Hello,

I need to simulate a flowing of substance, but I needn't(and can't) do it in right (completely physical-proved) way. My system will being used as education system, so I can do some fakes in simulation process.

Examples of flowing of substance\energy, that I need to simulate: flowing of smoke\dust\..., expansion of noise, propagation of explosive wave, heat conduction, etc.

Again, I do not simulate the real physical process, I can do it approximately(to get students only common understanding of the real process, to illustrate the process).

So I'm seeking for the convenient middleware or engine to solve my task. I glance over the CFD-Online links(both software and libs), but don't find anythink suitable for my task.

For example for particle modelling I intend to use Particle System API(www.particlesystems.org). May be this lib don't modelling particles' dynamic properly in terms of physics, but it suits me.

Is there anythink like that in CFD systems area? So called, CFD fake tool Could anybody advises me such a tool?

Thank you.

I see many peoples look-see my message, but there isn't any answer.

Does that mean that there isn't any well known CFD-tool suit my request or simply my English is so bad, that peoples simply don't understand what I asking for ?(If last, sorry for my English I'm from Russia)

Thank you.

Saw an article in a US 'general audience' newspaper or magazine about a year ago.

The folks in Hollywood had developed a system to produce fluid motion movies good enough for animated films. The article said that it wasn't physics but suited the purpose.

Maybe you can track that down with a search engine, but don't know exactly where you should start.

Your English is fine.

Not sure that it's what you're looking for, but take a look at this http://www.plasmapong.com/ Also look for the publications of its author, Steve Taylor (I remember there were a couple)

I saw an example of what you are looking for at a conference in Austin Texas (8th US National Conference on Computational Mechanics) a couple of years ago. It was not in a formal session, but sitting around a table during a break, this guy was showing movies of things like breaking waves on his laptop. I had the impression that it was not Navier-Stokes driven, but it looked very realistic, and I think relatively cheap to calculate. Sorry, I don't remember his name or affiliation, but I think he was academic.

I think instead of 'fake CFD', you are really looking for 'fast CFD' or 'real-time CFD'. The article Real-Time Simulation contains a discussion of real-time physics simulations including fluid flow that you might find interesting.

>>I think instead of 'fake CFD', you are really looking for 'fast CFD' or 'real-time CFD'.

Yes, you are right and thanks a lot for useful link.

In fact now I see four ways to achieve my goal (to create such fast\fake CFD-system):

1. Investigate algorithms utilized in game development (for example game www.plasmapong.com show us, that there are some methods in gamedev to achieve realistic fluid dynamic.

2. Try to simulate fluid dynamic by cellurar automates(CA)

3. Try to use particles (e.g. www.particlesystems.org) to simulate fluid dynamics (also there are some papers in net)

4. Try to find simple method of roughly solving real CFD problem (probably it is impossible to find such method)

"4. Try to find simple method of roughly solving real CFD problem (probably it is impossible to find such method)"

Actually the method used by plasmapong (see the reference to Stam's paper on their page) is a valid CFD method. Basicallly the speed comes from not enforcing convergence of the pressure equation (and similarly the implicit viscous terms); i.e. do a fixed number of iterations within the linear solver and don't bother checking the residual error!

Hi, Alexander,

I developed Flow Illustrator for this purpose. It works and it is available online at

http://chernyshenko.sesnet.soton.ac.uk/FIHome.htm

You upload an image of a body and get a movie of the flow past it. Enjoy .

And I would like to see if Russian works here . Just a few words:

Привет, Александр. Надеюсь, что мой веб-сайт вам поможет хотя-бы частично.

Regards

Sergei