I don't think I've ever checked shear friction at a monolithic joint
But we always provide top and bottom steel from beam into column, so it's never going to be an issue

Our code gives interface friction mu = 1.4 at monolithic interfaces with a section change, so you get 1.4 x As x fy = a very large number for your capacity
You compare that to your vertical shear reinforcement etc in your beam and I can't imagine it would ever be a problem for any realistic beam configuration

I could see it being important for some sort of transfer beam supporting a wall above as shear at the beam ends could be huge relative to flexure

This was a case where an intentional crack was induced, but where a potential crack would likely form anyway due to shrinkage restraint.

thread507-521962: Concrete Beam Repair

I think the key is detailing bar anchorage to be fully developed into the support, which isn’t always feasible or required by other parts of the code.

Note the requirement for shear friction for the reinforcement to be fully developed. Traditionally - bottom reinforcement is not fully developed at column supports.

I do not check shear friction at every monolithic pour. Agree with Greenalleycat that it is likely a better check at specific conditions and locations

Quote (Greenalleycat)

But we always provide top and bottom steel from beam into column, so it's never going to be an issue

I also provide top bars at the ends of simple supports. This was the way I was taught to detail and to me it's for the convenience of being able to hang the vertical stirrups. I believe there have been discussions here about providing a minimum 1/3 +Moment reinforcing as top bars, but I've searched through ACI 318-19 and cannot find anything on it. Simply, I'm not sure why we are putting top bars other than a) old detailing habits b)place to hang stirrups.

Now correct me if I'm wrong, simply putting the same bottom and top bars really "changes" the intended design of the beam. If you put equivalent bars, say (2)#5's top and bottom, you've really just detailed a beam with fixed ends, and not a simple span?

Quote (EZBuilding)

Note the requirement for shear friction for the reinforcement to be fully developed. Traditionally - bottom reinforcement is not fully developed at column supports.

I understand both sentences, but they seem to contradict each other, no?

If we design for shear reinforcing, are we not assuming the bar to reach fully yield? Therefore, the bar portion that extends into the column support is never fully developed in situation where it extends the 6" code minimum into a column.

Nowhere here does it say anything to do with shear friction:

Yes that contradiction is exactly my point - extension of bottom reinforcement at a column may provide some dowel action, but it does not provide shear friction.

Can we reduce the development length requirement by the amount of excess steel provided?

Good luck navigating this section:


Is basically says, yes you can reduce development length. But for every situation at the end of beam (i.e. non continuous supports) or where a column doesn't allow at least 12" development, you cannot reduced the development length into the column. I take it as, for every location where shear friction is to be developed at the end of a monolithic beam pour, you cannot realistically develop said shear reinforcing unless you have columns that are deep enough. For 90% of typical construction, this is not possible.

I am 95% sure that the shear friction provisions require full development of fy on each side of the section in question. Thus, no reduction in development length allowed.

Please note that is a "v" (as in Violin) not a "y".

Quote (StrEng007)

For 90% of typical construction, this is not possible.

You typically wouldn't need to rely on shear friction of the longitudinal bars for shear transfer at the supports of a concrete beam. We have other ways of taking care of that (i.e, Vc + Vs).

There may be unique situations like the thread I linked to earlier where you would need the shear friction dowel action of the longitudinal bars.

Quote (bones206)

You typically wouldn't need to rely on shear friction of the longitudinal bars for shear transfer at the supports of a concrete beam.

I understand and I'm not trying to be argumentative (quite the opposite actually, I really appreciate your help) but how do you know when you aren't suppose to rely on it? The way the code is written, it's so vague. They specifically say "existing or potential crack". As far as I'm concerned, you cannot eliminate any potential crack in concrete.

I like when these kind of topics come up because they make you think and challenge assumptions, which is great brain exercise and provides an opportunity to review fundamentals underlying our design practices. I think the code is meant to be vague here, because it's not necessarily mandating that the shear friction model be used. It's saying if you are going to use this method, these are the rules of the road. At least that's my interpretation. It's up to the designer to judge when it's applicable.

Shear friction theory is just that, a theory. It is a set of arbitrary rules loosely derived from testing.

In theory, any arbitrary plane cut through a concrete structure should satisfy a shear friction check. That's not to say we need to check every conceivable location.

Recently I had to do a deep dive into this whole thing, and found that when checking certain interfaces either through monolithic concrete or roughened construction joints, these will occasionally suggest a failure especially across unreinforced or lightly reinforced planes. The more places you start looking for potential failure planes, the more you will find, and I don't believe it is necessarily realistic.

I think the reason is that mu = 0.9~1.4 used in various codes for this situation is quite conservative. There is a meta study of hundreds of different shear tests called "Examination of the effective coefficient of friction for shear friction design" by Krc et al. 2016, which is an interesting read. The actual tested value is anywhere from mu = 1 up to around 7, with an average value in the order of 2-3.

Quote (bugbus)

In theory, any arbitrary plane cut through a concrete structure should satisfy a shear friction check. That's not to say we need to check every conceivable location.

So where do you draw the line? Where do we say a crack might happen here, so we need to satisfy the shear friction theory, and where can we dismiss it?

Quote (Tomfh)

So where do you draw the line? Where do we say a crack might happen here, so we need to satisfy the shear friction theory, and where can we dismiss it?

Typically at transitions from one element to another or at a change in section.

Quote (Deker)

Typically at transitions from one element to another or at a change in section.

Agree. But the codes say any “potential crack” location, which is pretty much anywhere in a monolithic concrete structure, eg a slab.

Yes, one could interpret that to mean any conceivable plane, but that seems unrealistic.

The “shall apply where appropriate” bit implies using judgment about when and where to apply the method.

I, and others, did massive amount of work on this back in 2014 that would be worth your time to review: Link

As a teaser, the thread includes the sketches below and some accompanying numerical work.

More from this thread: Link. Seriously, I probably lost two years of my life obsessing over shear friction. Frankly, it's a wonder that I'm still married.

Quote (Str007)

Does everyone do the shear friction check at monolithic pours?

Based on my travels, this is where I have landed on this:

1) Any plane you might draw anywhere on any member probably has to satisfy some version of shear friction. It's just equilibrium. The relevant question is whether or not its a version of shear friction that warrants evaluation.

2) Practically speaking, I don't believe that it is possible to generate a shear friction failure across any plane that is simultaneously:

a) Monolithically poured and;

b) Properly detailed for flexure at the same location.

This makes proper rebar detailing all the more important. You know, as if it weren't already FFS. Mission critical.

3) You can generate direct shear failures on planes parallel to the shear being considered. This is a different phenomenon from shear friction, however, and comes with a grossly higher capacity (10*SRT(f'c) in ACI parlance).

4) Hokie66 is right about everything and always has been. It's super annoying.

With reference to the above, it's useful to consider how that plays out in relation to modern hanger bar requirements for beam to girder connections:

1) There is no code requirement to check shear friction at such joints and, in my experience, no one does.

2) You probably will get a crack at or near the support interface.

3) We're cool with the crack so long as the joint is properly reinforced for flexure which is done by providing the required STM hanger reinforcement.

And since it's October, a pic of my 2014 pumpkin.

If you have ask, you're not qualified to ask.