Convergence issues
 

Hi guys,

I am running a structured Delta wing grid at low angles of attack (<5). I start off using laminar/steady/1st Order, and then after about 500 iterations, I switch over the S-A/Steady/2nd order.

The velocities conververge rapidly, but the continuity levels off at around 10e-3.

From what I understand in the segregated solver is that Fluent solves the Velocities first, then uses these to satisfy the continuity equation...so how can my velocities be converged but not my continuity?

Any advice? Should I change the under-relaxation factors or the Multi-grid options to help accelerate convergence?

Thanks for your time, Regards Riaan

Re: Convergence issues
 

try PISO , and if you do not see convergence try increasing PISO iterations by 1 . that should help

Re: Convergence issues
 

Thanks - I will try that.

Some further info on my grid : Structured with highest volume equiangle skew of 0.77, and highest aspect ratio = 6000 (but in a region where I don't expect significant gradients).

I dropped the under-relaxation on the momentum equation from 0.7 to 0.5 - and the continuity half, but after that it just leveled off again....

I did notice a separation from my Trailing edge and a reciculation bubble sitting there (from my 10e-3 continuity converged solution) - so I suspect the flow may be unsteady.

But 1st order should still solve it?

Re: Convergence issues
 

I used PISO, and it does seem to help convergence, but if I increase the PISO iterations (I just increased both) to 2, the solver (steady,segregated) always diverges.

Re: Convergence issues
 

well it should help ... but i also tried that .. what i did is i took one of my well working les calculation changed it to steady state k-e and started the calculations .. it was converging well .. then i put that piso with increased iterations and ..yes ..it started giving me warnings like viscosity limited to some values etc. .. i do not know at the moment what is wrong but i would brood over it.