Long Span Joist Parallel to Wall 1

[RFreund]

What is your favorite detail for a long span joist parallel to an exterior wall? A few options:

1. Space the joist off the wall, say 6' and use a ledger angle to pick up the metal deck. The angle will support the deck for dead loads. The deck top flutes are cut so that it can be manipulated and welded to the angle.

2. Same as #1 but direct the contractor to install the angle in segments to match the joist camber profile.

3. Place the end joist close to the wall, say 8" off the wall and cantilever the deck. Provide an angle with slotted holes. In this case the angle transfers out of plane lateral load and in plane lateral load.

For #1 and #2 the deflection of the joist will warp the deck and you might have problems with installation. Also you may need to install a truss system or some sort of additional out of plane bracing for the exterior wall.

For #3, you take care of deflection, but you don't have much room if you wanted to install an in-plane truss to help transfer out of plane loads from the wall. You need to rely on tension/compression of the deck.

Thanks!

EIT

http://www.HowToEngineer.com

 

[BAretired]

I've seen it done two ways (1 and 3). Method 2 makes a certain amount of sense but I don't recall ever seeing it done in practice.

In the case of roof framing, joist spacing will likely be in the order of six or seven feet. Usually, 22ga. x 1.5" deep steel deck is used and it is flexible enough to accommodate the difference in elevation due to camber. Floor joists are spaced more closely, but they have the additional dead load of a concrete slab which removes most of the camber.

In summary, method 1. seems to be the preferred choice of most engineers.

BA

 

[JLNJ]

I prefer #1 and don’t cut the deck flutes.

There is a lot of prying action where the ledger bolts to the wall. Make sure you have plenty of strength in the angle, the bolts to the wall, and the welds to the deck. The deck diaphragm action should keep you from needing a horizontal truss.

 

[RFreund]

Thanks for the responses. If you have really long span, say 100', you have 4.25" of camber and allowable deflection might be on the order of 4" for a roof. Does #3 start to become more attractive?

EIT

http://www.HowToEngineer.com

 

[BAretired]

For 100' span, Detail #3 would require very long slotted holes to accommodate a camber of 4.25". It might be better to place the angle at the highest elevation of the joist and splice the deck at the first joist. That way, when snow lands on the deck, the end span will tend to slope away from the wall.

BA

 

[r13]

No matter what method you use, you have to keep in mind and ask yourself, how much the elevation difference can be comfortably accommodated by the rigid edge support without detailing problems. For solid wall support, I might place the joist as close to the wall as possible, so you can relax the chamber for it supports fewer load and deflect less. For steel beams, I think all options are viable, depends on the deflection characteristic of the edge beam.

 

[JLNJ]

For longer spans you will likely have 3" roof deck. This will allow for a bit more warping for option #1.

For really long spans, I've seen it where the framing in the end bay is changed to accommodate the roof deflection. The last truss is set a ways from the wall and becomes a jack truss of sorts carrying framing which spans the opposite way.

If you place a joist near the wall and the deflection is large, you will rip the roofing and flashing apart. Check with your architect, but he or she will probably not be thrilled with large deflections up against the parapet.

We did a project where we put a new floor in up under a 120' span joist. I told the architect that he was going to need to account for something like 5" of deflection when he tied his interior light gauge framing partition walls off to the roof structure. Needless to say, he specified a regular slip track and was left scrambling at the end of the job when I reminded him of the large deflections in those long-span joists.

 

[RFreund]

Thanks again for the responses, mostly confirms my concerns/considerations and even more to think about. In general I suppose keeping the truss off the wall and maybe specifying tighter deflection requirement is what I'm leaning towards. However, one concern is then compression / tension of the roof deck for out of plan loads of the wall. The deck likely has a fair amount of axial strength, but you would need to consider some combination with bending stress. I don't see this check often (if ever). I'm not opposed to install some framing perpendicular to the wall, but this framing will also be subject to the same movement. Having said that, this movement is the same that would be found if you were to use a "jack truss" or any other case where you have joists that frame between wall and girder and you typically wouldn't think twice about that... Until you get to a 100' span.

EIT

http://www.HowToEngineer.com

 

[JAE]

We usually use perpendicular channels that serve to provide a positive connection to masonry or precast wall panels.

The deck axial strength is amazingly high if you calculate it, but I've never trusted "tin foil" members to brace heavier members so we use the channels.

 

[r13]

Note that when place the first joist close to support, the deflection is reduced due to reduction in tributary load. For example, if the joist span is 6', and the first joist is 1' to the support, that result in 1-(3.5/6)=0.417 reduction. For 4" anticipated deflection for interior joists, the deflection of the first joist is 4*(1-0.417)=1.668". If you provide 2" pre-camber, then it is almost level with the support when loaded. However, there is elevation difference between the first and adjacent joist due to different amount of chambering, this might not desirable on interior floors, but should be good for roof applications.

 

[dold]

I've done a lot of tiltup production work. Our client preferred option #3, but instead of cantilevering the deck, we used a 4x8 bent plate (LLH) that essentially runs overlapped on top of the joist along the entire length of the wall and welds to embeds similar to what JAE shows.. I never got a straight answer as to why they preferred this, but I think it had to do with erection, and this gave them a place to rest the cont plate while welding. We'd always spec the joists against the wall to have zero camber, and joist guys never seemed to have any issue with it. Now that i think about it, they may have liked that end joist just so they have something to hang their dropped ceiling from.

 

[r13]

Also easier to run the fire protection supply line, I believe.

 

[JAE]

Note that when place the first joist close to support, the deflection is reduced due to reduction in tributary load.

Don't forget that first joist may have much higher snow loading due to parapet drifting compared with interior joists - so that statement isn't always true.

 

[r13]

Yes, JAE has valid point for roof application with tall parapet. In such situation, one shall account for the additional deflection due to the snow built-up above the roof snow that is an uniform load for all interior joists. Note that the first joist carries only one half of the span between it and the edge support.

 

[jayrod12]

On longer span joists running adjacent to walls, I've seen the spacing get progressively tighter I believe to minimize said deflection differential etc.

For example in one of the warehouses I perform frequent work, the first joist is 12" away, second joist is 24" space, third joist is 36" space, then 48" and then whatever the standard spacing is after that. In that instance, there is built-up snow as well, which probably plays a part in the spacing, but realistically it's probably helping precisely the issue the OP is having.

 

[bones206]

Dold,

On your detail it seems like gravity loads would bypass the joist as the welded connections at the embeds would provide a gravity reaction. Is it designed that way?

 

[Brad805]

JVI makes a nice slotted joist connection if you are talking about attached in CIP/Tilt/PC. They come in different heights and configurations. Link You still have the roof membrane tear issue, but years ago I was told up to 5" was a solvable problem for most SBS membranes.

 

[XR250]

I have always placed the last joist 5 or 6 ft. from the parapet and let the deck be supported on an angle at the wall. I figure the deck would flex enough to accommodate the joist deflection. I guess you could also specify a lap joint in the deck at the joist location to enhance its flexibility.