Lateral Analysis Etabs - Stiffness Modifiers

[thrice]

Hello guys
Sorry in advance for the long winded question.
Relatively new (~6 months) to lateral analysis of taller structures and using etabs.

I've modelled a ten storey precast building in etabs and I'm using the shear wall designer to design the walls.
On the advice of senior collegaues I've split the shear walls to approximate the actual panel splits.
Set all in plane modifiers to 0.7. Ended up with a bit too much reo.
Said colleagues advised we can bring it down further to 0.4, reducing building stiffness and allowing the building to drift more, as long as within limits.

Reading on engtips some previous forums, I'm of the understanding that what I need to be doing is working out which walls are cracked, and assigning them a cracked modifier.

So I guess my first question is, is it safe to assume all shear walls are cracked like I have done? (Note I'm in Australia and the code stipulates N*/(Asf'c) ratios, see attached.)

Second question is, I have multiple cast-in plates at each level tieing each precast panel to the next. Especially in the cores I have some really heavy ones.
Can I model these cast-in plate connections in etabs? Is there anything else I should be considering since all the panels are stitched together?

Third question is can anyone recommend any resources that deal specifically with the design of tall buildings in precast?

Thank you.
thrice_a

 

[slickdeals]

Do not reduce stiffness to reduce rebar. If all walls are cracked, their relative stiffness and therefore, the loads they resist will be the same as what you have with a 0.7 modifier. Not quite sure on why reducing stiffness is changing wall rebar. Can you provide more detail?

You can model a rigid plate connecting them together separated by 1" or so to model the connection and determine the vertical shear in it due to them being connected.

 

[hardbutmild]

@slick
I believe that the OP meant that when he reduces the stiffness, his whole structure becomes less stiff and he enters the "long period" area of the response spectrum which in turn gives him a lower pseudoacceleration and base shear.

As to your question OP, I think that it depends on the person. I wouldn't do that, but I suppose some people would. In reality you can't really know what'll be cracked to what degree at the time of an earthquake. I'd especially like to add that most people calculate the stiffness based on the concrete alone, not taking rebar percentage into account. Imagine an example structure with only three walls, two are 2 meters long, the third one is 4 meters long. Now, when determining the base shear distribution, most people will say that the short walls will each take 1/10 of base shear and the long wall will take 8/10. But when you do that, you'll get for example 0,5% reinforcement in the short wall and 2,5% in the long wall. Now your long wall is even more stiff in comparison to the short wall, because it has more reinforcement. If you try to iterate the procedure you'll only get stiff elements stiffer which is absurd.

It seems to me like your code wants you to check what level of axial force your walls experience (I presume under the most probable vertical load, probably something like 1,0 DL + 0,3 LL) and use the stiffness according to the table.

As to your second question, I don't use etabs so I don't know.

Third question, I'd say you look here:

https://www.nehrp.gov/library/index.htm

https://www.fema.gov/emergency-managers/risk-management/building-science/earthquakes

https://peer.berkeley.edu/peer-reports

I think that nehrp has some guidelines on precast design, not really sure.

 

[Retrograde]

I would agree with your colleague that 0.4 is a more realistic modifier for a cracked wall than 0.7. Refer to Table 6.2.4 of AS3600:2018. Clause 6.2.4.2 does say that for uncracked sections you should use use 1.0.

 

[milkshakelake]

I don't know that particular code, but I don't consider rebar in the stiffness modifier and I don't use interpolation. I'd end up with successive iterative designs that keep changing the factor. You can use ETABS to check the shell stress and compare it to the modulus of rupture, and that's when you know if it's cracked. I don't remember the exact thing but I think it's F11 or S11 in ETABS. I'd say talk to your boss about establishing a stiffness modifier and use that. You can use cracked condition for lower floors and uncracked for upper floors, or use cracked for everything. 30% of lower floors cracked tends to work as a first guess.

One engineer I know uses 0.5 modifier for everything, regardless if it's cracked or not. He's been doing it for about 30 years, even before ETABS, so there must be some wisdom behind it. I don't take this approach but tend to crack the lower stories and uncrack the upper stories.