[sanjibdsharma]

Hi,

I am new to OpenFOAM. However, I am undertaking a difficult project which involves the following:

1) Steady state, 3D Model,
2) Single Phase Turbulent Flow
3) Multiple species, Multiple Reaction
4) Density change due to liquid-to-vapor flashing
5) Local Grid refinement

The mesh is already created using GAMBIT. So need to transfer that in OpenFOAM format.

Any suggestions will help me immensely. If there are similar cases solved already, please send me a link or a copy of that. My email address is:

sanjibdsharma@gmail.com

Best regards,

Sanjib

[Rachel]

Sanjib
I would suggest to read more about two solvers: reactingFoam and dieselFoam. I guess they are not steady state solvers.
Refer to Chalmers OpenFoam tutorials for some details.

Local grid refiniment is possible only with static mesh. Guess adaptive local mesh refinement is not implemented in them.

Kindly keep us updated with this thread about the progress, as it seems like an interesting topic.

Rachel

[sanjibdsharma]

Thanks Rachel. Where do I find the Chalmers OpenFoam Tutorials ? Please send me the link.

Regards,

Sanjib

[Rachel]

Hi Sanjib,

A little bit of googling would have revealed the result

Anyways, here are the link.
http://www.tfd.chalmers.se/~hani/kurser/OS_CFD_2007/
http://www.tfd.chalmers.se/~hani/kurser/OS_CFD_2008/

[sanjibdsharma]

Hi Rachel,

I tried to use the reactingfoam solver for a very simple case. Since I wanted to include only species transport without any reaction, so what made the changes in the code accordingly (turning off the chemistry header files and chemistryProperties input file). However, since the thermodynamic data are required, so I kept the thermophysicalProperties input file. But this file access data from chemkin database with chemkin.inp and therm.dat file. Although I could compile the code, but while running it the therm.dat file could not be accessed.
1) Do I have to compile the chemkin first ?
2) How do I use thermodynamic data without the chemkin format in the thermophysicalProperties input file ?

Thanks for your help,

Sanjib

[Rachel]

Can you post the output of your log file?

I do not think that chemkin has to be compiled. IT is just the Chemkin format.

OF-1.6 does not use Chemkin format for some solvers. You might want to check new solvers in OF, e.g.- rhoReactingFoam

I have some problems with this new solver
http://www.cfd-online.com/Forums/ope...ctingfoam.html

Let me know if you succeed with this solver.

[sanjibdsharma]

Hi Rachel,

Thanks for your reply. I understood what changes are necessary. Finally I could compile and run the reactingFoam module for a non-reactive species transport case.

However, there are more serious issues which I am not able to understand. I am copying an earlier mail below:

************************************************** *****
I am new to OpenFoam, however have been using FLUENT for a long time. To start a new project in OpenFoam, I thought it might be good to compare a benchmark case between OpenFoam and FLUENT. So, I took turbulent flow over a BACKWARD FACING STEP. Following are my conclusions:

1) Substantial difference in the velocity, turbulent-kinetic-energy and turbulent-dissipation fields.

2) The reattachment length is shorter in OPENFOAM prediction; While there are two circulation zones, top and bottom of the jet, predicted by OPENFOAM, there is only one predicted by FLUENT, the bottom one.

3) While I can use higher order discretization schemes in FLUENT, in OpenFoam I can use only "upwind" which is a first-order, bounded scheme. When I used 2nd order schemes, the results showed instability and no convergence was achieved. Does this mean we have to use only "upwind" and cannot use the higher order schemes ?

4) I also wanted to test the linear solvers by replacing default PCG/PBiCG with GAMG. However, this did not work and the simulation stopped abruptly. Does this mean we can use only PCG/PBiCG ?

These are critical to my project and would help immensely to understand how OpenFoam works.

************************************************** ******

Regards,

Sanjib