Copyright ©JCPDS-International Centre for Diffraction Data 2012 ISSN 1097-0002
MICROSTRUCTURE OF THE PLASTIC BONDED EXPLOSIVE KS32
1
M. Herrmann1, P. B. Kempa1, U. Förter-Barth1, W. Arnold2
Fraunhofer Institute for Chemical Technology ICT, Pfinztal, Germany
2
MBDA-TDW, Schrobenhausen, Germany
ABSTRACT
Plastic bonded explosives are highly filled polymers containing up to 90% high energy
crystalline solid loads. Different non-destructive diffraction techniques such as rocking curve
imaging were applied in order to characterize the microstructure and ingredients of PBX KS32,
including its behavior in shock-loading tests. The investigations revealed a significant change in
the amorphous binder, as well as damages to the embedded coarse HMX crystals during shockloading.
INTRODUCTION
Plastic-bonded explosives (PBX) are highly-filled polymers containing up to 90% high energy
crystalline solid load. The high loads are reached using bi- or poly-modal particle size
distributions, in which smaller particles fill the gaps between the larger ones. In recent years, the
microstructure of the solid ingredients came under the scrutiny of research when it was
discovered that incorporating meticulously recrystallized particles reduces the shock sensitivity
of PBXs. In such a context, factors such as particle size, shape, surface morphology, voids,
inclusions, impurities, dislocations and twinning have been discussed to influence the
mechanical sensitivity and the creation of hot spots during shock-loading. The crystallite size and
microstrain of high explosive powders were investigated by means of laboratory and synchrotron
X-ray diffraction (Herrmann et al., 2005a, 2005b, 2009a, 2009b). The investigations not only
revealed anisotropic diffraction peak broadening, but also correlated size/strain broadening with
particle processing and shock sensitivities.
For applications in penetrators, insensitive high explosives (IHE) are required in order to survive
heavy mechanical loading, e.g. during supersonic penetration of concrete or rock (Arnold , 2005a,
65
This document was presented at the Denver X-ray
Conference (DXC) on Applications of X-ray Analysis.
Sponsored by the International Centre for Diffraction Data (ICDD).
This document is provided by ICDD in cooperation with
the authors and presenters of the DXC for the express
purpose of educating the scientific community.
All copyrights for the document are retained by ICDD.
Usage is restricted for the purposes of education and
scientific research.
DXC Website
– www.dxcicdd.com
ICDD Website
- www.icdd.com
Copyright ©JCPDS-International Centre for Diffraction Data 2012 ISSN 1097-0002
2005b). Thus investigations were started with the aim to show how the microstructure of the
embedded crystals in the PBX KS32 is affected by shock-loading. However, the simultaneous
characterization of the ingredients in such a composite remains a challenging task, as it includes
a coarse and a fine crystal fraction of the same species, as well as the amorphous binder.
EXPERIMENTAL
KS32 consists of 85% HMX (C4H8N8O8), a crystalline high explosive, and 15% cured hydroxyterminated polybutadiene (HTPB) as the binder, with a total density amounting to 1.6 g/cm2. The
PBX was shock-loaded at MBDA-TDW in Schrobenhausen, Germany, and two loaded samples
were investigated at Fraunhofer ICT together with a pristine reference sample. Each sample was
measured twice on different surfaces. Figure 1 shows a cut sample as used for the measurements.
In order to collect separate information on each ingredient or individual fractions, different XRD
measuring and evaluation techniques were applied. Measurements were performed on a BraggBrentano Diffractometer D8 Advance from Bruker AXS, equipped with a copper tube, secondary
monochromator, scintillation counter and sample changer.
Figure 1. PBX-sample as measured. Cut surface dimensions are approximately 1 cm × 1 cm.
66
Copyright ©JCPDS-International Centre for Diffraction Data 2012 ISSN 1097-0002
An initial view of the crystalline fractions (HMX) was obtained from 2è scans evaluated by
whole-pattern fitting included in the program system TOPAS from Bruker AXS. The evaluation
was based upon the crystal structure data of HMX reported by Choi and Boutin (1970). The
evaluation yielded the unit-cell parameters, crystal density and the mean crystallite size
Cry size L (Balzar, 1999; Bruker, 2008).
The microstructure of the coarse fractions was probed by rocking curves combined with a
statistical evaluation (Herrmann et al., 2005b, 2007, 2008). The detector was therefore fixed at
selected peak positions and the samples were rocked around their symmetric orientations with a
step width of 0.01°ù and a measuring time of 5 s. The rocking curve patterns were decomposed
by means of peak fit using symmetrical Pearson VII functions, and the cumulative numbers of
peaks were then plotted against the peak widths of the rocking curve profiles. The run of the
curves and the median peak widths (X50) were interpreted as a quantity related to the internal
state of the crystals.
The amorphous parts were probed by peak fitting 2è scans, including the run of the base line of
the patterns. The amount of the amorphous content Camorphous was estimated from the relation of
the areas A under the halos and the crystalline peaks as related in the following equation:
Camorphous  Ahalos /( Apeaks  Ahalos )  100 .
RESULTS
The results of the measurements are summarized in Table 1 together with standard deviations
estimated from the four measurements of the loaded samples.
Fig. 2 shows 2è scan sections of two pristine KS32 samples. It should be noted that high
intensity fluctuations and jagged profiles occur due to the relatively coarse HMX crystals and
their poor orientation statistics, hindering a proper evaluation by whole-pattern fitting. Thus, the
results of the whole-pattern fitting yield only an initial assessment of the microstructure.
67
68
Copyright ©JCPDS-International Centre for Diffraction Data 2012 ISSN 1097-0002
Table 1: Results of diffraction experiments on two loaded KS32 samples compared to a pristine
reference. Values of two measurements for each sample are below, along with the average values
 and estimated standard deviations .
KS32 sample
Camorph
[%]
Cry size L
[nm]
Density
[g/cm3]
RC X50
[°ù]
17.9 / 15.2
16.6
122.8 /114
118.4
1.897 / 1894
1.896
0.085

91.8 / 97.6
77.1 / 91.6
84.4
8.7
1.894 / 1.89
1.892 / 1.892
1.892
0.0016
0.148
0.127


8.3 / 8.6
9.1 /10.8
9.2
1.1
pristine
loaded I
loaded II
1700
1600
1500
Lin (Counts)
1400
1300
1200
1100
1000
900
800
700
600
500
400
300
200
100
0
20.2
21
22
23
24
25
26
27
28
29
30
31
32
33
2-Theta - Scale
Figure 2. Sections of the 2è scans of two pristine KS32 samples. The patterns show high
intensity fluctuations and jagged peak profiles.
The whole-pattern fitting yielded densities of the incorporated HMX of 1.896 g/cm3 for the
pristine sample and a slightly smaller density of 1.892 g/cm3 for the loaded samples. The values
are well-distributed around the literature value of 1.894 with an estimated standard deviation of
69
Copyright ©JCPDS-International Centre for Diffraction Data 2012 ISSN 1097-0002
0.002 g/cm3, an insignificant difference well within experimental error range. The apparent
crystallite size Cry size L reduces from an average value of 118.4 to 84.4 nm upon loading,
which is considered significant, as the size had shifted more than three times the standard
deviation value of 9 nm.
Lin (Counts)
140
130
120
110
100
90
80
70
60
50
40
30
20
10
0
11
20
30
2-Theta - Scale
Figure 3. Magnification of the 2è scan background of a pristine KS32 sample (black) and two
loaded samples (blue and red). A halo occurs around 19 °2è, but with reduced intensity in the
Lin (Counts)
loaded samples.
7000
6000
5000
4000
3000
2000
1000
0
10
20
30
2-Theta - Scale
40
70
Copyright ©JCPDS-International Centre for Diffraction Data 2012 ISSN 1097-0002
Figure 4. Reference measurement of the binder HTPB (liquid).
An interesting detail was found by magnifying the background. A halo occurs around 19 °2è
(Fig. 3). The reference pattern in Fig. 4 shows that the halo belongs to the binder HTPB. After
loading of KS32, the intensity of the halo decreased significantly. The concentration of
amorphous HTPB, calculated from the areas under halos and peaks, reduced from 16.6 to 9.2%.
The binder concentrations of the pristine samples are acceptably close to the expected value of
15% of KS32, even without calibration. The decrease of the amorphous content of the binder is
not yet understood, but one possible explanation is the partial crystallization of the binder during
shock-loading.
54
52
50
48
46
44
42
40
38
36
34
32
30
28
26
24
22
20
18
16
14
12
10
8
6
4
2
0
-2
-4
12.6
12.8
13
13.2
13.4
13.6
13.8
14
14.2
14.4
14.6
14.8
Omega-Scale
Figure 5. Sections of a rocking curve of KS32 and pattern decomposition.
15
71
Copyright ©JCPDS-International Centre for Diffraction Data 2012 ISSN 1097-0002
Figure 5 shows a section of a rocking curve of a KS32 sample and its pattern decomposition
using symmetrical Pearson VII profiles. As the x-ray source and the detector are fixed in a
reflection position during the rocking curve measurements and only the samples are tilted, large
crystallites randomly come into reflection like differently oriented single crystals. Thus, up to
420 peak widths (full widths at half maximum) per sample were obtained and evaluated
statistically. The cumulative number of peaks was plotted versus peak width (Figure 6).
The plots show significant differences. The curves of the loaded samples are drawn to higher
peak widths indicating damage to the coarse HMX crystals after shock-loading. A quantification
of these changes is given by the median peak width X50 which is shifted from 0.085° for the
pristine samples to 0.148 and 0.127° for the loaded samples I and II, respectively. As the coarse
crystalline fraction is considered to dominate the sensitivity of a PBX, the rocking curve method
reveals crucial information on the microstructure of KS32.
normalized cumulative number of peaks
1.0
0.5
PBX KS32
pristine
loaded I
loaded II
0.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
FWHM [°Omega]
Figure 6. Cumulative number of rocking curve peaks as a function of the peak width.
Copyright ©JCPDS-International Centre for Diffraction Data 2012 ISSN 1097-0002
CONCLUSION
Various non-destructive diffraction techniques were applied in order to characterize the
microstructure of the PBX KS32 and its behavior upon shock-loading. The investigations
revealed a significant change in the amorphous binder concentration and a consequent damage to
the embedded coarse HMX crystals. Further investigations are in progress with a focus on
advanced sample preparation, a differentiation of defect types such as twinning and fracture, the
structural changes of the binder HTPB, and a separate characterization of the fine HMX fraction
in the PBX.
REFERENCES
Arnold, W. (2005a). “Significant Parameters Influencing the Shock Sensitivity, Part I:
Formulation and Binder,” 36th Int. Annu. Conf. of ICT, pp. 1–6
Arnold, W. (2005b). “Significant Parameters Influencing the Shock Sensitivity, Part II: Voids
and Defects,” 36th Int. Annu. Conf. of ICT, 188, pp. 1–14
Balzar, D. (1999). “Voigt Function Model in Diffraction-Line Broadening Analysis,” in
Microstructure Analysis from Diffraction, edited by Snyder, R.L., Bunge, H.J. and Fiala J.,
(Oxford University Press)
Bruker AXS (2008). “DIFFRACplus TOPAS, TOPAS 4 User Manual,” Karlsruhe, Germany.
Choi, C. S. and Boutin, H. P. (1970). “A Study of the Crystal structure of â-Cyclotetramethylene
Tetranitramine by Neutron diffraction”, Acta Cryst. B26, pp. 1235–1240.
Herrmann, M. and Fietzek, H. (2005a) “Investigation of the Microstructure of Energetic Crystals
by means of X-ray Powder Diffraction,” Powder Diffraction 20 (2), pp.105–108.
Herrmann, M. (2005b). “Microstructure of Energetic Particles Investigated by X-ray Powder
Diffraction,” Part. Part. Syst. Charact. 22, pp. 401–406.
Herrmann, M.; Kempa P. B.; Doyle S. (2007). “Microstructure of energetic Crystals – grain by
grain via rocking curve,” Z. Kristallogr. Suppl. 26, pp. 557–562.
Herrmann, M.; Förter-Barth, U.; Mathieu, J.; Patscheider, J. (2008).”Investigation of RS-RDX
Samples from the MSIAC-Round Robin by Means of X-ray Diffraction Rocking Curves,” 39th
Int. Annu. Conf. of ICT, 57, pp. 1–6
Herrmann, M.; Förter-Barth, U.; Kempa, P. B. (2009a). “Size/Strain Diffraction Peak
Broadening of the Energetic Materials FOX-7, RDX and ADN.” Central European Journal of
Energetic Materials (CEJEM) 6 (2), pp. 183–193.
Herrmann, M.; Förter-Barth, U.; Kempa P. B. (2009b). “Size/strain diffraction peak broadening
of the energetic materials HMX, CL-20 and FOX-12,” 40th Int. Annu. Conf. of ICT 32, pp. 1–11
72