Suitable scheme for node connectivity

Simbelmynë · March 25, 2015, 09:42

Hey,

If I wish to connect a set of randomly placed nodes using a five point stencil. It is important to retain the property that if nodeA-->nodeB then nodeB-->nodeA. What methods are suitable for this?

I can use a triangular mesh and connect triangle edges into this five point stencil (two connected triangles have 5 edges), but I was wondering if you have any other suggestions.

Cheers!

mprinkey · March 25, 2015, 11:01

Getting from random space points to a properly connected grid/mesh is a difficult problem and may be impossible under the constraint of a five-point stencil (I'm assuming 2D).

There are 2D paving algorithms that will give you quadrilateral meshes (four-sided cells all with four neighbors) but it is not a trivial process:

http://www.researchgate.net/publicat...ne_constraints

http://www.imr.sandia.gov/papers/imr20/Mukherjee.pdf

But in both of those papers, you start NOT with random points but with boundary points and then the algorithm fills in the domain with points/cells.

Triangular mesh algorithm is easier and gmsh can do this for you automatically.

March 25, 2015, 09:42	Suitable scheme for node connectivity	#1
Simbelmynë Senior Member Join Date: May 2012 Posts: 546 Rep Power: 15	Hey, If I wish to connect a set of randomly placed nodes using a five point stencil. It is important to retain the property that if nodeA-->nodeB then nodeB-->nodeA. What methods are suitable for this? I can use a triangular mesh and connect triangle edges into this five point stencil (two connected triangles have 5 edges), but I was wondering if you have any other suggestions. Cheers!

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
error message	cuteapathy	CFX	14	March 20, 2012 06:45
Is it suitable to use Weller's scheme in this case?	kaifu	OpenFOAM	5	June 22, 2011 08:09
Poisson Solver or suitable scheme	Tingting	Main CFD Forum	1	August 4, 2006 06:47
Which scheme is suitable for NS under vacuum?	Zhihang Zhang	Main CFD Forum	1	January 21, 2004 19:47
Regarding a suitable scheme for BFC	Abhijeet Vaidya	Main CFD Forum	4	December 18, 2001 12:15

March 25, 2015, 11:01		#2
mprinkey Senior Member Michael Prinkey Join Date: Mar 2009 Location: Pittsburgh PA Posts: 363 Rep Power: 25	Getting from random space points to a properly connected grid/mesh is a difficult problem and may be impossible under the constraint of a five-point stencil (I'm assuming 2D). There are 2D paving algorithms that will give you quadrilateral meshes (four-sided cells all with four neighbors) but it is not a trivial process: http://www.researchgate.net/publicat...ne_constraints http://www.imr.sandia.gov/papers/imr20/Mukherjee.pdf But in both of those papers, you start NOT with random points but with boundary points and then the algorithm fills in the domain with points/cells. Triangular mesh algorithm is easier and gmsh can do this for you automatically.