[ANSYS Meshing] Creating a good mesh on difficult geometry

Jipkrens1 · November 20, 2014, 10:52

Dear all,

For my Bachelors Thesis I'm analyzing a complex burner. I use the Ansys workbench / Fluent for this, and the last month I've been busy getting comfortable with the program. This forum has been really helpful with that.

At this point however I'm stuck with creating a good quality mesh in the workbench. The burner is a cylinder with a diameter of around 1 meter and a length of 2 meter. There is one inlet with pre-mixed combustion and one inlet for extra air. Inside the cylinder a conical liner is placed.

The liner is the problem: It's 1,20 meter long, but only 2 mm thick.

Whats the best way to create a good mesh for this geometry? I've been trying for over a week and can't seem to get it right.

Best regards,
Jip Krens

shivasluzz · November 21, 2014, 04:47

Try meshing it by yourself & tell what problems you are getting.
Some tips:
1. Simplify the geometry
2. Take only a sector or half or quarter section of the geometry.
3. Go for unstructured mesh first.

Jipkrens1 · November 21, 2014, 08:02

Dear shivasluzz,

Thanks for your reaction. I created a mesh with simpeler geometry (I left out the holes from the liner) and could create a good geometry this way. However, the most important part of my analysis is the mixture of air and fuel thru these holes in the liner.

I managed to create one mesh with the total geometry, but this mesh had 3,2 million elements, and the academic version of Fluent only supports 500,000 elements.

What do you mean by going for the unstructured mesh first?
I could post a photo of the geometry if that helps?

Thanks again,
Jip

shivasluzz · November 21, 2014, 08:17

Hope you are aware that there are two types of mesh 1. Structured mesh (Hex elements)
2. Unstructured mesh (tet elements)

If you want to reduce the number of elements then go for structured mesh. But it's time consuming.

Ok post the post

Jipkrens1 · November 21, 2014, 08:29

Dear shivasluzz,

The pictures are attached. The first picture is the burner room, the second one the liner which is placed inside the burner room.

Best regards,
Jip

shivasluzz · November 21, 2014, 09:03

First follow this simple method.
(hope you have already created the fluid model)
1. Insert "method" by right clicking on the 'mesh' in the tree view.
2. Set method to "multizone"
3. Generate mesh
4. Increase or decrease the "releavnce centre" in mesh detail to make the mesh finer or coarser.
Try this and tell me

Jipkrens1 · November 21, 2014, 09:36

I tried the multizone method, but got the following two errors:
- The mesher has an unhandled exception. Running out of memory may be the cause.
- The mesh generation was not successful.

Is there a way to solve this?

shivasluzz · November 21, 2014, 09:47

Decrease the "relevance centre" and try again.
If that also not works, then decompose your body into several bodies (1 part & many bodies under it) in design modeller. Split in such a way that each body is sweepable. For your geometry, split the main cylinder into 3 (before, after and at the location of liner), also split all the small cylinder which are created in the holes & etc.,
Now in the meshing software, right click on mesh & click "show sweepable" body. It should show many selectable bodies and then insert method, select "sweep" as your option.. Then mesh it!

Jipkrens1 · November 21, 2014, 09:50

shivasluzz, thanks again for your reply, I will try the sweep method aswell.

I get what you mean by splitting the main cylinder in 3 parts, but what do you mean by splitting the small cylinder / holes?

Thanks!

shivasluzz · November 21, 2014, 11:32

What I meant to say is.. If we make fluid model for your geometry, then we will get cylinders in the place of holes of liner rite.. That's wat I told..

solid particle erosion · January 12, 2017, 18:25

Hi all

I have an important question that will assist me with my final year project.

I am looking into solid particle erosion in pipes and would like to know how i can obtain a good quality mesh for a straight pipe (19.25 in in length and 2.25 inch in outer diameter) on ANSYS Fluent.

Thank you very much

et.one · January 30, 2018, 07:29

I am trying to make this body sweepable ( https://ibb.co/cRJC8m )

What I do is:

- slicing the two lateral cylinders --> which become sweepable
- slicing the bottom part, i.e. restriction + bottom cylinder --> which become sweepable
- slicing the hemisphere and the centrale part separately BUT they don't become sweepable in any way.

Any help would be much appreciated!

Thank you so much

November 20, 2014, 10:52	Creating a good mesh on difficult geometry	#1
Jipkrens1 New Member Jip Krens Join Date: Nov 2014 Posts: 5 Rep Power: 12	Dear all, For my Bachelors Thesis I'm analyzing a complex burner. I use the Ansys workbench / Fluent for this, and the last month I've been busy getting comfortable with the program. This forum has been really helpful with that. At this point however I'm stuck with creating a good quality mesh in the workbench. The burner is a cylinder with a diameter of around 1 meter and a length of 2 meter. There is one inlet with pre-mixed combustion and one inlet for extra air. Inside the cylinder a conical liner is placed. The liner is the problem: It's 1,20 meter long, but only 2 mm thick. Whats the best way to create a good mesh for this geometry? I've been trying for over a week and can't seem to get it right. Best regards, Jip Krens

January 12, 2017, 18:25	Meshing	#11
solid particle erosion New Member Join Date: Jan 2017 Location: Manchester Posts: 1 Rep Power: 0	Hi all I have an important question that will assist me with my final year project. I am looking into solid particle erosion in pipes and would like to know how i can obtain a good quality mesh for a straight pipe (19.25 in in length and 2.25 inch in outer diameter) on ANSYS Fluent. Thank you very much

January 30, 2018, 07:29	Making this geometry fully sweepable !	#12
et.one New Member Join Date: Jul 2017 Posts: 11 Rep Power: 9	I am trying to make this body sweepable ( https://ibb.co/cRJC8m ) What I do is: - slicing the two lateral cylinders --> which become sweepable - slicing the bottom part, i.e. restriction + bottom cylinder --> which become sweepable - slicing the hemisphere and the centrale part separately BUT they don't become sweepable in any way. Any help would be much appreciated! Thank you so much

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Inner geometry gets lost exporting mesh from ICEM CFD to CFX-Pre	powpow	CFX	3	December 20, 2012 10:14
[ICEM] Copy geometry then copy mesh	Anna Tian	ANSYS Meshing & Geometry	0	July 30, 2012 17:59
How to control Minximum mesh space?	hung	FLUENT	7	April 18, 2005 10:38
Mesh for 3 dim Geometry	Phil	FLUENT	9	July 12, 2000 05:39
unstructured vs. structured grids	Frank Muldoon	Main CFD Forum	1	January 5, 1999 11:09

November 21, 2014, 04:47		#2
shivasluzz Senior Member sluzzer Join Date: Jul 2014 Posts: 146 Rep Power: 12	Try meshing it by yourself & tell what problems you are getting. Some tips: 1. Simplify the geometry 2. Take only a sector or half or quarter section of the geometry. 3. Go for unstructured mesh first.

November 21, 2014, 08:02		#3
Jipkrens1 New Member Jip Krens Join Date: Nov 2014 Posts: 5 Rep Power: 12	Dear shivasluzz, Thanks for your reaction. I created a mesh with simpeler geometry (I left out the holes from the liner) and could create a good geometry this way. However, the most important part of my analysis is the mixture of air and fuel thru these holes in the liner. I managed to create one mesh with the total geometry, but this mesh had 3,2 million elements, and the academic version of Fluent only supports 500,000 elements. What do you mean by going for the unstructured mesh first? I could post a photo of the geometry if that helps? Thanks again, Jip

November 21, 2014, 08:17		#4
shivasluzz Senior Member sluzzer Join Date: Jul 2014 Posts: 146 Rep Power: 12	Hope you are aware that there are two types of mesh 1. Structured mesh (Hex elements) 2. Unstructured mesh (tet elements) If you want to reduce the number of elements then go for structured mesh. But it's time consuming. Ok post the post

November 21, 2014, 09:03		#6
shivasluzz Senior Member sluzzer Join Date: Jul 2014 Posts: 146 Rep Power: 12	First follow this simple method. (hope you have already created the fluid model) 1. Insert "method" by right clicking on the 'mesh' in the tree view. 2. Set method to "multizone" 3. Generate mesh 4. Increase or decrease the "releavnce centre" in mesh detail to make the mesh finer or coarser. Try this and tell me

November 21, 2014, 09:36		#7
Jipkrens1 New Member Jip Krens Join Date: Nov 2014 Posts: 5 Rep Power: 12	I tried the multizone method, but got the following two errors: - The mesher has an unhandled exception. Running out of memory may be the cause. - The mesh generation was not successful. Is there a way to solve this?

November 21, 2014, 09:47		#8
shivasluzz Senior Member sluzzer Join Date: Jul 2014 Posts: 146 Rep Power: 12	Decrease the "relevance centre" and try again. If that also not works, then decompose your body into several bodies (1 part & many bodies under it) in design modeller. Split in such a way that each body is sweepable. For your geometry, split the main cylinder into 3 (before, after and at the location of liner), also split all the small cylinder which are created in the holes & etc., Now in the meshing software, right click on mesh & click "show sweepable" body. It should show many selectable bodies and then insert method, select "sweep" as your option.. Then mesh it!

November 21, 2014, 09:50		#9
Jipkrens1 New Member Jip Krens Join Date: Nov 2014 Posts: 5 Rep Power: 12	shivasluzz, thanks again for your reply, I will try the sweep method aswell. I get what you mean by splitting the main cylinder in 3 parts, but what do you mean by splitting the small cylinder / holes? Thanks!

November 21, 2014, 11:32		#10
shivasluzz Senior Member sluzzer Join Date: Jul 2014 Posts: 146 Rep Power: 12	What I meant to say is.. If we make fluid model for your geometry, then we will get cylinders in the place of holes of liner rite.. That's wat I told..