solver non-convergence
Description
A carrier particle (D=200μm), coated with active drug particles (D=10μm), is studied in my case. When I build my case,something wrong occurs shown in figure below.
The .mfx and Particle_input.dat are attached.
Appreciate your help.
What is in the console before this message? The “MFiX Process has stopped” message is not an error (error messages are in red). This just means that the job ran to completion.
Without the .LOG files and with just the partial screenshot you submitted, it’s impossible to tell. Please use the “Submit bug report” menu item on the main menu which will create a .ZIP including all the job logs.
Also, take a look at the “Dashboard” pane to see what the stop time and current simulation time are. It looks to me like this job just ran to t=tstop
I set tstop=0.05s, however this case stop at Time: 0.11776E-05.
hopper_dem_3d_0831_2022-09-02T092338.448359.zip (324.2 KB)
Time: 0.0000 DT: 0.42821E-08 Remaining DEM NITs: 11676585
Wall Time - Elapsed: 0.02s Est. Remaining: 0.00s
WRITING VTP FILE : HOPPER_P_0000.vtp .
DONE.
Time: 0.0000 DT: 0.42821E-08 DEM NITs: 11676585
Time: 0.10705E-06 DT: 0.42821E-08 Remaining DEM NITs: 11676560
Wall Time - Elapsed: 0.02s Est. Remaining: *********s
Time: 0.21410E-06 DT: 0.42821E-08 Remaining DEM NITs: 11676535
Wall Time - Elapsed: 0.02s Est. Remaining: *********s
Time: 0.32116E-06 DT: 0.42821E-08 Remaining DEM NITs: 11676510
Wall Time - Elapsed: 0.02s Est. Remaining: -76048.05s
Time: 0.42821E-06 DT: 0.42821E-08 Remaining DEM NITs: 11676485
Wall Time - Elapsed: 0.02s Est. Remaining: -56874.59s
Time: 0.53526E-06 DT: 0.42821E-08 Remaining DEM NITs: 11676460
Wall Time - Elapsed: 0.02s Est. Remaining: -45377.41s
Time: 0.64231E-06 DT: 0.42821E-08 Remaining DEM NITs: 11676435
Wall Time - Elapsed: 0.02s Est. Remaining: -37712.54s
Time: 0.74936E-06 DT: 0.42821E-08 Remaining DEM NITs: 11676410
Wall Time - Elapsed: 0.02s Est. Remaining: -32241.44s
Time: 0.85641E-06 DT: 0.42821E-08 Remaining DEM NITs: 11676385
Wall Time - Elapsed: 0.02s Est. Remaining: -28134.30s
Time: 0.96347E-06 DT: 0.42821E-08 Remaining DEM NITs: 11676360
Wall Time - Elapsed: 0.02s Est. Remaining: -24942.95s
Time: 0.10705E-05 DT: 0.42821E-08 Remaining DEM NITs: 11676335
Wall Time - Elapsed: 0.03s Est. Remaining: -22380.06s
Time: 0.11776E-05 DT: 0.42821E-08 Remaining DEM NITs: 11676310
Wall Time - Elapsed: 0.03s Est. Remaining: -20211.41s
You have too much of an overlap between the small and large particle. This creates excessive force and an “explosion” of small particles that leave the domain.
Wall Time - Elapsed: 0.03s Est. Remaining: -20211.41s
negative remaining time looks weird - I’ll take a look at this
If you scroll up a bit in the console after the job terminates, you will see:
Program received signal SIGSEGV: Segmentation fault - invalid memory reference.
Backtrace for this error:
#0 0x7f552cdd4eff in ???
#1 0x7f54e0376fc0 in __desgrid_MOD_desgrid_neigh_build
at /usr/local/src/conda/mfix-solver-22.2.1/model/des/desgrid_mod.f:1060
#2 0x7f54e03a39e7 in __neighbour_mod_MOD_neighbour
at /usr/local/src/conda/mfix-solver-22.2.1/model/des/neighbour.f:56
#3 0x7f54e05d479d in __des_time_march_MOD_des_time_step
at /usr/local/src/conda/mfix-solver-22.2.1/model/des/des_time_march.f:219
#4 0x7f54e04138c1 in run_dem
at /usr/local/src/conda/mfix-solver-22.2.1/model/mfix.f:211
#5 0x7f54e04138c1 in run_mfix_
at /usr/local/src/conda/mfix-solver-22.2.1/model/mfix.f:146
Unfortunately, there were more messages printed after this and it’s easy to miss. It’s also not in the .LOG file. The SIGSEGV should be caught and turned into an error popup, this will be added to the next release.
Segfaults often occur when particles exit the domain.
Here’s a stack trace from gdb
#0 desgrid::desgrid_neigh_build ()
at /usr/local/src/conda/mfix-solver-22.2.1/model/des/desgrid_mod.f:1060
#1 0x00007f54e03a39e8 in neighbour_mod::neighbour ()
at /usr/local/src/conda/mfix-solver-22.2.1/model/des/neighbour.f:56
#2 0x00007f54e05d479e in des_time_march::des_time_step (nn=294)
at /usr/local/src/conda/mfix-solver-22.2.1/model/des/des_time_march.f:219
#3 0x00007f54e04138c2 in run_dem ()
at /usr/local/src/conda/mfix-solver-22.2.1/model/mfix.f:211
(gdb) list 1060
1055 do lj = ljc-jl_off,ljc+ju_off
1056 do li = lic-il_off,lic+iu_off
1057 lijk = dg_funijk(li,lj,lk)
1058 ltotpic =dg_pic(lijk)%isize
1059 do lpicloc = 1,ltotpic
1060 lneigh = dg_pic(lijk)%p(lpicloc)
(gdb) print lpicloc
$1 = 1
(gdb) print lijk
$2 = 45994
(gdb) whatis p
type = PTR TO -> ( char )
(gdb) print p
$3 = 0x0
I’m not sure where p is defined but it’s a null pointer, so attempting to dereference it p(lpicloc) causes a segfault.
Hope this helps,
– Charles
Thank you so much, this problem is solved.
I now have a new question, what should I do in order to get the number of detached particles after the collision, and the force between the particles in the collision?
Please start a new topic for a new question, this makes it easier for us to keep track of what’s been marked “solved”. Thanks.