I have previously advocated for computational chemistry as a ripe field that is well-accessible to an amateur of limited means. At least a handful of
members seem to agree that computational techniques are worthy of further discussion, and to promote that end I have created a new forum devoted to
computational methods.
It is not confined exclusively to computational chemistry; other topics of interest may include (for example) bioinformatics, expert system synthesis
planning, and finite element models of explosive phenomena.
I do realize that computational techniques are more obscure, and less widely used among amateurs, than their cost/availability alone would suggest.
It's my hope that a spotlight forum may help to slowly change that, at least among members here.FANTOMAS - 24-1-2007 at 12:05
I saw a software ,very good specs.,priced over 5500$ !That was a no-no! So , I think your proposal is eccelent.
But I would not know where to start , to contribuite to the above project .Ozone - 24-1-2007 at 16:57
Matlab and Femlab are usually made available at most universities under general license, viz. usually wither no password id required or it is provided
with the download. Both are relatively easy to use, and can produce excellent models.
Following this, there are many freeware and developmental "drunken walk" progs out their (Flory-Huggins lattices, leaves falling from trees, simple
CFD, etc.) that can be interesting. The only problem with these is that they are usually made fora specific purpose, so adapting it to certain needs
may be impossible.
For prediction of properties, Molecular dynamics (MM2) and Mopac are included in the Chemoffice suite (which is very handy, especially "minimize"
which can be used to iteratively reconform your structure to correspond with a given state temperature).
B3YLP Hartree-Fock would be neato.
It would be flat-out awesome to get something like Spartan or Aspen; you can draw your device, etc. and model heat, liquid, gas flows, etc. (big buck
and restricted licensing here).
I think it would also be worthy to add that decent graphing/analysis software such as Sigmaplot or Origin would also be a significant contribution to
the community (Excel is so limited and gets old in a hurry). I have yet to see any of these (except Chemoffice, which need not be cracked) available
in a non-activation-required form. Bummer.
At any rate, I could not agree more,
O3Polverone - 24-1-2007 at 17:56
I'm sure that more software will be available from MadHatter's FTP site. A recent version of Fujitsu CAChe is now there, for example. It includes
modeling, structure refinement, spectra predictions, and much more with your choice of Hartree-Fock, density functional theory, or advanced
semiempirical methods. It can also serve as a graphical interface to Gaussian 03 (though that software is not currently available on the FTP).
There are also a lot of good, free or low-cost programs out there for people who don't want to hoist the jolly roger. I particularly hope to
concentrate on these offerings when I get to doing some writeups for the subforum, though I do admit that the integrated commercial programs are
easier for a beginner to start playing with.chemoleo - 24-1-2007 at 18:22
Perhaps we could also include the more mundane calculation aspects of science?
Several threads come to mind here:
We could also include any sort of questions in natural sciences with emphasis on chemistry that can be addressed by mathematical approaches?JohnWW - 3-8-2009 at 13:22
Would that new thread include the subject of computational fluid dynamics (a branch of fluid mechanics), of importance in chemical engineering (and
also civil and mechanical)? Otherwise, I will post some ebooks on the subject on the Chemical Engineering thread in References.
[Edited on 3-8-09 by JohnWW]not_important - 3-8-2009 at 16:07
...
I think it would also be worthy to add that decent graphing/analysis software such as Sigmaplot or Origin would also be a significant contribution to
the community (Excel is so limited and gets old in a hurry)...
O3
While it's not the flashiest and takes a bit of learning, gnuplot does a decent job in many cases and is still used in publications. Multiplatform,
free, output to screen, plotter, printer, or image file, source code available.
).