CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (https://www.cfd-online.com/Forums/ansys-meshing/)
-   -   [ICEM] How to mesh a 3 lenz turbine blade in rotating domain (https://www.cfd-online.com/Forums/ansys-meshing/98413-how-mesh-3-lenz-turbine-blade-rotating-domain.html)

 CTan22 March 10, 2012 07:25

How to mesh a 3 lenz turbine blade in rotating domain

Hi,

I need to build a prototype VAWT for my final year project, hence I need to do some simulation before fabricate. But I need to know how to mesh a 3 lenz blade in a rotating circular mesh, so I need help and guidance. Below is a top view picture of the lenz turbine blade.

http://farm8.staticflickr.com/7179/6...e49e6ebb_b.jpg

 CTan22 March 10, 2012 09:30

anyone can help????:(

 PSYMN March 11, 2012 08:33

Draw a concentric circle (centered on the center of rotation)... That should divide your domain into two regions... The outer region and the rotating region... Make sure your geometry has two separate parts. Mesh everything. You probably want inflation layers (aka boundary layers or prism layers) on all those parts that represent your wind turbine...

Then go to solver setup and make sure your circular domain is rotating...

 CTan22 March 11, 2012 11:09

Quote:
 Originally Posted by PSYMN (Post 348778) Draw a concentric circle (centered on the center of rotation)... That should divide your domain into two regions... The outer region and the rotating region... Make sure your geometry has two separate parts. Mesh everything. You probably want inflation layers (aka boundary layers or prism layers) on all those parts that represent your wind turbine... Then go to solver setup and make sure your circular domain is rotating...
Hi Simon thank for your guidance, but I got one question on how to setup a mesh topology on three blade in a rotating domain. I always see is a single blade meshing, but if applied to multiple blade I'm wonder how does the mesh topology look like.

So if possible is there any reference I can read through, so I can do it my self. But once again thank for your effort to guide me. :)

 PSYMN March 11, 2012 17:52

4 Attachment(s)
Oh, you want a blocking strategy...

First, you will want to generate the mesh for just one of the 3 blades, but make it periodic and then copy rotate the mesh later...

In order for the mesh to be periodic, you need equal numbers of nodes upstream and downstream, so you can't just do a CGrid to capture the C shape of your blocking... I recomend an HGrid for the region between the hub and the rotating boundary (or a quarter Ogrid if you didn't have a hub), and then an Ogrid for your blade... Split the Ogrid in half and associate part of the middle of the Ogrid with you geometry where appropriate... This will give you nice boundary layer mesh on both sides of your mesh...

This is what I show in these pics... I also smoothed the mesh, although that isn't really necessary...

I suppose it could have been about 2 radial split simpler than this if I wanted also... but this was just a 5 minute first try...

But you could also go fancier than that... you could put another Ogrid behind the curved portion to better capture the strong swirl that will be there, etc...

 PSYMN March 11, 2012 17:57

Oh yea, this is just the mesh for the central rotating region... around that you would be doing a simple mesh of the flow domain with a circle in the middle of it... How you block that stationary zone will be based on the shape of your outer region, but it will involve an Ogrid for the circular inner boundary...

 CTan22 March 22, 2012 12:17

3 Attachment(s)
Quote:
 Originally Posted by PSYMN (Post 348831) Oh, you want a blocking strategy... First, you will want to generate the mesh for just one of the 3 blades, but make it periodic and then copy rotate the mesh later... In order for the mesh to be periodic, you need equal numbers of nodes upstream and downstream, so you can't just do a CGrid to capture the C shape of your blocking... I recomend an HGrid for the region between the hub and the rotating boundary (or a quarter Ogrid if you didn't have a hub), and then an Ogrid for your blade... Split the Ogrid in half and associate part of the middle of the Ogrid with you geometry where appropriate... This will give you nice boundary layer mesh on both sides of your mesh... This is what I show in these pics... I also smoothed the mesh, although that isn't really necessary... I suppose it could have been about 2 radial split simpler than this if I wanted also... but this was just a 5 minute first try... But you could also go fancier than that... you could put another Ogrid behind the curved portion to better capture the strong swirl that will be there, etc...
Hi Simon,

Sorry for the late reply as I try to figure out how to get the blocking you have suggested. But after I have done with the blocking, it doesn't look good. Can you kindly advice? Thank.

 Far March 22, 2012 12:44

blocking is good. Problem might be related to association. Could you plz also attach the .blk file

 CTan22 March 22, 2012 12:48

Quote:
 Originally Posted by Far (Post 350953) blocking is good. Problem might be related to association.
Hi Far thank for the advice, I will look on the associate issue and see whether can solve the problem.

 CTan22 March 22, 2012 12:56

1 Attachment(s)
Quote:
 Originally Posted by Far (Post 350953) blocking is good. Problem might be related to association. Could you plz also attach the .blk file
Here the file.

 PSYMN March 22, 2012 13:59

I don't have time to look at the blocking, but I agree with Far, it looks like you have associated internal edges with curves and some of the external edges (top left) that should be assoicated with curves are just surface associated.

I can tell this by the color...

Green => associated to curve

White/Black => Associated to surface

Cyan blue => Color the internal edges should be... Not associated to anything.

 Far March 22, 2012 14:01

1 Attachment(s)
It was association problem. Blocking is not good in few blocks as highlighted.
Updated blocking file is attached.

 CTan22 March 22, 2012 14:03

Thank for the advice, Simon. Now I know what went wrong.

 CTan22 March 22, 2012 14:05

Quote:
 Originally Posted by Far (Post 350968) It was association problem. Blocking is not good in few blocks as highlighted. Updated blocking file is attached.
Thank Far!! :) Appreciate for your effort to spot my mistake and correct it. Really thank alot.

 All times are GMT -4. The time now is 00:59.