xmd-8.html

来自「一个很好的分子动力学程序」· HTML 代码 · 共 1,092 行 · 第 1/4 页

for xdisp, ydisp and zdisp can contain the variables x, y and z, which
will equal the coordinates of each atom moved.  Thus you can apply a
displacement that is a function of individual atoms.  The optional
formula expression can initialize one or more variables for subsequent
use in the expressions xdisp, ydisp and zdisp.  As an example, suppose
we want to use the MOVE command to displace atoms away from the point
(x0,y0,z0) by an amount exp(-r/5), where r is the distance from point
(x0,y0,z0).  We can use the following code to do this,
<P>
<BLOCKQUOTE><CODE>
<PRE>
CALC  x0=5.0
CALC  y0=5.0
CALC  z0=5.0
MOVE  dx=x-x0 dy=y-y0 dz=z-z0 r=sqrt(dx^2 + dy^2 + dz^2) &sol;
      dx*exp(-r/5) dy*exp(-r/5) dz*exp(-r/5)
</PRE>
</CODE></BLOCKQUOTE>
<P>The first three CALC statements initialize values for x0, y0 and z0,
which will serve as the center for the atomic displacements.  The MOVE
command initializes four variables, dx, dy, dz and r, which are then
used to specify the displacements.  Note that the last three expressions
are the actual displacements.  Also note the use of the backslash
character, &quot; &sol; &quot; .
When this character is the last character on a line, the command is
continued on the next line.  The backslash character may be used for any
number of lines, as long as the resulting command does not exceeded 1024
characters.  Typically this is about 12 to 20 lines.
<P>
<P>
<DT><B>NRANGE ratio</B><DD><P>Set's the range needed by the neighbor search algorithm (see page 5).
This neighbor range is longer than the interatomic potential range.
The neighbor list includes all particle pairs whose separation is ratio
times the maximum cutoff distance for the interatomic potentials (ionic
and electron density for various species).  By default ratio is 1.1.
For a given potential the optimum value (in terms of computer time used)
may vary.
<P>
<DT><B>PARTICLE [ADD] np</B><DD><P>See the 
<A HREF="#command-position">POSITION command</A>.
<P>
<P>
<P>
<DT><B>PLOT  BOND {  OFF  |  ON  lo  hi  }</B><DD><P>Plots &quot;bonds&quot; between atoms.  All atoms between lo and hi
angstroms apart will be connected by lines.
<P>
<DT><B>PLOT  DEVICE { CANON | POSTSCRIPT }</B><DD><P>Specifies the printer to be used for plot output.  The default depends
on the installation.
<P>
<DT><B>PLOT DISP { ON | OFF | SCALE scale }</B><DD><P>Switches on/off displacement plotting.  If REFSTEP has been specified
previously in a program, then when the PLOT statement is given, atomic
displacements will be plotted (if switched on). scale specifies a
magnification scale for the displacement lines.
<P>
<DT><B>PLOT LAYER nlayer [layer1 [layer2 ..... ]]</B><DD><P>Chooses the layers for particle plot. <CODE>nlayer</CODE> is the number of
layers into which the box is divided (in the direction specified by PLOT
ORIENT).  If parameters layer1, layer2, etc. are specified, then only
these layers are plotted.  Otherwise all layers are plotted.
<P>
<DT><B>PLOT ORIENT { X/Y | Y/X | Y/Z | Z/Y | X/Z | Z/X }</B><DD><P>Chooses the orientation of the plot.  The default is y/x (which means y
in the vertical direction and x in the horizontal).  Orientation y/x
means that the y direction is plotted vertically, and the x direction is
plotted horizontally.  The z direction is used to divide layers (see
PLOT LAYER above).
<P>
<DT><B>PLOT PAGE page</B><DD><P>Number of pages used for plot.  Default is one.
<P>
<DT><B>PLOT SYMBOL [FILL] [ RGB red green blue | HSB hue saturation brightness ]
{ AST | CIR | CRO | DIA | ITR | NON | SQR | TRI }   radius</B><DD><P>Assigns a symbol to all of the selected particles.  The FILL option
causes the symbol to print as a solid block (valid for the circles,
diamonds, triangles and squares, ignored otherwise).  The RGB and HSB
options assign a color to the atom (the default color is black).  See
the section 
<A HREF="xmd-6.html#implementation-plotcolor">Using Color in Plots</A>
to learn about the meaning of
the color parameters. radius specifies the distance from the center of
the symbol to the furthest point.  When no SYMBOL command is given then
SYMBOL CIR 0.25 is assumed.  A list of symbol options follows:
<P>
<BLOCKQUOTE><CODE>
<PRE>
AST asterisk
CIR circle
CRO cross
DIA diamond
ITR inverted triangle (point downward)
NON no symbol
NUM index number for particle
SQU square
TRI triangle
</PRE>
</CODE></BLOCKQUOTE>
<P>The option NUM prints the number of the particle in place of a symbol.
This is useful for finding the location of a specific particle.  The NON
option prints no symbol.  This is useful when you want to plot bonds,
but don't want to have a symbol.
<P>
<DT><B>PLOT WRITE [SEL] [[+] file ]</B><DD><P>Places printer commands (either Canon LBP or PostScript commands
depending on DEVICE command) describing the lattice plot in either the
output file (specified on the command line) or in file.  If a + is
prefixed to the file name then the output is appended file.  If the SEL
option is specified than only the selected particles are plotted.
Before PLOT is called symbols must first be assigned to the atoms (see
the SYMBOL command).  Also one may further embellish the plot by
specifying DISP, PAGE or LAYER.
<P>
<P>
<P>
<A NAME="command-position"></A> <DT><B>POSITION [ADD] np</B><DD><P>Reads in a set of particle types and coordinates.  With the ADD option,
previous selections are preserved and the new particles are added to the
selection list.  Without the ADD option, all previous selections are
forgotten
and all the new particles are
selected.  The parameter np is the number of particles for the program
to read.  For each particle enter its type and coordinates one a
separate line after the POSITION statement as illustrated below
(you can substitute PARTICLE for POSITION).
<BLOCKQUOTE><CODE>
<PRE>
POSITION 4
1  1/4 1/4 1/4
2  1/4 3/4 3/4
2  3/4 1/4 3/4
2  3/4 3/4 1/4
</PRE>
</CODE></BLOCKQUOTE>
<P>
<P>
<DT><B>POSVEL  [ADD] np</B><DD><P>Reads in a set of particle types, coordinates and velocities.  
Coordinate units are angstroms, velocity units are cm/sec.
With the ADD option,
previous selections are preserved and the new particles are added to the
selection list.  Without the ADD option, all previous selections are
forgotten
and all the new particles are
selected.  The parameter np is the number of particles for the program
to read.  For each particle enter its type and coordinates one a
separate line after the POSVEL statement as illustrated below.
<BLOCKQUOTE><CODE>
<PRE>
POSVEL 4
1  1/4 1/4 1/4    0   0 0
2  1/4 3/4 3/4    0   0 0
2  3/4 1/4 3/4   20.0 0 0
2  3/4 3/4 1/4  -20.0 0 0
</PRE>
</CODE></BLOCKQUOTE>
<P>
<P>
<P>
<DT><B>POTENTIAL SET { EAM ntype | PAIR ntype | STILL | TERSOFF }</B><DD><P>Sets the potential type.  <CODE>EAM</CODE> selects the Embedded Atom Method,
and <CODE>ntype</CODE> specifies the number of elements (or atom types)
used in the simulation.  Similarly for <CODE>PAIR</CODE>, it is used for
pair potentials.  Subsequent commands must specify the exact form
of the potential functions.  See the section on Interatomic Potentials.
<P>The <CODE>STILL</CODE> option selects the Stillinger-Weber potential for
Si, the <CODE>TERSOFF</CODE> option selects the Tersoff potential for
Carbon/Silcon.
<P>
<P>
<DT><B>PSTRAIN e dx dy dz nx ny nz</B><DD><P>Performs an invariant plane strain on the selected particles.  The new
particle coordinates are given by
<P>
<BLOCKQUOTE><CODE>
<PRE>
r' = r + e d (n &middot; r)
</PRE>
</CODE></BLOCKQUOTE>
<P>where d is the vector (dx,dy,dz) and n is (nx,ny,nz). r is the original
particle coordinate measured from the origin and r' is the new particle
coordinate.
<P>
<P>
<A NAME="command-pressure"></A> <DT><B>PRESSURE ANDERSEN  xmass [ ymass [ zmass ] ] </B><DD><P>
<DT><B>PRESSURE CLAMP     bulkmodulus [ cstep ] </B><DD><P>
<DT><B>PRESSURE OFF</B><DD><P>
<DT><B>PRESSURE EXTERNAL pressureX [ pressure Y [ pressure Z ] ] </B><DD><P>
<DT><B>PRESSURE { ISOTROPIC | ORTHORHOMBIC }</B><DD><P>These commands control XMD's constant pressure algorithm
(see section 
<A HREF="xmd-5.html#theory-pressure">Constant Pressure</A>).
By default the constant pressure algorithm is off, consequently
the box size if fixed.
<P>The command PRESSURE ANDERSEN turns on Andersen's constant
pressure algorithm.  With this algorithm a one or more masses
must be specified.  If only <CODE>xmass</CODE> is specified,
then the volume mass is the same in all three directions.  If only
<CODE>xmass</CODE> and <CODE>ymass</CODE> is specified, then <CODE>zmass</CODE>
is taken equal to <CODE>ymass</CODE>.
<P>The command PRESSURE CLAMP turns on the pressure &quot;clamp&quot;.
It requires an estimate of the system bulk modules in units of Mbars.
An optional value <CODE>cstep</CODE> controls how quickly the box size
fluctuates in response to internal stress.
<P>The  PRESSURE OFF command switches off both the PRESSURE ANDERSEN and
PRESSURE CLAMP algorithms (note that only one algorithm can be used a
one time).
<P>The PRESSURE EXTERNAL command sets the external pressure of the
system.  The external pressure will have no effect until the PRESSURE
ANDERSEN or PRESSURE CLAMP algorithms are in use.  By default the external
pressure is zero.  You can also apply different pressures in the X, Y and 
Z directions when using PRESSURE CLAMP (you will get an error using
PRESSURE ANDERSEN on your first CMD or QUENCH command).  If only 
one pressure (pressureX) is specified, then all pressures are equal.
If two pressures are specified (pressureX pressure Y) then pressureZ is
set equal to pressureY.  You can specify three pressure to get independent
X, Y and Z pressures.
<P>The PRESSURE { ISOTROPIC | ORTHORHOMBIC } command controls the symmetry
of the repeating box.  With PRESSURE ISOTROPIC (the default) the box
contracts or expands uniformly, so a cubic lattice will remain a cubic
lattice, and a tetragonal lattice with maintain its a/c ratio, etc.
PRESSURE ORTHORHOMBIC allows the x, y and z directions to change size
independently, which can result in an initialize cubic lattice losing
its cubic symmetry, and a tetragonal lattice losing its a/c ratio, and
even its a/b ratio.
<P>
<P>
<DT><B>QUENCH nstep [nquench]</B><DD><P>Similar to the CMD command.  Performs CMD for nstep number of steps, but
applies one of two quench algorithms to the resulting velocities, as a
way to relax atomic configurations.  If nquench is 1 (the default), then
whenever the potential energy rises relative to the preceding time step,
all the velocities are set to zero.  If nquench is 2, then when any
individual particle has f.v&lt;0 where f and v are the force and velocity
for that particle, then the velocity for only that particle (not the
entire system) is set to zero.  This is equivalent to setting velocity
to zero whenever the energy of an individual atom increases.
<P>
<P>
<DT><B>RCV [file [ [run] step]]</B><DD><P>Reads an RCV step.  If file is specified then that file is read;
otherwise the lines following the RCV command are read.  If step is
specified then the data for that step is read; otherwise the first step
in the file is read.  If step is not in the file then an error message
is given and the program continues.  If run is specified (which implies
that step is also specified) then only the data that belongs to the run
and step are read.  If step is specified without run, and two RCV steps
have the same step number, then only the first is read.  When both run
and step are specified and no matching RCV step is found, then an error
message is given and the program continues.  The RCV command reads in
particle coordinates, as well as a Radial Distribution Function (RDF)
table.  The RCV file does not store atom types, so the RCV command sets
the atom types all to 1.
<P>
<P>
<DT><B>READ file</B><DD><P>Opens the file named file and reads the commands as if they were in the
current command file.  At the end of the file control returns to calling
file.
<P>
<P>
<DT><B>REFSTEP [CLEAR | COPY | SWAP]</B><DD><P>When used without options the current particles, displacements, types
and RDF are copied to the reference step.  The CLEAR option erases all
of the information previously stored in the reference step.  The SWAP
option switches the reference and current step.  The COPY option copies
the refstep state to the current state, while retaining the contents of
the reference step.
<P>
<P>
<DT><B>REMOVE</B><DD><P>Removes the selected particles from the list of current particles and
reference particles.
<P>
<DT><B>REPEAT { nrepeat | { COR | RCV } file  [  step1  [  step2  ]  ] } </B><DD><P>
<P>
<P>The first form of this command repeats nrepeat times each command
between a REPEAT statement and an END statement, for example
<BLOCKQUOTE><CODE>
<PRE>
REPEAT 10
   CMD 100
   WRITE COR +crack.cor
END
</PRE>
</CODE></BLOCKQUOTE>
<P>
<P>
<P>The second form reads a COR or RCV file, and for each step in the file
executes the commands between the REPEAT and END statements.  Note that