Discussion Forum

Hello everybody!

I'm new to the forum and glad about any help.

I have been working on a 3D cfd/heat-transfer problem consisting of a nozzle which ejects nitrogen gas into a "vacuum chamber" (diffrent chamber geometries). The geometry is quatered.

I'm using a compressible single phase flow with the Algebraic yPlus turbulence model (or k-eps).
At the chamber walls I use the Outlet pressure boundary condition (p_out = 1 mbar, suppress backflow).
At the Inlet I tried the mass flow (SCCM = 1250) and the pressure (p_in = 164 mbar, suppress backflow, normal flow) boundary conditions. (From experiment it is known, that at the particular flow rate of 1250 SCCM corresponds to an inlet pressure of 164 mbar).
The reference pressure is p_ref = 1 mbar.

1. What do you think is the better (more accurate) choice for the inlet bc (I experienced the mass flow bc is harder to converge)?

2. Do you have a hint for effective lowering of the reference and the outlet pressure at the bc since I try to approach 0.1 mbar absolute pressure in the chamber?

3. Downstream the nozzle exit the flow becomes supersonic with some oscillation in pressure and velocity (propably due to shock waves). There is a region were the pressure drops and stays constant at its minimum value of 1 mbar (p_ref). It seems that the pressure "would like to" drop further but can't because of the lower limit which I placed in the segregated solver. Does this constant minimum region negativly affect my simulation or is it "physically correct"? If not, do you have a hint how to avoid this behavior?

Thank you for your help.

Philipp

PS: Please led me know, if you need further information.