CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (http://www.cfd-online.com/Forums/ansys-meshing/)
-   -   [ICEM] Help needed in ICEM meshing (http://www.cfd-online.com/Forums/ansys-meshing/113768-help-needed-icem-meshing.html)

nauman55 February 26, 2013 07:18

Help needed in ICEM meshing
 
1 Attachment(s)
hi need help in meshing of an aircraft wing attached as picture.
regards

Far February 26, 2013 12:13

If you follow the wing body tutorial and some techniques to mesh the region inside the tip (a thread started by cfd-seeker and some other threads), you should be able to get a decent mesh.

If you want us to try, then share the files.

nauman55 February 26, 2013 14:00

Re
 
Far sir i have tried that wing body tutorial and changed my blocking strategy but the pre mesh i got had negtive volumes and uncovered faces. sir may i know in which format i shud upload the geomerty? ansys or the parasolid file?
regards

mraxx February 26, 2013 18:19

1 Attachment(s)
I have a similar problem. I spent hours on trying to mesh this geometry and almost threw my computer out of the window. The attached picture shows a medley of some of the worst outcomes of my attempts :)

Far, since you offered your help to naumann55, maybe you can have a look at my ICEM project and see if you can fix it. I do not really care about perfect mesh quality. Your help is greatly appreciated!
Here is the ICEM project file: http://www.file-upload.net/download-...irbus.zip.html


Also, I have some questions... maybe answers to this questions would bring me closer to generating the mesh by myself.

1. The ragged trailing edge shown in the picture below can be fixed by lowering the edge criterion, right? Because this did not seem to work for me but maybe I did not go low enough.

2. Many times, ICEM did not only mesh the FLUID domain but simply meshed through the surfaces. What am I doing wrong? What I did is I created a FLUD body and a SOLID body in the aircraft. When computing the mesh, I turned of the SOLID and selected Input Geometry: Visible.

3. Why are there almost never prisms on the surfaces although i specify prisms in the part mesh setup, in the prism meshign setup and check "create prism layers" in the computer volume mesh section?

4. I created a surface through the body of the aircraft as symmetry plane. ICEM created a surface mesh on this plane. But this surface mesh is also on the part of the surface that lies inside of the aircraft body. I would expect that to be a problem for the solver. Do I need to cut the inside of the aircraft out of the symmetry surface?

Thanks in advance for your answers!

Far February 26, 2013 23:56

wow classic. I will try on model as I have never seen such a Beauty in my life. Awesome


All day, every day :D

nauman55 February 27, 2013 05:46

sir sent you the geometry file on your email. request acknowledge reciept.
regards

Far February 28, 2013 12:44

There were many problems in geometry including outer surfaces, wing and fuselage. Corrected them but the fuselage surface cannot be rebuilt in ICEM and it must be corrected in your CAD software or design modeller.

http://imageshack.us/a/img526/1920/a001th.jpg
http://imageshack.us/a/img199/9428/a002.jpg
http://img560.imageshack.us/img560/9301/a003l.jpg
http://imageshack.us/a/img5/447/a004pl.jpg
http://imageshack.us/a/img546/5061/a005s.jpg
http://imageshack.us/a/img826/9423/a006r.jpg
http://imageshack.us/a/img191/411/a007a.jpg

Far February 28, 2013 12:58

Blocking
 
http://imageshack.us/a/img7/5029/a008tk.jpg

nauman55 February 28, 2013 13:48

thanx alot sir far
 
thanx alot sir i will do that. i request if you can email me the blocking and mesh you have created. get back to you after repairing the highlighted area.
regards

Far February 28, 2013 14:01

same blocking can be reused. Similar approach is used as used in Simon's three part tutorial on airfoil meshing.

Can you think of any other topology? like wing body tutorial? what do you think which one will be better?

Far March 1, 2013 14:30

Look at the pictures. I have used edit edge > Automatic linear and you can notice there is unexpected deflection of one edge in lower portion of fuselage.


http://imageshack.us/a/img191/9426/72891624.jpg

http://imageshack.us/a/img825/7170/b001r.jpg

http://imageshack.us/a/img41/4563/b002dz.jpg

Far March 1, 2013 15:00

Problem solved
 
Problem solved by dividing fuselage into two parts spanwise. Sending you files... And same blocking was re-used...

Quality > 0.61 and min angle > 18

Procedure to divide fuselage into two surfaces:

1. create iso-curve at 0.5 location on both sides (pressure and suction)
2. Segment surface by created curves.
3. update association and also make some manual association at newly created air-foil (edge to curve and vertex to point)

http://imageshack.us/a/img46/3314/c003gi.jpg

http://imageshack.us/a/img27/1086/c002vh.jpg

http://imageshack.us/a/img7/1779/12721270.jpg

Enjoy at RCMS with KP... ;)

Far March 1, 2013 15:11

@Max:

I took a look at your problem and I found it interesting. I remember similar problem at AIAA drag prediction workshop and meshing was done in ICEM hexa and tetra.

I can try to reproduce both hexa and tetra meshing if you are interested... :rolleyes:

Far March 1, 2013 15:40

No problem in Fluent. Please turn of "solid" when you are generating unstructured mesh.

http://img849.imageshack.us/img849/3545/93416776.png

nauman55 March 2, 2013 00:32

import to fluent
 
sir it is still not readable by fluent. the error message says of some missing faces. may be i have messed up with the boundary conidtions.
regards

Far March 2, 2013 01:10

As you can see in above pic, mesh has no problems in fluent.

cfd seeker March 2, 2013 06:07

Quote:

Enjoy at RCMS with KP... ;)
Hilarious :D

mraxx March 3, 2013 07:26

Quote:

Originally Posted by Far (Post 410931)
@Max:

I took a look at your problem and I found it interesting. I remember similar problem at AIAA drag prediction workshop and meshing was done in ICEM hexa and tetra.

I can try to reproduce both hexa and tetra meshing if you are interested... :rolleyes:

Yes, I am more than interested! :)

raj.singh September 27, 2013 06:54

1 Attachment(s)
is O Grid possible for such blocking???

cesarcg September 27, 2013 17:56

Dear mraxx,

I took a look at your geometry and saw that some surfaces at the tail of the fuselage were not cut, and thus these are outside of the fluid domain. The command "split surface by plane might be applied". I also saw that you model have many details at the inlet of the engine that can be omitted in a first approach.

Moreover, in my experience I think that the hardest part of the meshing will be at the regions where the wing, and stabilizers (vertical and horizontal) join the fuselage; as well as where the engine joins the wing. It is probably better to use tetra/prism in such regions as suggested by @Far.

Quote:

@Max:

I took a look at your problem and I found it interesting. I remember similar problem at AIAA drag prediction workshop and meshing was done in ICEM hexa and tetra.

I can try to reproduce both hexa and tetra meshing if you are interested...
I'll try to help during my free time.

Regards,
CÚsar


All times are GMT -4. The time now is 23:49.