Fluids

Fluids

Natural Convection in a Water Glass

    • Devanshi
      Subscriber

      Hello, I am trying to compute the flow pattern and temperature distribution in a glass of water using 2D axisymmetric non isothermal analysis. The transient heat transfer occurs through convection and conduction. Glass is of finite thickness of silica glass with a specified thermal conductivity. The water and the glass both are initially at 5°C and the ambient temperature is 25°C. I have set the following boundary conditions -


      1. Bottom wall - constant temperature of 25°C


      2. Outside wall - convective heat flux boundary condition. ( coefficient of heat transfer = 2)


      3. centerline - axial symmetry


      I created a coupled wall mesh interface between the water and the glass as i want to observe the temperature profile in glass wall too.


      I am facing problem defining the boundary condition for the free surface. Under material options in thermal conditions, i am only able to select solid materials for my water air interface. How can i rectify this?


      I assume this might be the reason for divergence in the simulation. Showing floating point exception error. 




       


       


       

    • raul.raghav
      Subscriber

      The glass separates water and air. The setting doesn't look wrong to me except that you mention a "finite thickness" in your statement and your wall thickness is set to 0? I don't think this is the reason for divergence of your solution.

    • Rob
      Ansys Employee

      Which density method are you using for water? If it's a sealed container and density changes the volume needs to go somewhere. Have you set the x-axis as the vessel centre line? 

    • Devanshi
      Subscriber

      The top surface is water air interface i haven't covered it with solid, i am not able to select fluid as the material for convection through top interface. For convection through the side of the glass, do i have to specify the wall thickness even though i have created a solid domain in cell conditions? 

    • Devanshi
      Subscriber

      Hi rwoolhou,


      I have used Boussinesq approximation method since i am considering density changes small. It's an open container, I hope I am not wrong in assuming negligible volume change as density change is small?
      I wasn't aware fluent rotates axisymmetric models only about x-axis. I'll correct that and see the result.

    • Rob
      Ansys Employee

      The top surface will be a wall, and that wall must be a solid material. However, if you set as zero thickness the material property will have no effect. As it's a free surface I'd suggest setting it as a slip wall. 


      The glass bound can be a thin wall, but you then need to assign it a thickness. I can't see from the image, and as staff cannot download files: we rely on Community volunteers like Rahul for that. 

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