[geod]

Hi all,

I am currently working on a project where I am required to simulate the heat exchangement from one fluid to another in concentric pipes. A hot fluid will pass through the inner tube, with a cold fluid passing through the outer tube.

I am really struggling to create the mesh for the two separate bodies. I have imported the model from Solidworks, created the block and associated the points with the outer pipeline. I am now wondering how do I then create a separate block for the inner pipeline? I have also tried an o-grid block for the inner pipeline but cannot seem to make any progress.

Could anyone suggest a tutorial or reading for something that can relate to this? Or even give me a suggestion for the next step I should take?

Any advice will be really appreaciated!

Thanks,

G

[geod]

Couple photos of geometry for reference.. just a simple tube within a tube.

geometry1.jpg

geometry2.jpg

[rolloblues]

Hi geod,

first of all for your kind of calculation you probably will need three domains: outer fluid, solid layer, inner fluid

To generate the mesh start with a block for the outer tube:
- associate the edges/project vertices
- create an O-Grid with the diagonal edges long enough to keep the square part inside the inner tube.
- split the O-Grid blocks along the diagonal according to the solid layer thickness (one split near the outer diameter and one near the inner diameter)
- associate the edges/project vertices
- create three parts (i.e. outer, inner, solid) and for each one select the related blocks.
- generate the mesh

[geod]

Hi Davide,

First of all, thanks for your interest and help. I tried out the steps that you suggested (as close as I could, I think I did ok), here is a couple photos of the pre-mesh along with quality and determinant 2x2x2 results, please let me know if this looks like a decent result:

premesh1.jpg

premesh2.jpg

One problem I encountered is it seems the inner mesh on one side of the pipe actually covers the outside diameter aswell, thought it might have been an association error but I repeated it and the same thing happened.

Regards,

George

[geod]

newpremes.jpg

So I've managed to a get a pre mesh looking like this.. But can't figure out why the Outerflow is condensing to one side, same with the inner pipe. Any ideas how to clean it up?

[rolloblues]

That's better than the first one but if I'm not mistaken now I see two O-Grids in your premesh.
I can tell it from the little blue layer between the outer circle and the yellow blocks. Besides two edges (up and right) of the yellow part have a similar shape of the inner purple square.

I think you should start once again paying attention to generate just one O-Grid after creating and associating the first block (for the outer pipe)

Anyway tomorrow I'll try to reproduce a similar case and post a few pics to give you the idea of what you should obtain in the end.

[rolloblues]

Here you have some figures that, I hope, will give you the idea of what you should come up with.

Cheers!

[geod]

Thanks for the photos, I have attempted a similar case to yours however when I do this my inner pipe ends up taking the inner flow, the outer flow takes it's own section and the inner pipe section, and the inner flow parts do not create a mesh:

hisattempt1.jpghisattempt2.jpg

I can't seem to find a way where the blocks end up in the same position as yours.

Here is an attempt I tried where the meshes were all in the positions I wanted them to be, however I think like you said I have multiple O-grids. Will this affect my simulation? Here is a photo:

nicegeom1.jpg

I really appreciate your interest! Thanks for the help.

[rolloblues]

Let's start with your last picture



This blocking could work if correctly associated but in my opinion it is unnecessarily complicated. My advice is to always keep things as simple as possible.

I think you should go back to your previous strategy:



This is good but incomplete, basically you are missing one split.
If you compare your blocking with mine you'll se that I have 13 blocks (5 violet, 4 purple, 4 green) where you only have 9 (by the way you are missing the purple blocks).
The 4 narrow blocks that you associated with the solid part actually should belong to the inner tube together with the square block. That's the reason why your mesh collapse onto the inner tube.

To mesh the solid part do another split along the diagonal of the Ogrid so that you obtain other four blocks (the purple blocks in my figure).

One last thing: why are keeping in the geometry also the thickness of the outer tube?
Since it is outside the fluid domain maybe you don't need it for your calculations, but if so, to mesh it you will need another split of the Ogrid just as for the purple blocks

[geod]

Excuse my amateurness... In regards to the complicated version picture I posted, I don't know where the unnecessary blocks came from. All I did was create a block, create the O-grid split the blocks and associate for each surface. But the mesh seemed to come out alright.

When I try your way I just can't seem to get it to work properly and I don't know why. Here are the steps I take:

Create block
Associate to inside of outer pipe + project vertices
Create o-grid
Move vertices to the inside of the inner pipe (They didn't automatically do that)
Split block for each remaining surface and associate, project vertices to the curves.
set pre-mesh params and then compute mesh.

However after completing these steps all the mesh seems to be a bit all over the place. Do the steps I took sound about right?

Thanks again for all your feedback.

[rolloblues]

Originally Posted by geod

Create block
Associate to inside of outer pipe + project vertices
Create o-grid
[COLOR="Red"][COLOR="Black"]Move vertices to the inside of the inner pipe (They didn't automatically do that)
Split block for each remaining surface and associate, project vertices to the curves.
set pre-mesh params and then compute mesh.

The procedure should work, but moving unassociated vertices can be tricky: if you don't put the right set of constraints you may end with vertices all over the place.

To ensure that the internal vertices of the Ogrid fall inside the inner pipe in the first place, you can play with the "offset" parameter in the Ogrid block tab. Basically with that parameter you are controlling the lenght of the diagonal edges of the Ogrid.

Once you generated the Ogrid, to modify that lenght you can use the "Modify Ogrid" under the "Edit Block" tab