Structural Rod Bracing

[erictlee]

When designing structural rod bracing, I feel like it's necessary to determine what compressive load results in plastic deformation of the shape to verify the shape can elasticity recover in addition to the other bracing related checks. However, I do not see this done often. What am I missing or am I being too conservative?

 

[Agent666]

Can you provide an example calculation as I'm struggling to envisage exactly what you're checking. In compression, rod bracing is generally so slender it buckles out of the way with ease with no plastic deformation.

 

[r13]

Steel rod is a tension only member, it is usually considered ineffective in compression.

 

[KootK]

What am I missing or am I being too conservative?

1) While I understand your concern, I don't believe that this is something that other engineers are checking routinely. I wouldn't be inclined to worry about this.

2) Given that it'll be tension only bracing, as retired13 pointed out, somewhere in the system there should be some opposing rod bracing that goes into tension when the rod(s) that you're considering go into compression. So long as normal drift limits are maintained for those tension rod braced bays, the imposed displacement should be small enough that the compression rods can easily buckle out of the way without going plastic.

3) Interestingly, if you run the numbers, a tiny amount of imposed axial displacement can produce a shocking amount of associated lateral displacement. So, in that respect, I can see how this has piqued your interest. Check out this thread for something interesting and related that we dealt with a while back. In particular, check out the sketch included in the originating post.

4) I might share your concern if you had an oddball system with a tension only brace in one direction but, say, a flexible moment frame in the other. Then you might see some more significant axial displacement imposed on the compression rods. That said, in decades of structural engineering I've never come across such a thing and I feel that, in many respects, it would be a poor system choice to begin with. If such things exist at all, perhaps they come about as retrofits etc.

5) Any pretension in the tension rods would ease this a bit as the compression rods wouldn't actually see compression until later in the frame load history. I've specified this for cables in the past but not for rods. Other than a general turn-of-the-nut specification, I imagine that it's pretty tough to exert any meaningful control over the pretension in a rod brace.

6) Even if you plastify the rod in bending/buckling, I'm not sure that's actually a problem other than aesthetically. For a stocky section made of ductile stuff, you should ultimately be able "pull through" any locked in flexural stresses and still achieve the full tensile yield value as planned. In a way, the locked in buckling stresses might actually serve to stiffen the braced bay when the load reverses and puts the brace back into tension. Although I suppose that this would in imply permanent set in your fame drift when what you'd surely prefer is a full elastic return to the neutral position.

 

[r13]

As a practice, calculate slenderness ratio (L/r) of the rod, and find the allowable stress from AISC table, then compare it to Fy, or multiply it to the rod cross sectional area, at this point you should realize why a rod is ineffective under compression - as the rod will have buckled before reached yield. To effectively brace a joint with reversible load, or a sway frame, a minimum of two rods need to be used. In real practice, the length of rod should be adjustable, as it would have excessive elongation (due to tension) and deformation (kink out of plane).

 

[erictlee]

Thank you for the responses.

I am aware rods are extremely ineffective in compression. That's essentially the reason for the post. Rods were merely used in this post because of the extreme compressive inefficiency as well as the fact structural rod bracing will see a compressive load unlike a cable. Realistically, my inquiry extended to all inefficient axial compressive shapes (angles, rods, etc.). It's always bothered me because it seems odd to determine deflection of a single roof support part of a cladding system, but not the most inefficient use of a shape used for the lateral force resisting system. Especially, if that deflection caused permanent deformation.

Retired 13 and Kootk, your explanations make a lot of sense with the assertion typical deformations will not likely cause plasticity.

 

[Tomfh]

What’s the actual question here?

Rod and slender angles are tension only members, so there is no compression to speak of. They’re effectively cables, and go slack under compression.

 

[phamENG]

Tomfh - I think the question is this: what actually happens to the tension only members when they are no longer in tension? It's not a question of contributing to the strength of the structure when they are in compression, but rather ensuring that they will elastically return to their original state and perform as expected when the lateral load cycles back the other way.

It's something I've thought of, but never really explored. I've mostly worried about it in cases where a slender tension brace is concealed behind some sort of finish - when it buckles (because it will), what happens to that finish?

Thanks for asking the question, erictlee.

 

[r13]

I don't have example can back me up, but let's try to understand the phenomenon by observing the behavior of a member loaded in compression.

Due to material or load path imperfection, the member tends to deflect, and a moment is induced by p-delta effect. At small load, the deformed shape would be a smooth curve, and the member remains elastic, but when it is stressed beyond compression limit, large, localized deformation (a kink/buckle) at the point of weakness will occur, at this stage, compression area is lost/reduced, and the member will never returned to its original state, as the tension area would have stressed beyond yield by the enormous secondary moment induced by the sudden change in deflected shape (V vs smooth curve).

This phenomenon is difficult to quantify, but, through careful modelling and proper analysis techniques, FEM maybe able to provide insightful information on this problem.