CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   OpenFVM project (http://www.cfd-online.com/Forums/main/15972-openfvm-project.html)

x-flow November 5, 2008 08:19

OpenFVM project
 
OpenFVM is a general CFD solver released under the GPL license. It was developed to simulate the flow in complex 3D geometries. Therefore, the mesh can be unstructured and contain control volumes with arbitrary shape. The code uses the finite volume method to evaluate the partial differential equations. As well as solving the velocity and pressure fields, the code is capable of solving non-isothermal multiphase flow.

Anyone willing to rollup their sleeves and help develop an open source CFD software is welcome to join the OpenFVM project. The code can run in parallel mode using PETSc. One object is to tackle fluid structure dynamics. The code is small and simple to understand. Reference manual is available and explains the main numerical methods used.


Ahmed November 7, 2008 03:10

Re: OpenFVM project
 
please read the following post http://www.cfd-online.com/Forum/main.cgi?read=62043


x-flow November 7, 2008 09:09

Re: OpenFVM project
 
Hi,

I understand the frustration it is to download a source code and then not being able to compile it. OpenFVM has been from the beginning supporting both windows and linux. The objective is to be cross-platform like Gmsh is. I have included also binary version of the solver at sourceforge. However, the source code is easy to compile on both systems. The major problem is getting the windows parallel version to work since PETSc does not maintain a windows version. But even for this, there are people working on that too.

Right now, the objective is also to develop an interactive GUI which is also cross-platform to eliminate the boring assignment of BC and editing of input files. I think this will allow new users to play around with the solver.

Another objective is too link VOFTools or another mesh cutting scheme to improve free surface computation. It would be nice to have fluid-structure interaction.


Ahmed November 8, 2008 01:05

Re: OpenFVM project
 
From the download page:

1. Download LASPACK and copy the source files (*.c, *.h) to directory "source/laspack" (required)

2.

3.

4. Edit LASPACK source files and replace all occurences of '#include "laspack/' with '#include "' or type: "./fixlaspack"

How about including the modified LASPACK source files in your download. make it simple for others to install your programme on any linux distro, to say it in a clear way, make your programme platform independent

Oliver November 8, 2008 12:15

Re: OpenFVM project
 
Why don't YOU do that and provide the solution back to the original author???

Ahmed November 8, 2008 12:18

Re: OpenFVM project
 
Very simple, he has done it, he claims that the code is working

x-flow November 10, 2008 11:13

Re: OpenFVM project
 
That is only version 0.1 or 0.2. That was because LASPACK was not initially included in the source.

Get the latest version from SVN and try to compile the serial version on Windows or Linux. It should compile with no problems.

The parallel version might be more difficult to compile on windows because it depends on PETSc and ParMetis.

Keep in mind that it is not perfect but I would like it to be always easy to compile and understand.


Fab November 10, 2008 13:49

Re: OpenFVM project
 
Hi x-flow,

great work! I like it a lot :) Though I am not helping developing it (yet)... it is a great program for self-education.

Greetings! Fab

joe shmo November 25, 2008 00:13

Re: OpenFVM project
 
Why would I work on this as opposed to OpenFOAM?

Please convince me.

x-flow December 5, 2008 05:37

Re: OpenFVM project
 
OpenFOAM is great code, I have worked with it myself. I would say OpenFVM would be very good starting point to learn finite volume methods and then if one would like to explore more they could go to OpenFOAM which is more flexible but also more complex.

To sum up, OpenFVM allows easier manipulation of the code and what you learn in the end is useful to get your hands dirty with OpenFOAM. If you are not CFD/C++ expert, I think it is best to start with OpenFVM. OpenFOAM is very generic in the range of CFX, Fluent and Star-CD, while OpenFVM is more focused on free-surface (VOF) tools in the range of Flow3D.


shirazbj April 12, 2010 09:48

I like the documentation, very detailed eqations, good to learn coding.

Is there any possible it will outputs VTK file for post-processing with ParaView?

Thanks

Hades December 14, 2010 12:29

I need some help...
 
Quote:

Originally Posted by x-flow
;57656
OpenFVM is a general CFD solver released under the GPL license. It was developed to simulate the flow in complex 3D geometries. Therefore, the mesh can be unstructured and contain control volumes with arbitrary shape. The code uses the finite volume method to evaluate the partial differential equations. As well as solving the velocity and pressure fields, the code is capable of solving non-isothermal multiphase flow.

Anyone willing to rollup their sleeves and help develop an open source CFD software is welcome to join the OpenFVM project. The code can run in parallel mode using PETSc. One object is to tackle fluid structure dynamics. The code is small and simple to understand. Reference manual is available and explains the main numerical methods used.

Hi, i'm a new user of the OpenFVM and I have some small problems with it. The first - how to run the environment (OpenFVM with all programs) on Win XP? It is possible? Please give me tips... Second - how to quickly operate on the OpenFVM - only from the command line? Perhaps will be better to change the operating system? Thanks for your time. :)

shirazbj December 14, 2010 22:04

control panel/system/advanced/environment variables

Hades December 15, 2010 04:25

Quote:

Originally Posted by shirazbj (Post 287482)
control panel/system/advanced/environment variables

Dear Cean, in this location I can't find OpenFVM... one more: I know how to generate grid in gmesh, but the next step... the program is generating some errors - it's looks like the program has problem with libraries - OpenFVM ver. 1.3 can work independently? Please, if you can, write me all sugestion how to run OpenFVM properly.
I have read the pdf manual and in my opinion all informations are connected with Linux OS.

shirazbj December 15, 2010 22:08

Have you downloaded the v1.3 windows binary openfvm.exe at http://sourceforge.net/projects/open...M.exe/download

1. Open a DOS terminal on xp and "cd\" to change to root dir
2. "md openfvm" to creat the dir
3. "cd openfvm" into it
4. save your downloaded openfvm.exe into this dir
5. then goto the environment variables, under "system variables" scroll down to Path, edit it to include ";c:\openfvm"
6. you can run openfvm in anywhere now.

Hades December 18, 2010 19:10

Thanks for your help...

Hades January 25, 2011 09:21

Quote:

Originally Posted by shirazbj (Post 287482)
control panel/system/advanced/environment variables

Dear Cean, perhaps I'm still doing somethig wrong but OpenFVM isn't workig properly. I can't see any results... when I try to use procedure "tutorial f 1", program stops. What else could be wrong?

shirazbj January 27, 2011 08:11

what is the displayed error when the program stopped?

you could also ask at http://sourceforge.net/projects/openfvm/forums

Hades February 21, 2011 04:31

Quote:

Originally Posted by shirazbj (Post 292449)
what is the displayed error when the program stopped?

you could also ask at http://sourceforge.net/projects/openfvm/forums

Thank you for your note and sorry for the absence. I signed up for a parallel forum ...
Here is the error:

Error : Surface 23 cannot be meshed using the transfinite algo - the same in four lines - 15, 19 and 27.

What it means? Can i fix it?
What happens if I ignore these errors?

gchnhn October 24, 2011 12:47

Thank you very much,x-flow! This is the right thing I have been seeking for several m
 
Yes! what you said is quite right! Not everyone who needs a fvm code is expert of C++ and have solid foundation of CFD theory. I am a newbie of CFD, I know the word "CFD" since April, 2011. I must finish understanding of the code and do some improvement to get my master degree. Should I have done this, I will leave CFD and this forum forever. I have zero interest in CFD.

X-flow,thank you again!


All times are GMT -4. The time now is 15:25.