I am working on modeling the adhesion force between silicon crystals using LAMMPS software. while I am doing literature research I found many potentials have been used to simulate silicon like (tersoff and reax). Also I am still not sure about the best way to build silicon crystal. Do I need to use another program to create the model or I can build it in LAMMPS.
Thank for your help,
It is fairly easy to build a silicon crystal using lattice diamond command within lammps..
The choice of potential depends on the kind of application .. In general, Kumagai or Screening functions are quite good...https://github.com/Atomistica/atomistica
For a beginner it may be tough to start these installations, so you may start with Erhart potential first which is provided within the LAMMPS potential folder.
Dear Saurav Goel , I find the potential you recommended is meaningful after read some of your articles. But I have some problems with install the atomisica. It seems the package can only be used in Linux. Would you kindly introduce your installation process as reference? Thanks for your time.
Dear Wang Sorry I saw your message just now after a while. MY simple suggestion is that use potentials available within LAMMPS such as Vasistha et al, or SW type and there is no much advantage chasing atomistica potential as it is computationally very expensive. So if you are not using HPC and thinking of simple workstation you better use faster potentials. We have described there here: https://www.sciencedirect.com/science/article/pii/S089069551400145X
If you are still keen to install atomisitca, it depends a lot on the architecture type, I can help you over teamviewer if you wish.
ok great. If you need any further inputs let me know. We are currently writing a paper on comparing these and find that SW is sort of a reasonable potential for study dislocation plasticity in silicon. For ideally for simulation such as cutting, I'd suggest use Erhart and Albe (Tersoff) potential to describe carbon and cross interaction between silicon and carbon and use SW potential to describe silicon. It would be faster and computationally good too.
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering
- Electrodes