[Linums]

Excuse me for not explaining well. I'll describe what I've done so you can see if it is right I did.
Pipe and fluid:

1. I made a circle on a new plane based on xy (Geometry).
2. I made a extruccion the circle.
3. I apply for faces by thin1 remove and apply a thickness to the pipe.
4. I ​​apply the material type bodyoperation Slice with 1 body and preserving bodies.
5. I apply a slice by slice the plane to divide the tube into two parts and to attach conditions to differ from each half of the tube boundary type.
6.I finally Select any internal diameters and I made a surface with 4 sides and 0 thickness.
7. I finished applying extraction (fill) by caps in all bodies with the option to preserve bodies capping "no" and if preserving solids.

It can see in Figure "pipe and fluid"

Pipe fins and fluid:
1. I delete the previous case and fill the surface because I've noticed that is not the same if you apply first finless fill and then you put them to do the pipe with fins and fill will apply.
2. To create the flap to rotate and I made a plan with a sketch and geometry.
3. I apply a extruccion (Extrude4)) to address asymmetric as you can see in the picture "pipe fins and fluid a"
4. I ​​built a pattern (parameterized) fin.
5. I suprficie and apply fill. The result can be seen in Figure "pipe and fluid"
6. Form new parts (fins (flippers) and casing (pipe)).
meshing.
Use Meshing with the options that you can see in the picture "meshing".
pipe fins and fluid.jpg

Meshing.jpg

pipe and fluid.jpg

pipe fins and fluid a.jpg

[Linums]

Fluent.

1. Use the 4-processors my computer in parallel as shown in fluent image.
2. I check the meshing and reporting of quality you can see in the picture "fluent"
3. Use k-o model and active energy.
4. I ​​check cell zone conditions that have the correct type fluid and fins as well as the pipe see picture "fluent 3"
5. In coditions boundary changes in heated and insulated conditions (heat flux parameterized) area and activate only in these areas that also parameterize shell conduction. You can see the result in "fluent 4"
6. the entrance area taken as the inlet and outlet presure as velocity inlet so the minus sign in the velocity also parameterized.
7. In reference values ​​take into account the heated area of the tube and the inner diameter of the tube inlet temperature and velocity input (output)
8. I did not change and I chose the more coupled method with all options without moverls
9. Initialize the program for hybrid
10. I run the program and see Figure converged "solution".

So far everything seems fine. but when reviewing the results averaged area especially heavy to get out 324.7454 k see figure "Area-weighted average".It turns out that this value varies only in tenths or hundredths or anything ... 1 fin or without fin or fins 15. and that's bad because a partner in its use Gambit meshing and geometry and such to the finless gives 317k and clear when it increases to 15 the result is greater fins.

The difencia is I put them to my geometries thick and think lines only.

Where do I can be wrong?

In meshing? any condition in fluent?

I really appreciate your attention and you can help me or tell me where my mistake ....I'm desperate !!!.

[Linums]

These are the latest images to which I referred.

I appreciate your help or comments may be where the error or should I change what I have.
Greetings.

[hwet]

i have not read the whole post I will do it later sometime when i am free, but just looking at the mesh, tet mesh on boundary will never give you a good result, not even an average result, you must have 10-15 layers of inflation at the boundary.

[Linums]

Originally Posted by hwet

i have not read the whole post I will do it later sometime when i am free, but just looking at the mesh, tet mesh on boundary will never give you a good result, not even an average result, you must have 10-15 layers of inflation at the boundary.

ok....




thanks for the comment, I'll work on the layers of inflation at the boundary.

[Gweher]

Well you need to use inflation layers near your fin areas, and avoid tet element in those inflation layers. In your case you can slice the fluid part and generate hexa elements. In the lower half of the pipe, just create a normal O-grid and for the upper part you can use a C-grid approach around the fin blocks and 2 quarter O-grids in the upper part. You can do this procedure with ICEM, and use a replay script to generate the mesh when you edit your parameters input.

[Linums]

Hi!
geometry upload my file so I can help with the mesh, I've only done basic exercises in ICEM and DM. Thank a lot to Gweher for their comments and responses, I'm starting with your suggestion but not much.

GREETINGS Gweher

[Linums]

[QUOTE=Linums;517168]Hi!
geometry upload my file so I can help with the mesh, I've only done basic exercises in ICEM and DM. Thank a lot to Gweher for their comments and responses, I'm starting with your suggestion but not much.

GREETINGS Gweher
[/QUOT

These are the images in GAMBIT meshing