I think your answer is in the commentary to 22.9.4.6. When bending cause tension in the reinforcing, it also cause an equal compression force in the concrete. This compression force can be used for the shear friction. So, no extra reinforcing is needed, the reinforcing can be counted for both flexure and shear friction.

Quote (ACI 318-14)

Where moment acts on a shear plane, the flexural compression and tension forces are in equilibrium and do not change the resultant compression A_vff_y acting across the shear plane or the shear-friction resistance. It is therefore not necessary to provide additional reinforcement to resist the flexural tension stresses, unless the required flexural tension reinforcement exceeds the amount of shear-transfer reinforcement provided in the flexural tension zone

This is one of those instances where shear friction makes me a bit nervous. Your compression face clamping force, and thus your shear friction capacity, depends on the vertical distribution of earth pressure being as you assumed it for your flexural design. While I'm not often a proponent of shear keys, here I think that it's cheap insurance. That or the CRSI diagonal bars.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

While I agree with what you're saying KootK it's also worth considering that traditionally we ignore any cohesion/friction on the compression area of the stem. AASHTO's LRFD bridge code considers this effect at flexural joints and it results in a much higher allowable shear strength, and I believe it to be quite applicable to situations such as the above.

Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries
www.americanconcrete.com

@TME: I know that you're quite knowledgeable in the this areaa so I'll be curious to know your thoughts on the following. I'm not familiar with the ASHTO stuff.

1) Canadian code separates cohesion as well. You'd get 0.5 MPA here. I'd wager about 2.5 MPA here from rebar. Is a 20% improvement of about the right order?

2) Since cohesion is not just rough contacted surfaces but molecular bond, is there a greater need for quality control and inspection of the mating surface?

3)Is the nature of cohesion such that it can be present when there is some degree of gap between surfaces? At ultimate flexural capacity, there may be a millimetre-ish physical gap over 80% of the cross section. Differential stem shrinkage might also open things up a bit.

4) Do you know if the ASHTO stuff is based on different testing from 318? I believe that the basic ACI testing was done in a setup producing nearly flexureless shear and no shrinkage restraint. Pretty cohesion friendly.

I agree though, if one gets cohesion and didn't account for it in design, there would be a pretty healthy buffer against my original concern. I like it.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

AASHTO treats shear friction fairly similar to ACI 318 in most respects. The general equation layout is:

• shear friction resistance = (cohesion factor, c) * (shear friction cross-sectional area) + (friction factor, μ) * (A_vf*f_y + permanent net compressive force)

There is also two limits placed on the shear friction strength, one based on shearing of the area and the other on compressive strength. A phi factor is also applied as per typical LRFD fashion and a minimum area of shear reinforcement is required. The "cohesion factor" is actually a combined cohesion and aggregate interlock factor. f_y is limited to 60 ksi.

Quote (KootK)

1) Canadian code separates cohesion as well. You'd get 0.5 MPA here. I'd wager about 2.5 MPA here from rebar. Is a 20% improvement of about the right order?

The cohesion factor and friction in AASHTO varies depending on the surface cast against. For example, against a clean surface roughened to 0.25 in. amplitude you get a whopping c = 0.24 ksi (1.65 MPa) and a friction factor of μ = 1.0. Without roughening you get c = 0.075 ksi (0.52 MPa) and μ = 0.6, very similar to yours.

Quote (KootK)

2) Since cohesion is not just rough contacted surfaces but molecular bond, is there a greater need for quality control and inspection of the mating surface?

The few times I've used the AASHTO shear friction equation to it's full potential the EOR required inspections of the mating surfaces prior to the secondary pours. This seems reasonable and appropriate to be a special inspection item.

Quote (KootK)

3)Is the nature of cohesion such that it can be present when there is some degree of gap between surfaces? At ultimate flexural capacity, there may be a millimetre-ish physical gap over 80% of the cross section. Differential stem shrinkage might also open things up a bit.

I believe that cohesion would be lost if a gap opened up. Though, of course it would be hard to tell if that was just a gap at the exterior of the concrete or whether it was through-thickness. However, AASTHO is quite clear that their "cohesion factor" also accounts for aggregate interlock so I believe they're considering more than just the straight cohesive bond between the two surfaces. Thus, I would think a crack/gap wouldn't be an outright cause for rejection but would be cause for further investigation.

AASHTO seems to think similar as "brackets, corbels, and ledges" have a specific cohesion factor of zero due to the potential for a vertically oriented crack.

Quote (KootK)

4) Do you know if the ASHTO stuff is based on different testing from 318? I believe that the basic ACI testing was done in a setup producing nearly flexureless shear and no shrinkage restraint. Pretty cohesion friendly.

AASHTO cites the following for their factors: Loov and Patnaik, 1994; Patnaik, 1999; Mattock, 2001; Slapkus and Kahn, 2004; Hofbeck, Ibrahim, and Mattock, 1969; Mattock, Li, and Wang, 1976; Mitchel and Kahn, 2001. These appear to be different from the references given in ACI 318-14 but I have not reviewed any of the AASHTO research listed. Definitely something on my "to-do" list, though.

And, yes, AASHTO also has the same requirement that shear reinforcement be "fully developed on both sides of the interface" and gives no credit for excess reinforcement quantities. However, the not that bars with f_y > 60 ksi to have an assumed f_y = 60 ksi is quite interesting.

Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries
www.americanconcrete.com

Quote:

The 2008 CRSI manual contains specific “D’ Bars to resist shear at this interface, but I have a 2014 CRSI retaining wall design guide that does not show these bars.

That caught my attention. Vewy intewesting...

My favorite part of the shear friction provisions. If they say it is more accurate, why not include it in 318?

Quote (ACI 318-14)

Other relationships that provide a more accurate estimate
of shear-transfer strength can be used under the requirements
of this section. Examples of such procedures can be found
in the PCI Design Handbook (PCI MNL 120), Mattock et al.
(1976b), and Mattock (1974).

"It is imperative Cunth doesn't get his hands on those codes."

Archie - not sure why they removed those bars. They provide the most predictable resistance due to the addition of cosine(45)*Fy to the shear-friction equation. Maybe, because they didn't want to imply that the code required them at all times.

"It is imperative Cunth doesn't get his hands on those codes."