Dear all,<br> <br> I did two md simulations of 200 particles each of a lennard-jones fluid. One of them gave me the correct pair distribution function for a lennard-jones fluid (converging to 1) and one did not (converging to zero). I have attached the .mdp files for both systems below. The barostats are different but I don't think this is the cause. I think that one worked because of the cut-off specifications (rlist, rvdw and rcoulomb), but I am not sure of the explanation of how the cut-off values can influence the shape of a pair distribution function. The fourier spacing in both parameter files are also different.<br>
Please, if someone knows how these cut-off values and maybe fourier spacing could influence the shape of a pair distribution function, let me know the explanation. <br><br>.mdp file which gave the plot which converges to zero:<br>
<br>title = NPT simulation of a LJ FLUID<br>cpp = /lib/cpp<br>include = -I../top<br>define = <br>integrator = md ; a leap-frog algorithm for integrating Newton's equations of motion<br>
dt = 0.002 ; time-step in ps<br>nsteps = 500000 ; total number of steps; total time (1 ns)<br>nstcomm = 1 ; frequency for com removal<br>nstxout = 500 ; freq. x_out<br>
nstvout = 500 ; freq. v_out<br>nstfout = 0 ; freq. f_out<br>nstlog = 50 ; energies to log file<br>nstenergy = 50 ; energies to energy file<br>
nstlist = 10 ; frequency to update neighbour list<br>ns_type = grid ; neighbour searching type<br>rlist = 1.0 ; cut-off distance for the short range neighbour list<br>
pbc = xyz ; Periodic boundary conditions:xyz, use periodic boundary conditions in all directions<br>periodic_molecules = no ; molecules are finite, fast molecular pbc can be used<br>
coulombtype = PME ; particle-mesh-ewald electrostatics<br>rcoulomb = 1.0 ; distance for the coulomb cut-off<br>vdw-type = Cut-off ; van der Waals interactions<br>
rvdw = 1.0 ; distance for the LJ or Buckingham cut-off<br>fourierspacing = 0.12 ; max. grid spacing for the FFT grid for PME<br>fourier_nx = 0 ; highest magnitude in reciprocal space when using Ewald<br>
fourier_ny = 0 ; highest magnitude in reciprocal space when using Ewald<br>fourier_nz = 0 ; highest magnitude in reciprocal space when using Ewald<br>pme_order = 4 ; cubic interpolation order for PME<br>
ewald_rtol = 1e-5 ; relative strength of the Ewald-shifted direct potential<br>optimize_fft = yes ; calculate optimal FFT plan for the grid at start up.<br>DispCorr = no ; <br>
Tcoupl = v-rescale ; temp. coupling with vel. rescaling with a stochastic term.<br>tau_t = 0.1 ; time constant for coupling<br>tc-grps = OXY ; groups to couple separately to temp. bath<br>
ref_t = 80 ; ref. temp. for coupling<br>Pcoupl = berendsen ; exponential relaxation pressure coupling (box is scaled every timestep)<br>Pcoupltype = isotropic ; box expands or contracts evenly in all directions (xyz) to maintain proper pressure<br>
tau_p = 0.5 ; time constant for coupling (ps)<br>compressibility = 4.5e-5 ; compressibility of solvent used in simulation<br>ref_p = 1.0 ; ref. pressure for coupling (bar)<br>
gen_vel = yes ; generate velocities according to a Maxwell distr. at gen_temp<br>gen_temp = 80 ; temperature for Maxwell distribution<br>gen_seed = 173529 ; used to initialize random generator for random velocities<br>
<br>.mdp file which gave the plot which converges to 1:<br><br>title = NPT simulation of a LJ FLUID<br>cpp = /lib/cpp<br>include = -I../top<br>define = <br>
integrator = md ; a leap-frog algorithm for integrating Newton's equations of motion<br>dt = 0.002 ; time-step in ps<br>nsteps = 500000 ; total number of steps; total time (1 ns)<br>
nstcomm = 1 ; frequency for com removal<br>nstxout = 1000 ; freq. x_out<br>nstvout = 1000 ; freq. v_out<br>nstfout = 0 ; freq. f_out<br>
nstlog = 500 ; energies to log file<br>nstenergy = 500 ; energies to energy file<br>nstlist = 10 ; frequency to update neighbour list<br>ns_type = grid ; neighbour searching type<br>
rlist = 0.3 ; cut-off distance for the short range neighbour list<br>pbc = xyz ; Periodic boundary conditions:xyz, use p b c in all directions<br>periodic_molecules = no ; molecules are finite, fast molecular pbc can be used<br>
coulombtype = PME ; particle-mesh-ewald electrostatics<br>rcoulomb = 0.3 ; distance for the coulomb cut-off<br>vdw-type = Cut-off ; van der Waals interactions<br>
rvdw = 0.7 ; distance for the LJ or Buckingham cut-off<br>fourierspacing = 0.135 ; max. grid spacing for the FFT grid for PME<br>fourier_nx = 0 ; highest magnitude in reciprocal space when using Ewald<br>
fourier_ny = 0 ; highest magnitude in reciprocal space when using Ewald<br>fourier_nz = 0 ; highest magnitude in reciprocal space when using Ewald<br>pme_order = 4 ; cubic interpolation order for PME<br>
ewald_rtol = 1e-5 ; relative strength of the Ewald-shifted direct potential<br>optimize_fft = yes ; calculate optimal FFT plan for the grid at start up.<br>DispCorr = no <br>
Tcoupl = nose-hoover; temp. coupling with vel. rescaling with a stochastic term.<br>tau_t = 0.5 ; time constant for coupling<br>tc-grps = OXY ; groups to couple separately to temp. bath<br>
ref_t = 80 ; ref. temp. for coupling<br>Pcoupl = parrinello-rahman ; exponential relaxation pressure coupling (box is scaled every timestep)<br>Pcoupltype = isotropic ; box expands or contracts evenly in all directions (xyz) to maintain proper pressure<br>
tau_p = 5.0 ; time constant for coupling (ps)<br>compressibility = 4.5e-5 ; compressibility of solvent used in simulation<br>ref_p = 1.0 ; ref. pressure for coupling (bar)<br>
gen_vel = yes ; generate velocities according to a Maxwell distr. at gen_temp<br>gen_temp = 80 ; temperature for Maxwell distribution<br>gen_seed = 173529 ; used to initialize random generator for random velocities<br>
<br>I appreciate your reply.<br><br>Lum<br>