The simulation box is divided into Nbin layers in the edim -direction. Every N steps, Nswap pairs of atoms are chosen in the -following manner. Only atoms in the fix group are considered. The -hottest Nswap atoms in the bottom layer are selected. Similarly, the -coldest Nswap atoms in the middle later are selected. The two sets of -Nswap atoms are paired up and their velocities are exchanged. This -effectively swaps their kinetic energies, assuming their masses are -the same. Over time, this induces a temperature gradient in the -system which can be measured using commands such as the following, -which writes the temperature profile (assuming z = edim) to the file -tmp.profile: +direction, where the layer 1 is at the low end of that dimension and +the layer Nbin is at the high end. Every N steps, Nswap pairs of +atoms are chosen in the following manner. Only atoms in the fix group +are considered. The hottest Nswap atoms in layer 1 are selected. +Similarly, the coldest Nswap atoms in the "middle" layer (see below) +are selected. The two sets of Nswap atoms are paired up and their +velocities are exchanged. This effectively swaps their kinetic +energies, assuming their masses are the same. Over time, this induces +a temperature gradient in the system which can be measured using +commands such as the following, which writes the temperature profile +(assuming z = edim) to the file tmp.profile:
compute ke all ke/atom variable temp atom c_ke/1.5 @@ -75,6 +76,13 @@ across a wide range of values, and the kinetic energy to be exchanged in large chunks or more smoothly. +The "middle" layer for velocity swapping is defined as the Nbin/2 + +1 layer. Thus if Nbin = 20, the two swapping layers are 1 and 11. +This should lead to a symmetric temperature profile since the two +layers are separated by the same distance in both directions in a +periodic sense. This is why Nbin is restricted to being an even +number. +
As described below, the total kinetic energy transferred by these swaps is computed by the fix and can be output. Dividing this quantity by time and the cross-sectional area of the simulation box diff -Naur lammps-25Jun08/doc/fix_thermal_conductivity.txt lammps-26Jun08/doc/fix_thermal_conductivity.txt --- lammps-25Jun08/doc/fix_thermal_conductivity.txt 2008-05-14 15:40:50.000000000 -0600 +++ lammps-26Jun08/doc/fix_thermal_conductivity.txt 2008-06-24 17:22:45.000000000 -0600 @@ -16,7 +16,7 @@ thermal/conductivity = style name of this fix command :l N = perform kinetic energy exchange every N steps :l edim = {x} or {y} or {z} = direction of kinetic energy transfer :l -Nbin = # of layers in edim direction :l +Nbin = # of layers in edim direction (must be even number) :l zero or more keyword/value pairs may be appended :l keyword = {swap} :l @@ -43,16 +43,17 @@ gradient is the system's response. The simulation box is divided into {Nbin} layers in the {edim} -direction. Every N steps, Nswap pairs of atoms are chosen in the -following manner. Only atoms in the fix group are considered. The -hottest Nswap atoms in the bottom layer are selected. Similarly, the -coldest Nswap atoms in the middle later are selected. The two sets of -Nswap atoms are paired up and their velocities are exchanged. This -effectively swaps their kinetic energies, assuming their masses are -the same. Over time, this induces a temperature gradient in the -system which can be measured using commands such as the following, -which writes the temperature profile (assuming z = edim) to the file -tmp.profile: +direction, where the layer 1 is at the low end of that dimension and +the layer {Nbin} is at the high end. Every N steps, Nswap pairs of +atoms are chosen in the following manner. Only atoms in the fix group +are considered. The hottest Nswap atoms in layer 1 are selected. +Similarly, the coldest Nswap atoms in the "middle" layer (see below) +are selected. The two sets of Nswap atoms are paired up and their +velocities are exchanged. This effectively swaps their kinetic +energies, assuming their masses are the same. Over time, this induces +a temperature gradient in the system which can be measured using +commands such as the following, which writes the temperature profile +(assuming z = edim) to the file tmp.profile: compute ke all ke/atom variable temp atom c_ke[]/1.5 @@ -65,6 +66,13 @@ across a wide range of values, and the kinetic energy to be exchanged in large chunks or more smoothly. +The "middle" layer for velocity swapping is defined as the {Nbin}/2 + +1 layer. Thus if {Nbin} = 20, the two swapping layers are 1 and 11. +This should lead to a symmetric temperature profile since the two +layers are separated by the same distance in both directions in a +periodic sense. This is why {Nbin} is restricted to being an even +number. + As described below, the total kinetic energy transferred by these swaps is computed by the fix and can be output. Dividing this quantity by time and the cross-sectional area of the simulation box diff -Naur lammps-25Jun08/doc/fix_viscosity.html lammps-26Jun08/doc/fix_viscosity.html --- lammps-25Jun08/doc/fix_viscosity.html 2008-05-15 13:42:47.000000000 -0600 +++ lammps-26Jun08/doc/fix_viscosity.html 2008-06-24 17:22:45.000000000 -0600 @@ -25,7 +25,7 @@
The simulation box is divided into Nbin layers in the pdim -direction. Every N steps, Nswap pairs of atoms are chosen in the -following manner. Only atoms in the fix group are considered. Nswap -atoms in the bottom layer with positive velocity components in the -vdim direction closest to the target value V are selected. -Similarly, Nswap atoms in the middle later with negative velocity -components in the vdim direction closest to the negative of the -target value V are selected. The two sets of Nswap atoms are paired -up and their vdim momenta components are swapped within each pair. -This resets their velocities, typically in opposite directions. Over -time, this induces a shear velocity profile in the system which can be -measured using commands such as the following, which writes the -profile to the file tmp.profile: +direction, where the layer 1 is at the low end of that dimension and +the layer Nbin is at the high end. Every N steps, Nswap pairs of +atoms are chosen in the following manner. Only atoms in the fix group +are considered. Nswap atoms in layer 1 with positive velocity +components in the vdim direction closest to the target value V are +selected. Similarly, Nswap atoms in the "middle" layer (see below) with +negative velocity components in the vdim direction closest to the +negative of the target value V are selected. The two sets of Nswap +atoms are paired up and their vdim momenta components are swapped +within each pair. This resets their velocities, typically in opposite +directions. Over time, this induces a shear velocity profile in the +system which can be measured using commands such as the following, +which writes the profile to the file tmp.profile:
fix f1 all ave/spatial 100 10 1000 z lower 0.05 vx &
file tmp.profile units reduced
@@ -81,6 +82,12 @@
adjusted across a wide range of values, and the momenta to be
exchanged in large chunks or more smoothly.
+The "middle" layer for momenta swapping is defined as the Nbin/2 + 1
+layer. Thus if Nbin = 20, the two swapping layers are 1 and 11.
+This should lead to a symmetric velocity profile since the two layers
+are separated by the same distance in both directions in a periodic
+sense. This is why Nbin is restricted to being an even number.
+
As described below, the total momentum transferred by these velocity
swaps is computed by the fix and can be output. Dividing this
quantity by time and the cross-sectional area of the simulation box
diff -Naur lammps-25Jun08/doc/fix_viscosity.txt lammps-26Jun08/doc/fix_viscosity.txt
--- lammps-25Jun08/doc/fix_viscosity.txt 2008-05-15 13:42:47.000000000 -0600
+++ lammps-26Jun08/doc/fix_viscosity.txt 2008-06-24 17:22:45.000000000 -0600
@@ -17,7 +17,7 @@
N = perform momentum exchange every N steps :l
vdim = {x} or {y} or {z} = which momentum component to exchange :l
pdim = {x} or {y} or {z} = direction of momentum transfer :l
-Nbin = # of layers in pdim direction :l
+Nbin = # of layers in pdim direction (must be even number) :l
zero or more keyword/value pairs may be appended :l
keyword = {swap} or {target} :l
@@ -46,18 +46,19 @@
system's response.
The simulation box is divided into {Nbin} layers in the {pdim}
-direction. Every N steps, Nswap pairs of atoms are chosen in the
-following manner. Only atoms in the fix group are considered. Nswap
-atoms in the bottom layer with positive velocity components in the
-{vdim} direction closest to the target value {V} are selected.
-Similarly, Nswap atoms in the middle later with negative velocity
-components in the {vdim} direction closest to the negative of the
-target value {V} are selected. The two sets of Nswap atoms are paired
-up and their {vdim} momenta components are swapped within each pair.
-This resets their velocities, typically in opposite directions. Over
-time, this induces a shear velocity profile in the system which can be
-measured using commands such as the following, which writes the
-profile to the file tmp.profile:
+direction, where the layer 1 is at the low end of that dimension and
+the layer {Nbin} is at the high end. Every N steps, Nswap pairs of
+atoms are chosen in the following manner. Only atoms in the fix group
+are considered. Nswap atoms in layer 1 with positive velocity
+components in the {vdim} direction closest to the target value {V} are
+selected. Similarly, Nswap atoms in the "middle" layer (see below) with
+negative velocity components in the {vdim} direction closest to the
+negative of the target value {V} are selected. The two sets of Nswap
+atoms are paired up and their {vdim} momenta components are swapped
+within each pair. This resets their velocities, typically in opposite
+directions. Over time, this induces a shear velocity profile in the
+system which can be measured using commands such as the following,
+which writes the profile to the file tmp.profile:
fix f1 all ave/spatial 100 10 1000 z lower 0.05 vx &
file tmp.profile units reduced :pre
@@ -70,6 +71,12 @@
adjusted across a wide range of values, and the momenta to be
exchanged in large chunks or more smoothly.
+The "middle" layer for momenta swapping is defined as the {Nbin}/2 + 1
+layer. Thus if {Nbin} = 20, the two swapping layers are 1 and 11.
+This should lead to a symmetric velocity profile since the two layers
+are separated by the same distance in both directions in a periodic
+sense. This is why {Nbin} is restricted to being an even number.
+
As described below, the total momentum transferred by these velocity
swaps is computed by the fix and can be output. Dividing this
quantity by time and the cross-sectional area of the simulation box
diff -Naur lammps-25Jun08/src/fix_thermal_conductivity.cpp lammps-26Jun08/src/fix_thermal_conductivity.cpp
--- lammps-25Jun08/src/fix_thermal_conductivity.cpp 2008-05-14 15:38:34.000000000 -0600
+++ lammps-26Jun08/src/fix_thermal_conductivity.cpp 2008-06-24 17:24:37.000000000 -0600
@@ -51,7 +51,8 @@
else error->all("Illegal fix thermal/conductivity command");
nbin = atoi(arg[5]);
- if (nbin < 3) error->all("Illegal fix thermal/conductivity command");
+ if (nbin % 2 || nbin <= 2)
+ error->all("Illegal fix thermal/conductivity command");
// optional keywords
@@ -105,8 +106,7 @@
// set bounds of 2 slabs in edim
// only necessary for static box, else re-computed in end_of_step()
// lo bin is always bottom bin
- // if nbin even, hi bin is just below half height
- // if nbin odd, hi bin straddles half height
+ // hi bin is just above half height
if (domain->box_change == 0) {
prd = domain->prd[edim];
@@ -115,8 +115,8 @@
double binsize = (boxhi-boxlo) / nbin;
slablo_lo = boxlo;
slablo_hi = boxlo + binsize;
- slabhi_lo = boxlo + ((nbin-1)/2)*binsize;
- slabhi_hi = boxlo + ((nbin-1)/2 + 1)*binsize;
+ slabhi_lo = boxlo + (nbin/2)*binsize;
+ slabhi_hi = boxlo + (nbin/2+1)*binsize;
}
periodicity = domain->periodicity[edim];
@@ -143,8 +143,8 @@
double binsize = (boxhi-boxlo) / nbin;
slablo_lo = boxlo;
slablo_hi = boxlo + binsize;
- slabhi_lo = boxlo + ((nbin-1)/2)*binsize;
- slabhi_hi = boxlo + ((nbin-1)/2 + 1)*binsize;
+ slabhi_lo = boxlo + (nbin/2)*binsize;
+ slabhi_hi = boxlo + (nbin/2+1)*binsize;
}
// make 2 lists of up to nswap atoms
diff -Naur lammps-25Jun08/src/fix_viscosity.cpp lammps-26Jun08/src/fix_viscosity.cpp
--- lammps-25Jun08/src/fix_viscosity.cpp 2008-05-15 16:42:47.000000000 -0600
+++ lammps-26Jun08/src/fix_viscosity.cpp 2008-06-24 17:24:37.000000000 -0600
@@ -56,7 +56,7 @@
else error->all("Illegal fix viscosity command");
nbin = atoi(arg[6]);
- if (nbin < 3) error->all("Illegal fix viscosity command");
+ if (nbin % 2 || nbin <= 2) error->all("Illegal fix viscosity command");
// optional keywords
@@ -116,8 +116,7 @@
// set bounds of 2 slabs in pdim
// only necessary for static box, else re-computed in end_of_step()
// lo bin is always bottom bin
- // if nbin even, hi bin is just below half height
- // if nbin odd, hi bin straddles half height
+ // hi bin is just above half height
if (domain->box_change == 0) {
prd = domain->prd[pdim];
@@ -126,8 +125,8 @@
double binsize = (boxhi-boxlo) / nbin;
slablo_lo = boxlo;
slablo_hi = boxlo + binsize;
- slabhi_lo = boxlo + ((nbin-1)/2)*binsize;
- slabhi_hi = boxlo + ((nbin-1)/2 + 1)*binsize;
+ slabhi_lo = boxlo + (nbin/2)*binsize;
+ slabhi_hi = boxlo + (nbin/2+1)*binsize;
}
periodicity = domain->periodicity[pdim];
@@ -154,8 +153,8 @@
double binsize = (boxhi-boxlo) / nbin;
slablo_lo = boxlo;
slablo_hi = boxlo + binsize;
- slabhi_lo = boxlo + ((nbin-1)/2)*binsize;
- slabhi_hi = boxlo + ((nbin-1)/2 + 1)*binsize;
+ slabhi_lo = boxlo + (nbin/2)*binsize;
+ slabhi_hi = boxlo + (nbin/2+1)*binsize;
}
// make 2 lists of up to nswap atoms with velocity closest to +/- vtarget