Fluids

Fluids

FSI of Vibrating Channel

    • Joseph Seamons
      Subscriber

      I have a 2-way-coupled fsi simulation where I am trying to simulate flow through a very flexible channel (channel modulus 3600 pa) and apply a transverse vibrational displacement to the channel of about 100 hz (see pictures for explanation). The ends of this flexible channel are connected to much stiffer channels as well to simulate better what is happening in our experimental setup.

      I have the simulation setup and have tried running it, but the problem is that when I introduce the transverse vibration to the tube, I get a reverse flow warning from the fluid solver which cause the elements in the solid domain to warp in a very odd way (the channel essentially collapses on itself).I am hoping I can get some help from someone who would be able to look over my setup and perhaps help me identify what I need to adjust to get convergence. Thank you for your help!

       

    • Rob
      Ansys Employee

      What flow boundaries are you using? What stops the flow ballooning the flexible part? What is the initial flow condition? 

    • Joseph Seamons
      Subscriber

      I am using a pressure inlet and oulet. I do not have anything in the solid domain preventing the ballooning of the channel as that is part of the physics I would like to capture. 

      I should mention with the inlet pressure, I am slowing ramping it up over the course of 0.5 seconds.

    • Rob
      Ansys Employee

      OK, which makes the mass flow a result of the calculation. If the initial flow conditions aren't good the tube expansion could be fast enough to then really mess with the flow field. If you run with no deflection how does it behave?

    • Joseph Seamons
      Subscriber

      When I run with no deflection, I am able to get the solution to converge without any issues as long as my inlet pressure doesn't exceed 1500 pa (which is fine for my considerations).

      The problem comes when tube deflection is introduced. The largest deflections I have tried are about 1 mm.

Viewing 4 reply threads
  • You must be logged in to reply to this topic.