AS3600_2018 - clause 6.2.4.2 for seismic modelling 3

[GTD_18]

Hi All,

I am just curious on what other teams/offices are doing with clause 6.2.4.2 (in AS3600_2018) wrt to the effective section properties of walls for seismic events?
• Uncracked = tensile stress < mean char flexural tensile strength
• Cracked = tensile stress > mean char flexural tensile strength.
So, from the above it is assumed that the initial seismic modelling should be completed on Igross, then assess the tensions vs rupture capacity and crack the section based on applied axial load.

Personally, up to this point, I had been initially adopting the “uncracked” wall value of 0.7Ig from ACI-316 and cracking from from here if required.

Just curious to see the design approach from others?

 

[Agent666]

By ductility limit I'm assuming you mean local curvature?

I'm not sure how AS code implements curvature limits or if it even requires checking of the curvatures in potential plastic hinge regions, but in NZ code limits are prescribed. Just so I'm clear curvature is the plastic hinge rotation divided by the effective plastic hinge length. I'm assuming AS code has similar philosophy for applying different detailing requirements and it is not still tied to the global level of ductility? (though I'm not sure if that is actually the case?).

The level of detailing required (confinement and antibuckling detailing predominantly) depends on the local level of plastic curvature.

The curvqture limits in our codes leave sufficient margin to get to MCE (1 in 2500 event) with sufficiently low probability of collapse. Basically, follow all the code requirements and it should mean it performs adequately under higher than code events, without the specific need to assess that limit state.

Note the global ductility is not the same as the local ductility demands areas of structure may see. If AS code does not use the concept of curvature to drive the level of detailing required then it is fundamentally wrong in its approach. We used to have this up until 2006 but it was recognised as being incorrect. If you have access to the NZ concrete code commentary it has a good explanation of why the global ductility you might use to determine the overall base shear is a poor measure of the detailing required. Often you might have a structure designed for mu=1.25, but require a ductile level of detailing to comply with the curvature limitations.

https://engineervsheep.com

 

[rapt]

Agent666,

For beams with mu > 1.25 and <= 3 (> 3 requires checks by other codes eg NZS), we introduced minimum compression reinforcement requirements and have limited neutral axis depth in the latest code. An attempt at something similar to the NZS reinforcement ratio limits but we hate reinforcement ratios when combining reinforcing and prestressing and sections with non-rectangular shapes!

 

[Settingsun]

The curvqture limits in our codes leave sufficient margin to get to MCE (1 in 2500 event) with sufficiently low probability of collapse. Basically, follow all the code requirements and it should mean it performs adequately under higher than code events, without the specific need to assess that limit state.

Thanks, that answers my question. So basically the ductility isn't maxed out at the nominal ULS design earthquake. It does however seem to imply that design life and importance level become irrelevant for the ULS if everything is going to remain standing? Or does the meaning of 'should' perform adequately under MCE change - eg 70/30 for IL2, 80/20 for IL3, 90/10 for IL4?

 

[Agent666]

Might be wrong here, but irrespective of the importance level, the maximum credible earthquake (MCE) is still a 1:2500 year event.

So basically, for an IL4 building with 50-year design life in NZ, you're designing for the MCE level event as the actual design basis earthquake. So naturally you'd expect better performance, and potentially less damage at a design level event, and a correspondingly lower risk to life safety (even lower risk than normal IL2 building).

The wording in our code is "sufficiently low probability of collapse", I don't believe they actually say hey it's a 0.0004% chance of collapse, so it's a bit subjective to anyone but the code writers who might understand or be able to quantify that risk.

Stuff that's designed correctly hangs on for dear life in terms of not collapsing, stuff that's designed or detailed poorly gets found out and in the worst cases will potentially collapse.

The buildings that collapsed in the NZ 2011 Christchurch earthquakes resulting in significant loss of life had some severe structural weaknesses. The peak accelerations during these events were upwards of two times the design basis earthquake which is obviously close or even exceeding the MCE level event on paper.

Other buildings with lower seismic ratings (less than 34% of current code requirement) often survived shaking in excess of 2 to 4 times what they would have originally been designed for, albeit with significant damage that eventually required demolition. But they did their job in terms of the life safety objectives. In my opinion, structures typically have a lot of reserve capacity if designed appropriately.

We have to remember the code is always a bare minimum requirement, and often it may be appropriate to exercise some judgement and go further than just the minimum to improve performance further.

https://engineervsheep.com

 

[Settingsun]

Thanks, Agent.