M01 Introduction to MD++: Difference between revisions

From Micro and Nano Mechanics Group
Jump to navigation Jump to search
Line 111: Line 111:
By default, '''SYS=gpp''' which corresponds to the GNU c++
By default, '''SYS=gpp''' which corresponds to the GNU c++
compiler. If you have the intel c/c++ compiler '''icc''' in your system, use '''SYS=intel''' to activate '''icc''', which produces much faster executables than GNU c++ compiler. For more '''SYS''' values, take a look at the file '''~/Codes/MD++/src/Makefie.base'''.
compiler. If you have the intel c/c++ compiler '''icc''' in your system, use '''SYS=intel''' to activate '''icc''', which produces much faster executables than GNU c++ compiler. For more '''SYS''' values, take a look at the file '''~/Codes/MD++/src/Makefie.base'''.

You may get an TCL-related error during compilation. You can bypass this error using, ''e.g.''
You may get an TCL-related error during compilation. You can bypass this error using, ''e.g.''


Line 117: Line 116:


by suppressing TCL option.
by suppressing TCL option.
After compilation, you have executable files such as ''fs_gpp'', ''eam_gpp'', and ''sw_gpp'' in the '''bin''' directory.<ref>If you use '''SYS=cygwin''', the executable files would be ''fs_cygwin.exe'', ''eam_cygwin.exe'', ''etc''.</ref> If you just want to compile one executable file, such as '''fs_gpp''', use

After compilation, you have executable files such as ''fs_gpp'', ''eam_gpp'', and ''sw_gpp'' in the '''bin''' directory.<A NAME="tex2html5"
HREF="footnode.html#foot152"><SUP>5</SUP></A>If you just want to compile one executable file, such as '''fs_gpp''', use


$ make fs build=R
$ make fs build=R


The name of each execution file stands for the potential model used in
The name of each execution file stands for the potential model that will be used in
the atomisitic simulation. For example, '''fs''' stands for the Finnis-Sinclair (FS) potential<ref>M. W. Finnis and J. E. Sinclair, ''Philos. Mag. A.'' '''50''' 45 (1984)</ref> , '''lj''' for the Lennard-Jones (LJ) pair potential,<ref>J. E. Lennard-Jones, "Cohesion" ''Proceedings of the Physical Society'' '''43''' 461-482 (1931)</ref> '''eam''' for the embedded-atom method (EAM) potential,<ref>M. S. Daw and M. I. Baskes, "Embedded-atom method: Derivation and application to impurities, surfaces, and
the MD simulation. For example, <TT>fs</TT> stands for the
other defects in metals", ''Phys. Rev. B'' '''29''' 6443 (1984)</ref> and '''sw'''
Finnis-Sinclair (FS) potential&nbsp;[<A
for the Stillinger-Weber (SW) potential.<ref>F. H. Stillinger and T. A. Weber, "Computer simulation of local order in condensed phases of silicon", ''Phys. Rev. B'' '''31''' 5262 (1985)</ref><ref>H. Balamane, T. Halicioglu, and W. A. Tiller, "Comparative study of silicon empirical interatomic potentials", ''Phys. Rev. B'' '''46''' 2250 (1992)</ref> Each potential model is applicable to a group of atomic species. For example, the FS potential was designed for body-centered-cubic (BCC) metals such as Mo, Ta and
HREF="node6.html#fin:sin">2</A>], <TT>lj</TT> for the
⚫
W. The SW potential was designed for diamond-cubic semiconductors such as Si and Ge.
Lennard-Jones (LJ) pair potential&nbsp;[<A
⚫
By the way, the executable '''md_gpp''' does not contain any potential
HREF="node6.html#lj">3</A>], <TT>eam</TT> for the
embedded-atom method (EAM) potential&nbsp;[<A
HREF="node6.html#daw:bask">4</A>], and <TT>sw</TT>
for the Stillinger-Weber (SW)
potential&nbsp;[<A
HREF="node6.html#sti:web">5</A>,<A
HREF="node6.html#bal:hal:til">6</A>]. Each potential model is
applicable to a set of atomic species. For example, the FS potential
was designed for body-centered-cubic (BCC) metals such as Mo, Ta and
⚫
W. The SW potential was designed for diamond-cubic semiconductors such
as Si and Ge.
⚫
The executable, <TT>md_gpp</TT> does not contain any potential
model. Hence it cannot be used to compute any material properties but
model. Hence it cannot be used to compute any material properties but
it can still be used to display atomic structures.
it can still be used to display atomic structures.

<BR>
If you modify the source code (in the <TT>src</TT> directory), you
If you modify the source code (in the '''src''' directory), you
need to recompile MD++. we recommend you to clean the previous executables
need to recompile MD++. we recommend you to clean the previous executables
and rebuild by,
and rebuild by typing,

<BR><TT>
$ make clean; make all build=R
$ make clean; make all build=R

<BR></TT>
⚫
To see more help on '''make''', type

⚫
$ make help


To check if you successfully install and compile MD++, try to run one of example scripts in the '''Examples''' directory, ''e.g.''
<BR>
⚫
To see more help on <TT>make</TT>, type
<BR><TT>
⚫
$ make help
<BR></TT>


⚫
$ bin/fs_gpp scripts/Examples/example01-mo.script
<BR>
To check if the installation and compilation is successful, try to run
a script in the <TT>Examples</TT> directory, e.g.
<BR><TT>
⚫
$ bin/fs_gpp scripts/Examples/example01-mo.script
<BR></TT>


⚫
If you want to stop MD++ running, press '''ctrl+c''' or close
<BR>
⚫
If you want to stop an MD++ run, press <TT>ctrl+c</TT> or close
the visualization window.
the visualization window.
<HR>
<references/>


== MD++ Directories ==
== MD++ Directories ==

Revision as of 21:25, 25 November 2007


Manual 01 for MD++
Introduction to MD++

Keonwook Kang and Wei Cai

Nov 24 , 2007


Overview

MD++ is a molecular dynamics simulation package written in C++. It is originally developed by Wei Cai when he was a graduate student at MIT and still being updated to include more features. MD++ is supposed to run in a Unix/Linux environment. However, if you have cygwin,[1] a linux-like environment installed in your Windows machine, you can also enjoy MD++ there. Currently, this code is designed mainly for studying the atomic behavior in solid crystal materials. It can also be extended to study other systems, such as fluids, polymers and bio-molecules. This document is the first of a series of manuals written for those new to MD simulations. These manuals explain how to use MD++ with simple examples.


Installation

The latest version of MD++ code can be downloaded from the web site, http://micro.stanford.edu/~caiwei/Forum/2004-12-12-MD++/. Here we assume that the reader is familiar with basic Unix/Linux shell commands.[2] Before you download MD++ code, we recommend that you make a sub-directory Codes in your home directory. 

$ mkdir Codes 
$ cd Codes

Save the MD++ package in this directory(~/Codes). Suppose the downloaded file is md++-2007-01-07.tar.gz, then we can extract it by, 

$ tar -zxvf md++-2007-01-07.tar.gz

This command creates the MD++ sub-directory and extracts all the files into it. Go to the MD++ directory and make a runs directory if it does not exist. 

$ cd MD++; mkdir runs

To compile the codes in release(R) mode [3] in a Linux system, type 

$ make all build=R

In order to compile MD++ In a different system (such as Windows), you can specify the SYS variable in the command line, such as 

$ make all build=R SYS=cygwin

provided that cygwin is installed. You can also compile MD++ code in an Apple/Machintosh computer, using 

$ make all SYS=mac build=R

By default, SYS=gpp which corresponds to the GNU c++ compiler. If you have the intel c/c++ compiler icc in your system, use SYS=intel to activate icc, which produces much faster executables than GNU c++ compiler. For more SYS values, take a look at the file ~/Codes/MD++/src/Makefie.base. You may get an TCL-related error during compilation. You can bypass this error using, e.g. 

$make SYS=cygwin TCL=no

by suppressing TCL option. After compilation, you have executable files such as fs_gpp, eam_gpp, and sw_gpp in the bin directory.[4] If you just want to compile one executable file, such as fs_gpp, use 

$ make fs build=R

The name of each execution file stands for the potential model that will be used in the atomisitic simulation. For example, fs stands for the Finnis-Sinclair (FS) potential[5] , lj for the Lennard-Jones (LJ) pair potential,[6] eam for the embedded-atom method (EAM) potential,[7] and sw for the Stillinger-Weber (SW) potential.[8][9] Each potential model is applicable to a group of atomic species. For example, the FS potential was designed for body-centered-cubic (BCC) metals such as Mo, Ta and W. The SW potential was designed for diamond-cubic semiconductors such as Si and Ge. By the way, the executable md_gpp does not contain any potential model. Hence it cannot be used to compute any material properties but it can still be used to display atomic structures.

If you modify the source code (in the src directory), you need to recompile MD++. we recommend you to clean the previous executables and rebuild by typing, 

$ make clean; make all build=R

To see more help on make, type 

$ make help

To check if you successfully install and compile MD++, try to run one of example scripts in the Examples directory, e.g. 

$ bin/fs_gpp scripts/Examples/example01-mo.script

If you want to stop MD++ running, press ctrl+c or close the visualization window.

MD++ Directories

Script File

Visualization Window Control

Bibliography

About this document ...



Notes

  1. ↑ For download and installation, visit http://www.cygwin.com/. You must download and install the full package of cygwin. See http://micro.stanford.edu/~caiwei/Forum/2004-12-12-MD++/cygwin.html for installation instructions.
  2. ↑ For those who are not familiar with Unix/Linux environment, we suggest the book A practical guide to Linux commands, editors and shell programming by Mark G. Sobell. You can also easily find several good web sties about the shell commands.
  3. ↑ If build=R, MD++ is compiled with the -O3 option. By default(build=D), MD++ compiles as debug mode with the -g option. In release mode, the executabale files run faster but the compilation time is slightly longer than debug mode
  4. ↑ If you use SYS=cygwin, the executable files would be fs_cygwin.exe, eam_cygwin.exe, etc.
  5. ↑ M. W. Finnis and J. E. Sinclair, Philos. Mag. A. 50 45 (1984)
  6. ↑ J. E. Lennard-Jones, "Cohesion" Proceedings of the Physical Society 43 461-482 (1931)
  7. ↑ M. S. Daw and M. I. Baskes, "Embedded-atom method: Derivation and application to impurities, surfaces, and other defects in metals", Phys. Rev. B 29 6443 (1984)
  8. ↑ F. H. Stillinger and T. A. Weber, "Computer simulation of local order in condensed phases of silicon", Phys. Rev. B 31 5262 (1985)
  9. ↑ H. Balamane, T. Halicioglu, and W. A. Tiller, "Comparative study of silicon empirical interatomic potentials", Phys. Rev. B 46 2250 (1992)
Retrieved from "http://micro.stanford.edu/mediawiki/index.php?title=M01_Introduction_to_MD%2B%2B&oldid=330"