MSE 460 - Computational Materials Science

Course Description:
This course will cover fundamentals of atomic level modeling of the structure and properties of materials. Specifically it will cover metals, semiconductors, oxides and other ionic crystals. First, the description of atomic interactions will be introduced. This will include both basics of the density functional theory and approximations in terms of pair potentials, embedded atom method and tight-binding. The methods of computer modeling include molecular statics, molecular dynamics, Monte Carlo and lattice dynamics (phonons). Interpretations of results of such modeling in terms of structures, for example using the radial distribution function, thermodynamic and statistical physics analyses will be an important component of the course.

Instructor:
Professor Vaclav Vitek, Room 218 LRSM, tel. 215-898-7883, vitek@lrsm.upenn.edu

Teaching assistant:
Roman Gröger, Room 215 LRSM, tel. 215-898-9171, groger@lrsm.upenn.edu

Meeting sessions (optional):
Each Monday, 3pm - 5pm; come to my office at LRSM/215
Start working on your assignment early; bring it with you any time you need an assistance.


Course materials & other helpful documentation:
Here is your primary source of information about programming in F77 that I wrote for you. Please, read it carefully as it contains the necessary knowledge base to get started.
Introduction to programming in Fortran 77 for students of Science and Engineering

If you work with Windows, the following link might be of interest for you. It contains various hints to customize your Emacs like e.g. change the keybindings, colors, font, etc.
http://www.gnu.org/software/emacs/windows

Did you try Emacs for MacOS X in the command line terminal ? If you don't like it, have a look at http://www.webweavertech.com/ovidiu/emacs.htm to learn about Emacs which runs entirely as a native Mac application.

If you don't like using Matlab, Maple or Mathematica for plotting your data, consider using the GnuPlot. It is a command-line plotting program which is very popular among scientists. Check the following web site to learn more: http://www.cs.uni.edu/Help/gnuplot
A particularly readable introduction to gnuplot that is very useful also as a further reference can be found at: http://t16web.lanl.gov/Kawano/gnuplot/index-e.html


Problems & answers:

How can I generate random numbers in F77 ?
F77 provides an intrinsic function called RAND which gives a number from a pseudorandom sequence. If the input parameter is 1, the random sequence generator is restarted. If you give it 0, the function returns a number from the pseudorandom sequence. This is what you need to get a random number in num: 
      num = RAND(0)
Note that if you call RAND(1) and then you read the random numbers using RAND(0), you always get the same sequence because the random number generator in Fortran gives really something what people call "pseudorandom" numbers. If you give RAND other number than 0 or 1, it uses this number as a seed for the generation of the random sequence (if you want to learn details, check with some advanced math book).
The atoms in my block are plotted in Matlab as empty circles. How can I draw them filled ?
Most likely, the plotting of your atomic structure is done using the Matlab command plot. This command always returns a handle (or a pointer) to the properties of your plot. For example: 
      handle = plot(x,y,'ko');
plots the block of atoms with coordinates stored in two vectors x and y such that each atom is represented by a little black circle. Then, you just have to change the property of the circle to make it filled. Write this line below your plot command: 
      set(handle,'MarkerFaceColor','k');
This sets up black as the color which will be used to fill the circles.
The do-while loop does not work.
There is an error in my introductory text to programming in F77 above (thanks Aimee). The correct structure of the do-while loops is: 
      do while (condition)
         :
         :
      enddo
The loop is being executed as long as the condition is true. Once it becomes false, the program continues below enddo. An updated version (at least 1.2) of the introductory manual is already given above.
The syntax of my Fortran code is not displayed in color.
By default, Emacs for Windows and Mac has this coloring switched off. To turn it off, create in the c:\ directory (Windows) or your home directory (Mac) a file .emacs (don't forget to include the comma) and copy the following lines into it. 
(cond ((fboundp 'global-font-lock-mode)
       ;; Turn on font-lock in all modes that support it
       (global-font-lock-mode t)
       ;; Maximum colors
       (setq font-lock-maximum-decoration t)))
Then, save the file and restart Emacs.