yade-users team mailing list archive

[Bruno Chareyre <question660585@xxxxxxxxxxxxxxxxxxxxx>]

Question #660585 on Yade changed:
https://answers.launchpad.net/yade/+question/660585

Bruno Chareyre posted a new comment:
Hi,
Just to summarize a few ideas after a discussion with Robert:
- the duplicated "setPositionsBuffer()" in DFN was definitely an experimental hack and it is unsafe to keep relying on this.
- removing this duplicate would let two-way-coupling work as it should (since the inherited setPositionsBuffer is not conditional).

What is really needed is 1/ a way to update permeability data based on
cracks extension and aperture and, 2/  to rebuild the permeability
matrix after that. To be sure no remeshing will occur it is enough to
set FlowEngine's meshUpdateInterval and defTolerance negative.

Further: if permeability is only updated in the cracks (not in intact
matrix) some performance could be gained by using cholmod
update/downdate features (cholmod_updown), which updates factorization
without going through the complete Cholesky algorithm (even changing the
size of the problem for advancing cracks seems to be possible).

Bruno

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