Placement of Turning Vanes

AFAIK we always centered these.....

http://buildingengineer.wordpress.com/2009/12/11/turning-vanes-necessary-component-or-efficiency-reduction-device/

Most generally we usu square throat and square heel elbows

Now, how is Ethan Hunt supposed to crawl through the air ducts with those things in the way?

Well, he is small enough.

Still the same... centered on the corner....

The closer you get to a straight line, the less the friction loss. The larger the radius of bend, the better. The fewer the vanes, the better (less friction).

I vote for equally spaced across the elbow.

"As tight as possible into the heel" would be to laminate the first vane right to the outer metal of the elbow. Oops.

Same here. I've been inside the NASA 40x80 Wind tunnel and all the vanes in that thing are equally spaced. If it's good enough for rocket science then it's good enough for me.

Standing waves can induce turbulence so I'd say breaking the heel geometry with a vane there is a good idea. I would increase the gap linearly towards the other join. Testing this could be fun...

I can see putting at one of the vanes tight into the corner in order to eliminate the square corner and make the outside corner at least a smooth radius if you're talking about a square 90 degree duct corner. The rest should be equally spaced toward the inside corner and agree that the fewer the better.

I worked with an expert in aerodynamics on a vibration problem with a flue gas duct after the ID fan. It turned out that the vibration was caused by an imbalance in the fan but he told me that the start and finish of the vanes should be across the flow, that is the splitting should all take place at the same point in the flow; extending straight pieces from the tangent points achieved this. We didn't make this our practice because it takes more material and the design we were using worked satisfactorily.

I also say he's wrong. In fact it reduces turbulence and allows air to flow properly. The biggest advantage of turning vane is reduction in erosion of bend due to turning air velocity. If you provide 3 vanes , the erosion shall be due to only 1/4 of total air flow.