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/)
-   -   gambit mesh "T" tube (http://www.cfd-online.com/Forums/ansys-meshing/119783-gambit-mesh-t-tube.html)

steven1 June 25, 2013 04:02

gambit mesh "T" tube
 
hello everyone,

i am a beginner of gambit. now i am trying to creat a mesh on a "T" form tube. do you have some ideas pr suggestions about how to seperate the different parts?

thank you

-mAx- June 25, 2013 04:23

you are not enforce to decompose your domain, execpt if you want a full hexa mesh

steven1 June 25, 2013 04:52

i have used fluent to calculate, it showed that there were instability at the entrance of the small tube. so i have no choice to seperate the interval part to have a more precise part. but i always get a great number of mesh.

-mAx- June 25, 2013 04:58

post pictures, and explain your boundary conditions

steven1 June 25, 2013 05:14

Hello , this is the tube that i use,.
The boundry conditions are : turbulence with a velocity of 5m/s inlet.
Flow rate weighting 0.5 ; 0.5
http://www.cfd-online.com/Forums/dat...AASUVORK5CYII=

steven1 June 25, 2013 05:18

i don't knoz how to add pictures, to i make a link here. thank you
https://www.dropbox.com/s/s1ufa02vv9oka4k/0625.docx

-mAx- June 25, 2013 05:19

I don't see pictures
What kind of BC did you use for outlet?

-mAx- June 25, 2013 05:19

Quote:

Originally Posted by steven1 (Post 435761)
i don't knoz how to add pictures, to i make a link here. thank you
https://www.dropbox.com/s/s1ufa02vv9oka4k/0625.docx

you have to go in advanced mode

-mAx- June 25, 2013 05:21

please post picture with focus on your mesh, else we don't see anything.
Pleas post also picture of velocity and pressure distribution

steven1 June 25, 2013 05:48

this is the document; https://www.dropbox.com/s/5oqvv81gvb...0question.docx

sorry for making you waiting so long, i am not used to the computer with the french keyboard that i use.

-mAx- June 25, 2013 06:17

ok can you display clother at the T junction, especially velocity distribution. Something sounds weird left to the T junction.

Regarding to the mesh, you seem to have splited your domain, right?
Can you also list all the surfaces with their BC, I see lots of surfaces

And finally when I see the the velocity distribution, I assume your inlet is the pipe with smaller diameter. But if it is exact, then the pressure distribution is wrong

steven1 June 25, 2013 07:05

yes, you are right, i have splited the domaine.
the inlet is the left side of the larger pipe, the other side of it and the small pipe are the outlets. The diameter of the larger pipe is 20mm, that of the small one is 6 mm, and the lengh of all the three parts around the intersection is 500mm
https://www.dropbox.com/s/c7o9q4nrku...question1.docx


-mAx- June 25, 2013 07:13

can you provide me your dbs file from Gambit? (delete the mesh for reducing file size)

steven1 June 25, 2013 07:24

https://www.dropbox.com/s/nsfn9id1k1...lageRe10-5.dbs
here it is, there are some changes, i didn't save that version.

-mAx- June 25, 2013 07:48

ok, seems to be ok, but you don't have any surfaces in the BC.
I assume you fixed the problem.
Other point: what are your dimensions m,mm,... ?

steven1 June 25, 2013 08:13

the dimensions are mm. and in fact i am a beginer of fluent and gambit, i don't know which faces to add with BC

-mAx- June 25, 2013 08:21

ok, did you scaled your domain in fluent?
Gambit doesn t have any dimension.

steven1 June 25, 2013 08:36

ok, i find that i am totally fool, i thought the dimension was set by default as mm. i just searched the user's guide, it's set as Meter. however, i didn't scale the dimension

-mAx- June 25, 2013 08:42

you have to scale your domain:
in your case, select mm, and then click on scale
If you have a diameter of 20, Fluent will understand 20m
If you scale it, Fluent will convert and undestand 20*1e-3 = 0.02m

Once it is done, do a check mesh, to be sure that your mesh is ok.
Then compute.

steven1 June 25, 2013 08:45

OK, thank you very much. i think i have to rebuild the mesh as the skewness exceeded 0.98


All times are GMT -4. The time now is 01:45.