Ties for wall vertical reinforcement 9

[WARose]

I have a feeling what the answer is on this one......but I'll ask anyway ()......Does anyone know of anyway (via a code anywhere) that allows you to consider a single layer/curtain of vertical wall reinforcement as "tied' via the horizontal steel? The way I've always read ACI code.....you need to have 2 layers/curtains of vertical steel, with ties between them in order to call it tied as per code.

Basically what I have is: a wall with a single layer of vertical reinforcement (and a horizontal layer as well). I'd like to use that steel as compression steel (for a vertical load on the wall).....but it's not tied. It's a pretty short wall (4' high, 10" thick).

 

[JAE]

Basement walls and retaining walls are fine - but think of the immense numbers of precast or tilt-up walls out there that have axial load on them with a single mat of rebar.



Check out Eng-Tips Forum's Policies here:
faq731-376

 

[JAE]

...and no transverse ties...

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[haynewp]

I don't see how the second part can be correct, that if the longitudinal bars exceed the 1% then they have to be tied. This reads as if it is independent of the bars taking axial loads or not. The first part would seem to be now more in line with masonry.

 

[JAE]

Here's the question I sent to ACI:

[blue]In earlier sections of 318 (such as 318-11) there is a section 14.3.6 which allows a designer to avoid transverse ties in "vertical" wall reinforcement as long as EITHER of two separate conditions occur: 1. As required is less than 1% of Ag - OR
2. Where vertical reinforcement is not required as compression reinforcement.

This has been interpreted as a means for a designer to avoid transverse ties - especially in walls with a single mat of reinforcement.

However, 318-14, paragraph 11.7.4.1 seems to change this. It reverses the wording but fails to change the "or" to an "and" thereby now requires designers to have transverse ties in ANY case where the reinforcement is provided in axial conditions.

The words "longitudinal reinforcement" is not defined in the case of a wall - is it vertical or horizontal?

The condition "required for axial strength" is also rather vague.

Can you clarify if the millions of square feet of single mat reinforced concrete walls out there, with no transverse ties, are now in violation of 11.7.4.1 and a structural danger?

Finally - this appears to be a significant change in the code yet no commentary is provided and I find no discussion online or on your website about it. Is there a reason for this? Was the wording changed but no one realized the vast difference it makes in wall design? [/blue]

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[slickdeals]

@JAE
The attachment isn't exactly the point being discussed here, but I had reached out to ACI a while back about the issue of wall ties and attached is their response. I will be curious to see what they respond back to you.

 

[BridgeSmith]

Interesting reading slickdeals; thanks for posting.

The way it looks to me , by a strict reading of the code provisions, it would seem that the vertical bars cannot be counted as compression steel, since they are not restrained from buckling by transverse reinforcement.

However, as a practical matter in your situation, WARose, if nearly half the concrete in your wall breaks off, which is what it would take to allow the bars to buckle, the vertical steel will not be enough to keep the wall standing, anyway. So, from that perspective, you ought to be able to count it.

If that isn't an option you're comfortable with (I can't say I would be), if the heavy loading is only in discrete locations along the wall, you could consider local external confinement (anchor plates on the faces, bolted through the wall, etc.). If the load is uniform, testing the actual strength of the concrete may be the best way to determine if it's sufficient or not.

 

[haynewp]

slickdeals,

Your basement wall question is exactly what I was interested in.

 

[KootK]

if nearly half the concrete in your wall breaks off, which is what it would take to allow the bars to buckle, the vertical steel will not be enough to keep the wall standing, anyway. So, from that perspective, you ought to be able to count it.

I believe that there's a logical inconsistency in that statement. Since the tendency of the bars to buckle would be one of the causes of half of the wall breaking off, I don't think that we can claim that the bars can be relied upon to not buckle because of a perceived hierarchy of failure. There's interplay and feedback involved.

I do, however, share your sense that the concrete mass surrounding the bars would likely be enough to brace them against buckling. And I wish there was a formal way to check that. I've attempted it in the past looking at the force required to brace a bar and checking that the concrete splitting capacity exceeded that value. Unfortunately, it leads to difficult questions about how much concrete section contributes to the resistance and whether or not we should rely on it being uncracked.

When I get to 1000 stars, I get to meet BAretired for lunch. Current star count = 790. Just sayin'...

 

[JAE]

slickdeals,
Thanks for that posting. They appear to first reply on the concept of an assembly of bars near the end of a shear wall. I can see that situation and understand the need/desire to require ties at the wall ends.

Their other reply on the basement wall is consistent with 318-11 and earlier codes - they suggest getting below 1% by thickening the wall to avoid ties...but they then don't talk about 318-14 where even if you have < 1% steel - if you have bars for axial capacity you need to tie the bars.

Your inclusion of an older document shows that the key for ties is if the vertical bars have the possibility of yielding . This is at least a way to measure whether you need ties or not from the older codes....and perhaps 318-14.

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[bones206]

I might be in the minority here, but I don't see any difference in the 2011 and 2014 code provisions. Yes the wording has changed, but I don't see any difference in the intent.


My interpretation differs from yours in that you read it as a requirement triggered when a wall has axial load, but I read it as triggered when the vertical reinforcement is being counted on for axial strength by the designer. To me, that is a totally logical code provision because the bars will be seeing higher percentage of the transformed section stresses and could be susceptible to buckling with no redundancy from the brittle concrete component.

The 2014 version says "If longitudinal reinforcement is required for axial strength...". To me that means the designer is relying on the bar's compression strength to resist part of the load rather than the concrete alone. It does not indicate that it applies to any wall simply containing rebar that has an axial load. We just have to design single curtain reinforced walls using concrete strength alone, which is consistent with tradition. That's my take at least.

 

[BridgeSmith]

Agreed KootK. I approached that very poorly. I was thinking more of the effects of external forces, such as seismic damage, when I wrote it. If the stress on the wall causes buckling of the steel, because the concrete is not adequate to confine it, then it fails in spectacular fashion.

There should be a way to determine if the concrete confinement is adequate to ensure yielding of the steel, rather than buckling, but there doesn't seem to be any that we've found so far.

 

[slickdeals]

To me that means the designer is relying on the bar's compression strength to resist part of the load rather than the concrete alone. It does not indicate that it applies to any wall simply containing rebar that has an axial load. We just have to design single curtain reinforced walls using concrete strength alone, which is consistent with tradition. That's my take at least.

Expand that a little further and imagine a wall with a single layer of reinforcement (or two layers for that matter) and the reinforcement is primarily provided for flexural tension and exceeds 1%. You will end up with a wall needing ties in the compression zone because the code does not appear to give you an out if you are more than 1% (regardless of whether you are counting on the compression reinforcement or not).

 

[slickdeals]

You will all find this Q&A interesting as well.

I don't believe anyone is arguing that if compression reinforcement is needed for strength, then it must be tied. However, many shear walls have boundary zone reinforcement triggered by flexural tension and often can end up exceeding 1% measured in the boundary zone of the wall. A case in point could be an 8" thick wall with (2)#7 each end spaced at 6" o.c. (about 1.25%). Strictly speaking, the code would require this to be tied since it exceeds 1%.

 

[JAE]

bones206,
I disagree with you and here's why:

Take a wall you have with one layer of vertical reinforcement in it and with a situation where reinforcement is required for compression load AND you have LESS than 1% reinforcement.

Under 318-11:
The provision states that ties are NOT required if you are
[blue]less than or equal to 1%[/blue]
[red]OR[/red]
[blue]Vertical reinforcement is NOT required as compression reinforcement[/blue]

With that wall you have - since you are less than 1% the second provision doesn't matter. It says "OR". So if either one of those are true (not both) you can avoid transverse ties.

Under 318-14:
The provision states that ties ARE required if you are
[blue]greater than 1%[/blue]
[red]OR[/red]
[blue]Vertical reinforcement is required as compression reinforcement[/blue]

With that wall you have less than 1% and you satisfy the first provision - but there's an [red]OR[/red] in there - so you have to satisfy both provisions.
Since you DO NOT satisfy the second provision you must provide transverse ties.

So with the same wall situation - each code is a different result.






Check out Eng-Tips Forum's Policies here:
faq731-376

 

[bones206]

Ok, I see the distinction now. Thanks JAE for clarifying. But in the example you give, I think it’s proper that the code wouldn’t allow that design scenario.

I think the new wording clarified the original intent and seems consistent with that commentary snippet that you posted earlier.

 

[KootK]

I believe that I know the answer to the question of what the code intent is. And, like most things, it's best sorted by tracing it back to the theoretical fundamentals that underpin the provision in the first place. I'll never be able to prove this to any doubters so let's just call this the KKT (KootK Theory). I put it forward for consideration and discussion, not as a statement of fact. It's going to be a bit of a journey but, for those interested enough to tag along, it may be worthwhile.

1) Why are columns required to have 1% reinforcing? The reason is to prevent passive yielding of the bars under conditions of concrete creep. Effectively, you are providing enough steel that creep strains, and thus bar strains, are keep below the yield strain of the rebar.

2) I submit that the 1% requirement also serves as a demarcation point of a different, but related sort. Another way to limit creep strain is to simply keep concrete stresses low enough that you don't get much creep tendency in the first place. I propose that the load limit below which creep strain is considered small enough to prevent passive bar yield is thought to be the point at which a designer would require 1% reinforcing for the purpose of resisting axial compression.

3) I submit that the 1% requirement only applies when there are net axial forces applied to the cross section. Flexurually induced compression does result in creep in the compression block but it is not as severe as in true compression members because the creep strain is partially restrained by the horizontal shear connection to the parts of the cross section in tension and, even, lesser compression (strain gradient effect). I believe that this is why you can have incidental bottom steel near the columns of two way slabs that will be highly compressed yet still not blow out despite the lack of ties and 3/4" cover.

4) If you take the three points above, and apply them to walls in an identical fashion, I believe that it leads to the following, rational interpretation of when ties would be required within a wall:

A) When there is a net axial compression on the cross section AND;

B) when the net axial compression load is sustained AND;

C) when the resistance of the net axial compression load would require reinforcing in excess of 1%.

5) Applying this to some common situations, I get the following:

i) Basement & retaining walls. No ties. These elements fail A and C. They may well pass B as much of the load may indeed be sustained.

ii) Axially loaded walls where design for the sustained axial load alone would NOT require reinforcement in excess of 1%. No ties as this fails C.

iii) Axially loaded walls where design for the sustained axial load alone would require reinforcement in excess of 1%. Yes to ties as this passes A, B, & C.

iv) Shear wall compression zones where rebar exists primarily for load reversal induced tension and design for sustained axial compression would NOT require reinforcement in excess of 1%. No ties as this fails C.

v) Shear wall compression zones where design for sustained axial compression would require reinforcement in excess of 1%. Yes to ties as this passes A, B, & C.

vi) The last two were tricks. Technically, other than where seismic ductility is required, no shear wall compression zone would EVER require ties as it would fail the test for sustained load (B). Kinda by definition, lateral loads are not sustained. I know, CRAZY!! And yes, I realize that this would not be kosher according to code compliant design.

6) I believe that ACI provides an alternate method for determining the need for ties based on calculating the expected strain in the reinforcing. I feel that this method is inherently more rational than the 1% business. Unfortunately, it's much less production office friendly.

When I get to 1000 stars, I get to meet BAretired for lunch. Current star count = 790. Just sayin'...

 

[Guest090822]

Why not just analyze the wall as “plain concrete”? ACI does have a Plain Concrete section. You only need to calculate 3 points for a simple interaction diagram. You will get very little moment capacity though.

 

[bones206]

I think the ACI 318 committee can just cut and paste KootK’s post into the commentary for the next edition...

 

[Settingsun]

Can you clarify if the millions of square feet of single mat reinforced concrete walls out there, with no transverse ties, are now in violation of 11.7.4.1 and a structural danger?

Settle down, JAE. Structures fall out of compliance with each code revision. Assuming this change was intended, it's probably the step from good-enough to preferable on the code development journey. I know some would prefer the former to be the end of the line but that doesn't always happen.

I like to keep in mind that the ACI doesn't have the power to crash the economy, which is what it takes to cause the sky to fall on the construction industry.

 

[JAE]

Hey - just adding a little hyperbole to get their attention never hurts.
I'm not worried. Just pissed.

Check out Eng-Tips Forum's Policies here:
faq731-376