Limited Ductile Assumption & Plastic Hinge - AS3600-2018 1

If the walls are in compression with mu = 1 and sp = 1, then they will not crack, and the building will sway elastically under the full load. So any attached memebers will attract that same load due to their connection and will not experience plastic sway effects either.

Thanks rapt,

So in this scenario, would we be forced to design all the walls as non-ductile based on the full elastic earthquake load? What would happen if we designed and detailed the walls as limited ductile with the reduced load used for strength calculations, would that lead to brittle failure under compression?

Would it be recommended to intentionally reduce the overall lateral stiffness of the structure in order to encourage yielding and plastic hinge formation at the base of the core walls, and thereby provide some level of ductility in the system? This would also allow you to justify using the reduced load for strength calculations.

Yes, everything would be designed as non-ductile as the building cannot sway in a ductile manner, so the walls will attract the full earthquake load at a far smaller sway deflection than would be experienced in a semi ductile or ductile building.

Structurally there is no problem with this being non-ductile. You basically end up with an importance level 4 design and hopefully everyone is safe from earthquakes.

Thanks rapt, what are your thoughts on the second question from the first paragraph of my previous post?

What would happen if we designed and detailed these walls as limited ductile with the reduced load used for strength calculations, would that lead to brittle failure under compression?

I did not think I had to answer it after my last answer.

The compression on the wall would be about mu / Sp times what you have allowed for in design. I assume that would lead to some sort of failure if you thought you had designed it to somewhere near the limit for the reduced load!

rapt said:

If the walls are in compression with mu = 1 and sp = 1, then they will not crack,

Click to expand...

I think this is misleading. Flexural cracking is associated with the flexural demand being higher than the flexural cracking moment. Yes the compression/tension affects the cracking moment, enhancing or degrading it. But stating something like designing to mu=1/Sp=1 means no cracking is fundamentally wrong in my view. Care to explain your thinking here, given you're on the code committee, one can assume you know more than me about a load of things. But I'm not seeing your logic here?

Agent666,

I would think you have more experience and education in earthquake design then me! I have been playing catch-up over the last few years. Before that we did not worry about earthquakes in Qld, just cyclones.

He said in the first question that his wall is still in compression under mu = 1 and Sp = 1. So we would assume no tension so no cracking.

Drapes said:

under the full elastic earthquake load (with u=1) all the walls remain in compression

Click to expand...

Guess OP can clarify, but I took that to mean still under compressive axial load (no axial tension), not that the entire wall section under combined design axial load and moment was completely in compression?

With seismic moment added I'd find it unlikely that from a design perspective things were uncracked, higher in a cantilevered wall yes it is probable as demand drops off significantly, but not at the base unless you had way too many walls and very little seismic moment I don't see you ever using the full gross uncracked stiffness?

Guess I'd need to know more about the structure to say one way or the other, but I guess first we need to understand if we are understanding the same nomenclature from the OP in terms of what wall remains in compression means.

I was referring to all walls being in compression and uncracked under all load combinations (axial plus moment due to seismic) using u=1 and sp=1. This was simply a hypothetical scenario, appreciate this is highly unlikely unless there are too many walls and very little seismic moment.

The reason I asked the question was to understand if during design, a check first needed to be carried out using u=1, sp=1 to confirm that the base of the core walls do indeed crack, in order to justify proceeding with a design based on u=2, sp=0.77. It appears as designers we should be doing this check, but rarely will it be an issue.

Think of it this way, the earthquake doesn't care how walls are detailed, it will just push them until either they yield in flexure (ductile response), or they had sufficient elastic strength to resist the loads without yielding.

Load keeps increasing until such time as the concrete cracks, then stiffness changes, dynamic response changes (structure softens), then again load still increases mobilising available strength. If wall yields then starts to display some ductility, how much depends on how big the event is. The design earthquake is just that, an arbitrary level of load based on probabilistic methods to achieve acceptable levels against the risk of collapse occurring. Bigger or smaller events for a given location may occur in reality.

Reality is under modest earthquakes wall starts out uncracked (potentially though not the first event it might have seen to crack it) but at the higher stiffness you naturally have a higher base shear as period is lower due to the increased stiffness. So it will rapidly get to a point under the uncracked scenario where it cracks. Most standards just take you directly to this second scenario looking at an appropriate post-cracking effective stiffness because in all likelihood it will occur very rapidly at the onset of shaking, within a cycle or two with a rapid shift towards a softened dynamic response.

You can certainly look at what you are noting, however I think you'll find the cracking moment capacity is rapidly exceeded. If it isn't then its probably a very inefficient structure possibly? In NZ at least we are supposed to ensure that the stiffness we use is appropriate and takes account of regions where there may be no cracking (most notably higher up in the structure where seismic loads in walls might be quite a bit lower). To determine how far we apply stiffness reductions up the structure it is often a bit of an iterative approach. Keep applying the stiffness reductions up the structure for a walled structure until we reach some state where we are only left with regions that have no stiffness reductions applied due to the moment demand being lower than the cracking moment capacity.

Thanks Agent666, a very insightful and helpful response. Much appreciated!

Agent666,

AS3600 now requires designers to determine stiffness dependent on the degree of cracking similar to NZS. Not sure if all Aussie designers have figured that out yet.