hybrid 3D surface mesh generation

classic Classic list List threaded Threaded
6 messages Options
Kim
Reply | Threaded
Open this post in threaded view
|

hybrid 3D surface mesh generation

Kim
Dear CGAL developers,

I would like to use the 3D surface mesh generation capabilities of CGAL in a mode similar to the hybrid mode of the 3D mesh generator (i.e. use both an implicit domain and a polyhedral domain) but could not find anything about this being posible. Is there any way of achiving with the 3D surface mesher? The reason is I would like to provide a geometry given by a polyhedral domain with an infill given by an implicit domain.

Kind regards,
Kim



--
You are currently subscribed to cgal-discuss.
To unsubscribe or access the archives, go to
https://sympa.inria.fr/sympa/info/cgal-discuss

Reply | Threaded
Open this post in threaded view
|

Re: hybrid 3D surface mesh generation

Sebastien Loriot (GeometryFactory)
You can use the 3D mesher for that purpose. You can simply remove any
criteria on the tetrahedra and use
CGAL::facets_in_complex_3_to_triangle_mesh() to extract the surface facets:

https://doc.cgal.org/latest/Mesh_3/group__PkgMesh3Functions.html#ga68ca38989087644fb6b893eb566be9ea

Best,

Sebastien.

On 9/6/20 10:12 PM, "F.A. Perez" ([hidden email] via
cgal-discuss Mailing List) wrote:

> Dear CGAL developers,
>
> I would like to use the 3D surface mesh generation capabilities of CGAL
> in a mode similar to the hybrid mode of the 3D mesh generator (i.e. use
> both an implicit domain and a polyhedral domain) but could not find
> anything about this being posible. Is there any way of achiving with the
> 3D surface mesher? The reason is I would like to provide a geometry
> given by a polyhedral domain with an infill given by an implicit domain.
>
> Kind regards,
> Kim
>
>
>
> --
> You are currently subscribed to cgal-discuss.
> To unsubscribe or access the archives, go to
> https://sympa.inria.fr/sympa/info/cgal-discuss
>

--
You are currently subscribed to cgal-discuss.
To unsubscribe or access the archives, go to
https://sympa.inria.fr/sympa/info/cgal-discuss


Kim
Reply | Threaded
Open this post in threaded view
|

Re: hybrid 3D surface mesh generation

Kim
Dear Sebastien,

Thanks for the reply! I did consider this option but would this not still mean higer meshing times and memory use since I am still essentialy creating a volume mesh, albeit neglecting the quality of the inside mesh. Given that I expect meshing times to be in the order of days sometimes and with high memory use (especially if creating a volume mesh in the process) my hope was I could use the faster and less resource intensive 3d Surface mesher somehow. I am aware for speed I can go for parallel but the memory issue would remain. Or is the difference in memory use rather limited?

Regards,
Kim


From: [hidden email] <[hidden email]> on behalf of "Sebastien Loriot (GeometryFactory)" <[hidden email]>
Sent: Monday, September 7, 2020 6:53 AM
To: [hidden email] <[hidden email]>
Subject: Re: [cgal-discuss] hybrid 3D surface mesh generation
 
You can use the 3D mesher for that purpose. You can simply remove any
criteria on the tetrahedra and use
CGAL::facets_in_complex_3_to_triangle_mesh() to extract the surface facets:

https://doc.cgal.org/latest/Mesh_3/group__PkgMesh3Functions.html#ga68ca38989087644fb6b893eb566be9ea

Best,

Sebastien.

On 9/6/20 10:12 PM, "F.A. Perez" ([hidden email] via
cgal-discuss Mailing List) wrote:
> Dear CGAL developers,
>
> I would like to use the 3D surface mesh generation capabilities of CGAL
> in a mode similar to the hybrid mode of the 3D mesh generator (i.e. use
> both an implicit domain and a polyhedral domain) but could not find
> anything about this being posible. Is there any way of achiving with the
> 3D surface mesher? The reason is I would like to provide a geometry
> given by a polyhedral domain with an infill given by an implicit domain.
>
> Kind regards,
> Kim
>
>
>
> --
> You are currently subscribed to cgal-discuss.
> To unsubscribe or access the archives, go to
> https://sympa.inria.fr/sympa/info/cgal-discuss
>

--
You are currently subscribed to cgal-discuss.
To unsubscribe or access the archives, go to
https://sympa.inria.fr/sympa/info/cgal-discuss



--
You are currently subscribed to cgal-discuss.
To unsubscribe or access the archives, go to
https://sympa.inria.fr/sympa/info/cgal-discuss

Reply | Threaded
Open this post in threaded view
|

Re: hybrid 3D surface mesh generation

andreas.fabri

Hi Kim,


The surface mesher is essentially a volume mesher without quality constraints on tetrahedra.
So it won't be faster. 


Best regards,

Andreas


On 9/8/2020 8:13 AM, "F.A. Perez" ([hidden email] via cgal-discuss Mailing List) wrote:
Dear Sebastien,

Thanks for the reply! I did consider this option but would this not still mean higer meshing times and memory use since I am still essentialy creating a volume mesh, albeit neglecting the quality of the inside mesh. Given that I expect meshing times to be in the order of days sometimes and with high memory use (especially if creating a volume mesh in the process) my hope was I could use the faster and less resource intensive 3d Surface mesher somehow. I am aware for speed I can go for parallel but the memory issue would remain. Or is the difference in memory use rather limited?

Regards,
Kim


From: [hidden email] [hidden email] on behalf of "Sebastien Loriot (GeometryFactory)" [hidden email]
Sent: Monday, September 7, 2020 6:53 AM
To: [hidden email] [hidden email]
Subject: Re: [cgal-discuss] hybrid 3D surface mesh generation
 
You can use the 3D mesher for that purpose. You can simply remove any
criteria on the tetrahedra and use
CGAL::facets_in_complex_3_to_triangle_mesh() to extract the surface facets:

https://doc.cgal.org/latest/Mesh_3/group__PkgMesh3Functions.html#ga68ca38989087644fb6b893eb566be9ea

Best,

Sebastien.

On 9/6/20 10:12 PM, "F.A. Perez" ([hidden email] via
cgal-discuss Mailing List) wrote:
> Dear CGAL developers,
>
> I would like to use the 3D surface mesh generation capabilities of CGAL
> in a mode similar to the hybrid mode of the 3D mesh generator (i.e. use
> both an implicit domain and a polyhedral domain) but could not find
> anything about this being posible. Is there any way of achiving with the
> 3D surface mesher? The reason is I would like to provide a geometry
> given by a polyhedral domain with an infill given by an implicit domain.
>
> Kind regards,
> Kim
>
>
>
> --
> You are currently subscribed to cgal-discuss.
> To unsubscribe or access the archives, go to
> https://sympa.inria.fr/sympa/info/cgal-discuss
>

--
You are currently subscribed to cgal-discuss.
To unsubscribe or access the archives, go to
https://sympa.inria.fr/sympa/info/cgal-discuss



--
You are currently subscribed to cgal-discuss.
To unsubscribe or access the archives, go to
https://sympa.inria.fr/sympa/info/cgal-discuss

-- 
Andreas Fabri, PhD
Chief Officer, GeometryFactory
Editor, The CGAL Project

phone: +33.492.954.912    skype: andreas.fabri

--
You are currently subscribed to cgal-discuss.
To unsubscribe or access the archives, go to
https://sympa.inria.fr/sympa/info/cgal-discuss

Reply | Threaded
Open this post in threaded view
|

Re: hybrid 3D surface mesh generation

andreas.fabri
In reply to this post by Kim

Hi again,


Just out of curiosity, what are you meshing that you expect "meshing times in the order of days"?


Best,

Andreas


On 9/8/2020 8:13 AM, "F.A. Perez" ([hidden email] via cgal-discuss Mailing List) wrote:
Dear Sebastien,

Thanks for the reply! I did consider this option but would this not still mean higer meshing times and memory use since I am still essentialy creating a volume mesh, albeit neglecting the quality of the inside mesh. Given that I expect meshing times to be in the order of days sometimes and with high memory use (especially if creating a volume mesh in the process) my hope was I could use the faster and less resource intensive 3d Surface mesher somehow. I am aware for speed I can go for parallel but the memory issue would remain. Or is the difference in memory use rather limited?

Regards,
Kim


From: [hidden email] [hidden email] on behalf of "Sebastien Loriot (GeometryFactory)" [hidden email]
Sent: Monday, September 7, 2020 6:53 AM
To: [hidden email] [hidden email]
Subject: Re: [cgal-discuss] hybrid 3D surface mesh generation
 
You can use the 3D mesher for that purpose. You can simply remove any
criteria on the tetrahedra and use
CGAL::facets_in_complex_3_to_triangle_mesh() to extract the surface facets:

https://doc.cgal.org/latest/Mesh_3/group__PkgMesh3Functions.html#ga68ca38989087644fb6b893eb566be9ea

Best,

Sebastien.

On 9/6/20 10:12 PM, "F.A. Perez" ([hidden email] via
cgal-discuss Mailing List) wrote:
> Dear CGAL developers,
>
> I would like to use the 3D surface mesh generation capabilities of CGAL
> in a mode similar to the hybrid mode of the 3D mesh generator (i.e. use
> both an implicit domain and a polyhedral domain) but could not find
> anything about this being posible. Is there any way of achiving with the
> 3D surface mesher? The reason is I would like to provide a geometry
> given by a polyhedral domain with an infill given by an implicit domain.
>
> Kind regards,
> Kim
>
>
>
> --
> You are currently subscribed to cgal-discuss.
> To unsubscribe or access the archives, go to
> https://sympa.inria.fr/sympa/info/cgal-discuss
>

--
You are currently subscribed to cgal-discuss.
To unsubscribe or access the archives, go to
https://sympa.inria.fr/sympa/info/cgal-discuss



--
You are currently subscribed to cgal-discuss.
To unsubscribe or access the archives, go to
https://sympa.inria.fr/sympa/info/cgal-discuss

-- 
Andreas Fabri, PhD
Chief Officer, GeometryFactory
Editor, The CGAL Project

phone: +33.492.954.912    skype: andreas.fabri

--
You are currently subscribed to cgal-discuss.
To unsubscribe or access the archives, go to
https://sympa.inria.fr/sympa/info/cgal-discuss

Kim
Reply | Threaded
Open this post in threaded view
|

Re: hybrid 3D surface mesh generation

Kim
In reply to this post by andreas.fabri

Dear Andreas, 

 

>>The surface mesher is essentially a volume mesher without quality constraints on tetrahedra. 

>>So it won't be faster." 

In that case I will have to check if I am doing something wrong since in a fast comparison it seemed to be between 30-50% slower compared to the surface mesher. 

 

>>Just out of curiosity, what are you meshing that you expect meshing times in the order of days? 

I am using it to mesh heterogeneous TPMS (triply periodic minimal surface) lattices which very quickly becomes a very time consuming task. 

 

 

With regards to the hybrid mode of the 3D volume mesher it works perfectly if both my implicit function and the Polyhedra domain are closed. Now, I would actually like to provide my implicit function, i.e. the TPMS function, as the open infinite function it is, and have it only have an effect on the closed polyhedral domain. Is this at all possible at the moment? The reason for this is that the polyhedral domain may contain sharp edges which I could preserve in this way but not with the current approach I am using, i.e. create a poisson reconstruction to represent the polyhedra domain as an implicit function and then combine this with the TPMS to obtain a closed implicit description that can be meshed with the normal implicit mesh mode.  

Of course, I could scale my polyhedral domain to be a bit larger in the implicit description, mesh this, then clip it to get my sharp edges, and remesh preserving edges. But that is not a very elegant approach and it would make the process take much longer. 

 

Regards, 

Kim 



--
You are currently subscribed to cgal-discuss.
To unsubscribe or access the archives, go to
https://sympa.inria.fr/sympa/info/cgal-discuss