B.C.'s for Fan/Compressor
 

In general I have created performance maps instead of trying to gather data for direct use in specific application. But right now I am at a position where I have no volume/mass flow rate data I can use or decide on, just a fan and its geometry.
I do know that it is pulling from atmosphere, and pushing to atmosphere pressure almost immediately after exit (blowing through battery case).
Right now the B.C.'s I've chosen are total stagnation perssure inlet=0 and entrainment outlet at P=0. I don't know if this is approrpiate since total pressure inlet should include a velocity term and this only drops to 0 some large distance from the inlet. And the need for entrainment at the exit is because there is some recirculation between the widely spaced blades.
What are boundary conditions appropriate to use here to determine the flow rate and thrust load resultant when no flow rate at a face is provided? Static pressure at either end is not robust even with zero gradient choices.

To put static pressure at inlet, you need some clever strategy. For this you should place inlet far away (30-60 chord) and specify 101325 as total and static pressure (that is zero velocity) and fan will determine its flow rate by sucking air for different rpms. Vary static pressure at outlet to make the performance map for particular rpm

That sounds similar to what I have done. Added an inlet and outlet extension. I know it always vents to atmosphere since its just a fan.

For the inlet I used Stationary Frame Total Pressure and set it to 0 at a good distance from the Fan domain. However, I doubt it's 30-60 chord lengths, so I may have to extend it further. I also let it suck air in not only from the "cylinder end" of this extension, but also from the "cylinder walls" perpendicular to the inlet interface at the fan domain. Is this a good idea? It seemed physically reasonable as air can come in from any and all directions when a fan is blowing.

We also added a bellmouth to ease the flow...

http://imageshack.us/a/img833/2074/fanog.png

Abstract
http://arc.aiaa.org/doi/abs/10.2514/6.2007-23

Full length paper sent via pm

Wow, this is a great visual, thanks! I'll read up on choice of this inlet as well.

There is paper in J. turbomachinery where they added nozzle at outlet for each point on map. They just specified 101325 pressure at outlet as well and changed nozzle area to get another point on map. They have also shown that by this method (adding nozzle at outlet) , stall/surge can also be simulated.

Link is here :

http://turbomachinery.asmedigitalcol...icleid=1467000

But the inlet was not as ours, it was a simple extension as you have mentioned in previous post. They proposed that adding large domain (like ours) will better simulate the flow in turbomachinery.