error with restart file in 1.2.8

raguelmoon's picture
Submitted by raguelmoon on Mon, 04/11/2011 - 05:08

Hi,
I get error when I run simulation from restart file, the error message is :

......
ERROR: Could not locate a fix/property storing value(s) for coefficientRollingFriction

However, when I run simulation without MPI then it works fine (with restart).

Thanks,
Ram

Forums:

ckloss's picture

ckloss | Mon, 04/11/2011 - 06:58

can you elaborate a bit more on what kind of simulation you are running and what input scripts you are using? Otherwise it's impossible to tell why this is

Christoph

raguelmoon's picture

raguelmoon | Mon, 04/11/2011 - 07:44

Hi,
I just poured about 500000 particles in simulation box and created restart file, an then i continued the simulation with that restart file. It is working fine without mpi. I want to insert mesh after restart file. However, without mesh the error message is same. I have three types of particles in my simulation. I poured type 1 particles (with two different diameters). Type 2 would be used for mesh while type 3 is for wall.
The material properties:

fix m1 all property/global youngsModulus peratomtype 2.0e6 2.0e6 2.0e6
fix m2 all property/global poissonsRatio peratomtype 0.45 0.45 0.45
fix m3 all property/global coefficientRestitution peratomtypepair 3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3
fix m4 all property/global coefficientFriction peratomtypepair 3 0.1 0.27 0 0.27 0.27 0 0 0 0

And, i used walls as:

fix xwalls all wall/gran/hertz/history 1 0 xplane 0 0.25 3
fix ywalls all wall/gran/hertz/history 1 0 yplane 0 0.25 3
fix zwalls all wall/gran/hertz/history 1 0 zplane 0.0 NULL 3

Thanks,
Ram

Ram

raguelmoon's picture

raguelmoon | Tue, 04/12/2011 - 07:29

Hi,
Is it not possible to insert mesh after restart file? I did it but it is not dumping any force/stress to mesh.
Thanks,
Ram

Ram

andrea.pasquali | Tue, 04/12/2011 - 11:25

Hi Ram,
Do you want to insert mesh like .stl?
I inserted .stl after particles setting in my simulation to move it inside particles (with restart).
If I run the movecad directly it seem like the particles are "ghosts" and the .stl doesn't interact with them.
If I run first a "empty" run inserting the .stl only (just few iterations) and then (from the new restart) moving the cad the simulation is ok.
I suggest you to run first to read the .stl and then to move it

Regards

Andrea

raguelmoon's picture

raguelmoon | Tue, 04/12/2011 - 14:46

Hi Andrea!
Thanks for help, I will try this. But I can not run restart file with mpi, it stuck with following message:

ERROR: Could not locate a fix/property storing value(s) for coefficientRollingFriction

Thanks,
Ram

Ram

andrea.pasquali | Tue, 04/12/2011 - 14:57

Hi Ram,
I can't help you about "coefficientRollingFriction".
Have you defined it as "property/global"? Have you defined it also in in. file where you read the restart?

Regards

Andrea

ckloss's picture

ckloss | Wed, 04/13/2011 - 07:52

Hi Ram, are you using the rolling friction model in a pair style?

Christoph

raguelmoon's picture

raguelmoon | Wed, 04/13/2011 - 11:25

Hi,
I debugged the problem. We don't need to comment the following line (only with new version, old versions are valid for both options):

pair_style gran/hertz/history 1 0 #Hertzian without cohesion

This solved the problem.
Thanks...
Ram

Ram

raguelmoon's picture

raguelmoon | Wed, 04/13/2011 - 18:51

hi,
i mean when i dont write above statement in restart script (in older versions) then there is no problem, but in new version i have to write the above statement...
Best,
Ram

Ram