[here_for_help]

The system I'm trying to model involves two immiscible liquids (one on top of the other) and an inlet and outlet at the bottom of the container, which feeds and empties, respectively, the bottom liquid. The fluid flow is driven by a pump downstream of the outlet, which creates suction and pulls the liquid from the container and recirculates it into the inlet, thus having a constant flow rate. In my simulation, I am attempting to see how the flow profile changes by changing the suction pressure.
Currently, I'm modeling the inlet and outlet as pressure inlet and outlets. For the Gauge Total Pressure on the inlet, I am inputting the hydrostatic pressure at the top of the inlet. For the Gauge Pressure on the outlet, I am inputting a lower value so as to create a pressure difference between the inlet and outlet. I am also specifying that the backflow direction is normal to the boundary and based on total pressure, with a volume fraction of 1 for the flowing liquid.
The solution converges relatively well and the flow profile is generally as expected, except for the reversed flow at the inlet and outlet. To give more detail, ~ the top half of the inlet is flowing liquid into the container, while the bottom half is draining it, and at the outlet, only the bottom ~ 1/3 is actually draining liquid out, while the top 2/3 flows liquid in. I've tried changing the inlet's Gauge Total Pressure to the hydrostatic pressure at halfway the inlet and at the bottom of the inlet with the same results, making it 0, specifying a supersonic/initial gauge pressure, and nothing seems to work. Any ideas?