Colums release and columns to beam joint in etabs 3

[seny]

Hi everybody,
I’m modeling 30 storey concrete building w/ shear walls lateral system and columns responsible only for gravity.

As per ACI 318 section 8.8 we top of columns to be modeled as fixed and release can be assigned to bottom of the column.
Is it appropriate way to model column in Etabs for service and ultimate stage – hinge at the bottom & fixed at the far end? So all the columns will have no release below the floor and hinge right above at all the point of column & slab intersection.

One more question, I need to transfer at the second floor elevation on of column to two adjacent columns over the bridge (transfer) beam. What is the best way (technique) to model transfer beam to two columns below connections in Etabs? In case two columns below, three columns above and beam have no releases – all the columns (below and above) fail b/c of transferred moment from the beam and required enormous sections, what is not the case in real life. Assign releases at the beam ends? In this case haw Etabs takes in consideration that in real life there are no released beam & column joints in monolithically poured structures?
I am sorry for long post.
Any input on these matters are highly appreciated.
Best regards.

 

[Skicat]

Seny,
ACI 318 8.8 states that it is permitted to assume both ends of a column are fixed when computing gravity load moments, assuming the column is built integrally with the structure. I'm not sure why you would fix the top and release the bottom as you stated above.

If you are going to transfer out one column at the second floor into two columns below at the first floor, I would expect that the transfer beam would be enormous. If I understand what you are saying, the column you are transferring is carrying 28 floors of load. I would not release the ends of the beam because you need to transfer out the moment as well as the shear. Are you using post tensioned concrete floors? If so, the transfer beam could also be post tensioned, which would help. What program are you using to analyze and design your floors and beams? If it were me, I would take the forces generated from the etabs model and model a transfer beam in a concrete beam program and not rely on etabs to analyze / design my transfer beam. I hope this helps and I would definately consult other sources given the magnitude of the task you are undertaking.

 

[seny]

Skicat,
Thanks for respond.
I agree that bridge beam has to be fixed to the columns to guaranty condition of real moment connection reinforcement and that is what I have in my model.
Building is RC no PT.
I am using Etabs and there is no problem w/ transfer beam.
I have room for say 5’Wx6’H beam even more if req’d.
Technique w/ bottom hinge was advised by Etabs support and this way commonly in use when you do not want column participation in lateral stability of the building.
I have previously constructed buildings w/ the same transferring conditions.
The question w/ modeling in Etabs – all three columns above the transfer beam (middle "B" point transferred and two ends continues"A"&"C") modeled w/ fixed bottom point fail b/c of bending moment transferred to columns from beam (deflection & rotation).
A B C
|__|__|
| |

 

[Murali27]

Hi Seny

It may help by idealizing partial fixity to column B base (full moment capacity may not be achieved). I used to pin the base of floating column and design the beam to absorb all the moment/shear. Nominal moments will be transferred to columns above or below. (may not be ideal if this is in seismic region).

Please also consider the construction sequence analysis (realistic way) for this transfer. It would help.

Hope the response is useful

 

[seny]

Hi Murali27,

As per real life i do not have concern about all three columns above since i can consider this situation as column above elastic footing and bottom end can b pinned.
But as per Etabs modeling w/ all three columns above without release - all of them are failing and requiring big reinforcement to govern bending from column & beam joint rotation.

 

[ishvaaag]

I have used may pinned assumptions in many buildings as murali over a transfer beam, and also at end of transfer beams, with entire success, and also seen amazing vierendeel behaviours (inversion of moments when a column was lacking in a tall building) and well, that's the nice property of continuity delivered by RC buildings, one can see even bombed buildings partially standing through continuities.

Then the question brought by Murali about construction sequence is also entirely meaningful; where I live (Spain) except some intelligent assumption sometimes standing in some programs, and except the only recently mandatory checks to have proper formwork resistance, this is an issue that almost never appears. Given this, it seems that the current method used, classical one, seems enough to bring the buildings to proper safety IF you get to have them built. That is, a proper analysis of the sequence of loading may show things many would love not to see, but that to some extent are (and for some kinds of buildings and kinds of sequences of construction and detailings or are contemplated or will bring ruin).

Other than that, some programs seem to be quite esoteric in what they take care of or not (and the same happens with some codes, that seem to be partisan for example to a slab with included beams respect those that pass punching without such: a flat plate unacceptable on punching without beams becomes magically acceptable on including them).

Vierendeel action over transfer beams uses to be detected by whole 3D models having the continuity embedded, and you can even play with the size of the transfer beam to get different proportions of members in elevation. Yet one must NEVER forget that these things, useful as they are are just MODELS of the structure, and the models only have the intelligence we put within.

So, yes, I have also had cases of columns and beams failing in entire continuity and where the sometimes only amenable solution to the intent has been pinning something. It has proven to be a satisfactory solution as long the requirements of equilibrium and ways of passing the loads are met and any code issues of relevance have been contemplated.

 

[seny]

thanks fro respond, ishvaaag.
Building is flat slab, gravity column and shear walls taking lateral responsibility entirely - no beams, so there is no Vierendeel trusses condition. In addition, slab is modeled w/ no out of plane stiff.
What i topically do in this situation - ignore in model transferring of column and after based on forces obtained from say Etabs calculate simple one bay beam to get bottom reinforcement and based on result provide in percentage top reinforcement above the supporting ends columns and column calculation accordingly. It is enough conservative way and i have already built structure w/ 25 storey transferred column.
Beam is always overlapping supporting columns, so all the three columns above really can be pinned at the bottom and can be considered as above footing and do not require special reinforcement detailing efforts. I never used pinned transfer beam since i am not comfortable w/ condition when beam is not overlaps in model the columns. For me the main question was and is - in what way I can create in model the most "real" transferring conditions.
In addition i will really appreciate if you can input a couple more words about your detailing in conditions when you had pinned bottom of transferred (and ends?) columns or (and) pinned transfer beam.
One more time thanks for respond.