Fluids

Fluids

Periodic conditions in two dimensions

    • Amir
      Subscriber

      Hello everyone,

      I am going to study a 2d two phase problem in a periodic domain, from left to right periodic also bottom to up

      so defined two pairs of peridoc boundary conditions,

      there is no pressure gradient, between two sets of periodic boundaries (in x and y direction).

      in the tab of periodic conditions, there is only one pressure gradient option (which I set to 0), also one flow direction (I set x =1 and y=0, but my case is in two directions periodic and should let the droplet flows in two directions)

      My question is how I should do the right setup for my case? as it seems my droplet only flows in x-direction.

      Thank you in advance

    • Prashanth
      Ansys Employee

      Hi

      I think adding few screenshots of the model you are currently working on might be helpful to undertand the problem better. 

    • Amir
      Subscriber

      Hello, thank you for the reply. This image is for one case, for every other cases, I also need x direction and y direction periodic BCs.

    • Prashanth
      Ansys Employee

      Hi

      If you are trying to track multiple parallel patched bubbles, reconsider what you are trying to do here. Switch to a 3D domain section to actually study a bubble travel with two extreme sides reserved for inlet and outlet. Set the rest of the planes as periodic or symmetry.

    • Amir
      Subscriber

       

      thank you, in fact I am trying to study coalescence of some small droplets, this image above would be the final shape,

      here in this problem should not be any inlet and outlet, so the desired BCs are periodic which in the simulations using coding and methods like lattice boltzman method are common, i.e., two pair of periodic BCs.

      is there any way to use two pairs of periodic BCs horizontally and vertically without any pressure gradient horzontally and vertically so the droplet able to freely move in two directions of the so-called periodic domain?

       

    • Rob
      Ansys Employee

      You can only have one paired set of translational periodic boundaries. 

    • Amir
      Subscriber

      OK, thank you, so that is the software limitation,

      Is it possible to write a UDF for this purpose as an equivalence?

    • Rob
      Ansys Employee

      It's possible, but I'll leave you to write it. Please post the code once you've got it working so others can benefit. 

    • Amir
      Subscriber

      Ok thank you

Viewing 8 reply threads
  • You must be logged in to reply to this topic.