Theoretical question about nonlinear analysis for diagonal steel braces

[milkshakelake]

Is it possible to design a building using the axial compression/tension limits of braces? Let's say a bracing can take 100 kips. So I would put a hard limit of 100 kips axial tension/compression in software (I use ETABS). This would use a nonlinear analysis. When this limit is reached, the load would distribute to other braces. My thinking is that when a brace is overloaded and starts to fail, the load would naturally go to the next stiffest member.

I wouldn't do this kind of analysis for beams or columns because of several reasons. I think the reason for a column is obvious, but for beams, a deck failure would instigate before loads can be redistributed. But diagonal braces are fundamentally different in this regard. I see two huge downsides to this method:

1. Nobody uses this method.

2. Hand calculations to check the analysis are basically impossible.

 

[Settingsun]

Compression braces are questionable as they're usually slender. That aside, pay attention to the connections. You want them reliably stronger than the member.

 

[milkshakelake]

The method I'm thinking of would definitely exceed yield strength of some members. So it would violate the safe design theorem. Thanks for the input!

 

[Settingsun]

Yielding isn't exceeding yield - it's equalling yield strength and is part of the ductile redistribution. And you usually have the capacity reduction factor so you don't exceed yield in the internal stress distribution that's in equilibrium with the external loads. I think you could do what you're saying for tension braces in principle.

Seismic is not covered by the principle I posted. It's for static loading.

 

[dik]

Yielding can occur and the structure can still be stable and sustain additional loading... by redistribution, until a mechanism is formed.

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

-Dik

 

[Tomfh]

This is a plastic method of design. Some software packages will allow you to model members in this way.

Redistribution is also allowed for beams and columns in concrete and steel structures.

 

[bones206]

RISA has a Euler buckling setting that may be useful for this type of analysis:

https://risa.com/post/how-to-model-tension-only-bracing-in-risa-3d

 

[JoshPlumSE]

The "euler buckling" member in RISA is meant to because as a tension only member. However, it allows some compression in order to reduce the number of iterations required to converge the analysis. But, in it's core, it's intended to be a tension only member.

 

[milkshakelake]

Thanks all for the input. I don't use RISA but that's good to know. ETABS has tension and compression limits. I'll just leave compression limit at 0.

Seismic is not covered by the principle I posted. It's for static loading.

Why would this only apply to static loading? There are cases where tension straps or cables are used for lateral forces.

Redistribution is also allowed for beams and columns in concrete and steel structures.

For columns, wouldn't they buckle at some point and violate the theorem steveh49 posted? It would be of limited usefulness economically because there would be redundant columns, unless redundancy was the objective.

 

[Settingsun]

I guess you could consider it applying to EQ in a sense, but EQ doesn't apply a load but actually a motion. The internal stresses depend on the capacity so there isn't a single external set of forces to be in equilibrium with.

Tension straps aren't fantastic for EQ because they only cycle in one direction. They stretch, stretch again, and again without compressing in between.