Understanding Molecular Simulations:

From Algorithms to Applications

Daan Frenkel and Berend Smit

To run the program (on a UNIX system) use the following steps:

1. compile the program in directory CaseStudy_XX/Source/ by typing:
   make clean
   make
   All programs are precompiled for Redhat Linux 6.0 (Hedwig) on an i386. If you are using such a system you can skip this step.
2. make the file run in Run/ executable by typing:
   chmod +x run
3. run the program by typing:
   run
4. the output is written in the file out.

The Case Studies have been written in FORTRAN using a UNIX operating system. For non-UNIX operating systems some modifications should be made.

In each of the Case Studies a file out.test can be found. This file is provided to test the implementation of the program. This test simulation, however, is much shorter that the ones that are used to generate the results that are discussed in these Case Studies. To reproduce the Case Studies the reader should therefore modify some of the input parameters. Important to note is that we have not given all the details (for example, total simulation time, time step, interval between two sample, etc.) to make an identical run. We hope that all the essential parameters (for example, temperature, pressure, density, etc.) to reproduce (within statistical errors) are given in the Case Studies. Therefore, before one can reproduce a Case Study it is important to develop some `feeling' for these non-essential parameters by performing some short test runs, such as:

• what time step gives a good energy conservation in an MD run
• how often do I need to sample
• can I estimate the total length of a simulation
• what is the optimum maximum displacement in an MC run

Some Case Studies are performed with the same program. The source code is given only in one of these Case Studies. To perform the simulations of the other Case Studies it is important to keep the same directory names (eg. CaseStudy_7/Source/, CaseStudy_7/Run etc.)