Fluids

Fluids

Topics relate to Fluent, CFX, Turbogrid and more

Unable to patch Cell Registered Region to define phase for multiphase simulation

TAGGED: 

    • Allen P Varghese
      Subscriber

      I have set up a multiphase model in Fluent with 2 phases (gas and liquid nitrogen)

      I created a region in Adapt>Cell Registers>New Region

      I used Save/Display to see the defined region and everything looks good

      I initialized the domain using standard method with phase2 (liquid) volume fraction as 0 over the entire domain

      I then patched phase2 (liquid) volume fraction as 1 in the defined region. Please find attached screenshot of this step. Please note that I have only selected the region and I have not selected the zone. I get the message (certain number of cells have been marked).

      Next to verify if the patching was done correctly, I created a contour to show the volume fraction of phase2 (liquid) but the result shows the whole domain having 0 volume fraction (implying that the region patching did not work). There are no error messages.

      Attached Figures:

      Region that was defined earlier in Cell Register and was patched in Fig1

    • Rob
      Ansys Employee
      Put a slice down the domain, probably x=0 or z=0 and look at that. n
    • Allen P Varghese
      Subscriber
      Hi Rob,nI applied a slice on XY plane and I have attached it as a figure. It also shows that the domain did not get patched.nn
    • Allen P Varghese
      Subscriber
      I was able to fix this by using the marked region to separate the single cell zone within Mesh>Zone>SeperatenOnce I had two cell zones, I patched each zone with respective values of phase volume fraction and I could observe the result in the initial contour.nThis however still does not explain why I couldn't achieve the same result just by patching the region. Confused (maybe a bug?).nnnn
    • Rob
      Ansys Employee
      Patch should work, and I've been using the feature. Initialise with zero n2liquid and then patch the region. What is odd in the first plane image is the line across the image. n
    • Allen P Varghese
      Subscriber
      That line indicates a perforated plate which is part of the model. The region I created depicts the fluid volume below the plate and I was trying to patch that region with n2liquid to 1.nI did initialize with 0 n2liquid over the whole domain and then tried to patch the marked region with n2liquid 1 but it wasn't working. I tried different models and also different meshes to troubleshoot but for some reason, regions aren't getting patched. That is when I found that cell zones are getting patched successfully, thus I split the cell zone into two and patched them as explained in the previous post.n
    • Rob
      Ansys Employee
      OK. Looking at the mesh display what is the perforated plate defined as? n
    • Allen P Varghese
      Subscriber
      Initially I tried assigning the walls of the perforated plate as PerfWall named selection and the other boundaries as Wall named selection in DesignModeller. My intention was to apply a BOI on PerfWall once in Fluent meshing mode. I went ahead and created a region but it wasn't patchingnIn my next attempt, I named all boundaries as Wall and did not do a BOI in fluent meshing. Created the region in cell register but it wasn't getting patched either.n
    • Rob
      Ansys Employee
      What concerns me is the fact that the perforated plate isn't a distinct wall. I'd expect a wall & wall:shadow pair. If the plate has a finite thickness the cell count in that area is likely to be excessive. nIf you separate the wall by angle (85 degrees ought to do it) what happens? n
Viewing 8 reply threads
  • You must be logged in to reply to this topic.