yade-dev team mailing list archive

[Vincent Richefeu <richefeu@xxxxxxxxxxxxxxx>]

Bruno,

I am not as confident as you. Imagine spheres that do not overlap on a regular grid.
If you impose a shear rate at constant volume (without gravity) by making the displacement of each sphere proportional to their height (the total height of the cell remain constant), I think that the spheres will never interact (!) By imposing velocity proportional to the height, the displacement of each sphere will be different due to the different inertia. So, it seems not exactly equivalent in my opinion.

An other point:
Even if periodic boundary condition works fine for triaxial loadings, it seems to me that the framework offered by yade deserves more than "tinkering". Otherwise I do not understand the interest! For this reason, I think it would be interesting to look in more detail what Gael and I say about H, r and s...
This is not something complicated and it is well-tested.

Vincent

Le 10 déc. 2009 à 15:27, Bruno Chareyre a écrit :

> Displacement/velocity?
> 
> It is totally equivalent for me, but proportional displacement seems easier and more robust (no risk of resetting the speed in an engine before it is actually computed, no special cases in Newton's law,...).
> I would go for proportional instantaneous displacement.
> 
> So, the scheme is :
> 
> Define the strain rate e'_kk in each direction
> 
> - e'_kk = constante
> or
> - e'_kk(t+dt) = e'_kk(t) + (stress_error)/mass
> 
> Assign proportional displacement dr = e * r
> 
> We are still discussing this here, I'm anticipating a little, but I think we can converge to something compatible with that.
> 
> Bruno
> 
> 
> -- 
> _______________
> Chareyre Bruno
> Maître de Conférences
> 
> Grenoble INP
> Laboratoire 3SR - bureau E145
> BP 53 - 38041, Grenoble cedex 9 - France
> Tél : 33 4 56 52 86 21
> Fax : 33 4 76 82 70 43
> ________________
> 
> 
> _______________________________________________
> Mailing list: https://launchpad.net/~yade-dev
> Post to     : yade-dev@xxxxxxxxxxxxxxxxxxx
> Unsubscribe : https://launchpad.net/~yade-dev
> More help   : https://help.launchpad.net/ListHelp