How to force interface and fluid zone to move?

    • Excited

      Hello everyone,

       I want to push a fluid zone which consists of zones and interfaces to move, details have been shown in Figure 1, the whole zone consists of three zones(top, mid, and bot), and they are connected by interface, the zone colored by green is the one I want to exert a force y to push it move along y axis, and the bot zone will be continuously compressed, i.e., height of outlet will decrease. However, once I started to calculate, errors popped out, which is shown in Figure 2.

       In my opinion, I guess the problem may come from interface movement, I assigned the force y respectively to

      (1)fluid zone: top zone and mid zone;

      (2)faces used to form interface:

      bot face of top zone(interface_to_mid_inlet),

      mid top face(interface_to_inlet_mid),

      mid bot face(interface_to_film_mid),

      bot face 1(interface_to_mid_film);

      (3)wall: cylindrical face of top and mid;

      Details about dynamic settings are shown in Figure 3.

       UDF code is also presented.

       Any advice will be highly appreciated, thx!



    • Rob
      Ansys Employee
      Where are you applying the force? Why do you need several interface zones? How are you accounting for the zone compressing? Ie which remeshing schemes are you using?

    • Excited
      Hi, (1) I want to apply the force on ÔÇÿinletÔÇÖ boundary, and then then whole green area colored in Figure 1 will be drove by this force, however, I cannot select other faces as a whole, so I select each of them and applied force on them.
      (2) I want to obtain high mesh quality through structured grid, so I divided the whole fluid into three zones: top, mid and bot(film);
      (3) All of mesh methods (ÔÇÿSmoothingÔÇÖ, ÔÇÿLayeringÔÇÖ and ÔÇÿRemeshingÔÇÖ) are adopted in my simulation.

    • Rob
      Ansys Employee
      OK, so you're moving the top section as a rigid body? Ie it's not deflecting. If it's not deflecting you only need layering and may want to rethink which parts move relative to the others: have a look at the tutorials and focus on the in-cylinder one(s) as that's close to what you want.
      To add, a structured mesh may not be the best option in many cases. Here I'd go with a swept mesh where my source face(s) may be a mix of structured and unstructured quad and tri elements. We've not used full structured meshes for most tasks in years, but many text books (and lecturers) may be in need of updating as the technology has changed a lot in the 20 years I've been with Fluent and Ansys.
Viewing 3 reply threads
  • You must be logged in to reply to this topic.