problems with VERY simple mesh

October 17, 2006, 02:21

Hi!

I have a very simple mesh in gambit (2D, about 5000 tri cells, only 10 cells with an equiSize Skew higer than 0.22).

Now, exporting to and importing in Fluent is no problem. Performing a grid check is ok as well. Now, I turned the solver to axissymmetric and activate k-epsilon modell. After that, a new grid chek failed:

WARNING: non-positive volumes exist.

Waht happend???

ANY idea??

Ralf

October 17, 2006, 02:24

check if your revolution axis is absolute coincident with your x-axis

October 17, 2006, 02:35

and it should be only in positive direction on both axis.

October 17, 2006, 02:37

Hi, thx for answer...

that might be the problem, but: First, I had the axis BC parallel to the y-axis, now it is coincident to the x-axis. But the problem still appears when i turn on the axis-symmetric solver.

Ralf

October 17, 2006, 02:45

it is, but that doen't help...

October 17, 2006, 03:33

check if the revolution axis lie exactly on the x-axis. a small gap about 1e-3 under the x-axis will crash the grid check

October 17, 2006, 06:54

Hi!

in the domain extents it is visible: the y domain starts from -1.2 e-14 - so there are some nodes below the x-axis!!

I solved the problem - editing the msh file: exchanging all y values below 0 to 0. Another way: translating the whole mesh in y-direction.

It works fine now... but why does fluent (or gambit) produces that problem???

Ralf

October 17, 2006, 08:21

hey have you checked right handed ness or left handed ness problem

October 17, 2006, 02:21	problems with VERY simple mesh	#1
Ralf Schmidt Guest Posts: n/a	Hi! I have a very simple mesh in gambit (2D, about 5000 tri cells, only 10 cells with an equiSize Skew higer than 0.22). Now, exporting to and importing in Fluent is no problem. Performing a grid check is ok as well. Now, I turned the solver to axissymmetric and activate k-epsilon modell. After that, a new grid chek failed: WARNING: non-positive volumes exist. Waht happend??? ANY idea?? Ralf

October 17, 2006, 02:24	Re: problems with VERY simple mesh	#2
mAx Guest Posts: n/a	check if your revolution axis is absolute coincident with your x-axis

October 17, 2006, 02:35	Re: problems with VERY simple mesh	#3
Claud Guest Posts: n/a	and it should be only in positive direction on both axis.

October 17, 2006, 02:37	Re: problems with VERY simple mesh	#4
Ralf Schmidt Guest Posts: n/a	Hi, thx for answer... that might be the problem, but: First, I had the axis BC parallel to the y-axis, now it is coincident to the x-axis. But the problem still appears when i turn on the axis-symmetric solver. Ralf

October 17, 2006, 02:45	Re: problems with VERY simple mesh	#5
Ralf Schmidt Guest Posts: n/a	it is, but that doen't help...

October 17, 2006, 03:33	Re: problems with VERY simple mesh	#6
mAx Guest Posts: n/a	check if the revolution axis lie exactly on the x-axis. a small gap about 1e-3 under the x-axis will crash the grid check

October 17, 2006, 06:54	problem solved!	#7
Ralf Schmidt Guest Posts: n/a	Hi! in the domain extents it is visible: the y domain starts from -1.2 e-14 - so there are some nodes below the x-axis!! I solved the problem - editing the msh file: exchanging all y values below 0 to 0. Another way: translating the whole mesh in y-direction. It works fine now... but why does fluent (or gambit) produces that problem??? Ralf

October 17, 2006, 08:21	Re: problems with VERY simple mesh	#8
sach Guest Posts: n/a	hey have you checked right handed ness or left handed ness problem

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[ICEM] surface mesh merging problem	everest	ANSYS Meshing & Geometry	44	April 14, 2016 07:41
[ICEM] Using a hybrid mesh for a simple pipe	Udio_NT	ANSYS Meshing & Geometry	17	October 18, 2012 15:42
[GAMBIT] Error on simple mesh	Rich1979	ANSYS Meshing & Geometry	3	August 16, 2010 06:21
[snappyHexMesh] external flow with snappyHexMesh	chelvistero	OpenFOAM Meshing & Mesh Conversion	11	January 15, 2010 20:43
Icemcfd 11: Loss of mesh from surface mesh option?	Joe	CFX	2	March 26, 2007 19:10