Inlet Boundary condition
 

Dear friends:
what is the meaning of "flow enters into the duct from all directions".
for the duct flow, we have inlet bc
regards

for eg.the inlet to draft fans in power plant..It is open to atmosphere and the flow enters from all directions into the suction chamber..

for the inlet boundary condition, we must specify the magnitude and the direction of the speed.

If I understand correctly, you want to represent the flow dynamics at the duct inlet accurately. In that case, why don't you extend the computational domain little bit upstream of the duct and use the actual atmospheric/free conditions for inlet? Trying to formulate an accurate inlet is quite troublesome. It is better to keep this menace away from the physical problem of interest.

The best practice would be to make the duct inlet submerged in the flow. Have considerable grid points upstream and also a few grid points downstream of inlet. A similar approach applies to outlet as well.

Hope it helps.