Resources and ideas for the design of non-conventional steel-concrete connections 3

[brTCP]

Hi,
Do you know any good apps, books and papers dealing with the design of nonconventional steel to concrete connections? By "non conventional" I mean connections that are not currently handled by programs such as Hilti Profis Engineering or Peikko Designer. Two examples I face very commonly are:

Your opinions are welcome!
Thanks!

 

[Eng16080]

You probably won't get many responses without providing more information.

Are the drawings section/side views or plan views? In connection 1), are you hanging a steel beam from a concrete slab, or is the steel beam supporting the slab? Are you transferring vertical loads and/or axial? What are the fasteners and how are they connected? What, specifically, are you having trouble analyzing?

 

[271828]

Doubtful. You'll probably need to fall back on statics and rational assumptions.

 

[driftLimiter]

Also Strut and Tie model can be useful for these. But not easy and rarely worth the effort.

 

[271828]

AISC has a design guide for cladding connections. You might find an example that's similar enough.

 

[SWComposites]

Neither design looks very good. Both have too short edge margin in the concrete. Tensile strength of lower fastener in 2) looks bad. Lots of other potential issue depending on loading, deflections, etc.

 

[JStephen]

I ran across this a few days ago, don't know if it would relate to anything you're doing:

https://www.asce.org/education-and-...alternative-designs-for-anchorage-to-concrete

 

[hokie66]

Those sections are not uncommonly encountered. As 271828 suggested, fall back on good engineering principles. You can't depend on software to solve all your problems. Make sure to respect edge distances, and provide reinforcement to prevent breakout.

 

[JoshPlumSE]

Neither design looks very good. Both have too short edge margin in the concrete. Tensile strength of lower fastener in 2) looks bad. Lots of other potential issue depending on loading, deflections, etc.

It looks to me like you're trying to extend out a steel beam from the end of concrete slab. And, you want to transfer moment at that connection. Am I correct in that assumption.

I think #1 looks pretty good conceptually. It's just through bolting a concrete member. Though I'd prefer to think of this as a simple connection rather than a moment connection. Like you're spanning a steel beam between two sides of an opening in a floor slab. Since you've got a steel plate (or a beam flange) on either side of the slab. The biggest issue I can see (for an existing slab) is worrying about hitting / damaging existing reinforcement when you drill the holes. Maybe you have to be cautious when drilling holes close to the edge like SWComposite mentioned. But, since we don't have dimensions that isn't fully clear to me.

#2 looks a little strange to me. It's got a lot of welds and bolts and plates that all have to function well to transfer that moment. I also worry about rigidity.

 

[hokie66]

They are both moment connections for the steel cantilever. I see nothing wrong with either concept, as long as first principles are used in the design of the connections to the concrete.

 

[brTCP]

Thank you all for your answers. Yes, these are supposed to be moment connections. When the anchors are far away from each other, I can design these connections based on "first principles", but I am worried about cases commonly encountered in practice in small residential buildings, where the anchors are relatively close to each other and the space is limited. @hokie66: do you have any resources you could share?

 

[Eng16080]

Refer to sketches below:

For connection 1), I would look at using two triangular stress distributions to resist the moment and vertical force. The stress distribution described in AISC Design Guide 1, Appendix B may be applicable. If this distribution doesn't seem reasonable for the bolt in tension (top right stress distribution per diagram), you could assume the stress area is equal to the bolt head bearing area plus its distribution into the steel plate. This will probably give a much lower capacity.

For connection 2), assuming that the right end of the steel is loaded with a vertical (downward) load, I think two triangular stress distributions where the steel plates are bearing on the concrete is reasonable. The vertical force and compression component of the moment would be resisted by these stress distributions. The tension component of the moment would be resisted by the shear capacity, V, of the 2 anchors attached to the top of the beam. That shear capacity is calculated per ACI 318, Chapter 17. If you calculate this by hand, I recommend double checking it using software due to the complexity of the code requirements. If the calculated shear capacity of these anchors is too low, you may want to consider enclosing the anchors with rebar to increase capacity.

In terms of anchor spacings and edge distances, I would use ACI 318, Section 17.7 for minimum requirements.

 

[KootK]

This SCI document is a bit dated (1996) but is one of my favorite references for unconventional steel to concrete connection design.

This kind of thing is awkward these days in that, with modern anchorage provisions available:

1) Taking a fundamental approach to connection design feels, and surely is, less "street legal" than using codified anchorage.

2) Many connections designed using fundamental approaches test out poorly which tends to justify #1.

3) Still, sometimes you gotta do what you gotta do.

Still, I see value in the SCI document in that it's all about doing smart things for smart reasons. Respecting tolerances, stiffness, constructability, the need to accommodate various kinds of movement. It's a very thoughtfully prepared guide.

 

[brTCP]

Great answers KootK and Eng16080! Thanks!

 

[Eng16080]

That's a good call on the SCI document. I forgot about that one. The connection details drawn (by hand) in that document are really impressive.

 

[ALK2415]

"Rilem 2001 Connection Concrete Steel"
has great detail about this subject also ...
Link