mesh

91 views
Skip to first unread message

王艺康

unread,
Oct 22, 2024, 3:38:50 AM10/22/24
to deal.II User Group
 I would like to ask a question, I have drawn a u-shaped tube and meshed it (hexahedral mesh). I also have a stl file of a real heart model. I want to use dealii to map the mesh of the u-shaped tube to the real model how do I do that or are there any tutorials I can borrow. I'm very grateful for any suggestions!

Wolfgang Bangerth

unread,
Oct 22, 2024, 2:01:52 PM10/22/24
to dea...@googlegroups.com
On 10/22/24 01:38, 王艺康 wrote:
> **
You will of course need some function y=f(x) that for every point x in your
tube model returns the corresponding point in the heart model. If you have
such a function, you can use
GridTools::transform()
to transform one mesh into the other.

Best
W.

--
------------------------------------------------------------------------
Wolfgang Bangerth email: bang...@colostate.edu
www: http://www.math.colostate.edu/~bangerth/


王艺康

unread,
Oct 22, 2024, 10:02:50 PM10/22/24
to deal.II User Group
Dear Prof Wolfgang
Thank you very much for your advice!
I don't have a function for this mapping relationship, I just have two such models now (a simple u-tube and a real left ventricle), what should I do if this is the case. Thanks a lot!

王艺康

unread,
Oct 23, 2024, 11:14:41 PM10/23/24
to deal.II User Group
1) The ideal left ventricle is modelled as a semi-long-axis ellipse and I use gmsh to get the mesh of the ideal model.
2) I have a model of the real left ventricle, if I transform the surface point coordinates of the real model into a cylindrical coordinate system and then use the cylindrical coordinate system to move the ideal mesh
I would like to enquire about the feasibility of this

Wolfgang Bangerth

unread,
Oct 24, 2024, 9:23:47 AM10/24/24
to dea...@googlegroups.com
On 10/23/24 21:14, 王艺康 wrote:
> 1) The ideal left ventricle is modelled as a semi-long-axis ellipse and I use
> gmsh to get the mesh of the ideal model.
> 2) I have a model of the real left ventricle, if I transform the surface point
> coordinates of the real model into a cylindrical coordinate system and then
> use the cylindrical coordinate system to move the ideal mesh
> I would like to enquire about the feasibility of this

As I already said, there is a function in deal.II that *does* the
transformation, but you have to tell it *what the transformation is*. It is
not enough to just provide two geometries, you actually have to provide the
transformation from one to the other.
Reply all
Reply to author
Forward
0 new messages