SJI Joist as Lateral Bracing

[jheidt2543]

Roof joist bearing on a supporting beam can be counted on as providing lateral support for the beams top flange. As a consequence of this, the SJI roof joist receives an axial load.

This also could be the case where, at an exterior wall, the joist provides support for the perimeter beam, which could also transmit some wind load from wall panel to beam to joist.

And of course, there is the OSHA requirement that the joist at column lines be a bolted connection to provide lateral support for the top of a column during erection.

My question is, how do we determine how much axial load a joist can receive? I haven't seen a reference to this in joist load tables, surely there is a limit.

 

[JedClampett]

We sometimes end up with this problem when we want to use a joist for a drag strut. We post the load on the drawing and leave it to the joist manufacturer to design it. They sometimes add additional bridging to brace the joist and sometimes don't. Of course the loads are relatively small (<6000 lb.).
I would advise against trying to calculate joist capacities yourself. The joist manufacturers are notoriously stingy with section properties, much less material properties, so any calculation would be starting on shaky ground.

 

[Lutfi]

jheidt2543,

Your thought process is very true. However, it is more complex to determine the force. You have the diaphragm be it floor slab or roof deck that will also provide resistance (that is what the roof diaphragm theory is all based on). Theoretically, the joists bearing on exterior CMY wall, they will see the load and the roof deck will resist the lateral load to the shear walls or vertical lateral bracing system.

How much force does it take to laterally brace a steel beam compression flange? I have heard many stories. The one I respect a lot came from researchers at some national universities. They did flange bracing using trivial methods such as tape/cardboard combinations. It works to certain extent.

I practice in Florida, for roof trusses, I always ask contractors to supply me a truss seat/strap that will resist the following forces:

1. The uplift
2. The force parallel with the wall (normal to the truss)
3. The force normal to the wall (parallel to the truss).

For a steel joist, you can determine a tributary area for each joist; compute the wind load and request the joist designer to design for additional axial load.

Joists and joist girders can also be used as part of frame system that carried lateral loads (by anchoring the seat and the bottom chord). Dr. James Fisher done an outstaying job along with Vulcraft in the book titled: “Designing with Vulcraft steel joist, joist girders and steel deck". I recommend that all structural engineers get a copy of it by contacting Vulcraft. I think it is around $14.99, a barging in opinion. Actually I won my copy at the AISC annual conference few years back in Seattle. The link to their page is

http://www.vulcraft.com/

and click on the order link.

Good luck

 

[JAE]

Well, for horizontal steel joists that are perpendicular to the wall that they are bracing, keep in mind that the lateral force coming up through the wall and into the joist begins to diminish linearly as the attached metal deck begins to receive it.

A steel deck may have an attachment pattern that provides, say 200 plf shear. Or you can directly calculate the fastener capacity along the joist and determine a plf.

So you joist gets the "full" lateral force at its end, directly at the wall, and then linearly gives this axial load up to the deck as you proceed down the top chord of the joist. At midspan, where the joist bending chord compression is max., the load may be essentially "gone".

 

[haynewp]

I think that for example on a standing seam roof with cross bracing betweeen joists, you wouldn't have the diaphragm distribution effect and it would be best to specify the axial load to the joist top chord if it is significant.

When I have an existing joist that receives new axial load (like a drag strut at the roof edge of a new addition) I have taken the axial load to the top chord and found the equivalent moment about the joist that would cause that axial force. Then I would take that moment and get an equivalent distributed gravity load that would cause that moment and add it to the distributed gravity load the joist would see anyway.

I would then compare the total equivalent distributed load on the joist to the table value to see if joist reinforcing was required. Then if it was within the table value, I would calculate the allowable moment diagram and compare that to the new diagram with the constant equivalent moment along the joist due to the axial load plus actual gravity load moment.

 

[JAE]

Yes - I agree that with standing seam roofs you don't have the ability to transfer the load away from the joist.

good idea on the equivalent moment technique.

 

[Lutfi]

Good discussion. I agree with haynewp regarding the standing seam roofing. I have used and continue to use the equivalent moment method as well. On new designs, I indicated Free Body Diagram with loads for all special joists. It is a little tougher if an existing joist receives new or revised loads are being applied. I had a tough time getting the joist designers to go back and pull their files and have them perform calculations. To be fair, they did do that for me in some cases where I was able to supply them with data and joist tags.

 

[jheidt2543]

Thanks guys for some really thoughtful and thought provocing responses. I guess that I had two conditions in mind when I asked the question.

Condition 1: SJI joist perpendicular to the wall, whether the wall is masonry or metal panel with the joist bearing on a beam. I'm wondering why rely on the diaphram action of the roof deck in it's "weak direction" when you could resolve the wind load into the joist as an axial load?

Condition 2: SJI joist parrallel to the wall, either masonry or metal wall panel. In this situation, the roof deck edge is receiving the wind load in the "strong direction".

So, in essence, you would have a one-way deck diaphram, the joist are already there, why not make use of them for both lateral bracing of the perimeter beam and resisting the wind load in the one direction and the deck in the other direction?

I guess I like the idea of showing the axial load on the plan and having the joist manufacturer design the joist for it. I was just wondering how big that force could required to be.

 

[haynewp]

I guess it is just the added expense of designing the joists to take the load instead of relying on the diaphragm to take it when it is there. When properly fastened, the deck can take a lot of axial load in the strong direction. For the weak direction, apparently it has historically performed well enough with adequate capacity (I have never seen drawings that state to design the joists for axial load when you have a legitimate diaphram to take it).

If the load combinations you are using does not have a full dead+roof_live+wind case, then you could check your dead+roof_live moment versus the dead+wind moment diagram(if you do not have a net uplift on the joist).

If you do have net uplift, your top chord would be in tension to help reduce the axial compression load you are concerned with.

 

[JAE]

Weak direction vs. strong direction??????

Yes, I know the flutes are one way and if you push on the edge of a deck in that direction it will accordian on you. However, you are not pushing on the edge of the deck as though the deck were out in space, free of any constraints.

The deck would be part of a semi-rigid diaphragm that has a shear stiffness G' that is INDEPENDENT of shear direction. You are essentially dragging the load into a deck that resists shear deformation across its width (from brace to brace or shearwall to shearwall). I suppose you could get some degree of accordian action, but it would be limited by the adjacent joists and their connection to the deck (i.e. the accordian action only occurs between the joists, not across the full width of the diaphragm).

 

[connect2]

Haynewp has a very good point 'When properly fastened...',

 

[jheidt2543]

JAE:

I may not be asking my question in the right way.

The difference in deck resistance along the span edge vs the span end is real, but can be taken care of by the fastener spacing. So, I take your last comment to say essentially, if the deck is there use it.

I think I was asking, why not use the joist in one direction? It just seemed to me that the joist are there and should be able to handle an axial load (since we already rely on the joist for lateral bracing of the support beams) and if so, how do we arrive at that load. The answer seems to be if you do it that way let the joist manufacturer figure it out.

But, the answer I think I'm getting from everyone overall is just use the deck as a diaphragm in both directions and leave it at that.

From my reading of the SDI Diaphragm Design Manual, the shear strength of a metal deck diaphragm is the smaller of three limiting values: one along the panel edge, one along the sidelap and one along panel end laps. While the shear direction may not be a factor, the shear capacity of one of the three conditions above seems to be controlling.

Maybe I was thinking too much???

 

[JAE]

No - you are making sense. I just think that your comment - "the joists are there and should be able to handle an axial load" is true, but you still have to follow a load path from the tip of the joist down to the ground somehow.

If the deck wasn't there, then the joists and support beams become a big, horizontal pinned-pinned frame that cannot direct axial load anywhere. Very weak and loose and - boom - fall down.

I guess I think that axial load transfer into a joist is impossible without the deck, unless you have rod x-bracing as in a pre-engineered metal building with light metal roofing.

 

[haynewp]

I have heard diaphragms described as a beam where the deck is the web and the joists just act as stiffeners for the web. I don't know if that helps any.

I have had to think about this some myself and I imagine it like the exterior wall is hit with an impact force, say gust of wind, the whole roof system moves freely until the deck 'catches' the roof and is engaged after some deflection. Then the axial loads begin to be introduced into the roof system and it is no longer just a moving object. Going back to JAE's earlier post, whatever load that is introduced into the joist end at this point, is quickly dissipated by the deck acting as the counteracting support along the entire length of the joist.

The joist itself has no other lateral support since it is theoretically pinned at the ends and the columns are modeled as pin-pin. The deck itself is thought of as a soild body resisting deformation.