CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Simulation of rotating pistons (http://www.cfd-online.com/Forums/cfx/114511-simulation-rotating-pistons.html)

donalffons March 12, 2013 09:33

Simulation of rotating pistons
 
Hi everyone!

I'm a beginner with Ansys / CFX and after completing a bunch of tutorials and doing some simple analyses, I am struggling now with the following problem:

I want to simulate the rotation of the pistons of an axial piston pump (like this one). For simplicity, I broke down my problem to what you see in this picture:
http://s7.directupload.net/images/130312/me93j56m.png
The point is to simulate, how the rotating parts influence the fluid. The goal is to find out, how much drag force is created by the system. Until now, I fail to set up the system in CFX-Pre.

I tried several things here:
- Stationary fluid domain & rotating walls: This doesn't work, because the walls are not allowed to have normal velocity components
- Rotating fluid & stationary / rotating walls: I tried around with some boundary conditions, but the results were always not as expected
- Immersed solid (and a different mesh): This worked, but the results were quite incorrect. For example, I got result forces in X-Direction

So, I'm lost here. Can anyone give me a hint on how to solve this? Or should I take a somewhat different approach?

donalffons March 12, 2013 13:59

...for example: When I simulate the problem with a rotating fluid domain, the shaft and pistons as no-friction walls and the housing as a counter rotating wall, I get this as a result (velocity contour):
http://s1.directupload.net/images/130312/4ig6c4qf.png
... which is not what I expect, of course. Since the fluid should move much faster near the center, where the rotating parts are.
When I change the contour plot to velocity in Stn Frame it looks like this:
http://s14.directupload.net/images/130312/6yo54lue.png
...which kind of looks like what I am aiming for. I'm still not quite sure, if this is the right way to do such an analysis and if it actually captures what is happening physically. Can anyone tell me, if what I'm doing here is going into the right direction?

ghorrocks March 12, 2013 18:24

Have a look at the tutorials for the difference between velocity and vel in stn. frame.

Are you trying to simulation the region between the swash plate and the cylinder block?

donalffons March 12, 2013 18:42

ghorrocks, yes, the area between the swash plate and the cylinder block.

I will do have a closer look at the tutorials tomorrow. Probably the multiphase mixing vessel could be a good start...
I would appreciate any comments and advices on my case setup, since I'm totally not familiar with Ansys and CFD software in general.

After the simulation, I want to extract the resulting drag moments with the torque_x@... functions.

ghorrocks March 12, 2013 19:11

The motion will be tricky. You might need to use moving mesh to model it. But, depending on what results you are looking for you might be able to simplify it to a frozen rotor model. That would be much easier, simpler and quicker - if it is appropriate.

Do the tutorial examples on moving mesh and rotating franes of reference to get a feel for these techniques.

donalffons March 12, 2013 19:36

I played around with moving mesh simulations a few days ago, but I really had a hard time in getting the simulation to succeed (because of the negative volume error message). So I actually was hoping to avoid this kind of simulation...
I will investigate on the frozen rotor issue + rotating frames a little more. Thank you for your input!

ghorrocks March 12, 2013 19:40

Yes, movign mesh simulations are tricky to avoid folding the mesh. But that is going to be the most accurate way of modelling it.


All times are GMT -4. The time now is 21:39.