mm and L N = 26 mm) were used and 3.2 mm, which correspond to weights of 163, 317, and 360 mg.
for these tests, and three nozzle–sample distances were considered, As shown, a clear correlation can be seen between the sample thick-
including 3, 8, and 15 mm. Differences can be seen in both early ness and the ignition delay. The pressure curves differ from those
response of the propellant and in its overall burning characteristics. presented in Fig. 3 due to the use of the large combustion cham-
It appears that the response of the JA2 sample to plasma during the ber (48 cm3 ), which lowered the chamber pressures and thus also
pulse is negligible for the test of L n–s at 15 mm because its pressure the burning rates. Bulk temperature rise may explain the sample
trace is very similar to that of the test with no propellant. For the thickness effect; however, propellant fragmentation leading to in-
two shorter nozzle–sample distances, plasma–propellant interaction creased burning surface area may also be playing a role. As will
was evident within the pulse, an indicator of early conversion of the be discussed with the following photographic results, the thinner
propellant sample from solid to gaseous products due to ablation JA2 samples would break into smaller fragments under the severe
and decomposition. thermal and gas dynamic conditions induced by the plasma jet.
 50 LI, LITZINGER, AND THYNELL

Photography of Evolving Plasma Jets and Burning JA2 interaction. Although the 4.5-mm-thick JA2 sample survived the
In an earlier work,39 an attempt was made to photograph the normal impingement of the plasma jet, plasma interaction caused a
response of a JA2 sample to the plasma in an open-air configuration, 25-mg mass loss of the propellant. The images (Fig. 12) did not show
and a typical result is shown in Fig. 12. In this test, the sample holder observable evidence of ignition of this part of propellant, which is
brought the JA2 sample close to the nozzle exit with L n–s at 5 mm. likely attributed to a luminosity overlapping effect.
The core of the plasma maintained contact with the sample surface The present photographic study continued the previous work but
throughout the plasma discharge, but no self-sustained burning was was conducted in a closed chamber in an attempt to achieve ignition.
achieved due to the lack of confinement of the plasma–propellant Initial tests were conducted in hope to capture plasma ignition of
JA2 samples within the discharge pulse by taking advantage of the
high temporal resolution of the CCD imaging array. However, it
was found that the luminosity of burning JA2 was much less than
that of the plasma, and so no images were obtained. In principle,
appropriate filters can optically attenuate the plasma brightness but
may not ensure a solution to solve this problem. Alternatively, the
brightness of plasma can be separated from that of burning JA2 by
prolonging the delay between the first stage and second stage of
burning through the use of lower plasma energy, larger L n–s values,
or larger chamber volume. In addition, to maximize the viewing
area of the optical window, and to have a relatively low operating
Downloaded by UNIVERSITY OF QUEENSLAND on June 16, 2014 | http://arc.aiaa.org | DOI: 10.2514/1.5866

pressure to protect the window, a larger chamber volume is more

desirable. Therefore, the windowed chamber was designed to have
a relative large volume (48 cm3 ) for the photographic work.
It has been found that, when using a PE capillary, a considerable
amount of soot was formed when the plasma cools after entering
the chamber. Compared to PE, Lexan produces less soot due to the
presence of oxygen in its composition. Therefore, Lexan capillaries
were used to minimize soot contamination of the chamber window.
Fig. 11 Effect of sample thickness on burning behavior of JA2: Figure 13 presents a comparison of evolving plasma images in
Vc = 48 cm3 , E = 1.56 kJ, and Ln–s = 5 mm. the closed chamber to that in open air, with test setups and image

Fig. 12 Typical 1.56-kJ plasma jet impinging on JA2 propellant sample in open air at nozzle–sample distance of 5 mm.

Fig. 13 Evolving plasma jets (E = 1.56 kJ) in open-air (top) and closed-chamber (bottom) and discharge current and plasma pressure in the chamber.
 LI, LITZINGER, AND THYNELL 51
Downloaded by UNIVERSITY OF QUEENSLAND on June 16, 2014 | http://arc.aiaa.org | DOI: 10.2514/1.5866

Fig. 14 Plasma-ignition burning JA2 in a closed chamber, Vc = 48 cm3 , E = 1.56 kJ, and Ln–s = 5 mm (top); plasma-induced hot gases under the same
test conditions (bottom); and corresponding pressure traces from tests with and without JA2 (center).

Fig. 15 Plasma-ignition burning JA2 (1.8 mm thick, 200 mg) in closed chamber showing a fragmented piece: Vc = 48 cm3 , E = 1.56 kJ, and Ln–s = 5 mm.

sizes shown in the last image of each row. A discharge current trace test with an inert substitute sample. The top row in Fig. 14 shows
representative for both closed-chamber and open-air configurations that ignition occurred no later than 10 ms after firing the plasma,
is presented to illustrate the pulse duration and the correlation of the and evidence of ignition was seen at multiple spots, an indication
current with the plasma luminosity. The chamber pressure resulting of fragmentation of the propellant sample, which is possible given
from the evolving plasma is also included in Fig. 13 to illustrate the the small thickness (2.3 mm) of JA2 sample and the short nozzle–
pressure corresponding to each image in the closed chamber. As one sample distance (3 mm). The test with an inert sample shown in
can see, both test configurations generated plasma jets that exhibit the bottom row of Fig. 14 indicates that the brightness in images at
the features of underexpanded supersonic jets. However, differences 5 ms mostly comes from plasma-generated hot gases. Furthermore
are also evident. In open air, because the ambient pressure is constant Fig. 14 shows gas motion that accompanies the burning process, re-
at 1 atm, the jet keeps growing after emerging from nozzle exit sulting from expansion into the chamber. As a comparison, Fig. 15
until it reaches a fully developed state around t = 120 µs, after presents a test that employed an even thinner (1.8-mm) JA2 sample
which it varies little in its structure and size. However, in the closed to increase the likelihood of fragmentation. As can be clearly seen,
chamber, because the ambient conditions (pressure and temperature) from t = 10 ms, a fragment was captured, as was its movement with
are changing with time over the event, the jet size undergoes a the gas.
growing (in length and breadth) and then shrinking (in breadth)
process, and the shock system is also varying, that is, the barrel Conclusions
shock is shrinking and Mach disk is pushed back toward nozzle Experiments have been conducted in closed chambers to study
exit due to the increasing pressure in the chamber. This dynamic the behavior of JA2 propellant during exposure to an ETC plasma.
feature suggests that the increase of chamber pressure would lead The results from parametric studies indicated that plasma energy has
to a decrease in jet penetration through the charge bed of an ETC a significant influence on ignition and combustion of the propellant.
gun. In early times, however, for example, when t = 60 µs, because At low plasma energies, rapid, plasma-driven burning occurred, but
chamber pressure is still low, the jet grows very fast. In general, the self-sustained burning was not achieved. With moderate plasma en-
jet has a much smaller size in a chamber than in open air under the ergies, combustion of the propellant exhibited a two-stage burning
same test conditions. Accordingly, the overall density in the plasma behavior: one stage of plasma-driven rapid burning that occurred
within the jet should be greater in the chamber than in open air and during the plasma pulse and a second stage of slower self-sustained
a more aggressive plasma–propellant interaction can be expected. burning, which occurred with a clear delay after the first stage.
Figure 14 shows burning JA2 and plasma-induced high- When plasma energy was increased further, the two-stage behavior
temperature gases in the closed chamber; also included are the corre- became less distinct and eventually disappeared, leaving only one
sponding burning pressure trace and plasma pressure from a baseline stage of burning. Nozzle length and diameter affected the ignition
 52 LI, LITZINGER, AND THYNELL

and combustion characteristics as a result of energy losses from the 14 Perelmutter, L., Sudai, M., Goldenberg, C., Kimhe, D., Zeevi, Z., Arie,

plasma as it flows through the nozzle. The propellant burning be- S., Melnik, M., and Melnik, D., “Plasma Propagation and Ignition in the
havior is also affected by both nozzle–sample distance and sample Chamber of a SPETC Gun,” IEEE Transactions on Magnetics, Vol. 35,
thickness. No. 1, 1999, pp. 213–217.
15 Del Guercio, M., Stobie, I., Katulka, G., and Oberle, W., “A Pulse
The plasma luminosity is considerably brighter than that of burn-
Forming Network Design for Electrothermal–Chemical Combustion Char-
ing JA2, so that it was not possible to capture the ignition and plasma-
acterization of Solid Propellants,” Tenth IEEE International Pulsed Power
driven burning of the propellant during the pulse. Hence, the ignition Conference, Digest of Technical Papers, Vol. 1, Inst. of Electrical and Elec-
delay was sufficiently prolonged through using a large test cham- tronics Engineers, New York, 1995, pp. 279–285.
ber to separate the ignition and burning of the propellant from the 16 Koleczko, A., Ehrhardt, W., Kelzenberg, S., and Eisenreich, N., “Plasma
plasma pulse. Consequently, only burning images (Figs. 14 and 15) Ignition and Combustion,” Propellants, Explosives, Pyrotechnics, Vol. 26,
in the second stage were captured. The images revealed vigorous 2001, pp. 75–83.
17 Birk, A., Del Guercio, M., Kinkennon, A., Kooker, D. E., and Kaste,
motion of hot gases in the closed chamber, which was induced by
the plasma jet and are expected to enhance heat transfer and, thus, P., “Interrupted-Burning Tests of Plasma-Ignited JA2 and M30 Grains in
to enhance the burn rate. Under certain test conditions, fragments a Closed Chamber,” Propellants, Explosives, Pyrotechnics, Vol. 25, 2000,
of the propellant were clearly visible in the photographs, suggest- pp. 133–142.
18 Beyer, R. A., “Small Scale Experiments in Plasma Propellant Interac-
ing that fragmentation occurs before and/or during the ignition and
tions,” Proceedings of 37th JANNAF Combustion Subcommittee Meeting,
combustion processes. CPIA Publ. 701, Vol. 1, 2000, pp. 137–144.
Results from this experimental work should aid in the formation 19 Kohel, J. M., Su, L. K., Clemens, N. T., and Varghese, P. L., “Emission
of a more complete understanding of plasma–propellant interaction Spectroscopic Measurements and Analysis of a Pulsed Plasma Jet,” IEEE
Downloaded by UNIVERSITY OF QUEENSLAND on June 16, 2014 | http://arc.aiaa.org | DOI: 10.2514/1.5866

and provide information needed to optimize the development of the Transactions on Magnetics, Vol. 35, No. 1, 1999, pp. 201–206.
plasma igniter for ETC propulsion applications. 20 Li, J. Q., Zhou, H., Kudva, G., Thynell, S., and Litzinger, T., “Ex-
perimental Investigation of Plasma Propellant Interactions,” Proceedings
Acknowledgments of 37th JANNAF Combustion Subcommittee Meeting, CPIA Publ. 701,
Vol. 1, Chemical Propulsion Information Agency, Laurel, MD, 2000,
This work was supported by the U.S. Army Research Office un- pp. 109–121.
der Contract DAAG55-98-1-0519 with the management of David 21 Edwards, C. M., Bourham, M. A., and Gilligan, J. G., “Experimen-
Mann. The authors are also grateful to Johnny Yu at ATK (Alliant tal Studies of the Plasma-Propellant Interface for Electrothermal-Chemical
Techsystems)–Radford Army Ammunition Plant for the shipment Launchers,” IEEE Transactions on Magnetics, Vol. 35, No. 1, 1999,
of the JA2 propellant. pp. 404–409.
22 Taylor, M. J., “Formation of Plasma Around Wire Fragments Created by
Electrically Exploded Copper Wire,” Journal of Physics D: Applied Physics,
References Vol. 35, No. 7, 2002, pp. 700–709.
1 Fair,
H. D., “Electric Launch Science and Technology in the 23 Katulka, G. L., “Parametric Study of High Energy Plasmas for
United States,” IEEE Transactions on Magnetics, Vol. 39, No. 1, 2003, Electrothermal–Chemical Propulsion Applications,” IEEE Transactions on
pp. 11–17. Plasma Science, Vol. 25, No. 1, 1997, pp. 66–72.
2 Haugh, D. C., and Gilbert, S., “U.K. Electric Gun National Overview,” 24 Nusca, M. J., McQuaid, M. J., and Anderson, W. R., “Numerical Model
IEEE Transactions on Magnetics, Vol. 39, No. 1, 2003, pp. 18–21. of the Plasma Jet Generated by an Electrothermal–Chemical Igniter,” Journal
3 Jung, J.-W., Kim, S.-H., and Yang, K.-S., “Overview of ETC Re- of Thermophysics and Heat Transfer, Vol. 16, No. 1, 2002, pp. 157–160.
search in Korea,” IEEE Transactions on Magnetics, Vol. 39, No. 1, 2003, 25 Kim, J. U., Clemens, N. T., and Varghese, P. L., “Experimental Study
pp. 22–23. of the Transient Underexpanded Jet Generated by Electrothermal Capil-
4 Lehmann, P., “Overview of the Electric Launch Activites at the lary Plasma,” Journal of Propulsion and Power, Vol. 18, No. 6, 2002,
French–German Research Institute of Saint-Louis (ISL),” IEEE Transac- pp. 1153–1160.
tions on Magnetics, Vol. 39, No. 1, 2003, pp. 24–28. 26 Li, J. Q., Kwon, J., Thynell, S. T., and Litzinger, T. A., “Experimen-
5 Rutberg, P. G., Shvetsov, G. A., and Savvateev, A. F., “Results of Recent tal Investigations of Characteristics of Electro–Thermal–Chemical Plasma,”
Research on Electromagnetic Launch Technology in Russia,” IEEE Trans- AIAA Paper 2001-3855, July 2001.
actions on Magnetics, Vol. 39, No. 1, 2003, pp. 29–35. 27 White, K. J., Katulka, G. L., Khuan, T., and Nekula, K., “Plasma Char-
6 Weise, T. H. G. G., Maag, J., Zimmermann, G., Eisenreich, N., and acterization for Electrothermal–Chemical (ETC) Gun Applications,” U.S.
Derlich, H., “National Overview of the German ETC Program,” IEEE Trans- Army Research Lab., Rept. ARL-TR-1491, Aberdeen Proving Grounds,
actions on Magnetics, Vol. 39, No. 1, 2003, pp. 35–38. MD, 1997.
7 Wang, Y., Cheng, S., and Zheng, P., “Widely Developing Electric Launch 28 Powell, J. D., and Zielinski, A. E., “Capillary Discharge in the Elec-

Technology in China,” IEEE Transactions on Magnetics, Vol. 39, No. 1, trothermal Gun,” IEEE Transactions on Magnetics. Vol. 29, No. 1, 1993,
2003, pp. 39–41. pp. 591–596.
8 Katulka, G. L., and Dyvik, J. A., “Experimental Results of Electri- 29 Taylor, M. J., and Dunnett, J., “A Description of the Wire Explosion

cal Plasma Ignition in 120-mm Solid Propellant Tank Gun Firings,” Pro- Process for the ETC Plasma Generators,” IEEE Transactions on Magnetics,
ceedings of the 33rd JANNAF Combustion Meeting, CPIA Publ. 653, Vol. 39, No. 1, 2003, pp. 269–274.
Vol. 3, Chemical Propulsion Information Agency, Laurel, MD, 1996, 30 Perelmutter, L., Sudai, M., Goldenberg, C., Kimhe, D., Zeevi, Z., Arie,
pp. 103–110. S., Melnik, M., and Melnik, D., “Plasma Propagation and Ignition in the
9 Marinos, C., “ETC Ignition and Temperature Sensitivity,” Proceed- Chamber of a SPETC Gun,” IEEE Transactions on Magnetics, Vol. 35,
ings of the 32nd JANNAF Combustion Meeting, CPIA Publ. 631, No. 1, 1999, pp. 213–217.
Vol. 3, Chemical Propulsion Information Agency, Laurel, MD, 1995, 31 Pesce-Rodriguez, R. A., Beyer, R. A., Kinkennon, A. E., Del Guercio,
pp. 109–118. M., Kaste, P. J., and Newberry, J. E., “In-depth Chemistry in Plasma-Exposed
10 Dyvik, J. A., and Katulka, G. L., “ETC Temperature Compensation: M30 and JA2 Gun Propellants,” Proceedings of the 37th JANNAF Combus-
Experimental Results of 120-mm Test Firings,” Proceedings of the 33rd tion Subcommittee Meeting, CPIA Publ. 701, Vol. 1, Chemical Propulsion
JANNAF Combustion Meeting, CPIA Publ. 653, Vol. 1, Chemical Propulsion Information Agency, Columbia, MD, 2000, pp. 145–155.
Information Agency, Laurel, MD, 1996, pp. 111–119. 32 Beyer, R. A., and Pesce-Rodriguez, R. A., “Experiments to Define
11 Del Guercio, M., “Propellant Burn Rate Modification by Plasma In- Plasma–Propellant Interactions,” IEEE Transactions on Magnetics, Vol. 39,
jection,” Proceedings of the 34th JANNAF Combustion Meeting, CPIA No. 1, 2003, pp. 207–211.
Publ. 662, Vol. 1, Chemical Propulsion Information Agency, Laurel, MD, 33 Katulka, G. L., White, K. J., Oberle, W. F., Kaste, P. J., Pesce-
1997, pp. 35–42. Rodriguez, R. A., and Leadore M. L, “Experimental Characterization of
12 Chaboki, A., Zelenak, S., and Isle, B., “Recent Advances in Plasma Effects on Energetic Materials for Electrothermal–Chemical Launch
Electrothermal–Chemical Gun Propulsion at United Defense, L. P.,” IEEE Applications,” IEEE Transactions on Magnetics, Vol. 35, No. 1, 1999,
Transactions on Magnetics, Vol. 33, No. 1, 1997, pp. 284–288. pp. 197–200.
13 Wren, G. P., and Oberle, W. F., “Influence of High Loading Den- 34 Bourham, M. A., Gilligan, J. G., and Oberle, W. F., “Analysis of Solid
sity Charge Configurations on Performance of Electrothermal–Chemical Propellant Combustion Behavior Under Electrothermal Plasma Injection for
(ETC) Guns,” IEEE Transactions on Magnetics, Vol. 37, No. 1, 2001, ETC Launchers,” IEEE Transactions on Magnetics, Vol. 33, No. 1, 1997,
pp. 211–215. pp. 278–283.
 LI, LITZINGER, AND THYNELL 53

35 Taylor, M. J., “Evidence for the Hypothesis of Ignition of Propellants by 39 Li, J. Q., Litzinger, T. A., and Thynell, S. T., “Interaction of Capillary
Metallic Vapour Deposition,” Propellants, Explosives, Pyrotechnics, Vol. 27, Plasma with Double-Base and Composite Propellants,” Journal of Propul-
2002, pp. 327–335. sion and Power, Vol. 20, No. 4, 2004, pp. 675–683.
36 Taylor, M. J., “Consideration of the Energy Transfer Mechanisms 40 Kim, B. G., Yu, M. S., Cho, Y., and Cho, H. H., “Distributions of Re-
Involved in SPETC Ignition Systems,”IEEE Transactions on Magnetics, covery Temperature on Flat Plate by Underexpanded Supersonic Impinging
Vol. 39, No. 1, 2003, pp. 262–268. Jet,” Journal of Thermophysics and Heat Transfer, Vol. 16, No. 3, 2002,
37 Li, J. Q., Litzinger, T. A., and Thynell, S. T., “Experimental Inves- pp. 425–431.
tigation of Plasma-Ignition and Combustion of Solid Propellants,” Pro- 41 Lamont, P. J., and Hunt, B. L., “The Impingement of Underexpanded,
ceedings of the 39th JANNAF Combustion Subcommittee Meeting, CPIA Axisymmetric Jets on Perpendicular and Inclined Flat Plates,” Journal of
Publ. 780, Vol. 3, Chemical Propulsion Information Agency, Laurel, MD, Fluid Mechanics, Vol. 100, Pt. 3, 1980, pp. 471–511.
2003. 42 Williams, A. W., and White, K. J., “Plasma–Propellant Interactions
38 Kappen, K., and Beyer, R. A., “Progress in Understanding Plasma– Studies: Measurements of Heat Flux Produced by Hydrocarbon Ablation-
Propellant Interaction,” Propellants, Explosives, Pyrotechnics, Vol. 28, Supported Plasmas,” IEEE Transactions on Magnetics, Vol. 37, No. 1, 2001,
No. 1, 2003, pp. 32–36. pp. 203–206.
Downloaded by UNIVERSITY OF QUEENSLAND on June 16, 2014 | http://arc.aiaa.org | DOI: 10.2514/1.5866