yade-users team mailing list archive
-
yade-users team
-
Mailing list archive
-
Message #06360
Changes in timestep definitions
Dear all,
1/ The trunk version (daily snapshot package) no longer needs a
definition of dt in each script if GlobalStiffnessTimeStepper is present.
Before, PWaveTimeStep was generaly used to define dt manually for the
first iterations, until GSTS would find something.
Now, GSTS will use PWaveTimeStep as the default value automaticaly.
2/ There is now an option to use density scaling as developed by P.
Cundall to impose a fixed value of the timestep (dt=1 is convenient,
typically).
https://yade-dem.org/doc/yade.wrapper.html#yade.wrapper.GlobalStiffnessTimeStepper.targetDt
Disclaimer: as stated in the documentation, it makes the dynamics
completely unphysical
3/ The script example/TimeStepperUsage.py shows usage and speedup with
the different methods. The results for the compaction of the same
periodic cloud are given below.
I wish a good back to work to all of you who went for vacations. :)
Bruno
#Fixed dt = 0.8 * PWave timestep:
45k steps, 139 seconds
#dt dynamicaly set with GlobalStiffness timestepper:
13k steps, 47 seconds
#GlobalStiffness timestepper + density scaling:
8k steps, 27 seconds
--
_______________
Bruno Chareyre
Associate Professor
ENSE³ - Grenoble INP
Lab. 3SR
BP 53
38041 Grenoble cedex 9
Tél : +33 4 56 52 86 21
Fax : +33 4 76 82 70 43
________________