yade-users team mailing list archive

[Jan Stránský <question628632@xxxxxxxxxxxxxxxxxxxxx>]

Question #628632 on Yade changed:
https://answers.launchpad.net/yade/+question/628632

    Status: Open => Answered

Jan Stránský proposed the following answer:
Hi Tina,

> Actually, I want to model breakage of this brittle material whose
porosity is no greater than 0.3. Besides, I also tried regularHexa, but
this approach also got a lower solid fraction.

for monodisperse packing, periodic randomDensePack produces porosity approx. 0.36. RegularHexa is the most dense monodisperse packing with 0.26 (in infinite space).
Boundaries always make porosity of these methods higher.

> in those questions, triaxial compression was recommanded

triaxial compreesion is used internally in randomDensePack

> a irregular boundary was used, triaxial compression also works on
irregular packing??

randomeDensePack compresses a cubic sample and then "crop" it to desired
shape. You can use similar approach, instead of 6 cubic walls, you can
use arbitrary number of box walls to compress particles arbitrary into
arbitrary shape (*)

> I tried filterSpherePack

filterSpherePack is again internally used by randomDensePack

According to your need:
- you need monodisperse packing and "porosity is no greater than 0.3", then regularHexa is your only choice (possibly with shift and orientation optimization)
- if you can have polydisperse packing, than:
  1) you can use standard randomDensePack/cubic triaxial test and play with PSD to get good results.
  2) you creates something similar to randomDensePack, but with boundaries actually matching your shape, see (*) above
  3) Alternatively you can use some geometric method, taking into account the boundaries (again reduces the boundary effect by cropping).

cheers
Jan

-- 
You received this question notification because your team yade-users is
an answer contact for Yade.