Propulsion and Power Research 2018;7(3):218–230

HOSTED BY http://ppr.buaa.edu.cn/

Propulsion and Power Research

www.sciencedirect.com

ORIGINAL ARTICLE

Characterization and manufacturing of a paraffin

wax as fuel for hybrid rockets
F. Piscitellin, G. Saccone, A. Gianvito, G. Cosentino, L. Mazzola

CIRA, Italian Aerospace Research Centre, Capua, CE 81043, Italy

Received 14 September 2017; accepted 18 April 2018

Available online 31 August 2018

KEYWORDS Abstract The hybrid propulsion performed with paraffin waxes exhibits most attractive
capabilities compared to solid or liquid engines, e.g., throttleability and re-ignition, alongside
Paraffin wax;
Hybrid rockets; higher regression rates compared to the conventional hydroxyl terminated polybutadiene
Manufacturing; (HTPB) hybrid fuel. This is because the paraffin wax forms a thin and hydro-dynamically
SASOLs 0907; unstable liquid layer, and then enhances the regression rate with the entrainment of droplets
Regression rate from the liquid-gas interface. Nevertheless, some critical open points on the manufacturing of
the paraffin fuel grains still persist, because the paraffin wax exhibits high shrinkage during the
solidification phase, leading to the formation of cavities, cracks and internal rips, which may be
detrimental to the mechanical properties and the structural integrity of the fuel grain. In this
context, this paper deals with a wide calorimetric, thermo-mechanical and physical character-
ization of the paraffin wax selected to manufacture the hybrid rocket engines (HRE) fuel grain,
in order to gain a thorough knowledge of the material necessary to avoid the formation of
critical defects. Several manufacturing methods were investigated, and it was found that only
laboratory scale processes, based on the use of a heated circular mould-piston apparatus, are
able to avoid the formation of critical defects, with the application of both high temperature
and pressure.
& 2018 National Laboratory for Aeronautics and Astronautics. Production and hosting by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).

n
Corresponding author.
E-mail address: [email protected] (F. Piscitelli).
Peer review under responsibility of National Laboratory for Aeronautics and Astronautics, China.

https://doi.org/10.1016/j.jppr.2018.07.007
2212-540X & 2018 National Laboratory for Aeronautics and Astronautics. Production and hosting by Elsevier B.V. This is an open access article under the
CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Characterization and manufacturing of a paraffin wax 219

Nomenclature TRL technology readiness level

ac critical size of micro-cracks, voids, etc. Greek symbols

A burning area
AM additive manufacturing η* complex viscosity (unit: Pa . s)
DSC differential scanning calorimetry γ̇ shear rate (unit: Hz)
EVA ethylene vinyl acetate copolymer γ surface tension in the Owens-Wendt model
FDM fused deposition modeling μ melt layer viscosity (unit: Pa . s)
FE finite element ρ density (unit: kg/m3)
G’ shear storage modulus (unit: Pa) σ surface tension of the melt layer (unit: mN/m)
G” shear loss modulus (unit: Pa) σc critical stress
h thickness of the melt layer (unit: m) θ contact angle
HRE hybrid rocket engines
HTPB hydroxyl terminated polybutadiene
Subscripts
I impulse (unit: N . s)
KIC fracture toughness
entr entrainment
ṁ mass flow rate (unit: kg/s)
p dynamic pressure (unit: Pa) dyn dynamic
PE polyethylene f fuel
r average regression rate (unit: m/s) l liquid
T temperature (unit: K) s solid
TGA thermal-gravimetric analysis sp specific

1. Introduction freight cargo, low cost, and reusable (recycling possibili-

ties). In addition, it they possess the same energy per unit
Within the Italian national research program HYPROB- mass as kerosene, but their density is 16% greater, no scrap
NEW, a specific project is dedicated to hybrid propulsion. possibility, and a long shelf life [3–6]. Finally, since this
The objective is to increase the technology readiness level fuel is non-explosive, grains can be fabricated on-site and
(TRL) of this system through development and testing of a thus can save both manufacturing and launch operation
rocket engine demonstrator of 1 kN thrust class, based on costs [7,8]. In spite of the aforementioned advantages,
gaseous O2 and paraffin wax. which encourage the replacement of HTPB, paraffin waxes
Hybrid rockets combine several advantages of solid and have two disadvantages: poor mechanical properties and
liquid rockets engines, overcoming in some cases their complex manufacturing process.
disadvantages and limitations. Compared to solid rockets, In order to improve the poor mechanical properties,
they are much less sensitive to the propellant cracks and de- several authors have formulated blends of paraffin wax
bonds. Compared to a liquid system, the hybrid takes with polymers, i.e., polyethylene (PE) [9–12] or ethylene
advantage of a reduced explosion hazard and complexity vinyl acetate copolymer (EVA) [9,13,14]. In some cases,
due to the absence of regenerative cooling systems for both aluminum or other energetic additives were added, with the
the combustion chamber and nozzle. Furthermore, they aim of improving the reduced regression rate [10].
have high specific impulse Isp, which can be further Also the manufacturing process of fuel grains remains an
improved using suitable additives. Moreover, they are open point, especially when grains with large diameters
throtteable, including shutdown and restart on demand. have to be manufactured. In fact, the high shrinkage of
Additionally, they are safe compared to solid propellant about 15%–25% [7] occurring during the solidification of
rockets, because the oxidizer is not in contact with the fuel the paraffin wax, leads to grain deformations, internal
until after it is injected into the combustion chamber. One of stresses, rips, defects, micro-cracks and other microstruc-
the few disadvantages of the hybrid rocket engines (HRE) is tural discontinuities.
the low burning rate, which was overcome with the use of Additionally, the intrinsic brittle behaviour of this mate-
paraffin waxes. They form a thin, hydro-dynamically rial reduces its workability, and finally its thermal inertia
unstable liquid layer on the fuel-melting surface, thus hinders a uniform cooling of the whole fuel grain.
enhancing the regression rate with the entrainment of Obviously, the influence of these phenomena increases as
droplets from the liquid-gas interface [1–4] (see Figure 1). the diameter of the fuel cylindrical grain increases. In
Regression rates of paraffin waxes, in fact, are 3 to 5 times addition, during operations, the fuel grain is exposed to a
higher than conventional fuels for HRE, i.e., hydroxyl pressure of about 40 bar (for the 1 kN on-ground demon-
terminated polybutadiene (HTPB). Compared to the classi- strator of HYPROB-NEW HRE [15]), so cracks or voids
cal HTPB system, paraffin waxes also offer several other can potentially cause a detachment of large pieces of fuel,
advantages, i.e., non-toxic, non-hazardous, shippable as which could obstruct the rocket nozzle. For all these
220 F. Piscitelli et al.

Figure 2 Content of linear (n) and branched (iso) alkanes.

Some researchers experimented using the innovative

methods of additive manufacturing (AM), e.g., the layer-
by-layer fused deposition modeling (FDM), 3D printing
procedure, to manufacture homogeneous solid grains with
ports of complex geometry and consequent different thrust
profiles having higher consistency, lower costs and shorter
lead-times in comparison with the conventional casting
methods. Additional benefits of the 3D printing method
consist in the ability to produce large grain sections without
critical defects and with excellent accuracy (see Stratasys
for a 3D worldTM technology). Arnold [18] used a machine
originally designed to produce solid acrylic parts with
paraffin printed as a support to fill voids, to produce
paraffin fuel grain structured with honeycomb acrylic
features. Theoretically, this method allows the dispersion
of energetic additives, but due to lack of previous work on
the subject, the authors preferred to subdivide the manu-
facturing into two steps, i.e., the 3D printing of the tubular
honeycomb acrylic structure, and then the spin casting
Figure 1 Schematic comparison between (a) liquefying solid fuel process of molten paraffin wax and additives around the
(e.g., paraffin wax) and (b) non-liquefying solid fuel (e.g., HTPB). acrylic structure.
The goal of this paper is to find a method to fabricate fuel
grains of paraffin wax free of defects, voids and cracks. To
reasons, the fabrication of a uniform solid fuel grain is of
this aim, an extensive calorimetric, and physical character-
crucial importance. Some authors solved this issue by using
ization of the selected paraffin wax, i.e., SASOLs 0907,
a circular mould-piston apparatus able to compensate for
was performed and discussed here.
the volumetric shrinkage occurring during the paraffin
Firing test campaigns on intermediate sub-scale motors
phase transition [16]. According to this method, the forma-
with 200 N thrust class, are ongoing in order to verify
tion of voids is minimized by using a piston, since the air
characteristics of the selected paraffin wax (SASOLs 0907)
trapped during the melting process can be removed.
under combustion conditions, using the test bench of
The effectiveness of a suitable thermal control has been
Industrial Engineering Department of University of Naples
investigated by Gerbal et al. [17] by means of finite element
Federico II, located at Grazzanise (Italy).
analysis (FEA). In fact, they reduce the volumetric shrink-
age by decreasing the temperature gradients in the paraffin
grain. De Sain et al. [6] preferred the use of a special spin
casting machine with a variable rotational speed, according 2. Materials and methods
to which molten wax sealed in a cylindrical container is
spun around the desired port axis by a centrifugal force, and 2.1. Fuel selection
is solidified as the system cools down. The spinning rate is
conveniently changed in order to distribute the additives Four different waxes were considered as potential fuels for
homogeneously, e.g., carbon black, along the cross section hybrid rockets, i.e., H1, C80, 6003 and 0907, all manufac-
of the wax fuel grain. In addition, by forcing the liquid tured by the SASOLs. They are composed of linear (n), and
paraffin wax to grow on the lateral paraffin wall, the effects branched (iso) hydrocarbons, in different percentages, as
of the shrinkage are completely removed. Simultaneously, shown in Figure 2. In addition in Figure 3 the molecular
the centrifugal force forms the central port yielding a glass- weight distribution of linear and branched alkanes in
smooth finish, with tolerances up to 70.7 mm [6]. SASOLs 0907 paraffin wax (data provided by SASOL) is
Characterization and manufacturing of a paraffin wax 221

Figure 3 Molecular weight distribution of linear and branched alkanes in SASOLs 0907 paraffin wax.

Figure 4 Molecular weight distribution (higher than the 10 wt% vs the maximum abundance) as Carbon number in linear (n) and branched (iso)
fractions.

Table 1 Wax properties declared by the manufacturer SASOL.

Melting or droppling Congealing Oil Penetration at Kinematic viscosity at Brookfield viscosity at

point/℃ point/℃ content/% 25 ℃ /(1/10 mm) 100 ℃ /(mm2/s) 135 ℃ /cP

H1 112 96–100 1 6–8

C80 88 78–83 6 9.4
6003 60–62 0–0.5 17–20
0907 88–102 83–94 0–1.5 4–10 14–18

reported. Moreover, the distribution of their molecular distribution of the SASOLs 0907 branched alkanes is
weights is centred on different values, as shown in centred on the highest value.
Figure 4, where percentages higher than the 10% by weight The high fraction of branched iso-alkanes in SASOLs
of the maximum abundance is reported as a function of the 0907 may be more advantageous to form microcrystals,
carbon number, for both linear and branched fractions. which allows higher fracture toughness and thus a better
Table 1 shows the wax properties cited by the manu- workability during the manufacturing of the fuel grain in
facturer SASOLs in terms of melting and congealing comparison with the linear paraffin waxes. Moreover,
points, penetration and viscosity. It appears that the H1 paraffin solid wax is an intrinsically fragile material at least
formulation consists predominantly of linear alkanes having at the investigated operating conditions. For this reason and
a large distribution of the molecular weights, i.e., ranging according to Griffith theory, higher values of fracture
between C32 and C97 centered at C65. SASOLs 6003 has an toughness increase the tolerance of surface or internal
abundance of linear alkane chains with respect to the micro-cracks and other microstructural defects of increased
branched alkanes, but both linear and branched molecular radius, keeping constant, at the same time the applied
weight distributions are centered on the same value. Unlike thermo-mechanical stress level during operations.
the other paraffins, the formulation SASOLs 0907 is For all these reasons, the SASOLs 0907 was selected to
predominantly composed of branched alkanes, whose con- manufacture the fuel grains for the HYPROB-NEW HRE
tent is also the highest. Additionally, the molecular weight technology demonstrator.
222 F. Piscitelli et al.

2.2. Differential scanning calorimetry (DSC)

Differential scanning calorimetry (DSC) measurements

were performed using a TA instruments DSC Q20. Every
test was carried out by firstly cooling the paraffin wax down
to 0 ℃, to have a suitable baseline, and then heating it up to
120 ℃ with heating rates of 2, 10, and 40 ℃/min. A
possible modification of the paraffin crystal structure,
caused by the cooling step during the manufacturing
process, could affect its integrity and consequently the
mechanical performance of the fuel grain. Thus, to study
the effect of the cooling rate on the paraffin crystallization,
also the cooling steps at 2, 10, and 40 ℃/min and the re- Figure 6 DSC thermograms at different cooling rates
(endothermic down).
heating at 10 ℃/min of the SASOLs 0907 paraffin wax
were monitored.

2.3. Thermo-gravimetric analyses (TGA)

Thermo-gravimetric analyses (TGA) were performed

using the Discovery TGA by TA Instruments. The pellets
of paraffin wax were heated from 40 ℃ to 700 ℃ at 10
℃/min, using both nitrogen and air controlled atmospheres.

2.4. Surface tension measurements

The surface free energy was calculated according to the

Owens-Wendt method [19]. The Owens-Wendt approach is Figure 7 DSC thermograms at 10 ℃/min of samples cooled at
one of the most commonly used methods to calculate the different cooling rates (endothermic down).
surface free energy of materials [20]. The main assumption
of this method is that the surface free energy is the sum of
two components: the dispersion and the polar components
Table 2 Parameters obtained from DSC measurements.
[21,22]. The Owens-Wendt model is described by the
geometric mean relationship: Heating/cooling rate/ Tm,1/ Tm,2/ Tm,3/ Tc,1/ Tc,2/ Tc,3/
1
1
1 (℃/min) ℃ ℃ ℃ ℃ ℃ ℃
ð1 þ cos θÞγ L ¼ γ D
S U γl
D 2
þ γ PS U γ Pl 2 ð1Þ
2 2 60 78 102.5 54 74 85
where θ is the contact angle between the liquid droplet and 10 67 81 97 54 74 83
40 87 95.5 51 67 77
surface, γ L the liquid total surface tension, γ D
l the dispersion
component of liquid surface tension, and γ Pl the polar
component of liquid surface tension. The unknown terms The value of the total surface free energy of the solid is
in the equation are γ D
S , which is the dispersion component of
1
the solid surface free energy, and γ PS , namely the polar obtained as γ S ¼ γ D
S þ γ S . Dividing by γ l , the Eq. (1) can
P P 2

component of the solid surface free energy. be re-written as:

 D 12
2 ð1
1
þ cos θÞγ L γ
12
1

12 ¼ lP U γ D þ γ PS 2 ð2Þ
γl
P γ l
S

The left-hand side of Eq. (2) contains quantities which

are measured experimentally (θ) or which are available in
the literature ðγ Pl ; γ L Þ, so that a plot of the left-hand side of
 D 12
γ
Eq. (2) versus quantity γlP gives a straight line with slope

12
1
l

γD
S and intercept γ P 2
S .
To assess the surface free energy of SASOLs 0907
Figure 5 DSC thermograms at different heating rates paraffin wax, measurements of contact angle using liquids
(endothermic down). with different polarities, i.e., water, formamide, and
Characterization and manufacturing of a paraffin wax 223

Figure 8 Thermogravimetric analysis in nitrogen atmosphere. Figure 9 Thermogravimetric analysis in air atmosphere.

diiodomethane, were performed. For each liquid the contact

angle values are averaged over ten measurements.
2.5. Rheological characterization

The rheological characterization was performed using a

TA Instruments HR-2 Hybrid Rheometer with 40 mm
parallel plate geometry and 500 μm gap distance. Measure-
ments of storage and loss moduli, and complex viscosity, in
temperature sweep mode were performed on both solid and
melted paraffin waxes. In detail, a heating rate of 2 ℃/min
was set for the solid state between 30 and 80 ℃, and
20 ℃/min for the temperature sweep from 100 to 200 ℃. A
frequency of 1 Hz was set for all tests. In addition, a Figure 10 Surface tension as a function of the temperature.
frequency sweep ranging from 0.1 to 100 Hz, i.e., from
0.6 to 628.5 rad/s, at 150 ℃ was carried out. This was
because the melt layer viscosity has a great influence on the different peaks and a shoulder at the highest temperatures
regression rate [1], since, as previously reported [20], it are apparent, as evident at the lowest heating rate, i.e., 2
affects the entrainment contribution of the mass flow rate, ℃/min. The affinity between the shape of thermograms and
pαdyn hβ
ṁentr p μγl σ π
where, according to Karabeyouglu et al. [1,2]. the molecular weight distribution, shown in Figure 3,
pdyn is the dynamic pressure, ℎ the thickness of the melt suggests that there may be a relationship between them.
layer, σ the surface tension and μl the melt layer dynamic Hence, by analyzing the DSC thermogram obtained at 2
viscosity. Exponents values α, β, γ, π can be found in the ℃/min, the peak at 60 ℃ could correspond to the melting
literature from Gater and L’Ecuyer [23], Nigmatulin et al. of the molecular weight fraction centred at about C37,
[24], and Karabeyouglu [1]. whereas the peak at 78 ℃ could correspond to the melting
of the molecular weight fraction centred at about C50, and
finally, the peak at 102.5 ℃ could be related to the melting
2.6. Fractographic analysis
of the molecular weight fraction centred at C83. Differences
in sharpness of peaks among thermograms at different
Fracture surfaces of solid grain paraffin wax samples,
heating rates can be attributed to the thermal inertia of the
manufactured according to various methods, were inspected
paraffin wax. Figure 6 shows the thermograms correspond-
by means of a stereo microscope the LEICA MZ-12X
ing to the different cooling rates of the molten paraffin.
whereas the micro-fractography was conducted using an
Again, no relevant differences can be observed apart from
optical microscope the LEICA DM-RXE in bright field and
the delay in temperature and the sharpness of peaks.
using an optical magnification of 550x.
Figure 7 shows the thermograms related to the heating
scans at 10 ℃/min after three different cooling rates. Some
3. Experimental results differences appear only at 90–100 ℃, temperatures at which
the highest molecular weights of branched alkanes are able
3.1. Differential scanning calorimetry (DSC) to melt, as the molecular weight distribution of Figure 3
suggests. Probably, this fraction is able to organize itself
The DSC results at different heating rates of paraffin into different crystalline structures when it undergoes
pellets are shown in Figure 5 (endothermic down). Two different cooling rates.
224 F. Piscitelli et al.

Figure 11 The three different shapes of the liquid droplets ((a) distilled water, (b) formamide, and (c) diiodomethane) in contact with the paraffin
sample.

Table 3 Surface tension parameters.

Wettability [1] 8777

Surface free energy [mJ/m2] 35
Dispersion component [mJ/m2] 33
Polar component [mJ/m2] 1
Work of adhesion paraffin-water [mJ/m2] 77

Figure 13 Dynamic viscosity of paraffin pellets in solid and liquid

states.

Figure 12 Comparison between storage modulus at 1 Hz and heat

flow at 10 ℃/min as a function of the temperature.

In Table 2 all data provided by the DSC measurements

are summarized. They differ slightly from those reported in
the literature [25]. In fact, at a heating rate of 10 ℃/min a
value of Tm,1 equal to 67 ℃ instead of 59.7 ℃ was Figure 14 G” and viscosity measured in frequency sweep mode at
measured. Similar differences were observed for Tm,2 equal 150 ℃.
to 81 ℃ instead of 82.1 ℃, and a Tm,3 equal to 97 ℃
instead of 98.9 ℃. Slight differences appear also during the liquefied, independently of the heating rate. The cooling
cooling ramp, particularly a value of 54 ℃ instead of step of the manufacturing process of the fuel grain has no
47.1 ℃ for the Tc,1, was measured and similarly a value relevant effects on its final crystal structure, since it affects
of 74 instead of 70.0 ℃ for the Tc,2, and a value of 83 only the highest branched molecular weight fraction, which
instead of 83.8 ℃ for the Tc,3. These differences can represents only a small percentage of the total.
presumably be attributed to the different instruments used.
In general, the wide melting and solidification ranges 3.2. Thermo-gravimetric analyses (TGA)
observed are probably due to the high percentage of
branched alkanes, which gradually undergo the phase The thermogravimetric results illustrated in Figures 8 and
transition [25]. 9 highlight that the weight losses at temperatures lower than
To summarize, regarding the manufacturing of the fuel 110 ℃, the maximum temperature at which the investigated
grain, these analyses suggest that at 110 ℃ and at atmo- paraffin wax is subjected, during the fuel grain manufacture,
spheric pressure the paraffin wax pellets are completely are negligible. In fact, only at 190 ℃ the weight losses
Characterization and manufacturing of a paraffin wax 225

Figure 15 CAD of paraffin wax solid fuel grain sample configuration. (a) Cross section, (b) longitudinal section, and (c) whole shape of the
paraffin grain.

Table 4 Size of solid paraffin wax grain configurations.

Configuration Height H/mm External diameter De/mm Internal diameter Di/mm Aspect ratio

1 80 65 Not present 0.8

2 95 50 Not present 0.5
3 35 50 4 1.4

become significant reaching the value of 0.1 wt%. This is energy is expressed in [J/m2], which is equal to [Nm/m2],
evidence that volatile elements are almost absent at whereas the surface tension is measured in [N/m], conse-
temperatures involved in the production process of the fuel quently the two performance indexes can be compared. In
grain, so the process should be considered completely safe. order to obtain the surface tension value at high temperature,
The thermo-gravimetric analysis performed in a nitrogen i.e., corresponding to the liquid state of SASOLs 0907
atmosphere (see Figure 8) shows a narrow and well-defined paraffin wax, a measure at 20 ℃ of the surface tension of
thermal event, which corresponds to the boiling point of the the solid paraffin wax was carried out, giving in output a value
paraffin wax occurring at about 430 ℃. In addition, the of 34.78 mN/m. This value is in good agreement with the
analysis performed in an air atmosphere (Figure 9) high- corresponding value of 33.8 mN/m obtained by linearly
lights that, after the main oxidative degradation at about extrapolating (see Figure 10) the surface tension values
300 ℃, there is a significant residual, ascribed to the measured by Kobald et al. [26,27], meaning that the value
formation of ashes, which are difficult to oxidize further of 27.5 mN/m at 110 ℃ can be used.
at temperatures below 550 ℃. The formation of these ashes The results in Figure 11 show the shape of liquid droplets
could affect the re-ignition process of the fuel grain during on the paraffin wax surface. All results were processed and
the firing test. the performance indexes are summarized in Table 3.
It is worth noting that water droplet on the SASOLs
3.3. Surface tension measurements 0907 paraffin wax surface describes a partially hydrophobic
behavior. This was confirmed also by analyzing the contact
The surface free energy was measured on solid paraffin angle measurements on formamide, partially polar liquid
pellets of SASOLs 0907, by putting them in a silicone mold with respect to the water. Whereas the paraffin wax high-
and melting then at 110 ℃. The paraffin wax sample was lights a hydrophilic behavior with diiodomethane due to its
solidified very slowly in order to avoid the formation of apolar nature.
possible macro and micro-cracks due to the liquid-solid
transition [26]. Through this procedure, it is possible to realize 3.4. Rheological and viscosity measurements
a paraffin sample with a very smooth surface, where the
measured surface free energy will depend only on the physical The storage modulus and complex viscosity of paraffin
state of the paraffin, and not on its morphology. In fact, during wax in the solid state are shown in Figure 12 and Figure 13,
the melting process, it is evident that a very smooth layer of respectively. In particular, Figure 12 shows the comparison
melted paraffin pellets is obtained due to the surface tension of between the heat flow during the DSC heating scan at
the liquid paraffin. It is worth noting, that the surface tension is 10 ℃/min and the storage modulus, highlighting that
correlated to the liquid paraffin wax, whereas the surface free paraffin wax has a storage modulus of 1.7  106 Pa, in
energy is related to the solid paraffin wax. The surface free the solid state, up to the first melting peak at about 70 ℃,
226 F. Piscitelli et al.

Figure 19 Optical microscope images obtained in bright field using a

magnification of 550x of paraffin solid grains manufactured with the
laboratory method at 80 ℃ as peak temperature.

Figure 16 Paraffin wax solid grain sample manufactured with the

laboratory method at 80 ℃ as peak temperature.

Figure 20 Paraffin wax solid grain sample manufactured with the

Figure 17 Fracture surfaces of paraffin wax solid grains manufac- laboratory method at 110 ℃ as peak temperature.
tured with the laboratory method at 80 ℃ as peak temperature.
In the ṁentr formula a value of viscosity at an average
temperature, ranging between the melting and boiling
(about 450 ℃) points, calculated using equation from
Marano and Holder [28,29] must be introduced. Since it
is not possible to perform measurements at same high
temperatures, the value of viscosity must be extrapolated
starting from values of complex viscosity obtained in the
range 100–200 ℃, using the asymptotic behavior correla-
tions (ABC) attributed to Kreglewski and Zwolinski [30].
This is a method which predicts PVT, thermal and transport
properties of n-paraffins and n-olefins; and it was also
Figure 18 Optical microscope images of paraffin wax solid grain
demonstrated that it could be successfully applied to
samples manufactured with the laboratory method at 80 ℃ as peak
branched alkanes if the length of branched chains are
temperature.
limited [30]. The Newtonian behavior of the liquid
SASOLs 0907 paraffin wax suggests that the branched
temperature at which the SASOLs 0907 paraffin wax
length chains are short enough to not affect the rheological
ceases to preserve its elastic response anymore. Similarly,
outputs where in this case, the viscosity errors are limited to
the complex viscosity (Figure 13) ranges between 2.8 and
8% [30]. According to the ABC method, the viscosity can
2.2  105 Pa . s before melting, whereas in the liquid
be assessed using the following equations:
phase, the viscosity decreases as the temperature increases,
as expected, with an Arrhenius-type behavior. In Figure 14 Y ¼ Y 1;0 þ ΔY 1 ðn n0 Þ ΔY 0 expð βðn n0 Þγ Þ ð3Þ
the loss modulus G” and dynamic viscosity measured in
frequency sweep mode at 150 ℃ are illustrated. It should be ΔY ¼ A þ B=T þ ClnT þ DT 2 þ E=T 2 ð4Þ
noticed that in the frequency range from 0.6 to 30 rad/s, the where Y is the property value, e.g., viscosity, at a carbon
melted paraffin wax displays a Newtonian behavior. number n, Y0 is the property pseudo-value at n ¼ 0, Y1 is
It is concluded that, the viscosity does not change with the limiting property value as n becomes very high, β and γ
angular frequency ω, whereas the loss modulus G” increases are adjustable parameters, A, B, C, D, and E, are tempera-
linearly as the angular frequency ω increases. ture dependent ABC parameters, listed in Ref. [30]. The
Characterization and manufacturing of a paraffin wax 227

Figure 21 Fracture surfaces of paraffin wax solid grains manufac-

tured with the laboratory method at 110 ℃ as peak temperature.

Figure 24 Scheme of in-house device used for paraffin wax fuel

solid grains manufacturing.

Figure 22 Optical microscope images of paraffin wax solid grain

samples manufactured with the laboratory method at 110 ℃ as peak
temperature.

Figure 25 Paraffin wax solid grain manufactured with the die casting
method.

Figure 23 Optical microscope images obtained in bright field using a third configuration, the central co-axial port was realized,
magnification of 550x of paraffin wax solid grain samples manufac- after solidification, by means of mechanical machining.
tured with the laboratory method at 110 ℃ as peak temperature. In order to highlight the presence of voids and cracks, the
paraffin fuel grains, once divided into two parts, were
viscosity was assessed at the temperature of 280 ℃, inspected and fractographically analyzed.
assumed as the initial temperature of the combustion Initially, basic laboratory procedures were used for the
chamber, which is the average temperature between the first preliminary manufacturing tests of the solid fuel grains.
melting and the boiling (i.e., about 450 ℃) points of the This method consisted in weighing the required amount of
paraffin wax. In particular, a melting point equal to 110 ℃ paraffin wax pellets, inserting the material into a suitable
(see Figure 5) was considered to ensure that all crystals mould, heating to a predetermined temperature, stirring the
were molten. As reference temperature, the highest avail- molten liquid and finally slowly cooling it down to room
able experimental value, i.e., 200 ℃ was used. The value of temperature.
viscosity at 280 ℃, calculated using the ABC method, is The first experiments were performed using an aluminum
equal to 0.3747  10-3 Pa  s. cylindrical vessel of the size and shape corresponding to
configuration 1 in Table 4. Pellets were heated to 80 ℃
using an electric oven under high vacuum conditions, in
3.5. Grain manufacturing order to eventually de-gas any entrapped air. This tempera-
ture was selected by the DSC scan (see Figure 5) corre-
Several configurations of solid paraffin wax grains were sponding to the temperature at which the majority of
fabricated having a cylindrical shape with a coaxial central hydrocarbon chains melt. The liquid paraffin wax was
port, as indicated in the technical drawing shown in slowly cooled down at about 1 ℃/min to room temperature.
Figure 15. The low cooling rate should have reduced the formation of
Samples of different sizes and aspect ratios were manu- internal and surface cracks, and voids during solidification.
factured. The height, external diameter and internal port In this first experiment, no external device for applying a
diameter of every configuration are listed in Table 4. For the mechanical pressure on the molten wax during transition
228 F. Piscitelli et al.

From these experiments it is possible to deduce, that

80 ℃ is not sufficiently high to realize homogenous fuel
grains. As shown in Figures 3 and 5, 80 ℃ corresponds to
the temperature at which the majority of linear paraffin
chains melt. This result suggests that higher temperatures
are needed in order to partially melt also the branched
fractions of paraffin. On the other hand, the presence of
internal flaws can be attributed to the spontaneous, volu-
metric shrinkage occurring during cooling, and transition
Figure 26 Fracture surfaces of paraffin wax solid grain samples from liquid to solid state.
manufactured with the die casting method. To improve the quality of the manufactured solid grains,
further tests were carried out using a very similar method,
but heating the pellets of SASOLs 0907 up to 110 ℃
inside a polyamide mould, of the size and shape corre-
sponding to the configuration 2 in Table 4. This temperature
was chosen because, as suggested by DSC analysis
illustrated in Figure 5, this is the temperature at which all
the paraffin wax crystals melt. A polyamide mould was
selected to increase both thermal insulation and detaching
capacity of the solid grains. Photographs of the paraffin wax
solid grain obtained in this way are shown in Figure 20. The
results of the microscopic analysis are shown in Figure 21,
Figure 22, and in Figure 23.
Figure 27 Optical microscope image of paraffin wax solid grain The photographs demonstrate that the temperature of
samples manufactured with the die casting method. 110 ℃ is adequate to melt all crystals, but an internal, large
area of macro-porosity was detected, again attributed to the
spontaneous volumetric shrinkage. Although the very slow
cooling was aided by the polyamide mould, it was not
sufficient to manufacture homogeneous grains. Indeed, the
shrinkage needs to be adequately compensated by an
externally applied pressure.
To improve the relatively poor quality of paraffin wax
solid grains manufactured with the laboratory method, a
second, alternative, more sophisticated procedure was
investigated and experimented, based on the simultaneous
application of thermal and mechanical energies, i.e., heat
flux and pressure.
The starting point was the raw material in the form of
Figure 28 Optical microscope image obtained in bright field using a commercially available pellets of SASOLs 0907 paraffin
magnification of 550x of paraffin wax solid grain samples manufac- wax. The pellets were introduced inside a cylindrical
tured with the die casting method. mould, equipped with an embedded electrical resistance
heating system and temperature measurements through a K
from liquid to solid state was used and a very slow cooling type thermocouple. The paraffin wax was heated at 110 ℃
rate was the only method employed to decrease the thermal for 20 min until it was completely melted. The molten wax
gradients inside the bulk material. These cylindrical sam- was stirred, then the electrical resistance heating system was
ples were separated along the longitudinal axis and the switched off and simultaneously a pressure of about 10 bar
resulting fracture surfaces were microscopically analyzed in was applied by means of a cylindrical piston, fitted with a
order to investigate the possible presence of internal micro- load cell for monitoring the exerted compressive force.
cracks and micro-voids. Finally, the material was slowly cooled down to room
Figure 16 illustrates an image of fracture surfaces of one temperature.
sample manufactured according to this method, whereas This manufacturing apparatus was accomplished by the
Figures 17, 18, and 19 illustrate the most important features use of specially designed equipment, which is illustrated in
revealed by microscopic analysis. It highlights the edges are Figure 24. A load of 300 kg was set. Considering the surface
homogeneous, but the core region shows traces of pellets, area of the melted paraffin wax, this value is equivalent to a
which were only partially liquefied, and large and numerous pressure of about 10 bar, kept constant by means of the
internal micro-voids and micro-cracks. hydraulic system. The tight coupling tolerances among the
Characterization and manufacturing of a paraffin wax 229

mould components allowed the expulsion of air entrapped Acknowledgements

inside the paraffin wax during the heating and stirring steps.
Using this method homogeneous, consistent, solid par- This work was carried out within the HYPROB-NEW
affin grains corresponding to the configuration 3 in Table 4 project, funded by the Italian Ministry of University and
were fabricated (Figure 25). In addition, in this case, the Research (MIUR) whose financial support was much
internal microstructure of the samples were microscopically appreciated.
analyzed and the resulting images are shown in Figure 26, The authors gratefully thank Dr. A. French and
Figure 27 and in Figure 28. Mr. M. Fragiacomo for their very precious support to
It is worth noting that using this method traces of pellets this work.
which were not completely liquefied, and surface and
internal micro-cracks, and micro-porosities of critical size References
were not detected.
The lesson learned from all the experiments reported in [1] M.A. Karabeyouglu, D. Altman, B.J. Cantwell, Combustion
the present investigation is that simultaneous application of of liquefying hybrid propellants, Part 1: general theory,
external mechanical pressure of about 10 bar and of thermal J. Propuls. Power 18 (3) (2002) 610–620.
loads corresponding to a bulk temperature of at least 110 [2] M.A. Karabeyouglu, D. Altman, B.J. Cantwell, Combustion
℃, are necessary to avoid the abovementioned manufactur- of liquefying hybrid propellants, Part 2: stability of liquid
films, J. Propuls. Power 18 (3) (2002) 621–630.
ing problems, i.e., incomplete melting of the pellets and the
[3] C. Dunn, G. Gustafson, J. Edwards, T. Dunbrack, C.
presence of micro-porosities.
Johansen, Spatially and temporally resolved regression rate
measurements for the combustion of paraffin wax for hybrid
rocket motor applications, Aerosp. Sci. Technol. 72 (2018)
371–379.
4. Conclusions [4] E.T. Jens, B.J. Cantwell, G.S. Hubbard, Hybrid rocket
propulsion systems for outer planet exploration missions,
A commercial paraffin wax, i.e., the SASOLs 0907, Acta Astronaut. 128 (2016) 119–130.
possessing a high amount of branched alkanes, was [5] C. Boros, P. Konecny, Development of wax fuel grain for
selected for its capacity to form micro-crystals, which hybrid rocket motor, Adv. Mil. Technol. 4 (2) (2009) 5–11.
allow high fracture toughness and workability. A wide [6] J. De Sain, B.B. Brady, K.M. Metzler, T.J. Curtiss, T.V.
Albright, Tensile tests of paraffin wax for hybrid rocket fuel
thermal and rheological characterization was carried out
grains, in: Proceedings of the 45th AIAA/ASME/SAE/ASEE
supporting the manufacturing process. The thermal ana-
Joint Propulsion Conference & Exhibit, Denver, Colorado,
lyses, i.e., DSC and TGA, gave information about the U.S.A., 2–5 August 2009, AIAA 2009–5115.
phase transitions and thermal and thermo-oxidative sta- [7] A. Karabeyouglu, G. Zilliac, B.J. Cantwell, S. De Zilwa, P.
bilities. The results show that at atmospheric pressure and Castellucci, Scale-up tests of high regression rate paraffin-
110 ℃ the paraffin pellets are completely melted, and based hybrid rocket fuels, J. Propuls. Power 20 (6) (2004)
that the thermal treatment used to manufacture the fuel 1037–1045.
grain once melted essentially does not influence the [8] D. Altman, Hybrid rocket development history, in: Proceed-
crystals structure. Additionally, the formation of ashes ings of 27th Joint Propulsion Conference, 1991, pp. 1–5.
during the thermo-oxidative degradation of paraffin wax [9] Y. Tang, S. Chen, W. Zhang, R. Shen, L.T. DeLuca, Y. Ye,
has to be taken into account for the re-ignition tests of the Mechanical modifications of paraffin-based fuels and the
effects on combustion performance, Propellants Explos.
fuel grain. Rheological characterizations gave informa-
Pyrotech. 42 (11) (2017) 1268–1277.
tion on the storage and loss moduli, and viscosity of both
[10] Y. Pal, V.R. Kumar, Physical and ballistic characterization of
solid and liquid paraffin wax. In particular, the viscosity aluminum-loaded paraffin hybrid rocket fuels, Energy Fuels
value at high temperature, i.e., 280 ℃, was assessed 31 (2017) 10133–10143.
using an extrapolation method, in order to establish the [11] S. Kim, H. Moon, J. Kim, J. Cho, Evaluation of paraffin-
value to be introduced in the mass flow rate equation. The polyethylene blends as novel solid fuel for hybrid rockets,
wide characterization was useful to address the satisfac- J. Propuls. Power 31 (2015) 1750–1760.
tory manufacture of the fuel grain. Several manufacturing [12] Y. Pal, V.R. Kumar, Thermal decomposition study of
methods were investigated and it was found that the use paraffin-based hybrid rocket fuel containing aluminum and
of a heated circular mould-piston system hinders the boron additives, Thermochim. Acta. 655 (2017) 63–75, http:
formation of voids and critical defects. This tool consists //dx.doi.org/10.1016/j.tca.2017.06.002.
[13] S. Maruyama, T. Ishiguro, K. Shinohara, I. Nakagawa, Study
of a cylindrical metallic box, heated through electrical
on mechanical characteristics of paraffin-based fuel, in:
resistance and applying pressure via an oleo-dynamic
Proceedings of 47th AIAA/ASME/SAE/ASEE Joint Propul-
piston as the paraffin pellets melt. Preliminary firing tests sion Conference Exhibition, American Institute of Aeronau-
demonstrated the effectiveness of this manufacturing tics and Astronautics, 2011.
methodology. The next step of the authors will be the [14] R. Kumar, P.A. Ramakrishna, Studies on EVA-based wax
scaling-up of the fuel grain sizes and eventually the fuel for launch vehicle applications, Propellants Explos.
addition of nanofillers. Pyrotech. 41 (2016) 295–303.
230 F. Piscitelli et al.

[15] G. Ranuzzi, D. Cardillo, M. Invigorito, Numerical investiga- [23] R.A. Gater, M.R.L. L’Ecuyer, A fundamental investigation of
tion of a N2O-paraffin hybrid rocket engine combusting the phenomena that characterize liquid film cooling, Int.
flowfield, in: Proceedings of 6th European Conference for J. Heat. Mass Transf. 13 (3) (1970) 1925–1939.
Aeronautics and Space Sciences - EUCASS, Krakow, [24] R. Nigmatulin, B. Nigmatulin, Y.A. Khodzaev, V. Kroshilin,
Poland, 29 June–2 July 2015, pp. 1–15. Entrainment and deposition rates in a dispersed-film flow,
[16] G. Bernard, Development of a Hybrid Sounding Rocket Motor Int. J. Multiph. Flow. 22 (1) (1996) 19–30.
(Master thesis), College of Agriculture, Science and Engineer- [25] M. Boiocchi, E. Toson, S. Gelosa, A. Sliepcevich, L.
ing, KwaZulu-Natal Univ, Durban, South Africa, 2013. Galfetti, L. De Luca, Pure paraffin waxes analysis for hybrid
[17] N. Gerbal, G. Durand, M. Avril, L. Lefranc, C. Vautier, B. rocket solid fuels: rheological, thermal and mechanical
Chergui, M. Diop, G. Feroldi, M. Deroo, J. Diakuma, P. Le characterization, in: Proceedings of 5th EUCASS, European
Moguedec, Q. Tahan, A. Ansart, T. Diallo, G. Sire, G. Conference for Aeronautics and Space Sciences, Munich,
Poirey, Formulation of a mechanically improved paraffin Germany, 1–5 July, 2013.
fuel, in: Proceedings of the 5th EUCASS European Con- [26] Y. Pal, J. Baskar, S. Murugesan, Testing of paraffin-based
hybrid rocket fuel using gaseous oxygen oxidiser, Def. Sci.
ference for Aeronautics and Space Sciences, Munich, Ger-
J. 62 (2012) 277–283.
many, 1–5 July 2013, pp. 1–12.
[27] M. Kobald, E. Toson, H. Ciezki, S. Schlechtriem, S. di Betta,
[18] D.M. Arnold, Formulation and Characterization of Paraffin-
M. Coppola, L. De Luca, Rheological, optical and ballistic
based Solid Fuels Containing Swirl Inducing Grain Geome-
investigations of Paraffin-based fuels for hybrid rocket
try And/or Energetic Additives (Master Thesis), Mechanical
propulsion using a 2D slab-burner, in: Proceedings of 5th
Engineering, Dept., Mechanical and Nuclear Engineering, EUCASS European Conference for Aeronautics and Space
Pennsylvania State Univ., State College, PA, U.S.A, 2014. Science, Munich, Germany, 1–5 July, 2013.
[19] D.K. Owens, R.C. Wendt, Estimation of the surface free [28] J.J. Marano, G.D. Holder, A general equation for correlating
energy of polymers, J. Appl. Polym. Sci. 13 (1968) the thermophysical properties of n-paraffins, n-olefins, and
1741. other homologous Series. 3. Asymptotic behavior correla-
[20] M. Kobald, E. Toson, H. Ciezki, S. Schlechtriem, S. di Betta, tions for thermal and transport properties, Ind. Eng. Chem.
M. Coppola, L. De Luca, Rheological, optical, and ballistic Res. 36 (1997) 2399–2408.
investigations of paraffin-based fuels for hybrid rocket [29] J.J. Marano, G.D. Holder, General equation for correlating
propulsion using a two-dimensional slab burner, Prog. the thermophysical properties of n-paraffins, n-olefins, and
Propuls. Phys. 8 (2016) 263–282. other homologous Series. 1. Formalism for developing
[21] M. Zenkiewicz, Methods for the calculation of surface free asymptotic behavior correlations, Ind. Eng. Chem. Res. 36
energy of solids, Polym. Test. 26 (2007) 14–19. (1997) 1887–1894.
[22] A. Rudawska, E. Jacniacka, Analysis of determining surface [30] A. Kreglewski, B. Zwolinski, A new relation for physical
free energy uncertainty with the Owens-Wendt method, Int. properties of n-alkanes and n-alkyl compounds, J. Phys.
J. Adhes. Adhes. 29 (2009) 451–457. Chem. 65 (1961) 1050.