[lzhaok6]

Dear all,

I've had a stability problem simulating a transient fluid-structure interaction problem (internal flow in pipe) using FLUENT intrinsic FSI solver. Whenever the dynamic mesh is turned on, instability occurs with growing continuity residual and sudden and drastic change of pressure at each time step. To demonstrate this problem, I've made a very simple pipe section (aluminum) with water inside flowing at 6m/s. k-omega SST model is used. Please see the pictures attached for the weird pressure behavior. If the dynamic mesh is turned off (only exchange pressure and displacement without moving the mesh), the pressure behavior becomes stable and normal (positive pressure at inlet and gradually decreased to zero at outlet) and the simulation is stable. So I'm pretty sure the instability comes from the dynamic mesh module.

Things I have tried so far but did not solve the problem:
1. Refine the mesh.
2. Decrease the residual.
3. Using implicit mesh update.
4. Decrease the time step size (smaller dt seems to be more unstable).

I've also attached the screenshots for my other configurations (fluid_sym is symmetry surface of the pipe section which I set as deforming).

Thanks for your attention!