yade-dev team mailing list archive

[Václav Šmilauer <eudoxos@xxxxxxxx>]

> I was expecting comparisons using physical x/y/z actually, but it
> sounds consistent.
> Is the "angle" taking into account mixed modes of deformation (imagine
> shearing between two parallel planes with variable distance between
> planes)?
It only needs the current state (Cell::strain matrix and Cell::size
vector), nothing else. Any deformation is possible.
>
> What would you think of defining boxes like this, without distorsion
> of halfSize :
>
> aabb->min = shear*wrap(unshearPt(se3.position))-halfSize
>
> It would need a small change in the collider, but for something
> simpler (no wraping).
I don't think it would work. The collider needs one value for wrapping
coordinates between zero and that number (separately for each axis). I
didn't check carefully, though.

v