Session O2: TMS: First-principles and Molecular Dynamics VII

Time dependent boundary conditions for large scale atomistic simulations of Richtmyer-Meshkov instabilities

Shock induced Richtmyer-Meshkov instabilities at perturbed metal vacuum/gas interfaces result in metal material ejecta. For strong shock waves, the material ejected is initially in the form of fluid sheets when the machining grooves are two-dimensional. These sheets ultimately break up to form a distribution of droplets. Large-scale non-equilibrium molecular dynamics simulations of Richtmyer-Meshkov instabilities allow the investigation of the dynamics of the breakup process but are limited in length and time scales by rarefaction wave reflections at the boundaries, which put the material behind the perturbed interface under tension eventually leading to spall. A time-dependent boundary condition based on the self-similar character of the release wave is shown to mitigate boundary reflections and reduce unwanted tensile waves behind the perturbed interface zone and thereby can be used to increase the wavelength and time scale for breakup simulations. We discuss the details of this method and results of NEMD simulations of a shocked Cu interface with a single mode perturbation characterized by a wavelength of $\lambda$ = 13 nm and a wavenumber amplitude product kh_0 = 1/2.   

Validation and Calibration of Metal Strength Models Using Richtmyer-Meshkov Instability Measurements and Simulations

The RM instability provides unique experimental insight into the arrest of spike growth due to material strength properties. This is especially important at high strain rates which cannot be probed using Hopkinson bar experiments. Data from experiments with a cylindrical copper flyer impacting a copper target with sinusoidal perturbations has been collected from Argonne National Laboratory's Advanced Photon Source (dynamic compression sector). Simulations have been compared against RM experimental data to address questions such as: - Can the experimental results be modeled using a 2D approximation or is 3D required? - What impact does uncertainty in the target geometry and material parameters have when comparing to the data? - Can a Lagrangian calculation adequately capture the jet growth rate or is Eulerian required? - How well do simple and more sophisticated material strength models capture the data? - Do we arrive at the same strength parameters for 2D and 3D simulations? - Should we examine another machined perturbation geometry? These results are preparatory for future experiments which will examine the strength of polymer coatings.   

Semi-analytic treatment of the Rayleigh-Taylor instability in a material with strength

Dynamic loading of samples with rippled interfaces has been used to infer material strength via measurement of the Rayleigh-Taylor instability driven growth. Typically, this analysis relies on two dimensional simulations of the driven assembly's response, varying strength models or other assembly parameters until measurements are reproduced. For materials exhibiting viscosity rather than strength, the growth rate can be expressed analytically. This approach has previously been extended to materials with strength by interpreting it as an effective viscosity. Here we report an alternative approach, incorporating the flow stress directly into the analytic expression for instability growth. Idealized two dimensional simulations were used to test and calibrate the growth rate. The rate relation can then be used to predict instability growth given the history of density ratio and acceleration at the interface, which can be obtained from one-dimensional simulations.   

Full multiphase description of materials: application on tin.

When a solid material is subjected to dynamic loading, a shock wave propagates from the impacted to the free surface where it is reflected. The behavior at free surface depends on the local state of matter and its roughness. When the material is in solid phase, spall appears in the bulk and the free surface remains intact. However, when a liquid or mixed solid/liquid phase appears, micro-spall occurs. Combined with roughness effects, the free surface crumbles into a particles cloud and leads to micro-jetting. Describing these processes in hydrodynamic codes remains a challenge since the equation of state (EoS), phase change kinetics, strength and damage modelling are all together involved in this description. Using tin as a school material, we initiated a program of work to improve our understanding of these mechanisms. It is based on a full modelling, implemented in a hydrocode. Here, we introduce this model based on a simple way to connect EoS, strength law and damage modelling. We show that the material description in the hydrocode calculation is improved. Details on stochastic analysis techniques used to determine the EoS (Pooh code) and the strength (CALiXt code) parameters will also be provided.   

Observation of phase transitions in shocked Tin by molecular dynamics

In classical shock theory, a phase transition leads to more or less marked discontinuities of the Hugoniot curve and can cause the split of the shock front. These characteristics are easily detected by the usual shock experiments, but the measurements are only indirect observations of phase transitions and give no information about the corresponding microscopic processes. Recent developments in new facilities (synchrotrons, free-electron X lasers) open up interesting new perspectives: the direct observation of the structural state behind the shock front becomes now possible. In parallel with these experiments, it is important to have proven simulation methods operating on the same scale as the basic phase transition processes, namely the microscopic scale. In this talk, we present a set of large-scale molecular dynamics simulations in a tin monocrystal, in which several phase transitions are expected under shock. The results are then compared with existing time resolved X-ray diffraction experiments.