[lammps-users] Simulation of chips sliding on Graphene

Hi all,

I want to move a molecule with a hemispherical diamond structure on graphene and measure the coefficient of friction.

Both molecules are treated as rigid bodies, and a moving virtual atom and a diamond tip are connected by a spring to apply a load to the tip. However, the diamond tip penetrates graphene.

What are some tips for doing this simulation? Also, please let me know if there are any other points that need to be corrected.


are you saying that this input does not do what you want or that you want people to review it before you run it?

Thank you for your reply, Axel.

I want the tip to move smoothly over the graphene, but the tip penetrates the graphene.


2020年12月7日(月) 22:48 Axel Kohlmeyer <[email protected]>:

Thank you for your reply, Axel.

I want the tip to move smoothly over the graphene, but the tip penetrates the graphene.

this is likely by construction. your input is rather complex and thus difficult to read.
you will have to debug it yourself. please note the following:

  • to have atoms immobile you simply don’t have to include them in time integration. using fix rigid for that is almost always a mistake and applying a temperature with fix rigid/nvt even more so.
  • if you want atoms to be thermally excited, but still restrained to a given location, consider using fix spring/self or create a more realistic model, where you place the sheet on top of a slab of a solid.
  • if you want the sheet to be somewhat mobile but restrain the center of mass, consider using fix spring.
  • if you use fix move the atoms will disregard all forces and move atoms exactly as prescribed
  • applying multiple fixes to do time integration of the same atoms is almost always a mistake (it is only meaningful if those different fixes work on different directions only). LAMMPS will warn about this.
  • it is difficult to debug a complex input. it is usually best to start over with an empty input and then add components in steps, check that everything works as expected, then add some more until the step where the system starts behaving in an undesired way is narrowed down. what to do in that situation depends on the specifics of what one wants to learn from the simulation. thus there can be no generic advice
  • there are multiple ways to drags an object “along”
  • if your intent is to model something like an AFM experiment, be aware that those are not as smooth and gentle as you might expect due to the relatively large mass of the AFM tip compared to a graphene sheet and the mechanics associated with moving the tip.