[covertdh]

Howdy,


I am attempting to model the flow inside of a closed circuit wind tunnel that is powered by a fan (as apart of my design process) and I am having some difficulties. I want to model the entire wind tunnel and not just the test section. I have tried a multitude of things and I am now seeking guidance.


I have tried making a cut in the fluid volume to apply a mass flow rate inlet condition and mass flow rate outlet condition. This seemed to get decent results however it did not seem that inlet and outlet shared the same state properties.



To remedy this I tried to instate a periodic boundary condition where I made a cut in the volume and specified a mass flow rate. However, this mass flow rate seemed to only be applied for the initialization of the CFD simulation. Nothing was keeping the air circulating and the solution did not seem to make sense.


Next I rejoined the volume (making it fully enclosed again) and tried fixing the velocity values where the fan would be and this worked well for the inviscid solution, however, when I applied a single equation turbulence model the solution would blow up. When initializing with no inlet or outlets the hybrid initialization just makes the initialization 0 m/s everywhere and and 0 gauge pressure. So when fixing the values in the fan chamber there is a velocity gradient that causes the solution to blow up.


Next I tried using the fan boundary condition on an internal face which seemed to work okay, however, the company that makes the fan we have has pretty sparse information on the fan so this has proven to not be so accurate.


Lastly, I have tried adding a momentum source to where the fan would be. I calculated the values based on mass flow rate * the velocity in the fan zone / the volume of the cell fan zone. This gives the units of N/m^3. When doing this I am not getting the expected test section velocities and I am at loss of what to do. I have tried everything that I could think of so any guidance would be greatly appreciated.