Computational modeling of energetic materials under impact and shock compression
thesisposted on 22.11.2021, 18:01 authored by Camilo Alberto Duarte CordonCamilo Alberto Duarte Cordon
Understanding the fundamental physics involved in the high strain rate deformation of high explosives (HE) is critical for developing more efficient, reliable, and safer energetic materials. When HE are impacted at high velocities, several thermo-mechanical processes are activated, which are responsible for the ignition of these materials. These processes occur at different time and length scales, some of them inaccessible by experimentation. Therefore, computational modeling is an excellent alternative to study multiscale phenomena responsible for the ignition and initiation of HE. This thesis aims to develop a continuum model of HMX to study the anisotropic behavior of HE at the mesoscale, including fracture evolution and plastic deformation. This thesis focus on three types of simulations. First, we investigate dynamic fracture and hotspot formation in HMX particles embedded in Sylgard binder undergoing high strain rate compression and harmonic excitation. We use the phase field damage model (PFDM) to simulate dynamic fracture. Also, we implement a thermal model to capture temperature increase due to fracture dissipation and friction at both cracks and debonded HMX/Sylgard interface. In our simulations, we observe that crack patterns are strongly dominated by initial defects such as pre-existing cracks and interface debonding. Regions with initial debonding between HMX particles and the polymer are critical sites where cracks nucleate and propagate. Heating due to friction generates in these regions too and caused the formation of critical hotspots. We also run simulations of a HMX particle under high-frequency harmonic excitation. As expected, higher frequencies and larger amplitudes lead to an increase in the damage growth rate. The simulations suggest that the intensity of the thermal localization can be controlled more readily by modifying the bonding properties between the particle and the binder rather than reducing the content of bulk defects in the particle.
Second, we present simulations of shock compression in HMX single crystals. For this purpose, we implemented a constitutive model that simulates the elastoplastic anisotropic response of this type of material. The continuum model includes a rate-dependent crystal plasticity model and the Mie-Gruneisen equation of state to obtain the pressure due to shock. Temperature evolves in the material due to plastic dissipation, shock, and thermo-elastic coupling. The model is calibrated with non-reactive atomistic simulations to make sure the model obeys the Rankine-Hugoniot jump conditions. We compare finite element (FE) and molecular dynamic (MD) simulations to study the formation of hot spots during the collapse of nano-size void in a HMX energetic crystal. The FE simulations captured the transition from viscoelastic collapse for relatively weak shocks to a hydrodynamic regime for strong shocks. The overall temperature distributions and the rate of pore collapse are similar to MD simulations. We observe that the void collapse rate and temperature field are strongly dependent on the plasticity model, and we quantify these effects. We also studied the collapse of a micron size void in HMX impacted at different crystal orientations and impact velocities. The simulation results of void collapse are in good agreement with a gas gun void collapse experiment. While the void size and crystal orientation do not affect the area ratio rate, they strongly affect the void collapse regime and temperature. Also, increased plastic activity when the crystal is impacted on the plane (110) renders higher temperature fields.
Finally, we studied shock compression and dynamic fracture in polycrystalline HMX using the same model implemented for shocks in single crystals. The goal of this study is to understand the role of crystal anisotropy and how it affects other hotspot formation mechanisms such as frictional heating. To simulate fracture, we used a phase field damage model implemented for large deformations. We first perform simulations of sustained shocks in polycrystalline HMX, where the grains are perfectly bonded to understand the effect of plastic deformation and hotspot formation due to plastic heating. Then, we simulate shocks in polycrystalline HMX with dynamic fracture. Simulations capture fracture evolution and frictional heating at cracks. In the polycrystalline case, we study heat generation due to shock and plastic deformation. A heterogeneous temperature field forms when the shock wave travels in the material. Temperature increases more in crystals that showed a higher magnitude of accumulated slip. When weak grain boundaries are included in the simulations, frictional heating becomes the dominant hotspot formation mechanism. As the crystals' interfaces break and crack surface sliding occurs, temperature increases due to friction at cracks. Hotspots tend to form at cracks oriented 45 deg from the shock direction. For this case, crystal anisotropy does not play an important role in temperature generation due to plastic dissipation. However, the random orientation of the crystals creates heterogeneous deformation and stress fields that cause the formation of a higher number of hotspots than the case where all the grains are oriented in the same direction.