code_saturne User's Forum

Dear all,

I am trying to simulate buoyant flow in an enclosed rod bundle, but the simulation is not converging. Looking at the results it appears there are a couple of cells near the domain boundary where I have non-physical velocities. I have tried switching on EXTRAG for pressure, but the problem persists. Furthermore, I tried using least squares for gradient reconstruction but again same issue occurs. Checking the mesh quality criteria shows no obvious mesh quality issues within this region. (I noted the aspect ratio is rather large, tried improving it but to no avail).

Any help/advice you can give is much appreciated.

Many thanks

Hello,

How does the flow look if you visualize it with vectors ? This might simply be a local recirculation, though in this case it would probably move with time.

How does the time-averaged velocity behave ?

Best regards,

Yvan

Hi Yvan,

Thank you for the prompt reply. I have looked at the cross flow vectors and the region does not appear to be a re-circulation zone. Additionally, I ran the simulation for longer and it appears to move across the domain(please see attached). My courant number also peaks within this zone(Oscillates to values upto 2.3).

I had not started computing the averages, as initially I was hoping to run the case for a few cycles, before computing them.

Regards,