Shear transfer on short pier

[CDLD]

I've got a small column pier (9" high) sitting on an 8" elevated slab (dowels post-installed into slab).
Curious how others justify shear transfer at the slab/pier joint?
In a normal pier/footing case, I would develop the bars above/below the pier/footing joint and use the shear friction clauses in ACI.

Since I don't have adequate development length into the slab, would it be acceptable to assume the bottom tie transfers shear equally to the 4 pier vertical bars and design them as dowels into the slab (using concrete breakout clauses)?

Can you design the 4 vertical pier bars as anchors (with shear and tension component) and check for breakout from the slab.

 

[KootK]

Yeah, the short pier thing is an annoying condition

would it be acceptable to assume the bottom tie transfers shear equally to the 4 pier vertical bars and design them as dowels into the slab (using concrete breakout clauses)?

Meh, unless there is lots of axial here concurrently, I suspect that most of the shear would get delivered through the leading edge regardless of the presence of the low tie.

Curious how others justify shear transfer at the slab/pier joint?

1) Assume that all of the shear gets transferred at the leading edge.

2) Avail myself of the clamping that the flexure will provide at the leading edge.

3) Use shear friction methodology to design the rebar at the trailing edge.

4) Not sweat the code requirement for full development wrt SF reinforcing (yes, I'm naughty).

5) Do everything reasonable to get as close as possible to full development. Small bars, good epoxy, etc.

 

[CDLD]

Meh, unless there is lots of axial here concurrently, I suspect that most of the shear would get delivered through the leading edge regardless of the presence of the low tie.

Thanks for the reply Kootk.
OK, let's say the low tie doesn't adequately transfer the shear to the rear bars, can you still design the front bars as dowels?
So the front bars would be designed for breakout from the slab and the rear bars designed for the tension breakout (or bond failure) from the moment.

1) Assume that all of the shear gets transferred at the leading edge.

2) Avail myself of the clamping that the flexure will provide at the leading edge.

3) Use shear friction methodology to design the rebar at the trailing edge.

4) Not sweat the code requirement for full development wrt SF reinforcing (yes, I'm naughty).

5) Do everything reasonable to get as close as possible to full development. Small bars, good epoxy, etc.

Something like this?:
Let's ignore any contributions from axial load on the frictional resistance for now.
To be clear the numbers in this post are theoretical but directly applicable to a problem I am facing.

The bond strength of the dowels is problematic in this example.

 

[KootK]

OK, let's say the low tie doesn't adequately transfer the shear to the rear bars, can you still design the front bars as dowels?

That, nowadays, is somewhat of a matter of philosophy.

1) Historically, no. You would design the compression zone as a clamped surface with the rear dowels and flexure providing that clamping.

2) HILTI currently uses a European model in which the front dowels are used and true dowels.

 

[TLHS]

I had to do shear friction stuff with overlays and short dowels at one point. I ended up using the Palieraki et al method as one of my justifications. It's also what Hilti is leaning on now for short embedment dowel/friction stuff. It's pretty reasonable to use.

Interface Shear Transfer Design using the Hilti Method for Pos...

Research-based interface shear design for concrete-to-concrete connections

ask.hilti.com

I've also just looked at the adhesion component of shear friction in the past and if it deals with the load then I'm pretty comfortable being aggressive with dowelling.

Meh, unless there is lots of axial here concurrently, I suspect that most of the shear would get delivered through the leading edge regardless of the presence of the low tie.

Koot,

I agree with this if you're talking about shear friction and the real function of the system. I think the rotation from the bending moment will push load to the front. I don't think I agree in the context of what he was asking, which is treating them as dowels/anchorage. I think if we're going to just use dowel action by treating them as anchors then it's reasonable to use all of them, and it's a mechanism that would have to be overcome before actual failure.

I think in this situation you have two situations:
1. Loads are small and your concrete basically ends up being pretty much monolithic with minor cracking
2. Things start to crack, in which case your top tie is going to be the one that's really taking in load from the anchor rods at the top of the pedestal. Then you get a strut to the bottom tie and the bottom tie pulls it back again to the other side. From here it can be in either anchor.

It doesn't feel like a dowel breakout is going to be significantly more rigid that the rebar transfer mechanism in step 2, expecially if it involves bending in the dowel. So I think both case 1 and 2 would engage all the dowels. I basically can't picture a dowel scenario where breakout happens in the bottom slab without engaging a pretty significant load in the rear dowels as well.

 

[KootK]

I think if we're going to just use dowel action by treating them as anchors then it's reasonable to use all of them, and it's a mechanism that would have to be overcome before actual failure.

Trouble is, if you use the rear dowels as "anchors" as you suggest, then they'll have very little shear capacity because of he atrocious edge distance problem in the pier at that location. So, yeah, they are active at some point in the load history. The rear anchors will just be knocked out of the game before you get around to mobilizing the front anchors fully.

and it's a mechanism that would have to be overcome before actual failure.

It's not enough that both mechanisms have to be overcome before failure. They both need to near their peak capacity at the same point in time, otherwise it doesn't make sense to combine their capacities. This is similar to the fundamental String/Spring thing.

1. Loads are small and your concrete basically ends up being pretty much monolithic with minor cracking

Except for the significant crack that is the cold joint.

I basically can't picture a dowel scenario where breakout happens in the bottom slab without engaging a pretty significant load in the rear dowels as well.

Well no, it doesn't happen in the bottom slab unless there is an edge distance issue there. Rather, it happens in the pier.

 

[TLHS]

Ha, I actually deleted a sentence at the end of my last post talking about how now we just have to detail for breakout on the back anchors on the top but deleted it to avoid the sidetrack conversation

Trouble is, if you use the rear dowels as "anchors" as you suggest, then they'll have very little shear capacity because of he atrocious edge distance problem in the pier at that location. So, yeah, they are active at some point in the load history. The rear anchors will just be knocked out of the game before you get around to mobilizing the front anchors fully.

Yeah, I agree with this. I was reading your reply as being an internal load distribution issue rather than this. I personally tend to be pretty generous with throwing in some extra internal anchors for shallow embedments like this and then ignore those with edge distance issues. Adding extra anchors at this depth is super cheap if you're already doing a couple, so I've totally just thrown them in at 3xembedment spacing for things in the past when it makes sense. Even when I don't do a dense grid, I try to have a few bonus anchors for when they hit stuff or when they end up having put the thing in four inches from where it was supposed to be.

You can't necessarily do this here because it's too small but it might be reasonable to pull the vertical bars in a couple of inches so you can get load into the ties and then you're back at normal industrial pier design concepts for lateral loads.

It's not enough that both mechanisms have to be overcome before failure. They both need to near their peak capacity at the same point in time, otherwise it doesn't make sense to combine their capacities. This is similar to the fundamental String/Spring thing.

Agreed

Except for the significant crack that is the cold joint.

Actually, this is what I was alluding to. I was saying that the upper section above the crack is reasonably monolithic as a unit.

Well no, it doesn't happen in the bottom slab unless there is an edge distance issue there. Rather, it happens in the pier.

View attachment 2134

I agree with this.

 

[KootK]

Actually, this is what I was alluding to. I was saying that the upper section above the crack is reasonably monolithic as a unit.

Ok, I'm on board with that.

 

[KootK]

If I could do as I wished w/o fear of being stabbed, maybe something like this as a plausible high performer. Widianto-esque.

Only use in shear the dowels without edge distance issues.

 

[CDLD]

I don't see bars breaking out from the pier as an issue (at the slab/pier joint).
The way I see it, breakout is precluded from the diagonal concrete strut to the top tie.

Looking at this problem with some fresh eyes, I think you can break up the steps like this:

Top of pier mechanisms:
1.a. Check breakout of anchor bolts from top of pier, or
1.b. Use strut and tie model to transfer shear to ties and preclude breakout (at top of pier)

Bottom of pier mechanims:
2.a. Check breakout of front vertical rebar at bottom side of pier (likely very low capacity from small edge distance), or
2.b. Make use of diagonal strut and preclude breakout at bottom side of pier, or
2.c. Use KootK modified shear friction method with partial development of rear bars.

Embedded bars:
Check for breakout if bars used as dowels or check for tension if using shear friction.

 

[KootK]

2.b. Make use of diagonal strut and preclude breakout at bottom side of pier, or
2.c. Use KootK modified shear friction method with partial development of rear bars.

2b and 2c are not an OR condition but, rather, an AND condition. That, because taken together they form the basis of a single method: shear friction. And that takes us back to the top with the concern over partial development in shear friction applications.

I don't see bars breaking out from the pier as an issue (at the slab/pier joint).
The way I see it, breakout is precluded from the diagonal concrete strut to the top tie.

It's not an issue precisely because the diagonal concrete strutting that you mention represents the shear friction method rather than treating the vertical rebar as true dowels.

 

[CDLD]

2b and 2c are not an OR condition but, rather, an AND condition. That, because taken together they form the basis of a single method: shear friction. And that takes us back to the top with the concern over partial development in shear friction applications.

In method 2b, I am picturing the diagonal concrete strut bearing directly on the vertical pier bars at the leading edge (at the cold joint). The shear is transferred through the leading bar in direct shear. The leading rebar is then designed for breakout/pryout/etc. from the slab as a dowel carrying shear.
This mechanism would be very similar to how we design anchor reinforcement for shear:


I agree the concrete strut is also used in the shear friction model, however in this case the shear is resisted from friction from the normal force on the strut.

 

[KootK]

This mechanism would be very similar to how we design anchor reinforcement for shear:

Yup, that's why I mentioned Widianto above. His seminal paper was the genesis of all that I believe.

In method 2b, I am picturing the diagonal concrete strut bearing directly on the vertical pier bars at the leading edge (at the cold joint). The shear is transferred through the leading bar in direct shear.

This confuses me. I thought that you considered one of the benefits of the strutting mechanism to be that I did not use the vertical bars as true dowel and, therefore, did not create breakout concerns. Moving shear through the dowels = breakout concerns.

In the context of shear friction in flexural applications, some folks kind of consider the leading dowels to be a form of "roughening".

 

[KootK]

I agree the concrete strut is also used in the shear friction model, however in this case the shear is resisted from friction from the normal force on the strut.

What you have described is precisely how shear friction is interpreted in flexural applications. It's also precisely how we typically design shear wall cold joints where the walls are interrupted by the slabs.

 

[CDLD]

This confuses me. I thought that you considered one of the benefits of the strutting mechanism to be that I did not use the vertical bars as true dowel and, therefore, did not create breakout concerns. Moving shear through the dowels = breakout concerns.

In the strut and tie model I posted above (I think you call this the Widianto method), we pass the shear to the dowels and preclude breakout from the ties anchoring them.

Your comments on shear friction: the friction from the normal force on the strut, the drag tie, the partial development of the rear bar - I agree with you and picture it the same way.

In a shear friction model, the rear bar needs to be anchored adequately to resist the tension from the moment or the tension required to mobilize the shear friction clamping force (whichever is greater). In the case of a short embedment, it can be difficult to anchor the tension that is needed to develop the shear friction force, which is why I'm trying to explore this dowelling option, which I am hoping will transfer shear through anchorage only and the tension from the moment.

I may be off base here, but wouldn't breakout be precluded at both the top and bottom of the pier in this sketch?

 

[TLHS]

Man, everyone's sketches are so much better than mine.

 

[KootK]

I agree, @CDLD knocked that out of the park. I used to do nice sketches. Now it's just Bluebeam scribbles...

 

[KootK]

Your comments on shear friction: the friction from the normal force on the strut, the drag tie, the partial development of the rear bar - I agree with you and picture it the same way.

So, then, do you not also see the rear vertical bar as being necessary shear friction reinforcement? I've been harping on that aspect because your original presentation suggested that you could use this compression strut idea as a way to both avoid break out and to avoid the the need for developed shear friction reinforcement. And I don't think that the compression strut accomplishes the later of those two objecties.

 

[KootK]

In the case of a short embedment, it can be difficult to anchor the tension that is needed to develop the shear friction force, which is why I'm trying to explore this dowelling option, which I am hoping will transfer shear through anchorage only and the tension from the moment.

Oh, I get it. And I'm not suggesting that there is some great answer to this that I understand and you do not. Rather, I feel that there simply is no great answer to this and that most every approach is some kind of kludge. I feel that we do ourselves a bit of disservice when we fail to acknowledge that openly.

Actually, this is what I was alluding to. I was saying that the upper section above the crack is reasonably monolithic as a unit.

That nags at me precisely because I suspect that it is true.

In the extreme, I think that one could envision the action of one of these mini-pier things as just an expanded version of a grout bed. And that, obviously, would completely suck when you think about the mickey mouse, low capacity resistance available from a grout bed. Yeah, the mini pier surely is better than a grout bed. But it also lacks some of the competencies inherent in a larger pier.

 

[KootK]

In the strut and tie model I posted above (I think you call this the Widianto method), we pass the shear to the dowels and preclude breakout from the ties anchoring them.

Yeah... I'm a bit hesitant to start pulling at the loose threads of the Widianto method because:

1) It's really all that we've got to lean on when it comes to reasonable pier sizing and;

2) I fear that pulling at the threads will risk the overall integrity of the sweater.

3) It really has been a significant contribution, which I respect.

None the less, there are holes in the Widianto method. I mention a few below for sport.