Torsion in Dapped Beam

[JustUseSteel]

I have a precast, prestressed inverted tee beam with significant torsion. It needs to dap onto another beam to receive it. Does torsional reinforcing need to extend into the nib? I believe it should, however my senior has told me otherwise - the crack will form after the re-entrant corner, so stirrups "before" the bearing pad are not needed / would not engage.

I have not been able to figure out in my head whats happening, conceptually. Given the nib/ledge have a depth - say 6" - we must assume the bearing stresses are somewhat uniform over that 6" x nib width section. So torsional stresses would be maximum at the re-entrant corner, and fade over the bearing pad. But would that stress/crack engage a stirrup within the nib?

 

[bugbus]

I would not count on a half joint like that transmitting any significant amount of torsion. I would have thought it should be modelled with the torsion degree of freedom released.

 

[JustUseSteel]

I would not count on a half joint like that transmitting any significant amount of torsion. I would have thought it should be modelled with the torsion degree of freedom released.

How would you accomplish that? Not sure I understand.

In other cases where the "dapping beam" has torsion, I am using connections from the beam to the precast slabs to remove it; therefore the beam itself does not experience torsion. However, there are a few cases where it is not possible use that type of connection. This beam is one of them.

 

[bugbus]

I would suggest there needs to be another way for that torsion to be resolved, other than through the half joint.

It would be really unusual in my opinion to rely on a joint like that to transmit torsion, unless the torsion were relatively small and only resulted in a slightly greater compression through one side of the bearing pad compared to the other. If the torsion were large enough to cause decompression through the bearing pad I would abandon that idea altogether.

Either way I think you would need to design this by strut-and-tie.

 

[JustUseSteel]

Yes, its significant enough to cause decompression in the bearing pad - something like 120k*ft of torsion and 140k reaction. Senior proposed using a welded connection at the top of the nib and a mechanical connection at the bottom to create a couple and remove/reduce the bearing eccentricity. That all sounds nice but... that to me means the torsion absolutely found its way through the nib, and therefore it must be reinforced for the torsion. (If even feasible)

 

[JustUseSteel]

Any other thoughts on this?

How it's detailed now, the dap joint is grouted all around, so the torsion would be resolved "at" the dap via a compression couple on either side of the beam cross section. In theory then, this is similar to a fixed end beam. It shouldn't allow significant torsion to exist over the dapped surface.

 

[KootK]

FWIW, I've been a part-time precast engineer for spells during my career. And most of that work was in WI. I've got a pretty good sense for what is common / practical in the midwest US.

In my opinion, the general handling of torsion in these systems is the shameful secret of the industry. Thoughtful EOR's know that there's a problem but leave it to the precasters so as not to hold up their projects. And precasters mostly turn a blind eye to torsion because they can't be seen to chicken out on work that they've bid when their competitors would let it fly and problems don't seem to manifest themselves.

Yeah, one can tell some creative stories about how erection techniques limit torsion and some / all of it can be resolved into plank flexure. Much of that strikes me as sketchy, however. And, unless someone has been hiding a mountain of relevant information that I've been unable to find, nobody has a rigorous way of quantifying these benefits.

Case in point would be what you describe below. I don't buy that for a minute given the amount of torque coming out of the dap relative to the stiffness of the diaphragm that would be needed to form the resisting couple. More to the point, even if the stiffness is there, good luck proving it. It's really just a bunch of wishful thinking from folks who have been cornered into no having any decent options with which to resolve the problem.

How it's detailed now, the dap joint is grouted all around, so the torsion would be resolved "at" the dap via a compression couple on either side of the beam cross section. In theory then, this is similar to a fixed end beam. It shouldn't allow significant torsion to exist over the dapped surface.

 

[KootK]

In my mind, the proportions of the dap will be such that you've no hope in hell in developing anything resembling a conventional, spiral torsion crack. Instead, I'd recommend treating the dap as a pair of arms that resist the torsion as a weird shear couple thing. This is really to transmogrify St.Venant torsion into a version of warping torsion locally. My sketch sucks something fierce but I wager you'll get the idea. The "arms" slide past one another vertically, they don't separate laterally.

When you imagine little STM models of each of the arms:

a) The arm pushing down into the support probably has the tie that it need via conventional dap detailing.

b) The arm lifting up from the support may lack its proper tie. It wants to have top steel that is localized near the coil rod and, most critically, run beneath the anchor plate and properly anchored past it. This, in my experience, represents some atypical -- but doable -- localized reinforcing.

Don't expect this method of evaluation to yield a torsional capacity anywhere that of the un-dapped beam beyond though.

 

[JustUseSteel]

FWIW, I've been a part-time precast engineer for spells during my career. And most of that work was in WI. I've got a pretty good sense for what is common / practical in the midwest US.

In my opinion, the general handling of torsion in these systems is the shameful secret of the industry. Thoughtful EOR's know that there's a problem but leave it to the precasters so as not to hold up their projects. And precasters mostly turn a blind eye to torsion because they can't be seen to chicken out on work that they've bid when their competitors would let it fly and problems don't seem to manifest themselves.

Yeah, one can tell some creative stories about how erection techniques limit torsion and some / all of it can be resolved into plank flexure. Much of that strikes me as sketchy, however. And, unless someone has been hiding a mountain of relevant information that I've been unable to find, nobody has a rigorous way of quantifying these benefits.

Case in point would be what you describe below. I don't buy that for a minute given the amount of torque coming out of the dap relative to the stiffness of the diaphragm that would be needed to form the resisting couple. More to the point, even if the stiffness is there, good luck proving it. It's really just a bunch of wishful thinking from folks who have been cornered into no having any decent options with which to resolve the problem.

Yes I've seen this as well. I believe we are one of few designers who try to account for torsion within the precast industry. The apparent lack of issues with our competitors who don't, though, seems to imply their logic is sound and the torsion is being resolved thru other mechanisms than within the beam/at bearing surfaces. In fact much of the time we add haunches for overturning moments, the architects complain that "no other precaster does this, whats going on?"