[fan123]

Hi everyone,

For my thesis i have to investigate how well a specific CFD simulation matches reality. The simulation is done with UG/NX 7.5. The problem is the cfd simulation underpredicts the reality quite a lot! I have tried everything to improve the simulation but the simulation never comes close to the reality. So i decided to ask my question here, i hope someone can help me.

So its a simulation of an axial fan (metal sheet blades, 1mm thickness).
The K-epsilon model is used (recommended and it gives the best results).
The fan has a max diameter of 1,2m. The mesh around the fan is unstructured. The mesh size near the blade is 4-6mm, y+ is between 11-200.
To simulate the rotation a rotating frame of reference is used, rotational speed 700 /min.

I hope somene can help me.
Many thanks in advance!

[fan123]

i forgot to say the fan is simulated in a square wind tunnel, 10,5 m long, wich is located in a room, no opening / inlet / outlet BC's are used

[fan123]

How good can CFD normally predict a fan curve?