yade-users team mailing list archive

[Klaus Thoeni <klaus.thoeni@xxxxxxxxx>]

Hi Bruno!

> >> I realize : by doing that, you will basically do exactly the same as I
> >> did with cylinders. David's model was based on spherical nodes
> >> connected by cylinders, right? So, it will need more work to
> >> (re)implement this from scratch than to use what is already in Yade.
> > 
> > No, David just used particles with a remote interaction model. Particles
> > only exist where wires intersect.
> 
> Well, nothing exists in a numerical model, right?
> David really simulated contacts between stones and cylinders (wire) linking
> the nodes IIRC, even if PFC was not "really" simulating the cylinders. It
> is similar to my geotextile model (link in previous mail).

David did consider contact between spheres only (see papers). The lines you 
can see in his pictures are not cylinders just lines introduced by the post-
processor.

> This is the the part implementation that you should use instead of starting
> from 0 (contact detection, sphere-cylinder contact geometry, ditributing
> forces on the nodes, it's all in the chained cylinder model already).

I will just use sphere-sphere/facet contact geometry for now and move on the 
cylinder approach once the first approach is working.

> For piecewise linear behaviour of steel, you are right, there is nothing
> like that at the moment.
> Perhaps have a look at HertMindlin law to see what it needs to implement a
> non-linear law (though it not fundamentaly differs from linear
> elastic-plastic in terms of code)

You think I can use an ElastMat material and pack everything which concerns 
the non-linear behaviour in the contact physics or should I create a new 
material?

Thanks

Klaus