Session D3: Grain Scale to Continuum Modeling I: Coarse Graining

DPDE Particle Method as a Generic Tool to Simulate the Mesoscale Response of HE Composites

Further developments and simulation results are presented that take the DPDE method closer to becoming a generic multi-scale computational method for the simulation of the shock response of HE at micron scales. In our initial application, the Hardy's averaging method yielded an in situ density in the shock state dependent on the particle size and inter-particle separation. The method was augmented to retrieve a density independent of these two variables. Second, the impactor-sample was modeled as a monolith with no interfacial separation. This was relaxed by introducing a contact algorithm to impose impenetrability and surface friction conditions. Simulation results show that the DPDE method predicts the quasi-static response and 1D transient heat conduction in agreement with the analytical solution. The simulated shock response of RDX is in reasonable agreement with shock propagation theory with contact interactions and separation at the impactor-sample interface. It is concluded that the DPDE method, as envisioned, may provide a unified multi-scale computational framework with inherent heat transport solution to simulate the shock response of HE that is independent of the particle size and inter-particle distance.   

Coarse-Grain Modeling of Energetic Materials

Mechanical and thermal loading of energetic materials can incite responses over a wide range of spatial and temporal scales due to inherent nano- and microscale features. Many energy transfer processes within these materials are atomistically governed, yet the material response is manifested at the micro- and mesoscale. The existing state-of-the-art computational methods include continuum level approaches that rely on idealized field-based formulations that are empirically based. Our goal is to bridge the spatial and temporal modeling regimes while ensuring multiscale consistency. However, significant technical challenges exist, including that the multiscale methods linking the atomistic and microscales for molecular crystals are immature or nonexistent. To begin addressing these challenges, we have implemented a \textit{bottom-up} approach for deriving microscale \quad coarse-grain models directly from quantum mechanics-derived atomistic models. In this talk, a suite of computational tools is described for particle-based microscale simulations of the nonequilibrium response of energetic solids. Our approach builds upon recent advances both in generating coarse-grain models under high strains and in developing a variant of dissipative particle dynamics that includes chemical reactions.   

A Molecular Dynamics simulation of Hugoniot curves of HMX using ReaxFF and its application in SPH modeling of macroscale terminal effects

HMX is a widely used high explosive. Hugoniot curve is a valuable tool for analyzing the equations of state, and is of importance for all energetic materials including HMX. The Hugoniot curves serve as one of the key character in continuum modeling of high explosives. It can be obtained from experimental measurements, and recently also from computational studies. In this study, the Hugoniot curve of HMX is calculated using a multi-scale shock technique via Molecular Dynamics (MD) simulations, where the reactive force field ReaxFF is obtained from Quantum Mechanics calculations and tailored for HMX. It is found that our MD Hugoniot curve of HMX from the optimized ReaxFF potential agree well with experiments. The MD Hugoniot curve of HMX is also incorporated in our in-house Smoothed Particle Hydrodynamics (SPH) code for the modeling of the macro-scale explosive behaviors of HMX explosives and HMX cased in a 3D cylinder.   

Hierarchical Multi-Scale Framework for Materials Modeling: Equation of State Implementation and Application to a Taylor Anvil Impact Test of RDX

In order to progress towards a materials-by-design capability, we present work on a challenge in continuum-scale modeling: the direct incorporation of complex physical processes in the constitutive evaluation. In this work, we use an adaptive hierarchical multi-scale (HMS) framework running in parallel on a heterogeneous computational environment to couple a fine-scale, particle-based model computing the equation of state (EOS) to the constitutive response in a finite-element multi-physics simulation. The EOS is obtained from high-fidelity materials simulations performed via dissipative particle dynamics methods. This HMS framework is progress towards an innovation infrastructure that will be of great utility for systems in which essential aspects of the material response are too complex to capture by closed form material models. The design, implementation, and performance of the HMS framework are discussed. Also presented is a proof-of-concept Taylor anvil impact test of non-reacting 1,3,5-trinitroperhydro-1,3,5-triazine (RDX).