Programs in Physics & Physical Chemistry
|[Licence| Download | New Version Template] ackz_v1_0.gz(2 Kbytes)|
|Manuscript Title: Identification of clusters in computer experiments with periodic boundary conditions.|
|Authors: H. Bunz|
|Program title: CLUSTER IDENTIFICATION|
|Catalogue identifier: ACKZ_v1_0|
Distribution format: gz
|Journal reference: Comput. Phys. Commun. 42(1986)435|
|Programming language: Fortran.|
|Computer: IBM 3090.|
|Operating system: MVS/SP 2.1.3 (XA).|
|RAM: 1000K words|
|Word size: 32|
|Keywords: Molecular physics, Experiment, Cluster size Distribution, Re-ordering of molecules, Determination of Topologically Connected patterns.|
Nature of problem:
In many computer programs simulating many-body problems, patterns topologically connected have to be determined. This problem arises especially in Monte-Carlo and molecular dynamics calculations simulating nucleation or cluster formation, respectively.
The molecules recognized as belonging to a specific cluster are re- ordered in such a way that all molecules in this cluster are ordered sequentially. The molecules have to be numbered. This sequential order of the molecules facilitates the subsequent calculation of any cluster properties.
The program is only applicable for a limited number of molecules in a primary cell of cubic shape and limited size. But it may easily be extended to any other rectangular primary cell of any dimension. It is especially designed for periodic boundary conditions allowing extension of the cluster over the boundary of the primary cell. But difficulties may arise for an infinite cluster as the program will give only the number of molecules of this cluster per cell.
The computer time consumed increases with the square of the number of molecules but decreases with increasing density of molecules and decreasing velocity of changing the consecutive configurations. Running time of sample problem is approximately 0.1 sec on Siemens 7890.
|Disclaimer | ScienceDirect | CPC Journal | CPC | QUB|