Re: change in box geometry for simulation of pure water

From: Axel Kohlmeyer (akohlmey_at_cmm.chem.upenn.edu)
Date: Thu Aug 28 2008 - 14:06:47 CDT

On Wed, 27 Aug 2008, Chris Harrison wrote:

CH> Maria,
CH>
CH> Can you confirm for me that the box becomes thinner in the z-axis? And: was
CH> the water box minimized and pre-equilibrated prior to turning-on
CH> usecontantratio and useflexbilecell?

hi,

couldn't resist to chime in...

i've seen this behavior before with other implementations.
to the best of my knowledge the regular ewald or PME should
only work well for exactly cubic simulations cells (you are
practically computing an infinitely large spherical crystal
_with_ a dielectric boundary and there are some small and
cell size independent non-vanishing terms in the potential).
a fully variable cell would thus be meta-stable. as soon as
your box becomes significantly flat or tilted it will not
go back.

if i remember correctly the effect can be mediated by using
a tighter convergence, but ultimatly enforcing isotropic cell
changes is probably the cleanest solution.

cheers,
   axel.

CH>
CH>
CH> -Chris
CH>
CH>
CH>
CH> On Wed, Aug 27, 2008 at 7:48 AM, maria goranovic
CH> <mariagoranovic_at_gmail.com>wrote:
CH>
CH> > Hello,
CH> >
CH> > I ran a simulation of a box of water with a standard config file using
CH> > useconstantratio and useflexiblecell. However, the simulation box becomes
CH> > thinner in one dimension, until the simulation crashes. Why does this
CH> > happen?
CH> >
CH> > Thanks for the suggestions, config file below:
CH> >
CH> > !!!!!!!!!!!!!!!!!!!!!!!!!
CH> > set MOL wat2
CH> >
CH> > set temp 310
CH> >
CH> > set xbox 39.97
CH> > set ybox 39.94
CH> > set zbox 39.94
CH> >
CH> > structure ${MOL}.psf
CH> > coordinates ${MOL}.pdb
CH> > temperature 0
CH> > parameters par-combined.prm
CH> > paraTypeCharmm on
CH> >
CH> > outputEnergies 5000
CH> > outputTiming 5000
CH> > xstFreq 5000
CH> > dcdFreq 5000
CH> > wrapAll on
CH> > margin 5
CH> >
CH> > rigidbonds on
CH> > timestep 2
CH> > nonBondedFreq 1
CH> > fullElectFrequency 1
CH> > stepsPerCycle 20
CH> >
CH> > exclude scaled1-4
CH> > 1-4scaling 1.0
CH> > switching on
CH> > switchDist 10
CH> > cutoff 12
CH> > pairlistdist 14
CH> >
CH> > cellBasisVector1 $xbox 00.00 00.00
CH> > cellBasisVector2 00.00 $ybox 00.00
CH> > cellBasisVector3 00.00 00.00 $zbox
CH> >
CH> > #########################################
CH> > cellOrigin 0.05952 0.09806 -0.1598
CH> >
CH> > Pme on
CH> > PmeGridsizeX 40
CH> > PmeGridsizeY 40
CH> > PmeGridsizeZ 40
CH> >
CH> > #########################################
CH> > # PRESSURE AND TEMPERATURE CONTRON
CH> > #########################################
CH> >
CH> > # temperature
CH> > langevin on
CH> > langevinDamping 5
CH> > langevinTemp $temp
CH> > langevinHydrogen no
CH> >
CH> > useflexiblecell yes
CH> > useconstantratio yes
CH> > langevinPiston on
CH> > langevinPistonTarget 1.01325
CH> > langevinPistonPeriod 100
CH> > langevinPistonDecay 50
CH> > langevinPistonTemp $temp
CH> >
CH> > binaryoutput on
CH> > outputname ${MOL}-equil
CH> >
CH> > # INITIAL ENERGY
CH> > minimize 0
CH> >
CH> > minimize 5000
CH> > output ${MOL}-min1
CH> >
CH> > #########################################
CH> > # HEAT
CH> > #########################################
CH> >
CH> > set tem 310
CH> > langevinTemp $tem
CH> >
CH> > for {set p 1 } { $p < 11 } { incr p } {
CH> > run 1000000
CH> > output ${MOL}-$p
CH> > }
CH> >
CH> >
CH> >
CH> > --
CH> > Maria G.
CH> > Technical University of Denmark
CH> > Copenhagen
CH> >
CH>
CH>
CH>
CH>

-- 
=======================================================================
Axel Kohlmeyer   akohlmey_at_cmm.chem.upenn.edu   http://www.cmm.upenn.edu
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:49:47 CST