[newuser!]

Hi there,

my model consists of a straight, waterfilled pipe. One end of the pipe is closed by a thin rubber membrane, where the fluid interacts with the structure. The other end is open and "loaded" by a pressure of 1000Pa.
The aim is to calculate the transient solution of the vibration of the membrane, e.g. at the centernode of the structure.

Performing a multifield Ansys-CFX simulation, I had a "kick-off" model with coarse mesh, without partitioned bodies, .... . There, the FSI worked well and I got a physical solution.

Now, I refined the mesh, partitioned the fluid and structure bodies and tried again to perform the simulation, but it failed, see files attached.

1) As I guess, the simulation aborts since structure elements are distorted. Am I right or is it the CFX-part that fails?

2) Can anyone give me advice what could be the reason for these problems? I read already the thread " FSI and size of time-steps" but I could not get rid of the problem.

Best regards
Philipp

ANSYS.txt

Fluid_001.txt

[mvoss]

The Ansys output shows a relatively high z-force after the very first time step (5e-4) you should check your initial fields.The error is due to the Ansys part, since CFX isn´t starting at all.

[newuser!]

Dear Mr. Voss,

thank you very much for this hint. I applied the pressure of 1000Pa instantaneous and used an incompressible fluid at the same time....
Maybe that was the reason for failure.

Using a slight ramp for pressure load (also reduced to 250Pa) and a slightly compressible water, I still get high values of Fz, but the simulation does not abort.
Do you have an explanation for this?
The pressure acts normal to the xy-Plane....

[newuser!]

I appended the Ansys-out-file....

[zzura]

may i know how you ramp the pressure

[ghorrocks]

This is basic stuff which is covered in the CFX tutorials. Have you looked at them?

[hellowqy]

Originally Posted by newuser!

Dear Mr. Voss,

thank you very much for this hint. I applied the pressure of 1000Pa instantaneous and used an incompressible fluid at the same time....
Maybe that was the reason for failure.

Using a slight ramp for pressure load (also reduced to 250Pa) and a slightly compressible water, I still get high values of Fz, but the simulation does not abort.
Do you have an explanation for this?
The pressure acts normal to the xy-Plane....

Dear Mr. Nnewuser
Thank you very much for you information. I recently also do the FSI simulation (the interaction between water and membrane), which is very similar with your case. Here I have some questions: 1. Why my simulation stops at the beginning even though I set the load to be zero. The error report said that there some elements of highly distortion. I improved the element quality and decreased the time step, still the same results. someone said my be the rubber material for the membrane is the problem. the data for the rubber material is density=1100kg*m^-3,young's modulus=1e7 Pa
poission's ratio=0.45.
2. How to make the liquid(like water) to be incompressible.
thanks a lot!