AS 11.5.2 b) - Plain concrete Walls on De or Ee Soil Classification

[tmac7285]

The relevant clause states: The simplified design methods for walls subject to vertical compression forces is b) not to be constructed on sites with soil classifications De or Ee, as defined in AS 1170.4, and where subject to earthquake design.

Is this essentially saying that stability walls can't be designed using this clause and should use say Cl 11.7.4 and be designed as columns, and that plain concrete walls can still use this clause provided that they aren't a stability element? Or is this clause trying to say that plain concrete walls are not appropriate at all for De or Ee sites and we should have columns/ligatures instead of plain concrete walls?

The part where it states 'and subject to earthquake design actions' could be confusing. We often don't consider blade walls (say 2000x200) contribute to building stiffness and don't include these in our stability designs and that the core/dedicated shear walls take all shear/moments related to the stability design, particularly when the slab system is a flat plate. However, these walls are still subject to the drifts from the earthquake and would still be subject to some earthquake action regardless of what we assume.

What are your thoughts?

 

[Retrograde]

I wrote this in another thread recently:

"My understanding is that the philosophy of the new edition of AS3600 is that you cannot just assume some walls are not part of the earthquake system and then ignore any lateral requirements on those walls. Although this has been standard practice in Australia for decades the code is moving away from this approach."

Having said that, if this is the intent of the code writers they should have been a lot clearer in their language. Maybe the new Amendment will fix the ambiguities.

 

[QSIIN]

Tmac,

Part of the issue is that (supposedly) section 11 was never intended to be used for isolated blade elements.

Unfortunately, 3600 doesn't clearly define the difference between a section 11 wall and a section 10 column. And as the D&C market has pushed leaner and leaner designs, thin&long blade columns have become very common and popular. In a building with several of them, the difference in relative stiffness between a core and the blade walls becomes small, so you can't just ignore them without thought.

If you only have a small core, but several 2000long blades, the combined stiffness of the blades might be enough to attract significant lateral forces away from the core, in which case they must comply with the new section 11 requirements, especially if they go into tension in your seismic analysis and you've assumed ductility. You might say "the blade columns will crack and the lateral load will go back to the core". Maybe this is true. But what is the consequence of an isolated blade failing in a lateral event. Catastrophic collapse?

Relatively small columns don't have this issue, as the lateral analysis will show that they don't attract much lateral load, small enough to ignore and throw back onto the core. They are usually more rigorously designed, so can probably handle even the full elastic seismic load they might attract (but still check for drift).

In my opinion, the answer is to proportion element size based on their intended function. If you don't want to follow the seismic provisions for walls, then they must be small enough not to attract the lateral forces (or detailed not to via slip joints). If they attract lateral forces they must be designed accordingly. If you need a long blade element for architectural proposes, or to reduce slab spans, or other reasons, be prepared for them to start attracting lateral loads you must design for.

 

[rapt]

I cannot see how this clause can be made more clear.

It specifically says that walls on sites with soil classifications De or Ee cannot be designed using the simplified wall design method in section 11 and must be designed using section 10 as columns!

Where is the ambiguity?

 

[rapt]

Retrograde

It is not "the philosophy of the new version of AS3600". It is standard earthquake design philosophy. It was the philosophy behind earlier versions of AS3600 also, ever since Structural Ductility and Performance Factors were introduced.

The problem in the past has been that most Australian designers have not understood earthquake design requirements and have not practiced them correctly. I will include myself in that group that did not understand (but I have not designed a building for 35 years so it was not so much of a problem for me), until I started reading up on earthquake design and the logic behind it.

The Australian philosophy tended to be that wind load governed strength so earthquake did not matter. And the vertical elements taking 90-95% of the sway effects were assumed to take 100% and the lesser elements were assumed to be gravity load controlled. And as wind design accounted for the full ultimate design loading from the wind the structure did not rely on ductility of the structure to account for 50-80 of the sway that happens after ultimate strength is reached as happens in earthquakes.

So the strength comparison justifying a wind design only was not valid, they are completely different design scenarios and both need to be checked.

If you want to design for earthquake with mu = 1 and Sp = 1, then you can design the old way. Because you are then designing both wind and earthquake for the maximum load they will experience.

Australian designers need to put their post graduation study hours to good use and study up on earthquake design philosophy. AS3600 is not going to teach that, it just defines limits. At the same time, they need to study up on fire design logic if they want to circumvent the AS3600 deemed to comply Fire rules as some appear to be trying to do these days without really understanding the underlying logic.

 

[tmac7285]

Thanks guys.

QSSIN,
I agree that with a small core, long blade elements will become a non-trivial part of the lateral stability system. I don't know exactly where the trivial vs non-trivial line lies, but in the building I am considering the blades will add <3% of stiffness to the cores currently considered. I think that this would be small enough to be negligible for overall building movements. The real answer is to run the analysis with these elements included and assess the magnitude of the forces attracted to the blades and design accordingly.

RAPT,
My initial interpretation of the clause is as you state; we can't use the simplified design for De or Ee soil classification generally. I assume I would then use Cl 11.7.4, if the elements are to avoid ligatures, noting that ligatures cannot be avoided for higher strength concrete (>50MPa).
The main point of confusion is why the writers of the code bothered to modify the main part of the clause of 'not be constructed on sites with soil classifications of De or Ee, as defined in AS 1170.4' with 'and where subjected to earthquake design actions'. If the intent is the first part of the clause, there is no need to state 'and where subjected to earthquake design forces' given we are meant to leave behind some older practices where the contribution of smaller elements are ignored in the stability design, particularly for blade walls. Essentially all walls are subject to earthquake design actions, so it seems redundant. Another engineer I spoke to interpreted the clause as that we can use the simplified method for De or Ee soil classifications provided it isn't a stability element(similar to previous practice), so perhaps it isn't as clear-cut as you suggest.

The Priestly and Paulay book on seismic design for RC does note the practice of not including columns in the stability design, if I recall correctly. I don't recall the book saying it was inherently wrong, though probably relied on the assumption of ligatures anyway.

 

[rapt]

tmac7285

Why would you use 11.7.4 and why do you have to assume something when the last sentence in the clause specifically says "Otherwise design as a column in accordance with section 10"?

The wording was done like that to try to cover all building types and all possible interpretations from engineers who apparently think they have done law degrees and try to read the code like a lawyer.

Not all buildings have to be designed for earthquake forces. This clause is in the walls section, not the earthquake section and can be used for the design of walls that are not subject to earthquake forces. This was especially the case before the latest earthquake code where there were areas in Australia where earthquake conditions did not have to be considered. And is still the case to a lesser extent with the new earthqauake code for smaller buildings that do not have to be designed for earthquake loads.

This clause is covering the case with those soil classifications and where the building (not just the wall) has to be designed for earthquake actions. As soon as the building has to be designed for earthquake actions, any wall in it has to also.

If we had not done that, then walls in buildings that do not have to be designed for earthquake forces would have still had to apply the more onerous rules.

 

[QSIIN]

Tmac,

If the 2000x200 blades are small enough that they only attract <3% of the lateral load (total or each?), is it even fair to say they are "walls"?
They sound like "columns" to me, that should be designed to section 10. If the gravity loads aren't particularly large, and they aren't high strength concrete, you might be able to take advantage of some relaxed reinforcement detailing.

A lot of the issues come from vertical elements that are vaguely defined as columns or walls. My thinking, especially considering the definition of structural walls in section 14, is that if an element is large enough to attract lateral forces, it's a wall and must be designed accordingly. Otherwise it's a column. The alternative is a wall that is intentionally slip jointed to avoid attracting lateral loads, such as a non-load bearing concrete partition/feature wall

 

[rapt]

`QSiin,

That logic does not hold either except in buildings with a combination of stiffer "walls" and less stiff "columns".

What if you have a moment frame without walls. The columns are all taking their share of the lateral forces. But they are not walls.

Just because a member takes lateral action does not make it a wall.

The problem is that the code has differentiated by giving different rules but cannot define what the difference is! Now that we have computers we would all be a lot better off if the code simply had a design method for members with combined compression/tension and bending that covered all cases, and then an allowance that certain of those members do not require restraint ties on every vertical bar.

OH, we do already have that in section 10. But then we also have a simplified wall design rule in section 11.

 

[QSIIN]

Rapt,

Tmac clearly discusses 2000x200 elements being ignored in a core/shear wall system, the discussion isn't about every possible framing system.

But even still, if the structure only has a series of 2000x200 "columns", they would attract the lateral forces and would have to be designed appropriately.My point being, the behaviour of an element isn't dictated by the name to which the engineer allocates it.

I think engineers need to understand that there's not a black and white line between a "wall" and a "column" or even a "braced" structure and a "moment sway frame". There's a whole grey are in the middle as you transition from one behaviour to another, depending on geometry and relative stiffness.

The main issue is that section 11 has previously given far, far better capacities for slender elements than section 10, and less reinforcement requirements to boot (2000x200 "walls" for example). Engineers and builders are used to the size and reo rates of these elements and are trying to find ways to keep achieving them to stay competitive by exploiting loop holes in the code.

It's all good and well to say engineers need to study more, and that the code isn't a guide/how to book. But unless the code is more clear and explicit, people will continue to exploit it.

I think the code has definitely taken steps to rule out walls that don't need to be designed for lateral loads, however some of these definitions are still vague enough for people to poke holes in them.

 

[rapt]

the behaviour of an element isn't dictated by the name to which the engineer allocates it

That is exactly the point I was making in my previous post! There should not be different rules for walls and columns, especially when the code does not define the difference.

I have discovered that no matter what wording is used in the code, someone somewhere will come up with their own interpretation that is completely at odds with the words in the code and the logical intent of those words but suits the way they want to interpret it. It is a no win situation for code writers with some users of the code. Some of the discussion above is a perfect example. The code is very specific in the case discussed that it must be designed as a column to section 10, but at least 2 people were saying they were using section 11 rules still.

The wall rules came from early British codes from what I can see and they were dealing with very low concrete strengths. As concrete strengths have increased over time, those rules have not been reviewed to ensure they are still logical. I know someone who compared a slenderness calculation for a wall as a column and it failed miserably. But the wall section of the code allowed it. Forcing 2 layers of reinforcement helped but still not enough to make it work by calculation.

Before the 2001 code, those same rules were limited to 50MPa concrete. Then it increased to 60, then 100 and now 120MPa. Without reviewing the simplified rules!

If you want an example of stupid code rules as technology changes, look at BS8110 minimum reinforcement rules for slabs and walls. Independent of concrete strength, it is .0013! That was a reasonably good number for concrete strength of 20MPa Cube strength, so 16MPa cylinder, which was what was generally used in the 1950's when it was written. The code still allows the same minimum now (pre Eurocode) with 90MPa concrete!

 

[QSIIN]

I agree with that: there should not be different rules for walls and columns. They are both compression elements with bending. The magnitude of the load, and slenderness are the variables.

Unfortunately though, the code clearly gives very different rules for each, and if the code will not explicitly state when to use one or the other, the engineer is left to make that decision themselves. And until it is more clear, 200x2000 high-strength columns will continue to be walls.

If the code can't be more explicit, is there a solution other than telling engineers to do better? Because a lot of them aren't listening/don't care.

I haven't had a chance to look over the new 3600 draft, but hopefully it's making some additional steps in the right direction.

 

[rapt]

QSIIN,

Unfortunately getting something out of the code after it has been accepted for a period is often very difficult.

About as hard as changing designers mindset over historic design approaches.

The important thing, and this is where the simplified wall rule is most problematic, is that a simplified deemed-to-comply rule should always be more conservative that then general rule, especially in a case where it is a strength/collapse issue rather than serviceability. In this case, it is not.

 

[tmac7285]

rapt,

CL 11.7.4 is designing the wall as per chapter 10 column design, but allows for no ligatures, provided that the conditions are met (low stress, low reo, low concrete strength etc). I'm not trying to be facetious, but why would I think I can't do that if it is not precluded specifically given that it is using the general design principles of chapter 10? I'm not trying to weasel my way around the code, hence the original question, which I wouldn't have bothered to ask if the intent was to get around the code to carry out previous practice.

I take your point about some buildings not having to be designed for EQ; the work I carry out essentially always considers EQ and I wasn't aware.

So in general, is your suggestion to increase the blade width to make for detailing with ligatures practical (say 250 minimum) and never have plain concrete walls (mesh each face only) on a De or Ee soil classification site? I'm trying to make decisions and make the case to my superiors as to why we need to change our practice, particularly for this project. I can't go empty handed and I want to have it clear in my head why we are changing our practice and back it up with as much as possible. Ultimately though, it's not my company and the over-arching decisions will be made by the owners.

In this building for instance, are you suggesting the core walls would need to be designed according to the principles in chapter 10 as it is a De soil classification and clearly subject to EQ actions, hence having core walls with ligatures all the way up the building for the full height? That seems excessive for lowly stressed areas towards the top of the building. Speaking to the New Zealanders in the office, they don't have ligatures in core walls for the full height when using N40MPa and N50MPa, only in boundary elements, granted they design to a different concrete code. I accept that for N65MPa and above there are serious concerns with the brittle nature of higher strength concrete (as was previously addressed in AS3600:2009 with additional confinement for high strength columns), which require ligatures without a doubt in N65MPa or greater walls designed with limited ductile or ductile response.

QSIIN,

Yes, these are essentially columns in that they are discrete elements and not long lengths, however the intent is to detail these as plain concrete elements with mesh each face currently on the drawings. The core is also relatively bigger hence any contribution to building stiffness would be small from these blades. At 200 thick, it's not really practical to have ligatures in them and it gets back to your original point of the AS 3600 not clearly defining what a wall vs column. As above, I'm considering increasing the thickness to allow for detailing down the track in the design process.

I've had a flick through the draft amendment, but the definition of wall vs column doesn't feature. Cl 11.7.4 is re-written though, but doesn't mention soil classification.

 

[captain_slow]

OP,

I think that 2000x200 elements should be treated as walls - they are slender enough to experience flexure and certainly stiff enough to attract a fair amount of in-plane shear so it's easy to see how the boundaries get fuzzy.

I know that there is somewhat common practice out there of calling these sort of structural elements "gravity walls/columns" and then excluding them from lateral analysis to design the core for full lateral loading. The reality is that whilst this sort of approach is easy and scalable (and gives a seemingly conservative core design) it is also not in touch with what actually happens to the building. These elements that are typically fully cogged into at least one floorplate attract shear by the very nature of being there and so they must be designed for the kind of design actions that they experience. If these checks are not carried out then you are simply faced with possibility of structural failure.

A good visualisation test it so include these elements into the lateral model and see if they go into flexure and how much shear they attract; if they do then simplified wall design formula from Section 11 cannot be used for them - it says so at the beginning of Section 11 (as you have already correctly identified). You have to use Section 10 and the only code provision that may potentially spare you from specifying ligs is that clause in the middle of the page at the end of Section 11 (apologies as I don't have a copy of the code on me atm). This includes core walls and all of the 2000x200 elements - this process is code mandated. If I were your external certifier I would query you very thoroughly on these blade elements.

One more thing to bear in mind are provisions of Section 14 - these have a very high impact on design and detailing of walls. If you need any further reason to abandon the Section 11 wall formula simply look at the beginning of wall-related part of section 14; if you want to use the Section 11 wall formula for design checks then you have to treat your structure as non-ductile... . Another thing to bear in mind - part of Section 14 that addresses limited ductile walls also explicitly prohibits use of non-ductile reinforcement. I know that great majority of reinforcing mesh available out there is class L so that would be another thing that I would be querying if I were your certifier.

As to where to draw a line between a column and a wall; I tend to use H/L = 2 as my line in the sand. That's where AS3600-2018 seems to draw the line between flexural-dominated and shear-dominated elements. I know there is also a line somewhere in Section 5 that differentiates between walls and columns.

 

[blihpandgeorge]

i agree with captain_slow in that i cant see how, in typical scenarios, a 2m x 200thk element can be called a column. Unless it is under a very tall storey of say 12-15m or next to a monster core, like at the bottom of tall tower, it is going to attract load. For a storey height of say 6m, it has an aspect ration of 3 which is not a bending element to my mind, and shear mechanism are going to dominate its behaviour. To my mind,this reinforces (pardon the pun) that ties are more likely.

back to the original question - i read the clause to require " is this clause trying to say that plain concrete walls are not appropriate at all for De or Ee sites and we should have columns/ligatures instead of plain concrete walls?"

FWIW : I am Aus based and having recently been thru some condensed earthquake engineering from first principles for a project in NZ, there is some very fundamental principles of earthquake engineering that are not widely understood in many Aus offices, and i can see how this is leading to confusion in the standards. To my mind, understanding these principles makes a massive difference of how to interpret the standard, a specific example is that the role of mu and ductility on the assumption of what they building is experiencing and that many seismic provisions are not necessarily for strength against a load but for confinement for ductility for load reversals etc. I know i will be preaching to the converted to some...

 

[tmac7285]

rapt,

Further to my other response above I've done a bit more reading and clause 11.2.1 specifically allows for clause 11.7.4 to override the requirements of clause 10.7.4, so the code explicitly allows for my previous assertion.

captain slow,

I agree it's worth seeing what the impact of including the the 2000x200 elements in the stability model and assessing the forces that they attract and using this as my base point for making detailing/design decisions. We have actually looked at designing several other jobs as non-ductile walls where the wind forces dominate by significant margin, so this is actually feasible in some buildings.

blihpandgeorge,

I agree, it will attract load unless specifically detailed to allow for the building movement, which we will not be doing for these walls typically. It's going to be a steep learning curve for structural engineers in Australia in general for earthquake design. All the information exists out there for us to improve our knowledge.

I think it's a bit of an issue how most university course ((back when I went through) did not spend almost any time on earthquake design. Speaking with my NZ colleagues, they spent years having the basic principles drummed into them. It doesn't help that most of the office grey beards came through engineering in Australia before earthquake design was really a thing. We're catching up with accepted practice elsewhere and headed in the right direction, but there might be a few painful jobs in the interim until a new 'normal' is accepted.

 

[rapt]

AS3600 Section 12 defines when a member is considered to be non-flexural. For a cantilever it is considered to be 1.5. It does not matter whether the cantilever is a horizontal member or a vertical member. It was pushed it a little further to 2 in the wall section. In the 2009 code, it was left to the designer to check the requirements of section 12 to decide between using non-flexural and flexural design methods "as appropriate".

The section 5 definition of 4:1 to differentiate between wall and column is purely for section 5 and relates to the effect of end conditions for fire. The bars in the corners lose significantly more strength under fire than the bars along a long face as the corner bars are heated from 2 sides.

It is assumed that at a 4:1 ratio or more, there are sufficient side face bars at higher reduced strength to overcome the corner bar effect.

if you look at the column tables, they are very dependent on column dimension because of this, whereas walls are assumed to have no reduction for corner bar effects.

Beams have a similar width effect compared to slabs in fire.