Michael I. Zeifman

Combined molecular dynamics–direct simulation Monte Carlo computational study of laser ablation plume evolution

A two-stage computational model of evolution of a plume generated by laser ablation of an organic solid is proposed and developed. The first stage of the laser ablation, which involves laser coupling to the target and ejection of molecules and clusters, is described by the molecular dynamics (MD) method. The second stage of a long-term expansion of the ejected plume is modeled by the direct simulation Monte Carlo (DSMC) method. The presence of clusters, which comprise a major part of the overall plume at laser fluences above the ablation threshold, presents the main computational challenge in the development of the combined model. An extremely low proportion of large-sized clusters hinders both the statistical estimation of their characteristics from the results of the MD model and the following representation of each cluster size as a separate species, as required in the conventional DSMC. A number of analytical models are proposed and verified for the statistical distributions of translational and inter...

Direct Simulation Monte Carlo Modeling of Homogeneous Condensation in Supersonic Plumes

A particle simulation method to model water condensation process in a supersonic rocket plume is proposed and developed. Classic nucleation theory is used to predict nucleation, condensation, and the evaporation rates for water clusters. Microscopic kinetic models are developed to simulate collision processes between water clusters and monomers, between water clusters and foreign molecules, and evaporation of monomers from water clusters. These models are integrated into the direct simulation Monte Carlo method to simulate an axisymmetric multispecies gas expansion coupled with condensation. The developed computational scheme, first verified by empirical scaling laws of condensation in supersonic microjets, is then applied to predict the spatial distributions of water cluster number density, size, and temperature in a rocket exhaust plume. An empirical equation is used to correct classical nucleation rate, condensation results are compared between the original and corrected nucleation rate, and the impact of the nucleation rate on a flow with condensation is discussed in detail.

Nonequilibrium numerical model of homogeneous condensation in argon and water vapor expansions

A computational approach capable of modeling homogeneous condensation in nonequilibrium environments is presented. The approach is based on the direct simulation Monte Carlo (DSMC) method, extended as appropriate to include the most important processes of cluster nucleation and evolution at the microscopic level. The approach uses a recombination-reaction energy-dependent mechanism of the DSMC method for the characterization of dimer formation, and the RRK model for the cluster evaporation. Three-step testing and validation of the model is conducted by (i) comparison of clusterization rates in an equilibrium heat bath with theoretical predictions for argon and water vapor and adjustment of the model parameters, (ii) comparison of the nonequilibrium argon cluster size distributions with experimental data, and (iii) comparison of the nonequilibrium water cluster size distributions with experimental measurements. Reasonable agreement was observed for all three parts of the validation.

Kinetic Model of Condensation in a Free Argon Expanding Jet

The direct-simulation Monte Carlo (DSMC) method has recently been developed to simulate homogeneous condensation in a free-expansion rocket plume. However, cluster‐monomer and cluster‐cluster collision models as well as the determination of cluster size were simplified in the previous work, and the effect on the accuracy of the numerical simulation results was not quantified. In this work, the molecular-dynamics (MD) method is used to simulate collision and sticking probabilities for argon clusters and the results are compared with the hard-sphere model. These improved models are then integrated into a DSMC code to predict the Rayleigh scattering intensity in a free-expanding argon condensation plume, and numerical results are compared with experimental data along the plume centerline. Nomenclature A = constant B = constant b = impact parameter c =v elocity d = diameter E = evaporation rate or energy I = intensity i = number of atoms J = nucleation rate K = intensity constant k B = Boltzmann’s constant L = specific latent heat M = cluster mass m = molecular mass N = number density nc = number of simulated nuclei particles ps = saturation pressure q = sticking coefficient R = ideal gas constant r = distance or radius T = temperature V = interaction potential or specific volume α = species polarizability � t = time step � V = cell volume � = potential constant λ0 =w avelength ρ = density σ = surface tension or potential constant Subscripts

Sensitivity of Water Condensation in a Supersonic Plume to the Nucleation Rate

The direct simulation Monte Carlo (DSMC) method has recently been developed to simulate homogeneous water condensation in a free expansion rocket plume. A nucleation rate, based on the classical nucleation theory (CNT), was used in the DSMC simulation to predict initial nuclei in the condensation region. However, recent experimental research suggests that the CNT nucleation rate should be corrected and the magnitude of the correction is a function of the vapor temperature. Because the nucleation rate is one of the most important factors that impacts the accuracy of the numerical simulation results for a condensation plume, the impact is investigated of the corrected nucleation rate on cluster growth processes and flow macroparameters in an expanding flow using the DSMC method to model both the gas and condensate flow.

Analysis of Numerical Errors in the DSMC Method

The direct simulation Monte Carlo (DSMC) method is one of the most popular numerical methods used to model rarefied gas environment flows. In order to predict the accuracy of a solution obtained by the DSMC method we have to be able to estimate its accuracy. In the work presented here we have developed a technique to estimate the numerical accuracy of the DSMC method. This paper presents a derivation of expressions of the variance of the DSMC estimators of number density and translational temperature, and the corresponding comparison with the empirical variance. A discussion of the deterministic numerical errors corresponding to typical DSMC parameters such as the time step, cell volume, and total number of simulated particles is given. Moreover, a comparison of two different DSMC schemes, No Time Counter (NTC) and Majorant Frequency (MF), is made.

Direct simulation of condensation in a one-dimensional unsteady expansion: Microscopic mechanisms

We apply a molecular dynamics (MD) technique to the simulation of a quasi-one-dimensional unsteady free expansion to determine the dominant microscopic mechanisms of condensation in supersonic flows. In this way, it is possible to reproduce the basic physics of the coupled condensation flow with a moderate computational effort. The MD results confirm that the fundamental mechanism for the initiation of condensation is through dimer formation in two-stage ternary collisions of monomers.

Multiscale simulation of laser ablation of organic solids: evolution of the plume

A computational approach that combines the molecular dynamics (MD) breathing sphere model for simulation of the initial stage of laser ablation and the direct simulation Monte Carlo (DSMC) method for simulation of the multi-component ablation plume development on the time- and length-scales of real experimental configurations is presented. The combined multiscale model addresses different processes involved in the laser ablation phenomenon with appropriate resolutions and, at the same time, accounts for the interrelations among the processes. Preliminary results demonstrate the capabilities of the model and provide new insights into complex processes occurring during the ablation plume expansion. The spatial distribution of monomers in the plume is found to be strongly affected by the presence of large clusters. Interaction between the clusters and monomers can result in splitting of the monomer distribution into faster and slower components. The overall spatial mass distribution is found to have little relation with the monomer distribution.

Modeling of Argon Condensation in Free‐jet Expansions with the DSMC Method

Homogeneous argon condensation is modeled in a free expansion jet with a particle method. Microscopic kinetic models are used to simulate nucleation process, collisional processes among clusters and monomers, and evaporation of monomers from clusters. These models are integrated into the direct simulation Monte Carlo (DSMC) method to simulate an axisymmetric argon expansion coupled with condensation. Experimental data is used to correct the nucleation transient time, nucleation rate, and the condensation coefficient in the nucleation model. The corrected model is then, for the first time, applied to predict the distribution of Rayleigh scattering intensity in a free expansion argon condensation plume.

A Hybrid MD‐DSMC Model of Picosecond Laser Ablation and Desorption

A two-stage computational model of the evolution of a plume generated by laser ablation of an organic solid is presented and discussed. The first stage of the laser ablation involves laser coupling to the target and ejection of the plume and is described by molecular dynamics (MD) simulations. The following stage of a long-term expansion of the ejected plume is modeled by the direct simulation Monte Carlo (DSMC) method. The results of the MD simulations demonstrate that the physical mechanism of material ejection at sufficiently high laser fluences is a phase explosion of the overheated material followed by a homogeneous decomposition of the expanding plume into a mixture of liquid droplets (molecular clusters) and gas phase molecules. The extremely low proportion of large-size clusters hinders both statistical description of their parameters from the results of MD simulations and the following representation of each cluster size as a separate species, as required in the conventional DSMC. Therefore, a new computational scheme, which treats the size of large clusters as a random variable, is developed. The results of the hybrid model demonstrate that even for low laser fluences and short pulse duration, the evolution of the plume differs considerably from that predicted by pure thermal desorption models. For high fluences, the phase explosion of the target material and intensive processes of particle interactions within the plume are responsible for dramatic changes in the plume evolution as compared to that at low fluences.