Closed Box Fluid filled at Increasing Temperature Lose Mass – Ideal Gas Law

    • Andrea

      Hi everyone,

      I'm simulating a closed aluminum box filled with air, 2D setup.

      I increase the temperature of the air inside the box and given that the volume is closed I would like to observe increasing temperature and pressure at constant density.

      What actually happens is that the temperature rises but the weighted-average-area density (and the mass as consequence) of the fluid drops. Slightly changing the setup I can observe an increase in pressure but still a considerable density (mass) reduction.

      The ways I tried heating the gas inside are the following(used every one alone):

      • heat source in the gas;

      • heat source in the solid (thick box surrounding the fluid);

      • heat flux across the interface solid-fluid;

      • convection of heat from the outside of the box enclosure.


      • pressure-based/transient/velocity formulation:absolute/Planar

      • viscous model SST k-omega(to simulate natural convection inside the box)/energy: on/ S2S radiation

      • BC: in a compatible way with the above heating "strategies" I set: thermal coupling in the solid-fluid interface, adiabatic condition in the outside wall of the enclosure

      • IC: (standard) T=273K, operating pressure=1bar, turbulent kinetic energy and specific dissipation rate: 1

      • air density: ideal gas law

      • The mesh is done with quads, bi-exponential growth ratio of 1.15 both in solid and fluid, so close to the wall I have a nice boundary layer


      • scheme: coupled [also tried Simple]

      • gradient: green-gauss

      • Pressure: PRESTO! [also tried Second order]

      • others: first order upwind

      IN the attached picture the heat sources is located in the fluid.

      I have no Idea why I'm losing mass (not even displayed in the mass balance flux in the postprocessing section).

      Any guess, advice is very welcome! It's so frustrating not be able to verify the IGL with fluent.

      PS: I was long, but I hope exaustive 

    • DrAmine
      Ansys Employee

      Switch on floating operating pressure and use a sensible time step size like sqrt of a minimum of cell thickness divided by a thermal gravity. Thermal gravity is isobaric thermal expansion times maximum temperature differences times the earth acc. 

    • Andrea

      Thank you very much Amine!

      It works perfectly with the ideal gas! Unfortunately the checkbox for the Floating Operating Pressure disappears using a real gas from the NIST library.

      Do you know by chance how to deal with this?

      Thanks again for the professionality!  

    • DrAmine
      Ansys Employee

      It would require some UDF programming to mimic Fluent behavior when working with real gas: you need to adjust the operating pressure every iteration like Fluent.  

    • Andrea

      Ok, so in this case I would get the density properly computed with the real gas model and I could adjust the operating pressure with the ideal gas law for example.

      It seems quite challenging but I will try for sure. Thank you again!

      Best regards,


Viewing 4 reply threads
  • You must be logged in to reply to this topic.