[TUFAN]

hii...
i have just started using this fluent software, and i am facing some problems.
i need to develop MHD modelling of plasma flow over a flat plate........if anyone has a mhd code..(better if related to plasma// otherwise will also do)...could you plzz mail the document to me at my email address.....
....would be really greatful for your help...
tufan

[kel85uk]

Hi Tufan,
From what I've learnt from my own undergrad thesis project, it's better to just write your own code. The different regimes of flow in MHD alone can lead to different equations and different methods in writing the code. Most codes are ideal MHD, while some have joule heating effects (resistive MHD), and some such as the MACH code (it's not available to non-US citizens btw) have a single-fluid/two temperature formulations, viscous, and compressible (very nice to have if you're trying to model pulsed plasma thrusters!). All depends on the nature of your flow problem. That's why in this case, it's better to just write your own code, based on the governing equations which are required to solve your problem, saves a lot of time and hassle also to try and decipher another person's source.

If you must, here're some of the codes (note you'll need access to a Linux OS with their support libraries and compilers installed since they're all research codes):

CLAWPACK (ideal MHD, MHDCLAW module written by rossmanith)
BEARCLAW (similar to clawpack)
FLASH (Hydrodynamics code which has general resistive MHD)
ZEUS and ATHENA
Code_Saturne (For arc simulations, incompressible, resistive MHD)
OpenFOAM
MATLAB (Try to code yourself. ;-) )

And on that note also, the add-on MHD module in FLUENT is junk. The results you get if you specify using the electric potential method will lead you to crappy solutions and using the magnetic induction method which most codes do, is rather like using a black box. We don't know how they're ensuring a divergence free magnetic field, and also how they're ensuring that the pressure remains positive. Are they using Powell's 8-wave, or are they using constrained transport by evans & hawley? We just don't know, and these are the methods which sometimes can make or break your solution to the problem.

All the best to you. It's gonna be a long project if you're trying to simulate DBD plasma actuators.

Regards,
Kelvin

[TUFAN]

thanks kelvin for your reply...i will definitely pay heed to all you said. you saved me from a lot of trouble and time, since i had no idea that the mhd add-on of fluent was crapy.....thanx...