CFD Online Discussion Forums - [ICEM] what is the strategy to block this geometry?

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)

- ANSYS Meshing & Geometry (https://www.cfd-online.com/Forums/ansys-meshing/)

- - [ICEM] what is the strategy to block this geometry? (https://www.cfd-online.com/Forums/ansys-meshing/107660-what-strategy-block-geometry.html)

llrr	October 3, 2012 09:10

what is the strategy to block this geometry?

1 Attachment(s)

Hi everyone, I am learning ICEM CFD (v14.0) now. Can you help to see how I can get high quliaty mesh with this geometry? Thank you.

cfd seeker

October 3, 2012 09:47

attach .tin file so that we can have a try on your geometry

diamondx

October 3, 2012 12:27

initiliaze your blocling first. do two vertical split each one at the end of the deflectors.
same for horizontal split. because of deflector, you will need y-grid there. good luck.
Share your tin, i might take a look at it late at night...

llrr	October 4, 2012 02:31

Quote:

Originally Posted by diamondx (Post 384751)

Thank you guys, here is the tin file.

llrr	October 4, 2012 02:33

I will also try it according diamondx's suggestion, thank you.

llrr	October 4, 2012 05:17

2 Attachment(s)

Hi Diamondx and cfd seeker, I enclosed the picture to show the determinant and angle histogram (I was going to upload .blk file, but it is not allowed). I do not think the mesh quality is good especially when I look at the Angle histogram. Can you help me to see how I can improve it? Thank you.

stuart23

October 4, 2012 09:05

Hi llrr, try creating a zip archive of the files, then upload

llrr	October 5, 2012 04:05

1 Attachment(s)

Quote:

Originally Posted by stuart23 (Post 384917)

Hi llrr, try creating a zip archive of the files, then upload

Hi Stuart, thank you for the suggestions. Here it is.

diamondx

October 5, 2012 13:36

sorry for not replying, i will take a look at it when i'll have time, i'm little bit busy... sorry again i hope you're not in a rush

llrr	October 7, 2012 03:34

Quote:

Originally Posted by diamondx (Post 385130)

sorry for not replying, i will take a look at it when i'll have time, i'm little bit busy... sorry again i hope you're not in a rush

Hi Diamondx, I appreciate your help. Take your time and I will also keep learning:-) Thank you.

Far	October 7, 2012 03:39

tin file is missing

llrr	October 7, 2012 03:43

1 Attachment(s)

Quote:

Originally Posted by Far (Post 385260)

tin file is missing

Hi Far, Thank you for the information. I guess it is because I didn't found that I could not attach .tin file last time. Here is the .zip file of .tin. Thank you.

Far	October 7, 2012 07:52

1 Attachment(s)

Here is the solution. Is this what you asked for? ;)

llrr	October 7, 2012 08:40

Quote:

Originally Posted by Far (Post 385323)

Here is the solution. Is this what you asked for? ;)

Hi Far, can you explian to me how you made the block? Thank you.

Far	October 7, 2012 10:35

1. one horizontal split for the upper part. second and third split for the two inclined plates.

2. three vertical splits for the upper part.

3. Change the material of two blocks on the right and left side of upper part to dead.

4. Vertical splits for the inclined plates

5. associates edges and vertices

6. associate faces to part for four inclined plates.

Far	October 7, 2012 13:18

This was really a very simple and basic blocking. No Y-block, or O-grid.

http://imageshack.us/a/img339/2828/86685757.png

http://imageshack.us/a/img26/9853/31826919.png

llrr	October 7, 2012 13:37

Quote:

Originally Posted by Far (Post 385343)

Thank you very much for your help. I will study it carefully.

llrr	October 8, 2012 03:19

1 Attachment(s)

Quote:

Originally Posted by Far (Post 385356)

This was really a very simple and basic blocking. No Y-block, or O-grid.

http://imageshack.us/a/img339/2828/86685757.png

http://imageshack.us/a/img26/9853/31826919.png

Hi Far, I did what you told me and I made it. But when I open the window on the roof and I have a problem with the mesh quality again. Would you please help me to see it again what the problem is of my block? Enclosed please find the .tin and .blk file. Thank you.

llrr	October 8, 2012 03:24

I tried to make a y-block at the wedge of the roof, but it looks not right. What did I do wrong?

energy382

October 8, 2012 06:17

you insert some new geometry, called roofwinwall. what's that?!

nevertheless, it's not a closed surface, so if you want to insert y-block, you've to insert additional surfaces.

Far	October 8, 2012 06:21

Oh, it is open. In old file there was no extended surface on roof. If there is extended surface then energy382 is suggesting the right approach.

llrr	October 8, 2012 07:07

Quote:

Originally Posted by energy382 (Post 385488)

you insert some new geometry, called roofwinwall. what's that?!

nevertheless, it's not a closed surface, so if you want to insert y-block, you've to insert additional surfaces.

Hi energy382, the problem is solved. I need to add six surfaces to enclosed it. Thank you.

energy382

October 8, 2012 07:50

1 Attachment(s)

Quote:

Originally Posted by llrr (Post 385510)

Hi energy382, the problem is solved. I need to add six surfaces to enclosed it. Thank you.

No, your blocking is destroyed. Need to do a new one.

You've to improve mesh quality (edge parameters, smoothing etc). Don't have time to do this too.

llrr	October 9, 2012 02:34

Quote:

Originally Posted by energy382 (Post 385520)

No, your blocking is destroyed. Need to do a new one.

You've to improve mesh quality (edge parameters, smoothing etc). Don't have time to do this too.

Hi energy382, thank you for the help. I made the new one yesterday and generate relatively good mesh quality. May I ask why the block is destroyed since I have the same steps to operate the block? I also would like to ask if there should not be any single surface (not closed surface) in the geometry? Thanks a lot.

energy382

October 9, 2012 06:06

Quote:

Originally Posted by llrr (Post 385645)

Yes, you should always use a closed volume. If there's a non-closed volume in your real geometry, this is always either an inlet, opening or outlet (and therefore, you have to create a new surface)

Even if you have periodic boundaries (e.g. turbine blade passage), you should at least create a part with a single point within, where you project your faces to. Otherwise, ICEM always project edges to closest surface.

nkme2007

October 10, 2012 05:54

Hello All,

I want to do analysis of heat transfer from water flowing through pipes submerged inside concrete. I am modelling in GAMBIT and wish to analyse it on Ansys FLUENT.

Can anybody help me out, how to model and simulate?

Does any tutorials exist?

llrr	October 10, 2012 10:20

Quote:

Originally Posted by energy382 (Post 385672)

Hi energy382, thank you for the explaination. After I check the mesh, I got some error information of multiple edges and non-manifold vertices. May I ask what cuases these errors and how should I fix it? Thank you.

Far	October 10, 2012 10:28

did you output mesh to fluent?

energy382

October 11, 2012 08:02

Quote:

Originally Posted by llrr (Post 385965)

I guess, you've forgotten to merge vertices after creating y-grid!?

llrr	October 12, 2012 03:13

Quote:

Originally Posted by energy382 (Post 386145)

I guess, you've forgotten to merge vertices after creating y-grid!?

Should I merge the vertex again after creating the y-block? How to do that? Thank you.

llrr	October 12, 2012 03:14

Quote:

Originally Posted by Far (Post 385968)

did you output mesh to fluent?

Yes, I will output the mesh to Fluent. but what is the relation between the error information and Fluent solver? Thank you.

llrr	October 12, 2012 03:54

Quote:

Originally Posted by energy382 (Post 386145)

I guess, you've forgotten to merge vertices after creating y-grid!?

Hi energy382, what I did is:
(1) split the block
(2) merge the vertex
(3) create y-block
(4) generate the mesh

After I create the y-block, should I merge the vertex again? What vertex should I merge? I tried to merge vertex again, but it gives me the information that the vertex is unique. If there is something that I did wrong, please let me know. Thank you.

energy382

October 15, 2012 09:32

Quote:

Originally Posted by llrr (Post 386237)

you've to do two y-blocks right? (right and left side)
you've to merge the vertices, where the two edges (vertical) of the y-blocks match.

compare my blocking with your blocking at this area.

llrr	October 18, 2012 08:37

Quote:

Originally Posted by energy382 (Post 386681)

you've to do two y-blocks right? (right and left side)
you've to merge the vertices, where the two edges (vertical) of the y-blocks match.

compare my blocking with your blocking at this area.

Hi energy382, I do the two y-blocks at the same time. I tried to merge the vertex where these two y-blocks meet, but it still gives me the information that the vertex is unique? Is it possible that the vertex already merged when I generate the two y-blocks at the same time? Thank you. BTW, I cannot see your block.

llrr	October 18, 2012 08:48

Quote:

Originally Posted by energy382 (Post 386681)

you've to do two y-blocks right? (right and left side)
you've to merge the vertices, where the two edges (vertical) of the y-blocks match.

compare my blocking with your blocking at this area.

Hi energy382, the problem has been solved. The mesh is fine. I did not merge the vertexs where the two y-block meets (I generated them at the same time, I could not perform the 'merge vertex' operation at all). Thank you for the help.

All times are GMT -4. The time now is 16:10.