Re: [ESPResSo-users] questions about speed electrostatic on espresso

[Georg Rempfer]

Please provide a complete simulation script, including particle placement and your integration loop. These runtimes seem unusually high, indeed.

Greetings,

Georg

On Wed, May 11, 2016 at 10:52 AM, <address@hidden> wrote:

Hi

I simulation a polymer with counterion in a box.

The size of simulation cell is Lx=5Rg   Ly=5Rg  Lz=10 Rg

Chain length is 46 and number of counterion is 138.

A  p3m algorithm is used to calculate the full electrostatic interactions between charged momomer and counterions.q_momomer =-3  q_counterin =+1 for 3000 step it takes about 50 minute.It is very slow.How do i improve it?

parametrs:

time_step 0.01

skin 0.4

lj_sigma 1.0

 lj_epsilon 1.0

 bjerrum [expr 0.7 * $lj_sigma]

d_mon $lj_sigma

 d_ion [expr 0.425*$lj_sigma]

 monomer_momomer_epsilon $lj_epsilon

 monomer_momomer_sigma $d_mon

monomer_momomer_cutoff [expr $monomer_momomer_sigma * pow(2.0,1./6.)]

 monomer_momomer_shift [calc_lj_shift $monomer_momomer_sigma $monomer_momomer_cutoff]

 monomer_momomer_offset 0.0

 monomer_ion_epsilon $lj_epsilon

 monomer_ion_sigma [expr ($d_mon+$d_ion)/2.0]

 monomer_ion_cutoff [expr $monomer_ion_sigma * pow(2.0,1./6.)]

 monomer_ion_shift [calc_lj_shift $monomer_ion_sigma $monomer_ion_cutoff]

 monomer_ion_offset 0.0

 ion_ion_epsilon $lj_epsilon

 ion_ion_sigma $d_ion

 ion_ion_cutoff [expr $ion_ion_sigma * pow(2.0,1./6.)]

 ion_ion_shift [calc_lj_shift $ion_ion_sigma $ion_ion_cutoff]

 ion_ion_offset 0.0

bond_K 30.

 bond_rcut [expr 1.2 * $lj_sigma]

temperature 1.0

 gamma 15.34

interacyions:

inter 0 fene $bond_K $bond_rcut

inter 0 0 lennard-jones $monomer_momomer_epsilon $monomer_momomer_sigma $monomer_momomer_cutoff $monomer_momomer_shift $monomer_momomer_offset

inter 0 1 lennard-jones $monomer_ion_epsilon $monomer_ion_sigma $monomer_ion_cutoff $monomer_ion_shift $monomer_ion_offset

inter 1 1 lennard-jones $ion_ion_epsilon $ion_ion_sigma $ion_ion_cutoff $ion_ion_shift $ion_ion_offset

inter coulomb $bjerrum p3m tune accuracy 1e-3

Thanks