Hi,

I have a slug flow problem. Can anyone tell me how initial conditions and boundary conditions are usually specified? Thanks!

Pei-Ying Hsieh

(1). I would say, do whatever you like. (2). Initial condition is something you specify first for the transient flow calculations. For the steady state calculations, you have the initial guess of the flow field. (3). The boundary condition is something for the wall boundary, inlet, outlet, or symmetry boundaries. These determine uniquely the problem you are trying to solve. (4). By the way, what is a slug flow problem?

Hi, John,

Thank you for your reponse.

For oil industry (although I am not in the oil industry), they transport oil through pipes. In the oil pipes, air and water also exist. I believe some people call this slug flow (or, in waste transport industry, they have the same slug flow problem).

In my problem, surface tension effect is strong and it is a transient problem. If I consider oil as the primary fluid, how to specify the shapes of air and water initially since the shape depends on flow (geometry, velocity) and the surface tension (the radius of the air/liquid bubbles are greater than the radius of the pipe, so that they are not spherical).

As for BC, the oil/air interface moves into the domain of interests. For each grid point at the inlet, whether it is a interface varies with time and the curvature of the interface. Again, the curvature of the interface is unknown (depends on flow and surace tension). I tried uniform inlet velocity and a vertical stright line for interface (the pipe is horizontal), I got strange results and I don't think this is reasonable anyway.

Any suggestion? Thanks!

Pei-Ying

(1). Very good. At least now I know that you are dealing with flow in a pipe. The geometry is a simple one. (2).You are not looking at small air bubbles, instead , a very large air bubble which is larger than the pipe raduis. Perhaps, what I should say is, certain portion of the flow in the pipe is air, and certain portion of the flow is oil. And there is always interface between the oil and the air region. (3). If you have a pipe connecting to a tank. The pipe is initially filled with air, and there is a gate at the entrance of the pipe which separate the oil from the air in the pipe. (4). So, you have the initial conditions, with velocity =0 everywhere, and the pressure in the tank is set to a value greater than that in the pipe. (5). At t=t0+, you open the gate (assuming that it does not disturb the flow pattern at the time the gate is open), and start the flow. So, I think, it is possible to assume that the initial interface is a straight line or a flat plane surface.

It looks like you are interested in three-phase flow (oil, water and gas). There is a review article in Multiphase Science and Technology (1998). It is a very difficult problem and little is known on multiphase flows (if multi > 2). I would suggest to review this article as well as a book on two-phase flows (e.g. http://www.cfd-online.com/Books/show...hp3?book_id=58). Generally, slug flow is limied to a range of flow conditions and history. In a horizontal pipe it can have an axisymetric shape of elongated bubble or a nonsymetric plug depending of gas and liquid flow rates as well as on other parameters (e.g. inert gas or evaporated light components of oil). Generally, for a fixed flow conditions slugs travel in a series with only statictically repeatable shape and size. For transient formation of slug flow you may want to check "The Bubble-Slug Flow Pattern Transition and Instabilities of Void Fraction Waves" Int. J. Multiphase Flow, Vol. 13, No. 2, pp. 199-217, 1987.

Andrzej

Hi, Andrzej,

Thank you very much for the info. They are very helpful to me.

Pei-Ying

Hi, Andrzej,

Could you please send me the title of the review article in Multiphase Science and Technology? Thanks! (There are 4 volumes in 1998, also, there is an article in 1999 talks about the view of the future).

Pei-Ying

"Multiphase Pipeline Flows in Hydrocarbon Recovery", Volume 10, Number 1 (1998) by Arne Valle. Check

http://www.begellhouse.com/mst/mst_previous.html#10_1

You may find there some other interesting staff on multiphase flows.

Andrzej