The Effects of TATB Ratchet Growth on PBX 9502

PBX 9502 is a plastic‐bonded explosive that contains 95 wt.‐% TATB, a graphitic‐structured high explosive known to undergo “ratchet growth,” i.e., irreversible volume change that accompanies temperature excursions. Earlier studies have reported changes in TATB‐based composites as a function of thermal cycling and density change, however, a clear distinction between density and ratchet‐growth effects has not been made. In the work reported here, an “as‐pressed density” baseline for the mechanical response of recycled PBX 9502 is established over a density range of interest, then high‐density specimens are thermally cycled between −55 and 80 °C to achieve “ratchet‐grown” parts in the same low‐density region. As‐pressed and ratchet‐grown specimens with identical densities are then analyzed using microX‐ray computed tomography and USANS techniques to obtain information about pore‐size distributions. Data show that after ratchet‐growth, PBX 9502 specimens contain, in general, more numerous and smaller voids than specimens that were pressed with lower compaction pressures to match the same density. The mechanical response of the ratchet‐grown material is consistent with damage, showing lower tensile stress and modulus, lower compressive modulus, and higher tensile and compressive strain, than as‐pressed specimens of the same density.     

Microstructural Differences between Virgin and Recycled Lots of PBX 9502

PBX 9502 is an insensitive high explosive formulated comprised of 95 wt% TATB and 5 wt% Kel‐F 800^TM binder. Due to the relatively high cost of manufacturing TATB (triaminotrinitrobenzene), methods for reclaiming TATB from PBX 9502 machine cuttings were previously developed. Reclaimed PBX 9502 was mixed with ~ 50% virgin PBX 9502 to produce “recycled” lots of PBX 9502. Several studies have shown significant differences between the mechanical properties of virgin and recycled lots of PBX 9502, and postulated that the differences were related to various aspects of TATB particle size and distribution. The purpose of this study is to show that these differences in mechanical properties are related to differences in the distribution of TATB within the microstructure of PBX 9502. Ultimately, a better understanding of these properties may lead to selected formulation changes for future rebuilds, Lifetime Extension Programs (LEP) and/or candidate replacements to enhance engineering and physics performance.     

Diameter Effect and Failure Diameter of a TATB-Based Explosive

The detonation velocity of PBX-9502, an explosive consisting of 95 wt% TATB and 5 wt% Kel-F 800, was measured precisely over a range of charge diameters at 75°C, 24°C, and −55 °C. The diameter-effect curves obtained by plotting detonation velocity versus the reciprocal of charge radius were found to differ from those reported in the literature for other solid and liquid explosives, being concave upward at large diameters. The curve at 75°C was found to be a straight line at small diameters and thus simulates the behavior of a homogeneous explosive. At intermediate charge diameters, the effect of varying the temperature by 130°C was quite small. The failure diameter varied from 5.85 ± 0.15 mm at 75°C to 10.5 mm at −55 °C.     

The Microstructure of TATB‐Based Explosive Formulations During Temperature Cycling Using Ultra‐Small‐Angle X‐Ray Scattering

TATB (1,3,5 triamino‐2,4,6‐trinitrobenzene), an extremely insensitive explosive, is used both in polymer‐bound explosives (PBXs) and as an ultra‐fine pressed powder (UFTATB). Many TATB‐based explosives, including LX‐17, a mixture of TATB and Kel‐F 800 binder, experience an irreversible expansion with temperature cycling known as ratchet growth. Additional voids, with sizes hundreds of nanometers to a few micrometers, account for much of the volume expansion. Measuring these voids is important feedback for hot‐spot theory and for determining the relationship between void size distributions and detonation properties. Also, understanding mechanisms for ratchet growth allows future choice of explosive/binder mixtures to minimize these types of changes, further extending PBX shelf life. This paper presents the void size distributions of LX‐17, UFTATB, and PBXs using commercially available Cytop M, Cytop A, and Hyflon AD60 binders during temperature cycling between −55 and 70 °C. These void size distributions are derived from ultra‐small‐angle X‐ray scattering (USAXS), a technique sensitive to structures from about 2 nm to about 2 μm. Structures with these sizes do not appreciably change in UFTATB. Compared to TATB/Kel‐F 800, Cytop M and Cytop A show relatively small increases in void volume from 0.9 to 1.3% and 0.6 to 1.1%, respectively, while Hyflon fails to prevent irreversible volume expansion (1.2–4.6%). Computational mesoscale models combined with experimental results indicate both high glass transition temperature as well as TATB binder adhesion and wetting are important to minimize ratchet growth.     

Dead Zones in LX‐17 and PBX 9502

Pin and X‐ray corner turning data have been taken on ambient LX‐17 and PBX 9052, and the results are listed in tables as an aid to future modeling. The results have been modeled at 4 zones/mm with a reactive flow approach that varies the burn rate as a function of pressure. A single rate format is used to simulate failure and detonation in different pressure regimes. A pressure cut‐off must also be reached to initiate the burn. Corner turning and failure are modeled using an intermediate pressure rate region, and detonation occurs at high pressure. The TATB booster is also modeled using reactive flow, and X‐ray tomography is used to partition the ram‐pressed hemisphere into five different density regions. The model reasonably fits the bare corner turning experiment but predicts a smaller dead zone with steel confinement, in contradiction with experiment. The same model also calculates the confined and unconfined cylinder detonation velocities and predicts the failure of the unconfined cylinder at 3.75 mm radius. The PBX 9502 shows a smaller dead zone than LX‐17. An old experiment that showed a large apparent dead zone in Composition B was repeated with X‐ray transmission and no dead zone was seen. This confirms the idea that a variable burn rate is the key to modeling. The model also produces initiation delays, which are shorter than those found in time‐to‐detonation.     

The Effect of Shear Strain on Texture in Pressed Plastic Bonded Explosives

The insensitive explosive PBX 9502 contains 95 wt‐% of TATB crystals and a plastic bonding agent (Kel‐F). The TATB crystals have plate‐like morphology, similar to that of graphite or boron nitride. We have used X‐ray diffraction to measure the preferred orientation (texture) of the TATB crystals in parts fabricated by pressing PBX 9502 powder. Independently, we have used finite‐element calculations to derive the direction and magnitude of the shear imposed during the consolidation of this composite material. Based on our results, we propose that the texture develops because the applied shear causes the TATB crystals to rotate such that their (002) basal planes are parallel to shear planes. The texture predicted by this model agrees qualitatively with that measured at various locations within the PBX 9502 compact. Further validation of this model is obtained by the measurement of the thermal expansion coefficient of PBX 9502, which is highly anisotropic.     

Ignition and Growth Modeling of LX‐17 Hockey Puck Experiments

Detonating solid plastic bonded explosives (PBX) formulated with the insensitive molecule triaminotrinitrobenzene (TATB) exhibit measurable reaction zone lengths, curved shock fronts, and regions of failing chemical reaction at abrupt changes in the charge geometry. A recent set of “hockey puck” experiments measured the breakout times of diverging detonation waves at ambient temperature LX‐17 (92.5% TATB plus 7.5% Kel‐F binder) and the breakout times at the lower surfaces of 15 mm thick LX‐17 discs placed below the detonator‐booster plane. The LX‐17 detonation waves in these discs grow outward from the initial wave leaving regions of unreacted or partially reacted TATB in the corners of these charges. This new experimental data is accurately simulated for the first time using the Ignition and Growth reactive flow model for LX‐17, which is normalized to detonation reaction zone, failure diameter and diverging detonation data. A pressure‐cubed dependence for the main growth of reaction rate yields excellent agreement with experiment, while a pressure‐squared rate diverges too quickly and a pressure‐quadrupled rate diverges too slowly into the LX‐17 below the booster equatorial plane.     

Initiation Pressure Thresholds from Three Sources

Pressure thresholds are minimum pressures needed to start explosive initiation that ends in detonation. We obtain pressure thresholds from three sources. Run‐to‐detonation times are the poorest source but the fitting of a function gives rough results. Flyer‐induced initiation gives the best results because the initial conditions are the best known. However, very thick flyers are needed to give the lowest, asymptotic pressure thresholds used in modern models and this kind of data is rarely available. Gap test data are in much larger supply but the various test sizes and materials are confusing. We find that explosive pressures are almost the same if the distance in the gap test spacers are in units of donor explosive radius. Calculated half‐width time pulses in the spacers may be used to create a pressure‐time curve similar to that of the flyers. The very‐large Eglin gap tests give asymptotic thresholds comparable to extrapolated flyer results. The three sources are assembled into a much‐expanded set of near‐asymptotic pressure thresholds. These thresholds vary greatly with density: for TATB/LX‐17/PBX 9502, we find values of 4.9 and 8.7 GPa at 1.80 and 1.90 g/cm³, respectively.     

Time‐Evolution of TATB‐Based Irreversible Thermal Expansion (Ratchet Growth)

TATB is an insensitive high explosive, attractive for use because of its safety aspects. TATB compactions, with or without binder, undergo irreversible volume expansion (or ratchet growth) upon thermal cycling. In the past, experimental elucidation of this phenomenon has focused on irreversible expansion as a function of the number of thermal excursions over a given temperature range, where growth is asymptotic with increasing cycle number. In this paper, we demonstrate that ratchet growth also occurs as a function of time at constant temperature, and that growth is substantial at elevated temperatures. We have measured strain response in PBX 9502, a TATB‐based composite, by performing thermal‐cycling tests with different durations at high temperature. Irreversible growth arises from the thermal ramps themselves (increasing and decreasing), as well as from the subsequent isotherms. PBX 9502 specimens with previously‐identified TATB texture/orientation were used in order to eliminate and/or evaluate texture as a variable. Measurements were also performed on dry‐pressed TATB (no binder) to confirm that expansion as a function of time (constant temperature) is not caused by the binder. A simple analysis of the time‐response data demonstrates consistency in the results. We propose that the primary driving force for irreversible expansion is the proximity of the current strain value (due to thermal history) to the strain saturation point of the current cycle (i.e. strain at infinite high‐temperature hold times or an infinite number of cycles). Such tests should aid in the understanding and modeling of ratchet growth response in these materials.     

Changes in Pore Size Distribution upon Thermal Cycling of TATB‐based Explosives Measured by Ultra‐Small Angle X‐Ray Scattering

Hot‐spot models of initiation and detonation show that voids or porosity ranging from nanometer to micrometer in size within highly insensitive energetic materials affect initiability and detonation properties. Thus, the knowledge of the void size distribution, and how it changes with the volume expansion seen with temperature cycling, are important to understanding the properties of the insensitive explosive 1,3,5‐triamino‐2,4,6‐trinitrobenzene (TATB). In this paper, void size distributions in the 2 nm to 2 μm regime, obtained from small‐angle X‐ray scattering measurements, are presented for LX‐17‐1, PBX‐9502, and ultra‐fine TATB formulations, both as processed and after thermal cycling. Two peaks were observed in the void size distribution: a narrow peak between 7–10 nm and a broad peak between 20 nm and about 1 mm. The first peak was attributed to porosity intrinsic to the TATB crystallites. The larger pores were believed to be intercrystalline, a result of incomplete consolidation during processing and pressing. After thermal cycling, these specimens showed an increase in both the number and size of these larger pores. These results illuminate the nature of the void distributions in these TATB‐based explosives from 2 nm to 2 μm and provide empirical experimental input for computational models of initiation and detonation.     

Dynamic mechanical signatures of a polyester‐urethane and plastic‐bonded explosives based on this polymer

The complex shear moduli of the segmented polyurethane Estane 5703p, Livermore explosive (LX)‐14, and plastic bonded explosive (PBX)‐9501, which use this polymer as a binder, have been investigated. Segmented polyurethanes, such as Estane 5703, contain microphase‐separated hard segments in a rubbery matrix of soft segments. LX‐14 is composed of 95.5% 1,3,5,7‐tetranitroazacyclooctane (HMX) explosive with 4.5% Estane 5703 binder. PBX‐9501 is composed of 94.9% HMX, 2.5% Estane 5703p binder, 2.5% nitroplasticizer (NP), and about 0.1% antioxidant Irganox 1010. In the temperature range from ?150 to 120°C, two relaxations were observed as peaks in the loss modulus and tangent delta in Estane 5703p and LX‐14. A third relaxation was found in PBX‐9501. The low temperature relaxation associated with vitrification of the poly(ester urethane) soft segment occurred in the shear loss modulus (G″) at ?29 and ?26°C in Estane and LX‐14, respectively, at 1 Hz. In PBX‐9501 the Estane soft segment glass transition peak, T_g(SS), in the loss modulus occurred at ?40 ± 3°C at 1 Hz. The reduction in soft segment glass transition in PBX‐9501 is clear evidence of plasticization of the soft segment by NP. The apparent activation energy of the maximum in the loss modulus for LX‐14 and PBX‐9501 over the frequency range from 0.1 to 10 Hz was 230 kJ/mole (55 kcal/mole). The hard segment glass transition, T_g(HS), was observed as a peak in the loss modulus at about 70°C. In LX‐14 the transition was observed at lower temperatures (56–58°C at 1 Hz) depending on thermal history. There was a low temperature shoulder on the T_g(HS) of Estane 5703 associated with soft segment crystallinity. Modulated differential scanning calorimetry (MDSC) was used to verify the T_g(HS) in Estane and 50/50 mixtures of Estane with NP. In PBX‐9501 the hard segment glass transition occurred between 65 and 72°C. The presence of NP in PBX‐9501 gave rise to a new transition, T_eu(NP), between 8 and 15°C. This peak is believed to be associated with the eutectic melting of the plasticizer. Returns of fielded PBX‐9501 that were 6 and 11 years old were also measured. Small variations in T_g(SS) and the rubber plateau modulus were observed in these aged samples, consistent with migration of plasticizer and/or very low levels of chain scission. © 2002 Wiley Periodicals, Inc. J Appl Polym Sci 83: 1009–1024, 2002     

Air Gaps,Size Effect,and Corner‐Turning in Ambient LX‐17

Various measurements under ambient conditions are presented for LX‐17. The size (diameter) effect has been measured with copper and Lucite confinement, where the failure radii are 4.0 and 6.5 mm, respectively. The air well corner‐turning has been measured with an LX‐07 booster, and the dead‐zone results are comparable to the previous TATB‐boosted work. Four double cylinders have been fired, and dead zones appear in all cases. The steel‐backed samples are faster than the Lucite‐backed samples by 0.6 μs. Bare LX‐07 and LX‐17 charges of 12.7 mm radius were fired with air gaps. Long acceptor regions were used to truly determine if detonation occurred or not. The LX‐07 booster crossed a 10 mm gap with a slight time delay. Steady‐state LX‐17 crossed a 3.5 mm gap but failed to cross a 4.0 mm gap. LX‐17 charge with a 12.7 mm radius run after the booster crossed a 1.5 mm gap but failed to cross a 2.5 mm gap. Timing delays were measured where the detonation crossed the gaps. The Tarantula model is introduced as embedded in reactive flow JWL++and Linked Cheetah V4, mostly at 4 zones mm⁻¹. Tarantula has four pressure regions: off, initiation, failure, and detonation. The physical basis of the input parameters is considered.     

Aspects of the Tribology of the Plastic Bonded Explosive LX‐04

The dynamic coefficient of friction, μ_d, of the plastic bonded explosive (PBX) LX‐04 was measured on stainless steel, aluminum, Teflon and the explosive itself as a function of temperature between ambient and 135 °C at a rotational speed of 0.0025 rad/s⁻¹. An optical profilometer was used to analyze surface roughness. LX‐04 is a composite of the explosive 1,3,5,7‐tetranitroazacyclooctane (HMX) and Viton A in an 85/15 weight ratio. For LX‐04 on stainless steel, μ_d decreased from 0.38 at ambient to 0.18 at 95 °C, then was nearly constant to about 125 °C, where the coefficient began to increase again. The opposite behavior was observed for aluminum. Against Teflon μ_d was nearly constant from ambient to 65 °C at 0.43, and then decreased to 0.17 from 100 °C to 135 °C. Against LX‐04 itself the coefficient of friction averaged 0.64 at temperatures between 35 °C and 95 °C, but tended to increase during the measurement, probably due to adhesion of the Viton to itself. Above 95 °C the coefficient dropped off and became nearly constant again at 0.16 up to 135 °C. Measurements on stainless steel with the mock explosive RM‐04‐BR, a composite of cyanuric acid and Viton A, and with the same weight ratio as the actual explosive, compared reasonably well with the explosive itself.     

Detonation in TATB Hemispheres

Streak camera breakout and Fabry‐Perot interferometer data have been taken on the outer surface of 1.80 g/cm³ TATB hemispherical boosters initiated by slapper detonators at three temperatures. The slapper causes breakout to occur at 54° at ambient temperatures and 42° at −54 °C, where the axis of rotation is 0°. The Fabry velocities may be associated with pressures, and these decrease for large timing delays in breakout seen at the colder temperatures. At room temperature, the Fabry pressures appear constant at all angles. Both fresh and decade‐old explosive are tested and no difference is seen. The problem has been modeled with reactive flow. Adjustment of the JWL for temperature makes little difference, but cooling to −54 °C decreases the rate constant by 1/6th. The problem was run both at constant density and with density differences using two different codes. The ambient code results show that a density difference is probably present, but it cannot be quantified.     

Thermal Damage on LX‐04 Mock Material and Gas Permeability Assessment

RM‐04‐BR, a mock material for the plastic‐bonded HMX‐based explosive LX‐04, is characterized after being thermally damaged at 140 °C and 190 °C. We measured the following material properties before and after the thermal experiments: sample volume, density, sound speed, and gas permeability in the material. Thermal treatment of the mock material leads to de‐coloring and insignificant weight loss. The sample expanded, resulting in density reductions of 1.0% to 2.5% at 140 °C and 190 °C, respectively. Permeability in the mock samples was found to increase from 10⁻¹⁶ to 10⁻¹⁵ m², as the porosity increased. The permeability measurements are well represented by the Blake‐Kozeny equation for laminar flow through porous media. The results are similar to the gas permeability in PBX‐9501 obtained by other researchers [1, 2].     

Growth of 1,3,5-Triamino-2,4,6,-Trinitrobenzene (TATB). II. Control of growth by use of high Tg polymeric binders

The remarkable safety characteristics of the high explosive 1,3,5-triamino-2,4,6-trinitrobenzene (TATB) are revolutionizing the design and deployment of nuclear weapons. Kel-F 800 is used as the binder to obtain high-density, mechanically stable billets of TATB that can be machined into desired shapes. However, repeated thermal cycling between − 54 °C and 74 °C of high density, pure, and plastic-bonded TATB billets causes a permanent volume expansion (growth) of about 1.5 vol% to 2.0 vol%. Debonding of the Kel-F 800 binder occurs during growth, causing a reduction in the mechanical properties of the plastic-bonded explosive. The coefficient of thermal expansion (CTE) of these TATB billets between ambient temperature and 74 °C is 67.0 × 10⁻⁶/°C. TATB undergoes a secondary mechanical relaxation just above room temperature, coinciding with the onset of a high CTE, above the glass transition temperature (T_g) of Kel-F 800. Thus, by judicious selection of a high-T_g binder, we have essentially eliminated growth, stopped the degradation of mechanical properties after thermal cycling, suppressed the secondary mechanical relaxation, and lowered the CTE to 50.0 × 10⁻⁶/°C between ambient temperature and 74 °C.     

Detonation Measurements on Damaged LX‐04

We have applied thermal insults on LX‐04 at 185 °C and found that the material expanded significantly, resulting in a bulk density reduction of 12%. Subsequent detonation experiments (three cylinder tests) were conducted on the thermally damaged LX‐04 samples and pristine low‐density LX‐04 samples and the results showed that the fractions reacted were close to 1.0. The thermally damaged LX‐04 and pristine low‐density LX‐04 showed detonation velocities of 7.7–7.8 mm μs⁻¹, significantly lower than that (8.5 mm μs⁻¹) of pristine high‐density LX‐04. Detonation energy densities for the damaged LX‐04, low‐density pristine LX‐04, and hot cylinder shot of LX‐04 were 6.48, 6.62, and 6.58 kJ cm⁻³, respectively, lower than the detonation energy density of 8.11 kJ cm⁻³ for the high density pristine LX‐04. The break‐out curves for the detonation fronts showed that the damaged LX‐04 had longer edge lags than the high density pristine LX‐04, indicating that the damaged explosive is less ideal.     

LX‐17 Corner‐Turning

We have performed a series of highly‐instrumented experiments examining corner‐turning of detonation. A TATB booster is inset 15 mm into LX‐17 (92.5% TATB, 7.5% kel‐F) so that the detonation must turn a right angle around an air well. An optical pin located at the edge of the TATB gives the start time of the corner‐turn. The breakout time on the side and back edges is measured with streak cameras. Three high‐resolution X‐ray images were taken on each experiment to examine the details of the detonation. We have concluded that the detonation cannot turn the corner and subsequently fails, but the shock wave continues to propagate in the unreacted explosive, leaving behind a dead zone. The detonation front farther out from the corner slowly turns and eventually reaches the air well edge 180° from its original direction. The dead zone is stable and persists 7.7 μs after the corner‐turn, although it has drifted into the original air well area. Our regular reactive flow computer models sometimes show temporary failure but they recover quickly and are unable to model the dead zones. We present a failure model that cuts off the reaction rate below certain detonation velocities and reproduces the qualitative features of the corner‐turning failure.     

Tailoring the irreversible thermal expansion of 1,3,5-triamino-2,4,6-trinitrobenzene crystals by bioinspired polydopamine coating

Bioinspired polydopamine (PDA) was selected for the fabrication of 1,3,5-triamino-2,4,6-trinitrobenzene (TATB)-based microcapsules to improve the thermal stability via a facile in situ polymerization of dopamine on the surface of explosive crystals in a weak alkaline aqueous solution. The effects of experimental conditions, including TATB core size, PDA shell content, elevated-temperature hold time, hold temperature, and test stress on the irreversible thermal expansion of TATB crystals, were comprehensively and comparatively studied. After coating, the strain change at each cooling and heating stage in a thermal cycling test from −54 to 74 °C visibly decreased, attributing to the fact that the highly crosslinked and dense PDA shell acted as a rigid pressure vessel to constrain the expansion of energetic crystals. The irreversible expansion strain at room temperature after a 23–113 °C cycle decreased with the increasing of PDA shell thickness. Compared with raw fine grains TATB (FTATB) crystals, FTATB enabled in 1.5 wt % PDA showed a dramatically drop in the irreversible expansion strain at room temperature by 27.7% (from 0.520 to 0.376%). The results demonstrated the excellent ability of PDA to alleviate irreversible thermal expansion of anisotropic energetic crystals. © 2019 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2019 , 137, 48695.