6
$\begingroup$

As far as I understand, there are two related concepts called potential and forcefield. A forcefield is essentially a sum term of more than one potential.

My question is, is the term "Forcefield" only applicable to MD (Molecular Dynamics) simulations?

Can MC (Monte-Carlo) simulations use "Forceforeld"s?

Improve this question
$\endgroup$
4
  • 1
    $\begingroup$ I'm not sure about MC simulations, but in geometry optimizations and conformational searches, one can definitely use the name "force fields". $\endgroup$
    – wzkchem5
    Jun 23 at 6:09
  • 2
    $\begingroup$ The basic wiki page answers your question : en.wikipedia.org/wiki/Force_field_(chemistry). A force field is needed for a very fast computation compared to ab initio, in a simple MC algorithm the transition rate $i \rightarrow j$ depends on $e^{\beta\Delta E_{i,j}}$, $E_{i/j}$ is issued from a force field or whatever. $\endgroup$
    – M06-2x
    Jun 23 at 10:11
  • 3
    $\begingroup$ The whole distinction between MD and MC can be a bit blurry anyway when you start looking at things like force-bias Monte Carlo or Langevin Thermostats in MD $\endgroup$
    – Ian Bush
    Jun 23 at 19:02
  • 3
    $\begingroup$ @IanBush Another good one is hybrid MC/MD where you do Monte Carlo sampling but use MD to propose the moves for acceptance/rejection. $\endgroup$
    – jheindel
    Jun 23 at 20:32

1 Answer 1

Reset to default
6
$\begingroup$

My experience is that people use the term "potential" and "force field" nearly inter-changeably. This is because the force at a particular point on a potential energy surface (i.e. a potential) is $F=-\nabla V(\mathbf{r})$. That is, the force points in the direction of steepest descent on the potential.

Because the two concepts are so closely related, people use them interchangeabley.

Strictly speaking, one only needs the potential to sample a particular ensemble, and this is what is usually done in monte carlo simulations. However, using the forces as a means of sampling tends to be quite efficent when there aren't large barriers between certain parts of configuration space.

Basically, whichever term you use, people will know what you mean in the context of molecular dynamics or monte carlo.

Improve this answer
$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .