Fig 6-9, ASCE 7-02

Thats what I do. But you bring up a good point. I think you would not have to. It makes it easy to do in a model since the roof loads are generally combined in the same loads case as the Wind/Leeward walls. But I think that maybe overly conservative....

Is it conservative?

The 0.75 case is (I believe)for wind at 45 degrees to the walls(each component is 1/sqrt(2) which is 0.71 - rounded up to 0.75.

If you have this 45 degree wind then you have a 45 degree component acting on the eaves of both walls. Do your force vectors and see what you get.

By the way, they never take this into diagonal wind into account in Australia, and buildings dont seem to fall down because of it (they have winds up to 140mph too).

csd72 is correct....I'm not sure why you would just say - we don't need to do that case when it is clearly a required condition.

I am little confused by the question. The case is based on wind blowing at a diagonal to the building. It is shown broken down into components on the ASCE figure, so if you have a flat roof, you have to look at each component separately for suction on the roof.

For example, if the specified wind load is 20psf, that is 15 psf in the X direction and 15 psf in the Y direction, but I wouldn't add those to get 30 psf to design the roof by. Along the diagonal, the load would be sqrt(15^2+15^2)=21.21 psf required for design.

I'm not saying that we don't need to do this case, I am just questioning its application.

Haynewp, you stated with great clarity what I was trying to get at, but not very clearly. For a flat or nearly flat roof, the resultant uplift pressure shouldn't change with changes in the direction of the wind. Let's say that when the wind blows in the X or Y direction the uplift pressure on the roof is 20 psf. If the wind blows on a 45 degree diagonal you have 0.707*20=14.14 psf components in the X and Y directions. Combining them as vectors, you have sqrt(14.14^2 + 14.14^2)=20 psf. When I was looking at the figure in ASCE 7 earlier, I interpreted it to mean that the components had to be added algebraically on the roof, and that didn't make sense to me.