Fluids

Fluids

gas distributor with vacuum outlet

    • lorenzo_2
      Subscriber

      Hi,

      I need to simulate the flow of gas inside a ring distributor. In order to reduce the computational cost I created a portion of the chamber. The flow passes from an initial conduit into a larger volume by passing through a very small hole which results in a choked flow. Then the flow, after expanding in the chamber, follows the outer and inner ducts, passes through slots and finally exits the distributor in a channel facing the vacuum.

    • Rob
      Ansys Employee
      What is the density in the system?
    • lorenzo_2
      Subscriber
      In the case with 1 Pa:
      in the VOF the density varies from 1.185e-5 to 8.347e-2 kg/m^3 . Near the outlet section the density is about 5e-4/1e-4 .
      In the case with 1e-4 Pa:
      in the VOF the density varies from 5,380e-10 to 8,340e-2 kg/m^3 . Near the outlet section the density is about the same as in the previous case.

      I used a pressure of 4000 Pa as input and the laminar model with low pressure boundary condition. Also, the distributor walls are at 900 K while the chamber walls are at 1000 K.
      Thanks for the availability.

    • Rob
      Ansys Employee
      Is there any symmetry in the domain? Just wondering how much mesh you've got in the interesting region.

    • lorenzo_2
      Subscriber
      I used the watertight geometry meshing workflow and I added a body of influence near the first hole.

      I have defined the lateral surfaces, which delimit the portion of the ring, as planes of symmetry (even if it is not really so but I think that as an approximation can go well).

    • Rob
      Ansys Employee
      It might be mesh related, can you adapt in the nozzle unit without hitting the cell cap?
    • lorenzo_2
      Subscriber
      Unfortunately, no. I've also tried creating a submodel where I use only the volume downstream of the slots, using a denser mesh at the point of interest, but I get the same result; this makes me think that the mesh is not the problem.
      Thanks a lot anyway for the advice
    • Rob
      Ansys Employee
      Double precision?
    • lorenzo_2
      Subscriber
      yes
Viewing 8 reply threads
  • You must be logged in to reply to this topic.