Water Turbine meshed in TurboGrid: Geometry and Topology issues
 

Hi all,
Please I have a couple of question regarding my geometry and topology.

Geometry:




I am simulating a water turbine operating in a river. The turbine has 3 untwisted blades with constant cross-section. The turbine geometry was created using matlab and solidworks. the blade sections was 2D expect the lower one which is attached to the hub. An extra foil was created and attached to the shroud to help trace the shroud curve.

• Geometry problem
  when the hub, blade, and shroud was imported to TurboGrid, and the shroud tip was created, the blade tip was curved and not flat as i expected. this happened even if I cancel the extra curved foil that attached to the shroud (please see file 1)
• Topology problems
  - I tried to use the traditional topology (H/J/C/L-Grid), but an error popped up and I do not know why (please see file 2 for the message)
  - I changed to ATM optimized and the topology went with no error, but the topology tip mesh looks shifted form the shroud. The tutorial for the axial fan also has shifted topology but not to much (Pease see file 3).

Please I have another questions regarding using TurboGrid to mesh a water turbine. I have read in another discussion that I can use General Grid Interface (GGI) to take account of the shroud in order to simulate the water turbine but I do not know what is that means.
(http://www.cfd-online.com/Forums/cfx/22798-can-i-use-turbogrid-mesh-wind-turbine-blade.html)

please my last question is can I incorporate another domain with the domain created in TurboGrid e.g. can I surround the turbine passage with cylindrical domain after I import it to ICEM or any other meshing software this to simulate the real water altitude above the water turbine. Also can I extend the rotor passage length; in other words, make the outlet farther downstream to be attached to the outer cylinder.

I appreciate you suggestions

Thank you,

Aziz

error solved if blade number increased
 

The error message and the topology mesh shift disappeared are caused when I set One-to-One Interface Ranges> Periodic to anything rather than Full. I am still looking for answers regarding the last two question.

Please any ideas are appreciated

thank you

Do not attach the curved profile at shroud. Just extrude geometry at hub and shroud linearly.


This is done to give you best quality. So no worry, after all it is a periodic mesh ;)

[QUOTE] ''Do not attach the curved profile at shroud. Just extrude geometry at hub and shroud linearly.''

Thank you Far for your reply. I extracted linear 3D data for both the hub and shroud. No foil was considered at the shroud, but I still have the same curve blade tip.

thank you though for the suggestion

is it possible to show some images at the blade tip with and without shroud? Please make the shroud transparent and turn on the lines as well...

[QUOTE]''is it possible to show some images at the blade tip with and without shroud? Please make the shroud transparent and turn on the lines as well.''

Hi Far Please see attached

Thank you

geometry seems ok to me

Hi Far I have more couple of questions can I contact you on your email


Thank you,

Hi all I still have problem with the topology mesh.

I used the Traditional Topology method and the topology mesh seems deflected and overlapped at some positions as you can see in the images. This probably because of the small number of blade which caused increase in the passage width . I tried to use control points to fix that but The best I could get is in the images.

Please any suggestion are highly appreciated :confused:

I do not know why the control points cannot be created exactly where I place the cursor. For example, if I tried to position the cursor at a node to fix the distortion it would usually jump either to the upper or lower areas.:mad:

:(I have waist a lot of time I think I will go back and use ICEM.

even your problem is not solved using ATM topology?

No the ATM give you a small errors percent but it yielded a negative volume when generating the mesh.