Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Transient Heat Transfer in Radial Direction

  1. Aug 13, 2011 #1
    My geometry is bit complex. I want to make a heat transfer model based on FDM in the radial direction. But it has become very difficult to discretize the domain because of multiple materials. As the segments are symmetric, I want to model the heat transfer in one segment. My schematic is given in the attachment. I missed a steel shell around the copper shell in the schematic. I am seeking some expert advice on approaching this problem.


    Attached Files:

    • heat.JPG
      File size:
      22.2 KB
  2. jcsd
  3. Aug 20, 2011 #2
    Why use the FDM and not finite element method? That would be a trivial FEM problem.
  4. Aug 21, 2011 #3
    I am not that familiar with FEM that's why i chose FDM. Also, I am trying to simulate it using heat transfer model in COMSOL. Will it be really easy to use FEM? Will it take lot of time?
  5. Aug 21, 2011 #4
    Making the geometry would be harder than everything else. I guess you'll already have the geometry and materials and boundary conditions. Seems like you just need to press a couple of buttons to automesh and solve it. But I never used COMSOL much so can't quite remember the details.
  6. Aug 22, 2011 #5
    I wanted to use FDM to check the temperature profile at different radial points and compare with my experimental results. You are right, its the geometry that is the hardest part in comsol. I am still struggling to create the geometry. :(
  7. Aug 24, 2011 #6
    conjugate Heat Transfer in cylindrical geometry

    I am beginner to comsol 4.2, Can someone provide me some easy examples to inplemente
    navier stokes and energy equation in comsol. I would be very grateful for any kind of help
  8. Aug 24, 2011 #7
    Re: conjugate Heat Transfer in cylindrical geometry

    I think you should start a new thread for a new kind of problem. Did you solve this steady state thermal one?
  9. Aug 30, 2011 #8

    Please look at my geometry and suggest how can I determine the temperature profile at these nodes. I am still struggling to build up a mode with FDM because of multiple material properties. :(

    Also in comsol having some difficulties with the boundary condition.

    Attached Files:

    • FDM.JPG
      File size:
      11.7 KB
  10. Aug 30, 2011 #9
    Sorry I don't have a clue about FDM. I still think you should be using FEM. Comsol can automesh it and you can specify materials and boundary conditions pretty easily. Don't remember how, just that it's easy enough.
  11. Aug 31, 2011 #10
    I have created the geometry and also computed results using comsol but, as the outer shell is extracting heat from air surrounding it, I have specified heat flux as the boundary in the outer shell. This is the only boundary condition I have set. For that I need heat transfer coefficient for the convective heat transfer taking place between steel shell and hot air. How to approximate the heat transfer coefficient for my case? I looked at several sources but could not find anything to calculate heat. trans. coeff.
  12. Aug 31, 2011 #11

    I don't know much about estimating heat transfer coefficients, but there are probably handbooks full of empirical data. It depends on things you haven't mentioned, like the geometry outside, orientation wrt gravity, etc.
  13. Sep 10, 2011 #12
    Why am I getting an uniform temperature profile at different radial positions (please above pic) i.e. the temperature difference between one node to another is very small, when using fin? But without fin (only steel shell) the temperature difference is much higher between nodes, especially the node near the wall has a greater temperature.

    Please explain someone.
  14. Sep 10, 2011 #13
    Is this a transient or steady state analysis? In steady state you should of course expect a uniform temperature everywhere.

    Can you post the color plots of the results showing the mesh?

    Do you have mesh convergence, time step convergence, etc.?
  15. Sep 11, 2011 #14
    It was for my experimental results. My simulation result is also similar to that. But incase of my simulation, i have used constant heat flux and also used a guessed value of h. So results differ from experiment but trends have some similarities.

    Why the experimental result is so uniform in trend?

    Attached Files:

  16. Sep 11, 2011 #15
    It looks like the experimental result matches the simulation except for different heat flux/etc. I can't quite read the scales on the color plots, but if the heat flux is low enough then you should expect it to be nearly uniform.

    Isn't the purpose of the fins to make the temperature more uniform? Which is what you're seeing.
  17. Sep 12, 2011 #16
    Yes the purpose of the fins is to make uniform temperature profile. How can I explain this phenomenon with my geometry? Could you please help? Is is beacuse of a shorter heat path for heat conduction from the walls?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook