← Back to team overview

yade-users team mailing list archive

Re: Initial time step

 

> for the triaxial I want to use the PWaveTimeStep() as a first estimate
> and then going on with the GSTS. But if I include in the python script
> o.dt=.5*utils.PWaveTimeStep()
> then it seems it is only using this value although the TT generator
> file is using GSTS. How is that I give an initial estimate if not as
> above?
It is a "feature" that I recall now. If you set timestep via O.dt, it
will disable timesteppers (otherwise the change would be only effective
until the next run of the timestepper). You can re-eanble timesteppers
with
http://beta.arcig.cz/~eudoxos/yade/sphinx/yade.wrapper.html#yade.wrapper.Omega.usesTimeStepper .

Not very logical, as I see now (and not documented either). I would
change it as follows: if you set O.dt to a non-positive  value, then
timestepper will be enabled (and exception raised if there not one). If
you set O.dt to a positive value, timestepper will be disabled. Does
that sound better?

O.usesTimeStepper will then be redundant.

For your question now,

O.dt=.5*utils.PWaveTimeStep()
O.usesTimeStepper=True

should set the initial Δt value and hand control over to GSTS.

Cheers, v.






Follow ups

References