I developped a free CFD program, called "CellFlow":

http://www.xs4all.nl/~westy31/CellFlow/CellFlow.html

A funky feature is that it can 4 dimensional flow.

Apart from that, I am not sure if it actually better than other free CFD
codes. My guess it is probably quite handy if you have a simple geometry.

Gerard

Bruce Scott TOK wrote:
Gerard Westendorp wrote:

| I developped a free CFD program, called "CellFlow":
|
| http://www.xs4all.nl/~westy31/CellFlow/CellFlow.html
|
| A funky feature is that it can 4 dimensional flow.
|
| Apart from that, I am not sure if it actually better than other free CFD
| codes. My guess it is probably quite handy if you have a simple
| geometry.

Looks good though I guess it windows-only. What's your numerical scheme?


Thanks.
My numerical scheme is perhaps a bit unusual. It originated from
acoustic simulations, based on electric circuit equivalents. I also have
a web page on that:

http://www.xs4all.nl/~westy31/Electric.html

When I was doing acoustic simulations quite a few years ago, I was
wondering if I could adapt them so they also did Navier Stokes. The
acoustics were networks of masses an springs, or inductors and capacitors.
Eventually, I got it to work, I was quite pleased when i saw the first
vortex appear (acoustics has no vortices). So my code was always
time-dependent and compressible from the start.
A disadvantage is that the acoustic time step is very small, it is about
dx/c, so for a cell of 1 cm and c=300, you have timesteps of 30
microseconds.
But if you are interested in more or less incompressible flow, you might
as well drastically decrease the speed of sound in you model. This is
what I usually do, the velocities will be at something like Mach 0.1 in
the fluid.

Gerard

[..]
Looks good though I guess it windows-only.
[..]
Forgot to answer that.
It is written in Visual basic 6.0
I do not know if there are Linux or Mac versions of Visual basic.
If so, a Linux or Mac version could be made.

So yes, I guess it is windows only. If everyone really likes it, I might
make it cross platform.

Gerard

Gerard Westendorp wrote:
[..]

Looks good though I guess it windows-only.

[..]
Forgot to answer that.
It is written in Visual basic 6.0
I do not know if there are Linux or Mac versions of Visual basic.
If so, a Linux or Mac version could be made.

So yes, I guess it is windows only. If everyone really likes it, I might
make it cross platform.

Gerard
Gerard,


Where is the source code?

You said on the website it was opensource.

With the source some people on this group might be able to help you with
a cross platform transition.

Kind regards,


Jan Gerrit

Jan Gerrit Kootstra wrote:

[..]


Where is the source code?

I have just put it at the end of the web site.

With the source some people on this group might be able to help you with
a cross platform transition.

Well, that would be nice of course!
I think the actual computation would be not too hard translate.
The graphics might be more work.


Kind regards,


Jan Gerrit

Bruce Scott TOK wrote:
[..]

Thanks for that... I'll look at it in more detail after meeting season
in June. Shouldn't be too hard to turn the basic kernal into F90
runnable using g95.

cool!
I considered Fortran, but I found (using a stopwatch and a large number
of test computations) that the speed was about the same as VB. The
graphics seemed easier in VB.

I saw you have a check for negative densities...
wouldn't negative temperatures be more of a danger (for compressible
cases)?

Usually they both get negative very quickly, if velocities exceed the
speed of sound. The speed of sound is artificially lowered to allow
courser time steps. But sometimes, you overdo it.

Gerard