Tension only bracing 8

[BST-98]

Good day,

My query is regarding designing bracing (cross bracing) as tension only members, mainly in low rise warehouses. Why is this the norm as bracing could be compressed leading the member to fail due to slenderness. Is it purely for economical efficiency?

Thanks for your time

 

[GeorgeTheCivilEngineer]

Yes economic efficiency - also you can use flat bars for tension bracing which doesn't get in the way as much.

Circular hollow sections are often used for compression-tension, but the connections are a bit more complex and the section itself costs more.

 

[StrEng007]

This got me thinking. For the assumption of tension only bracing, does anyone check their braces to make sure they are slender enough to buckle under minimal compression loading?
Obviously metal rods (tension rod bracing) are a safe assumption for tension only. Once you start getting into circular HSS, is this something to be checked?

 

[BridgeSmith]

Checking tension-only bracing for compression resistance would be prudent if compression resistance would be detrimental to the functioning of the structural system. I doubt there's many cases where having 'extra' lateral resistance would be a problem.

 

[jayrod12]

I've stopped using hollow sections as tension only bracing due to the potentially large compression values they are able to take before buckling. It just got to the stage on a few of the projects where the tension only assumption was so far off I felt it to be incorrect.

For angles and rods, tension only all day. And I make sure they don't exceed the tension slenderness, but I've never compared it to the compression slenderness. Usually I pick based on the smallest radius of gyration value allowed for tension members and there's rarely a capacity issue. Granted we're in a moderate wind, low (like embarrasingly low but they're now making us check it after years of neglecting it) seismic area.

 

[WARose]

At one point (not sure if they still do) AISC limited the slenderness ratio for tension only bracing to not more than 300. Not sure if they had the issue you were talking about in mind when they did that though.

Tension only bracing is prohibited in certain seismic design categories.

 

[WinelandV]

WARose,

At least as far back as the 2005 Specification, Section D1 states that there is "no maximum slenderness limit for members in tension", but then adds a user note that limiting the slenderness ratio to less than 300 is preferrable.

IIRC, the CISC has an enforced max slenderness of 300 for tension members.

StrEng007, when I analyze a steel tower in RISA, instead of using "Tension Only" for the x-bracing, I use "Euler-buckling" - this lets the brace take compression load until it exceeds the buckling capacity, at which point RISA "turns off" that member, and re-runs the analysis.

An aside, I hate doing cross-braced tubes and try to avoid it if at all possible.

Please note that is a "v" (as in Violin) not a "y".

 

[jayrod12]

StrEng007, when I analyze a steel tower in RISA, instead of using "Tension Only" for the x-bracing, I use "Euler-buckling" - this lets the brace take compression load until it exceeds the buckling capacity, at which point RISA "turns off" that member, and re-runs the analysis.

A note regarding the euler buckling analysis option, it does not check the member taking compression loads. So back to the comment about hollow sections, they can take a potentially large value of compression, potentially undersizing the tension only portion of the bracing without notifying you about the amount of load the euler buckling tagged members are actually taking. Not saying you aren't paying attention to it, but flagging it for others that haven't noticed that phenomenon.

 

[phamENG]

If you're looking for actual behavior (like a seismic design), then tension only bracing isn't a good idea. For wind design and general stability, though, it can be okay.

For me, it's ensuring that the member is controlled by elastic buckling. As long as that's the case, the member should (in theory) return to its original position without any reduction in strength or ductility after the load passes. If it's compact enough to get to inelastic buckling, you'll have problems after the load is removed and/or reversed.

The other thing to consider - how much drift will be permitted by the tension only bracing? Because as the brace in compression buckles, its stiffness drops quickly.

 

[WinelandV]

jayrod12,

Pretty sure RISA isn't hiding load in the EB members. If the compressive load in the EB member is lower than the buckling capacity, it reports the load and gives a code check. If the load is greater than the buckling load, it re-forms the stiffness matrix with the buckled member omitted, forcing the axial that it was carrying to be picked up elsewhere (usually the other leg of the x).

Note: I wouldn't tag a member as EB if it's a tube/pipe. That way lies madness.

As to the OP's question, x's tend to be designed as if the other side of the X isn't even there. For rods and small angles, this is fine, and your drift is calc'd solely on the tension bracing.

Please note that is a "v" (as in Violin) not a "y".

 

[StrEng007]

For me, it's ensuring that the member is controlled by elastic buckling.

Right, so per my statement to make sure it buckles under minimal loading (ie it's slender enough to do the old Euler bendy bendy)

 

[canwesteng]

I've never worried about this as long as the building system is ok for tension only bracing, there are many seismic systems that won't allow it. For a moderate ductile concentric braced frame, you are required to evaluate the effect of the possible compression resistance of the tension brace anyway (as least in Canada).

 

[driftLimiter]

The AISC SCM User note D1 is reccomends a maximum slenderness of 300 and a minimum of 200 (usernotes in D1 and E2). I use this criteria on my tension only bracing similar to the example in AISC SDM for OCBF Tension only).

Note that tension only bracing is only permitted in OCBF, which is a system with little expected inelastic deformation.

 

[WARose]

Note that tension only bracing is only permitted in OCBF, which is a system with little expected inelastic deformation.

The thing about that is: I've never been too clear as to why it is a issue seismically. I have a few theories (i.e. permanently bent/damaged when it tries to go into compression during one phase of the motion, damaged when slammed harder than models predict when it is "engaged" for tension, etc, etc).....but I've never been sure.

 

[driftLimiter]

I would suppose that it is due to compression damage. When we do a special braced frame they have us consider a lot more things in the compression load path and connection.

 

[jayrod12]

Pretty sure RISA isn't hiding load in the EB members. If the compressive load in the EB member is lower than the buckling capacity, it reports the load and gives a code check.

If this is truly the case, this is a new development. Their help file used to specifically flag using Euler buckling members as ones that no code check was performed on.

 

[WinelandV]

jayrod12,

I was mistaken and you are correct - even when it's still taking compressive load, it won't run the code check for anything but tension only. This seems to be an odd choice to me.

Please note that is a "v" (as in Violin) not a "y".