README for TRIELAS      CWP Proprietary code   U-43          31 March 1997

This package is proprietary to the Consortium Project on Seismic
Inverse Methods for Complex Structures until 15 July 2000.

TRIELAS - Programs for seismic model building and seismic modeling in
	Delaunay triangulated 2D anisotropic Elastic media.

ELARAY - ray tracing for elastic anisotropic models
ELACHECK - get elastic coefficients of model
ELATRIUNI - convert TRIangulated ELAstic models to UNIformly sampled models
ELAMODEL - make piecewise homogeneous anisotropic model
RAYDATA - simple program to display ray data from elaray
ELABEAM - Elastic seismograms from triangulated models

Demos are located in the directory "Demos".

Of these, "ELABEAM" is not quite finished, and should be used with
care.
Author: Andreas Rueger CWP  March 02, 1997

The codes depend on the CWP/SU (Seismic Unix) free software package
available from CWP at:

ftp.cwp.mines.edu
http://www.cwp.mines.edu/cwpcodes

Once SU is installed, this package will install, simply by typing
make in this directory.

## Comments from Andreas Rueger.

Comments on:  -ELAMOD 

This is a 2D raytracing program for transversely iotropic
media with in-plane rotated axis of symmetry. I wrote this
code in 1994 to get started in anisotropy. It is a helpful
tool to display rays/wavefronts in anisotropic media to
visualize many wave phenomena that are unknown in the 
isotropic world. Most of the raydata such as traveltimes,
polarizations, etc can be plotted and a simple seismogram
calculator such as elabeam.c can be used to generate synthetic
data.

Warning:
While this code is a useful and educating tool, it is 
a research algorithm and not fully developped and tested.
It is not a production strength algorithm! I only release
the code on public demand and hope that you keep me posted
on any additional modification that you may develop.


Strengths and limitations:
The models can be rather complex. Triangulation allows
to generate models composed of homogeneous blocks. Extension
to inhomogeneous blocks should be straightforward, at least
algorithmically. Basically all ray-modes can be traced. Each
interface in the model can be defined to reflect, transmit
and convert the wave. The rays can also be stopped at any
interface in the model, so many ray experiments can be
simulated. Wavefronts can be plotted at any traveltime or 
traveltime-increment.


This code does not include amplitude computation, although 
I did some rough calculations and support most of the 
necessary data structues. So this needs more work. The same
is true if you want to handle cusps, shadowzones or two-point
raytracing in general. The <elabeam.c> code in the modif-directory
can give you a start on that.

The code was written on a NeXT maschine and hence I did not
generate X-Windows output. All graphics are in PostScript. 

The problem to continue the tangential slowness across a 
boundary is solved using a rootfinder that Tariq Alkalifah
and I designed. It allows to only compute one root in a
rather fast manner. Just recently, I had some cases there
the rootfinder did not converge for converted waves. If
you experience this, plan of linking a standart rootfinder
for Snell's law instead.

I include a set of demos to demosntrate some of the nice
features of this algorithm.

Finally, I want to thank Dave Hale who wrote the acoustic
version of this program.


______ Andreas Rueger CWP  March 02, 1997



John Stockwell	 | jstockwe@mines.edu
Center for Wave Phenomena (The Home of Seismic Un*x)
Colorado School of Mines
Golden, CO 80401 | http://www.cwp.mines.edu/cwpcodes
voice: (303) 273-3049  |    fax: (303) 273-3478.
