Bug Reports for CFDEM®coupling, LIGGGHTS®, and ParScale

(Possible) bugs / suspicious behavior should be reported as a new thread here, not in the user forum. This should give both you and the developers a clear idea about the status of your bug report (submitted/assigned/fixed)

cfdemSolverIB problems with CFD time-step

Submitted by paulaalejandrayo on

Hello, I am using cfdemSolverIB in a simple case, trying to measure effective permeability of a pack of spheres. I need to resolve the flow behaviour explicitly. I based my case on the twoSpheresGlowinskiMPI tutorial.
I have two cases with the exactly same configuration except for the time-step for the CFD part (and the coupling step) (DEM time-step is kept constant). Although the behaviour in both cases looks physically plausible, the pressure drop behaviour change dramatically with the value of the time-step for CFD solver.

Bug in fix insert/stream for the insert direction

Submitted by fengyi on

As said in website: https://www.cfdem.com/media/DEM/docu/fix_insert_stream.html , "The insertion volume is generated by extruding the insertion face as specified via insertion_face in the direction of the face normal. "

Howerver, when I use the cfdemSolverPiso slover, I just invert the face normal of STL file as insert_face, and the insertion direction didn't change according to the postprocessing results.

Restarting Multisphere with fix wall/gran and changes to fix property/global

Submitted by estefan31 on

I think this is a bug with multisphere and read_restart when either fix property/global is used or when fix wall/gran is used. If fix property/global is changed between runs and then a restart file is saved, then the atoms will not be reloaded when read_restart is called in a second input file. If the second input file also includes a fix wall/gran sometime after read_restart, then the code will either crash or throw an error. I notice this happens when multisphere is run in parallel and on a single processor.

Bug in cfdemSolverPiso for nonNewtonian Viscosity Models

Submitted by AliBlues on

Hi,
There seems to be a bug in the cfdemSolverPiso as it does not update viscosity of the flow based on the selected viscosity Model. For example, I’ve selected the powerlaw viscosity model in the periodic channel tutorial with the appropriate model parameters. But the solver just doesn’t update the viscosity. I went to the powerLaw model class definition and in the calcNu() function of the class added the line (Info<<”COMPUTING VISCOSITY").

Error when using make auto on version 3.6.0 (OS RedHat)

Submitted by mschramm on

When using the make auto, I receive the following error;

[mschramm@mschramm-precision-tower-5810 src]$ make auto
mv: cannot stat `style_contact_model.tmp': No such file or directory
ls: cannot access /usr/include/vtk*: No such file or directory
make[1]: Entering directory `/home/mschramm/LIGGGHTS/LIGGGHTS_3.6.0/LIGGGHTS-PUBLIC/src/Obj_auto'
Makefile:468: *** Could not obtain VTK_MAJOR_VERSION. Stop.
make[1]: Leaving directory `/home/mschramm/LIGGGHTS/LIGGGHTS_3.6.0/LIGGGHTS-PUBLIC/src/Obj_auto'
make: *** [auto] Error 2

(Possible ) bug when multi-sphere and periodic boundary are used together

Submitted by yidongxia on

I've observed some strange behaviors of multi-sphere particle motions when periodic boundaries are used.

For example, when I inserted multi-sphere particles (with gravity) into a hopper, and assign periodic boundaries in the direction normal to the cross-section, all particles are supposed to drop down due to gravity. But lots of particles strangely appear on the very top of the vertical boundary after a while, and they never move! (see the attached figure).

Versions are right, but it says my OF version is not supported.

Submitted by Zekun Wang on

************************************
ERROR from your CFDEM code!
ERROR: your OpenFOAM(R) version is not supported!
Please use a version that is supported, i.e., see the bashrc file in the source directory!
************************************

bug in massflow/mesh for clumps when mesh is near processor border

Submitted by richti83 on

When the mesh used in massflow/mesh is to close to a processor border the particles are counted more than once because the counter is not immediately communicated to the particles which belong to the same clump but are on the other processor. I would like to communicate the property "ms_counter_" after every change of it (line 446 and 509). How can one achive this ?
Testscript and screenshot attached. Works with one core or when mesh is move 200mm in Z direction, but not with 4 procs and mesh in 0/0/-0.5.

compute_inertia_molecule.cpp

Submitted by robin on

Good afternoon to all of you,

[A little bit of context: I've modifyed compute_inertia_molecule.cpp to compute the inertia of a group of particles. Verifying my new file, I think I've spotted a little typo in the original file.]

In the loop summing the contributions from all the atoms (lines 190, 191, 192) we should use the distance between each atom and the global center of mass of the molecule (called comall in the file). In compute_inertia_molecule.cpp, another variable is used instead of comall, leading to a wrong result.