Built-up column w/ Slender Elements

[strangeoddity]

Here's a structure type you don't see everyday...for good reason.

Brief Description: I am looking to analyize a "stand-alone" column structure (k=2.0) built up of a series of 18--1" dia. rods. in a circular footprint. The column tapers as it gets taller. The structure will be "banded" together at set intervals. The architect would prefer to limit the banding locations... currently showing 3 interior bands for a column 18'-6" in height. There is is a horizontal force applied at the top of the structure.

Questions:

1.Does anyone know of a good reference that touches on this topic?
2.Obviously it is the slenderness of the individual members that has me concerned. If these members were just loaded with pure axial loads, then the bands would serve as tension rings and absorb the buckling force created by these slender rods. Unfortunately, due to the loading condition...some of the rods will be in compression... others in tension. With this loading condition, I don't believe the "banding" is serving as a tension ring anymore. So now what?

 

[Bagman2524]

I would check each rod individually for buckling between the connection points. And the entire member for a comination of bending and compression and bending and tension.

 

[strangeoddity]

Using a k=1.0 for each individual member I would need banding at 2'-0" increments...which maybe the way to go.

However, if the banding is a collector for the buckling forces of the compression members...how are these forces resolved? do you think they are just resolved internally because they are rings?

 

[UcfSE]

Have you checked the built-up column provisions in the AISC specification, or whatever steel spec you are using?

 

[strangeoddity]

Unfortunately, AISC isn't overly clear on the subject matter of built-up columns (at least not in the ASD manual). the commentary says "Requirements for detailing of built-up members, which cannot be stated in terms of calculated stress, are based upon judgement, tempered by experience". Plus the built-up section doesn't exactly imply built-up members with rods. AISC seems to limit maximum connection attachement to 24" o.c.

 

[UcfSE]

You might want to check the newer spec for soem additional information involving built-up columns. You'll find a modified KL/r ratio to use in E2 and E3.

Given that your column is tapered however it doesn't appear you can use chapter E anyway. The first paragraph states the chapter applies to prismatic members. You might be looking at using FEM to design this.

 

[swearingen]

This is definitely out of the built-up column bounds mentioned in AISC. A built-up column must have continuous (or nearly so) shear connection between the pieces. These rods are acting independently until they get to the bands. The entire thing should be modeled with the load at the top to get the loads in each rod. Then the maximum compressive load diverted to a single rod should be used to come up with a band spacing. You will find that you'll need many bands to make that work - he's not going to like that...