According to 11.5.3 ACI318-99 "Stirrups and other bars or wires used as shear reinforcement shall extend to a distance 'd' from extreme compression fiber and shall be anchored at both ends according to 12.13 to develop the design yield strength of reinforcement".
mannyg - are you talking about shear across (perpendicular) to the reinforcing? If so, ACI does not offer any "steel" shear values, rather, the shear friction section that whymrg referred to is the way to go. It assumes the the reinforcing is extending across a joint or shear plane and allows shear to transfer across that plane via friction between the surfaces. The rebar that extends across the plane just keeps the surfaces together to allow the friction to occur. The bars MUST be developed to their full Fy on either side of the plane...thus you cannot reduce the development by Asreq'd / Asprovided.
In any case this question is pertinent also for stirrups, since at limit strength the stirrup are to bear -for the practice in Spain fyd (only up to 400 MPa fyk steel is used since to use higher strength would mean spoil of material, since no higher than such strength is allowed to be considered for stirrups). In all, a quite close to the nominal limit strength is to be anchored... and I would say this is not being generally respected nor by designers nor by fabricators.
Traditional practices used (maybe a bit less now) to fall short of the lengths required to properly anchor such stresses. Focusing at corners we would be able -more than once in a life- to see even cracked stirrups of the so sharp corners detailed. In much more cases, fully plastified regions at corners are present, then sometimes even welded, hopefully with welded rebar... and, er, with some times even visible sectionreduction of the main rebar!
It is for this surely that welded cages are not allowed in bridges, with the exception of non-critical meshes.
Some of these failures you can see as well in the detailing books.
So ample consideration by the code specifiers to these issues would be very welcome, from what my eyes have seen.
To clarify, I was referring to shear-friction action of a dowel bar. Jae, you are absolutely correct. I should have reviewed UBC-97 Section 1911.7.8 on shear friction which clearly states "...shear-friction reinforcement shall be anchored to develop the specified yield strength on both sides by embedment, hooks or welding to special devices." Therefore, the development length requirements for tension shall apply from both sides of the shear plane in the direction of the bar.
Since I do not fulfill these development lengths on one or both sides of the shear plane, I will simply reduce the shear capacity of the bar. Is that common practice?
You know, mannyg, I'm not sure. A very puzzling issue arose in our office a few months ago concerning this issue with developing rebar from a hollow core precast slab into a supporting concrete masonry wall.
ALL of the details that NCMA and others show across the U.S. usually indicates a bar grouted into the cells of the plank and then turned down into the masonry wall. The purpose of the rebar, of course, is to tie the floor/roof to the wall, taking lateral forces perpendicular to the wall as well as parallel to the wall.
The problem is that NONE OF THE DETAILS MEET CODE. (in our humble opinion)
The shear friction plane is located at the end of the plank and the bars generally extend out of the plank only about 2 to 4 inches before turning down into the wall. Therefore, the detail suggests that you are trying to develop a hook beyond the shear friction plane and your hook ldh is only 2-4 inches...which is obviously not the required amount for any bar.
I notified the NCMA on this issue and the response was that they would look into it at their next summit conference.
The way I read the code, you either meet the fy condition, or you have zero capacity in terms of shear friction. I don't see how you can reduce the capacity (this is what Asreq'd/As provided was doing...which is disallowed).
Can this situation be compared to an expansion bolt anchor?
If you grout (with epoxy) a Hilti anchor (using rebar) into concrete, you get a shear capacity, even though the embedded length is much less than the development length.
The difference would be that the HILTI epoxy is bonded to the concrete and you develop a pure shear cone effect. With simple embedded rebar, you are depending upon the bond stress between steel and concrete, not between concrete and epoxy and then epoxy to steel.
The zero (or diminished) capacity referred by JAE above also happen in Spain's EHE code in the negatives of beams in bending at the outer extremes...there's no embedment enough to develop 20 mm rebar -without complementary joint messing detailing- that then theoretically against decades of practice may need to be undersued, or lesser diameter bars be selected.
Joint design is still much to be worked out, so thes problems are likely to stay for some time, it is my view.
The RC buildings made the last decades are behaving well mainly on that they only have been facing vertical loads, if earthquake other thing might develop.
Development and slice lengths as per code stay 2 and more likely 2.t times those strictly required. With live loads usually present only in a percent of the capacity, the short as per the code embedments work well but for the real test in which the code requirement of limit strength becomes real.
Something of the same happens with excessive redistribution of negatives towards positives.
And respect masonry I had the experience of trying to support a 35 cm thick slab with I think about 7 m in the masonry wall and even without any wind loads it was not feasible to meet the code nor thick nor thin the wall would be (I think I discarded centering the load forcibly, don't exactly remember)... my conclusion was that the code is targeted to produce exclusively lightloaded shortspan structures (for masonry units), and make everybody look elsewhere even for normal spans.
Jae, look at this case. If I had two rebars, a #4 rebar that fully developed at the shear-friction plane and a #5 rebar that did not fully develop, both with the same length. To say, the #4 rebar has a shear capacity and the #5 rebar has no shear capacity is hard to swallow. Not to mention, the larger surface area of the #5 rebar will yield a higher tensile capacity at the shear plane assuming the bond stress in both cases were the same. What reasons would invalidate this rationale?
mannyg - I'm not a particular expert on this subject but I have a guess as to the philosophy that created it. One of the primary concerns in ACI 318 is the avoidance of abrupt failures in concrete (it is a brittle material, after all).
The concept of shear friction is that the steel you install must cross the potential crack, keep the crack closed, and therefore engage the friction between the concrete surfaces. If you use a larger bar, and do not develop its full fy, then the potential exists for an overload condition that does not allow a gradual failure of the system.
You know that they do this throughout the code...for example, flexural capacity (phi Mn) uses a phi of 0.9 while for shear (which is abrupt) they use 0.85. For columns (even more non-redundant) we have phi = 0.7.
In this case, you want the potential crack to open wide at first, giving ample warning of a problem. With your #4 bar vs. #5 bar example, the #4 bar, if overstressed, would enter the yield zone and allow the crack to open. You'd be warned and you'd run out and hire a structural engineer to fix it.
With the #5 bar, your unanticipated overstress would at first stretch the bar, but the limit state is now the bar pulling out of the concrete....and it never gets a chance to yield. BOOM
The commentary in ACI 318 section 11.7 talks through shear friction pretty well...but it doesn't really address this rationale.
The above is my "guess" that seems to make sense to me. What do you think?
