AIMD Supercell Size for Perfect Crystals.

[hszhao.cn@gmail.com]

Hello VASP community,

I'm planning to run some Ab Initio Molecular Dynamics (AIMD) simulations on perfect crystal systems. I have a question regarding the use of supercells: Is it always necessary to use an appropriately sized supercell for AIMD simulations, even when dealing with perfect crystals?

I understand that supercells are crucial for studying defects and can help with issues like phonon sampling and finite size effects. However, I'm unsure if these considerations are as important for a perfect crystal structure. Specifically, I'd appreciate insights on:

1. The main advantages of using a supercell for perfect crystals in AIMD.
2. Potential drawbacks or inaccuracies that might arise from using a smaller cell.
3. Guidelines for determining an appropriate supercell size for AIMD of perfect crystals.
4. Any computational tricks or settings in VASP that could help balance accuracy and efficiency when choosing supercell size.

Thank you in advance for your expertise and advice!

Regards,
Zhao

[christopher_sheldon1]

Hi Zhao,

This is quite a broad question. If you're new to AIMD, then I'd recommend taking a look at the VASP tutorials. There's one on MD that might be helpful to go through: https://vasp.at/tutorials/latest/md/ . There's also a wiki page with more information about MD generally: wiki/index.php/Category:Molecular_Dynamics .

Does this help?

Best wishes,

Chris

[hszhao.cn@gmail.com]

See here for the related discussion.

[christopher_sheldon1]

Thank you. It seems that the link you shared has answered your question sufficiently, so I will lock this topic.

Best,

Chris