[jpanchog]

Hello all

I'm doing for my investigation proyect the simulation of 3D laminar flow in a square duct (Lx = Lz = 5 (cm), Ly = 50 (cm)).

The problem that we are experiencing is that the z-velocity component (transversal to the flow) grows unrealisticaly, provocating global divergence after a few iterations (see attached figure).

We are using the compresible unsteady SIMPLE algorithm with power law scheme, fluid used is air and inlet velocity (y=0) is 2.5 (m/s).

After one month we have failed to encounter a solution to this problem, can you give me an idea please??

Thanks in advance!!

Best regards

[FMDenaro]

Hi,

1) the Re number is O(10^3) for the reference lenght (Lx,Lz) but increases along the axis, are you sure that for a full confined duct your field produces a steady laminar solution?
2) you compressible solver can produce stable and oscillation-free solutions for low Mach flows?