Retaining Wall - Flexural Reinforcement from Stem Into Footing 21

[CWEngineer]

I am trying to get some clarification regarding the flexural reinforcement of the stem of a retaining wall into the footing.

Does the flexural reinforcement in the stem of a wall, need to be developed into the toe, such as show in Figure 1 of the attached document. Or is providing a standard hook (12db), sufficient, such as that show on Figure 2 of the attached document? If providing a standard hook is sufficient, can the hook be turned towards the heel?

Thanks in advance

 

[CELinOttawa]

https://www.concrete.org/publications/internationalconcreteabstractsportal.aspx?m=details&ID=19949

 

[Tomfh]

Standardisation of hooks in AS 3600

Thanks for the link. This includes some useful info alright. My take aways from it:

-Transverse bar does improve pull out capacity, especially for small round bars.
-One you are into real bars (say 12mm deformed), a standard hook with its corresponding small radius bend is sufficient to break a bar, even though we only use 50% of that in practice. Which raises the question - why bother with these big radius bends designed to limit bearing stresses if we know a tight bend will break the bar? Why not maximise the length of straight bar into the concrete?

 

[ATSE]

Kook has made excellent arguments.
Note that 2006 and 2010 Caltrans standard retaining wall details show larger radii for the stem to toe reinforcing hook. If you can look beyond the politics and disfunction of Caltrans, realize that they still have more than a few very good senior engineers and good standard details, with a lot of full scale testing. They obviously changed the ACI std hook detail with intention.

 

[Tomfh]

ATSE,

Do you know why they changed to large radius? What do their results show? Stronger joint? Better ductility? Etc

I'm interested to know why gentle bends are so beneficial when tight bends alone happily break bars.

 

[ATSE]

Tom,
Not sure about the 9" radius - seems very large. But 3 diameters (ACI 318 Table 7.2) seems fairly tight.
Kook and others above can articulate the "why" better. No doubt that concrete crushing stresses are lower at the inside of larger radius bend.

 

[KootK]

Note that 2006 and 2010 Caltrans standard retaining wall details show larger radii for the stem to toe reinforcing hook.

Thanks for sharing the Caltrans detail. It's really the first time that I've seen evidence of anyone in North America considering bar diameter in practice.

I'm interested to know why gentle bends are so beneficial when tight bends alone happily break bars.

Perhaps it's not a big deal. Certainly, I'm no expert on whatever body of related testing exists out there. That said, I see the following problems when it comes to applying the Wheeler results to the retaining wall condition.

1) Wheeler looked a 135 hooks rather than the 90 hooks that we use in retaining walls. That, presumably, was because Wheeler seems to me most interested in the rapid development of beam stirrups.

2) Because Wheeler's primary interest was beam stirrups, his testing regimen stopped at #6 bars. Concrete crushing inside bends is known to be more critical with larger diameter bars as one might encounter in larger retaining structures.

3) Wheeler did not test hook anchorage in tension zones. The area where stem bars enter footings is a flexural tension zone in many applications.

Probably the most important difference is simply this:

1) In a simple anchorage situation, you've got 1 x As X fy contributing to concrete bearing stresses.
2) In the retaining wall situation, you've got 2 X As X fy contributing to concrete bearing stresses because you're yanking on both ends of the bar.

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.

 

[CELinOttawa]

"In the retaining wall situation, you've got 2 X As X fy contributing to concrete bearing stresses because you're yanking on both ends of the bar."

How so? I do not see a difference... A developed bar cannot have double the stress in the bar. Are you sure you've thought this one through?

 

[KootK]

How so? I do not see a difference... A developed bar cannot have double the stress in the bar. Are you sure you've thought this one through?

Pretty sure. The point that I've been somehow failing to make since January is that the joint mechanics are not about rebar being developed. Rather, they are about rebar transferring tensile force around a corner. In the general case where entry and exit bar tension is unequal, some force transfer does occur via bond stress. That's just a small part of the greater story however.

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.

 

[CELinOttawa]

I think you are doubling your forces for no reason. If we are talking about whatever force the bar carries when entering the joint, I agree, but the way you show it is more analogous to a detailed pulley problem (where you include friction and other internal forces).

Whetherwe are talking about a developed bar or a bar who's total forces are going around a corner (at which point As isn't really representative), you always have equal and opposite reactions.

Yes, I understand the internal joint does take out some load in compression and/or bar development. I still think you have the wrong end of the bar on this subtle detail, but must commend you for your masterful treatment of the subject otherwise (above).

Going to give this one further thought...

 

[CELinOttawa]

I think this is what's happening:

And I think the fact that we are all so used to showing the stress in a bar as AsFy is actually standing in the way of a productive conversation. Unless I'm mistaken, I don't think you're saying that both ends of the bar carry AsFy, but that's what you showed. What you're getting at is that you see the corner "take out" the load. I think you're both right and wrong...

The corner is subject to the stress which the bars carry. The bars still manage to get around the corner, and they carry the load that they carry due to the applied actions. The applied actions can be thought of as coming from the Resistance, or (as I have shown it) from the Input Forces. While I agree wholeheartedly and enthousiastically that the stress in the concrete is roughly 2xFxCosTheta, I do not see that as being the forces carried by the bar.

Do you see the distinction? I appreciate that we are splitting hairs, but I think it is important. Also, you like a debate and I love to refine my understanding of Structural Engineering.... Thoughts?

 

[Tomfh]

I agree with CELinOttawa that the tension load is resolved by equal and opposite reactions. There isn't a doubling of the load.

1) Wheeler looked a 135 hooks rather than the 90 hooks that we use in retaining walls. That, presumably, was because Wheeler seems to me most interested in the rapid development of beam stirrups.

I agree this is a factor.

2) Because Wheeler's primary interest was beam stirrups, his testing regimen stopped at #6 bars. Concrete crushing inside bends is known to be more critical with larger diameter bars as one might encounter in larger retaining structures.

Is it? According to the AS3600 tests the bigger bars worked better. Look at the graph. The bigger bars killed it - even with a short extension length. It was the small soft bars that needed the cross bar.

3) Wheeler did not test hook anchorage in tension zones. The area where stem bars enter footings is a flexural tension zone in many applications.

I agree this is important.

 

[CELinOttawa]

Actually I now think Kootk is entirely wrong about the development into the footing. I'll do a sketch and send it through later today... There is no doubling of the force at all.

 

[Tomfh]

Actually I now think Kootk is entirely wrong .

Now I'm really confused. Didn't you think that already?

 

[CELinOttawa]

Not entirely! Lol... I thought there might be an influence of the bar continuing past the corner, but now I do not...

Here:

 

[KootK]

...but must commend you for your masterful treatment of the subject otherwise (above)

Thanks. Not sure how much of this still holds but, regardless, it was kind of you to say.

I agree, but the way you show it is more analogous to a detailed pulley problem (where you include friction and other internal forces).

Yes, the joint is in fact very much like that.

Unless I'm mistaken, I don't think you're saying that both ends of the bar carry AsFy, but that's what you showed.

That is correct. The tension on the vertical leg could potentially take on any value from zero to As x fy. Same goes for the tension on the horizontal leg. I would argue that, in many conventionally proportioned retaining wall situations, both the vertical and horizontal tension forces would approach As x fy (economical proportioning).

If you examine my sketch in greater detail, you'll see that I included a starred note at the top indicating that I'd shown a simplified condition and acknowledging that As x fy represented the upper limit of applied bar tension (equal utilization. I also said as much in the verbiage accompanying that post:

In the general case where entry and exit bar tension is unequal, some force transfer does occur via bond stress.

Let's not get hung up on the 2X business as that is not the salient point here. The salient points here, in my estimation, are these:

1) The concrete bearing stress that would accompany a standard, development only situation, would reflect a zero value for horizontal leg rebar tension.

2) Your typical retaining wall closing joint will have a non-zero value for horizontal leg rebar tension.

3) The non-zero value fro horizontal leg rebar tension in #2 will add concrete bearing stress in addition to that associated with #1.

4) #3 implies that, just because a standard hook is good enough to break the bars in a pure development situation, standard hooks may not fully address concrete bearing issues in a closing joint situation.

Do you see the distinction?

I'm afraid that I do not. As I see it, flexure in the footing demands horizontal bar tension at the joint just as flexure in the stem demands vertical bar tension at the joint.

Also, you like a debate and I love to refine my understanding of Structural Engineering....

I know it. You've been missed.

Is it? According to the AS3600 tests the bigger bars worked better. Look at the graph.

Not sure to be honest. I don't have any readily available research to support my claim. Nor can I remember where I picked up my assumption. Before proceeding further, I should mention that I only have access to a partial google doc on this: Link. I don't have the original proceedings document nor Wheeler's original research paper. If you're working from better information, do let me know.

The data that I have access to is shown below. To me, it appears that the larger bars did not outperform the smaller bars. Rather, all of the deformed bars tested (#4, #5, #6) reached the same ultimate stress. That being the breaking stress of course. Moreover, because the failure mode was generally bar tensile fracture, the tests don't really say anything about differences in concrete bearing stress. One would either need to measure the bearing stress directly or modify the tests to induce bearing failures in order to make inferences about the relative performance of the various bar and hook geometries with respect to concrete bearing stresses.

On January 14th, I referenced Klein's curved bar node work. On September 13th, CEL referenced the same. I've included a snippet from that document below which, in my opinion, pretty much says it all with regard to how and "exiting tension" exacerbates the concrete bearing stresses inside the rebar corner.

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.

 

[jayrod12]

Off topic post...

Discussions like these are why I cruise eng-tips. Seeing the heavyweights here discuss topics in more detail than a lot of us even knew existed.

I have now forwarded this thread to many people in our office when they ask about retaining wall stuff. I think it blows many of their minds.

 

[CELinOttawa]

Hey Kootk; My thanks. The compliment stands, just as a quick FYI...

As to the discussion at hand: I think you're conflating seismic and blast type work (ie: A proper frame, which can be closed by force that actually exists on both sides, and a simple anchoring situation, which is what I continue to believe exists here.

I may break out the hydrocode this weekend. You've gotten me very keen to know what an artificial physics environment might say about the stresses at the bar in a retaining wall. I suspect we may have just stumbled, or rather started to trample upon, something that our profession does not yet understand fully.

BUT: If you're right, why don't all those crappy, poorly detailed retaining walls fail? I think they don't fail because, while they are much poorer examples than would come from under your hand, they only really *require* the development of the bar to work in practice...

 

[KootK]

BUT: If you're right, why don't all those crappy, poorly detailed retaining walls fail?

1) In the wild, they probably don`t fail because of material and load factors and, in my estimation, the low probability of retaining walls actually seeing the design soil pressure. You know, the usual stuff that makes structural engineering an oddly consequence free space and makes it nearly impossible for clients to distinguish good design from bad. Additionally, retaining walls seem to usually experience overturning stability failures prior to material failures for whatever reason.

2) In the lab, the joints do fail prematurely. I posted a good deal of information on that above. Of course, if you don`t accept my arguments regarding the demands on the joints, you will also be unlikely to accept the relevance of the testing.

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.

 

[Tomfh]

Before proceeding further, I should mention that I only have access to a partial google doc on this: Link.

Yep, that's the one I was looking at.

To me, it appears that the larger bars did not outperform the smaller bars.

Regarding the specimens without crossbars, NONE of the SMALL bars broke, MOST of the MEDIUM bars broke, and ALL of the LARGE bars broke.

pretty much says it all with regard to how and "exiting tension" exacerbates the concrete bearing stresses inside the rebar corner.

Yeah but it doesn't work like that in reality. The bar isn't a rope draped over a pulley. It's a sticky bar bend around a sticky surface.

 

[CELinOttawa]

I see the research you've posted and it is interesting and compelling, but I do not believe it applies. Overall my position on this is becoming more entrenched and, I think, simpler...

Like a column that someone omits from a design calculation, but installs, which later results in a failure of a footing or beam below upon which it rests, you cannot decide you don't have a load path when you clearly do.

You want to treat the corner as a pulley, or at least as analogous to one... But like Tomfh has pointed out, it isn't a good comparison.

Once we have sufficient concrete around the full development length of the hook, the stresses have been transferred into the surrounding concrete, whether or not the bar then continues or does double duty in the footing. The load path is there, and the stress field will be created, unless you place bond breaking materials. That's why so many of those tests include bond breakers and jacketing tubes... Because they are only getting any result at all in their testing by forcing an artificial situation. The stresses actually transfer into the surrounding concrete, and then I believe they act as I outlined in my last sketch.