NATO UNCLASSIFIED

Ballistic Properties of Scavenged Solid Rocket Propellants

Fabio Cozzi, Stefano Manenti, Alba L. Ramaswamy

Andrea Olivani, Luigi T. DeLuca ECE Department
Solid Propulsion Lab, Dipartimento di Energetica University of Maryland
Politecnico di Milano College Park
34 Via La Masa MD 20742
20158 Milan USA
ITALY

ABSTRACT
Ignition and steady burning of scavenged composite solid rocket propellants were experimentally investigated
at the Solid Propulsion Laboratory of Politecnico di Milano, using a variety of strand burners. The unusual
ignition behavior of some of these compositions, tailored for burning at low burning rates and relatively low
chamber pressures (11 to 35 atm), was found to be associated with a peculiar multiphase flame structure.
Video recording of ignition and steady burning tests of the scavenged compositions revealed the presence
of a scattered liquid film and the formation of large spherical particles at the reacting surface layer,
which hindered a smooth flame development. A break in the slope of burning rate vs. pressure plots was
observed for the scavenger compositions. The break, occurring at pressures above 50 atm for one composition
and above 25 atm for another composition, appears to be due to different combustion regimes. In the low
pressure region, the burning rate is determined by the NOx oxidation reactions; in the high pressure region,
ClOx oxidation reactions prevail.

NOMENCLATURE
Latin Symbols Abbreviations
Isp = specific impulse, s AP = Ammonium Perchlorate, NH4ClO4
Ivol = volumetric specific impulse, s g cm-3 E00 = propellant, reference composition
n = Vieille law pressure exponent E101 = propellant, scavenged composition No. 1
p = pressure, atm E125 = propellant, scavenged composition No. 2
q = radiant heat flux, W/cm2 HTPB = Hydroxyl Terminated PolyButadiene
R = correlation coefficient N = number of experimental runs
rb = burning rate, cm/s SN = Sodium nitrate, NaNO3
t = time, s

1.0 BACKGROUND
Although the environmental impact of chemical rocket propulsion is overall negligible[1], during the last few
years attention has focused on ammonium perchlorate (AP)-based propellants with the intent of improving the
performance whilst reducing the hydrogen chloride HCl fraction in the exhaust products. New strategies in the
development of “energetic” and “clean” solid compositions include a variety of options spanning from

Paper presented at the RTO AVT Specialists’ Meeting on “Advances in Rocket Performance Life and Disposal”,
held in Aalborg, Denmark, 23-26 September 2002, and published in RTO-MP-091.

RTO-MP-091 32 - 1

NATO UNCLASSIFIED
 NATO UNCLASSIFIED

Ballistic Properties of Scavenged Solid Rocket Propellants

propellants containing chlorine scavengers to chlorine neutralizers to energetic oxidizers. Chlorine scavenging
formulations, whereby AP is partially replaced by sodium nitrate NaNO3 (SN), are the closest to the relatively
well-known class of conventional AP/HTPB composite propellants commonly used by the propulsion
industries. This technique, leading to a large fraction of HCl scavenged directly in the combustion chamber,
to form the common NaCl salt, allows a good compromise between performance, cost, safety, reliability,
and enhanced environmental protection.

For many years FiatAvio has been investigating several innovative classes of propellants[2,3].
Novel formulations were developed and fully characterized to address different applications of solid rocket
motors. In particular, compositions containing different amounts of aluminum, binder, and AP/NaNO3 were
qualified at the laboratory and motor firing levels. Excellent results were achieved for propellant compositions
burning at relatively high pressures (say, above 40 atm), leading to small-scale and full-scale motors with a
scavenging effect in the range of 50 to 75% (see Table 1)[4]. Several motor configurations were tested.
Most runs were performed in a standard 2 inches small-scale test motor; MTM is a special small-scale motor
used to test propellants, thermal protections, and nozzles; E00 data are for a full-scale motor. Combustion
chamber pressures cover a range from 45 to 200 atm (from E-2”-2012 to E00) and related burning rates could
be tailored for the proper motor requirements. Burning rates higher than 0.6 cm/s were obtained with high gas
flows in side-burning motor configurations; only one end-burning propellant grain, E-MTM-2, was fired.
All of the tested motors ignited regularly and burned uniformly, featuring a pressure/time plot in agreement
with the corresponding motors, using the conventional propellants.

Table 1: Motor Configurations and Propellant Formulations

Propellant Scavenging Chamber Burning Propellant

Al Binder
Effect* Pressure rate** mass
(%) (%)
(%) (atm) (cm/s) (kg)
E-2”-2012 20 12 70 45 0.743 0.4
E-2”-2013 20 13 70 50 0.875 0.4
E93 1 14 75 100 2.13 0.4
E93 1 14 75 200 3.138 0.4
E91 1 14 55 100 2.180 0.4
E91 1 14 55 200 3.172 0.4
E-MTM-1 20 12 50 45 0.780 6.5
E-MTM-2 20 12 50 45 0.810 7.0***
E-81 1 14 75 100 2.20 ≈ 8.0
E00 5 14 - 100 2.00 ≈ 60
E101 4 14 75 15 0.530 0.4
E125 4 14 75 15 0.460 0.4

* Computed reduction of HCl, at nozzle exit with p = 1 atm, wrt the corresponding
conventional propellant.
** Burning rate suitably tuned within the required range at the operating pressure.
*** Motor configuration is end-burning, while all other configurations are side-burning.

32 - 2 RTO-MP-091

NATO UNCLASSIFIED
 NATO UNCLASSIFIED

Ballistic Properties of Scavenged Solid Rocket Propellants

Due to the higher density of SN as compared to AP and depending upon formulation, a small increase
(1-2%) in the volumetric specific impulse can be achieved by using scavenged propellant, even if the
delivered specific impulse decreases a little bit (see Table 2)[3,4,5]. Moreover the costs (raw material and
manufacture) of scavenged compositions is essentially the same of conventional composite propellants[5].
Thus, the scavenging technique produces a good compromise between performance, cost, safety, reliability,
and enhanced environmental protection.

Table 2: Impulse of Scavenger Propellants (Taken from Ref. 3)

Al/Binder Isp Ivol HCl

Formulation AP/NaNO3
(HTPB) (s) (sgcm-3) (%)
Ariane V 18/14 68/- 265.3 464.7 21
1814scav1 18/14 32/29 244.4 443.7 1.4
1814scav07 18/14 50/18 252.3 451.8 8.4
1814scav05 18/14 54/14 255.0 454.3 10.5
2013scav1 20/13 36/31 244.8 457.1 1.9
2013scav07 20/13 49/18 249.8 462.2 5.4
2013scav05 20/13 53/14 253.7 465.9 9.2
2012scav1 20/12 32/29 244.2 461.0 2.0
2012scav07 20/12 50/18 249.7 467.1 5.4
2012scav05 20/12 54/14 254.3 471.6 9.7

Reduced aluminum content compositions tailored for burning at low burning rates and chamber pressures,
say in the range of 11 – 35 atm, showed however some anomalies in terms of ignition and burning behavior
under small-scale motor testing. All propellant formulations are multimodal in AP and monomodal in the
scavenging agent. Two of these specific compositions, respectively called E101 and E125, are reported as the
last two entries in Table 1. Tests at the motor level conducted by FiatAvio showed unexpected pressure/time
combustion profiles. A detailed testing of transient and steady burning of E101 and E125 was subsequently
conducted at the Solid Propulsion Laboratory of Politecnico di Milano, over the same range of operating
conditions. In order to obtain a better understanding, experimental results were compared with those obtained
from a reference or base-line composition called E00 (a conventional AP/HTPB composite propellant
containing about the same Al fraction), even if the combustion requirements of E00 (high burning rate and
high pressure) are drastically different from those of E101 and E125. The measured densities of the three
composite propellants under examination were found to be:
• 1.679 ± 0.009 g/cm3 for E00;
• 1.743 ± 0.024 g/cm3 for E101;
• 1.740 ± 0.009 g/cm3 for E125.

The objective of this paper is to report the findings of the experimental investigations on transient and steady
state burning measurements for the above compositions.

RTO-MP-091 32 - 3

NATO UNCLASSIFIED
 NATO UNCLASSIFIED

Ballistic Properties of Scavenged Solid Rocket Propellants

2.0 LABORATORY EXPERIMENTAL INVESTIGATIONS

Radiative ignition was tested by means of a continuous wave, closed circuit water-cooled, CO2 laser with
200 W maximum power output and 11 mm beam diameter (1/e2 Gaussian beam). Full opening of the
mechanical shutter required about 3 ms. The resultant thermal flux impinging on the surface samples was in
the range of 100 to 500 W/cm2. Testing was performed in the combustion chamber depicted in Fig. 1, over the
range of 11 to 35 atm of nitrogen, with an internal chamber volume of 2.8 lt. A small brass dish was set under
the propellant sample to collect the combustion residues. The residues were analyzed by scanning electronic
microscopy (SEM).

Laser beam
CO2 Laser Mirror
Beam Splitter
IR Detector
Pressure Transducer
High-Speed Color Video Camera
Photodiode

Propellant sample
N2 in Gas out

Figure 1: Sketch of the Ignition Apparatus.

This kind of apparatus[6,7] is becoming rather common nowadays thanks to its easy control of the ignition
stimulus (both in terms of the energy flux and the exposure time). Data concerning the appearance of the first
flame and its development were systematically collected using a standard infrared (IR) photodiode. The latter
has a view angle of 30° and a spectral sensitivity in the range of 0.35 to 1.1 µm with maximum response at
0.8 µm. The photodiode’s, 5 mm2 active surface, was placed in front of the combustion chamber at a distance
of about 70 mm from the propellant sample. In addition, a high-speed color video camera (up to 2000
frames/s) was systematically used to supplement or corroborate the IR photodiode information. The pressure
was monitored/controlled with a piezoresistive transducer, 20 kHz natural frequency and 68 atm full scale,
calibrated before testing and mounted flush with the chamber wall. Propellant samples consisted in
parallelepipeds, 3.5 x 3.5 mm2 in cross-section and 5 mm in thickness. The burning surface was freshly cut
and no lateral surface inhibitor was used. All runs were performed in a nitrogen atmosphere at ambient
temperature. The measured ignition delay, defined as the time of first appearance of the flame from the shutter
opening, was reported in a standard time (t) vs. radiant flux intensity (q) bilogarithmic plot. The total error in
measuring the ignition delay has been estimated[8] to be smaller than 1 ms.

Ignition testing, with at least five runs at each investigated point, was performed under both continuous and
go/no-go radiation. As expected[9,10], no difference could be detected as far as the first flame appearance is
concerned. However, experimental results point out a relatively large scattering of the data probably due to the
heterogeneity of the propellant samples and multiphase effects at the burning surface layer. The complete set
of results obtained for the investigated compositions were reported[8] in a previous work. Figure 2 shows the
ignition delays measured at 35 atm. Best linear fitting laws were used to find the associated log t vs. log q
plots by a least squares method. The latter are reported below, along with the correlation factor R and the
total number of experimental points N. For all the tested compositions, the slope of the straight lines of the

32 - 4 RTO-MP-091

NATO UNCLASSIFIED
 NATO UNCLASSIFIED

Ballistic Properties of Scavenged Solid Rocket Propellants

log t vs. log q plot is much larger than the theoretical -2 value[11], as commonly found for AP-based composite
propellants irradiated with a CO2 laser, essentially due to radiation penetration[6,7]. In summary, the following
results were obtained for the time of first flame appearance (t being in s and q in W/cm2):

at 11 atm E00 t ign = (39.79 ± 9.870) ⋅ q −1.262 ± 0.046 R2=0.969 N=50

E101 t ign = (65.29 ± 18.42) ⋅ q −1.336 ± 0.050 R2=0.980 N=15

E125 t ign = (46.04 ± 16.27) ⋅ q −1.271 ± 0.065 R2=0.953 N=20

at 25 atm E00 t ign = (43.21 ± 10.80) ⋅ q −1.282 ± 0.047 R2=0.969 N=25

E101 t ign = (58.72 ± 12.17) ⋅ q −1.314 ± 0.037 R2=0.989 N=15

E125 t ign = (45.19 ± 16.10) ⋅ q −1.281 ± 0.065 R2=0.953 N=40

at 35 atm E00 t ign = (71.83 ± 17.58) ⋅ q −1.381 ± 0.045 R2=0.980 N=20

E101 t ign = (42.96 ± 10.48) ⋅ q −1.252 ± 0.044 R2=0.983 N=15

E125 t ign = (51.31 ± 17.16) ⋅ q −1.308 ± 0.063 R2=0.948 N=25

1009
8
7
6
5
4

2
time, t, s

10-19
8
7
6
5
4
E00
3
E125
2 E101

10-2 2 3 4 5
102
radiant flux, q, W/cm2

Figure 2: Ignition Results for the Tested Propellants under 35 atm of N2.

RTO-MP-091 32 - 5

NATO UNCLASSIFIED
 NATO UNCLASSIFIED

Ballistic Properties of Scavenged Solid Rocket Propellants

The ignition delay times (in terms of the first flame appearance) can be considered similar for all tested
compositions; only slightly shorter for the reference composition E00.

For all three tested compositions the influence of the pressure appeared negligible, as shown in Fig. 3 for
propellant E101. Likewise, ignitability is similar for all tested compositions and only slightly easier for the
reference composition E00; see Fig. 4 for propellants E125. In summary, the ignition delay (in terms of time
of first flame appearance) and ignitability, features common trends for all tested compositions and therefore
cannot explain the poor performances under the small-scale motor testing manifested by E101 and E125 with
respect to E00.

A second series of tests was carried out in the same ignition apparatus, see Fig. 1, to try to elucidate
the transient flame development from the appearance of the first flamelet to full flame development,
rather than just detecting the first flame appearance. Although it has been known for a long time that flame
development during ignition transients can follow a wide range of patterns leading to all sorts of combustion
pathologies[6-12,13], this is usually not a problem for motor applications due to the implementation of
well-behaved propellants under well-controlled operating conditions. However, in this instance, detailed
visual analyses immediately confirmed a scattered occurrence of the first flame with the formation of hot
spots on the burning surface, likely due to the combined effect of non-uniform laser heating and
heterogeneous mixture composition, as clearly manifested from the experimental results reported in Fig 2.
Surprisingly, both the high-speed color video camera and the IR photodiode also revealed a flame
development which is sluggish in time and erratic in space for E101 and E125 as compared to the reference
composition E00, see Fig. 7-5. Results obtained under continuous irradiation at 11 atm are reported in Fig. 5;
flame development is obviously even slower under go/no-go testing (see E125 at 11 atm in Fig. 5).
The overall flame structure appears typical of compositions with low aluminum content for E00, but strongly
irregular and heterogeneous for E101 and E125.

1009
8
7
6
5
4

2
time, t, s

10-19
8
7
6
5
4

3 11 atm
25 atm
2
35 atm

10-2 2 3 4 5
102
radiant flux, q, W/cm2

Figure 3: Ignition of Propellant E101 under 11, 25, and 35 atm of N2.

32 - 6 RTO-MP-091

NATO UNCLASSIFIED
 NATO UNCLASSIFIED

Ballistic Properties of Scavenged Solid Rocket Propellants

E125 11 atm
E125 25 atm
E125 35 atm
ignitability, qc, J/cm2

101
9

2 3 4 5
102
radiant flux, q, W/cm2

Figure 4: Ignitability of Propellant E125 under 11, 25, and 35 atm of N2.

0.8

0.7

0.6

0.5
time, t, s

E00 11 atm
0.4 E101 11 atm
E125 11 atm
E125 11 atm go/nogo
0.3

0.2

0.1

0.0
100 200 300 400 500
2
radiant flux, q, W/cm

Figure 5: Sluggish Flame Development for E101 (Continuous Irradiation) and E125 (Continuous and
Go/No-Go Irradiation) at 11 atm over the Whole Range of Radiant Flux.

A third series of experiments was conducted, under steady-state operating conditions, in a windowed
Crawford burner to verify if and how the sluggish flame development observed for compositions E101 and
E125 under ignition would affect the steady-state burning. Testing was performed in a different combustion

RTO-MP-091 32 - 7

NATO UNCLASSIFIED
 NATO UNCLASSIFIED

Ballistic Properties of Scavenged Solid Rocket Propellants

chamber, 0.8 lt of internal volume, under a low nitrogen flow and at ambient temperature. Propellant samples
were placed over a small brass dish allowing the collection of some solid combustion residues. The residues
were subsequently analyzed by infrared spectroscopy to obtain some qualitative data about their chemical
composition. Pressure was controlled by a pneumatic feedback system. At least three runs were performed at
each investigated point and five readings were obtained at each run. Propellant samples consisted in
parallelepipeds about 3.5 x 3.5 mm2 in cross-section and 30 mm in thickness, with a freshly-cut burning
surface and a lateral surface inhibitor. Ignition was initiated with a hot Nichrome wire. Burning rate
measurements were carried out by analyzing the digital video records. The total error in burning rate
measurements was estimated[8] to be about 0.5%. While it was possible to burn the reference composition E00
at a steady rate from 70 atm down to 1 atm, E101 and E125 could not be burnt steadily below respectively
about 9 and 7 atm. In addition, as shown in Fig. 6, steady burning rates of compositions E101 and E125 were
systematically lower compared to the reference composition E00, while exhibiting a relatively larger pressure
exponent. Compositions E101 and E125 show a break in the pressure exponent located respectively at 50 and
25 atm, in both cases the pressure exponent values decreased from about 0.8 to 0.5 and 0.65 respectively,
approaching values of 0.48 measured for the reference composition E00. In summary, the following results
were obtained for steady burning rates (rb being in cm/s and p in atm):

E00 rb = (0.149 ± 0.003) ⋅ p 0.480 ± 0.007 R2=0.991 N=21 p= 1-70 atm

E101 rb = (0.035 ± 0.002) ⋅ p 0.791 ± 0.016 R2=0.994 N=15 p= 11-50 atm

E101 rb = (0.106 ± 0.010) ⋅ p 0.504 ± 0.024 R2=0.985 N= 9 p= 50-70 atm

E125 rb = (0.029 ± 0.001) ⋅ p 0.821 ± 0.018 R2=0.996 N= 12 p= 11-25 atm

E125 rb = (0.049 ± 0.001) ⋅ p 0.652 ± 0.007 R2=0.998 N= 18 p= 25-70 atm

E00
E101
E125
100
0.9
0.8
0.7
burning rate, rb, cm/s

0.6

0.5

0.4

0.3

0.2

10-1
100 7
101 50
102
pressure, p, atm

Figure 6: Steady Burning Rates of Tested Propellants.

32 - 8 RTO-MP-091

NATO UNCLASSIFIED
 NATO UNCLASSIFIED

Ballistic Properties of Scavenged Solid Rocket Propellants

Figure 7: Flame Development of E00 (top) and E125 (bottom) Compositions.

The solid residues collected both during ignition and steady burning tests primarily consisted of big spherical
particles, about 100 µm in diameter. The simple collection technique implemented most probably did not
allow the collection of smaller particles. Particles from the ignition tests were analyzed by SEM microscopy,
see Fig. 8. The particles had a relatively smooth external surface, with the interior of some of the broken
particles showing a porous and coral like structure. The SEM observations also allow analysis of the surface
composition in terms of the elemental chemical species, see Fig. 9, indicating that for the scavenged
propellant the residue surface is mainly composed of Na, Cl, with some traces of Al. Analyses conducted by
infrared spectroscopy show the presence of the NO3 and NO2 functional groups, likely derived from
the thermal decomposition of SN. No differences could be observed between the combustion residues of
propellants E101 and E125. Video recordings show that the residues originate from the liquid layer,
which partially covers the burning surface, and seems to originate at the burning surface, or just below it,
from the melting/decomposing SN. Droplets appear in the liquid layer and grow in dimensions, as combustion
proceeds, until ejected from the burning surface. The growth is due to agglomeration with other droplets and
to collection of material from the liquid layer. Sometimes a “small explosion” with both a luminous flash and
the ejection of small luminous particles (likely burning Al) from the surface layer, accompanies the droplets
ejection. Bigger particles appear to be liquid just prior to impact on the collection dish; as a matter of fact,
the impact side of the big residues is flat. Also, from the ignition video records, it was observed that the
diameter of the residues decreases as either the pressure or the radiant heat flux increases.

RTO-MP-091 32 - 9

NATO UNCLASSIFIED
 NATO UNCLASSIFIED

Ballistic Properties of Scavenged Solid Rocket Propellants

Figure 8: Combustion Residues of Propellant E101 observed by SEM.

A Qualitative Chemical Composition of Region 1 and 2 Is Shown Fig 9.

Figure 9: Qualitative Chemical Composition of Residues Surface, Region 1 (left) and

Region 2 (right) of Fig. 8, obtained by SEM Microscopy.

32 - 10 RTO-MP-091

NATO UNCLASSIFIED
 NATO UNCLASSIFIED

Ballistic Properties of Scavenged Solid Rocket Propellants

3.0 DISCUSSION OF RESULTS

The effect of pressure and composition was found to be negligible, under the investigated set of operating
conditions. Overall, ignition delays of the tested compositions fall in the general class of AP-based composite
propellants[6,7,14,15,16]. However, this may be misleading for practical applications. Detailed visual observations
and testing under a variety of operating conditions revealed a slow time development and non-uniform
spatial propagation of E101 and E125 flames compared to the baseline E00. Likewise steady burning rates of
E101 and E125, with and without laser radiation assistance, are decreased compared to the baseline E00.
The difference decreases with either increasing pressure or increase in the radiant flux. A much larger value of
the pressure deflagration limit (PDL)[17] of compositions E101 and E125 with respect to the reference E00 was
observed.

A slope break occurs in the burning rate vs. pressure plots for propellants E101 and E125, while the baseline
composition E00 in the range 1-70 atm did not show any, see Fig. 6. A possible explanation of the scavenger
composition ballistic behavior has been given by Kondrikov et al.[18].The unusual form of the rb ( p ) plot for
propellants E101 and E125 is connected with two different forms of burning wave propagation which are
related in turn to the presence of two different subsystems in the propellant: subsystem 1 consisting of big
particles of SN and AP, and subsystem 2 (called the matrix) which includes HTPB, Al, and small and medium
sized particles of AP. At low pressures, the rate of gaseous reactions over the burning surface is low and the
reaction layer is relatively thick. The reaction determining the burning rate of the propellant proceeds as an
interaction of the overall mass of the reactants in the upper layers of the condensed phase, including the liquid
layer, on the one hand, and the first flame zone, on the other. Since the pressure exponent of E101 and E125 is
quite high in this pressure region, n = 0.8, it is argued that the reaction proceeds primarily in the kinetic
regime, in a well-stirred layer due to molecular diffusion and gas bubble formation.

The chemical reactions, taking place both in the liquid layer and in the hot gas, may be described only in an
approximate and generalized form. The elementary steps of this complex sequence of chemical reactions are
not fully known, but the main overall processes appear to be the following:
(1) NH4ClO4 + NaNO3 → NaClO4+NH4NO3 → NaClO4+2H2O + 3/4N2 +1/2NO2
(2) NH4ClO4 → ClOx, H2O, HCl, Cl2, O2
(3) (ClOx, Cl2, HCl ) + NaNO3 → NaCl + (NOx, H2O, O2)
(4) HTPB → Gaseous fuel (hydrocarbons) + solid carbonaceous residue
(5) Fuel + NOx → NO + (CO,CO2, H2O, N2)
(6) Fuel + ClOx → HCl, Cl2, CO,CO2, H2O
(7) NO + 1/2O2 ↔ NO2
(8) NaClO4 → NaCl + 2O2

Reaction (1) between solid AP and liquid SN may proceed, in particular, in the reaction zone that defines the
burning rate of the scavenged propellant in the low pressure regime. The sodium perchlorate formed in the
course of this reaction is a relatively stable solid, and, at low pressures, reaction (1) is just one of the different
ways to eliminate the active chlorine oxides from the reaction zone thereby preventing interaction of
the chlorine oxides with the fuel. Ammonium nitrate, on the contrary, decomposes quickly at 300-400 °C,

RTO-MP-091 32 - 11

NATO UNCLASSIFIED
 NATO UNCLASSIFIED

Ballistic Properties of Scavenged Solid Rocket Propellants

giving rise to nitrogen dioxide, the common oxidizer produced from the burning of nitrocompounds and
double-base propellants.

Reaction (2) outlines the process of AP degradation. The latter primarily produces the chlorine oxides, ClOx,
the main oxidizing agents from the burning of AP composite propellants. It also produces HCl, Cl2, H2O and
some O2. The chlorine oxides react with the liquid SN (reaction (3)) forming the most important final product,
NaCl, and a mixture of nitrogen oxides and Cl2. In the case of HCl, acting as a reagent, a small quantity
of water is also formed. The reactions of Cl2 and HCl with sodium nitrate proceed more slowly than the
ClOx + NaNO3 reactions, but they also take part in forming the corresponding quantities of NaCl, and, what is
especially important, the production of the NOx species, in the low pressures burning processes.

The main heat generating reactions are reactions (5) and (6). In the sequence of oxidizing agents that take part
in the reactions (5) and (6), the first one, NOx, is much less active as oxidizer than the second, ClOx.
But owing to the reactions (1) and (3), the NOx concentration in the liquid and in the first flame under low
pressure burning grows, whereas that of the chlorine oxides, as well as chlorine itself, reduces. Moreover,
the product of the nitrogen dioxide reaction, nitrogen oxide, may produce NO2 in the course of reaction (7).
Correspondingly, at low pressures, the NOx is the oxidizer, which is suggested to be primarily responsible for
the rate of heat evolution.

Under higher pressures, the combustion mechanism is drastically changed as compared to the low pressure
regime. The flame reactions responsible for the overall burning wave propagation are located in narrow
regions, quite close to the surface of the small matrix inclusion sites. In these small flames the reaction of
the chlorine oxides with the fuel is completed within the classic framework of composite propellants:
the diffusion regime results in the formation of the combustion products, CO, CO2, H2O, etc. and the chlorine
in the form of Cl2 and HCl, before noticeable quantities of NOx evolve in the course of the decomposition of
the large SN particles. In the first approximation, only very small concentrations of nitrogen oxides are
present in these small flames, and, accordingly, only traces of the reaction (5) are assumed. The burning
reactions satisfy the well-known laws of AP/HTPB composite propellants combustion, with the only
exception that burning rate is smaller due to cooling effects for the large SN particles.

In the high pressure combustion regime, the layer of liquid mixture of sodium nitrate and sodium chloride has
a thickness much greater than the thickness of the flame generated from burning of mixture composed by
HTPB and small AP particles. The diameter of the particles is approximately on the order of 0.01 mm and the
thickness of the gas-zone where mixing of the gaseous oxidizing agents and fuel completes, is approximately
of the same order of magnitude. Presumably, the average thickness of the liquid layer is at least one order of
magnitude greater. This offers an additional explanation to the fact that at pressures higher than the transition
point, burning of the matrix inclusions may proceed independently and nitrogen oxides would only slightly
affect it.

At the same time the thickness of the gaseous layer at low and moderate pressures, where mixing of the
oxidizing agents produced by decomposition of the large AP particles is complete, should be about the same
order of magnitude as the thickness of the liquid layer on the surface. Consequently, the layer takes part in all
of the processes taking place in the burning zone, first of all, generating nitrogen oxides instead of oxides of
chlorine as the main oxidizing agents and consequently suppressing the velocity of flame propagation.

Attention has to be placed on the potentially corrosive action of the NaCl product from the scavenging
reactions, on the internal walls of the combustion chamber. For this purpose various solutions are being
developed. The sodium chloride is potentially corrosive due to the existence of the chlorine in the highly

32 - 12 RTO-MP-091

NATO UNCLASSIFIED
 NATO UNCLASSIFIED

Ballistic Properties of Scavenged Solid Rocket Propellants

reactive Cl- ionic state. The chlorine must thus be reduced to a less reactive state. This can take place through
reaction with the gaseous combustion products from the propellant burning, to incorporate the chlorine into a
less reactive functional group or chemical species. Furthermore, the pressures and temperatures together with
efflux of the exhaust fumes can be optimized to increase the expulsion of the sodium chloride and reduce its
potential accumulation or deposition on the walls of the combustion chamber.

4.0 CONCLUSIONS
Ignition and steady burning of scavenged composite solid rocket propellants were experimentally investigated
at the Solid Propulsion Laboratory of Politecnico di Milano. A digital high speed camera allowed a visual
analysis of both the ignition and steady burning tests, revealing a liquid layer flowing over the burning surface
originating droplets, and large spherical particles. The liquid layer and particles appear to disrupt the regular
gas flow from the burning surface, hindering flame propagation of E101 and E125 compared to E00. A slope
change in the burning rate vs. pressure plot was evidenced for the scavenged compositions only. In the high
pressure region the pressure exponent decreases to about 0.5-0.65 from about 0.8 observed under low
pressure. The suggested explanation for this phenomenon is that two competing processes participate in
defining the propellant burning rate at low and moderate pressures, reactions (3) and (5), on one side,
and reactions (2) and (6), on the other. The first pair of reactions is responsible for the propellant burning at
lower pressures whereas the second pair, reactions (2) and (6), are more effective at higher pressure where
ClOx is evolved immediately near the burning surface. Under high pressures, where the high temperature zone
is close to the surface, reaction (6) taking place in the diffusion regime becomes a dominant factor of the
process as a whole. In addition, effects of regression rate and extent of reaction for the participating large
sodium nitrate particles at different pressures need to be considered.

One of the results which is apparent from this model and which seems useful for practical applications,
consists in the fact that one may move the transition point toward lower pressures by just increasing the
volume fraction of the HTPB/AP/Al matrix in the overall mixture. This might be achieved by increasing the
mass fraction of the small and/or medium particles of AP. It might be useful to vary also the sodium nitrate
particle size. For this purpose, future tests will be conducted with different sodium nitrate particle sizes.

From the point of view of performance, scavenged propellant are comparable to conventional AP-based
composition. Due to the higher density of SN as compared to AP a slight increase in the volumetric specific
impulse can be achieved, while a small loss in the specific impulse has to be paid[3,4]. By using scavenged
propellant a reduction of HCl emission can be achieved[3,4] up to 75%. Being SN a cheap raw material and
being the propellant manufacture essentially the same of AP-based composite propellants, the cost of
scavenged composition is equivalent to the cost of classical composite propellants[5].

Finally, attention has to be placed on the potentially corrosive action of the NaCl on the internal walls of the
combustion chamber.

5.0 ACKNOWLEDGMENTS
The financial support by FiatAvio–Comprensorio BPD and ASI, contract I/R/37-00, for this investigation
is gratefully acknowledged. The authors wish also to thank Dott. B. D’Andrea and Dott.sa F. Lillo
(FiatAvio–Comprensorio BPD) for providing important data and helpful discussions.

RTO-MP-091 32 - 13

NATO UNCLASSIFIED
 NATO UNCLASSIFIED

Ballistic Properties of Scavenged Solid Rocket Propellants

6.0 REFERENCES
[1] “Environmental Aspects of Rocket and Gun Propulsion”, AGARD Conference Proceedings No. 559,
Paris, France, February 1995.

[2] B. D’Andrea and F. Lillo, “Industrial Constraints as Evaluation Criteria in Developing Solid Space
Propellants using Alternative Energetic Materials”, AIAA paper 97-2795.

[3] B. D’Andrea, F. Lillo, A. Faure, and C. Perut, “A New Generation of Solid Propellants for Space
Launchers”, Acta Astronautica, Vol. 47, Nos. 2-9, pp. 103-112, 2000.

[4] B. D’Andrea, F. Lillo, and R. Massimi, “Performance Investigation of Advanced Solid Propellants
Formulations by Conventional and Innovative Methodologies”, 27th International ICT Conference,
paper P94, 1996.

[5] C. Perut, V. Bodart, B. Cristofoli, “Propergols solides pour lanceurs spatiaux generant pas ou peu de gaz
chlorhydrique”, AGARD Conference Proceedings No. 559, Paris, France, Paper 4, February 1995.

[6] L.T. De Luca, T.J. Ohlemiller, L.H. Caveny, M. Summerfield, “Solid Propellant Ignition and Other
Unsteady Combustion Phenomena induced by Radiation”, Technical Report, Department of Aerospace
and Mechanical Sciences, Princeton University, Princeton, New Jersey, 1976.

[7] C.E. Hermance, “Solid-Propellant Ignition Theories and Experiments”, in “Fundamentals of Solid
Propellant Combustion”, edited by K.K. Kuo and M. Summerfield, AIAA Progress in Astronautics and
Aeronautics, 1984, Vol. 90, chapter 5, pp. 239-304, AIAA, Washington, DC, USA.

[8] L.T. DeLuca, F. Cozzi, S. Manenti, and A. Olivani, “Ballistic Testing of Clean Solid Rocket
Propellants”, 32nd International ICT Conference, Paper V10, 2001.

[9] L.T. DeLuca, L.H. Caveny, T.J. Ohlemiller, M. Summerfield, “Radiative Ignition of Double-Base
Propellants: I. Some Formulation Effects”, AIAA Journal, 1976, Vol. 14, No. 7, pp. 940-946.

[10] L.T. DeLuca, T.J. Ohlemiller, L.H. Caveny, M. Summerfield, “Radiative Ignition of Double-Base
Propellants: II. Pre-Ignition Events and Source Effects”, AIAA Journal, 1976, Vol. 14, No. 8,
pp. 1111-1117.

[11] H.S. Carslaw and J.C. Jaeger, “Conduction of Heat in Solids”, Oxford University Press, London, UK,
2nd edition, 1971.

[12] T.J. Ohlemiller, L.H. Caveny, L.T. De Luca, M. Summerfield, “Dynamic Effects of Ignitability Limits
of Solid Propellants Subjected to Radiative Heating”, in 14th Symposium (International) on Combustion,
The Combustion Institute, Pittsburgh, PA, USA, 1972, pp. 1297-1307.

[13] B.N. Kondrikov, S. Cristoforetti, I.V. Grebenyuk, and L.T. DeLuca, “Gasification of Solid Propellants
and Propellant Ingredients under Influence of Thermal Radiation”, 32nd International ICT Conference,
paper P98, 2001.

[14] A.D. Baer and N.W. Ryan, “Ignition of Composite Propellants by Low Radiant Fluxes”, AIAA Journal,
1965, Vol. 3, No. 5, pp. 884-889.

32 - 14 RTO-MP-091

NATO UNCLASSIFIED
 NATO UNCLASSIFIED

Ballistic Properties of Scavenged Solid Rocket Propellants

[15] E.W. Price, H.H. Jr. Bradley, G.L. Dehority, and M.M. Ibiricu, “Theory of Ignition of Solid
Propellants”, AIAA Journal, 1966, Vol. 4, No. 7, pp. 1153-1181.

[16] G. Lengellé, A. Bizot, J. Duterque, and J.C. Amiot, “Ignition of Solid Propellants”, La Recherche
Aérospatiale, 1991, pp. 1-20, No. 1991-2.

[17] C. Bruno, G. Riva, C. Zanotti, R. Dondé, C. Grimaldi, and L.T. DeLuca, “Experimental and Theoretical
Burning of Solid Rocket Propellants Near the Pressure Deflagration Limit”, Acta Astronautica, 1985,
Vol. 12, No. 5, pp. 351-360. See also IAF Paper 83-367, 1983.

[18] B.N. Kondrikov, L.T. DeLuca, F. Cozzi, A. Olivani, and S. Manenti, “Steady-State Burning of a
Scavenged Solid Composite Propellant”, 33rd Annual Conference of ICT, Paper P136, 2002.

RTO-MP-091 32 - 15

NATO UNCLASSIFIED
 NATO UNCLASSIFIED

Ballistic Properties of Scavenged Solid Rocket Propellants

This page has been deliberately left blank

Page intentionnellement blanche

32 - 16 RTO-MP-091

NATO UNCLASSIFIED