[fanaticreader95]

I am simulating a data center in steady state. The server racks are modelled using the equilibrium porous media model(I intend to change this to non-equilibrium as I believe that would make more physical sense). Each rack has a heat source inside of it. I am using a conformal mesh for the hot aisle and server racks. There is an internal fan boundary placed at the interface of the server racks and the hot aisles. The characteristic fan curve was obtained experimentally and is being modelled as a 5th order polynomial. There are 3 velocity inlets each 1.2 m wide with an inlet velocity of 1.3 m/s. Between the inlet and the main room there is a porous jump boundary. Similarly at the vent entrances there are also porous jump boundaries. I am using an outflow boundary condition. The fluid I am using is air and I am utilizing the boussinesq approximation. The energy equation is being used and I am using the Standard k-epsilon model and including effects from buoyancy and turbulent dissipation. When running the simulation I have my results converge but get very strange results with regards to the static and total pressure. I don't have a photo of the static pressure contour at this time(will get one in a few days) but it effectively looks like a uniform contour that is constant through the central room yet has a range of 2 or 3 atm.

[LuckyTran]

Looks fine

[fanaticreader95]

The only other thing I am curious about is to whether it would make sense to use a non-equilibrium porous media model or an equilibrium porous media model? I think both could make sense depending on the circumstances.