From 6d4435a9bc21ae9bc0f3b547325e00a5817be484 Mon Sep 17 00:00:00 2001 From: sjplimp Date: Fri, 20 Apr 2007 23:59:34 +0000 Subject: [PATCH] git-svn-id: svn://svn.icms.temple.edu/lammps-ro/trunk@514 f3b2605a-c512-4ea7-a41b-209d697bcdaa --- doc/displace_atoms.html | 19 ++++++++++++++----- doc/displace_atoms.txt | 19 ++++++++++++++----- doc/fix_heat.html | 24 +++++++++++++----------- doc/fix_heat.txt | 24 +++++++++++++----------- 4 files changed, 54 insertions(+), 32 deletions(-) diff --git a/doc/displace_atoms.html b/doc/displace_atoms.html index 8b333beb3d..b33ea38965 100644 --- a/doc/displace_atoms.html +++ b/doc/displace_atoms.html @@ -63,11 +63,20 @@ spacings. The lattice command must have been previously used to define the lattice spacing.

Care should be taken not to move atoms on top of other atoms. After -the move, atoms are remapped to the periodic simulation box. In -parallel, atoms should not be moved so far that they cross more than -one processor's sub-domain, else they may be lost. If this is a -problem, successive displace_atom commands can be used to move atoms a -large distance. +the move, atoms are remapped into the periodic simulation box if +needed. +

+

Atoms can be moved arbitrarily long distances by this command. +However if the box is non-periodic, this can change the shape of the +simulation box. This is not a problem, except that the mapping of +processors to the simulation box is not changed from its initial 3d +configuration by this command. In other words, the number of +processors assigned to each dimension of the simulation box so as to +form a topological 3d grid of processors is not changed; see the +processors command. Thus if the box shape changes +dramatically, the simulation may not be as well load-balanced (atoms +per processor) as the initial mapping tried to achieve. This can be +adjusted for by using the processors command.

Restrictions:

diff --git a/doc/displace_atoms.txt b/doc/displace_atoms.txt index ad941edf35..33f2c51abd 100644 --- a/doc/displace_atoms.txt +++ b/doc/displace_atoms.txt @@ -55,11 +55,20 @@ spacings. The "lattice"_lattice.html command must have been previously used to define the lattice spacing. Care should be taken not to move atoms on top of other atoms. After -the move, atoms are remapped to the periodic simulation box. In -parallel, atoms should not be moved so far that they cross more than -one processor's sub-domain, else they may be lost. If this is a -problem, successive displace_atom commands can be used to move atoms a -large distance. +the move, atoms are remapped into the periodic simulation box if +needed. + +Atoms can be moved arbitrarily long distances by this command. +However if the box is non-periodic, this can change the shape of the +simulation box. This is not a problem, except that the mapping of +processors to the simulation box is not changed from its initial 3d +configuration by this command. In other words, the number of +processors assigned to each dimension of the simulation box so as to +form a topological 3d grid of processors is not changed; see the +"processors"_processors.html command. Thus if the box shape changes +dramatically, the simulation may not be as well load-balanced (atoms +per processor) as the initial mapping tried to achieve. This can be +adjusted for by using the "processors"_processors.html command. [Restrictions:] diff --git a/doc/fix_heat.html b/doc/fix_heat.html index 2ae0da7579..70229dc08b 100644 --- a/doc/fix_heat.html +++ b/doc/fix_heat.html @@ -27,20 +27,22 @@ fix 4 qout heat 1 -1.0

Description:

-

Add non-translational kinetic energy (heat) to the a group of atoms -such that their aggregate momentum is conserved. Two of these fixes -can be used to establish a temperature gradient across a simulation -domain by adding heat to one group of atoms (hot reservoir) and -subracting heat from another (cold reservoir). E.g. a simulation -sampling from the McDLT ensemble. Note that the fix is applied to a -group of atoms, not a geometric region, so that the same set of atoms -is affected wherever they may move to. +

Add non-translational kinetic energy (heat) to a group of atoms such +that their aggregate momentum is conserved. Two of these fixes can be +used to establish a temperature gradient across a simulation domain by +adding heat to one group of atoms (hot reservoir) and subracting heat +from another (cold reservoir). E.g. a simulation sampling from the +McDLT ensemble. Note that the fix is applied to a group of atoms, not +a geometric region, so that the same set of atoms is affected wherever +they may move to.

Heat addition/subtraction is performed every N timesteps. The eflux parameter determines the change in aggregate energy of the entire -group of atoms. If heat is subtracted from the system too -aggressively so that the group's kinetic energy goes to zero, LAMMPS -halts with an error message. +group of atoms. Since eflux is in units of energy/time, this means a +larger value of N will add/subract a larger amount of energy each +timestep the fix is invoked. If heat is subtracted from the system +too aggressively so that the group's kinetic energy goes to zero, +LAMMPS halts with an error message.

Fix heat is different from a thermostat such as fix nvt or fix temp/rescale in that energy is diff --git a/doc/fix_heat.txt b/doc/fix_heat.txt index ce6ffc6b41..e0abba0a44 100644 --- a/doc/fix_heat.txt +++ b/doc/fix_heat.txt @@ -24,20 +24,22 @@ fix 4 qout heat 1 -1.0 :pre [Description:] -Add non-translational kinetic energy (heat) to the a group of atoms -such that their aggregate momentum is conserved. Two of these fixes -can be used to establish a temperature gradient across a simulation -domain by adding heat to one group of atoms (hot reservoir) and -subracting heat from another (cold reservoir). E.g. a simulation -sampling from the McDLT ensemble. Note that the fix is applied to a -group of atoms, not a geometric region, so that the same set of atoms -is affected wherever they may move to. +Add non-translational kinetic energy (heat) to a group of atoms such +that their aggregate momentum is conserved. Two of these fixes can be +used to establish a temperature gradient across a simulation domain by +adding heat to one group of atoms (hot reservoir) and subracting heat +from another (cold reservoir). E.g. a simulation sampling from the +McDLT ensemble. Note that the fix is applied to a group of atoms, not +a geometric region, so that the same set of atoms is affected wherever +they may move to. Heat addition/subtraction is performed every N timesteps. The {eflux} parameter determines the change in aggregate energy of the entire -group of atoms. If heat is subtracted from the system too -aggressively so that the group's kinetic energy goes to zero, LAMMPS -halts with an error message. +group of atoms. Since eflux is in units of energy/time, this means a +larger value of N will add/subract a larger amount of energy each +timestep the fix is invoked. If heat is subtracted from the system +too aggressively so that the group's kinetic energy goes to zero, +LAMMPS halts with an error message. Fix heat is different from a thermostat such as "fix nvt"_fix_nvt.html or "fix temp/rescale"_fix_temp_rescale.html in that energy is