[ICEM] Meshing a Rocket Injector Assembly

AshisGhosh · January 6, 2014, 17:30

Hi There,

I'm creating a mesh for an injector assembly to be used in a hybrid rocket. I will eventually export the mesh to use with OpenFoam.

With the mesh featured in the pictures I achieved a non-orthogonality of 85, which is much to high to obtain usable results.

To create the mesh I set up the line mesh for all the edges to achieve that orthogonal appearance that you can see on the outside. However, I do not know how I can replicate that for the inside. I've tried the various types of meshing methods available between octree and hexa dominant.

As this is my first serious project with ICEM-CFD I am still highly inexperienced. Any advice or tips would be welcome.

cutmesh.PNG

fulloutside.PNG

closeoutside.PNG

model.jpg

Thanks.

Far · January 7, 2014, 01:09

you are using ICEM HEXA?

AshisGhosh · January 18, 2014, 15:29

Upon trying that, I get the following:

Cut view:
full.jpg

Close up:
close.PNG

The max non-orthogonality increased from 85-89, but the average went down to 17 from 29.

However, this time it failed to checkMesh checks from OpenFoam for skewness.

Any suggestions?

Thanks!

Far · January 19, 2014, 04:06

i would recommend you to use hexa method

FJSJ · January 19, 2014, 06:13

Can you share your geometry?

AshisGhosh · January 19, 2014, 12:54

I used the hexa-dominant method - is that different from the HEXA method? I am a bit confused with all of this.

Here's my geometry as a .STEP file.

InjFlow.zip

Thanks!

Far · January 19, 2014, 13:03

yes both are different.

Hexa dominant is for FEA and ICEM CFD HEXA is for CFD and uses blocking.

raghav · January 21, 2014, 06:06

Hi Ashis,

I have uploaded the files in the dropbox

https://www.dropbox.com/s/spsxctmlz9zui38/InFlowShared.zip

You can refine further if u want.

Regards,
Raghav

AshisGhosh · February 5, 2014, 01:52

Thank you very much - this will be a great help!

Where would I be able to learn how to do it to this degree? I tried many times to achieve the same thing. Could you explain your blocking strategy?

Again, much thanks!

raghav · March 7, 2014, 01:40

Hi Ashis,

It will come by practice. I used
1. Combo of Bottom-Up and Top-Down approach.
2. Used O-Grids in Tandem

January 6, 2014, 17:30	Meshing a Rocket Injector Assembly	#1
AshisGhosh New Member Ashis Ghosh Join Date: Jan 2014 Posts: 4 Rep Power: 12	Hi There, I'm creating a mesh for an injector assembly to be used in a hybrid rocket. I will eventually export the mesh to use with OpenFoam. With the mesh featured in the pictures I achieved a non-orthogonality of 85, which is much to high to obtain usable results. To create the mesh I set up the line mesh for all the edges to achieve that orthogonal appearance that you can see on the outside. However, I do not know how I can replicate that for the inside. I've tried the various types of meshing methods available between octree and hexa dominant. As this is my first serious project with ICEM-CFD I am still highly inexperienced. Any advice or tips would be welcome. cutmesh.PNG fulloutside.PNG closeoutside.PNG model.jpg Thanks.

February 5, 2014, 01:52	Thank you!	#9
AshisGhosh New Member Ashis Ghosh Join Date: Jan 2014 Posts: 4 Rep Power: 12	Thank you very much - this will be a great help! Where would I be able to learn how to do it to this degree? I tried many times to achieve the same thing. Could you explain your blocking strategy? Again, much thanks!

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Need help with a rocket nozzle simulation	badboyz31	CFX	14	September 22, 2014 01:06
Doubt regarding 3-D meshing of a rocket nose	nakul	OpenFOAM Pre-Processing	0	April 3, 2010 16:54
Singularity of grid?Volume meshing vs face meshing	Ken	Main CFD Forum	0	September 4, 2003 12:09
Volume Meshing & Face Meshing? singularity of grid	ken	FLUENT	0	September 4, 2003 12:08

January 7, 2014, 01:09		#2
Far Senior Member Sijal Join Date: Mar 2009 Location: Islamabad Posts: 4,558 Blog Entries: 6 Rep Power: 54	you are using ICEM HEXA?

January 18, 2014, 15:29		#3
AshisGhosh New Member Ashis Ghosh Join Date: Jan 2014 Posts: 4 Rep Power: 12	Upon trying that, I get the following: Cut view: full.jpg Close up: close.PNG The max non-orthogonality increased from 85-89, but the average went down to 17 from 29. However, this time it failed to checkMesh checks from OpenFoam for skewness. Any suggestions? Thanks!

January 19, 2014, 04:06		#4
Far Senior Member Sijal Join Date: Mar 2009 Location: Islamabad Posts: 4,558 Blog Entries: 6 Rep Power: 54	i would recommend you to use hexa method

January 19, 2014, 06:13		#5
FJSJ Senior Member Javi Join Date: Jan 2013 Posts: 276 Rep Power: 16	Can you share your geometry?

January 19, 2014, 12:54		#6
AshisGhosh New Member Ashis Ghosh Join Date: Jan 2014 Posts: 4 Rep Power: 12	I used the hexa-dominant method - is that different from the HEXA method? I am a bit confused with all of this. Here's my geometry as a .STEP file. InjFlow.zip Thanks!

January 19, 2014, 13:03		#7
Far Senior Member Sijal Join Date: Mar 2009 Location: Islamabad Posts: 4,558 Blog Entries: 6 Rep Power: 54	yes both are different. Hexa dominant is for FEA and ICEM CFD HEXA is for CFD and uses blocking.

March 7, 2014, 01:40		#10
raghav Member Raghav Join Date: Jul 2010 Location: India, Karnataka Posts: 47 Rep Power: 18	Hi Ashis, It will come by practice. I used 1. Combo of Bottom-Up and Top-Down approach. 2. Used O-Grids in Tandem