Fluids

Fluids

vof and eulerian-eulerian rotating domain – mass conservation violation?

    • desouzadacosta
      Subscriber

      Hello All

      I am trying to simulate a rotating domain with 2d axisymmetric swirl (in the horizontal x direction) with either vof

      and eulerian eulerian approach. I am trying to reach a stable paraboloid surface.

      However the liquid phase keeps being added with no apparent reason (the domain is closed like

      a rotating recipient).

      This is the initial state

    • Rob
      Ansys Employee
      Can you check the mass integral of the phases? What is the source attached to?n
    • desouzadacosta
      Subscriber
      Hi Rob,nThanks for the output. The simulation is 2D axissymetric swirl.nThe fixed value is defined by the udf below, to impose the velocity profiles(_av (axial velocity), _rv (radial velocity), _tv (tangential(swirl) velocity) in the region of the impeller :n#include udf.hnDEFINE_PROFILE(l_b_av, t, i)n{n   cell_t c;n   real pv[ND_ND];nn   begin_c_loop(c, t)n   {n      C_CENTROID(pv, c, t);n      F_PROFILE(c, t, i) = -1.7505*pv[1] + 0.2425*pv[1] + 0.009;n   }end_c_loop(c, t)n}nDEFINE_PROFILE(l_b_rv, t, i)n{n   cell_t c;n   real pv[ND_ND];nn   begin_c_loop(c, t)n   {n      C_CENTROID(pv, c, t);n      F_PROFILE(c, t, i) =7.2332*pv[1]*pv[1] + 0.8786*pv[1] - 0.0679;n   }end_c_loop(c, t)n}nDEFINE_PROFILE(l_b_tv, t, i)n{n   cell_t c;n   real pv[ND_ND];nn   begin_c_loop(c, t)n   {n      C_CENTROID(pv, c, t);n      F_PROFILE(c, t, i) = 12.858*pv[1]*pv[1] - 10.529*pv[1] + 0.1524;n   }end_c_loop(c, t)n}nnAny ideas are very welcome.n
    • DrAmine
      Ansys Employee
      UDF are wrong if applied on cell zone. You need to use C_PROFILE.n
    • desouzadacosta
      Subscriber
      Hi DrAminenThanks for the clue. Should C_PROFILE aplicable for porous zone only ?nThe udf code in the section 2.3.24.5. Example 3 - Fixed Velocity UDF, pretty much intends to do the same as above.n
Viewing 4 reply threads
  • You must be logged in to reply to this topic.