forked from lijiext/lammps
git-svn-id: svn://svn.icms.temple.edu/lammps-ro/trunk@3032 f3b2605a-c512-4ea7-a41b-209d697bcdaa
This commit is contained in:
parent
516081334a
commit
631f5b5f63
|
@ -11,6 +11,10 @@
|
||||||
See the README file in the top-level LAMMPS directory.
|
See the README file in the top-level LAMMPS directory.
|
||||||
------------------------------------------------------------------------- */
|
------------------------------------------------------------------------- */
|
||||||
|
|
||||||
|
/* ----------------------------------------------------------------------
|
||||||
|
Contributing author: Carolyn Phillips (U Mich), reservoir energy tally
|
||||||
|
------------------------------------------------------------------------- */
|
||||||
|
|
||||||
#include "mpi.h"
|
#include "mpi.h"
|
||||||
#include "math.h"
|
#include "math.h"
|
||||||
#include "string.h"
|
#include "string.h"
|
||||||
|
@ -26,12 +30,12 @@
|
||||||
#include "respa.h"
|
#include "respa.h"
|
||||||
#include "comm.h"
|
#include "comm.h"
|
||||||
#include "random_mars.h"
|
#include "random_mars.h"
|
||||||
|
#include "memory.h"
|
||||||
#include "error.h"
|
#include "error.h"
|
||||||
|
|
||||||
using namespace LAMMPS_NS;
|
using namespace LAMMPS_NS;
|
||||||
|
|
||||||
enum{NOBIAS,BIAS};
|
enum{NOBIAS,BIAS};
|
||||||
enum{MASS,RMASS};
|
|
||||||
|
|
||||||
/* ---------------------------------------------------------------------- */
|
/* ---------------------------------------------------------------------- */
|
||||||
|
|
||||||
|
@ -41,6 +45,10 @@ FixLangevin::FixLangevin(LAMMPS *lmp, int narg, char **arg) :
|
||||||
if (narg < 7) error->all("Illegal fix langevin command");
|
if (narg < 7) error->all("Illegal fix langevin command");
|
||||||
|
|
||||||
time_depend = 1;
|
time_depend = 1;
|
||||||
|
scalar_flag = 1;
|
||||||
|
scalar_vector_freq = 1;
|
||||||
|
extscalar = 1;
|
||||||
|
nevery = 1;
|
||||||
|
|
||||||
t_start = atof(arg[3]);
|
t_start = atof(arg[3]);
|
||||||
t_stop = atof(arg[4]);
|
t_stop = atof(arg[4]);
|
||||||
|
@ -63,6 +71,7 @@ FixLangevin::FixLangevin(LAMMPS *lmp, int narg, char **arg) :
|
||||||
// optional args
|
// optional args
|
||||||
|
|
||||||
for (int i = 1; i <= atom->ntypes; i++) ratio[i] = 1.0;
|
for (int i = 1; i <= atom->ntypes; i++) ratio[i] = 1.0;
|
||||||
|
tally = 0;
|
||||||
|
|
||||||
int iarg = 7;
|
int iarg = 7;
|
||||||
while (iarg < narg) {
|
while (iarg < narg) {
|
||||||
|
@ -74,6 +83,12 @@ FixLangevin::FixLangevin(LAMMPS *lmp, int narg, char **arg) :
|
||||||
error->all("Illegal fix langevin command");
|
error->all("Illegal fix langevin command");
|
||||||
ratio[itype] = scale;
|
ratio[itype] = scale;
|
||||||
iarg += 3;
|
iarg += 3;
|
||||||
|
} else if (strcmp(arg[iarg],"tally") == 0) {
|
||||||
|
if (iarg+2 > narg) error->all("Illegal fix langevin command");
|
||||||
|
if (strcmp(arg[iarg+1],"no") == 0) tally = 0;
|
||||||
|
else if (strcmp(arg[iarg+1],"yes") == 0) tally = 1;
|
||||||
|
else error->all("Illegal fix indent command");
|
||||||
|
iarg += 2;
|
||||||
} else error->all("Illegal fix langevin command");
|
} else error->all("Illegal fix langevin command");
|
||||||
}
|
}
|
||||||
|
|
||||||
|
@ -81,6 +96,10 @@ FixLangevin::FixLangevin(LAMMPS *lmp, int narg, char **arg) :
|
||||||
|
|
||||||
id_temp = NULL;
|
id_temp = NULL;
|
||||||
temperature = NULL;
|
temperature = NULL;
|
||||||
|
|
||||||
|
flangevin = NULL;
|
||||||
|
nmax = 0;
|
||||||
|
energy = 0.0;
|
||||||
}
|
}
|
||||||
|
|
||||||
/* ---------------------------------------------------------------------- */
|
/* ---------------------------------------------------------------------- */
|
||||||
|
@ -92,6 +111,7 @@ FixLangevin::~FixLangevin()
|
||||||
delete [] gfactor2;
|
delete [] gfactor2;
|
||||||
delete [] ratio;
|
delete [] ratio;
|
||||||
delete [] id_temp;
|
delete [] id_temp;
|
||||||
|
memory->destroy_2d_double_array(flangevin);
|
||||||
}
|
}
|
||||||
|
|
||||||
/* ---------------------------------------------------------------------- */
|
/* ---------------------------------------------------------------------- */
|
||||||
|
@ -101,6 +121,8 @@ int FixLangevin::setmask()
|
||||||
int mask = 0;
|
int mask = 0;
|
||||||
mask |= POST_FORCE;
|
mask |= POST_FORCE;
|
||||||
mask |= POST_FORCE_RESPA;
|
mask |= POST_FORCE_RESPA;
|
||||||
|
mask |= END_OF_STEP;
|
||||||
|
mask |= THERMO_ENERGY;
|
||||||
return mask;
|
return mask;
|
||||||
}
|
}
|
||||||
|
|
||||||
|
@ -110,7 +132,7 @@ void FixLangevin::init()
|
||||||
{
|
{
|
||||||
// set force prefactors
|
// set force prefactors
|
||||||
|
|
||||||
if (atom->mass) {
|
if (!atom->rmass) {
|
||||||
for (int i = 1; i <= atom->ntypes; i++) {
|
for (int i = 1; i <= atom->ntypes; i++) {
|
||||||
gfactor1[i] = -atom->mass[i] / t_period / force->ftm2v;
|
gfactor1[i] = -atom->mass[i] / t_period / force->ftm2v;
|
||||||
gfactor2[i] = sqrt(atom->mass[i]) *
|
gfactor2[i] = sqrt(atom->mass[i]) *
|
||||||
|
@ -124,9 +146,6 @@ void FixLangevin::init()
|
||||||
if (temperature && temperature->tempbias) which = BIAS;
|
if (temperature && temperature->tempbias) which = BIAS;
|
||||||
else which = NOBIAS;
|
else which = NOBIAS;
|
||||||
|
|
||||||
if (atom->mass) massflag = MASS;
|
|
||||||
else massflag = RMASS;
|
|
||||||
|
|
||||||
if (strcmp(update->integrate_style,"respa") == 0)
|
if (strcmp(update->integrate_style,"respa") == 0)
|
||||||
nlevels_respa = ((Respa *) update->integrate)->nlevels;
|
nlevels_respa = ((Respa *) update->integrate)->nlevels;
|
||||||
}
|
}
|
||||||
|
@ -147,11 +166,27 @@ void FixLangevin::setup(int vflag)
|
||||||
/* ---------------------------------------------------------------------- */
|
/* ---------------------------------------------------------------------- */
|
||||||
|
|
||||||
void FixLangevin::post_force(int vflag)
|
void FixLangevin::post_force(int vflag)
|
||||||
|
{
|
||||||
|
if (tally) post_force_tally();
|
||||||
|
else post_force_no_tally();
|
||||||
|
}
|
||||||
|
|
||||||
|
/* ---------------------------------------------------------------------- */
|
||||||
|
|
||||||
|
void FixLangevin::post_force_respa(int vflag, int ilevel, int iloop)
|
||||||
|
{
|
||||||
|
if (ilevel == nlevels_respa-1) post_force(vflag);
|
||||||
|
}
|
||||||
|
|
||||||
|
/* ---------------------------------------------------------------------- */
|
||||||
|
|
||||||
|
void FixLangevin::post_force_no_tally()
|
||||||
{
|
{
|
||||||
double gamma1,gamma2;
|
double gamma1,gamma2;
|
||||||
|
|
||||||
double **v = atom->v;
|
double **v = atom->v;
|
||||||
double **f = atom->f;
|
double **f = atom->f;
|
||||||
|
double *rmass = atom->rmass;
|
||||||
int *type = atom->type;
|
int *type = atom->type;
|
||||||
int *mask = atom->mask;
|
int *mask = atom->mask;
|
||||||
int nlocal = atom->nlocal;
|
int nlocal = atom->nlocal;
|
||||||
|
@ -163,42 +198,7 @@ void FixLangevin::post_force(int vflag)
|
||||||
|
|
||||||
// apply damping and thermostat to appropriate atoms
|
// apply damping and thermostat to appropriate atoms
|
||||||
|
|
||||||
if (massflag == MASS) {
|
if (rmass) {
|
||||||
if (which == NOBIAS) {
|
|
||||||
for (int i = 0; i < nlocal; i++) {
|
|
||||||
if (mask[i] & groupbit) {
|
|
||||||
gamma1 = gfactor1[type[i]];
|
|
||||||
gamma2 = gfactor2[type[i]] * tsqrt;
|
|
||||||
f[i][0] += gamma1*v[i][0] + gamma2*(random->uniform()-0.5);
|
|
||||||
f[i][1] += gamma1*v[i][1] + gamma2*(random->uniform()-0.5);
|
|
||||||
f[i][2] += gamma1*v[i][2] + gamma2*(random->uniform()-0.5);
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// invoke temperature since some computes require it to remove bias
|
|
||||||
// test v = 0 since some computes mask non-participating atoms via v = 0
|
|
||||||
|
|
||||||
} else if (which == BIAS) {
|
|
||||||
double tmp = temperature->compute_scalar();
|
|
||||||
|
|
||||||
for (int i = 0; i < nlocal; i++) {
|
|
||||||
if (mask[i] & groupbit) {
|
|
||||||
gamma1 = gfactor1[type[i]];
|
|
||||||
gamma2 = gfactor2[type[i]] * tsqrt;
|
|
||||||
temperature->remove_bias(i,v[i]);
|
|
||||||
if (v[i][0] != 0.0)
|
|
||||||
f[i][0] += gamma1*v[i][0] + gamma2*(random->uniform()-0.5);
|
|
||||||
if (v[i][1] != 0.0)
|
|
||||||
f[i][1] += gamma1*v[i][1] + gamma2*(random->uniform()-0.5);
|
|
||||||
if (v[i][2] != 0.0)
|
|
||||||
f[i][2] += gamma1*v[i][2] + gamma2*(random->uniform()-0.5);
|
|
||||||
temperature->restore_bias(i,v[i]);
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
} else {
|
|
||||||
double *rmass = atom->rmass;
|
|
||||||
double boltz = force->boltz;
|
double boltz = force->boltz;
|
||||||
double dt = update->dt;
|
double dt = update->dt;
|
||||||
double mvv2e = force->mvv2e;
|
double mvv2e = force->mvv2e;
|
||||||
|
@ -240,14 +240,182 @@ void FixLangevin::post_force(int vflag)
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
} else {
|
||||||
|
if (which == NOBIAS) {
|
||||||
|
for (int i = 0; i < nlocal; i++) {
|
||||||
|
if (mask[i] & groupbit) {
|
||||||
|
gamma1 = gfactor1[type[i]];
|
||||||
|
gamma2 = gfactor2[type[i]] * tsqrt;
|
||||||
|
f[i][0] += gamma1*v[i][0] + gamma2*(random->uniform()-0.5);
|
||||||
|
f[i][1] += gamma1*v[i][1] + gamma2*(random->uniform()-0.5);
|
||||||
|
f[i][2] += gamma1*v[i][2] + gamma2*(random->uniform()-0.5);
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
// invoke temperature since some computes require it to remove bias
|
||||||
|
// test v = 0 since some computes mask non-participating atoms via v = 0
|
||||||
|
|
||||||
|
} else if (which == BIAS) {
|
||||||
|
double tmp = temperature->compute_scalar();
|
||||||
|
|
||||||
|
for (int i = 0; i < nlocal; i++) {
|
||||||
|
if (mask[i] & groupbit) {
|
||||||
|
gamma1 = gfactor1[type[i]];
|
||||||
|
gamma2 = gfactor2[type[i]] * tsqrt;
|
||||||
|
temperature->remove_bias(i,v[i]);
|
||||||
|
if (v[i][0] != 0.0)
|
||||||
|
f[i][0] += gamma1*v[i][0] + gamma2*(random->uniform()-0.5);
|
||||||
|
if (v[i][1] != 0.0)
|
||||||
|
f[i][1] += gamma1*v[i][1] + gamma2*(random->uniform()-0.5);
|
||||||
|
if (v[i][2] != 0.0)
|
||||||
|
f[i][2] += gamma1*v[i][2] + gamma2*(random->uniform()-0.5);
|
||||||
|
temperature->restore_bias(i,v[i]);
|
||||||
|
}
|
||||||
|
}
|
||||||
|
}
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
/* ---------------------------------------------------------------------- */
|
/* ---------------------------------------------------------------------- */
|
||||||
|
|
||||||
void FixLangevin::post_force_respa(int vflag, int ilevel, int iloop)
|
void FixLangevin::post_force_tally()
|
||||||
{
|
{
|
||||||
if (ilevel == nlevels_respa-1) post_force(vflag);
|
double gamma1,gamma2;
|
||||||
|
|
||||||
|
// reallocate flangevin if necessary
|
||||||
|
|
||||||
|
if (atom->nmax > nmax) {
|
||||||
|
memory->destroy_2d_double_array(flangevin);
|
||||||
|
nmax = atom->nmax;
|
||||||
|
flangevin = memory->create_2d_double_array(nmax,3,"langevin:flangevin");
|
||||||
|
}
|
||||||
|
|
||||||
|
double **v = atom->v;
|
||||||
|
double **f = atom->f;
|
||||||
|
double *rmass = atom->rmass;
|
||||||
|
int *type = atom->type;
|
||||||
|
int *mask = atom->mask;
|
||||||
|
int nlocal = atom->nlocal;
|
||||||
|
|
||||||
|
double delta = update->ntimestep - update->beginstep;
|
||||||
|
delta /= update->endstep - update->beginstep;
|
||||||
|
double t_target = t_start + delta * (t_stop-t_start);
|
||||||
|
double tsqrt = sqrt(t_target);
|
||||||
|
|
||||||
|
// apply damping and thermostat to appropriate atoms
|
||||||
|
|
||||||
|
if (rmass) {
|
||||||
|
double boltz = force->boltz;
|
||||||
|
double dt = update->dt;
|
||||||
|
double mvv2e = force->mvv2e;
|
||||||
|
double ftm2v = force->ftm2v;
|
||||||
|
|
||||||
|
if (which == NOBIAS) {
|
||||||
|
for (int i = 0; i < nlocal; i++) {
|
||||||
|
if (mask[i] & groupbit) {
|
||||||
|
gamma1 = -rmass[i] / t_period / ftm2v;
|
||||||
|
gamma2 = sqrt(rmass[i]) * sqrt(24.0*boltz/t_period/dt/mvv2e) / ftm2v;
|
||||||
|
gamma1 *= 1.0/ratio[type[i]];
|
||||||
|
gamma2 *= 1.0/sqrt(ratio[type[i]]) * tsqrt;
|
||||||
|
flangevin[i][0] = gamma1*v[i][0] + gamma2*(random->uniform()-0.5);
|
||||||
|
flangevin[i][1] = gamma1*v[i][1] + gamma2*(random->uniform()-0.5);
|
||||||
|
flangevin[i][2] = gamma1*v[i][2] + gamma2*(random->uniform()-0.5);
|
||||||
|
f[i][0] += flangevin[i][0];
|
||||||
|
f[i][1] += flangevin[i][1];
|
||||||
|
f[i][2] += flangevin[i][2];
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
// invoke temperature since some computes require it to remove bias
|
||||||
|
// test v = 0 since some computes mask non-participating atoms via v = 0
|
||||||
|
|
||||||
|
} else if (which == BIAS) {
|
||||||
|
double tmp = temperature->compute_scalar();
|
||||||
|
|
||||||
|
for (int i = 0; i < nlocal; i++) {
|
||||||
|
if (mask[i] & groupbit) {
|
||||||
|
gamma1 = -rmass[i] / t_period / ftm2v;
|
||||||
|
gamma2 = sqrt(rmass[i]) * sqrt(24.0*boltz/t_period/dt/mvv2e) / ftm2v;
|
||||||
|
gamma1 *= 1.0/ratio[type[i]];
|
||||||
|
gamma2 *= 1.0/sqrt(ratio[type[i]]) * tsqrt;
|
||||||
|
temperature->remove_bias(i,v[i]);
|
||||||
|
flangevin[i][0] = gamma1*v[i][0] + gamma2*(random->uniform()-0.5);
|
||||||
|
flangevin[i][1] = gamma1*v[i][1] + gamma2*(random->uniform()-0.5);
|
||||||
|
flangevin[i][2] = gamma1*v[i][2] + gamma2*(random->uniform()-0.5);
|
||||||
|
if (v[i][0] != 0.0) f[i][0] += flangevin[i][0];
|
||||||
|
else flangevin[i][0] = 0;
|
||||||
|
if (v[i][1] != 0.0) f[i][1] += flangevin[i][1];
|
||||||
|
else flangevin[i][1] = 0;
|
||||||
|
if (v[i][2] != 0.0) f[i][2] += flangevin[i][2];
|
||||||
|
else flangevin[i][2] = 0;
|
||||||
|
temperature->restore_bias(i,v[i]);
|
||||||
|
}
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
} else {
|
||||||
|
if (which == NOBIAS) {
|
||||||
|
for (int i = 0; i < nlocal; i++) {
|
||||||
|
if (mask[i] & groupbit) {
|
||||||
|
gamma1 = gfactor1[type[i]];
|
||||||
|
gamma2 = gfactor2[type[i]] * tsqrt;
|
||||||
|
flangevin[i][0] = gamma1*v[i][0] + gamma2*(random->uniform()-0.5);
|
||||||
|
flangevin[i][1] = gamma1*v[i][1] + gamma2*(random->uniform()-0.5);
|
||||||
|
flangevin[i][2] = gamma1*v[i][2] + gamma2*(random->uniform()-0.5);
|
||||||
|
f[i][0] += flangevin[i][0];
|
||||||
|
f[i][1] += flangevin[i][1];
|
||||||
|
f[i][2] += flangevin[i][2];
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
// invoke temperature since some computes require it to remove bias
|
||||||
|
// test v = 0 since some computes mask non-participating atoms via v = 0
|
||||||
|
|
||||||
|
} else if (which == BIAS) {
|
||||||
|
double tmp = temperature->compute_scalar();
|
||||||
|
|
||||||
|
for (int i = 0; i < nlocal; i++) {
|
||||||
|
if (mask[i] & groupbit) {
|
||||||
|
gamma1 = gfactor1[type[i]];
|
||||||
|
gamma2 = gfactor2[type[i]] * tsqrt;
|
||||||
|
temperature->remove_bias(i,v[i]);
|
||||||
|
flangevin[i][0] = gamma1*v[i][0] + gamma2*(random->uniform()-0.5);
|
||||||
|
flangevin[i][1] = gamma1*v[i][1] + gamma2*(random->uniform()-0.5);
|
||||||
|
flangevin[i][2] = gamma1*v[i][2] + gamma2*(random->uniform()-0.5);
|
||||||
|
if (v[i][0] != 0.0) f[i][0] += flangevin[i][0];
|
||||||
|
else flangevin[i][0] = 0.0;
|
||||||
|
if (v[i][1] != 0.0) f[i][1] += flangevin[i][1];
|
||||||
|
else flangevin[i][1] = 0.0;
|
||||||
|
if (v[i][2] != 0.0) f[i][2] += flangevin[i][2];
|
||||||
|
else flangevin[i][2] = 0.0;
|
||||||
|
temperature->restore_bias(i,v[i]);
|
||||||
|
}
|
||||||
|
}
|
||||||
|
}
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
/* ----------------------------------------------------------------------
|
||||||
|
tally energy transfer to thermal reservoir
|
||||||
|
------------------------------------------------------------------------- */
|
||||||
|
|
||||||
|
void FixLangevin::end_of_step()
|
||||||
|
{
|
||||||
|
if (!tally) return;
|
||||||
|
|
||||||
|
double **v = atom->v;
|
||||||
|
int *mask = atom->mask;
|
||||||
|
int nlocal = atom->nlocal;
|
||||||
|
|
||||||
|
energy_onestep = 0.0;
|
||||||
|
|
||||||
|
for (int i = 0; i < nlocal; i++)
|
||||||
|
if (mask[i] & groupbit)
|
||||||
|
energy_onestep += flangevin[i][0]*v[i][0] + flangevin[i][1]*v[i][1] +
|
||||||
|
flangevin[i][2]*v[i][2];
|
||||||
|
|
||||||
|
energy += energy_onestep*update->dt;
|
||||||
}
|
}
|
||||||
|
|
||||||
/* ---------------------------------------------------------------------- */
|
/* ---------------------------------------------------------------------- */
|
||||||
|
@ -283,14 +451,53 @@ int FixLangevin::modify_param(int narg, char **arg)
|
||||||
strcpy(id_temp,arg[1]);
|
strcpy(id_temp,arg[1]);
|
||||||
|
|
||||||
int icompute = modify->find_compute(id_temp);
|
int icompute = modify->find_compute(id_temp);
|
||||||
if (icompute < 0) error->all("Could not find fix_modify temperature ID");
|
if (icompute < 0) error->all("Could not find fix_modify temp ID");
|
||||||
temperature = modify->compute[icompute];
|
temperature = modify->compute[icompute];
|
||||||
|
|
||||||
if (temperature->tempflag == 0)
|
if (temperature->tempflag == 0)
|
||||||
error->all("Fix_modify temperature ID does not compute temperature");
|
error->all("Fix_modify temp ID does not compute temperature");
|
||||||
if (temperature->igroup != igroup && comm->me == 0)
|
if (temperature->igroup != igroup && comm->me == 0)
|
||||||
error->warning("Group for fix_modify temp != fix group");
|
error->warning("Group for fix_modify temp != fix group");
|
||||||
return 2;
|
return 2;
|
||||||
}
|
}
|
||||||
return 0;
|
return 0;
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/* ---------------------------------------------------------------------- */
|
||||||
|
|
||||||
|
double FixLangevin::compute_scalar()
|
||||||
|
{
|
||||||
|
if (!tally) return 0.0;
|
||||||
|
|
||||||
|
// capture the very first energy transfer to thermal reservoir
|
||||||
|
|
||||||
|
double **v = atom->v;
|
||||||
|
int *mask = atom->mask;
|
||||||
|
int nlocal = atom->nlocal;
|
||||||
|
|
||||||
|
if (update->ntimestep == update->beginstep) {
|
||||||
|
energy_onestep = 0.0;
|
||||||
|
for (int i = 0; i < nlocal; i++)
|
||||||
|
if (mask[i] & groupbit)
|
||||||
|
energy_onestep += flangevin[i][0]*v[i][0] + flangevin[i][1]*v[i][1] +
|
||||||
|
flangevin[i][2]*v[i][2];
|
||||||
|
energy = 0.5*energy_onestep*update->dt;
|
||||||
|
}
|
||||||
|
|
||||||
|
double energy_me = energy - 0.5*energy_onestep*update->dt;
|
||||||
|
|
||||||
|
double energy_all;
|
||||||
|
MPI_Allreduce(&energy_me,&energy_all,1,MPI_DOUBLE,MPI_SUM,world);
|
||||||
|
return -energy_all;
|
||||||
|
}
|
||||||
|
|
||||||
|
/* ----------------------------------------------------------------------
|
||||||
|
memory usage of tally array
|
||||||
|
------------------------------------------------------------------------- */
|
||||||
|
|
||||||
|
double FixLangevin::memory_usage()
|
||||||
|
{
|
||||||
|
if (!tally) return 0.0;
|
||||||
|
double bytes = atom->nmax*3 * sizeof(double);
|
||||||
|
return bytes;
|
||||||
|
}
|
||||||
|
|
|
@ -27,20 +27,30 @@ class FixLangevin : public Fix {
|
||||||
void setup(int);
|
void setup(int);
|
||||||
void post_force(int);
|
void post_force(int);
|
||||||
void post_force_respa(int, int, int);
|
void post_force_respa(int, int, int);
|
||||||
|
void end_of_step();
|
||||||
void reset_target(double);
|
void reset_target(double);
|
||||||
void reset_dt();
|
void reset_dt();
|
||||||
int modify_param(int, char **);
|
int modify_param(int, char **);
|
||||||
|
double compute_scalar();
|
||||||
|
double memory_usage();
|
||||||
|
|
||||||
private:
|
private:
|
||||||
int which,massflag;
|
int which,tally;
|
||||||
double t_start,t_stop,t_period;
|
double t_start,t_stop,t_period;
|
||||||
double *gfactor1,*gfactor2,*ratio;
|
double *gfactor1,*gfactor2,*ratio;
|
||||||
|
double energy,energy_onestep;
|
||||||
|
|
||||||
|
int nmax;
|
||||||
|
double **flangevin;
|
||||||
|
|
||||||
char *id_temp;
|
char *id_temp;
|
||||||
class Compute *temperature;
|
class Compute *temperature;
|
||||||
|
|
||||||
int nlevels_respa;
|
int nlevels_respa;
|
||||||
class RanMars *random;
|
class RanMars *random;
|
||||||
|
|
||||||
|
void post_force_no_tally();
|
||||||
|
void post_force_tally();
|
||||||
};
|
};
|
||||||
|
|
||||||
}
|
}
|
||||||
|
|
Loading…
Reference in New Issue