基于深度学习势HMX炸药的燃面推移机制研究

… is validated via neural network potential-enabled molecular … ,3,5,7-tetrazocane (HMX), are considered here to evaluate the … and surface temperatures support the EM-HyChem model. …

Molecular simulations of high energetic materials (HEMs) are limited by efficiency and accuracy. Recently, neural network potential (NNP) models have achieved molecular simulations of millions of atoms while maintaining the accuracy of density functional theory (DFT) levels. Herein, an NNP model covering typical HEMs containing C, H, N, and O elements is developed. The mechanical and decomposition properties of 1,3,5-trinitroperhydro-1,3,5-triazine (RDX), hexahydro-1,3,5-trinitro-1,3,5-triazine (HMX), and 2,4,6,8,10,12-hexanitrohexaazaisowurtzitane (CL-20) are determined by employing the molecular dynamics (MD) simulations based on the NNP model. The calculated results show that the mechanical properties of α-RDX, β-HMX, and ε-CL-20 agree with previous experiments and theoretical results, including cell parameters, equations of state, and elastic constants. In the thermal decomposition simulations, it is also found that the initial decomposition reactions of the three crystals are N-NO2 homolysis, corresponding radical intermediates formation, and NO2-induced reactions. This decomposition trajectory is mainly divided into two stages separating from the peak of NO2: pyrolysis and oxidation. Overall, the NNP model for C/H/N/O elements in this work is an alternative reactive force field for RDX, HMX, and CL-20 HEMs, and it opens up new potential for future kinetic study of nitramine explosives.

The discovery and optimization of high-energy materials (HEMs) face challenges due to the computational expense and slow iteration of traditional methods. Neural network potentials (NNPs) have emerged as an efficient alternative to first-principles simulations. This study presents EMFF-2025, a general NNP model for C, H, N, and O-based HEMs, leveraging transfer learning with minimal data from DFT calculations. The model achieves DFT-level accuracy, predicting the structure, mechanical properties, and decomposition characteristics of 20 HEMs. Integrating EMFF-2025 with PCA and correlation heatmaps, we map the chemical space and structural evolution of these HEMs across temperatures. Surprisingly, EMFF-2025 uncovers that most HEMs follow similar high-temperature decomposition mechanisms, challenging the conventional view of material-specific behavior. EMFF-2025 offers a versatile computational framework for accelerating HEM design and optimization.

… sample the electronic free energy surface, we require the total … we trained HMX- or DAAF-specific ML potentials (panel b). … sample size (ESS), as a function of burn-in time choice (t 0 …

Reactive force fields for molecular dynamics have enabled a wide range of studies in numerous material classes. These force fields are computationally inexpensive compared with electronic structure calculations and allow for simulations of millions of atoms. However, the accuracy of traditional force fields is limited by their functional forms, preventing continual refinement and improvement. Therefore, we develop a neural network-based reactive interatomic potential for the prediction of the mechanical, thermal, and chemical responses of energetic materials at extreme conditions. The training set is expanded in an automatic iterative approach and consists of various CHNO materials and their reactions under ambient and shock-loading conditions. This new potential shows improved accuracy over the current state-of-the-art force fields for a wide range of properties such as detonation performance, decomposition product formation, and vibrational spectra under ambient and shock-loading conditions.

The response of high-energy-density materials to thermal or mechanical insults involves coupled thermal, mechanical, and chemical processes with disparate temporal and spatial scales that no single model can capture. Therefore, we developed a multiscale model for 1,3,5-trinitro-1,3,5-triazinane, RDX, where a continuum description is informed by reactive and nonreactive molecular dynamics (MD) simulations to describe chemical reactions and thermal transport. Reactive MD simulations under homogeneous isothermal and adiabatic conditions are used to develop a reduced-order chemical kinetics model. Coarse graining is done using unsupervised learning via non-negative matrix factorization. Importantly, the components resulting from the analysis can be interpreted as reactants, intermediates, and products, which allows us to write kinetics equations for their evolution. The kinetics parameters are obtained from isothermal MD simulations over a wide temperature range, 1200-3000 K, and the heat evolved is calibrated from adiabatic simulations. We validate the continuum model against MD simulations by comparing the evolution of a cylindrical hotspot 10 nm in diameter. We find excellent agreement in the time evolution of the hotspot temperature fields both in cases where quenching is observed and at higher temperatures for which the hotspot transitions into a deflagration wave. The validated continuum model is then used to assess the criticality of hotspots involving scales beyond the reach of atomistic simulations that are relevant to detonation initiation.

Reactive force fields have enabled an atomic level description of a wide range of phenomena, from chemistry at extreme conditions to the operation of electrochemical devices and catalysis. While significant insight and semi-quantitative understanding have been drawn from such work, the accuracy of reactive force fields limits quantitative predictions. We developed a neural network reactive force field (NNRF) for CHNO systems to describe the decomposition and reaction of the high-energy nitramine 1,3,5-trinitroperhydro-1,3,5-triazine (RDX). NNRF was trained using energies and forces of a total of 3100 molecules (11,941 geometries) and 15 condensed matter systems (32,973 geometries) obtained from density functional theory calculations with semi-empirical corrections to dispersion interactions. The training set is generated via a semi-automated iterative procedure that enables refinement of the NNRF until a desired accuracy is attained. The root mean square (RMS) error of NNRF on a testing set of configurations describing the reaction of RDX is one order of magnitude lower than current state of the art potentials.

… , and reactive and quantum molecular dynamics simulations, can … We have used feed-forward neural network with three hidden … dynamics simulations from neural networks and deep …

Detonation of energetic materials (EMs) is of great importance for military applications, while the understanding of detailed events and mechanisms for the detonation process is scarce. In this study, the first deep neural network potential NNP_Shock for molecular dynamics (MD) simulation of shock-induced detonation of EMs was generated based on a deep potential model, providing DFT accuracy but 106 times the computational efficiency. On this basis, we employ our deep potential to perform MD simulations of shock-induced detonation of high-performance EM material 2,4,6,8,10,12-hexanitro-2,4,6,8,10,12-hexaazaisowurtzitane (CL-20, C6H6N12O12) and obtain the theoretical Chapman-Jouguet (C-J) detonation velocities and pressures directly by multiscale shock technique (MSST) for the first time, which are in good agreement with experiment. In addition, the Hugoniot curves and initial reaction mechanisms were successfully obtained. Therefore, the NNP_Shock potential is competent in research of the detonation performance and shock sensitivity of CL-20, and the method can also be transplanted to studies of other EMs.

… of HMX based on the combined thermogravimetric (TG) measurements and chemical reaction neural network (… The potential extension of CRNN to kinetic modeling of other energetic …

… the dynamic behavior of propellant combustion (HMX/GAP). … deep neural network surrogate model was also developed to predict propellant combustion characteristics, such as burning …

This study presents NeuroFire, a neural network-based chemical solver framework for multi-dimensional combustion modeling of solid composite propellants. The framework integrates …

… combustion characteristics of nitrogen-containing hazardous chemicals, such as Hexogen (RDX) and Octogen (HMX), … to predict the behavior of intermediate combustion products. Key …

… V burn (conductive burn rate for HMX 18). Features associated with hot spot mechanisms are used in an attempt to infuse the neural network … rates are possible using neural networks. …

… combustion involves intricate chain reaction networks that demand comprehensive understanding for practical applications. In this study, we train a high-fidelity neural network potential (…

We present a complete set of isothermal third‐ and fourth‐order elastic coefficients, as well as higher‐order thermal stress coefficients, for the monoclinic molecular crystal β‐1,3,5,7‐tetranitro‐1,3,5,7‐tetrazocane (β‐HMX) in the P21/n space group setting at 300 K and atmospheric pressure. These higher‐order coefficients are obtained from a neural network model. Leveraging the spectral bias property of neural networks, the model is trained directly on temporally fluctuating stress data homogenized from isothermal molecular dynamics (MD) simulations of the mechanical response of single‐crystal samples during imposed isothermal strain to failure. Once trained, the model accurately reproduces not only the stress history of the samples but also their second‐order tangent stiffness tensor, specific heat, and Grüneisen parameter, all while satisfying the stress‐free and energy normalization conditions associated with a single reference configuration. We further employ a higher‐order Sobolev norm to enforce a specific heat consistent with MD and explore the ability of several different activation functions to produce smooth, positive definite elastic moduli, which, in turn, enable the model to perform well in predicting the adiabatic response for unseen, nonisothermal MD loading paths, despite being trained solely on isothermal data acquired at various discrete temperatures. Although this expressive neural network model is suitable for direct use in high‐fidelity hydrocodes, we provide an analytical polynomial approximation, constructed via Taylor expansion, that respects the monoclinic symmetry of β‐HMX, is practically as accurate as the learned model, and is guaranteed to be faster for inference.

Abstract Solid propellant burning rate, gas phase temperature, and condensed phase thickness depend on combustion chamber pressure, laser intensity, and propellant compositions. The ignition and combustion of solid propellants occur in three phases namely solid phase, condensed phase, and gas phase. In this study, moving boundary modeling was applied to each of the phases by coordinate transformation. This research includes modeling and dynamic simulation of the ignition and combustion of HMX/GAP, a high-energy material in the ratio of 8:2, with the gas phase of the combustion model consisting of 50 species and 234 reactions. The mathematical modeling used mass, energy, and momentum conservation equations, as well as constitutive equations for the moving boundary. Parametric studies were run under different operating conditions, with an initial temperature of 300 K, a pressure of 10–100 atm and a laser intensity of 100 W/cm2. A burning rate of 2.2 cm/s and a gas phase temperature of 2700 K were obtained under an operating pressure of 100 atm. Extinguishment of the solid propellant was rigorously analyzed in terms of dynamic simulation with various depressurization rates during combustion. This was carried out by depressurizing the solid propellant from 70 atm to 40 atm. Important factors of the extinguishment were discussed based on the mathematical model and various depressurization rates with parametric studies. At a depressurization rate of −8000 atm/s, the solid propellant was fully extinguished. From this study, one can identify the phenomenon for the extinguishment of HMX/GAP propellant using fast depressurization with rigorous mathematical model used for ignition and combustion.

… inside HMX or embedded on the surface of AP particles. The decomposition of AP and HMX … This paper provides new insight into the regulation of combustion performance of solid …

… systems, we conducted machine learning potential-based molecular dynamics simulations … production, partially compromising the clean combustion advantage intrinsic to nitrogen-rich …

A novel neural network potential (NNP) model is developed to analyze the thermal decomposition mechanism in multi-density Cyclotrimethylenetrinitramine (RDX). The model …

Chemical kinetic mechanisms are crucial for modeling the combustion processes of solid propellants, but the specific impacts of these mechanism’s parameters on combustion have not been fully assessed. This study conducted a comprehensive sensitivity analysis on kinetics, thermodynamics, and transport parameters affecting solid propellant mechanisms, exemplified with HMX as a case study. A onedimensional steady-state numerical model incorporating gas and liquid phase mechanisms of HMX was developed and validated against experimental data. This model enables a thorough sensitivity analysis to evaluate the influence of various parameters, including the reaction constant (k) of each elementary reaction, enthalpy of formation (hf), entropy (s), heat capacity (cp), collision diameter (σ), and potential well-depth (ε) of each species, on key combustion characteristics over a wide range of pressure. The analysis revealed that gas kinetics predominantly govern the HMX combustion compared to the liquid kinetics, particularly at high pressures. Notably, the decomposition reactions of H2CNNO2 and N2O in the gas phase were identified as highly sensitive reactions that control the r and the pressure exponent of HMX. By calculating the normalized sensitivity coefficients of all parameters, the cp values of small gaseous molecules were found to be the most significant factors affecting combustion, indicating a role played by the thermodynamic properties of small species. This research could enhance our understanding of HMX combustion mechanisms and underscore critical areas for future development and refinement of detailed kinetic mechanisms of solid propellants.

… In this study, a quantum chemistry-level neural network potential (NNP) was developed to simulate the thermal decomposition of TKX-50. Throughout the simulation, we tracked the …

… force fields and the prohibitive computational cost of density functional theory (DFT), this study developed a machine learning force field (ML-FF) for b-HMX … of b-HMX with accuracy …

… articles and news from researchers in related subjects, suggested using machine learning. … /lg force field is highly appropriate for the reactive molecular dynamics of HMX and HTPB. …

… a machine learning force field (ML-FF) for β-HMX using a high-… outperforms traditional empirical force fields, achieving near-DFT … Our study reveals that the β-HMX crystal exhibits strong …

… using machine learning. … force field was chosen as COMPASS force field [27, 28], mainly because that both of CL-20 and HMX belonged to nitramine explosive, and COMPASS force …

We present a data-driven blended equation of state (EOS) approach for condensed phase high explosive materials. We first calibrate four different high explosive materials (Nitromethane, HMX, PETN and TATB) using a single or blending multiple Fried Howard Gibbs (FHG) EOS by an ad hoc trial and error method that has been used in the past, and which leads to a predictive model that can be used in engineering calculations. This ad-hoc calibration is then re-calibrated based on Bayesian optimisation via Gaussian Process regression. The two calibrations are then compared qualitatively and quantitatively and are shown to be in good to excellent agreement.

2,6-Diamino-3,5-dinitropyrazine-1-oxide (LLM-105) is a relatively new and promising insensitive high-explosive (IHE) material that remains only partially characterized. IHEs are of interest for a range of applications and from a fundamental science standpoint, as the root causes behind insensitivity are poorly understood. We adopt a multitheory approach based on reactive molecular dynamic simulations performed with density functional theory, density functional tight-binding, and reactive force fields to characterize the reaction pathways, product speciation, reaction kinetics, and detonation performance of LLM-105. We compare and contrast these predictions to 1,3,5-triamino-2,4,6-trinitrobenzene (TATB), a prototypical IHE, and 1,3,5,7-tetranitro-1,3,5,7-tetrazoctane (HMX), a more sensitive and higher performance material. The combination of different predictive models allows access to processes operative on progressively longer timescales while providing benchmarks for assessing uncertainties in the predictions. We find that the early reaction pathways of LLM-105 decomposition are extremely similar to TATB; they involve intra- and intermolecular hydrogen transfer. Additionally, the detonation performance of LLM-105 falls between that of TATB and HMX. We find agreement between predictive models for first-step reaction pathways but significant differences in final product formations. Predictions of detonation performance result in a wide range of values, and one-step kinetic parameters show the similar reaction rates at high temperatures for three out of four models considered.

Abstract The energy release of aluminized explosives was impeded by the aggregation of aluminum powders, especially for nano-aluminum powders. The core-shell HMX@(Al@GAP) explosives with 5 wt.% to 15 wt.% aluminum were fabricated, with systematical investigation of morphology, mechanical properties, thermal decomposition, combustion and detonation properties. Compared with the physical mixtures, the aluminized explosives with core-shell microstructure showed better creep resistance and mechanical strength. The thermal decomposition of HMX in core-shell HMX@(Al@GAP)was slightly advanced due to good dispersion of Al@GAP particles. Delayed but more vigorous and longer burning during combustion process was witnessed for the core-shell explosive. For HMX@15wt%(Al@GAP) aluminized explosives, the detonation velocity and specific kinetic energy was 8567 m/s and 1.412 kJ/g, which was 1.3% and 5.1% higher than the corresponding physical mixtures. The reaction of the core-shell HMX@(Al@GAP) and the physical mixtures were calculated by ReacFF-/g force field, showing remarkable dependent of the average temperature on the microstructure. Present research might provide new insight for synergistically improvement of mechanical and detonation properties in aluminized explosives.

… The reaction force field that appropriates for the system is the most crucial part of reaction … , 35] have shown that the ReaxFF/lg force field is optimal for HMX and poly-NIMMO. Therefore, …

… Consequently, we engaged the ReaxFF/lg force field modality to delve into the thermal … , the deflagration of DNAN's nitro groups initiates at 102 ps, succeeded by the deflagration of DNB…

In this study, a 3D-DEM was employed to model an HMX explosive containing a large void oriented at 45°. The effects of hammer height, particle size, and void size on HMX detonation were systematically investigated. The detonation was observed at hammer drop heights of 0.1 m and 0.3 m, indicating that moderate impact energy enhances stress concentration, thereby promoting hot spot formation. Particles with a size of 150 μm were found to be the only ones capable of triggering detonation, indicating that larger particles improve stress concentration and energy transfer. Detonation could be achieved with void sizes of 64 and 100 discrete elements, as larger voids increase stress concentration and lower the detonation threshold. The reaction time of the detonation aligns with experimental results. However, the detonation pressure was found to be lower than experimental values, which is attributed to the void defect in the HMX explosive model. Overall, these findings demonstrate that an optimal void size and moderate striker height can maximize stress concentration and energy focusing, resulting in a rapid local temperature rise and the propagation of detonation waves.

Molecular dynamics (MD) simulations are currently widely used to study large-scale displacement cascades based on massive simulation trajectories. However, when the irradiation process involves the complex chemical reactions, effectively analyzing and extracting features from these data becomes a challenge. Here, we introduce a new cross-platform toolkit, HlightReaxMD, designed to directly obtain information about the irradiation damage process and chemical reactions from MD trajectories and further achieve the prediction of irradiation damage. The analysis tools in HlightReaxMD include chemical reaction analysis and calculation of reaction kinetic parameters, analysis of the collision cascade process, and calculation of necessary physicochemical properties. HlightReaxMD supports the analysis of all elements used in reactive force fields by reading ReaxFF potential file parameters and provides an automated solution for tracking atomic-scale collision events and analyzing chemical reaction mechanisms through constructing cascade trees and reaction network paths. A machine learning-driven model using the analysis results has been included in HlightReaxMD, which can predict irradiation damage by considering various factors, rather than relying solely on the Norgett-Robinson-Torrens displacements per atom (NRT-dpa) model. It enables researchers to automatically obtain dynamic processes and reaction information from atomic to microscale defects from terabyte-level trajectory data. Thus, HlightReaxMD can promote systematic research on irradiation effects in materials science.

The reactive molecular dynamics using ReaxFF provides an effective means to generate global reactions for pyrolysis of realistic fuel mixtures. The reactions from large-scale pyrolysis simulations of a fuel mixture may be characterized by multiple reaction sites, explosion of intermediate species structures, and scattered contribution of diversified pathways to product species. This work proposes an approach of SRG-Reax aiming at generating skeleton reaction networks based on reaction patterns or classes of reaction centers from huge reactions obtained from ReaxFF MD simulations of realistic fuel pyrolysis. SRG-Reax (Skeleton Reaction network Generation for ReaxFF MD) is implemented through building a semi-supervised machine learning model of tri-training for predicting the reaction classes of pyrolysis reactions based on an extended reaction center. Three different reaction center descriptions of reaction features and reaction transformation fingerprints are employed as inputs for developing the tri-training classifier. Major reaction pathways can be identified based on reaction class ratios and product species ratios calculated by merging reaction pathways of the same reaction class. The SRG-Reax approach was applied in skeleton reaction network generation for RP-3 pyrolysis based on the ReaxFF MD simulations of a high-fidelity 45-component RP-3 fuel model. The skeleton reaction networks for n-paraffins, iso-paraffins, cycloparaffins, olefins, and aromatics in RP-3 pyrolysis were obtained. The reaction class ratios and product species ratios in the obtained skeleton reaction network provide comprehensive intuitive insight into global pyrolysis chemistry. SRG-Reax has the potential to obtain relatively complete skeleton reaction networks for the pyrolysis of hydrocarbon fuel, polymers, biomass, coal, and more.

… SMX 2- ), universal machine-learning force field (MLFF) models … , a series of machine-learning-based molecular dynamics (… This study overcomes the limitations of molecular dynamics …

… a molecular model of kerogen from Huadian oil shale. Subsequently, reactive molecular dynamics … For these structures, we utilized deep learning combined with quantum chemistry …

… Reactive molecular dynamics … HMX nanoparticle (NP) and core–shell structured HMX@Al NP. The results indicate that the decomposition of HMX@Al NP is earlier than that of the HMX …

With ultrashort duration and ultrahigh energy, femtosecond laser (fs-laser) pulses are very promising for the precision machining of energetic materials. Compared with the mechanical machining meth...

In this work, we report the discovery of energy cocrystals using an efficient iterative workflow combining an evolutionary algorithm and a machine learning potential (MLP). The compound 2,4,6,8,10,12-Hexanitro-2,4,6,8,10,12-hexaazaisowurtzitane (CL-20) has attracted significant attention owing to its higher energy density than traditional energetic materials. However, the higher sensitivity has limited its applications. An important way to reduce its sensitivity involves cocrystal engineering with traditional explosives. Many cocrystal structures are expected to be composed of these two components. We developed an efficient iterative workflow to explore the phase space of CL-20 and 1,3,5,7-tetranitro-1,3,5,7-tetraazacyclooctane (HMX) cocrystals using an evolutionary algorithm and an MLP. The algorithm was based on the Universal Structure Predictor: Evolutionary Xtallography (USPEX) software, and the MLP was the reactive force field with neural networks (ReaxFF-nn ) model. A set of high-density cocrystal structures was produced through this workflow; these structures were further checked via first-principles geometry optimizations. After careful screening, we identified several high-density cocrystal structures with densities of up to 1.937 g/cm3 and HMX: CL-20 ratios of 1:1 and 1:2. The influence of hydrogen bonds on the formation of high-density cocrystals was also discussed, and a roughly linear relationship was found between energy and density.

… use of AIMD in complex reactive system with large amount of … learning algorithm [24] and the dynamic machine learning … tensors) computed by machine learning AIMD, and then uses …

Energetic materials undergo chemical bond cleavage and structural reorganization under external stimuli such as heat and irradiation, leading to performance degradation or change. This study employs molecular dynamics simulations to investigate the atomic-scale mechanisms of irradiation-induced damage in β-HMX and its subsequent thermal decomposition mechanism. Results revealed that irradiation defects are primarily localized within collision cascade regions. Irradiation-generated nitrogen oxides and HNO x radicals provide new reaction sites during HMX thermal decomposition, significantly altering pyrolysis pathways and increasing the activation energy. Concurrently, irradiation-induced defects restructure the reaction network, modifying the propagation efficiency of radical chain reactions. This work provides atomic-level insights into the microscopic mechanism through which irradiation influences the thermal decomposition of HMX, offering a theoretical foundation for understanding the chemical evolution and stability of energetic materials under extreme conditions.

Accurate description of detonation performance for explosives remains a challenge for current experimental and theoretical methodologies. Herein, we address this issue through combining a multi-scale shock technique and a first-principles based deep neural network potential. This approach enables us to conduct molecular dynamics simulations encompassing over a thousand atoms and extending for several nanoseconds, allowing us to evaluate the detonation performance of the insensitive explosive NTO crystal. Utilizing the ZND model, we successfully determine the detonation velocity (7.9 km s-1), and detonation pressure (33 GPa) of the NTO crystals at the C-J state, which align well with experimental results. Additionally, we predict the detonation performance of three host-guest materials: NTO/H2O2, NTO/CO2, and NTO/N2O, all of which exhibit higher detonation temperatures compared to the NTO crystals in the present model. Moreover, we proposed the time to reach the C-J state as a shock sensitivity descriptor for explosives. Our findings reveal that the order of shock sensitivity for these materials is NTO/H2O2 > NTO/N2O > NTO/CO2 > NTO, and the trend can be explained in terms of bulk modulus, electronic band gap and oxygen balance. The enhanced shock sensitivity by embedded small molecules arises not only from the reduction in initial reaction barriers, but also from the faster evolution rate of final products and the release of more heat. Our research might present a cutting-edge framework for precisely, quickly, and safely evaluating and modulating the detonation performance and shock sensitivity of explosives.

… The recent advances in machine learning interatomic … prototypical crystalline explosives (CL-20, HMX, RDX, TNT, NTO, and … quantum-accurate reactive dynamics and the macroscopic …

We demonstrate the use of quantum molecular dynamics to identify the β- to δ-molecular structure transition in bulk-phase HMX, which has been considered as the primary reason for the increased sensitivity in the thermal decomposition of HMX. Both physical and chemical changes accompany this transition, but no previous study has shown conclusively which specific change, or set of changes, is responsible. We find that the initial decomposition mechanism of HMX can explain this sensitivity issue. Our DFT simulations of the periodic system followed by detailed finite cluster calculations of the transition states find two distinct initial unimolecular reaction pathways in β-HMX that operate simultaneously. (1) For the HONO release reaction, β-HMX first transformed to an intermediate, in which one parallel N–NO2 group transitions from chair to boat conformations with a low +1.2 kcal/mol barrier, followed by unimolecular HONO release (+42.8 kcal/mol barrier, rate-determining step). (2) For the NO2 cleavage react...

… to elucidate its liquid-phase and gas-phase decomposition … re-examined for liquid-phase decomposition, and early ring-opening … The proposed pathways for decomposition of HMX are …

A neural network potential (NNP) is developed to investigate the complex reaction dynamics of 1,3,5-trinitro-1,3,5-triazine (RDX) thermal decomposition. Our NNP model is proven to possess good computational efficiency and retain the ab initio accuracy, which allows the investigation of the entire decomposition process of bulk RDX crystals from an atomic perspective. A series of molecular dynamics (MD) simulations are performed on the NNP to calculate the physical and chemical properties of the RDX crystal. The results show that the NNP can accurately describe the physical properties of RDX crystals, such as the cell parameters and the equation of state. The simulations of RDX thermal decomposition reveal that the NNP could capture the evolution of species at ab initio accuracy. The complex reaction network was established, and a reaction mechanism of RDX decomposition was provided. The N-N homolysis is the dominant channel, which cannot be observed in previous DFT studies of isolated RDX molecule. In addition, the H abstraction reaction by NO2 is found to be the critical pathway for NO and H2O formation, while the HONO elimination is relatively weak. The NNP gives an atomic insight into the complex reaction dynamics of RDX and can be extended to investigate the reaction mechanism of novel energetic materials.

… –related processes in the front of the shock precursor wave … all the energy terms of the ReaxFF potential function, and E lg … their impact on the decomposition characteristics of HMX. The …

To elucidate the detailed thermal decomposition mechanism of the CL-20/DNT cocrystal material, a neural network potential (NNP) model was developed. This NNP model has been demonstrated to exhibit excellent computational efficiency while retaining the accuracy of ab initio. Molecular dynamics simulations over a wide temperature range were performed based on the trained NNP model. The results show that the intermolecular hydrogen bond interaction formed between DNT and CL-20 during the initial decomposition process and the DNT capture of the NO2 group released from CL-20 together enhance the stability of the cocrystal system. The detailed formation pathways for the major products, including N2, H2O, CO2, and HNCO, were obtained by product analysis. Furthermore, temperature was found to influence the quantity variation of CO2, and part of CO2 is converted to CO at high temperature. By analyzing the clusters, it is found that the C rings or C-N rings are more easily formed at the temperature of 2200 K, and some O atoms will combine with the clusters at the temperature of 3000 K. This study investigates the complex reaction dynamics of CL-20/DNT from an atomic perspective, and the methodology can be extended to a broader range of energetic material systems.

… , momentum, and energy across the shock front. Instead of simulating a shock wave within a … site was the most possible reaction pathway for the initial decomposition of HMX. And that …

… Figure 6 compares the potential energy change of the HMX/… also shows that HMX in the mixture will decompose at a faster … thermolysis of HMX at any stage, and the insensitivity of HMX/…

… to the decomposition of BC-HMX. Since the transformations in the combustion front commence at … [26] theoretically analyzed the possible primary reactions for BC-HMX using density …

This work investigated the thermal decomposition process of the HMX/DNAN melt-cast explosive system at high temperatures using reactive molecular dynamics. The initial reaction paths of the system and the effects of DNAN on the thermal decomposition of HMX were revealed. The findings indicated that the H atoms and nitro groups produced by the decomposition of DNAN were reacted with HMX and its intermediates, which in turn promoted the pyrolysis of HMX, while the OH radicals produced by the decomposition of HNO2 continued to induce the decomposition of HMX. Calculations of activation energies showed that the addition of DNAN enhanced the sensitivity of HMX to thermal stimuli. In terms of products, the addition of DNAN significantly impacted the final product H2O of HMX, while its impact on CO2 is relatively small. At 2000 K-3500 K, the maximum number of clusters in mixed system showed a general decreased trend with increased temperature.

… A new neural network potential (NNP) is constructed to reveal the thermal decomposition … , which allows the investigation of the entire decomposition process of NM from an atomic …

… on two fundamental aspects of thermal decomposition [4]: (1) a … For a smaller molecule, such as HMX, the infinite network is … 2B shows eight possible intermolecular attractions per HMX …

CL-20 (2,4,6,8,10,12-hexanitro-2,4,6,8,10,12-hexaazaisowurtzitane, also known as HNIW) is one of the most powerful energetic materials. However, its high sensitivity to environmental stimuli greatly reduces its safety and severely limits its application. In this work, ab initio based neural network potential (NNP) energy surfaces for both β-CL-20 and CL-20/TNT co-crystals were constructed. To accurately simulate the thermal decomposition processes of these two crystal systems, reactive molecular dynamics simulations based on the NNPs were performed. Many important intermediate species and their associated reaction paths during the decomposition had been identified in the simulations and the direct results on detonation temperatures of both systems were provided. The simulations also showed clearly that 2,4,6-trinitrotoluene (TNT) molecules in the co-crystal act as a buffer to slow down the chain reactions triggered by nitrogen dioxide and this effect is more significant at lower temperatures. Specifically, the addition of TNT molecules in the CL-20/TNT co-crystal introduces intermolecular hydrogen bonds between CL-20 and TNT molecules in the system, thereby increasing the thermal stability of the co-crystal. The current reactive molecular dynamics simulation is performed based on the NNP which helps in accelerating the speed of ab initio molecular dynamics (AIMD) simulation by more than 3 orders of magnitude while preserving the accuracy of density functional theory (DFT) calculations. This enabled us to perform longer-time simulations at more realistic temperatures that traditional AIMD methods cannot achieve. With the advantage of the NNP in its powerful fitting ability and transferability, the NNP-based MD simulation can be widely applied to energetic material systems.

Against the backdrop of insufficient research into the microscopic reaction mechanisms of pentazole anion ( N5−$$ {\mathrm{N}}_5^{-} $$ ) salts, the present study developed a deep neural network potential (DNNP) model calibrated with first principles data. On this basis, large‐scale molecular dynamics (MD) simulations were performed to conduct an in‐depth investigation into the thermal decomposition mechanism and kinetic processes of hydroxylamine pentazole (NH3OHN5) at the atomic scale. A highly precision DNNP model was constructed using an active learning strategy, whose predictions for energy and atomic forces showed excellent agreement with Density Functional Theory (DFT) results. MD simulations revealed that the thermal decomposition of NH3OHN5 initiates with a hydrogen transfer reaction. The protonation of the N5−$$ {\mathrm{N}}_5^{-} $$ reduces its ring‐opening energy barrier from 125.45 to 112.13 kJ/mol, significantly promoting the ring‐opening decomposition process. The final decomposition products were predominantly N2, H2O, and NH3. This research elucidates the decomposition pathways and reaction mechanism of NH3OHN5 at the atomic scale, demonstrating the exceptional capability of the DNNP in simulating the reaction dynamics of energetic materials and providing a theoretical foundation for the subsequent molecular design of high‐performance, green energetic materials.

A neural network potential (NNP) is developed to investigate the decomposition mechanism of RDX, AP, and their composites. Utilizing an ab initio dataset, the NNP is evaluated in terms of atomic energy and forces, demonstrating strong agreement with ab initio calculations. Numerical stability tests across a range of timesteps reveal excellent stability compared to the state-of-the-art ReaxFF models. Then the thermal decomposition of pure RDX, AP, and RDX/AP composites is performed using NNP to explore the coupling effect between RDX and AP. The results highlight a dual interaction between RDX and AP, i.e., AP accelerates RDX decomposition, particularly at low temperatures, and RDX promotes AP decomposition. Analyzing RDX trajectories at the RDX/AP interface unveils a three-part decomposition mechanism involving N-N bond cleavage, H transfer with AP to form Cl-containing acid, and chain-breaking reactions generating small molecules such as N2, CO, and CO2. The presence of AP enhances H transfer reactions, contributing to its role in promoting RDX decomposition. This work studies the reaction kinetics of RDX/AP composites from the atomic point of view, and can be widely used in the establishment of reaction kinetics models of composite systems with energetic materials.

As a nitrogen-rich metal-free energetic salt, crystalline N2H5N5 with a cyclopentazolate anion has been recently synthesized, but its decomposition mechanism is far from clear. In this study, a neural network potential (NNP) has been developed to investigate the complex reaction dynamics of N2H5N5. Large-scale ab initio-quality molecular-dynamics simulations driven by this model demonstrate that the NNP accurately captures the time-resolved emergence and disappearance of intermediate species in the reaction of the bulk crystal. The triggering event is synchronous: the N2H5+ cation initiates a step-wise dehydrogenation while the N5- ring undergoes either (i) direct cleavage or (ii) hydrogen-assisted opening. Once these two "ignition" steps have occurred, the subsequent chemistry is that successive H-stripping from N2H5+ and rapid rearrangement of the opened N5- chain lead almost exclusively to the formation of N2 and NH3 with only trace amounts of radicals. This simplicity stands in sharp contrast to CHON explosives such as TNT or RDX, where initiation is dominated by nitro-group scission and the ensuing NO2-driven redox cascade spawns a tangled network of hundreds of intermediates. Furthermore, we have calculated the apparent activation energy of the decomposition of N2H5+ and N5-, revealing that N2H5+ is consumed significantly faster than N5-. This work not only clarifies the decomposition mechanism of N2H5N5 but also provides a reliable approach for studying energetic ionic salts, contributing to a deeper understanding of their chemistry.

The temporal evolution of crucial products and their kinetics features are important for understanding the reaction behaviors of high explosives pyrolysis. We perform the large-scale and long-duration reactive force field...

… Typically, equations of state used for HMX in numerical simulations treat CV as constant … [51, 52]: eg one for solid HMX and a second for liquid HMX. This could be used to predict the …

We present a physics-informed machine learning framework for predicting shock initiation criteria for reactive metamaterials and study shock propagation through a one-dimensional laminate structure. The laminate material was composed of an HMX bed with equally distributed 2 mm thick copper pillars. The Wide-Ranging equation of state (WR EOS) was used to model HMX while the Romenski EOS was used for the elastic regime of copper, with the assumption of perfect plasticity. The shock was initiated by using an aluminum impactor and gauges were placed at the entry of the first copper pillar and exit of the last pillar. A modified machine learning model was then developed to predict the shock initiation criteria for the laminate structure. The proposed model only uses short-time measurements for predicting this behaviour, leading to large reductions in computational cost at higher dimensions. This framework suggests a data-driven guideline for the design of optimal laminate structures (e.g. number of copper pillars, thickness, and distribution).

… of HMX microstructures of interest in this work. In their approach, inert meso-scale simulations … 2.4, we present the methods for propagating the microstructural uncertainties to the macro-…

… ) simulations with machine learning (ML) to investigate crack nucleation, propagation, and coalescence in a prototypical HMX-… This study develops a multiscale modeling framework for …

… behaviour of HMX and ATC–HMX, as well as the effect of ATC on HMX decomposition, we … from HMX decomposition intensified the convective mechanism in flame propagation during …

Shock wave interactions with defects, such as pores, are known to play a key role in the chemical initiation of energetic materials. The shock response of hexanitrostilbene is studied through a combination of large scale reactive molecular dynamics and mesoscale hydrodynamic simulations. In order to extend our simulation capability at the mesoscale to include weak shock conditions (< 6 GPa), atomistic simulations of pore collapse are used to define a strain rate dependent strength model. Comparing these simulation methods allows us to impose physically-reasonable constraints on the mesoscale model parameters. In doing so, we have been able to study shock waves interacting with pores as a function of this viscoplastic material response. We find that the pore collapse behavior of weak shocks is characteristically different to that of strong shocks.

… The 1-D reactive Euler equations are updated to include multiscale state variables for HMX … Initial work into multiscale modeling was presented and showed that modeling the …

Shock initiation of plastic-bonded explosives (PBX) begins with the formation of so-called "hot spots", which are energetic reactions localized in regions where the PBX microstructure concentrates the input shock wave energy. We developed a model PBX system to study hot spots which consists of a single crystal of the high explosive HMX (cyclotetramethylene-tetranitramine) embedded in a transparent polyurethane binder (J. Phys Chem. A, 2020, 124, 4646-4653). In the current work we use this model system to study the influence of input shock pressure (12-26 GPa) on hot spot generation using micrometer-resolved high-speed imaging and nanosecond-resolved optical pyrometry. By shocking ∼100 HMX single crystals (HMX-SC), two distinct shock pressure thresholds were observed. The threshold for producing single hot spots in some crystals was 15 GPa. At 23 GPa, hot spot density was sufficiently high to lead to rapid deflagration of the entire HMX-SC. It takes about 25 ns after the shock passes for the hot spots to appear to our visible-light detection apparatus which has a noise floor at about 2000 K. That indicates the shock produces nascent hot spots that undergo a thermal explosion that reaches temperatures >2000 K in 25 ns. The initial hot spot temperature is roughly 3800 K which settles down to 3400 K, the adiabatic flame temperature of HMX. The higher initial temperature is attributed to release of stored interfacial strain energy produced by the shock. An initial estimate for the velocity of the flame front originating at an HMX hot spot is 550 m/s.

… Analysis revealed that recrystallized TATB and HMX were interconnected, forming micro-nanostructured microspheres. The enhanced interfacial contact and reaction area resulted in a …

Nanowear-resistant coatings are critical for extending the service life of mechanical components, yet their performance optimization remains challenging due to the complex interplay between atomic-scale defects and macroscopic wear behavior. While experimental characterization struggles to resolve transient interfacial phenomena, multiscale simulations, integrating ab initio calculations, molecular dynamics, and continuum mechanics, have emerged as a powerful tool to decode structure–property relationships. This review systematically compares mainstream computational methods and analyzes their coupling strategies. Through case studies on metal alloy nanocoatings, we demonstrate how machine learning-accelerated simulations enable the targeted design of layered architectures with 30% improved wear resistance. Finally, we propose a protocol combining high-throughput simulation and topology optimization to guide future coating development.

… mentioned before as possible ignition sites and the current simulations illustrate their … burning process into the "cold" regions of the HMX which would not be consistent with a burn rate …

… of these materials under combustion or detonation conditions. … calculated four possible decomposition pathways of the α-HMX … unit cell parameters and atomic positions of δ HMX. The …

… of HMX/GAP pseudo-propellant combustion has been established to predict the propellant burning rate and detailed combustion … The thermochemical parameters of HMX and GAP are …

… Ignition delay times and burn times of TKX-50 and HMX with different particle … simulations are performed to analyze the thermal decomposition of TKX-50 and HMX in terms of potential …

Abstract We have developed a continuum modeling approach, grounded in classical physical chemistry, based on the following assumptions: (1) that the states in the material can be represented by local stationary averages of the pressure (stress), temperature, and mass fractions computed from atomistic simulation, (2) and that the mixture has well-defined molecular components, each with a complete equation of state. The continuum model, “Gibbs formulation”, applies to near-atomic length and time scales, which we identify as the scales where the high frequency, high energy phonons equilibrate in molecular mixtures, (about six atomic radii and six to ten vibrational periods). Phase changes and chemical changes due to reaction are not in (asymptotically, long-time) equilibrium, and changes are assumed to occur on much longer time scales than those required for stress and temperature equilibration. Recently in the Journal of Chemical Physics, J. Chem. Phys. 144, 184111 (2016), we carried out both atomistic molecular dynamics (MD) simulations and “mirrored” continuum simulations to model, thermal ignition of a nano-sized cube of explosive RDX. The NVE ensemble simulations of a constant volume explosions of RDX were performed using reactive molecular dynamics (RMD), that use ReaxFF as model chemical changes in the MD simulation. The MD simulation was regarded as the exact molecular system. The continuum simulation was regarded as an interpretation and measurement of the average chemical changes between a set of identified chemical components of that molecular system. In this work, we extend these ideas to include spatial averaging to study wave propagation and spatially distributed transport, combined with chemical reaction and compare continuum based simulations with recent RMD simulations of a sustained spatially distributed deflagration in a nano-scale slab of HMX carried out by Joshi and Chaudhuri, similarly done in RDX. [3] Both atomistic and continuum simulations show a hot spot ignition followed by a structured deflagration that propagates through the HMX slab and are compared with good to excellent agreement.

… In this paper, the thermal decomposition process and cluster growth law of the HMX/LLM-… molar ratio of 1:1 was investigated. The results show that clusters can be formed between HMX …

… to obtain an improved model for combustion and detonation. … However, the potential energy decreases from the beginning … atomic insight into the whole process of thermolysis of HMX/…

Reactive molecular dynamics (RMDs) calculations were used to determine, for the first time, the process of thermolysis of the mixed explosives, including 3-nitro-1,2,4-triazol-5-one (NTO) and octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazoline (HMX). Significantly, this is the first time that a layered model for mixed explosives, which is an extreme innovation of mixed explosive models was adopted. It is shown that a large amount of NO2 in the HMX and OH groups generated by the decomposition of HNO2 has a favorable effect on the thermolysis of NTO, as further validated by a reduction in the activation energy of NTO/HMX. The amount of H2O and N2 in the resulting products increased significantly, but the amount of NH3 changed slightly. The analysis results correspond to the change in chemical bonds. Whenever there is an increase in temperature, the time for the maximum number of clusters to appear is shortened accordingly. In addition, the acidity of NTO has been considered. An independent gradient model based on Hirshfeld partition (IGMH) and atoms in molecule (AIM) analysis of NTO/HMX was implemented. The relatively strong hydrogen bonds indicate that HMX can inhibit the acidity of NTO and is beneficial for the wide application of NTO/HMX-based plastic-bonded explosives (PBXs).

… a systematic molecular dynamics simulation to study the reaction kinetics of HMX. A set of … HMX. The heats of formation, heats of reaction, and reaction rate parameters are obtained. …

… (MD) simulations have been performed to investigate AP/HMX (… bond lengths of HMX in the AP/HMX composite have been … The volume thermal expansion coefficient of the AP/HMX …

… The laser-induced ignition response of HMX has been investigated using a detailed … Ignition occurs in the gas phase and the flame propagates back toward the surface of the HMX in …

… For this exploratory study, we focus on analyzing the potential benefits in terms of ignition … Hotspot size–temperature threshold for plain HMX and HMX in the HMX/Al composite. The …

… We use molecular dynamics simulations with the reactive potential ReaxFF to investigate the initial reactions and subsequent decomposition in the high-energy-density material α-HMX …

… Temperature at ignition is found to be a function of heating rate. The ignition process for fast cook-off is consistent with laser ignition of bare HMX samples. Simulation results depend on …

Abstract In this work, we have for the first time, applied a rapid compression machine (RCM) to generate initial uniformly high temperature environment to investigate the auto-ignition behaviors of solid energetic material (HMX/NC/NG). Pressure evolution recording and high speed visualization were synchronized to reveal the response of energetic material to thermal stimulus in a time scale of the order of 100 ms. Results show that at sufficiently high end of compression (EOC) pressure and temperature, auto-ignition of solid energetic material is observed after certain period of induction time upon EOC. The ignition delay time defined by the first flame spot observation from high speed imaging (IDTI) is smaller than defined by the maximum pressure rise rate instant (IDTP). However, both IDTI and IDTP decrease with the increase of EOC pressure and temperature. In addition, the burning duration also decreases, indicating a faster burning rate. By tuning the EOC pressure and temperature, the critical thermodynamic condition that separates the auto-ignition region and the non-ignition regime is obtained. The present method of thermal stimulus generation is believed to provide a new approach for evaluating the thermal stability of energetic materials, beyond the literature on cook-off test approaches.

This study utilizes the discrete element method (DEM) to model the ignition and combustion of cyclotetramethylene-tetranitramine (HMX) particles under drop hammer impact. Key …

HMX/Al is an energetic composite formed during the manufacturing and processing of mixed explosives, and its safety has been explored under the action of force and electrostatic field. In order to investigate the impact response and ignition characteristics of HMX/Al energetic composites based on the force–electric coupling method, a self-designed force–electric coupling impact test device was adopted to carry out the impact sensitivity test of composite explosives under the influence of different factors. The model of HMX/Al energetic composites was established under the effect of force–electric coupling, and the critical ignition voltage and the response ignition mechanism were obtained under the conditions of the force–electric coupling test. The results show that under the condition of fixed drop height, the critical ignition voltage of the impact response of HMX/Al first decreases and then gradually increases with the increase of Al powder content; when fixing the content of Al powder, the greater the strength of the applied electric field, the more intense the impact ignition response of HMX/Al energetic composites; the greater the strength of the restraining shell, the more significant the degree of the ignition response of the HMX/Al energetic composites. The F2602 binder used for coating HMX has an inhibitory effect on the ignition response of explosives under the force–electric coupling test; the critical ignition voltage of composite powder is smaller than that of the pressed tablets with the same formula. Under the influence of external impact loads, hot spots form inside the explosives. Simultaneously, the Joule heating effect generated by an external power source continuously heats, causing heat accumulation and ultimately leading to an explosion. Therefore, HMX/Al energetic composites are more dangerous under force–electricity coupling, and the results of the study have a certain reference significance for the safety assessment of energetic materials under compound stimulations.

… the system has achieved a dynamic equilibrium situation. The average flame propagation velocity of the HMX dust cloud does not monotonically vary with the ignition energy, but there …

… the ignition of granular HMX and pressed HMX pelle.ts. Porosity appears to increase … the laser ignition of HMX are also investigated. Voids or cracks in a damaged HE can potentially be …