CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Siemens (
-   -   HELP! TETRA and HEXA boundary (

sunlight007 August 4, 2004 12:00

HELP! TETRA and HEXA boundary
Hello, Geometry: Impeller on top a cylindrical mesh with velocity (0,0,constant) and the impeller is bounded to a spiral. The space inbetween is thin and is tetra too. The spiral mesh is also tetra. At spiral outlet i have tried extrusion with a thin hexa layer NEGATIVE VOLUME of these cells How to repair??? Later i tried only tetra i.e. no hexa at all in the mesh following boundaries: 1) fixed outlet 2) pressure condition

I tried stationary calculation which diverges (UD and SIMPLE and also MARS) 3) i tried only Tetra (i.e. no hexa at all) and it only works to small presuure at outlet. 4) i know the mass flow: how can i use it for outlet? What could the problem be? How to repair? Thanks

Anton Lyaskin August 4, 2004 14:20

Re: HELP! TETRA and HEXA boundary
Try "Fix left-handed" button in Check tool - probabli orientation is the cause of the negative volume

sunlight007 August 5, 2004 04:18

Re: HELP! TETRA and HEXA boundary
no, there are no left-handed cells but cracks. how can i repair? thanks

Anton Lyaskin August 5, 2004 05:36

Re: HELP! TETRA and HEXA boundary
If you have cracks than this is probably because your extruded mesh is not connected to the other part. Try VMERGE command

sunlight007 August 5, 2004 08:12

Re: HELP! TETRA and HEXA boundary
i did ccom, vcom and vmerg it does not help. In ICEM i created the tetra mesh, exported to STAR-CD, put hexa layer on onlet and outlet and around 1000 cracks are found in the around 300000 cell-mesh. I tried pure tetra mesh i.e. no hexa on inlet and outlet but there is hardly any (u,v,w) although inlet has (0,0,7.96) velocity. i then tried inlet and pressure but this also gives no the expected answer. I rotate some parts of the geometry and expect velocity greater than inlet but there it is around zero. How to better the mesh? What is the problem??? Thanks

Anton Lyaskin August 5, 2004 13:29

Re: HELP! TETRA and HEXA boundary
Did you try to plot your cells with cracks to see what's the problem? There can be two ways: 1) if there are no actuall crack visible (well, looks like not your case because you have problems with running the simulation) then you get the message because the cells are to thin, 2) if there are reall cracks there then you have to try higher values for VMERGE tolerance

Helge August 6, 2004 06:15

Re: HELP! TETRA and HEXA boundary
You may try to extrude your mesh within ICEM and do all checks there

All times are GMT -4. The time now is 09:32.