[kebsiali]

Hello Dear foamers

I'm modelling flow inside a wind tunnel "a real experimental wind tunnel"
The wind tunnel has a devised geometry to generate a huge recirculation bubble.
I had the geometry meshed and ran using Spallart-Allmaras turbulence model it converged in simpleFoam and then i ran it in pisoFoam for around 20seconds of physical time but the physics sought is not seen due to the lacking of an LES or DNS abilities.
I jumped into using IDDES models either Spallart-Allmaras IDDES in OF3 or 4 or k-omega SST-IDDES in OF3+ or 1606+
I dont know if these models have been validated yet or not
and it is because of this that I'm here

when I run my simulation using the IDDES models the calculation diverges
I've tried piso and pimple, backward and crankNicolson 0.9, tried to increase nCorr, nOuterCorr, nNonOrthCorr, but it doesnt improve
Im using the LUST scheme for the div(phi,U), tried to play with the solver parameter forcing lower tol and relTol increasing the sweeps of the GAMG p solver without success

What I notice is big vortices being generated downstream of the main recirculation.

I have tried a full 3d case and an extruded 2d case (2.5d) but have the same problems.

any ideas please???