Fluids

Fluids

meshing a coil heat exchanger

    • mounaraha
      Subscriber

      Hello,

      i need your help for meshing a rectangular plate heat exchanger circulated by a serpentine path. I want to have an hexahydral mesh on fluid domain flowing in the serpentine and tetra on the other domain of a phase change material.

      Because i'm not familiar with ICEM meshing, I would like to know if it is possible, in ansys meshing, to do a tetra mesh for all the systen even though it is less accurate than hexa for fluid domain or the second method is to do an hexahydral mesh in the serpentine with the method of slicing the tube into 5 bodies and then couple them in one part and apply multizone mesh and for the rectangular domain i will keep a terta meshing as mentionned.

      for the second approach, i'm not sure if the slicing of the sepentine would affect the interface between the two domains and if it is applicable for transient state. thank you

      here are the necessary links to understand my questions;

      the link of the system modelling is below; https://www.cjoint.com/c/KFrqoQgqt8j

      here is the link of slicing the serpentine for a simple reduced similar system with a Ubend tube, into 5 parts; https://www.cjoint.com/c/KFrsBXr6AKj

      here is the link of the preliminary automatic mesh to show that there are 4 contact regions and not just one; https://www.cjoint.com/c/KFrsEd7JUVj

    • Keyur Kanade
      Ansys Employee
      As ANSYS Staff, we can not download attachments. Please upload images using upload image functionality.
      For hex meshing please check following.
      Ansys Meshing Sizing:
      " target="blank">


      DM Share topology:
      " target="blank">


      SpaceClaim Share topology:
" target="blank">


Ansys Meshing: Decomposition for hex meshing:
" target="blank">


Ansys Meshing: Meshing Methods:
" target="blank">


Ansys Meshing: Inflation:
" target="blank">


Regards Keyur
How to access Ansys Online Help Document
How to show full resolution image
Guidelines on the Student Community
How to use Google to search within Ansys Student Community
  • mounaraha
    Subscriber
    thank you Keyur.
    Here are the images;
    1) the system modelling in designmodeler
    2)here is the image of slicing the serpentine for a simple reduced similar system with a Ubend tube, into 5 parts. It is just done for testing what will the second approach give me

    3)here is the image of the preliminary automatic mesh to show that there are 4 contact regions and not just one contact region
    thank you



  • Rob
    Ansys Employee
    I suspect the four contact regions are because you have four surfaces on the outside of the serpentine channel. As you're not familiar with ICEM CFD, why not use Workbench Meshing as per the last image. Use Share Topology in SpaceClaim and selective meshing: ie mesh the serpentine before meshing the block. You'll not need to do all the blocking but I would split the serpentine to help the mesher keep the quality around the bends. Read up on how to sweep a mesh with inflation as the set up is slightly different to how you'd expect.
  • mounaraha
    Subscriber
    thanks Rob.
    for the meshing, i did it in Workbench Meshing.
    When splitting the serpentine, i think i will always have 4 contact regions and that could normally affect the results . Is it roght?
    Also, as i understand, you suggest using Spaceclaim for meshing. I read that it is used for finite element analysis not finite volume analysis as i want.
    thanks for clearing
  • Rob
    Ansys Employee
    Not SpaceClaim, use Workbench Meshing. You can get a decent mesh out of SpaceClaim but I'd use either Meshing or Fluent Meshing.
  • mounaraha
    Subscriber
    Okey, and for splitting the serpentine, would it affect the flow and the boundary layers in contact with the rectangular domain filled with phase change material (it will undergoes a solidification )? thanks
  • Rob
    Ansys Employee
    The splits you have will give an O-grid, very popular in the 90s but not used much now. If you use inflation with a reasonable aspect ratio freezing the liquid in the channels will be OK. Just watch you retain enough space for the moving fluid and enough mesh resolution.
    The splits I'm suggesting cut across the channel so you'll have a series of straight and curved sections. It just gives the mesh something to anchor too, an axial split in the xy plane will help with that too.
  • Viewing 7 reply threads
    • You must be logged in to reply to this topic.