yade-users team mailing list archive

[Václav Šmilauer <eudoxos@xxxxxxxx>]

> 1- "Investigation of impact fracture behavior of automobile laminated
> glass by 3D discrete element method" (M. Y. Zang · Z. Lei · S. F.
> Wang,2007)
> http://www.springerlink.com/content/9w724n3p43420211/?p=24a11cb5f9bd408a8041e03961c7ec13&pi=0
> but Kn and Ks in this paper is not correct.
Ok, they use 3-DOF contacts (normal and shear forces; no torsion and
bending) with cohesion and simple failure criterion.

> If it is possible I want use Kn , Ks in paper of Wang , Mora
> 2- "Macroscopic elastic properties of regular lattices" (Yucang Wang ,
> Peter Mora,2008)
> http://dx.doi.org/10.1016/j.jmps.2008.08.011
Hm, they're publishing what's been known for quite some years I think
(e.g. http://www.springerlink.com/content/k1rctjk0vqwf9juq/ ). The fact
that macroscopic E is proportional to Kn and macroscopic ν to Kt/Kn can
be shown by trivial dimensional analysis as well (there are 2 parameters
describing contact-level elastic behaviro, i.e. Kn and Kt/Kn; and 2
macroscopic properties with same dimensions, E and ν).

> 3- A finite deformation method for discrete modeling: particle
> rotation and parameter calibration(Yucang Wang · Fernando
> Alonso-Marroquin,2009)
> http://www.springerlink.com/content/412t27w5955643jl/

I read this paper; Anton mentioned it was one of his priorities to
implement this kind of contact deformation, but I don't know how he
feels about it now.

As of now, there is no class to compute 6 DOF relative displacement of 2
spheres (both ScGeom and Dem3DofGeom compute only normal + shear, but
not bending and torsion), though following that paper, it shouldn't be
difficult to implement.

It seems that you don't need that for the Zang+Lei+Wang paper (they use
just 3 dofs)

> I saw the files of yade but I don't underestand how the Kn , Ks are
> calculated,
> I know that Cpm is for Concrete but in the other law it seems that
> there is no tension, 
> the model in the above papers is not for particular material and Kn,
> Ks is only depend E, poisson' ratio and sphere radius. 
> You think I can use your models? which model?
> or I have to write my code?

There is no use for Cpm in yout case, it is too complicated. You
definitely have to write you code.

For constitutive law itself, it is rather easy; you essentially need to
give the failure criterion, and write a few lines that will compute
forces from displacements.

If you need to use their formulas to compute contact stiffnesses
(42,43), you will have to write an Ip2 functor (not difficult either).

Before you get into that, make sure you understand how Yade works.
Mailing list cannot replace you reading the documentation thoroughly
(https://www.yade-dem.org/sphinx/, work in progress).

Cheers, Vaclav

v