yade-dev team mailing list archive

[Bruno Chareyre <bruno.chareyre@xxxxxxxxxxx>]

> For that, it is necessary to start reasoning in total shear strain
> (precomputed in by summing increment or otherwise)
I have really no fundamental problem with "total" and "strain", but you 
will not get me to reason that way (like it is apparently impossible for 
you to just read and understand ScGeom and related code, while it would 
take no more than 10min).
For "total", I'm used to different algorithms, that are (for me) simpler 
and more versatile. I've read you dem3dof code, I know how it works, it 
is I think consistent but personnaly I don't need all this complexity. 
For "strain", it is not a relevant paradigm for the type of modeling I'm 
interested in (plus I still don't understand how strain can be related 
to rotation of bodies without enriched media or Cosserat-like theory).

>   + have one plasticity
> interface for use in Law2 via virtual functions that will be specific
> for each formulation.
>    
Plasticity interface? If (f>fmax) f=fmax, this is plasticity in 
incremental algorithm. Do we need a virtual function call for this? On 
this aspect I start to see a benefit in the inheritance proposal : we 
don't bring the complexity of Dem3Dof into ScGeom.

Bruno