|
[Sponsors] |
April 18, 2019, 17:45 |
symmetry considerations in a turbine
|
#1 |
Senior Member
Svetlana Tkachenko
Join Date: Oct 2013
Location: Australia, Sydney
Posts: 416
Rep Power: 15 |
Hello All,
When modelling a turbine I would like to have periodicity in the mesh. For example if I have six blades in a turbine I would like to mesh one sixths of a cylinder-shaped region and then copy the mesh to the remaining space (like copying a piece of pizza six times to create the pizza). I am not sure whether ANSYS Meshing supports this, and if so, then how it is called and how to use it. Additionally when modelling, is it possible to run cfd simulation only on the one piece of the pizza and then copy the results to the remaining pieces. This is like symmetry but not on a face, rather on three dimensional bodies. As I am using steady state simulation with frozen rotor, can this be useful when the turbine is moving forward without an angle...? Regards, Svetlana |
|
April 23, 2019, 03:27 |
|
#2 |
Senior Member
M
Join Date: Dec 2017
Posts: 703
Rep Power: 13 |
I don't know if/how it works in ansys meshing, but Turbogrid is a good choice to mesh turbomachinery geometry. It is very common to model only one blade because it saves alot of time and resources.
I think the term you are looking for is rotational periodicity |
|
April 23, 2019, 22:59 |
|
#3 |
Senior Member
Svetlana Tkachenko
Join Date: Oct 2013
Location: Australia, Sydney
Posts: 416
Rep Power: 15 |
Thanks AtoHM!
1) It appears that ansys design modeller and Turbogrid has problems with this particular geometry imported from cad, with errors like parts of the geometry may not be validated and subsequent operations may fail. Turbogrid says it does not know where is the leading edge. I tried repairing the geometry in design modeller and in ansys space claim but both of these have not been successful. I also tried to split the top and the bottom surfaces of the blade so that they are separate, but this has not been of use either and I continue to get the same error message. If there is any software that I could use to simplify the geometry aggressively to make it more smooth, it would be great to check. 2) Rotational periodicity is an interface in ansys cfx pre and I think it works well, but in the case the propeller is moving at an angle (forward motion) it no longer applies as the fluid movements are no longer periodic like before. I would be interested in knowing the available modelling options for this situation. |
|
April 24, 2019, 03:09 |
|
#4 |
Senior Member
M
Join Date: Dec 2017
Posts: 703
Rep Power: 13 |
Regarding 1) Yes, you won't be able to use CAD geometry directly in Turbogrid. You need to set up the fluid path in design modeler. This might help https://www.youtube.com/watch?v=2951JXtWJ4w (I haven't watched it in full, but I think it explains the right thing).
2) You mean the incoming flow is not in axial direction? Yes, then you can't use it. However, you can still mesh only one blade and duplicate it to get the full circle. Turbogrid allows you to create 1:1 mesh on the periodic boundaries. The dublication can be done in CFX Pre. |
|
April 24, 2019, 03:50 |
|
#5 |
Senior Member
Svetlana Tkachenko
Join Date: Oct 2013
Location: Australia, Sydney
Posts: 416
Rep Power: 15 |
Thank you AtoHM.
I will check this video tutorial and see if I can get turbogrid to work with this geometry. Just in case it would be great to know whether the 1:1 matching can be accomplished without turbogrid in the case turbogrid does not work with this particular geometry. I'll let you know of the progress. |
|
April 26, 2019, 19:48 |
|
#6 |
Senior Member
Svetlana Tkachenko
Join Date: Oct 2013
Location: Australia, Sydney
Posts: 416
Rep Power: 15 |
Is there a button to automatically split a cylinder shaped geometry into two, three, four, etc equal slices like a pizza? It would probably simplify my modelling considerably.
|
|
April 27, 2019, 18:23 |
|
#7 |
Senior Member
Svetlana Tkachenko
Join Date: Oct 2013
Location: Australia, Sydney
Posts: 416
Rep Power: 15 |
If I have two frozen rotors one straight behind another, do I put frozen rotor interface between them or mixing stage? What is the difference? Do not see any tutorial online about this
|
|
April 29, 2019, 02:13 |
|
#8 |
Senior Member
M
Join Date: Dec 2017
Posts: 703
Rep Power: 13 |
||
April 29, 2019, 02:27 |
|
#9 |
Senior Member
Svetlana Tkachenko
Join Date: Oct 2013
Location: Australia, Sydney
Posts: 416
Rep Power: 15 |
Thanks, I already checked these. I don't see any discussion of the specific use case of two frozen rotor domains being one next to another though.
|
|
Tags |
propellers, turbines |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Problem with torque output from a turbine simulation. | pa-dundas | FLUENT | 5 | July 17, 2022 06:27 |
Symmetry boundary considerations | mrwts | Main CFD Forum | 3 | August 23, 2016 05:45 |
6dof for rotation of turbine: unable to fix the center | Havinh | Fluent UDF and Scheme Programming | 0 | August 5, 2015 06:35 |
Possible Bug in pimpleFoam (or createPatch) (or fluent3DMeshToFoam) | cfdonline2mohsen | OpenFOAM | 3 | October 21, 2013 10:28 |
Cyclic Symmetry - transient- axial turbine case. | Gowrav | Main CFD Forum | 1 | April 13, 2013 09:15 |