home *** CD-ROM | disk | FTP | other *** search
- This is release 1.0 of Ami2D and is the first official release. Therefore,
- any comments and or suggestions would be appreciated.
-
- Russell Leighton
- leighton@ccwf.cc.utexas.edu (valid through Dec 1993)
-
- ---------------------------------------------------------------------------
-
- REQUIREMENTS
-
- The minimum system requirements for running the codes are: A 68020
- processor or better, 68881 floating point processor or better, AmigaDOS
- 2.04 or higher, ARexx and at least 2 megabytes of free RAM. Optionally, a
- hard drive and lots more memory are highly desirable. Additional software
- to enhance the operation of the codes is also highly recommended. This
- includes WShell (a commercial replacement for the AmigaDOS shell), GNUPlot
- (a freely-distributable general purpose plot package), AmigaGuide (a
- freely-distributable hypertext utility from Commodore), CyberCron (a
- freely-distributable task scheduler and pseudo-queue manager), and
- PowerSnap (a freely-distributable utility that allows cutting and pasting
- of text.) None of these programs are required, but definitely increase the
- interactivity and flexibility of Ami2D. Check for the latest versions of
- the freely-distributable programs on the latest Fred Fish disks or on
- AMINET.
-
- I have compiled a version that does not require a 68020 or better or the
- math coprocessor, but found that run times were not acceptable (even on my
- A3000 performance was on average 20% of the performance for the 68020/68881
- optimized version.) However, if enough people feel they must have this
- version I will post it.
-
- INSTALLATION
-
- To install this release just copy the Ami2D directory (keeping names and
- contents intact) to the desired location. Of course, the programs will run
- fine from a floppy disk. I recommend that you copy the rexx scripts in
- ami2d:rexx to your rexx: directory (except for rexxcon.rexx and
- rexxbar.rexx ). If you do not have rexxarplib.library and/or
- rexxmathlib.library then copy them from ami2d:libs to your libs:
- directory.
-
- Most of these codes are heavily dependant on ARexx, therefore, make sure
- that the ARexx interpreter is running. I recommend including the following
- line in your s:user-startup file:
-
- rexxmast >NIL:
-
- This will insure that ARexx is always available (it is a good idea to do
- this as a lot of applications require ARexx).
-
- For a quick test to see if everything is functional try double clicking on
- the ami2d icon. A screen should appear with two windows, one with buttons
- the other with a prompt. Type the following in the window with the prompt.
-
- -> !cd ami2d:test
- -> !beam1
-
- Now click on the lin2d button and wait for the model to be replotted (which
- indicates that the job finished.) Then type the following in the console.
-
- -> fill
- -> fill deform 20
-
- If you have a few spare megabytes (at least 2) you can try srm4 instead of
- beam1. However, you will have to run the model through Opt2D (unless you
- have at least 4 megabytes to spare.) To do this just click on the opt2d
- button before running lin2d as above. In a few seconds the model will be
- replotted and should be optimized (watch how the nodes are plotted to get
- an idea how the optimizer works.) Run Lin2D as above. While Lin2D is
- running you can check on the status by typing the following.
-
- -> !solstat
-
- On a 25Mhz A3000 this problem will take around 2 minutes to run (longer if
- you did not run the model through the optimizer.) Be sure to save the model
- if you want to keep the results. You can also do fill plots with the
- elements colored according to result level. For this problem type the
- following.
-
- -> post sigd
- -> fill result max 0 84
-
- This will result in a fill plot of the maximum deviatoric stress within
- each element. Other stress components and combinations are defined in
- ami2d:rexx/init.rexx which is read in by the ami2d.rexx script. Of course
- you can define your own.
-
- To try the nonlinear solver, Nln2D, try ami2d:test/rotate.rexx. This model
- is a simple beam pinned at one corner with displacements applied to the
- opposite corner such that the resulting displacements should result in a
- pure rotation. The solution procedure is iterative, therefore, the model
- must be run through the solver several times to converge on the desired
- solution. The first run through the solver will give the linear solution
- which shows the gross error that can result if a linear solution procedure
- is used when large displacements are expected.
-
- To quit out of the session type stop in the window.
-
- COMMENTS ?
-
- This initial release is a freeware version and can be freely distributed
- (provided all files in the original distribution remain intact), however, I
- would appreciate any comments/suggestions on the use of these codes. If
- you find a bug then please give me the details necessary to duplicate the
- bug (such as an ARexx script). This will insure greater likelyhood that
- the bug will be squashed. My e-mail address is given above (valid through
- Dec 1993.)
-
- Possible future enhancements include:
-
- - Advanced materials
- - composite materials (orthotropic, specially orthotropic laminae,
- laminates, etc.)
- - near incompressible, incompressible material behavior
- - viscoelastic (time-temperature dependant) material
- - nonlinear material behavior (rubber elasticity, plasticity, etc.)
- - Enhanced interface
- - Advanced modeler
- - Enhanced postprocessing
-
-