AEC Application, Diffuser coanda effect

[Anonymous]

Dear Forum

I am simulation a small office with solar radiation, two 4 ways diffusers are used to condition the space, the simulation run well and converge, but i  notice the resultant is not like i expect, i don't see the coanda effect typical of this terminals.

I have tried with k epsilont turbulent model and mix model and the results are what i want, y also the expected temperature is what i expect according thermal calculations, but the airflow is falling down not horizontal near the air outlet.

Trying with SST-Hellsten or the SST-Smirnov-Menter the simulation fail because excited.

I think the mesh is enough to produce valid result because i have tried different scenarios with different sizes and the result are similar, even i have tried improving the boundary layer.

Do you have some idea how to improve the coanda effect in diffusers?

Note: I have checked  previous post about the topic.

Thanks.

[David.Short.]

Hi @Anonymous,

Thanks for reaching out on the forum. You should be able to capture the Coanda effect using k-omega and ADV 5 however it requires very fine and uniform mesh on the diffuser surfaces and also along the ceiling. A good technique is using a refinement refine incorporating the diffuser and the full area of ceiling on which the Coanda is expected.

Give that a go and let me know how you get on!

All the best,

David