On 3 Apr 2022, at 11:36 am, Gao, Ankang <ankang.gao@imperial.ac.uk> wrote:

Hi Daniel,

Thanks for your interest in Nektar++.

Nektar++ doesn't support elements with zero Jacobian. You still need to smooth the cusp geometry.

Cheers

Ankang

---

From: nektar-users-bounces@imperial.ac.uk <nektar-users-bounces@imperial.ac.uk> on behalf of Daniel Lester <daniel.lester@rmit.edu.au>
Sent: 31 March 2022 00:15
To: nektar-users <nektar-users@imperial.ac.uk>
Subject: [Nektar-users] Meshing of random packed spheres

 

Hi Nektar Users,

I’m not a Nektar user but I am considering using this code for a project involving computation of flows across the interface of a porous medium comprised of random packed spheres (RPS) - i.e. flow within the medium, the boundary layer and the turbulent bulk. Previous finite volume models of RPS involved smoothing of the cusp-shaped geometry at the contact points between spheres.

Would this geometry be feasible to mesh and run in nektar if the # of spheres was order 100's, or would this be getting too fiddly/complex? Any feedback greatly appreciated,

cheers

Daniel