[wiiketi]

Hi

I have several questions while doing my research. These questions bother me a lot.

I am modelling the fluid flow in a porous media. My domain is shown in the attachment.

The big cylinder is the porous medium (rock). The blue face is the inlet boundary. Water with constant flow rate (I specified the velocity-inlet in the fluent) is injected from this face. The red face is the outlet boundary, with constant pressure (I specified the pressure-outlet in the fluent). The green part is a tube, connected with the inlet surface. This part is not porous medium, but pure tube. Basically, this model contains two flow regimes, one is fluid flow inside a tube governed by the laminar NS equation, the other one is fluid flow in the porous media governed by the Darcy's law.

I did not show the interior of the rock, just want you to see the tube in the center.

In my experiment, I injected water to this porous medium with a tube in the center, with a constant flow rate. The rock is 1.5-in. diameter by 8-in. length. The tube length is 4-in, diameter is 4mm. The pressure inside of the tube and pressure at the inlet surface is the same.

I need to calculate the pressure and velocity distribution in the domain. This problem is steady state.

My question is, for the inlet boundary condition, should I specify a constant velocity, or should I specify a constant pressure at the inlet surface and tube wall?

Below is what i am thinking.

If I specify a constant velocity, every node at the inlet surface will have this velocity. however, in the experiment, I think most of the flow rate will go into the tube, but not flow on the inlet surface. In other words, we donot know the flow rate distribution on this inlet surface.

If I specify a constant pressure on the inlet surface and tube wall, the domain will not include the tube interior. We are sure that the pressure on the inlet surface and tube interior are the same.

I personally prefer the latter one.Anyone who has some opinions?

Thanks

kai

[wiiketi]

can anyone help me? Thanks