Unbraced Length of Compression Members in Lattice Towers

[MSL93]

I am looking at a steel derrick structure (lattice tower) constructed with tubular members. The point of concern is the unbraced lengths used for the main braces that have secondary braces connected at intermediate points. The two main braces that the secondary brace is connected to can be in compression at the same time.

The EN 1993-3-1 Figure H2(b) (attached to this thread) states that this secondary brace (the code refers to it as a "corner stay") will have limited effect if both braces are in compression. From this note on the figure, it would sound like the secondary brace cannot be used to reduce the unbraced length of the main braces if there both main braces have compression load in them simultaneously. I have a few issues with this note.

1. There is no commentary to explain this note.
2. Even though the two main braces could be in compression at the same time, they may not be very heavily loaded and still have a lot of reserve capacity.
3. Both main braces will be in compression just when the structure is only subjected to dead load (i.e. most of the time).
4. What would be the purpose of having this secondary bracing if it can't be used to reduce the unbraced length of the main brace? The nature of it being a secondary brace will be such that it does not carry much load.
5. Secondary bracing is generally considered "stiff" enough to provide buckling restraint to the main brace if it has a compressive capacity of at least2.5% of the maximum compression force in the main brace.

I don't find this note in any other lattice tower design code or standard (TIA-222, ASCE 10, CSA S37). Can anyone please provide some explanation or a thought experiment on the merit of this note and if it reasonable?

 

[KootK]

The trick here is to recognize that the thing doing the bracing needs to, itself, be prevented from moving. The system that you're looking at relies on the a compression member bracing being coupled to a tension -- or zero force -- member to to stabilize it. If the member stabilizing the brace is also prone to buckling, then the strategy looses it's effectiveness since both of the braced members just buckle unison. I've done a poor job of trying to illustrated this below where both braces would tend to just rotate about the axis of the primary tower leg.

2. Even though the two main braces could be in compression at the same time, they may not be very heavily loaded and still have a lot of reserve capacity.

I'm sure that would help matters. But, then, the difficulty comes in quantifying the degree of that help.

 

[Lomarandil]

If you're looking to make this work, the concept is called "lean-on" bracing, and is discussed in (at least several editions of) the Guide to Stability Design Criteria of Metal Structures:

https://www.thriftbooks.com/w/guide...-structures-5th-edition/342238/item/54333597/

 

[BAretired]

The bracing shown below does not prevent buckling. Wrong!!! See next post.

 

[Flotsam7018]

This is presented in ASCE 10, requiring the crossing member to have a tension force equal to at least 20% of the compression force to provide a bracing point, see Example 9 in ASCE 10-15. You can always consider the different unbraced lengths as appropriate for each load combination.

 

[JoshPlumSE]

I think it's a really good point. So much so that I wonder why TIA hasn't mentioned it in their code. That's usually the "go to" code for me for discussions on this type of lattice tower. But, I don't remember any discussions like this at all.

That being said, TIA does require the tower to be analyzed (considering geometric non-linearities) for many different load directions. Therefore, they are still likely to catch any elastic buckling that would tend to occur if those braces simultaneously go into compression.

 

[BAretired]

The rotation described in my previous post cannot take place. Each latticed leg is stable because they are each composed of four triangles.

 

[BAretired]

Member 1 is redundant unless it is needed to brace the vertical member, in which case, only half of it is required.