Hi, i am a beginner with Phoenix.I want to know how to simulate a gas-solid pipe flow with Phoenix 3.4 in cartesian coordinates . When I use the Ipsa two phase flow ( air- sand), the domain material is phase 1 (air), phase 2 ( sand), the problem is how to deal with the pipe material. I have not seen a demo on the same thing. thank you

you should not model the pipe as blockage but perhaps as thinpipe or plate. hope this help you moh

hi,

thank you for your advice, how can I discard the computational domain outside the pipe. for the two phase flow inside the pipe, I have an inlet, the pipe is simulated as ( thinpipe, blockage), and an outlet. the method is ipsa full. thank you

tahar,

If you are looking at pipe flow try using polar co-ordinates.

Otherwise, make the intlet and outlet the same size as the pipe, keep the area around the pipe as aminimum. This should speed up computational time and keep the flow inside the pipe.

Hope this helps - Leon

Leon,

Thank you for your advice. Because the geometry of the system we are studying consist of two horizontal pipe, and one vertical. It'a difficult to use polar coordinate.

I began to simulate the flow in one horizontal pipe. The model consist of one inlet, one outlet, and a pipe ( thinpipe, blockage). The inlet and outlet size are the same as the pipe. Even if the area around the pipe is minimum, the computational is performed inside and outside the pipe. So my question is can I limit my computational domain to the flow inside the pipe itself.

tahar,

Unfortunately, not that I am aware of. Some one may correctly me on this one if I am wrong.

Try looking a fine gri volumes. This way you can refine your mesh inside the pipes and have a very course mesh outside of the pipe.

Another option maybe to look at BFC. I dont use this system often but it maybe of some use to your problem.

Hope this helps.

Leon

Leon

Once again I than you for your suggestion. I think that with the thinpipe object geometry there is a problem since the wall thickness is much more smaller than the grid cell size and any grid refinement is difficult. so I will try BFC. regards,

Tahar