Roof Beam to Tilt up panel Connection

[civeng80]

This is a roof beam to tilt up concrete panel connection that I am proposing on a job. I thought this was a fairly typical connection detail at least in Australia.

Comments from checking engineers are that it will not act as a pin connection (even though it is designed as a pinned connection) and that significant moment will be transfered possibly damaging the panel. I always thought that this was the closest (or one of the closest) type of panel to beam connections to represent a pin.

Any ideas or suggestions or modifications on this would be appreciated.

 

[civeng80]

aaron

Yes it is necessary because

1. Of what Hokie said about transfering the strut load to the shear wall if you go with this type of connection.

2. Over time the nut on the ferrule may loosen (even slightly) and the clamp drop off.

 

[aaronPTeng]

civeng80,

1. This depends on the design, An I am guessing in your case for your situation you need this.
2. I dont think this is really the case As the base of the clamp is fixed hard up against the precast. I don't see how it can loosen. either way just a suggestion I have done plenty of connections this way and it is fairly typical, However each to their own

enjoy,


"Structural Engineering is the Art of moulding materials we do not wholly understand into shapes we cannot precisely analyse, so as to withstand forces we cannot really assess, in such a way that the community at large has no reason to suspect the extent of our ignorance." Dr. Dykes, 1976

 

[civeng80]

These structures stay up for 50 years or more.

Many things happen in 50 years.

I always like to weld them (tack welded) then forget them.

Cheers

 

[rowingengineer]

I don't like the cast in plate detail, N20 bar cranked into a 150mm panel for what purpose? N20 sounds large and the crank is going to be fun. I would suggest N16's would be better, and probably N12's would cut the mustard.

maybe something more like this???

http://www.reid.com.au/Reid_Services/Reid_products/Connections1/Cast-In_Plates.aspx

http://www.nceng.com.au/

"Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning."

 

[CELinOttawa]

I'm probably a bit late to the conversation, but I can tell you that I've seen a lot of this detail in New Zealand, along with a few failures. This detail failed in a fire scenario, costing the owner the ability to reconstruct, as well as plenty of failures under EQ load. Both due to the moment, but this is very rare indeed.

Frankly in Aussie I understand you have much less significant EQ loads, and I can see why these are common and no causing issues. I agree that there will be a moment induced, but I doubt it would be in any way significant. Let's not forget that what we are looking at in 2D is separated by a meter in real life... Small rotations and deflections will eat/dissipate a great deal of load in this case. P-Delta is not always our enemy.

FYI: Where I need to be able to predict behaviour accurately (ie: In NZ Capacity Design), I use the shelf angle for construction (only) and introduce a 15mm X 50mm X 200 or 300mm (depending on length required to extend past pair of or 4-bolt set) plate and guarantee pin behaviour.

That's likely over the top for Aussie, with your Cyclonic loading needing more "hold onto the weight" design and less "flex without breaking" seismic design.

 

[civeng80]

Thats an interesting comment.
Can you elaborate how the connection caused failure in a fire due to the moment ?
Also could you place a drawing of the connection you use in NZ ?

Very interested in your comments.

 

[CELinOttawa]

So the failure which occured due to fire was in a large warehouse. Piles of stored wares were involved, and the heat caused the rafters to sag and twist. Once the beams were effectively counter-cambered, they were unable to keep from continuing to rotate, and the top bolts in the affected (discussed & shown above) connection blew out. The lower bolts never failed, and the total rotation at the knee was never beyond the capacity of a true pin connection. While I cannot know a true pin condition would have saved the day, it would have stood a much better chance, and would have likely saved the owner from replacing panels and rafters both.

I'm attaching a sketch of the detail I've used in NZ; I'm on parental leave at the moment and don't have a scanner at home, even if I did have a drawing of the detail. Note that this also comes in very handy where you are close to the limit on the shear value of the connection. You can't neglect the induced moment in the panel, but you can limit the design values for the bolts to pure shear.

<a href="

http://tinypic.com?ref=1zd8y7d"

target="_blank"><img src="

http://i40.tinypic.com/1zd8y7d.jpg"

border="0" alt="Image and video hosting by TinyPic"></a>

 

[civeng80]

Thanks for that CEL.

So you use a shelf angle for erection only (then remove) and connect to what seems to be some steel packers to the panel ?

But what about the eaves channel which is the main culprit in inducing the moment ?

 

[CELinOttawa]

If the contractor wants the shelf angle to make their job a little easier, yes, it is for construction only. Most of the time I've found the panels are propped and the rafter is crane lifted, so they don't bother with the shelf angle (though I have seen it used, and required that it was removed afterwards).

For the eaves channel, the pin details means this issue is also solved. They are in line, or nearly so, therefore the moment is now zero and/or negligible.

Note that where required structurally, I have used back to back Diamond CFS channels powder fastened to the panel. Cheaper and quicker to build. If the loads are too high for powder fastening, I try to change the load path or rearrage to minimise load. Never been a big fan of too many post-fix anchors or chemsets. That many embeds is just going to lead to tears/site work/money.

Make sense?

 

[civeng80]

Thanks again.

This topic interests me alot.

Ive been looking for a better more physical pin type connection for a long time.

The one you presented sounds good, but is it possible to see it in a bit more detail ?

Questions like is the butt plate to flange welded to shim on panel?

How do you pack up exactly to the beam butt plate etc.

Your comments greatly interest me.

 

[CELinOttawa]

The shim plate is welded onto the end plate, and then the gap (15mm to 25mm typically) is left to be dry packed.

I've also extended the end plate up to be able to add a final purlin where the panels are not kept tall (ie: no fire condition or other architectural requirement), thus having it do double duty. Otherwise you often see an akward detail with an extended cleat to hold a final purlin out and away from the knee. Looks like garbage and is even tougher to fabricate and install. Terrible detail, frankly!

 

[civeng80]

To be honest I dont like this detail and I think it would be rejected by Aussie Engineers.
I haven't seen one detail that fits the bill yet.

 

[CELinOttawa]

Well, keep looking and keep working towards figuring out something you do like... Each detail should be thought through in great *cough* detail *cough*.

I love introducing the rocker (what you called a steel shim) because it lets me control the behaviour and I know that I won't get unexpected prying forces.

Let us all know what you choose to do in the end, but I really can't see why anyone would reject a detail that introduces a true pin condition... Would it be rejected on the basis of undue complication? Certainly can't be rejected for any reason of safety or predictability of behaviour.

 

[civeng80]

Please don't be upset by my comments, maybe I didn't express myself well.

Its probably a good detail when its up and functioning.

Construction of of it may also be OK once you get a system going.

Are the bolts from panel to beam cast in ferrules or Trubolts or chemsets and what happens if you need 4 bolts for shear?

 

[CELinOttawa]

Hi CivEng80,

I'm not at all bothered by the comments, if anything having another engineer criticise your work makes you think again and build your skills!

The anchors are embeds (cast in elephant foot ferrules), and in the cases where I've needed four bolts, I've placed two to each side of the flange. Then you need to ensure you get the load spread, so I use a diamond shape end plate welded onto the end of the rafter to spread the load, rather than trying to check the rocker plate as a beam. The most I've used was eight ferrules in two rows, which then get checked for moment as a group, but permitted the rafter to see a true pin and minimised the design loading required at the connection.

 

[a2mfk]

See the attached sketch and calcs. This is how I have done tilt-panel and other concrete connections that support steel beams, and how I have analyzed them in considering the moment from the eccentricity of the connection (whether a seated angle, single shear plate, etc.)

At least in Florida, USA, welded headed studs have been one of the most popular forms of these types of embed plates. To deal with the reduced values of headed studs and other connections in concrete that are the result of ACI Appendix D (complex code that greatly reduced the capacity of concrete anchors), some engineers now use a shear lug concept, which I prefer for higher loads. This gets you out of Appendix D (for the shear lug), and then you can design using standard shear and bearing methodology. I may add additional horizontal or vertical rebar for crack control depending on the loads and type of concrete member, and have used additional rebar and the strut and tie method to reinforce the concrete under the shear lug to prevent a diagonal crack and shear failure.

Connections using epoxy or expansion bolts are useful in existing concrete, similar to the connection shown in the lower right corner of the sketch. The manufacturers now have some pretty good design programs in the USA at least, which can make design a lot easier and faster (multiple iterations). In any case, remember you may have shear and tension in the upper fasteners simultaneously.

Back to your connection, I may consider turning the TOP hooked bars horizontally, and then providing supplemental vertical bars to reinforce the possible shear failure of the upper rebar in tension. This would also give you a couple of additional vertical bars in the panel in the area of your connection where you have the highest stresses. Though an additional horizontal bar I indicated in the sketch with your detail would do about the same thing.

Hope this helps. Like to hear what other engineers are doing out there with these types of connections. They become very problematic with thinner panels and especially near panel edges, door openings, etc.

 

[CELinOttawa]

A2mfk: Great post, but bear in mind that the added horizontal bar has nearly no effect on the SLS state and the initial (read: apparent Failure - initial cracking and the upset call from the client) and only very modest gains on the ULS. There is an excellent report on this from the University of Canterbury. I can post photos of my copy if anyone's keen.

 

[rowingengineer]

keen

http://www.nceng.com.au/

"Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning."

 

[CELinOttawa]

Here you go rowingengineer... Apologies for the watermark; Like I said above, I don't have access to a scanner atm...

 

[civeng80]

Back to the original question.

Been looking at a Bunnings building which is huge. It has the same connection as mine shown above but no bolts on the shelf angle that I could see at least. So the joint is completely free to rotate. Anyone involved in the design of these buildings here ?

Again any comments would be appreciated as the checking engineer is giving me a hard time on this.