Earthquake Analysis & Design to AS3600 & AS1170.4

[GRADSTRUCT]

Hi All,

I know similar topics have been discussed a lot here, and I've tried my best reading and understanding these threads to understand the intent of the different WALL/COLUMN design clauses in these codes. Could someone please let me know if my understanding on how to design a less than 4 Storey building is correct or make any suggestions? Also disclaimer, I'm not trying to find a loophole in the code to reduce my reo/tie requirements - this is just my understanding of the clauses.

For Context, I use Inducta - RCB to do my lateral analysis.

1) First, I model the full structure with only my shear walls and core walls fixed-fixed with everything else pinned-pinned. Based on this analysis (with ductility factor 1), I check my horizontal displacement, the shear, moment and axial capacity of the shear/core walls ONLY. If all okay, I proceed with ductility factor 1. Or else, I increase to ductility factor 2, check if axial stress in these walls are less 0.2f'c, and add the necessary boundary elements for my shear walls.

2) Then, I re run the model (with ductility factor 1) with all my other vertical load bearing elements as fixed-fixed and design my columns for these conservative forces. And all my blade walls I check if any of them have tension in any section by checking P/A +- M/Z. If any of them are in tension I design them as columns and see if I can get them to work with only mesh both face. By keeping f'c < 50MPa and my vertical reo ratio < 0.01Ag I could get away with the tie requirements. The rest of the blade walls I design to simplified method in Sec 11.

3) In the end, I would have designed the shear walls for ductility factor 2 forces (in plane shear capacity of these walls to ductility factor = 1) with additional detailing for ductility. Columns as fixed-fixed for ductility factor 1 or 2 forces. Blade walls in Tension as Columns with only mesh with ductility factor 1 forces. Blade walls in Compression also with only mesh with ductility factor 1 forces. Also any squat walls (one storey walls) as only mesh with ductility factor 1 forces.

Is this correct? If I model every column and blade wall as fixed-fixed, my design would end up being very conservative. If my columns and blade walls are precast and I only have a dowel connection, I'm essentially forcing these vertical elements to be pinned-pinned right?

Thanks in Advance for your comments and suggestions.

 

[T9809]

Hi,

Just my two cents:

For the purposes of designing the core walls/assessing displacement it is generally ok to "pin" your columns. However, when you design your columns, this would be an unconservative assumption as it would be all be built monolithically and you would be ignoring any actions that eq/wind impose on the columns. Particularly if you have long blade columns.

RCB has a function to reduce the stiffness of column based on the axial load present in the column if you want to reduce your moments

Generally not in favor of designing blade walls to section 11 as it does not account for the moment interaction and underestimates the slenderness effects (compared to if you were to check it as a column).

Centrally reinforced walls with mesh should not be done as it offers no ductility.

For precast columns/walls you are still able to develop moment depending on the dowel arrangments and hence do not agree with the assumption that it is a "gravity only" element.

 

[GRADSTRUCT]

Thank you for replying,

When I design my columns, I do run my model as fixed-fixed and design the columns and walls for these loads.

I generally always provide mesh on both faces for the blade walls. Do you not recommend designing them to Sec 11 if they're fully in compression under ductility 1 loads?

 

[Greenalleycat]

Not sure what Aussie allows for walls, but mesh is a big no-no in walls in NZ, particularly if designing for ductility
Unless you're just meaning 'mesh' as in 'a mesh of deformed steel in both directions'

 

[rscassar]

I think any structure which are designed for earthquakes, which includes all buildings that I can think of, L-class reo cannot be used. That pretty much limits you to the secondary elements like retaining walls or facade panels.

Mesh still has it's place though. It's wonderful for crack control in slabs.

 

[Just Some Nerd]

AS3600 commentary doesn't prohibit low ductility for mu = 1 scenarios, as a matter of rule

I will make the point however that at least one code writer, Menegon, is not a fan of mesh/low ductility reinforcement in general for walls as there is significant risk to do so. There's no room for error in the modelling when it comes to load distribution as the elements are not ductile - if a wall experiences a load greater than its capacity the load won't just redistribute elsewhere like it would in a ductile slab

----------------------------------------------------------------------

Why yes, I do in fact have no idea what I'm talking about

 

[li0ngalahad]

I would add that no matter what ductility you adopt, if your slab/columns system acts as moment resisting frame, you will need to detail the columsn and slabs as such (Clauses 14.4.3 , 14.5.4 and 14.5.5). The commentary indicates that the column/slab system is to be considered lateral laod resisiting if it takes 20% or more of the total lateral load (unfortunately the code and commentary is silent on how to do this check. Personally what I do is an overturning moment comparison between the whole building vs all the shear walls combined and check if it's under 80% of the total).

Generally I leave everythign fixed-fixed and design all for the same ductility level, if I can. Note that for ductility >1 you cannot use the simplified wall design and cannot use L-class reinforcement. I take it that if the main lateral resisting structre is designed assuming a higher level of ductility, then I cannot use simpilified formula and L-class reinforcement in any element of the structure even if I am considering the full elastic forces for those elements. This is because the drift experienced by your columsn might be a lot higher than what your model shows due to additional second order effects applied to a structure where plastic hinge regions have formed, and therefore you may be still underestimating your columns forces even if you are mutliplying eveything by mu/Sp.

Also, if I use ductility 2, column detailing is required only for stocky columsn where D>Lu/5.