[ZoeSydney]

Hi there

I've simulated Transient Boussinesq laminar natural convection with a 2D rectangular cavity with air as the working fluid. The hot wall is set to be at a constant temperature 295 K and ambient temperature is at 293 K. The results are in great agreement with theoretical values. The flow is buoyancy-driven.

Now I am trying to set density, thermal conductivity and dynamic viscosity as functions of temperature with everything else as the same as the Boussinesq simulation. (Specific heat is constant =1006.24 j/kg-k). But there are negative y velocities in the domain with an increasing magnitude as time marching forward.

Please find the detailed setup in the attachment.detailed questions.docx

Could anyone please let me know how to get rid of the negative y velocities?
Thanks and regards,
Zoe

[ZoeSydney]

Any thoughts?