Fluids

Fluids

Reseversed flow

    • hugo CFD
      Subscriber

      I am simulating a flow along a heat exchanger, quasi counter flow, as the image represents. For each channel, in the cross-flow zone, the fluid only enters on one side, the other side being closed.



      During the simulation, during the iterations the message "reversed flow in 16 faces on pressure-outlet 24" appears to me. How do I solve this problem, which affects the convergence of the solution?

    • seeta gunti
      Ansys Employee

      Hello,


      If the number of faces are reducing over the iterations, you can neglect the warning message and continue the run. 


      We need to be worry if the number of faces occuring the reverse flow are increasing over the iterations, you need to extend the outlet region to avoid reverse flow.


      Regards,


      Seeta


       

    • hugo CFD
      Subscriber

      I can not change the geometry of the heat exchanger.


       


      The reserved flow affects the result of the fluid outlet temperature, right?


       


      regards


      Hugo

    • DrAmine
      Ansys Employee

      Post a picture of your fluent case or better pictures of the geometry.


      If you have unphysical reversal flow it will affect the overall result.  

    • hugo CFD
      Subscriber

      my geometry is the same as the image above

    • seeta gunti
      Ansys Employee

      Can you send us the screen shot of your mesh, models, solver settings and methods and residuals etc.


      You did not mention whether the number of faces that reverse flow occurring are increasing or decreasing when the solution progressing.

    • hugo CFD
      Subscriber

      The mesh of all geometry, fluid and solid



      Mesh only of the solid



      And, the mesh only the fluid



      Now, i will show the setup parameters


      Models



      Methods



      and controls



      The number of reverse flow faces decreases to a certain value, then holds. The image shows the minimum face value that is reached.



      Ps: If I change the discretization method to second order upwind, it does not reach convergence and the residuals of the continuity equation stabilize at 10 ^ -3.

    • Karthik R
      Administrator

      Hello,


      Have to tried changing the pressure velocity coupling from SIMPLEC to Coupled?


      Have you tried using Pseudo-Transient?


      Have you tried to let your simulation run longer? Does the reversed flow resolve itself?


      Thank you.


      Best Regards,


      Karthik

Viewing 7 reply threads
  • You must be logged in to reply to this topic.