help needed - wind loads on football shaped building

huangyhg

help needed - wind loads on "football" shaped building
hi all.  i am starting schematic design concepts on a building which has a "football" shaped glass atrium, about 90 feet tall, 185 long and 75 wide.  by this i mean if you took a section cut long ways through a football, you end up with a section which curves at constant radius (about 135'), mirrored about itself, sort of like the cross section of an airplane wing.  the atrium cuts through the middle of an office building, which is narrower and shorter than the atrium, but much longer.  so the atrium will see wind loads from all directions, to some extent, less whatever is shielded by the main structure.
i would like to hear what others think about figuring out how the appropriate wind loads (125 mph hurricane zone) would be calculated, for both c&c and mwfrs.  there are no seismic requirements.
when wind hits perpendicular to wide direction, i would figure its conservative to treat like flat wall, and apply pressures normal to surface, ignoring curvature.  this may be critical load case for c&c loads for glass panels and out of plane bending of the vertical supports.
but when wind hits perpendicular to narrow direction, it can simply deflect to one side or the other, and the building will cut through the air much like an airplane wing.  i am thinking right now about using the equations for round walls (like on domed structures) from asce7 to calculate the drag force, which i can then apply uniformly for design of mwfrs.  but for c&c loading of glass panels and vertical structure out of plane, not sure what to think.  if you look at domed roofs, they end up with suction over the entire roof.  i wonder if that is what i would end up with here too.  
i have already suggested we recommend a wind tunnel test based on the complex geometry, but my senior engineer doesn't think it is waranted or cost effective.  
any help any of you can give is much appreciated.
thanks.

wind tunnel sounds good to me....  be cheaper than a collapsed bulding...unless you want to figure some emperical loads and then double or triple them..
i did one of these - what a dimensional nightmare.
i figured it was simlar to a barrel roof (though the radius of the barrel is varying). it seems to me that you can come up with some conservative wind estimates without spending $$ on the wind tunnel.  
i would treat the wide direction like a standard wall (like you said).  i would treat the narrow direction like a domed roof and would argue that having predominantly suction is the right way to go.  this shape seems similar to wind running along a gabled roof (and the pitch does not matter), which has predominantly suction as well.
gabled roof/barrel roof:  that's sort off what i was thinking as well.  
some of the tables for gable roofs have the leeward roof pressure being higher than the windward pressure. this would result in a net drag in the direction of the wind.  which would basically follow how i feel it should come out.
i may still try for the wind tunnel test.  need to try to find out costs, as the site around the building is not very busy.  so modeling the site may not take too much.  much less than a downtown high rise anyway.  and if we can save 5-10 psf on the glass, it may pay for itself anyway.