Slotted connection to avoid load transfer 1

[canadiancastor]

I'm working on a steel infill truss that will be used to attach thin natural stone panels to a concrete building. The building is 35 stories high with cantilevered concrete slabs. The infill structure is a "3D" box element that is 2' wide by 1' deep and will be suspended outside of the building envelope (see attached sketch). My plans is to cantilever steel channels off the slab face (using channel anchors or concrete screws) and attach the truss to these channels. The truss itself will be composed of steel angles, from a preliminary analysis, around L2x2x3/16. I plan on transferring 100% of the gravity load to the bottom slab (at each story), while lateral loads will be transferred at the top and bottom of each truss. I'm thinking of having the whole truss galvanized once fabricated.

My question is about the transfer of lateral loads at the top of the truss. I want to avoid transferring vertical loads, for a couple of reasons:
[ol 1]
[li]I don't want my structure to support the concrete slab and its loads[/li]
[li]I want to permit thermal expansion to some degree[/li]
[li]Some room for on site adjustment, if possible, without welding[/li]
[/ol]

I'm trying to find an easy way to create connection that transfers loads in both lateral directions without transferring gravity loads. I came up with the slotted connection shown in my sketch. I know slotted connections are used all the time for light gauge steel walls at the top rail. Is there an issue with the bolt moving up and down in the slot? Like screeching noise? Wearing or loosening of the bolt? I guess I'm not used to having things moving around in my designs.

This will be my first time on this type of project, any extra tips would be appreciated. I have mandated an external firm for the design and drafting of the thin granite panels and the substructure (back face anchors to aluminum clips).

 

[DaveAtkins]

Slotted holes which allow movement are common in hot rolled structural steel, especially where thermal movement must be tolerated.

I like your detail, but I have a couple of recommendations:
[ul]Use two bolts at the connection; it is never a good idea to use just one bolt.
Specify how the nut is prevented from backing off. You can stake the threads, or use Loctite, or weld the nut to the bolt.[/ul]

DaveAtkins

 

[milkshakelake]

Slotted holes don't allow movement if the screws are tightened too much. Specify "tighten bolts finger tight." I couldn't find anything in the AISC code about this, but here's an article. Link

 

[TLHS]

If it's critical that the nuts not be too tight, I would tend to say that you double nut the bolt with them both finger tight and then back off the nut closest to the joint tight against the rear nut. You have the inner nut at the tightness you want, but have the two nuts tightened against each other so that they can't vibrate or otherwise work themselves off the bolt.

 

[1503-44]

Better off with a tack weld if there are significant temperature extremes.

 

[SwinnyGG]

If I'm understanding your details correctly, I'd be concerned about the assembly generating noise- if the slotted connections are truly loose enough to allow movement and not transfer any vertical load, you'll also be relying on bearing on the bolts to absorb lateral load. If you're hanging panels on the outside, I imagine a condition where the CG of the assembly is outside the mounts, meaning that under zero wind pressure the assembly will rest to the outside (axial tension in the top support) and then when wind blows against the building, slam into a hard stop as the assembly is pushed toward the wall (axial compression in the top support).

You could turn the top connection into a flexure connection. Instead of channel for the top connection, use a flat plate (long side parallel to the ground) that can absorb some axial load, but would be soft enough in bending on that axis to not apply significant vertical load back into the structure. You could slot that connection for constructability if you want to, but you'd want the connections to be slip critical.

 

[canadiancastor]

Thanks for the tips everyone.

@1503-44 I would rather not have to weld on site, I think staking the thread would be preferable. Do you think the temperature difference is something that could loose bolts?

@SwinnyGG Without even getting into the C.G. of the structure, the wind loading is in both directions (pressure as well as suction), so if wobbling creates noise, it will be a problem. I like your idea of a flat plate, it removes the need for any sliding joints. I will look into hold much load it can transfer.