Flexural behaviour of walls as a function of design approach

[BentEng]

Hi

I'm wondering what the difference is in wall flexural behaviour, stiffness, neutral axis depth etc if I design a lateral resisting wall in the following ways (AS3600), considering axial and moment loading:

1. Cracked section analysis using an equivalent stress block and lumped end tension steel, designing the wall as one whole element to Section 8 or 10.
2. Linear stress distribution using P/A + M/Z, and dividing the wall into segments for tension and compression, as shown in the commentary of AS3600-2009, and designing each segment separately to section 10 or 11.

What are the main behavioural difference? Is one more correct or more preferable to the other? Would this change the analysis, as in the stiffness modifiers for the wall?

I've heard conflicting opinions in the past about this that option 1 is more theoretically correct, as the concrete cracks and linear stress distribution doesn't occur. But I've also been told that option 2 more aligns with the linear elastic analysis model (etabs), and is gives a more conservative design.

Any input is appreciated.

 

[rapt]

AS3600-2018 does not allow 2 if the wall goes into tension! The new commentary will reflect this.

Not sure what you define as "lumped" end tension steel.

 

[BentEng]

Hi RAPT

I suppose this is a more theoretical question about how the wall would behave if you designed one way or the other.

I understand that the 2018 3600 states explicitly that walls are to be designed as per option 2, if detailed for ductility. In this case, I believe it's to avoid excessive strain in the reinforcement as the wall goes x2.6 into yielding.

But for the sake of argument, if the wall isn't detailed for ductility, say a pure elastic wind design, how does the behaviour differ? Is option 2 the better/only/code compliant approach?

And by lumped steel, I mean concentrated bars at one end, proportioned to take the full tension load, as opposed to a distribution of bars to match the elastic distribution of tension.

Ty

 

[rapt]

The new rule is not just for ductility, it applies to all sway cases. The 2009 rules were misleading, eg saying to use section 8 or 10 to design the wall. The only way I could see that 8 applies is if there is very small axial force, or it is a path to getting to Deep Beam design in section 12 but left the decision on which way to go to the designer who was supposed to realise a squat wall was a deep beam. The new code says 10 or 12 depending on H/L ratio to help out those who do not understand the flexural/non-flexural distinction and were designing squat walls as columns by section 10.

If the wall is squat, H/L less than about 2, it will act as a non-flexural member and should be designed by strut/tie methods (section 12) which is similar to what you are suggesting.

If it is flexural, H/L > 2 (there is a grey area around 1.5 to 3) then it should be designed as a column.

H in this is the total wall height, not the floor to floor height.

That is what the new rules are spelling out.