Design of Walls - Simplified Method - Restraint of Vertical Reinforcement - AS3600-2018 1

[Drapes]

I refer to the new concrete code AS3600-2018 under "Section 11 - Design of Walls, clause 11.7.4 - Restraint of Vertical Reinforcement" - see attached excerpt.

With reference to the last statement, does this only apply for walls that are designed as columns, or for walls in general whether they are designed as columns or designed using the simplified method for walls?

The reason I ask is because the preceding statement under this clause relates to walls designed as columns, so unsure if this last statement is simply a follow up requirement, or if it should be read in isolation.

 

[Retrograde]

I read it as applying to walls in general.

 

[rapt]

Walls with concrete strength > 50MPa, like it says.

 

[milkshakelake]

I read it in isolation (i.e. walls in general) because the preceding statement has a clearly defined a and b. This isn't c, or preceded with "Exception:".

 

[Drapes]

Great, I thought that might be the case. Thanks everyone

 

[QSIIN]

The ambiguity is left there to lessen the number of engineers who notice it (as you have), and subsequently lessen the backlash at the code committee, all while covering themselves.

These changes aren't even listed in the front of the code in the changes list! Just a sentence here or there to the column and wall sections, but I'm sure once the contractors start hearing about these changes to wall ties and confinement, they're not gonna be happy!

But again, I'm sure many engineers won't interpret it the way as mentioned above, and their designs won't change.

And they're the engineers the D&C contractors will choose.

 

[Drapes]

Agree QSIIN,

I have already shown it to some of my colleagues who interpreted the last sentence as only applying to walls designed as columns given the preceding context - they justify it as being like a closing remark to the preceding statement - and they had me somewhat convinced as well. But as far as they're concerned if they have a gravity only col/wall (say 300x1500) designed using the simplified method for wall design, even at 65 or 80MPa they can design the col/wall without resorting to ligs (assuming all other requirements for using the simplified method are met of course).

For clarity and to avoid ambiguity, I believe the code should really say something like - "For walls designed using the simplified method for wall design or as columns with a concrete strength exceeding 50MPa..." instead of "For walls with a concrete strength exceeding 50MPa..."

 

[Drapes]

The new statement also references clause 14.5.4 for confinement requirements for strengths > 50MPa, which is the section for earthquake design.

Does that mean that for say a 300x1500 or 150x800 blade col/wall which will not attract earthquake loads (or only a negligible amount), there is still scope to use the simplified method for wall design and adopt > 50MPa concrete without any confinement steel? I suspect not, but was hoping to hear what others think.

 

[degenn]

11.7.4 is very badly drafted as the final paragraph appears to contradict (b) since the strength must be <50Mpa, if the last paragraph is read in isolation.

The reference to 14.5.4 should only be relevant, reasonably, if mu > 1.
If the design is based on mu=1, then the higher moment is probably not a problem if the column is in a core building. Presumably you can then design using the first part 11.7.4.

In any case, clarification is needed.

 

[Drapes]

Thank you for your thoughts degenn.

Based on my interpretation of the new code, and comments made herewith, I have prepared a brief summary on wall confinement requirements using 6x different wall scenarios. Still a little unclear on how to proceed with wall type 3 and 5 - I welcome any feedback on this, or if I have overlooked anything.

For the purposes of my summary below, "shear wall" refers to walls subject to both gravity and lateral loads, and "gravity wall" refers to walls subject to gravity loads only (assumes negligible lateral loads).

1. Shear wall - fc' > 50MPa and limited or moderately ductile (mu>1). Simplified method cannot be used (cl 14.4.4.1) so wall must be designed as a column. But must be confined irrespective since fc' > 50MPa

2. Shear wall - fc' <= 50MPa and limited or moderately ductile (mu>1). Simplified method cannot be used (cl 14.4.4.1) so wall must be designed as a column, however there is scope to avoid confinement if cl 11.7.4 is satisfied.

3. Shear wall - fc' > 50MPa and non-ductile (mu=1). Wall can be designed using simplified method or as a column. Unclear if confinement steel is required or can be avoided here irrespective of the method used, given fc' > 50MPa.

4. Shear wall - fc' <= 50MPa and non-ductile (mu=1). Wall can be designed using simplified method or as a column. If designed as a column, there is scope to avoid confinement if cl 11.7.4 is satisfied. If designed using simplified method, no confinement required by default.

5. Gravity wall - fc' > 50MPa and struc ductility factor (mu) irrelevant. Wall can be designed using simplified method or as a column. Unclear if confinement steel is required or can be avoided here irrespective of the method used, given fc' > 50MPa.

6. Gravity wall - fc' <= 50MPa and struc ductility factor (mu) irrelevant. Wall can be designed using simplified method or as a column. If designed as a column, there is scope to avoid confinement if cl 11.7.4 is satisfied. If designed using simplified method, no confinement required by default.

For walls 3 and 5 above, the code is not clear enough, although the initial consensus from this thread was that for ANY wall with fc' > 50MPa, confinement steel would be required regardless of the circumstances or the design method adopted.

 

[Drapes]

What are other peoples thoughts on this?

 

[QSIIN]

Drapes,

I agree there are still some holes if you look close enough.
One thing though, it's my opinion that you can't really have "gravity wall only" with mu negligible. 1170.4 states that any stiff element (concrete or precast wall) tied into your floor slab must be included in the lateral system and designed for accordingly, even if it attracts small lateral loads. And as such, if designed to chapter 11, you're whole structure is now non-ductile and must be mu and sp of 1.0 for global analysis, regardless of the rest of the structure.

What I'm unsure of is, if you check a compression element with pm diagram, considering slenderness as well, and are below 0.5phiNu (therefore can be without confinement reinforcement) is your element considered ductile? My gut says no. Which leads me to think unless you're below 0.15f'c you need confinement.

 

[Drapes]

Thanks for your feedback QSIIN.

Re: "gravity walls", I was more referring to walls where the max ultimate compression load would be dictated by gravity (eg load combo 1.2G+1.5Q), and not by load combinations involving equake (eg load combo G+0.3Q+EQx+0.3EQy). Of course, as you noted the equake loading would still be considered in the global anaylsis, but it would have little discernible effect and would not govern the design of the wall.

If this type of wall had to be categorized for equake though, I agree it would be treated as non-ductile.

 

[QSIIN]

Hi Drapes,

You're very welcome, and thank you for your thoughts.

I guess for a purely theoretical discussion, we could say gravity only, but I guess my point is that even if 1.2G + 1.5Q (or 1.35G, or 1.2G + 0.4Q + Wu) results highest load combination, they do not necessarily overlap with EQ, even if EQ stress are significantly lower.

If you're relying on your walls be ductile, and they are subject to some sort of EQ load, and are in excess of 0.15f'c, confinement may be required, regardless of your gravity/wind design.

 

[rapt]

Drapes and others

You do not design for the worst of Gravity/Wind loading and Earthquake loading.

Earthquake is a completely separate design and is very dependent on detailing to achieve the elasticity of the structure that you have assumed. Nu/Mu might be less for Earthquake than gravity after you divide by mu/Sp, but the problem is the building continues to sway further after Nu/Mu is reached. The actual actions will reach at least 1 / Sp times the design values and depending on the situation, especially with compression members, a much high load possibly up to the full elastic earthquake load.

That is why in some sections on wall design, we have required the check be based on mu = 1 and Sp = .75, and in one case mu = 1 and Sp = 1. For instance, under the design load after reduction by mu/Sp, a wall might be fully in compression, you you might think you can detail it for compression only loads. But if the wall is in compression, it cannot crack, so it continues to attract the full elastic effects. But the analysis and design assumption has been that the wall is very cracked and plastic hinges have formed and are capable of performing under overload.
If it is in compression still at mu = 1 and Sp = 1, then ok.

So in the clause in question, Nu should not be Neq / mu * Sp, it probably should be Neq. That is the one I mentioned we missed in the Opal tower thread.

The last sentence was meant to over-ride the conditions above. I still think it does. It will be changed so that there is no possibility of Code Lawyers reading it any other way!

It would be nice if designers recognized the intent of the code and its rules and try to meet that rather than try to pick holes it it.

The wall rules were introduced in AS1480 when the maximum concrete strength was 50MPa and highest used was much less in most cases. When concrete strengths were increased to 65 and then 100, the column rules were updated to consider the confinement requirements for the much more brittle concrete but the wall rules were not. No-one in his right mind would use 80MPa concrete in a compression member centrally reinforced. Unfortunately some have been. If you look at the stress strain diagrams for 80-100Mpa concrete there is no "plastic" section after the peak is reached. There is no softening. It is brittle and will not perform well in plastic hinges under sway unless it is well confined. And it cannot be confined in a 150-180 thick wall with central reinforcement. And it will not be confined by normal gravity load wall detailing.

RE Columns/walls in a building under earthquake action not attracting much sway/earthquake load.
If a vertical element is attached to the main sway resisting elements (e.g. core walls) via diaphragms (i.e. slabs) they all sway the same amount and have to be designed for that sway effect (clause 14.4.2). It is no use having a wall/column that you are not relying to resist sway failing due to sway and not being able to resist gravity loads! That results in collapse.

Just about all buildings taller than 3 stories and a lot that are lower than that in Australia are now required to be designed for at least a minimum earthquake loading. If someone has a wall with 50MPa concrete, that building is going to have to be designed for earthquake loading and detailing. Read the latest Earthquake loading code! Who in his right mind is going to use >50Mpa concrete in a building that is of so low importance that it now does not have to be designed for earthquake loading.

 

[Drapes]

Noted QSIIN, but my thoughts are why would you design these walls as ductile if you can design them as non-ductile and still find that gravity load combinations (or even wind load combos) govern? I'm talking about walls ("gravity walls") that don't attract significant compression load even under non-ductile conditions for equake (or under wind loads for that matter).

As I understand it, the only reason you would design the wall to exhibit any sort of ductility would be to reduce the design load on the wall due to equake, if for example the load attracted by the wall was far too high based on non-ductile parameters. In this instance, agree that confinement may be required to accommodate the resulting ductility.

 

[Drapes]

Thanks for your feedback rapt, appreciated.

Re: cols/walls designed for the sway effect (inter-storey drift), are you simply referring to the additional in-plane moments induced on the section due to sway? If yes, and having considered this sway effect, would there still be scope to avoid confinement steel for non-ductile and ductile walls with fc' <= 50MPa if:

-clause 11.7.4 is satisfied (where walls are designed as columns) - can apply to both ductile or non-ductile walls
-the wall is still subject to compression over the entire section (where walls are designed using simplified method, as per cl 11.2.1a) - applies to non-ductile walls only

 

[rapt]

Drapes,

As I said elsewhere I cannot preempt what will be the modifications to the code.

But the 11.7.4 (a) limitation is a problem with earthquake.
It is relating the need for confinement to Nu. Nu for a complete wall may be minuscule but under sway the stresses at one end could be very high. The rule was written initially on the assumption that you would be looking at small sections of the wall and the compression in each section. So it is really looking at the stress at the end section of the wall, not the average axial force in the entire wall in making this decision.

We will have to look at that for the next amendment as well as the mu and Sp values that must be used for the check.

If you look at the last paragraph in 11.2.1, it requires the check to be done for mu = 1 and Sp = 1. So non ductile.

 

[QSIIN]

Haha many engineers out of their right minds!

My understanding of 11.7.4 was for walls designed as columns, to plot an interaction diagram with 0.5phiNu and phiMu, taking into account slenderness, or in-plane moments (or both). And if you're below that lower boundary, then you can consider doing away with ties. as opposed to simply averaging loads over the ENTIRE wall (which I've seen done before) and comparing to CL11 phiNu for the entire wall.

Even so, the stresses at the extreme compression fibre could well be in excess of 0.15-0.2f'c, therefore requiring confinement for EQ, even if below 0.5phiNu on the Interaction curve.

Is that then to say, if you do an elastic model, with mu and sp of 2 and 0.77, and you haven't reduced wall stiffness due to cracking, and your extreme compressive stresses are below 0.15f'c, that would be the only case you could do away with ties? Or is that still considered non-ductile requiring mu and sp of 1?

 

[rapt]

Not as I understand it as you have reduced the ultimate design actions by a factor of 2.67.

If it is uncracked then it has to be designed for the full elastic earthquake action. The idea of the reduction for mu and Sp is that firstly it cracks and sways to the reduced strength limit and then plastic rotation and over-strength allow it to sway further. If it is uncracked it cannot form plastic hinges so you cannot make the reduction.

There is one area of terminology I have a problem with in earthquake design, the word elastic.

Action effects for an uncracked section are called Elastic. But that results in an inelastic ductile structural response.

A building that cracks and forms plastic hinges and is really well detailed is then said to have an elastic response!