[gmx-users] energy minimisation of carbon nanotubes.

[Justin Lemkul]

On 9/15/12 12:49 PM, Elie M wrote:
>
> Dear all,
> I have a small question regarding energy minimization. I am minimizing a carbon naotube in vaccuo and my (mdp )input file is: ;
>
> ;     SWNT (10,0) on its own
>
> ;     Elie Moujaes, SWNT+polymers project
>
> ;     Energy minimization input file
>
> ;
>
> cpp                 =/usr/bin/cpp
>
> define              =
> -DFLEXIBLE
>
> constraints         =
> none
>
> integrator          =
> cg
>
> nsteps              =
> 2000
>
> ;
>
> ;     Energy minimizing stuff
>
> ;
>
> emtol               =  0.00001
>
> emstep              =
> 0.03
>
>
>
> nstcomm             =
> 1
>
> pbc                 =  xyz
>
> periodic_molecules  =  yes
>
> ns_type             =
> grid
>
> rlist               =  1
>
> rcoulomb            =  1.0
>
> rvdw                =  1.0
>
> Tcoupl              =
> no
>
> Pcoupl              =
> no
>
> gen_vel             =
> no
> I have put emstep very loww toa chieve accuracy.In the output i got:
> Reading file SWNT-EM.tpr, VERSION 4.0.5 (single precision)Polak-Ribiere Conjugate Gradients:   Tolerance (Fmax)   =  1.00000e-05   Number of steps    =         2000   F-max             =  3.48055e+03 on atom 799   F-Norm            =  9.67779e+02
>
> Stepsize too small, or no change in energy.Converged to machine precision,but not to the requested precision Fmax < 1e-05
> Double precision normally gives you higher accuracy.
> writing lowest energy coordinates.
> Polak-Ribiere Conjugate Gradients converged to machine precision in 359 steps,but did not reach the requested Fmax < 1e-05.Potential Energy  =  8.7646348e+03Maximum force     =  1.6186214e+01 on atom 701Norm of force     =  5.4860659e+00
> It seems that I have to use the double precision version but as the single precision is concerned,Is the minimization fine (as the maximum force on atom 701 is 16...)?

The outcome seems reasonable.  I doubt that any algorithm, in single or double 
precision, would ever reach Fmax < 1e-05.  Values of Fmax < 10 are generally 
suitable for just about any purpose and Fmax < 1000 will generally result in 
stable simulations for most systems.

-Justin

-- 
========================================

Justin A. Lemkul, Ph.D.
Research Scientist
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================