Process exit ! how to solve it

Hi, everyone

when i launched my case, a confusing problem stopped it.

“Process exit D:\anaconda\envs\mfix-20.1.2\Scripts\mfixsolver.exe -s -f D:\MFiX\fluidizedbed\fluidizedbed.mfx”

Did you meet it before? how do you solve it? i can’t find any problem in my setup, though it maybe seems very simple and rough.

Please help me fluidizedbed.mfx (19.7 KB)

Hi Wuming.

Some versions of MFiX had an issue where backtraces were not being printed when the solver hit a problem at runtime, leading to these unexplained “Process exit” messages. Was there anything in the console, or the *.LOG files?

I ran your case on my machine, using the most recent version of the code.
After a few warnings, it produced this error:

Program received signal SIGSEGV: Segmentation fault - invalid memory reference.

Backtrace for this error:
#0  0x7f9192e55c02 in ???
#1  0x7f9192e54e45 in ???
#2  0x7f91df9c5b7f in ???
#3  0x7f9193306064 in __set_ic_dem_mod_MOD_set_ic_dem
	at /home/cgw/Work/NETL/mfix/model/des/set_ic_dem.f:86
#4  0x7f91932dc5cc in __make_arrays_des_mod_MOD_make_arrays_des
	at /home/cgw/Work/NETL/mfix/model/des/make_arrays_des.f:242
#5  0x7f91931b39ba in __main_MOD_initialize
	at /home/cgw/Work/NETL/Cases/fluidized_bed/build/pymfix/main.f90:389
#6  0x7f919334849e in run_mfix_
	at /home/cgw/Work/NETL/mfix/model/mfix.f:132
#7  0x7f91931b3c42 in __main_MOD_run_mfix0
	at /home/cgw/Work/NETL/Cases/fluidized_bed/build/pymfix/main.f90:81
#8  0x7f91931a6206 in f2py_rout_mfixsolver_main_run_mfix0
	at /home/cgw/Work/NETL/Cases/fluidized_bed/build/f2pywrappers/mfixsolvermodule.c:1270

so something is going wrong with the simulation. We are looking into it. If it’s an MFiX bug we will try to include a fix in the upcoming 20.2 release.

  • Charles

The array that stores particles species is only allocated if the energy equation is turned on. We will move the allocation in the 20.2 release. In the meantime, please turn the energy equation on and makes sure all ICs and BCs are set properly for the energy equation. Use appropriate Initial and BC values for k and epsilon (zero epsilon is not good). Your mesh is also not good. I personally like to extend the geometry past the inlet and outlet planes to get a proper intersection.

Hi, Charles

Thank you for your reply. Today, i ran my case again on another PC after make some revising to it, according to Jeff’s advice. The process does not exit any more.

I guess it may be caused by my poor mesh. After all, all settings are perfect, it is impossble not to run.

wuming

Thank you, Jeff.
According to your advice, it ran successfully. But i still have a question, why there are not default values set for turbulent models, which seems constant for normal use.

I have the same problem and I need to know how I can check the mesh quality and how I can improve it?
Can anyone helps me on what the important mesh parameters to check are and how I can improve them?

Hi, please read the user manual of new version of MFIX. There are some description. To learn nore, you could read some relevant book about mesh generation.