steel portal frame - roof bracing question 1

[greznik91]

I apologize for poor English.

QUESTION: Im wondering is it necessary to connect roof bracing at the top of a ridge - apex haunch?

I have seen some portals with bracing all the way to the top, and some that ends a bit lower.

I think when bracing is not connected all the way to the top that there is some bending moment (around minor axis) of a primary steel beam (IPE 270)in order to transfer forces...
I think there should be at least an additional hollow section connection at the ridge (yellow in image). What do you think about this?

 

[klaus.]

Basically many different Systems are possible......

 

[jayrod12]

I don't feel as though there will be any appreciable minor axis bending due to the roof purlins.

 

[greznik91]

I don't feel as though there will be any appreciable minor axis bending due to the roof purlins.

yeah, but there are not at the very top - ridge

 

[Ron247]

Following your entire load path while considering the "movement" of the roof will give you a better idea of where you need rods. The two most common configurations are 2-sided where you have rods on each sidewall and 1-sided where you have rods on only one sidewall. 1-sided is generally used on smaller width buildings say 60' or less. 1-sided also loads the endwall bracing on both ends of the building. In 2 sided you may or may not need rods at the ridge. The movement somewhat starts at/near the ridge and proceeds in both directions to the foundation. In 1-sided you generally must have rods at the ridge. The movement starts at the wall with no rods, and proceeds across the entire roof to the other side. Therefore, it cannot have a break in the load path.

For 2-sided, you also tend to consider where your endwall posts are at. The wind blows on the wall panels, that are supported by the wall girts, that are supported by the endwall columns. The force from the top of the endwall column is what pushes on the roof and makes it want to move. There is some minor loading from the upper wall panel where it meets the gable but most load is via the column connections. If there is an endwall column right at the ridge as shown in your 2nd picture, the roof movement starts at the ridge. If the column is not aligned with the ridge as shown in your last picture, you can start the bracing where it is.

My guess is that the hollow section you show is a strut that are sometimes used on larger buildings. When the roof purlins near the rods can "strut" the load across the bay to the other frame, struts are generally not used. In those cases, the rods tend to align close to the purlins.

The entire standard rod bracing system is a truss. The 2 rigid frames the rods connect to are the top and bottom chords. The Z-purlins near the rods are the verticals that connect the chords. Half of the rods are the diagonals that stabilize the truss from racking. Only visualize the rods that are in tension. The half of the rods in compression are along for the ride until the wind or seismic acts in the opposite direction. The tops of the endwall columns are the point loads onto your truss chord.

 

[jayrod12]

yeah, but there are not at the very top - ridge

But based on your picture, they align with the bracing connections. So again, you've got the roof purlins.

 

[JAE]

I think the first photo is a condition that "may" be a problem.

If you think about the roof diaphragm X-bracing as a truss, then the main roof girders are the chords of that truss.
If the rod bracing does not align with the ridge, where you essentially have a kink in your chord members, you will induce major axis bending in the beam/chords...not minor axis bending.

This bending amount may be manageable, and that is what I suppose the designer had in mind. You just need to check the beams for that combined gravity + wind + kink-induced bending.

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[Ron247]

I can't tell by the 1st picture if this is a 1-sided or 2-sided brace system. I see rods on the right wall but can't tell about the left wall. If it is 1-sided, that would explain the center rods that actually span across the ridge rather than joint on the ridge.

This is the width building that sometimes use 1-sided bracing. Due to its narrow width, the rod size would not tend to indicate which direction the brace rods are "accumulating" load since the rod load at one side is not much different than the other side.

 

[greznik91]

I have better understanding now.

and @Ron247 great post! tnx