[GerhardHolzinger]

I try to simulate an air lift loop reactor with twoPhaseEulerFoam-2.3.x. My domain is roughly cylindrical with a separating tube (draft tube) within. This tube creates two regions within the reactor. The riser is above the inlet patch at the center of the bottom. The ascending gas bubbles entrain surrounding water, thus creating an upward movement of the liquid phase in the riser. The outer region between the draft tube and the outer walls is the downcomer. There, the liquid phase flows downward towards the bottom.

With twoPhaseEulerFoam-2.2 the simulation produced the expected flow. However, with twoPhaseEulerFoam-2.3 I seem to be unable to reproduce my own results. I post the results of the 2.2 simulation in an answer to this post to avoid confusion.

With the newer version of the solver, the gas phase shows high velocities in the downcomer region, thus reversing the liquid flow. What I don't know is what accelerates the air in this region, where no air is present.

Maybe someone here can enlighten me.

[GerhardHolzinger]

Here is the phase fraction field and the liquid velocity of the same case simulated with twoPhaseEulerFoam-2.2.

Here the liquid circulation is as it is expected. Rising in the riser and coming down in the downcomer.