[Hale]

hi,

I want to create a rectangular cyllinder inside another rectangular cylinder and define a wall boundary on the faces of the inner cylinder. Is there any easy way to do it in blockMesh? If not please give me some examples to do it in other ways because I looked at som similar cases in snappyHexMesh but I couldn't understand how it works.

I really appreciate your help.
Hal

[startingWithCFD]

What do you mean by "rectangular cylinder"? A cylinder with a hexahedral mesh?
It would help to present a picture with the desired output to describe it more clearly.

[Hale]

thanks for the reply...

I actually wanted to make a hollow circular cylinder as shown in the figure below but I realized that it is not an easy task in openfoam using blockMesh tool. Then I tried to make it as a hollow rectangular prism which was also not easy at all. I managed to get the rectangular prism but it wasn't the most optimal way to do it. So I want to know if there are some other smart and easier ways to generate such a rectangular prism or hollow cylinder with. Such like a circular or rectangular pipe!

NB. the inner faces should then be defined as walls...

[romant]

Which part do you want to simulate? The flow in an annulus (a), the flow inside the pipe (b), or conjugate heat transfer or similar (c)? Shown in the attached picture.

[Hale]

Originally Posted by romant

Which part do you want to simulate? The flow in an annulus (a), the flow inside the pipe (b), or conjugate heat transfer or similar (c)? Shown in the attached picture.

Hi, Thanks for your time.

Actually I only need part b shown in your pic but the system I want is a little bit more complicated. I have attached a sketch of the geometry.

I have a big circular tank where the water is coming from the sides of this tank. When the water level reaches the smaller cylinder in the middel of the big tank, it begins to fall into the cylinder in the middel and continues downstream. So I need the faces of the inner cylinder to be walls.

As far as I know it is not possible to define inner boundaries in openfoam so I have to cut the space between these two cylinders and define them as walls. And this is really confusing in blockMesh.

[romant]

Originally Posted by Hale

Hi, Thanks for your time.

Actually I only need part b shown in your pic but the system I want is a little bit more complicated. I have attached a sketch of the geometry.

I have a big circular tank where the water is coming from the sides of this tank. When the water level reaches the smaller cylinder in the middel of the big tank, it begins to fall into the cylinder in the middel and continues downstream. So I need the faces of the inner cylinder to be walls.

As far as I know it is not possible to define inner boundaries in openfoam so I have to cut the space between these two cylinders and define them as walls. And this is really confusing in blockMesh.

The best thing is to do it the following way.

Split your geometry vertically into 3 pieces as shown in the picture. In radial you split it into 3 regions. Start with a cylinder in the center (where the cylinder contains a hexagonal cylinder, in blue). The different colors show where the block are defined. In section 2, you have to leave the 2nd ring from the outside without a mesh, in section 1, only mesh the hexagon in the center and the adjoining ring with mesh and in section 3, just give all the blocks a mesh. It is alot of work with blockmesh, but you should end up with a fairly good mesh. just remember to use the same number of cells along the adjoining edges in each block. My suggestiong, for speed and accountability, use 5 cells or something along each of the edges. makes the blocking faster and more visible. Use paraFoam -block to look where you places points in blockmesh, and blockMesh after every new block, otherwise looking for errors will be a pain in the butt.

[Hale]

Thank you very much roman...I will definitely try this method

[immortality]

hi
Could you do that?how?

[Hale]

Originally Posted by immortality

hi
Could you do that?how?

Hi,

yes. The way roman suggested worked fine I don't know if I can describe it in a better way than roman did in previous post but I can try. I have attached a sketch of the way that the blocks have to be defined. I hope this helps you to understand.

I have also attached the code and the final geometry in paraview.

[immortality]

Thanks Hale
I'm glad you could do that with the help of Roman,
a side question,how you made the images with explanations?with what software in Linux?
@Roman:Hi
I want to ask same question to you!

[romant]

Originally Posted by immortality

Thanks Hale
I'm glad you could do that with the help of Roman,
a side question,how you made the images with explanations?with what software in Linux?
@Roman:Hi
I want to ask same question to you!

For the pictures that I made, I used Inkscape, which is a vector drawing program available to windows, mac, and linux systems. For visualisation of blockmesh files, you can use parafoam from your installation of openfoam with the command

paraFoam -block

This will give you a visualisation of the different points and lines you create with blockmesh.

[immortality]

do you have a guide for Inkscape?

[ngj]

Well, if you had tried using your friend Google, then the following link would be the top of the list:

http://inkscape.org/doc/

Kind regards

Niels

[Mihiran]

Originally Posted by romant

Which part do you want to simulate? The flow in an annulus (a), the flow inside the pipe (b), or conjugate heat transfer or similar (c)? Shown in the attached picture.

Dear Romant,

Could you please give the methodology to case 'c'.

best regards,
Mihiran

[romant]

Originally Posted by Mihiran

Dear Romant,

Could you please give the methodology to case 'c'.

best regards,
Mihiran

Hej,

The solution has already been given. When you take a look at the splitting in post http://www.cfd-online.com/Forums/ope...tml#post447388 you can See multiple rings, which can all be used for the purpose of conjugate heat transfer in an annulus. The outer ring would be the outer shell. The next ring traveling inwards is the fluid domain and the next ring can be the inner cylinder. The rectoid in the center is not meshed and the inner wall are also curved.

I hope that helps.