with k-epsilon turbulence

Is it incompressible and steady state then good choice

if not bad choice

it is incompressible and steady state

theoretical results shows velocity gradient generated is small so we can assume that

I cant say anything about that

Your problem is ill-posed: you cannot have both velocities and pressure be Neumann-type (zeroGradient) at the same boundary.

And boundary conditions are the problem once more

Feels like a good 90% of noob errors are BCs

I tried giving totalpressure bc to both inlet and outlet along with pressureinletoutletvelocity bc for velocity

Still I face the same bounding error problem

Because of this documentation only I gave totalpressure to both inlet and outlet

Which are the boundaries for U in your domain? We know that inlet/outlet are Neumann, What about the others?

Inlet have pressureinletvelocity and outlet have zerogradient

My main prob is bounding of k/epsilon

What about the other boundaries of the domain?

There is an utily called boxTurb in ofv21 «Create a box of divergence free turbulence conforming to a given energy spectrum » but i don't know how to use it.

dnsFoam has an example for that I believe

i am out now i will send all the boundary condition data tomorrow

It's not clearly indicated but your mesh has to be regular to use this as it use FFT technique. This is why I was having issue. Of course my first reflex was to grep the whole tutorial folder for this word ^^

Ok so I changed my epsilon and k conditions

dimensions [0 2 -3 0 0 0 0]; internalField uniform 50; boundaryField { inlet { type turbulentMixingLengthDissipationRateInlet; mixingLength 5.96e-3; value uniform 50; } outlet { type zeroGradient; value uniform 50; } walls { type epsilonWallFunction; value uniform 50; }

For k dimensions [0 2 -2 0 0 0 0]; internalField uniform 1; boundaryField { inlet { type turbulentIntensityKineticEnergyInlet; intensity 0.05; value uniform 1; } outlet { type zeroGradient; value uniform 1; } walls { type kqRWallFunction; value uniform 1; } }

Initially I still get bounding epsilon but it goes away after couple 100 of iterations

Unfortunately I had to stop iterations after 500 iterations due to low disk space but now I have again started it

Hopefully it will converge fully I will keep you all posted

your U and p BC?

For P dimensions [0 2 -2 0 0 0 0]; internalField uniform 0; boundaryField { inlet { type totalPressure; rho rho; psi none; gamma 1; p0 uniform 50; value uniform 50; } outlet { type fixedValue; value uniform 0; } walls { type zeroGradient; gradient uniform 0; value uniform 0; }

For U dimensions [0 1 -1 0 0 0 0]; internalField uniform (0 0 0); boundaryField { inlet { type pressureInletOutletVelocity; value uniform (0 0 0); } outlet { type inletOutlet; inletValue uniform (0 0 0); value uniform (0 0 0); } walls { type noSlip; }