yade-users team mailing list archive

[chiara modenese <c.modenese@xxxxxxxxx>]

2010/4/1 Václav Šmilauer <eudoxos@xxxxxxxx>

> > I see in Ig2_Sphere_Sphere_ScGeom that part of the condition we create
> > a ScGeom instance is:
> > penetrationDepthSq>0 || c->isReal()
> > Is this condition accessed only for potential interactions previously
> > created? I mean, potential interactions are created because overlap
> > occurs between bounding boxes, right? So has this class the power to
> > create the interaction alghough no potential one exists?
>
> No, it doesn't have such power: the functor is called for all
> interactions that exists already (both potential and real); it is the
> collider that creates potential interactions for overlapping bboxes. Ig2
> functors in general can turn potential interaction into a real one (as
> their purpose is exact collision detection + updating existing
> geometry); they however cannot (|| c-> isReal()) erase an interaction
> that already exists, even if spheres are far apart (it is Law2 that is
> supposed to handle this).
>

Ok, so at least a potential interaction should exists before this functor is
called. BTW, do we usually enlarge the bboxes? Is that needed if I use this
detectionFactor? I mean, if the bboxes perfeclty fit the spheres, and say
the boxes have the same coordinates along one axis, the potential
interaction is not created if we do not have a real geometrical indentation,
right?
cheers, Chiara


>
> > I need a distance of separation (that will depend on various
> > parameters) between particles for the creation of the interaction.
> > Should I derive a new Ig2 for sphere-sphere interaction? Now I see
> > that there is a detection factor that would do something similar but I
> > need something just a little more complicated.
>
> If you derive from the geometry functor, you will end up duplicating
> fairly complicated geometrical code; I would advise against that.
>
> If I understand, you will have no interaction unless there is some
> distance between the spheres?
>
Yes

>
> The easiest would be to set the detection factor to minimum value that
> satisfies your needs (e.g. 1.5 if you need sometimes 1.2, sometimes 1.3
> and sometimes 1.5) and then, in the Law2 functor, make the decisions. If
> the particles are too close, you will just exert no forces on particles.
>
mmh, I actually have a precise formulation that tells me how much should be
this distance (depeding for instance on the equivalent radius of particles
in contact and other variables). But agree that deriving a new Ig2 would
lead to a duplication.

>
> Cheers, Vaclav
>
>
> _______________________________________________
> Mailing list: https://launchpad.net/~yade-users<https://launchpad.net/%7Eyade-users>
> Post to     : yade-users@xxxxxxxxxxxxxxxxxxx
> Unsubscribe : https://launchpad.net/~yade-users<https://launchpad.net/%7Eyade-users>
> More help   : https://help.launchpad.net/ListHelp
>