Fluids

Fluids

Condensation modelling in Fluent

    • Hassan_Kashif
      Subscriber

      Hi,


      I posted to this forum a few week ago but was removed since it was attached to another thread and haven't posted a follow-up so here it is after diagnosis.

      I am trying to model condensation in a 3D domain. I am using the VoF model with Evaporation-Condensation Lee model and have followed set ups from various papers and Youtube tutorials which go through their set-ups (in varying detail) and have been unsuccessful. I have refined my mesh in the ways recommended (inflation layers, control, skewness) and the used a variety of turbulence modelling, boundary conditions and simulation controls (correct patching, time step < 0.005, large iteration number). I noticed that all of the work I have come across which details condensation set ups set up the phases changes in the following way:

      From Phase - Water

      Species - H2O

      To Phase - Water Vapor

      This initially struck me as incorrect since I am trying to model condensation so I should be going from Vapor to Water and not the other way around. So when my simulation failed to show condensation, I adjusted this model to:

      From Phase - Water Vapor

      Species - H2O

      To Phase - Water

      When I hit initialise I get the warning: Latent Heat cannot be less than 0! I tried running the programme anyway to see what would happen and I encountered a floating point exception. After googling on this forum about both the floating point exception and the Latent Heat cannot be less than 0, I made adjustments to my mesh (for the floating point exception) but could not figure out what to do for the Latent Heat is less than zero. I read that is has something to do with material properties not fully defined in fluent but as far as I could see I didn't see anything missing. I will add I am very new to this type of modelling but have a small grasp after a couple months of (mostly) failing.


      TLDR:

      If I am trying to model condensation. In the Evaporation - Condensation model is the arrangement of the "From phase" and "To Phase" important?

      From Phase - Water

      Species - H2O

      To Phase - Water Vapor

      shows no condensation.

      From Phase - Water Vapor

      Species - H2O

      To Phase - Water

      gives a warning of Latent Heat cannot be less than Zero! I am not sure what I am missing in the material properties.


      Screenshots showing Temperature and Volume Fractions in post processing of my most recent set up:

    • Rob
      Ansys Employee
      There are rules surrounding how the from and to phases are set: since the model is evaporation AND condensation the order matters, but the DIRECTION is accounted for. VOF and phase change are not ideally suited as you're tracking the free surface. Either you will want a very fine mesh, or review the multifluid VOF which is part of the Eulerian multiphase model. Note, check 2022R1 and in about 5-6 months 2022R2 as the advice may change.
    • Hassan_Kashif
      Subscriber
      Hi Rob,
      I appreciate the reply.
      I selected the VoF model since most of the literature I found used this model over the Eulerian model due to simplicity and time efficiency compared to the Eulerian multiphase.
      I have a question regarding what you said about the evaporation condensation model, could you clarify what you mean by ORDER and DIRECTION? My understanding from what you said is that the order of my "to" and "from" phase matters, so for condensation I should state "Water vapor" to "Water liquid" and investigate the "Latent Heat cannot be less than zero" warning.
      I also have not enabled the Eulerian film modelling for the model shown as that would be part of my model development if any condensation of vapor occured.
      My mesh refinement is limited by the number of nodes I have also since I am currently using the student software.


    • Rob
      Ansys Employee
      Look in the documentation, regarding the to and from phases. For direction, the fluid moves from liquid to gas and gas to liquid. Both are covered in the model, in one case energy is absorbed into the fluid with no change in temperature, in the other energy is released.
    • DrAmine
      Ansys Employee
      Is the condensation taking place a the free surface or as in classic evaporator along the walls? For free surface condensation: Lee Model as it is is not enough. For wall driven condensation: I remember I had hard times doing things like this involving pure fluids.


    • Hassan_Kashif
      Subscriber
      Hi DrAmine,
      My project is concerned with "Atmospheric water harvesting using advanced adsorbent materials" so in essence I'm trying to model an AWH as a container filled with vapor where The walls of this container are below the dew point of the water vapor so condensation can occur at the wall. Once I can model condensate water forming at the walls I want to explore ways of improving condensate formation through surface optimisations.
      I too have been struggling with this wall driven condensation since November of last year but seemed to have gotten some minor results from an adjusted model on this YouTube tutorial I found:
      " target="blank">

      I changed the model from VoF to Eulerian multiphase as suggested by Rob:
      The geometry is a flat 2D Plate with the following conditions:
      2 Horizontal Walls - Original wall conditions (Material = Copper)
      Top wall - Condenser wall (T = 273 K and Material = Copper)
      Bottom wall - Boiler wall (T= 400 K and Material = Copper)
      I have also changed the initialisation conditions from the linked tutorial so that the fluid domain is filled with water vapor rather than water liquid so that I can immediately and more easily observe condensation at the condenser surface. I have patched the temperature of the liquid and vapor phase to 300 K. However I obtain these large values of temperature in my transient run at time step = 0.0005 with 2500 iterations:
      I don't understand why the temperature is so high for water.l even though there is no water present in this zone:
      As you can see from the above image I'm getting very minor levels of water condensate forming at the edges however surprisingly more is present at the bottom edge (boiler boundary condition) than the top edge (condenser boundary condition)
      Amount of condensate present at lower timesteps increases also but then decreases. Most condensate is present around the time step shown below:
      This is the corresponding temperature plot:
      When running this model under VoF I get a floating point exception error at around 400 iterations due to divergence in continuity equation. I am unsure what is wrong with the set up I have shown you screen shots of and have not yet been able to translate any condensation in a 3D domain which is my goal.
      Sorry for the long post. Any insight into wall driven condensation will be greatly appreciated.


Viewing 5 reply threads
  • You must be logged in to reply to this topic.