Learning Forum

[Gabriel Earle]

Hello Community!

I’m having issues with my simulation converging in certain situations. I am trying to find the drag and lift coefficients at different angles of incidence towards a pentagonal cylinder for my university project. I'm new to Ansy’s fluent and I am self-learning which I must admit is proving quite difficult. For some angles closer to a corner orientation the solution converges (see the figure with the flat line for drag coefficient) well but towards the face orientation the solution does not converge well (see other figure of drag coefficient) and produces a strange discontinuity when the different angles are plotted together (see the figure with polynomial fitting).

I have two questions:

1) How can I get results that fits the polynomial fitting and converges rather than the discontinuity.

2) Is there a way I can automate the process of doing these simulations for different angles?

Thanks in advance, any help would be greatly appreciated!

[Rob]

Assuming you're correcting the reference area in Fluent and the mesh is good enough you should be OK. Review the flow field around the problem area, are you getting separation/stall effects?  Velocity vectors are useful to do this. 

You can automate some of the work, but depending on goals and compute resource it may not be worth the effort. For different angles are you altering the geometry or the inlet velocity vector components? 

[Gabriel Earle]

Hi Rob,

Thanks for getting back to me!

I am using 2D analysis and using the inner circle diameter as my reference length so that it is independent of the angle. This is fairly well established in literature for this research so I don't think this is the issue. I conducted a grid sensitivity study so I think that the mesh is fine enough. I will look into the problem area with velocity vectors - any hints on what i should be looking out for?

For changing the angles I am altering the inlet velocity vector components so that I do not have to remesh etc. each time.

[Rob]

Look for separation, ie stall. If you've followed the usual y+ guides you may find you need to refine the streamwise mesh too as flow cross inflation just sees high aspect ratio cells.