What do you do when first mode being in torsion cannot be avoided? 1

[Ed818]

I am looking for opinions from fellows here in regards to seismic design especially torsional effect.
Say I have a floor plan as below. Two stories (all flat slab) only and in low seismic zone (but I still want to enhance the its seismic performance as much as I can)

1. What will you do to enhance the structure in general if torsion cannot be avoided?
2. If the above layout cannot be changed. What can I do to achieve a better performance to earthquake? Which elements that I need to look carefully into can enhance its capacity? (My opinion is the columns in upper left corner considering the floor will rotate about the core, causing more displacement (especially for non-linear) in upper left corner and thus require these columns to have better capacity to the drift. Am I wrong?)
3. If I don't perform non-linear analysis. How do I check if torsion is acceptable with a given structure? (My understanding is, with torsion, the structure will drift more especially when it becomes non-linear. Do I increase the the drift that I got from linear analysis and times a certain factor and then check all the columns' F-displacement capacity to this factored drift with given reinforcement?)
4. Is there any books that you would recommend especially those ones that cover detailed explanations or examples for torsional effect?

Appreciated if you don't mind share your opinions.

 

[KootK]

It might be desirable to ameliorate the torsion issues by attempting to create moment frames on all of the column lines.

 

[HTURKAK]

Apparently, the system has extreme torsional irregularity. In general,it is not desirable that the torsional mode is the first or second mode . Some codes also do not allow the use of flat slab system for SFRS.

I will suggest you,
-check your local code for applicability of flat slab system,
-try to minimize eccentricity with adding some shear walls around column at upper left corner,
- If possible,at least use drop beams at perimeter,

There are some limitations and penalty clauses for extreme torsional irregularity. Check your code for the modelling and analysis limitations, penalty clauses ( the use of redundancy etc).

 

[Ed818]

Thank you for your help.

The code doesn't prohibit the use of flat slab.
In fact it doesn't require the first mode not being in torsion as well. The only thing regarding irregularity is to consider 10% accident torsion.
I have also seen structure like this everywhere here but personally I want to achieve better seismic performance regardless.

I had tried but architect insisted MO since there will be curtain wall all around.

I was thinking the same thing too but only band beams can be used. Not sure how much it will help. And even if I put perimeter frames the first mode will still be torsional.
The column layout is more irregular than the above one so also hard to arrange perimeter frame.

Now my question are,
1. Put code aside, if the torsional mode is the first or second mode, how do you decide whether it is acceptable or not? ie. how much torsion is acceptable or how much is not in terms of analysis?
2. If the current layout cannot be changed, what elements need to be enhanced the most and why?

Really appreciated if you can answer these two questions.
Thank you.

 

[MIStructE_IRE]

Even for means of escape does the architect not require another stair or lift core? Architecturally this just doesn’t look right.

Still, If i had to do it, I’d look at moment frames.

 

[HTURKAK]

1- Perform 3D analysis considering inherent torsion and accidental torsion ( 10% code specified) additive. Assume shallow beams ( embedded ) having 3t width. Make sure that the columns are stiff enough so the drifts are in the allowable limits.

2- Provide closed stirrups for the embedded beams to enhance frame behavior.

 

[Ed818]

Thanks for your reply.
I did perform 3D analysis with accidental torsion considered and every thing looks 'fine'. The drift is not much so the column should be able to resist this drift. However, this is linear analysis only, with non-linear the drift will increase a lot more. And I believe this is the reason why torsion should be avoided since it will increase drift much more when structure becomes non-linear. (please correct me if I am wrong on this). So when you said make sure the columns are in the allowable drift, what drift do you compare to? Is it linear? (I don't think the linear drift will increase too much if a structure goes from 'regular' to torsional irregular). If it is non-linear, how do you get the non-linear drift without performing non-linear analysis? Is there any multiplier method?

What about the columns? I would like to have closed stirrups for columns as well. Which columns are the one that need to enhance the most at first glance? (are they the ones in the top left corner since they will drift more?)

 

[Ed818]

It is a carpark. The shear wall is where the ramp is. Maybe thats why?

 

[HTURKAK]

-This structure can be analized with Equivalent Lateral Force Procedure,Modal Response Spectrum Analysis, or Response History
Analysis. But with the 3D model.
- I am not sure how you modelled the flat slab.. I proposed a 3D frame with embedded beams having width 3xt. If this is the case with stiffness modifiers applied, you will get the elasto plastic drifts with applying multiplication factor ( Deflection Amplification Factor, Cd. ).
- The analysis should be repeated with applying torsional amplification factor. ( Ax= δmax/1.2δavg)^2))

My favorite number for Allowable Story Drift is Δa = 0.01 h and for individual column ma. 0.015 h .

Does this respond answer to your questions ?. If not, i will suggest you to consult with a local senior engineer.

 

[FirstPrinciple]

It would be interesting to know how the torsion is resisted. Is it primarily through a force couple between the long shear wall and the core or through torsion of the tubular core wall? The shear wall may be creating a bigger issue than it is solving as it pulls the center of rigidity away from center of mass. I would even check modeling in the slab frames and eliminating the wall to see if the wall is helping or hurting.

You noted it was a car park so some other items to be aware of that have not been mentioned. If the torsional resistance is from a force couple between the shear wall and core walls, and the shear wall is on the ramp, pay attention to how the diaphragm links these two element together as it may not be as simple as a continuous flat plate. Additionally, there is often a building joint where the ramp transitions from on grade to elevated deck to prevent a ramp strut load path and minimize shrinkage cracking. In low seismic zones these items may not be critical but they can be in high seismic zones.

 

[MSUK90]

My favorite number for Allowable Story Drift is Δa = 0.01 h and for individual column ma. 0.015 h

Each individual column shall be checked for a limiting drift of 0.015h??
Is this a codal provision or your own check?
Also, I would like to know that is it possible to have a flat slab system in low to moderate seismic zone without shear walls??
If yes, what type of system it will be to calculate 'R'??

 

[HTURKAK]

Dear MSUK90 (Structural),

Please find my responds to your questions asked to me ;

Q-Each individual column shall be checked for a limiting drift of 0.015h??
A= You may check but by visual inspection ,the most critical columns upper left columns ..

Q=Is this a codal provision or your own check?
A= The valid code in my zone does not allow the use of flat slab systems alone..The flat slabs can only be used if the SFRS is structural wall system.. So , that is my personnel opinion ..


Q=Also, I would like to know that is it possible to have a flat slab system in low to moderate seismic zone without shear walls??
If yes, what type of system it will be to calculate 'R'??
A=IMO, Flat slab systems without shear walls should be used for low seismic zones for two -three storey bldgs. R should not be more than 4.0



GOOD LUCK..

 

[MSUK90]

R should not be more than 4.0

I think none of the building codes define 'R' value for flat slab system?
Do they?

 

[HTURKAK]

I do not remember at least and do not know any code ..THat is because of you can not assign a hysteresis loop for flat slab system ..The joint and column strip without confining stirrups will deteriorate immediately..

But for flat slab system,you may still design the column strips assuming shallow depth beam and provide link reinforcement and assign a limited R value..

Let me explain in more detail;
- I will not suggest, encourage the use of flat slab system for moderate and high seismic zones except the SFRS is shear walls,
- If the use of flat slab is inevitable at low seismic zones, and the bldg is low rise , you may use flat slab system but you are required to design the column strips as the beam elements of the frame system..
- The use of drop beams at perimeter will enhance the seismic resistance..
- This approach could be assumed circumventing the codes but if you strictly follow the detailing rules, the SFRS system can be classified as MOment Frame..(IMO)..


I hope this respond answers to your question..

 

[MSUK90]

I hope this respond answers to your question..

Very well. Thanks a lot. Cheers.