CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (https://www.cfd-online.com/Forums/ansys-meshing/)
-   -   [GAMBIT] meshing stirred tank (https://www.cfd-online.com/Forums/ansys-meshing/113024-meshing-stirred-tank.html)

jamalf64 February 10, 2013 15:48

meshing stirred tank
 
4 Attachment(s)
Hi Dear Friends
I strongly need your help
I have a geometry that is stirred tank. I want to mesh periodic geometry. but I dont know how should I start? Please give your suggestions.
I attached 4 pictures of my problem: 2 pictures are periodic domain that I want to mesh that.

Thank you so much

-mAx- February 11, 2013 03:40

http://aerojet.engr.ucdavis.edu/gamb...e/mg04.htm#777

jamalf64 February 11, 2013 06:14

Quote:

Originally Posted by -mAx- (Post 407055)

Dear Max
Thank you so much for reply
I can specify zone types but I cant mesh geometry. please help me about meshing strategy.

-mAx- February 11, 2013 06:20

and why can't you mesh your geometry?
Can you point out the warning or errors you got?

jamalf64 February 11, 2013 06:32

2 Attachment(s)
Quote:

Originally Posted by -mAx- (Post 407085)
and why can't you mesh your geometry?
Can you point out the warning or errors you got?

I can mesh that but mesh quality is low. I do auto mesh(gambit default mesh) by tet/hybrid/Tgrid in inner zone and hex/wedge/cooper in outer zone.

-mAx- February 11, 2013 07:46

max skewness about 0.75 is good, limit is 0.9-0.95

jamalf64 February 11, 2013 13:08

2 Attachment(s)
Quote:

Originally Posted by -mAx- (Post 407106)
max skewness about 0.75 is good, limit is 0.9-0.95

Thanks in advance
But answers is wrong and I know this is due to the bad quality of mesh.
I've already got the correct answers by Another type of impeller by the same parameters in fluent. Its solution in each time step(10^-3) after every 30 iterations would converge but in this case the solution is converge after 80 iterations in each time step(time step is 10^-6).
I attached 2 pictures of previous problem.

Best Regards

-mAx- February 12, 2013 02:27

and what gives you a check Mesh?
But for sure your skewness is not guilty

jamalf64 February 12, 2013 06:08

Quote:

Originally Posted by -mAx- (Post 407293)
and what gives you a check Mesh?
But for sure your skewness is not guilty

Check mesh no problem and its OK

-mAx- February 12, 2013 06:11

Then you don't have problem with mesh.

jamalf64 February 12, 2013 06:17

Quote:

Originally Posted by -mAx- (Post 407342)
Then you don't have problem with mesh.

But when mesh quality is low, solution takes too much time. I know that to solve this problem, mesh quality must be high.

Best Regards

-mAx- February 12, 2013 06:28

skewness = 0 means perfekt
skewness = 1 means degenerate
:rolleyes:
....
http://aerojet.engr.ucdavis.edu/gambithelp/html/users_guide/ug0304.htm#gc_exmesh

jamalf64 February 12, 2013 07:18

Quote:

Originally Posted by -mAx- (Post 407348)
skewness = 0 means perfekt
skewness = 1 means degenerate
:rolleyes:
....
http://aerojet.engr.ucdavis.edu/gambithelp/html/users_guide/ug0304.htm#gc_exmesh

Thank you so much for replies
But I'm not convinced

-mAx- February 12, 2013 07:24

Quote:

Originally Posted by jamalf64 (Post 407369)
But I'm not convinced

I will convince you:
generate an awfull mesh with skewness greather than 0.97
Then import in fluent, and do a check mesh

jamalf64 February 12, 2013 07:30

Quote:

Originally Posted by -mAx- (Post 407370)
I will convince you:
generate an awfull mesh with skewness greather than 0.97
Then import in fluent, and do a check mesh

I know in this case, the software displays an error message(failed mesh). But when mesh quality is low(not bad) solution takes a lot of time

-mAx- February 12, 2013 07:40

no
the value of your skewness is low, which means your quality is good!
Again:
skewness ~0 means excellent!
skewness ~1 means very poor!

While you compare your both models,what is smallest cell size in both meshes.
And how many cells they have. (cf your post #7)

jamalf64 February 12, 2013 08:00

Quote:

Originally Posted by -mAx- (Post 407375)
no
the value of your skewness is low, which means your quality is good!
Again:
skewness ~0 means excellent!
skewness ~1 means very poor!

While you compare your both models,what is smallest cell size in both meshes.
And how many cells they have. (cf your post #7)

My PC is hanging and I dont know why GAMBIT does not run. I will contact you later.

THANKS IN ADVANCE


All times are GMT -4. The time now is 11:09.