Last update: Feb.24 2009 15:30 JST
Software packages: NAOPC and MTGRP
Author: Yasuhiko Okada
(with contributions from Alexander A. Kokhanovsky)
In this page, a software package to obtain optical properties (cross sections, phase matrix) of a group of spherical particles based on the numerical orientation averaging using Quasi-Monte-Carlo method is given. The calculation for a fixed orientation is based on the T-matrix method.
Also presented is a package to generate coordinates of a group of evenly distributed spherical particles.
|
Packages for 1CPU
1. A package for numerical simulations of optical properties using numerical orientation averaging : NAOPC for 1 CPU
[tgz file]
2. A package for generating of coordinates of a group of evenly distributed spherical particles : MTGRP
[tgz file]
3. Combined package of NAOPC and MTGRP for 1CPU
[tgz file]
In each package, please refer to [readme.pdf] on how to use it.
|
Packages for multiple CPUs
4. NAOPC for multiple CPUs
[tgz file]
5. Combined package of NAOPC and MTGRP for multiple CPUs
[tgz file]
In the packages 4 and 5, please refer to [readmeMultiCPU.txt] on how to use the package on multiple CPU system. Other detailes are found in [readme.pdf]
|
Acknowledgment:
The software packages given above use the T-matrix code for clusters of spheres for
a fixed orientation of a cluster with respect to the incident light direction
located at
ftp://ftp.eng.auburn.edu/pub/dmckwski/scatcodes/index.html.
The fixed orientation code (scsmfo1b.for) used in our software packages was developed by Daniel Mackowski.
We acknowledge his contribution with many thanks.
At the same location, users can find the code for the analytical averaging with respect to orientation.
The results of the averaging code presented here have been compared with the case of analytical averaging and the differences smaller than 1% have been obtained, if appropriate input parameters are chosen. The analytical averaging gives an exact result. Also it is faster in many cases. However, users may find useful to use numerical averaging codes presented here due to the following reasons: 1) the calculations on several CPUs (e.g., 10-20) can be performed, which leads to smaller calculation time as compared to 1CPU analytical averaging code; 2) the approximate results for the cross sections and phase matrix elements (e.g., 1-5% error) can be quickly obtained, if the sparse integration grid is selected.
|
Point of contact:
Dr. Alexander A. Kokhanovsky
Institute of Environmental Physics
University of Bremen
Otto Hahn Allee 1, D-28234 Bremen, Germany
Phone: 00 - 49 - (0) - 421 - 218 - 2915
Fax: 00 - 49 - (0) - 421 - 218 - 4555
E-mail: alexk@iup.physik.uni-bremen.de
|
|