2D simulation turned out to be 3d!!

zeitistgeld · April 29, 2009, 07:04

Hello, every one. I was simulating the flow field around an airfoil using 2 dimensional geometry and with one element in the third dimension. However when I check the y plus or wall shear distribution of the foil surface, it was not 2d at all! See at http://img207.imageshack.us/img207/9940/yplus.png
how could that happen. i set the front and back sides as symmetry boundaries. Could it be the element aspect ratio that caused the problem? The first grid node was placed at 0.01 mm from the surface and the 3rd dimensional thickness was 1 cm. Or can I force the symmetry boundaries as translational periodic ones? And how to? I am using CFX 10.0.

ghorrocks · April 29, 2009, 19:35

Hi,

The thickness of the 3rd dimension should be approximately equal to the smallest element edge length. This effect can also be caused by your mesh not being perfectly 2D (ie the planes are not flat or angled or the extrusion is not straight).

Glenn

zeitistgeld · April 30, 2009, 00:48

Hi.Glenn. Thanks for your reply. I extrude the mesh in ICEM with one cell thickness in Z direction. If this thickness is of the magnitute order of the boundary cell, it is very cumbersome to set boundaries condtions, and I have to zoom in constantly. Before extrusion, it might be better to set tolerance first, am I right?

ghorrocks · April 30, 2009, 01:39

Hi,

Yes, making the extrusion thickness equal the smallest element edge length can make working with the geometry difficult. But sometimes it can eliminate problems like what you see so I would give it a try.

While you are at it, put in a feature request to your support person for a true 2D model in CFX. It must be the most requested feature for years in CFX but still you have to do this single element thickness thing. The more requests there are for it the more likely it is for it to happen.

Glenn Horrocks

LSC · May 10, 2009, 03:11

Hi Zei, it is nice to find somebody doing the same thing. I am trying to do airfoil simulations in CFX too. Do you create your mesh in CFX pre-pro or GAMBIT? Is there any good links that provide guides to newbie like me in using CFX?

zeitistgeld · May 10, 2009, 03:56

Hi, LSC.
I an using ANSYS ICEM and GRIDGEN to generate the mesh, and they are both good meshers. For begginners, the best way is to read carefully the tutorials and go through all by yourself to get a quick understanding of the whole procedures of a simulation.

LSC · May 10, 2009, 04:44

Hi Zie, thanks for replying. I am wondering when you did your meshing, does your trailing edge as a finite thickness or the Upper and lower surface joins at a common vertex? Because I was facing problem with 0 thickness at trailing edge when I do meshing in GAMBIT..

zeitistgeld · May 10, 2009, 05:05

http://courses.cit.cornell.edu/fluent/airfoil/index.htm
if you are using gambit, read the link above will be much help I think. it's a c-type mesh. So if your airfoil is zero thickness at the trailing edge, just use this type of mesh.

LSC · May 10, 2009, 06:10

because I understand that fluent solves 2D which CFX aren't capable of..I read that for 2D simulations I can have a trailing edge of zero thickness but for 3D the zero thickness will cause problems during mesh which I encountered too..do you have any problems like this where the trailing edge is of a finite thickness? pls advice..

zeitistgeld · May 10, 2009, 06:33

If you can generate 2D mesh first, then in the third dimension you just extrude one element thickness then you will get the 3d mesh which CFX can read.

LSC · May 10, 2009, 07:06

ok..thanks for your advice..I'll try that out..one element thickness which mean? usually I sweep with 1 unit which is I think 1 metre in the z direction

zeitistgeld · May 10, 2009, 07:11

if the first layer grid is of milimeter magnitude or even much smaller, you sweep 1 meter in the Z direction, that will lead to high aspect ratio. You'd better sweep one element as thick as the smallest element in the boundary layer.

LSC · May 10, 2009, 09:13

ok..have you managed to get the lift coefficients and drag coefficients?

cfang · July 15, 2009, 08:19

..............................

ghorrocks · July 15, 2009, 18:30

Well the tutorial examples provided with CFX are a good place to start. Wasn't that obvious?

cfang · July 15, 2009, 22:36

.................

ckleanth · July 16, 2009, 02:16

sometimes I wonder ....

Lance · July 16, 2009, 02:32

Quote:

Originally Posted by cfang

Tutorial of cfx? Where can i get it? Anyway... wat is the difference between cfx n cfd?

LOL, best post ever!

Did you notice the sticky post on the main cfx thread?

cfang · July 16, 2009, 12:31

thanx......

ckleanth · July 16, 2009, 13:08

which university are you attending? taking an aerodynamic degree and if you have no previous exposure to cfd or other numerical methods I'd be worried if I were you.

To be honest the way you ask your questions "I need to do a final year project to get my degree.. The title is 3d airfoil simulation.. However, i dont have any idea what to do and how to start.. Anybody can help me?"

seriously.. what response did you expect to get?

You'd also have a supervisor that comes with the said project title. Did you ask him/her what you should be doing? I assume to ask questions here you would have access to ansys academic license ...

consult your supervisor about your project, and get someone to teach you the fundamentals. one this is done and if CFX is still the tool you will use you should start working through the tutorials then on your problem

surely as you said you learn how to use matlab and c++ so you know the drill... be prepared for long working hours and...

good luck

April 29, 2009, 07:04	2D simulation turned out to be 3d!!	#1
zeitistgeld Member zeitistgeld Join Date: Apr 2009 Posts: 37 Rep Power: 16	Hello, every one. I was simulating the flow field around an airfoil using 2 dimensional geometry and with one element in the third dimension. However when I check the y plus or wall shear distribution of the foil surface, it was not 2d at all! See at http://img207.imageshack.us/img207/9940/yplus.png how could that happen. i set the front and back sides as symmetry boundaries. Could it be the element aspect ratio that caused the problem? The first grid node was placed at 0.01 mm from the surface and the 3rd dimensional thickness was 1 cm. Or can I force the symmetry boundaries as translational periodic ones? And how to? I am using CFX 10.0.

April 30, 2009, 01:39		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	Hi, Yes, making the extrusion thickness equal the smallest element edge length can make working with the geometry difficult. But sometimes it can eliminate problems like what you see so I would give it a try. While you are at it, put in a feature request to your support person for a true 2D model in CFX. It must be the most requested feature for years in CFX but still you have to do this single element thickness thing. The more requests there are for it the more likely it is for it to happen. Glenn Horrocks happy likes this.

July 15, 2009, 08:19	3D airfoil simulation..	#14
cfang New Member Chih Fang Join Date: Jul 2009 Posts: 18 Rep Power: 16	.............................. Last edited by cfang; July 20, 2009 at 22:28.

July 15, 2009, 22:36		#16
cfang New Member Chih Fang Join Date: Jul 2009 Posts: 18 Rep Power: 16	................. Last edited by cfang; July 20, 2009 at 22:28.

July 16, 2009, 02:16		#17
ckleanth Senior Member George Join Date: Mar 2009 Location: Birmingham, UK Posts: 257 Rep Power: 18	sometimes I wonder .... __________________ Top 4 tips 1. Knowledge is everything and Ignorance is dangerous. 2. Understand your limitations and try to eliminate them. 3. Get yerself a bike and hoon the chuffer. You will soon learn why dogs like to hang their heads out the car window. 4. Please before asking any questions on how to run simulations in CFX, go though all the tutorials

April 29, 2009, 19:35		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	Hi, The thickness of the 3rd dimension should be approximately equal to the smallest element edge length. This effect can also be caused by your mesh not being perfectly 2D (ie the planes are not flat or angled or the extrusion is not straight). Glenn

April 30, 2009, 00:48		#3
zeitistgeld Member zeitistgeld Join Date: Apr 2009 Posts: 37 Rep Power: 16	Hi.Glenn. Thanks for your reply. I extrude the mesh in ICEM with one cell thickness in Z direction. If this thickness is of the magnitute order of the boundary cell, it is very cumbersome to set boundaries condtions, and I have to zoom in constantly. Before extrusion, it might be better to set tolerance first, am I right?

May 10, 2009, 03:11		#5
LSC Member LSC Join Date: May 2009 Posts: 58 Rep Power: 16	Hi Zei, it is nice to find somebody doing the same thing. I am trying to do airfoil simulations in CFX too. Do you create your mesh in CFX pre-pro or GAMBIT? Is there any good links that provide guides to newbie like me in using CFX?

May 10, 2009, 03:56		#6
zeitistgeld Member zeitistgeld Join Date: Apr 2009 Posts: 37 Rep Power: 16	Hi, LSC. I an using ANSYS ICEM and GRIDGEN to generate the mesh, and they are both good meshers. For begginners, the best way is to read carefully the tutorials and go through all by yourself to get a quick understanding of the whole procedures of a simulation.

May 10, 2009, 04:44		#7
LSC Member LSC Join Date: May 2009 Posts: 58 Rep Power: 16	Hi Zie, thanks for replying. I am wondering when you did your meshing, does your trailing edge as a finite thickness or the Upper and lower surface joins at a common vertex? Because I was facing problem with 0 thickness at trailing edge when I do meshing in GAMBIT..

May 10, 2009, 05:05		#8
zeitistgeld Member zeitistgeld Join Date: Apr 2009 Posts: 37 Rep Power: 16	http://courses.cit.cornell.edu/fluent/airfoil/index.htm if you are using gambit, read the link above will be much help I think. it's a c-type mesh. So if your airfoil is zero thickness at the trailing edge, just use this type of mesh.

May 10, 2009, 06:10		#9
LSC Member LSC Join Date: May 2009 Posts: 58 Rep Power: 16	because I understand that fluent solves 2D which CFX aren't capable of..I read that for 2D simulations I can have a trailing edge of zero thickness but for 3D the zero thickness will cause problems during mesh which I encountered too..do you have any problems like this where the trailing edge is of a finite thickness? pls advice..

May 10, 2009, 06:33		#10
zeitistgeld Member zeitistgeld Join Date: Apr 2009 Posts: 37 Rep Power: 16	If you can generate 2D mesh first, then in the third dimension you just extrude one element thickness then you will get the 3d mesh which CFX can read.

May 10, 2009, 07:06		#11
LSC Member LSC Join Date: May 2009 Posts: 58 Rep Power: 16	ok..thanks for your advice..I'll try that out..one element thickness which mean? usually I sweep with 1 unit which is I think 1 metre in the z direction

May 10, 2009, 07:11		#12
zeitistgeld Member zeitistgeld Join Date: Apr 2009 Posts: 37 Rep Power: 16	if the first layer grid is of milimeter magnitude or even much smaller, you sweep 1 meter in the Z direction, that will lead to high aspect ratio. You'd better sweep one element as thick as the smallest element in the boundary layer.

May 10, 2009, 09:13		#13
LSC Member LSC Join Date: May 2009 Posts: 58 Rep Power: 16	ok..have you managed to get the lift coefficients and drag coefficients?

July 15, 2009, 18:30		#15
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	Well the tutorial examples provided with CFX are a good place to start. Wasn't that obvious?

July 16, 2009, 12:31		#19
cfang New Member Chih Fang Join Date: Jul 2009 Posts: 18 Rep Power: 16	thanx...... Last edited by cfang; July 20, 2009 at 22:28.

July 16, 2009, 13:08		#20
ckleanth Senior Member George Join Date: Mar 2009 Location: Birmingham, UK Posts: 257 Rep Power: 18	which university are you attending? taking an aerodynamic degree and if you have no previous exposure to cfd or other numerical methods I'd be worried if I were you. To be honest the way you ask your questions "I need to do a final year project to get my degree.. The title is 3d airfoil simulation.. However, i dont have any idea what to do and how to start.. Anybody can help me?" seriously.. what response did you expect to get? You'd also have a supervisor that comes with the said project title. Did you ask him/her what you should be doing? I assume to ask questions here you would have access to ansys academic license ... consult your supervisor about your project, and get someone to teach you the fundamentals. one this is done and if CFX is still the tool you will use you should start working through the tutorials then on your problem surely as you said you learn how to use matlab and c++ so you know the drill... be prepared for long working hours and... good luck __________________ Top 4 tips 1. Knowledge is everything and Ignorance is dangerous. 2. Understand your limitations and try to eliminate them. 3. Get yerself a bike and hoon the chuffer. You will soon learn why dogs like to hang their heads out the car window. 4. Please before asking any questions on how to run simulations in CFX, go though all the tutorials

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
FSI TWO-WAY SIMULATION	Smagmon	CFX	1	March 6, 2009 13:24
Continuous vs interrupted simulation	sega	OpenFOAM Running, Solving & CFD	4	November 3, 2008 14:29
strange simulation error	Ralf Schmidt	FLUENT	2	May 4, 2007 13:02
Fire simulation using FDS from NIST	Jens	Main CFD Forum	1	January 22, 2004 01:53
3-D Contaminant Dispersal Simulation	Apple L S Chan	Main CFD Forum	1	December 23, 1998 10:06