indoor simulation convergence problem
 

Hi, guys.

Now, I'm conducting the indoor air simulation.
As an initial simulation model, I'm analyzing the empty room.
Room size is 5.62 (m) x 5.62 (m) x 3 (m). The dimension of inlet is 0.6 (m) x 2.7 (m), and outlet is 0.6 (m) x 0.8 (m). In my simulation, the RNG K-e turbulence model and enhanced wall treatment model were used.
About 2 milion hexahedral mesh was used, and the mesh quality is enough.
The air was supplied 0.31 m/s velocity.

Because the indoor situation of simulation is very simple, I expected to converge easily. However, the residual does not go down e-3 even though the iteration is about 500 times. Because it is empty room in the simulation, it is enough to go down to e-5 about 500 times, but I don't know why it is hard to reach the convergence.Is the size of the inlet too large?? I would appreciate your opinion. Thank you.