home *** CD-ROM | disk | FTP | other *** search
- ###############################################################################
- # #
- # Dust V2.00 - Copyright ©1994 by A.Maschke #
- # All rights reserved. #
- #-----------------------------------------------------------------------------#
- # #
- # DustMD v0.2 - Visualization of Molecular Dynamics #
- # #
- ###############################################################################
-
- INDEX
- 1. Introduction
- 2. Description of md2dust
- 3. Description of md2rpl
- 4. What is MDAr ?
- 5. Supplied files
- 6. Some more notes
- 7. Author, rights and literature
-
- 1. Introduction
- -------------------------------------------------------------------------------
-
- MD is a new external particle-effect for Dust and Real3D. The program
- mdar calculates the particle-positions and outputs data you can
- plot using gnuplot. md2dust generates a script which
- will drive Dust to generate the final objects using this data-files.
- The program md2real writes rpl-programs to create the objects *and*
- the frames.
-
-
- IT'S STRONGLY RECOMMENDED TO HAVE A LOOK AT THE SUPPLIED ANIMATION
- BEFORE READING THIS TEXT.
-
- Please look for more than 5 seconds at the animation otherwise you can't
- recognise the complicated process
- (The motion of every atom depends on the motion of all other atoms.)
-
- 2. Description of md2dust
- -------------------------------------------------------------------------------
-
- You can use md2dust to create your own effects, too. It reads
- in <Nnum> data-files "md.1", "md.2", ... .
- Every file contains the coordinates of one particle over
- the whole time.
- The number of data-points of the files is equal to the number
- of final particle-objects. After reading the points md2dust
- creates a Dust-script which loads the base-particle-object
- and applies the new coordinates to it.
-
- Example:
-
- 3 particles:
- motion of particle 0: (0,1,2)->(2,3,4)->(4,5,7)
- motion of particle 1: (3,4,2)->(1,3,2)->(1,2,6)
- motion of particle 2: (4,5,7)->(3,1,2)->(4,5,4)
-
- *You have to create the following data-files:
- (Normally your solver should create such data-files)
- file "md.1":
- 0 1 2
- 2 3 4
- 4 5 7
- file "md.2":
- 3 4 2
- 1 3 2
- 1 2 6
- file "md.3":
- 4 5 7
- 3 1 2
- 4 5 4
-
- *After that you have to create a particle-object of 3 particles (spheres).
- Name it "part0".
-
- *Now you can run md2dust:
-
- md2dust md 3 3 mdar.dust part0 100.0
-
- ("md" is the base name for your data-files,"3" the number of particles, the
- next "3" the number of points (=objects), "mdar.dust" ist the Dust-script,
- "part0" is the base-particle and "100.0" the scale-factor for all points.)
-
- *You get a script "mdar.dust" looking like this:
- load(1,part0)
- setppos(1,0,0,100,200)
- setppos(1,1,300,400,200)
- setppos(1,2,400,500,700)
- savepobj(1,part.0001)
- load(1,part0)
- setppos(1,0,200,300,400)
- setppos(1,1,100,300,200)
- setppos(1,2,300,100,200)
- savepobj(1,part.0002)
- load(1,part0)
- setppos(1,0,400,500,700)
- setppos(1,1,100,200,600)
- setppos(1,2,400,500,400)
- savepobj(1,part.0003)
-
- *Create the objects using Dust:
- dust mdar.dust .
-
- 3. Description of md2rpl
- -------------------------------------------------------------------------------
-
- The program md2rpl creates two rpl-programs. The first file (the one with the
- "test"-extension) is for adjusting the scene. It creates the first particle
- object and refreshes the windows. The render-file (with no extension)
- creates all objects sequentially, renders the pictures which are saved
- as "images/md/pic.1, images/md/pic.2, ..." - be sure to have this
- directory !
-
- Usage:
- md2rpl <data-files> <particles> <frames> <rpl-file> <scale-factor>
- Example:
- md2rpl md.posI 256 200 mdar.rpl 0.8
- (this options I used to create the included animation)
-
- After creating the rpl-files you have to create the scene:
- -start Real3D, rename your root-level as "root" (not "Root")
- -open a RPL-Window
- -enter "mdar.rpl.test" LOAD"
- -select the level "particle"
- -open a render-screen (maybe DBLPAL HAM8) with the
- name "Real.1", set public-screen-flag (yes,the name and the
- public-screen-flag are important)
- -open a View-window on this screen
- -choose the Menu "View/ViewCam/Autofocus"
- (now you should see the balls)
- -adjust the windowsize, renderflags, rotate the view,...
- (define your scene)
- -delete the "particle"-level
- -save the scene
-
- To render all frames simply load your scene, open a RPL-window
- and enter "mdar.rpl LOAD" - be sure to have no windows
- in front your render-window because the *screen* will be saved.
-
- NOTES: 1. After a crash you can continue the rendering after editing
- the script "mdar.rpl":
- Say, your last rendered frame is frame 123. Edit the file.
- Search for "frame.1". Choose the "Mark block"-function of
- your editor. Search for "frame.124". Choose the "Delete block"-
- function after moving the cursor to the previous line.
- Save it.
- 2. You can change the attributes of the spheres editing the
- MakeSphere-procedure of the generated rpl-programs.
-
- 4. What is MDAr ?
- -------------------------------------------------------------------------------
-
- For the theory consult the literature mentioned below.
- (If you have any detailed question send me mail - a description
- in this manual is impossible because I don't know where
- to start.)
-
- A short and easy description could look like this:
-
- MD computes the motion of Argon-atoms in a box at a constant
- temperature. The initial conditions are velocities satisfying the
- Maxwell-statistic. The boundary conditions are periodic,
- a Lennard-Jones-potential describes the forces between
- the atoms. After reaching the reference-temperature the box is opened.
-
- Because *every* atom interacts with all other atoms there are *lots* of
- calculations to be done.
-
- Before to run the program you have to create a job-file looking
- *exactly* like this:
-
- ----------START OF FILE------------
- ;comment
- N 32
- rho 0.636
- Tref 2.53
- L 10.229
- h 0.1
- sigma 0.3405e-9
- eps 119.8
- m 6.63382e-26
- rc 2.5
- timemx 300
- irep 5
- istop 100
- iseed 1234
- writestate 1
- writepos -1
- posind 1
- showstate 30
- usetbl 0
- crtbl 20
- calcg 0
- statename md.stateI
- posname md.posI
- gname md.gI
- dofcc 1
- tconst 0
- ----------END OF FILE-----------
-
- Parameters of interest:
-
- N : number of particles: if(fcc==0): any number between 32..256
- if(fcc!=0): 32,108 or 256
- rho : density (0.6..2.0)
- Tref : temperature (0.5..3)
- L
- h : time-step (0.01..0.1)
- sigma
- eps
- m
- rc
- timemx : (number of objects) * 2 \
- irep : (number of objects) / 40 (suggested) > (timemx-istop)=number of objects
- istop : number of objects (suggested) /
- iseed
- writestate
-
- writepos : -1 !!!!!
- posind
- showstate
- usetbl : some speedup (if set to 1)
- crtbl : 20
- calcg : 0 !!!!! (otherwise time-wasting)
- statename :
- posname : base-filename
- gname :
- dofcc : initial condition: create a fcc-lattice (if not set to 0)
- tconst : if(tconst!=0): keep periodic bounding conditions after calibration (looks bad)
-
- Keep all other values unchanged.
-
-
- 5. Supplied files
- -------------------------------------------------------------------------------
-
- *drawer doc:
- DustMD.doc - this file
-
- *drawer script:
- mdar.dust - first part of the Dust-script created by md2dust
- (this script creates only the 1st object.)
- md.script - shell-script which runs md2dust to create the
- complete mdar.dust-script
- mdar.rpl.test - the test-RPL-file, execute it by typing
- "mdar.rpl.test" LOAD
- into a RPL-window
- (the mdar.rpl-file is too large(1.5MB) to be included here)
- *drawer objects:
- 2 objects and the Dust-script which creates the base-particle-object
- using this objects
-
- *drawer src:
- private
-
- *drawer bin:
- executables for every Amiga (cc -O3 )
- executables with the extension "040" for turbo-machines (cc -O3 -m68040)
-
- *drawer data:
- md/md.#? - data-files created by mdar
- taski - job-file for mdar (start it typing "mdar taski")
-
- *MD.anim5 : the animation
-
- 6. Some more notes
- -------------------------------------------------------------------------------
-
- (i) I'm sure, this a very bad documentation.
- If you don't understand anything described here don't hesitate to
- send me a mail.
-
- (ii) If there where many people which are interested in this effect I
- could improve the user-interface, simplify the options, add some
- more features, ...
-
- 7. Author, rights and literature
- -------------------------------------------------------------------------------
-
- Author:
- Andreas Maschke
- Zenkerstraße 5
- 06108 Halle/Saale
- Germany
-
- Phone: ++49 (0)345/5170331
- EMail: epgbc@cluster1.urz.Uni-Halle.DE
-
- Rights:
- The DustMD-package belongs to the Dust-project. Dust is SHAREWARE
- with a fee of $25 or 25 DM + postage.
-
- Literature:
- A. Rahman, Physical Review 136, A405 (1964)
-
- D.W. Heermann, Computer Simulation Methods in Theoretical Physics,
- 2nd Edition, Springer-Verlag 1990
-
-
- (19 July 1995)
-
-