Steel beam on stud wall details

[AlexWong122]

Did any one encounter the situation where there is a steel beam on top of header like the sketch below? The reaction force on the steel beam end is around 20 kips so the bearing is not a problem. This steel beam is located on main floor and uplifting is not an issue. The span is 16 ft and the factored moment is 100 kip-ft, which is too large for any PSL beam. I am using W 10 x 45 steel beam (due to the floor height and being conservative). The other end of the steel beam is properly restraint to a PSL post. This beam will have solid blocking @ 7' O.C. with joists that runs parallel to it.

By visiting the different sites I have seen some engineer just put the steel beam on top of the top plate without any connection. But is it necessary to restraint the steel beam from moving laterally or axially in case of earthquake event? What is common practice for this type of joint?

Similarly I have also seen steel beam rest on stud wall with a few studs underneath the beam. No bolting or any connection between the steel beam and the top plate.

I am still trying to figure out all the industrial standard for wood frame construction so hopefully it is not a stupid question.

 

[Ron247]

I have always attached them down to the wall. I have also fastened the top flange against rotation with a diagonal angle. Friction is a "connection" but not a good one.

If you have a 20' long beam that has the compression flange braced at the ends only, the unbraced length is 20'. If you do not attach it at all on the ends, it doesn't seem right that the unbraced length is still 20'.

 

[KootK]

For me, it's a pair of stiffeners and nice wide plate welded to the bottom flange that gets lagged I to the wood.

 

[WARose]

I've seen it several times and it's not a good detail. Even a lot of engineers don't realize you are violating AISC's assumptions about the support provided (for allowable forces).

 

[jittles]

Similar to KootK, if the beam depth is adequate, simply put a couple of holes right in the flange and lag through to the wood below.

 

[civeng80]

Thru- Bolting the beam could be difficult as the lintel may be as wide as the top plate. Maybe coach screws and a wide plate welded to the flange to get a vertical hole.
Kootk I think stiffeners is probably a bit of an overkill.

 

[CANPRO]

I think this question is too general to get any specific answers.

I've sized W8 beams for floors where there is no net uplift and not much of an end reaction - in this case, I would just provide nail holes in the bottom flange and specify the beam is blocked tight laterally.

I've also sized very large beams to support roof loads where there is significant uplift. In these cases I've had extra plates welded to the beam so it could be lag screwed/bolted to the wood (if it isn't on a steel post).

Alex - it sounds like this question has come up as a result of a specific design you're working on...maybe you can provide some more details and get some more specific advice. Beam span, beam size, end reaction (uplift?)...

 

[XR250]

In wood construction, there's usually enough other connections to keep things braced that I don't worry about it. If it is a roof beam with a bunch of uplift then I might think differently.

 

[KootK]

Kootk I think stiffeners is probably a bit of an overkill.

Let's hope so. A bit of overkill is my money makin', easy sleepin' sweet spot.

If the steel is coming from a real supplier as opposed to someone's scrap yard, then I consider the added cost for a couple of stiffeners and a pair of SDS screws to be negligible in the grand scheme of things. Additionally, I've not yet created an automated tool for evaluating web sway buckling at the bearing points in the absence of the stiffeners so I'm never to excited about having to do the math on eliminating the stiffeners.

As CANPRO mentioned, much is a function of context. If beams are shallow and almost as wide as they are tall, or attached framing convincingly locks rotation close to the bearing, it's a pretty safe bet that stiffeners etc are not required for web sway buckling. On the other hand, if it's a W14 or a load involving a lot of torsion, stiffeners it is.

One thing to be wary of, in my opinion, is that there is a stiffness compatibility problem when it comes to bracing steel with wood, particularly at it pertains to more local forms of buckling.

 

[WinelandV]

I don't think the stiffeners are overkill. Especially with the end stacked above the support, my default is to put the stiffeners in.

Per AISC 360-16 J10.7:

7. Unframed Ends of Beams and Girders
At unframed ends of beams and girders not otherwise restrained against rotation about their longitudinal axes, a pair of transverse stiffeners, extending the full depth of the web, shall be provided.

To say that it's adequately restrained against rotation, you'd need to show that whatever is above the beam is going to prevent that top flange from rolling. If it's supporting a floor, maybe you could justify it b/c of diaphragm action (assuming some positive connection to the top flange). I for one wouldn't want to start trying to calc out the rotation restraint (nor does J7.10 state what strength and stiffness is required to be considered "restrained").

If this is just a "I want to know what to do in the future if I run across it" question, I'd definitely add the stiffeners to your standard detail. If this is already in the field and someone needs a stamp after-the-fact, you may have to run down the restraint question.

 

[AlexWong122]

Thanks for all the inputs. I have updated the threads with more detail information.

 

[Larryhd2]

20kips of bearing on wood top plates?... with a 2x6 wall that's 363psi assuming even bearing, double that if triangular.

I don't know what lumber you're using but I think you need to spread this load out a bit.