Chemics Engine?

That sounds like a lot of work. Even just representing the fluids in 3d is a feat unto itself.

I'm not aware of anything like that for games at the moment (since I suppose its use is sort of niche), I'd recommend looking at fluid dynamics to get an idea of how to visualise the chemicals. There might be professional software used by scientists to visualise chemical reactions, but I'd expect it to cost a lot..

Quote:

Original post by Gandi So maybe you create tow "ChemicsObject"'s in 3D space and tell the engine they are made of H² and O² and have a specific mass, and the engine automatically generates a new ChemicsObject of H²O with 2 times the mass.

2 times the mass?

Molar mass of H2 ~2 grams
Molar mass of 02 ~32 grams
Molar mass of H20 ~18 grams

2H2 + 02 -> 2H20

4g + 32g -> 36g

Have look at the "Multiple Interacting Liquids" on http://physbam.stanford.edu/~fedkiw/. It is the only example of such a simulation I have seen so far.

What you are looking for is called "chemical kinetics". It is a very complex subject. This paper, for example, shows a way to model the reaction of two molecules in 3D space. This is real hardcore stuff.

For a game, I would simply create a database of common reactions, and look up everything there.

Quote:

Original post by LessBread Quote: Original post by Gandi So maybe you create tow "ChemicsObject"'s in 3D space and tell the engine they are made of H² and O² and have a specific mass, and the engine automatically generates a new ChemicsObject of H²O with 2 times the mass. 2 times the mass? Molar mass of H2 ~2 grams Molar mass of 02 ~32 grams Molar mass of H20 ~18 grams 2H2 + 02 -> 2H20 4g + 32g -> 36g

What I meant what was 2 times the mass of h2o, but just forget about it^^

This chemical kinetics-stuff looks really scary, so I guess I'll go for the database approach.

Anyway, thanks for helping
gandi

It all depends on what you're trying to accomplish. If you want to replicate chemical reactions as they would occur in nature, I would recommend you change your goal. 8^) Computational chemistry is VERY complex and even the best possible simulations aren't quite right yet.

If you want something more simple, you can get away with rule-based systems that are limited in their scope and abilities but do indeed produce some artificial chemistries. Here are some references for you to check out. I realize some of them are paper, but with a teeny bit of effort you can find many of the papers online in PDF format for free.

-Kirk

http://www.mitpressjournals.org/toc/artl/15/1
This issue of the journal Artificial Life is devoted to artificial chemistries. Like I said, many of these articles can be found online for free.


http://organicbuilder.sourceforge.net/
Organic Builder is from Tim Hutton, a major researcher in artificial chemistries. The demo is very cool and gives you some insight into artificial chemistry's capabilities.


http://johnnyvon.sourceforge.net/
JohnnyVon is an artificial life/computational evolution application. Very cool.


http://www.ventrella.com/Darwin/darwin.html
Darwin Pond is another artificial life/evolution application.

Quote:

Original post by Yann L This paper

This is useless information unless Gandi has a supercomputer, or better (and by better, I mean this http://mags.acm.org/communications/200807/?folio=91&CFID=23814078&CFTOKEN=83211963)

Quote:

Original post by Prune Quote: Original post by Yann L This paper This is useless information unless Gandi has a supercomputer

No shit, Sherlock... [rolleyes]

Oh wow, I never thought it would be impossible to perform calculations per molecule on a normal computer xD

To make that clear - I don't want to make an exact simulation of reactions. It is just for a game and it does not need to be perfectly realistic. So I guess a database or rule-based system will fit best.

Gandi

Look at the artificial chemistry links a suggested. Organic Builder in particular should give you some good ideas.