yade-users team mailing list archive

[chiara modenese <c.modenese@xxxxxxxxx>]

> Even with linear law, it was giving strange results, see modified script
> attached. You need to at least damp the cell deformation.
>
Oh sure, it was zero in the script but I was actually changing it to see
what happened.

Can you get stable systems with similar properties, timestep, and with rigid
> boundaries?
>
I am only using rigid boundaries, ATM. This script with the PB was only a
try since I am having troubles with the time speed using TT class. HM works
fine with boxes, though I am facing few other issues I hope to solve soon.
BTW, calibration is needed anyhow. Reason: to maintain the computing time at
a reasonable level, I cannot (though ideally one could do it) assume the
real young modulus of the particles. It is possible to find out a threshold
value for the normal modulus which will not then have an influence on the
bulk properties. I would have a general question related to this:

What is a reasonable range of values for the overlapping in a quasi
static-simulation like the TT? 1% of particles sizes? If I reduce too much
the interparticle stiffness I fear I could have unacceptable overlappings
but I do not know starting from what value would be such. Thanks.

I would double check the values of relative velocities and see if they have
> something wrong.
>
Yep, I will do it. Indeed the problem arises when particles are free to
rotate. If rotations are not allowed then the system reaches a stable state.

Chiara