[gmx-users] Some questions: epsilon_r , DispCorr , g_velacc and PME

[Anton Feenstra]

David wrote:
> On Fri, 2003-08-15 at 17:32, Nuno R. L. Ferreira wrote:
> 
>>Dear * gmx's
[...]
>>3) I'm starting to work with autocorrelation functions, namely VACF. Can
>>anibody give me some hints on how to calculate diffusion coefficients from
>>this?
> 
> Doesn't g_velacc do that for you? Otherwise check some textbook like
> Allen & Tildesley (that never hurts!)

To get accurate diffusion from the vacc, you will need velocities on every
md step (i.e. every 2 fs!). Usually, that means writing a lot of data,
since you will still need, for most systems, several 100ps of simulation
to get converged systems. (I've had some students work on this once.)

If you work out the maths, you see that the integral of your VACF is exactly
the same as the slope of the positional MSD (or something similar), from both
you get your diffusion coefficient, which means that mdrun actually can do
the (accurate!) integration for you!


-- 
Groetjes,

Anton
  _____________ _______________________________________________________
|             |                                                       |
|  _   _  ___,| K. Anton Feenstra                                     |
| / \ / \'| | | Dept. of Pharmacochem. - Vrije Universiteit Amsterdam |
|(   |   )| | | De Boelelaan 1083 - 1081 HV Amsterdam - Netherlands   |
| \_/ \_/ | | | Tel: +31 20 44 47608 - Fax: +31 20 44 47610           |
|             | Feenstra at chem.vu.nl - www.chem.vu.nl/~feenstra/       |
|             | "If You See Me Getting High, Knock Me Down"           |
|             | (Red Hot Chili Peppers)                               |
|_____________|_______________________________________________________|