small steel canopy stability?

To make the structure stable, the two column-to-beam connections would need to be moment connections or knee braces would be necessary. Functionally, the structure will likely work considering the connections to the building. At worst, you might get racking in the direction parallel to the building entrance if the wind loads were high enough.

I know you said the columns are pinned at the base, but can the two posts be checked as flagpoles (cantilevered) with a fixed base if the connection allows? If they are stiff enough, with such a small area for wind loading, the deflection may be within acceptable limits.

If i look at the Photo i would say...no need to worry

but next time...go get a structural engenner to do the design

I'm with Klaus, the welded roof frame would be fairly stiff acting as a vierendeel frame back to the connections to the building, so recheck it as such inclusive of any loads it transfers through the connections to the building. I wouldn't think you would need the cross bracing as shown in the model (if that is what it is).

One thing to note would be that if the columns were also welded into a portal then the portal frame would need to be capable of accommodating any lateral drift from wind or seismic of the main structure at the point of connection (its goes along for the ride if you like). I assume they are pinned top and bottom, you note bolted, but not clear if the knee connection was intended to be a pin or more fixed as details of how it is bolted are not provided. If the details are such that it's pinned top and bottom then they just tilt over to accommodate this relative movement between the roof and ground and you don't have to worry about the deformation compatibility.

n3jc said:

In a calc. model Id use all connections as pinned which means structure is not horizontally stable. But since forces will be small is this the right approach?

Click to expand...

This is not correct, you need to model it so the connections provided represent the model and that model/structure is stable, you note you are modelling it all as pinned but its unstable. Model should need to be stable and have sufficient degrees of freedom restrained, I suspect its only solving because of the bracing that is present, but the final structure doesn't have this provided. As noted above the fact that it has now been welded would negate the need for any bracing to make it stable.