UBC & ACI - Reg

[Murali27]

Dear All

Under ultimate state, though wall is not cracked, 0.7Ig need to be used in the analysis. How do we modify the stifness in ETABS. Are they f11 & f22.
Please let me know your opinion

We are currently using UBC code as a reference for analysis purpose. Stiffness modifiers as stated in UBC 1997 are for ultimate. It doesn't clearly indicate what to be used for wind tunnel study and vibration studies. But ACI has got the right direction. But we need to satisfy the UBC requirements.

Do i get any reference or cross reference from UBC to ACI with respect to stiffness modifiers?

Thanks
Murali G

 

[WillisV]

Murali - setting f11, f22, and f12 to 0.7 will accomplish what you need to do.

 

[Murali27]

WillisV

Thanks for your reply. I am not convinced with this option of revising f11,f22&f12, because it changes axial & shear stiffness of the wall.

By setting 0.7 to f11,f22&f12, has increased the time period 15% more, which eventually increased wind loads/drifts also.

As per ACI 318, wind studies can be done with gross section properties for vertical elements.

Please comment

Thanks
Murali G

 

[WillisV]

Cracking your walls will result in an increase in period of the building no matter how you go about it - that is no suprise...

 

[Murali27]

Thanks WillisV

Yes, you are right. But for the wind tunnel studies, estimation of time period,etc as described in ACI 318, we may need to use gross section properties i.e 1.0 Ig.

Regarding setting the values of f11,f12 as 0.7 in ETABS doesn't seem to be correct as it reduces axial stiffness also.

I have a structure of 370 m height, structural properties were sent to wind tunnel based on gross section properties for walls and columns, fundamental time period is 7 sec.

If we set 0.7 to f11, f22 in ETABS, it goes to 8 sec, which will call for dampers for this building.

Please tell me your opinion and any further references on the above

 

[blihp]

i have been reading this thread and its earlier discussion (thread726-151169) and thought i may chip in..

i think a key criteria is that you need to use different properties and stiffness's for different analyis and thus i often have atleast 2 models at any one time, an ultimate model and a serviceability model

For the ultimate model used for element design, and seismic design, i would personally do as previously described by WillisV and set the walls to uncracked, then see which walls exhibit tension stresses greater than the code and then change these walls to a lower stiffness. I would imagine that these will be isolated walls on the corners of the cores at the bottom, and possibly at the very top (where there is insufficient compression preload from self weight).

For the serviceability model, used for storey drift under wind, natural frequency for wind, i would assume gross section capacities for the wall (1.0 Ig) but would still crack beams and slabs as these will nearly always be cracked in reality as soon as theses elements are depropped during construction. The key here is to keep the relative stiffness between elements about the same.

I fully concur with WillisV that reducing the stiffness of (nearly) any element down from its full gross stiffness WILL ALWAYS increase the period. There is no surprises here.

You make reference to need dampers if your period is 8seconds, i suspect you need these to keep occupant accelerations to within acceptable levels? A major assumption in the calculation of acceleration is the level of structural damping. The accelerations to be sensitive to damping levels, it may be worthwile pursuing this avenue if you wish to keep away from dampers.

Without knowing anything about your building, 8 second period at 370m sounds pretty stiff!

 

[Murali27]

Thanks bilhp for your thoughts.

I fully agree with your procedure that is what i have been following.

But the question is

How to assign the reduction of bending stiffness for shear walls in ETABS packate?. Reduction of f22 for walls doesn't seem to be correct. I wrote to CSI and i am not convinced with their answer.

Is there any clear provision in UBC for computation of wind load drift as given in ACI?

Acceleration is just above the limit. Here the peer reviewers are asking us to follow for serviceability also the same parameters as set forth in UBC 1910.11 (Ultimate state). This is the problem now. If we assign f22 & f11 as 0.7 irrespective of wall is cracked or not, we end up in higher time period, acceleration and structural frame loads.

I agree that damping matters but this too matters for the dynamic component of wind.

Please suggest your opinion on this

thanks
Murali g

 

[blihp]

obviously CSI are the best placed to comment on how industry use their package, and i do not wish to sound contradictory to their advice.

i have only previously reduced the bending stiffness of shear walls, ie M1, M2 and torsion, and not played with the axial stiffness, but this has been for mainly core stabilised or dual frame/core buildings

however, if your structural system uses outriggers and you are developing net tension in perimeter columns/shear walls then i would probably look to reducing their axial stiffness.

The decision on what elements to modify, and which properties require modification is for the designer to decide based upon their knowledge and judgement. Without knowing alot more about your particular building it is hard to give specific advice. (i am not requesting for more info, as the problem is starting to sound like a judgement call)

i have not come across the use of ultimate values of stiffness being used for the calculation of serviceability accelerations under wind before. Given that wind loads/gusts are only applied for such short durations, ie <3seconds, the ability for cracks to open up is limited. I should note that i have used ultimate values of stiffness to check storey drifts under seismic.

i hope this helps

"a wise man know his limits" anon.
blihp