flow past cylinder
 

hi forum members !
hope everyone is fine.
i am simulating flow past cylinder using a self developed code. I have used Riemann invariants based boundary conditions for inflow and outflow (subsonic flow). I am trying to simulate flow at Re =100. But the data as i visualize is physically not seeming ok. From the way the animation appears, could you guess and suggest what could be the problem ? I see a wave front starting from the inflow (left side) going to the right and then bouncing off the right (outflow). from the wave front it appears that the velocity at the inflow boundary appears to retard and then spreads to the domain. i have attached four stage-wise images of the x-velocity contours.
i have tried to cleary tell the whole situation , i will reply if anyone asks further detail.
pls help

nobody even have a hint ? :-)

I can't figure it out the inflow and outflow regions from your computational domain...is your code a low-Mach solver?

thank you much for your reply

of course it is a low mach number solver. but preconditioning has been disabled and solved at Re = 100.

alongside , since you have judged the situation , the question is , IS PRECONDITIONING NECESSARY AT THIS REYNOLDS NUMBER ?

looking forward.

Preconditioner is required for low mach number

...but at relatively higher Mach number, laminar NS equations wont stay valid. Turbulent treatment would be necessary in that case.
Aint it right?

No...at high Mach number the classical NS are always valid (at least for Ma<5). Laminar or turbulent is just a common way to see the state of the flow motion. Turbulence does not depend on Mach but on Reynolds number. Your case at Re=100 can be simulated without any turbulence model.

a little diverting question maybe, but could you give a few tips on dummy nodes/cells? i mean in case of unstructured grid, how would these be defined. at what distance etc. etc. or if there is any source from where one could get some guidance.