Interesting Roof Structure 1

[phamENG]

Reading through the news this morning and came across an article with a picture of an interesting roof. It's a grain storage building in Ukraine. I'll let you guess why there's a large hole in the roof.

But I'm interested in the framing. Looks like a castellated steel rafter system with a turnbuckle tension tie. Has anyone seen one of these before? Does it have a name? Any comments on the possible merits or faults with it?

 

[dik]

Looks great... first time I've seen one like this

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Do you feel any better?

-Dik

 

[phamENG]

I suppose we could have a discussion on castellated and cellular beams in general. I brought up the steel joist supply chain issues a while back - at what point can castellated and cellular start taking a bit of market share? They aren't complicated, and for a decent size fab shop with the right tools it seems like it would be a pretty straightforward process. They're still not quite as efficient as an OWSJ, but they do better than a sold wide flange joist system in terms of material usage and I imagine the lead times wouldn't be nearly as long.

 

[dik]

I think that in 50+ years I've only desiged 2 or 3 of them... not common.

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Do you feel any better?

-Dik

 

[phamENG]

dik - I agree. O'Hare International in Chicago is a pretty fantastic example of cellular beams. But they aren't used in the States very much anymore - I've always contributed it to the dominance of efficient and readily available OWSJ systems and the lack of worthwhile difference in cost between material for whole wide flange sections and material+labor for castellated/cellular beams. But with the bottlenecks so many are seeing, and long delays, and sky-high material costs, what's the tipping point? Is there a sweet spot where LB and CB sections start popping up more often? There are a handful of companies pushing them - like this one out of North Carolina, for example - but they still aren't common.

 

[KootK]

1) She gets some points for surviving substantially intact.

2) I'd call it a tied rafter roof, like we do in wood.

3) Do we think that the ties terminate at the walls or pass through them to hit the rafters concentrically? It's difficult for me to tell from the photos. Pass through would be niftier structurally but, obviously, a problem for envelope.

4) There will be axial demand in the castellated beams so something will need to brace those for both weak axis and lateral torsional buckling. That "something" could be either:

a) The roof panels themselves if those are strong and stiff enough to do that job and the connection detailing is suitable.

b) If I'm not mistaken, there's a stringback and/or bridging system in there at intervals.

5) It's a shallow roof and a pretty long span. unless it doesn't snow in the Ukraine, I doubt this quite works numerically. My money's on strong axis buckling under an unbalanced load case.

 

[KootK]

O'Hare International in Chicago is a pretty fantastic example of cellular beams.

O'Hare is fantastic but I'd have to think that choice was more about aesthetics than economy.

Photo for those who've not had 4 zillion connection flights through ORD. Mid way always felt like a large bus station in comparison.

 

[Brad805]

The castellated beams produce a nice looking structure, but I have never quite followed what is done with the waste? The pre-eng market succeeds by reducing waste by processing large quantities of orders where they can optimize how the shapes are cut from the plate steel. Here you end up with a ton of little circles??? Other than that, it is just another piece of steel. I would expect where one locates all the ties would be important in your example. I could see a lot of back and forth with the manufacturer unless you had a clear understanding of how they layout their openings.

 

[KootK]

Here you end up with a ton of little circles???

It's much cleverer than that for a true castellated beam (not the O'Hare example). They make the zig zag cut on the beam and then take the two beam halves and shift them over to produce a deeper section. It's pretty slick.

 

[XR250]

I

It's a shallow roof and a pretty long span. unless it doesn't snow in the Ukraine, I doubt this quite works numerically. My money's on strong axis buckling under an unbalanced load case.

I'm going with lateral buckling of the bottom flange

 

[KootK]

 

[phamENG]

2) I'd call it a tied rafter roof, like we do in wood.

3) Do we think that the ties terminate at the walls or pass through them to hit the rafters concentrically? It's difficult for me to tell from the photos. Pass through would be niftier structurally but, obviously, a problem for envelope.

4) There will be axial demand in the castellated beams so something will need to brace those for both weak axis and lateral torsional buckling. That "something" could be either:

a) The roof panels themselves if those are strong and stiff enough to do that job and the connection detailing is suitable.

b) If I'm not mistaken, there's a stringback[sic] and/or bridging system in there at intervals.

5) It's a shallow roof and a pretty long span. unless it doesn't snow in the Ukraine, I doubt this quite works numerically. My money's on strong axis buckling under an unbalanced load case.

2) I agree fundamentally. Curious if it's a proprietary system or a regional public domain framing system.
3) Looks concentric with the rafters, but I doubt they go all the way through. Maybe a tab off the bottom flange with a clevis on the tie?
4) I think it's both A and B.
5) If this website is to be believed, Kyiv gets some snow but not a lot, relatively speaking. Even in the snowiest month, it's not going to be significantly more than a typical roof live load. Unbalanced could do unpleasant things, but the ridge looks like an extended end plate moment connection which likely helps handle some of that.

I agree that O'Hare was for looks and not to save money. It's Chicago. But is there a point where it makes economic sense to use them in a more utilitarian fashion?

 

[KootK]

I'm going with lateral buckling of the bottom flange

That's braced every few feet though, right?

 

[phamENG]

Brad - the cellular beams are similar.

 

[KootK]

But is there a point where it makes economic sense to use them in a more utilitarian fashion?

For sure there is such a point. If I'm not mistaken, these thing originated during wartime when material was at a premium and labor was less so. That said, I don't understand the intricacies of the current market forces well enough to be able to say whether or not that time is now.

When I've used castellated beams in the past, it's been as the roof for pool structures. Painting a castellated beam is much easier than painting an OWSJ so, in those cases, that was the source of the economic logic.

 

[WinelandV]

A manufacturing company that we work with produces this fun bride for access in ag facilities out of a W14x22, "increased" to 20" deep with castellations.

The welded seam down the middle isn't shown, but mentally input it, and you can see the zig-zag shape that Koot mentioned.

 

[XR250]

That's braced every few feet though, right?

Ahh, yes, now I can see that. I wonder what that is kicked off to?
If it is braced properly, then, yes, I agree with you.

 

[Brad805]

Koot, I had always thought these were plate steel and not WF. Interesting. I have priced them before, and they never worked out, plus supply in Canada was terrible. They sure do suit the fancy of many archs for open ceilings.

 

[KootK]

Ahh, yes, now I can see that. I wonder what that is kicked off to?

If it's bridging and the bridging itself is tied off to something, all good. That's not the only option though. When I look at the detailing up close, I suspect that it's not actually bridging but, rather, a strongback system that would rotationally brace the beams without the need for bracing of the bracing. It's difficult to say with any certainty from just the one photo of course.

 

[phamENG]

If it's a concrete gable end, it may just die into the wall. The wall probably has enough stiffness and strength to resist the brace force.