Learning Forum

[fabian.hering]

Hello together,
since I am doing my master's thesis on an optimization of a ruhston-turbine I have encountered some issues regarding the mesh size.
To solve this issue I was thinking of doing 2D CFD simulations instead of 3D, since Discovery uses alot of cells for the 3D-Mesh.
Is there a way to do a 2D-CFD simulation is discovery or do I need to export it into Fluent?

Attached you will find some pictures of my project.

I hope someone can help me out.

 

.

[Rob]

I'm not sure how you'd optimise this in 2d given you're looking at 3d effects as the flow is pushed out radially by the Rushton Turbine: a quarter periodic model would have been used back when 1M cells was too many to run, ie the "good old days". Whilst not being as familiar with the Discovery solver (I cover Fluent) check whether you can use thin blades, disc and baffles as you're likely using a lot of mesh to resolve the thickness. 

[fabian.hering]

Thank you so much, by now I could save alot of memory by using symmetry on the outer volume.
Do you know if there is a way to use symmetry boundary conditions for the rotating fulid zone, since Discovery does not let me apply it to non-cylindrical shapes?

[Rob]

Not symmetry - how is a rotating system symmetric? In Fluent I'd use Rotational Periodic to model a sector of the tank, less straightforward here as you have 4 baffles and 6(?) blades. Actually I'd model the whole thing but I have a "grown-up" licence! :)