Thank you Nuwan Dewapriya it will be interesting for my colleagues. I will share your information with them. On the other hand, what is your opinion regarding Ansys LSDyna?

Thank you Vadym Slyusar. The finite element models are unable to reveal complex energy dissipation mechanisms at the nanoscale. In this post, I have briefly mentioned why molecular dynamics is suitable for these problems (or why FE is not) https://www.researchgate.net/post/Atomic-Scale_Investigation_on_the_Mechanical_Behavior_of_Ultrathin_Multilayers_Under_Shock_Loading. However, those MD models can be used to obtain more accurate constitutive relations for macroscale FE models.

Thank you, Nuwan Dewapriya . Your comment is very important.

Thank you Vadym Slyusar.

We recently published two more articles, and we made our MD (LAMMPS) and DFT (VASP) models freely available.

The first is “Quantum and classical molecular dynamics simulations of shocked polyurea and polyurethane” (https://doi.org/10.1016/j.commatsci.2021.111166). The models are available here: https://github.com/nuwan-d/quantum_md_of_shocked_polymers

The second paper is “Molecular dynamics study on the shock induced spallation of polyethylene”( https://doi.org/10.1063/5.0072249). The models are available here: https://github.com/nuwan-d/shock_response_pe

Hi Nuwan, Do you think it would be straightforward to tweak your MD for the propagation of SW in water? Thanks in advance for your response.

Hi Ramin M. Abolfath

I think MD can be used to simulate the SW propagation in water. Once the MD model of water is ready, proper displacements need to be applied to a selected set of water molecules.

Also, how long does it take for molecules in SW to relax to the thermal equilibrium?

I've never modeled water or SW in MD. I guess that the relaxation time could be a few picoseconds.

Thank you, Nuwan Dewapriya, for sharing the input scripts selflessly, which is beneficial for my research. I’d like to study the shocking behavior of the polymer networks with bead-spring models. I have learned your scripts and am still curious about equilibrating the sample with free surfaces. Specifically, in the shock model, the replicate sample was relaxed under the NVT ensemble. But the stress in the transverse directions may not be relaxed due to the lack of barostat. While in the impact model, you mentioned using indenter surfaces to compress the sample back to the specified length. After that , NVT ensemble was used to equilibrate the system. But the length may exceed the specified value during the relaxation. I’d like to get any suggestions on the properly equilibrated configuration, which is important for the following shock.

Thank you Ricky Zhao for your kind words.

Depending on the MD model (e.g., polymer structure, chain length, force field), the process of replicating the periodic cell can induce some residual stresses in the polymer sample. We experienced that in one of our polyurea modes. We got rid of the residual stresses by equilibrating the replicated sample at a temperature well above its glass transition temperature (using NVT). More information is available in this paper: https://doi.org/10.1557/s43579-021-00073-5

The main purpose of using an indenter surface was to prevent the protrusion of some polymer chains at the free surfaces. As you have mentioned, the equilibrated sample length exceeds the target value resulting in a somewhat loosely packed surface layer. However, the thickness of that layer is ~1 nm; Figure 2d of this paper: https://doi.org/10.1016/j.commatsci.2021.110504 shows the through-thickness density profiles of the bilayer impact models. I don’t think that this surface layer has a significant influence on the shock response of the targets.

Thank you Nuwan Dewapriya for your detailed reply! I still have a question of how to use an indenter surface to compress the system in lammps? Do you use fix indent command? Thanks!

That is correct Ricky Zhao . I use the fix indent command.

Hi Nuwan. I am a novice to MD simulation of SW with LAMMPS. I was wondering you or anyone in this thread would be interested in collaborating on a project? It would be simple for you. We can publish the results together. The application is on cancer therapy. Please contact me if interested. The system of interest is water molecules.

Hypervelocity impacts & shock waves can be cool

Analysis of hypervelocity impacts: The tungsten case. - YouTube

Hypervelocity dust impacts on plasma-facing materials through molecular dynamics simulations - YouTube

Hi Nuwan. Can you please help me how to plot temperature during the spallation process. Should I use temp/region command please let me know. I really need your help with this

Hi Srishti Mishra ,

The LAMMPS input file "spall_in.in" in this GitHub repository (https://github.com/nuwan-d/MD_model_JAM-21-1174) shows how to compute the density and the MATLAB file "x_t_density.m" shows how to plot the spatiotemporal variation in density. A similar procedure can be followed to plot temperature.

It should be noted that the center-of-mass velocity of a group of atoms needs to be subtracted when computing temperature. The "compute temp/com" command in LAMMPS can be used for that: https://docs.lammps.org/compute_temp_com.html

Hi Nuwan Dewapriya, I have been given your research model as a project and the alteration that we have to do for it is change it into single aluminum crystal. I made a similar model on OVITO and ran simulations on it but couldn't get that flow of atoms in x-direction after the piston movement. Is it possible to make changes in your code for material only and keeping the geometry file, if yes than how ?

Thank you in advance.

Hi Kevin Sangoi , As in this input file: https://github.com/nuwan-d/MD_model_JAM-21-1174/blob/main/spall_in.in

you need to properly define the piston (lines 49 & 50) by taking into account the geometry of your aluminum model.

Our latest MD (LAMMPS) and DFT (VASP) models of phase-separated polyurea are available here: https://github.com/nuwan-d/shock_response_of_polyurea

How can I apply the shock-front absorbing boundary condition in LAMMPS for gathering shock-induced attributes? I mean after shock when the cell reached maximum compression, I want to create a piston in opposite to shock which will disable the reflection of shock to capture the changes during shock. Nuwan Dewapriya

Md. Rifat Hossain

I am not sure how that can be done in LAMMPS. Perhaps, this paper might give you some ideas: Article Molecular dynamics simulation of the shock response of mater...