CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   Compressible flow - BC problem (

Luk_Fiz April 18, 2010 04:07

Compressible flow - BC problem
Hi all, I have following situation and dont know what BC to use.
The domain is filled with water in cylindrical tank, opened from the top - like the glass filled with water - set as compressible. There is "piston" of given mass hang over the surface of the water (assume that all is in vacuum so there is no air between) at given height. In the moment the cylinder is released and hits the surface. I am interested in observation of pressure waves and its damping.

I have solved this in field of Reimann theory on the paper but have to solve this with CFX. What BC on the top water surface should I use? Do I need to use FSI (I have no idea in this field) or CFX only? How to take into account kinetic energy of the cylinder - it can be hang on different heights?

Thanks in advance,

ghorrocks April 18, 2010 07:39

I am not precisely sure what you are trying to do. A diagram would help.

But it looks like a multiphase free surface simulation with the mass being an immersed solid with a 6-DOF solver should do it. You will need to talk to CFX support about examples of the 6DOF solver as it is a hidden beta feature in V12.

Luk_Fiz April 18, 2010 17:12

I was unable to paint it, and I was thinking that the description is enough. The situation is simple (and almost everyone is able to perform physical simulation). There is a glass of water and a cylindrical mass (of the diameter equal to diameter of the glass) hang above. The situation is started from the point where the mass is released and hits the surface of the water - so basically there is no free surface. The problem is basicaly one dimmensional, along vertical axis. I am interested in:
- amplitude of the mass movement, the maximum depression it follows (as the water is compressible it compreses and realeases and so on). It is similar to Helmholtz resonator problem,
- the maximum pressure in the glass,
- vibrations damping and so on.

Since I am not very much into FSI I am interested in solving this only within CFX with some given BC to the surface of the water, without modelling the mass.
In the second part of the task is to take a leak from the glass cylinder into account.


ghorrocks April 18, 2010 18:23


I was thinking that the description is enough.
It is obvious when that is your task, but not so obvious to us. A diagram is much easier than reading reams of text. Ask any engineer.


Since I am not very much into FSI
I was not proposing FSI, I was proposing a 6DOF simulation. Very different. However if you already know the piston motion you can model this as a simple immersed solid simulation with no need for 6DOF. There are examples of how to do this with CFX V12.

To do free surface simulations like this accurately you will need a very high quality mesh. Preferably hexs. If you need surface tension then you will need hexs with aspect ratio very close to 1. With a fine mesh and immersed solid approach you should be able to get the depression in the water, pressures, damping etc.

Luk_Fiz April 19, 2010 02:35

Sorry for missunderstanding my note - of course for me situation is obvious but for others it can be different.

Since I have cfx 10 I am not sure what "6DOF" mode is - as far as I know it also has not Immersed Bodies.

My problem is generally how to solve this task strictly in CFX, without usting any "outside" help. I have solved this problem analyticaly but dont want to put f.e. piston motion into CFX but rather to set that BC in CFX to have it all as output.


ghorrocks April 19, 2010 06:34

In CFX V10 you don't have the immersed solid option. A good reason to update if you ask me.

In V10 you will have to do this with moving mesh. Much trickier, but possibly more accurate.

All times are GMT -4. The time now is 13:01.