CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (http://www.cfd-online.com/Forums/ansys-meshing/)
-   -   create geometry out of centerline and radius (http://www.cfd-online.com/Forums/ansys-meshing/93632-create-geometry-out-centerline-radius.html)

Lilly October 21, 2011 05:13

create geometry out of centerline and radius
 
Dear all,

I want to create a geometry of a vessel with some bifurcations out of the points of the centerline of the vessel and the radius of the vessel at these points. Does anybody have an idea how I could do this? The software I can use is Fluent, Gambit, Tgrid and ICEM. I tried to do it using Gambit, but since my centerline has a lot of points, it would be really complicate and prone to errors to insert all values manually.
Thanks a million for your help and ideas!
With best regards,
Lilly

-mAx- October 21, 2011 06:01

*do you have a sketch of what you want?
I*f you also have a cad package you can build your geometry withi cad, then export in your mesher

Lilly October 26, 2011 06:28

3 Attachment(s)
Dear ĖmAx-,

thank you for your help! I donít have access to a commercial CAD software at the moment, but I try to get access to AutoCAD through our university. Do you think it would work with this software? I also added some pics of what I want to do. The plan is to create one long bended tube or some small tubes, which are merged afterwards, with the radius of the vessel radius and the centerline of the vessel centerline. I have several vertices defining this centerline and the value of the radius of the vessel at these points (pic vessel1). A friend and me tried to get such a vessel-geometry composed of small tubes using Wolframs Mathematica (pic Vessel2), but especially if the vessel bends there is a gap in between the single tubes and I donít know how to create a closed, smooth, single surface out of this. I would also be happy just to use the centerline of the vessel and to insert the radius at the beginning and the end of the single branches and have a linear variation of the radius in between. Do you have any idea, how I could do this?

Thanks a million for your help!
Lilly

-mAx- October 26, 2011 07:24

with Gambit, can you import all your vertex (from text file).
http://my.fit.edu/itresources/manual..._import_vertex
Then you just need to build surfaces: between 2 given center-vertex you create the circle with your given radius and then same with next centr-vertex. Now you will be able to build surface (and also volume)
You only need patience.
You are working on this geometry for a while, isn't it?

Lilly October 27, 2011 10:49

1 Attachment(s)
Hi -mAx-,

thank you!
I tried a similar method and I am afraid (if I understand you correct) I will have the same problem with your method: if the vessel is bent there appear gaps between the tubes (see pic) and I don't know how to close them :-(.

Yes, I have been working at this geometry for a while, but I realized now, that even if i segment the single pixels of the dicom data the surface of my model will be really uneven and by smoothing the vessel gets too thin (especially in some regions). Therefore I try to test this method of getting an "ideal" vessel, now.

Thanks a million for your help!
Lilly

Lilly October 27, 2011 12:02

Hi -mAx-,

I just realized my error in reasoning on my way to the train! I was using the "Revolve Edges" option when I was getting the gaps between the tubes :o. I will try your method tomorrow!
Thank you!
Lilly

-mAx- October 28, 2011 02:16

revolve vertice, then sweep surface (you need to create local axis to sweep along local axisw)
Don't hesitate to sweep more than needed, and then split the sweeped surface with edge or surface which intersects. It will close your surface (gap issue)
Be carefull not creating any virtual entity else, you will have less freedom withsplit and or boolean operations


All times are GMT -4. The time now is 12:49.