AISC Built Up Column Fasteners

[bookowski]

I'm looking at a cruciform column made up of four equal leg angles (4x4), 3/8" space between each angle, and want to make sure I'm understanding a few of the E6 (13th edition) provisions:
- Lower bound on fastener spacing such that kl/r of each shape < KL/r of the overall, easy enough
- Fasteners must also resist the shear transfer between parts. got it.
- Here's the part i'm not clear on "at the ends of built up compression members bearing on base plates all components in contact with one another shall be connected by a weld having a length not less than the maximum width of the member or by bolts spaced longitudinally not more than four diameters apart for a distance equal to 1-1/2 times the maximum width of the member". The "components in contact with one another"... since I have 3/8 spacers between my angles, I strictly read this as not applying since my components are not in contact. But if the intent is some minimum robust end shear transfer then it should apply I guess. The commentary states "in both cases the end connection must be welded or fully tensioned bolted", which would seem to imply that I do need them. If I assume that this does apply, despite not being in contact, and assume 3/4" bolts then I need end bolts @3" o.c. over a distance of 1.5 x 8 = 12", so say 4 end bolts on each cruciform leg top and bottom (so 32 bolts per column added). I've got loads of these so it's quite an uptick in fab/labor to add on 32 bolts per column if I don't need to. I have not run the numbers yet but I'm guessing that these end bolts far exceed what I need to transfer shear stresses in my column pure/mostly axial condition so it seems like a bit of a kick in the nuts to have to provide this. The commentary does note that the requirements "cannot be stated in terms of calculated stress but are based upon judgement and experience", I guess an acknowledgement that this may not make sense but do it anyway. (?)

 

[bones206]

Sounds like the intent is to ensure equitable load sharing between the individual elements at the point of bearing.

 

[ProgrammingPE]

I would provide the 16 bolts at each end of the column. Like bones206 said, it is to ensure that all of the parts of the built-up member get engaged equally. There isn't a reason why adding spacers should make this requirement unnecessary. The angles are technically not directly "in contact" with each other, but they are in contact with the spacers which are in contact with the other angles so each of these would still require the end connection.

Structural Engineering Software:

http://www.structuralcentral.com

Structural Engineering Videos:

http://www.youtube.com/channel/UCOj2mXXf3ZbZtgxTc6BtkGg

 

[bookowski]

What if you didn't need them to act compositely get the required capacity, i.e. what if 4x a single angle capacity is sufficient. Would you still provide 32 bolts to stitch them together? Or rather do you think you need to. It would of course be easier to provide them, but if you multiply this out times hundreds of columns (so thousands of bolts) it's not a small cost uptick.

 

[bones206]

I don't see why you couldn't design it as a single angle column, with 3 architectural angles just going along for the ride. If you consider it that way, you shouldn't be bound by the requirements specific to built-up members.

 

[ProgrammingPE]

If 4x the capacity of a single angle works then you wouldn't need the end bolts since you're not using the built up member capacity. I'm not sure how your load is being applied, but it is unlikely that all 4 angles will support it equally so make sure you account for that.

Structural Engineering Software:

http://www.structuralcentral.com

Structural Engineering Videos:

http://www.youtube.com/channel/UCOj2mXXf3ZbZtgxTc6BtkGg

 

[bookowski]

To poke at this further.

What if the 4 angle cruciform was not bolted per aisc end requirements (32 end bolts) but rather had some minimal stitching together along its length, lets say bolting at 24" o.c. (4 bolts up the length at each leg, so 5 spaces). Is there any quantifiable benefit that the angles offer each other in reducing the buckling length, or rather changing the buckling mode.

 

[bones206]

I'm not sure the answer to that, but it's an interesting question. Might need a buckling guru to weigh in on that one...

One way to look at it might be a single L-column with partial-length buckling reinforcement in the form of 3 other L-columns. Not sure how to quantify that though...

I'm pretty confident if you meet the maximum intermediate fastener spacing requirement you are still getting a good amount of composite section benefit. But I'm less sure about the effect the end slip would have if you aren't meeting the end fastening requirement. My guess is having the tighter end fastener spacing would improve the buckling capacity to a degree, but that the intermediate fastener spacing probably has a greater influence.

I did find this paper, but I'm not a member of ASCE so I'm not able to download and review:
ASCE Library: Buckling of Four Equal‐Leg Angle Cruciform Columns