241 lines
7.1 KiB
C++
241 lines
7.1 KiB
C++
/* ----------------------------------------------------------------------
|
|
LAMMPS - Large-scale Atomic/Molecular Massively Parallel Simulator
|
|
www.cs.sandia.gov/~sjplimp/lammps.html
|
|
Steve Plimpton, sjplimp@sandia.gov, Sandia National Laboratories
|
|
|
|
Copyright (2003) Sandia Corporation. Under the terms of Contract
|
|
DE-AC04-94AL85000 with Sandia Corporation, the U.S. Government retains
|
|
certain rights in this software. This software is distributed under
|
|
the GNU General Public License.
|
|
|
|
See the README file in the top-level LAMMPS directory.
|
|
------------------------------------------------------------------------- */
|
|
|
|
/* ----------------------------------------------------------------------
|
|
Contributing author: Pieter in 't Veld (SNL)
|
|
------------------------------------------------------------------------- */
|
|
|
|
#include "math.h"
|
|
#include "string.h"
|
|
#include "stdlib.h"
|
|
#include "fix_nvt_sllod.h"
|
|
#include "math_extra.h"
|
|
#include "atom.h"
|
|
#include "force.h"
|
|
#include "domain.h"
|
|
#include "group.h"
|
|
#include "update.h"
|
|
#include "respa.h"
|
|
#include "modify.h"
|
|
#include "fix.h"
|
|
#include "fix_deform.h"
|
|
#include "compute.h"
|
|
#include "error.h"
|
|
|
|
using namespace LAMMPS_NS;
|
|
|
|
enum{NO_REMAP,X_REMAP,V_REMAP}; // same as fix_deform.cpp
|
|
enum{NOBIAS,BIAS};
|
|
|
|
/* ---------------------------------------------------------------------- */
|
|
|
|
FixNVTSlodd::FixNVTSlodd(LAMMPS *lmp, int narg, char **arg) :
|
|
FixNVT(lmp, narg, arg) {}
|
|
|
|
/* ---------------------------------------------------------------------- */
|
|
|
|
void FixNVTSlodd::init()
|
|
{
|
|
FixNVT::init();
|
|
|
|
if (!temperature->tempbias)
|
|
error->all("Temperature for fix nvt/sllod does not have a bias");
|
|
|
|
// check fix deform remap settings
|
|
|
|
int i;
|
|
for (i = 0; i < modify->nfix; i++)
|
|
if (strcmp(modify->fix[i]->style,"deform") == 0) {
|
|
if (((FixDeform *) modify->fix[i])->remapflag != V_REMAP)
|
|
error->all("Using fix nvt/sllod with inconsistent fix deform remap option");
|
|
break;
|
|
}
|
|
if (i == modify->nfix)
|
|
error->all("Using fix nvt/sllod with no fix deform defined");
|
|
}
|
|
|
|
/* ---------------------------------------------------------------------- */
|
|
|
|
void FixNVTSlodd::initial_integrate(int vflag)
|
|
{
|
|
double dtfm;
|
|
|
|
double delta = update->ntimestep - update->firststep;
|
|
delta /= update->nsteps;
|
|
t_target = t_start + delta * (t_stop-t_start);
|
|
|
|
// update eta_dot
|
|
|
|
f_eta = t_freq*t_freq * (t_current/t_target - 1.0);
|
|
eta_dot += f_eta*dthalf;
|
|
eta_dot *= drag_factor;
|
|
eta += dtv*eta_dot;
|
|
factor = exp(-dthalf*eta_dot);
|
|
|
|
// update v and x of only atoms in group
|
|
// remove and restore bias = streaming velocity = Hrate*lamda + Hratelo
|
|
// thermostat thermal velocity only
|
|
// vdelu = SLLOD correction = Hrate*Hinv*vthermal
|
|
|
|
double **x = atom->x;
|
|
double **v = atom->v;
|
|
double **f = atom->f;
|
|
double *mass = atom->mass;
|
|
int *type = atom->type;
|
|
int *mask = atom->mask;
|
|
int nlocal = atom->nlocal;
|
|
if (igroup == atom->firstgroup) nlocal = atom->nfirst;
|
|
|
|
double h_two[6],vdelu[3];
|
|
MathExtra::multiply_shape_shape(domain->h_rate,domain->h_inv,h_two);
|
|
|
|
for (int i = 0; i < nlocal; i++) {
|
|
if (mask[i] & groupbit) {
|
|
temperature->remove_bias(i,v[i]);
|
|
vdelu[0] = h_two[0]*v[i][0] + h_two[5]*v[i][1] + h_two[4]*v[i][2];
|
|
vdelu[1] = h_two[1]*v[i][1] + h_two[3]*v[i][2];
|
|
vdelu[2] = h_two[2]*v[i][2];
|
|
dtfm = dtf / mass[type[i]];
|
|
v[i][0] = v[i][0]*factor + dtfm*f[i][0] - dthalf*vdelu[0];
|
|
v[i][1] = v[i][1]*factor + dtfm*f[i][1] - dthalf*vdelu[1];
|
|
v[i][2] = v[i][2]*factor + dtfm*f[i][2] - dthalf*vdelu[2];
|
|
temperature->restore_bias(i,v[i]);
|
|
|
|
x[i][0] += dtv * v[i][0];
|
|
x[i][1] += dtv * v[i][1];
|
|
x[i][2] += dtv * v[i][2];
|
|
}
|
|
}
|
|
}
|
|
|
|
/* ---------------------------------------------------------------------- */
|
|
|
|
void FixNVTSlodd::final_integrate()
|
|
{
|
|
double dtfm;
|
|
|
|
// update v of only atoms in group
|
|
// remove and restore bias = streaming velocity = Hrate*lamda + Hratelo
|
|
// thermostat thermal velocity only
|
|
// vdelu = SLLOD correction = Hrate*Hinv*vthermal
|
|
|
|
double **v = atom->v;
|
|
double **f = atom->f;
|
|
double *mass = atom->mass;
|
|
int *type = atom->type;
|
|
int *mask = atom->mask;
|
|
int nlocal = atom->nlocal;
|
|
if (igroup == atom->firstgroup) nlocal = atom->nfirst;
|
|
|
|
double h_two[6],vdelu[3];
|
|
MathExtra::multiply_shape_shape(domain->h_rate,domain->h_inv,h_two);
|
|
|
|
for (int i = 0; i < nlocal; i++) {
|
|
if (mask[i] & groupbit) {
|
|
temperature->remove_bias(i,v[i]);
|
|
vdelu[0] = h_two[0]*v[i][0] + h_two[5]*v[i][1] + h_two[4]*v[i][2];
|
|
vdelu[1] = h_two[1]*v[i][1] + h_two[3]*v[i][2];
|
|
vdelu[2] = h_two[2]*v[i][2];
|
|
//dtfm = dtf / mass[type[i]] * factor;
|
|
dtfm = dtf / mass[type[i]];
|
|
v[i][0] = v[i][0]*factor + dtfm*f[i][0] - dthalf*vdelu[0];
|
|
v[i][1] = v[i][1]*factor + dtfm*f[i][1] - dthalf*vdelu[1];
|
|
v[i][2] = v[i][2]*factor + dtfm*f[i][2] - dthalf*vdelu[2];
|
|
temperature->restore_bias(i,v[i]);
|
|
}
|
|
}
|
|
|
|
// compute current T
|
|
|
|
t_current = temperature->compute_scalar();
|
|
|
|
// update eta_dot
|
|
|
|
f_eta = t_freq*t_freq * (t_current/t_target - 1.0);
|
|
eta_dot += f_eta*dthalf;
|
|
eta_dot *= drag_factor;
|
|
}
|
|
|
|
/* ---------------------------------------------------------------------- */
|
|
|
|
void FixNVTSlodd::initial_integrate_respa(int vflag, int ilevel, int flag)
|
|
{
|
|
if (flag) return; // only used by NPT,NPH
|
|
|
|
// set timesteps by level
|
|
|
|
double dtfm;
|
|
dtv = step_respa[ilevel];
|
|
dtf = 0.5 * step_respa[ilevel] * force->ftm2v;
|
|
dthalf = 0.5 * step_respa[ilevel];
|
|
|
|
// atom quantities
|
|
|
|
double **x = atom->x;
|
|
double **v = atom->v;
|
|
double **f = atom->f;
|
|
double *mass = atom->mass;
|
|
int *type = atom->type;
|
|
int *mask = atom->mask;
|
|
int nlocal = atom->nlocal;
|
|
if (igroup == atom->firstgroup) nlocal = atom->nfirst;
|
|
|
|
// outermost level - update eta_dot and apply to v
|
|
// all other levels - NVE update of v
|
|
|
|
if (ilevel == nlevels_respa-1) {
|
|
double delta = update->ntimestep - update->firststep;
|
|
delta /= update->nsteps;
|
|
t_target = t_start + delta * (t_stop-t_start);
|
|
|
|
// update eta_dot
|
|
|
|
f_eta = t_freq*t_freq * (t_current/t_target - 1.0);
|
|
eta_dot += f_eta*dthalf;
|
|
eta_dot *= drag_factor;
|
|
eta += dtv*eta_dot;
|
|
factor = exp(-dthalf*eta_dot);
|
|
} else factor = 1.0;
|
|
|
|
// update v of only atoms in group
|
|
|
|
double h_two[6],vdelu[3];
|
|
MathExtra::multiply_shape_shape(domain->h_rate,domain->h_inv,h_two);
|
|
|
|
for (int i = 0; i < nlocal; i++) {
|
|
if (mask[i] & groupbit) {
|
|
temperature->remove_bias(i,v[i]);
|
|
vdelu[0] = h_two[0]*v[i][0] + h_two[5]*v[i][1] + h_two[4]*v[i][2];
|
|
vdelu[1] = h_two[1]*v[i][1] + h_two[3]*v[i][2];
|
|
vdelu[2] = h_two[2]*v[i][2];
|
|
dtfm = dtf / mass[type[i]];
|
|
v[i][0] = v[i][0]*factor + dtfm*f[i][0] - dthalf*vdelu[0];
|
|
v[i][1] = v[i][1]*factor + dtfm*f[i][1] - dthalf*vdelu[1];
|
|
v[i][2] = v[i][2]*factor + dtfm*f[i][2] - dthalf*vdelu[2];
|
|
temperature->restore_bias(i,v[i]);
|
|
}
|
|
}
|
|
|
|
// innermost level - also update x of only atoms in group
|
|
|
|
if (ilevel == 0) {
|
|
for (int i = 0; i < nlocal; i++) {
|
|
if (mask[i] & groupbit) {
|
|
x[i][0] += dtv * v[i][0];
|
|
x[i][1] += dtv * v[i][1];
|
|
x[i][2] += dtv * v[i][2];
|
|
}
|
|
}
|
|
}
|
|
}
|