Topics relate to Fluent, CFX, Turbogrid and more

Chillers on Rooftop – Unrealistic Results – Question about Boundary Conditions

    • onlycfd

      Hello Fluent Experts,

      Hope you are all doing great.

      Would very much appreciate your thoughts and guidance on a technical challenge which I have been facing for quite some time.

      A brief writeup on what is being performed in the simulation.


      We have two chillers on a roof top. Each chiller has got 3 independent V circuits. We modeled two chillers as there is an uneven loading on the V circuits of the chillers (Not modeling the tubes and water flow inside. Taking into account only the air side pressure losses across the V coils)

      Axial fans are modeled with the blade topography. Frame motion is assigned to the fan rotating domain, as well as to the blades of the fans. There is a mesh interface between rotating domains and stationary atmosphere region.


      So, we assigned the loss coefficients to the porous coils, rpm to the rotating fan domains.

      Mesh Orthogonal Quality is at 0.2

      Have assigned pressure inlet on the sides (equal to 0) and pressure outlet at the top (equal to 0), trying to replicate stationary air at the far regions. We have data only for the RPM of the fans.


      What is physically expected is due to the suction created by the rotating fans, atmosphere is sucked over the coils of the V, and pushed out in the upward direction to the atmosphere.


      But that is not what we are seeing in the results.


      First the velocity plot and the pressure plot do not look realistic. We see a high-pressure region till the far end of the outlet, which in turn is impacting the velocity field.


      Would appreciate your guidance and thoughts as to what we are missing in the simulation model, like any corrections to be made w.r.t to the boundary conditions setup, etc.


      Thanks a lot for your kind guidance and help


    • Federico Alzamora Previtali


      I would suggest making a couple of velocity vector plots to see how the air is flowing through the coils and fans. You could make these in the plane through mid fans, but also through a plane perpendicular to this (cutting mid-way through 3 fans). Finally, a vector plot normal to the axial direction of the fans to ensure that the frame motion is behaving as expected.

    • Rob
      Ansys Employee


      I suspect I know the reason: been there, made the same mistake. What is the operating density that you’ve set, what is the air density definition and please post a velocity vector plot (in plane) same surface as the others. 

      Nice to see someone modelling the baby fans, if fed properly they can grow...  https://www.lngindustry.com/magazine/lng-industry/march-2011/ 


    • onlycfd

      Thank You Thank You very much for your guidance and attention. 

      I am attaching below few images as requested. Still raking brains to get it going. 





    • Rob
      Ansys Employee

      OK, it's the operating density that's causing the problem. Look in https://ansyshelp.ansys.com/account/Secured?returnurl=/Views/Secured/corp/v231/en/flu_ug/flu_ug_bcs_sec_operating.html  Section  

      Assuming you're not using high speed fans (ie over about 0.3M) use incompressible ideal gas as you don't need to worry about pressure effects: read up on why! If you look at the equation why do you think the flow is a bit odd? 

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