Partial Tensile Lap

[asabender]

A building has been constructed with less than the full required development length of exposed (set) dowel between a foundation wall and a structural slab. My past experience has been that the required tensile lap can be determined per ACI 12.2 and then the fraction of the normal design strength which is provided by the partial development can be determined: e.g. if a #5 bar requires 21.2" of development in 5ksi mix, than a #5 with 19" of development has [19/21.2= 89%] strength.

I have been trying to find a source for this, and have combed the forums to the extent possible, all to no avail. I have seen some instances of people saying exactly what I think to be the case, but never a source... Can anyone point me in the right direction?

Thanks!

 

[JAE]

ACI Chapter 12 has a provision that the tension development length can be multiplied by the factor (As(required)/As(provided).

Thus, if you have more steel area than required by analysis then you can reduce the development length (from which your lap is calculated).
However, you cannot reduce it lower than 12" as a minimum.

A tension lap is either Class A (1.0 x dev. length) or Class B (1.3 times dev. length).



Check out Eng-Tips Forum's Policies here:
faq731-376

 

[wannabeSE]

Are you looking for ACI 318-11 section 12.2.5 for tensile reinforcing in flexural members?

 

[asabender]

@JAE: Interesting. It had escaped my attention that you can simply use that to limit the connection capacity. In this case I don't have more steel than I need, but the tensile connection is relatively unimportant. This is negative steel at the perimeter, which was also tied downward to help with curling as a non-governing function. In this case As = As (req-flexure), but As > As(req-tension); furthermore, the flexural As achieves development in the slab. I think that this will work for me, as that is clearly the intent of the equation ... multiply the ld by T(req)/T(provided). I can run with that. Thanks!

@wannabeSE: yes, I was.

 

[KootK]

Are we talking about dowels that are embedded into the wall and then bent horizontally into the slab to deal with negative moment? If you're relying on negative flexure at that location, then you're not out of the woods yet. Any moment assumed to be present in the slab at the joint will also be present in the wall. And that means that your tensile connection is very important.

Below, I've sketched two versions of how this ought to be done. The first is the conventional way that most designers approach this. The second is the way that I think it ought to be. For what it's worth, most designers that I know will model this joint as pinned and only provide top steel for crack control. The reason is primarily quality control. Wet set = kinda bad; field bending = kinda bad; dowels getting run over by forklifts = kinda bad. All tolled... bad.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[asabender]

@kootk: What an excellent response. I agree in general, but would point out a few things. First, I'll sweep all of this into the realm of "theoretical discussion" by mentioning that this is a slab on grade, with no real moment... it's just over-designed (designed for settlement?). Second, I cant really go around presuming that they haven't taken the wall vs slab depth difference into account due to my unique roll in the situation. The EoR asked for thus and such an extension length of the bar along the top of the slab, and the contractor screwed up... I was asked, as the third party owners rep, to document fixes and propose solutions. I can tell them that they have an available tensile lap capacity of X based on the break strength of the concrete and the existing lap length, but I cant really (read: "don't want to") tell the EoR that they have way too much strength called out, and that they should calm down. I assume that they will take the contractor to task to some extent anyway, to make a point about not screwing up.

Thanks for the response

 

[TXStructural]

If the design relies upon the dowels to hold the wall to the foundation to resist sliding of the wall across/off the edge of the slab, it probably uses shear friction. ACI 318 requires full development for these dowels if shear friction is intended. (Just thought I would stir the pot a little.)

 

[KootK]

Well, thanks for being patient with me asabender. I totally misconstrued your situation. I should have picked up, by your mention of curling, that we were talking about SOG.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[asabender]

@kootk: not at all! thanks for the response.

@txstructural: that's also interesting, but the slab shouldn't see much lateral force aside from thermal which it should resist fully with a tiny fraction of the steel. i think that the real answer is the As/As req. they can then gracefully turn to their calculations check the design for allowable tensile reduction.

thanks again for the replies. it's always interesting hear the amazingly varied directions other engineers' minds move in.