yade-dev team mailing list archive

[Bruno Chareyre <bruno.chareyre@xxxxxxxxxxx>]

> quick question about contact damping (hence defined at the level of 
> the contact law). I think the use of contact damping should 
> potentially be a possibility for every contact law. So if the law 
> functor operates on ScGeom functor, what would you suggest to do if 
> not copying into the contact law the whole code which aims to update 
> the shear force? Basically all it is needed for the contact damping 
> are shear and normal velocities at contact but they are computed every 
> time step in ScGeom.. Could we make those values available in the 
> contact law once requested?
One way would be to derive a ScGeom+velocities. It would inherit from 
ScGeom and add members incidentVn and incidentVs, which would be updated 
at the time they are calculated. It would still need to copy some parts 
of ScGeom code and include some "incidentUs=..." in it.
I'm sure there is a better way.
This question shows that each constitutive law is based on different 
concepts and needs different variables, which is currently the limiting 
factor in the unification of geometry classes.
Adding data incidentVel in ScGeom is not good since most laws don't need it.
Adding method getIncidentN/SVelocity() is not good either, since it 
would result in those quantities being computed twice at least.

Bruno

--
_______________
Bruno Chareyre
Associate Professor
Grenoble INP
Lab. 3SR
BP 53 - 38041, Grenoble cedex 9 - France
Tél : 33 4 56 52 86 21
Fax : 33 4 76 82 70 43
________________