Up-set wood beam - LTB issue?

[jittles]

Hi folks, I've got a wood-framed project where an Architect wants a clean uniform soffit, so we're being asked to up-set some wood beams, leaving their top flange unbraced.

14" deep I-joist framing hangered to beams typically, bottom flush with beams. I'm fairly confident that a small amount of up-set will be just fine (say an 18" deep GLB), but in one case we require a 36" deep GLB, and we only have framing on one side.

Assuming the beam is braced, a 5 1/8" x 36" deep GLB works. Unbraced and it's at about 185% stress. Of course, this one is up-set into a 2x6 parapet wall, so the preference is to avoid going with a wider beam for now.

Thoughts on how to justify LTB? Perhaps a few tension ties to the joists along the length would restrain any lateral translation at some discrete points?

My default position here is that the beam needs to work assuming completely un-braced, and I have been unable to find much guidance on up-set wood beams in my searching.

(Blocking is shown but that is probably going to be a ledger rather than blocking)

Tension ties - single? double to give a bit of moment/twist resistance?

 

[KootK]

1) I suspect that this is probably fine in real life.

2) I agree, the 36/14 combo is getting up there.

3) I imagine that the hanger install provides some restraint but that feels funny to rely on sometimes.

4) A ledger and it's fastening to the sheathing and glulam probably also provides some restraint but, again, somewhat sketchy when one considers cross grain bending etc.

5) Both #3 and #4 get sketchier for me when considering a very large beam like the 36". Obviously, bigger beams require stronger bracing.

6) Certainly, I'd be fine with your tension strap idea. May you alternate a top side strap every fourth joist and a bottom side strap every fourth joist staggered. Then call your unbraced LTB length 8' which is surely fine for something so beefy.

I've had this same concern where my beam was a three ply LVL. That was much worse as, by the book, you have to consider that three independent 1.75" members rather than a monolithic 5.25" member. Barf.

HELP! I'd like your help with a thread that I was forced to move to the business issues section where it will surely be seen by next to nobody that matters to me:

http://www.eng-tips.com/viewthread.cfm?qid=456235

 

[jittles]

Thanks for the reply KootK, those are the same sort of "probably OK but a bit sketchy" feelings I have. I just don't have full confidence to OK this one just yet.

I've asked the Architect if he can stomach a slightly thicker parapet and go with a 6 3/4" x 33" GLB that works assuming unbraced. That little bit of added width helps immensely.

 

[JoshPlumSE]

Obviously, we're talking about stability effects. Lateral displacement of the beam due to high compression stresses at the top.

This isn't really the same thing as you have for AISC where you have a very discrete compression flange. But, it's at least similar in concept. AISC requires that we have EITHER lateral restraint for the compression flange (which you don't have) or that you're braced against twist of the cross section. That is something I think you can manage.... I think your detail with the strapped blocking is sufficient to justify restraint against twist.

Given that these beams are reasonably stiff against torsion, I think you'd be okay. Diaphragm, topping slab, and the joist to girder connection should all contribute to a pretty solid restraint. At least how I see it.

 

[KootK]

Here's another option to add to the heap-o-sketchy.

HELP! I'd like your help with a thread that I was forced to move to the business issues section where it will surely be seen by next to nobody that matters to me:

http://www.eng-tips.com/viewthread.cfm?qid=456235

 

[jittles]

I'd be more comfortable if I did some of the strapping to restrain twist along the whole beam length more regularly than just a discrete midspan bracing. Or perhaps if I had joists on both sides of the beam.

At that point, however, cost just doesn't seem to make sense. Buy the slightly bigger beam and save the labor hassle.

Architect said no big deal to the thicker parapet, so I'm going with the bigger beam today and I'll file this discussion away for future reference.

KootK, good point on the load promoting the counterclockwise rotation. That leaves the twist restraint force couple as the beam bearing into the joist top flange and then the strap on the bottom is nicely located for the tension part of the couple. Not a bad idea at all.

Thanks for the input folks!

 

[txeng91]

I've had this same concern where my beam was a three ply LVL. That was much worse as, by the book, you have to consider that three independent 1.75" members rather than a monolithic 5.25" member. Barf.

Just out of curiosity can you point me in the direction of the “by the book reference”. I spent a few hours looking into this the other day on a 4-ply. Unless I missed something this is what I was able to find:

NDS - did not see anything related differentiating with between solid and built-up beams

Some Canadian code/report I can’t remember- take with as a single ply, tests have shown that only about 30% of composite action is obtained for out of plane bending.

Woodworks Sizer- takes width of single ply

RISA (apparently in cooperation with woodworms) - takes width as full beam

My opinion- single ply is too conservative for 3 plys and up. The real answer is somewhere in between. Specifying adequate interconnection dowels through all plys should get you something close to the full beam width.

 

[wannabeSE]

How is the parapet above braced. Is it relying on the torsional resistance of the GLB that has LTB concerns? If so, this exasperates the problem

 

[KootK]

Just out of curiosity can you point me in the direction of the “by the book reference”.

Sadly I cannot. This is something that I learned myself here on Eng-tips and, unfortunately, I didn't document that which I found convincing.

HELP! I'd like your help with a thread that I was forced to move to the business issues section where it will surely be seen by next to nobody that matters to me:

http://www.eng-tips.com/viewthread.cfm?qid=456235

 

[jittles]

wannabeSE, yes, the parapet loads the top edge of the beam and relies on the beam for 100% of its support.

It's a pretty short parapet. It's not being used as a railing/guardrail at this location, so it's a minimal projection above the top of the beam. In fact, with the height of this beam, it's probably just a few plys of 2x nailed flat on top, not even tall enough to warrant a stud wall.

At different locations along this line, this is in line with other parapet/railings, hence the concern on thickness. At the other locations, however, the beams do not have this large up-set issue (much shorter spans, much shallower beams, flush in most cases).

 

[Triangled]

i agree with your default mode.... beam must work by itself.

i think you need to make the architect move on this one, or, their roof will move in ways they'd prefer not.

36" deep / 14 : deep is so large that i doubt it has any significant effect on compression face stability.

the failure mode, as i understand it, is a buckling phenomena, which i think of being not slow, and i prefer slow with lots of warning

the buckling mode would be the top compression face rotating clockwise to the right, placing the hanger nails in "pullout", and depending how its all connected, placing the ledger in cross grain tension