Column-Flat slab connection 2

[Matt_Struc]

Dear All,
This is my first thread and I have heaps of questions, so please bear with me
Question 1:
I have recently read this clause in the Australian Standard AS3600 cl 10.3.1: " Le = effective length determined in accordance with Clause 10.5.3; or
alternatively may be taken as— (i) for a braced column restrained by a flat slab floor, Lu".
Does this mean connection of columns to flat slabs should be considered as pin. which means in my numerical models I pin the columns attached to flat slabs, and columns would be design for vertical and accidental moments only? ( That is when columns are not part of the lateral resisting system)

I know this might sound bezzar, but my doubts about this issue will be more obvious in my second question
Question 2:
If no, columns should be modelled as fixed connections to slabs, then can someone justify this practice for me which goes as follows. Every time, I check a transfer slab plans, the loads given by the structural engineer on columns would be only Dead and Live vertical loads as attached here.
Does this mean columns attached transferring should be modelled as pinned at their base ( maybe i am thinking because column starter bars are not developing into the slab)

http://files.engineering.com/getfil...-1ffe-408a-931f-3052fdaa49f5&file=Capture.PNG

Question 3: ( which is related to previous questions)
I also noticed the same thing on foundation plans where the engineer only specified column dead and live vertical loads for the geotechnical engineer to design, while for the lift core, full table of loads was given as attached.

http://files.engineering.com/getfil...39b-4ae7-9d38-6a180a734696&file=Capture1_.PNG

http://files.engineering.com/getfil...3ffc-4a36-9a96-8f6235cdba15&file=Capture3.PNG

 

[hokie66]

You have described a building with flat slab floors and a lateral resisting core. This is common. The designer has not used the columns as part of the lateral system, but only for axial loads.

The column design would have to include moments imposed by the flexure of the floors, or minimum prescribed moments, so not pinned. But these moments from the supported columns would likely be inconsequential for design of a transfer floor, therefore if no moments are given at the base of the columns, you would be justified in considering these to be pins.

Notwithstanding the above, beware of punching shear. All slabs must cope with punching shear, including unbalanced shear, which is often most critical at slab edges.

 

[Matt_Struc]

Dear Hokie66
Thank you for your answer. But how could the designer justify not designing for moments from columns above in transfer slab specially as you mentioned it will affect punching shear outcomes and even if one would want to check bearing stress under columns it would also be perhaps unconservative to ignore moments.
What about pad footings? Shouldn't moments be considering for both punshing and soil bearing capacity checks.

 

[WARose]

If it's done like a typical RC column to slab connection (see ACI 352), I certainly wouldn't model it as a "pin". It would be fixed.....and although it does not appear that it is part of the Lateral Force Resisting System....the connection will still see the resultant forces from compatibility, P-Delta, etc once it is set into motion......and your model should reflect that.

You should also be aware of the situation that hokie66 alluded to with a increase in shear from a pattern live loading. (That obviously would not occur simultaneously with a lateral load.)

As for your question on the foundations.....if it's a fixed base, you can model it and see what kind of forces you get. But in my experience (if you truly represent the flexibility of the base in such a situation), just about all unbalanced forces are dissipated in the columns.

 

[hokie66]

Moments are usually more important in the design of upper storey columns. In the lower storeys, the axial load rules. As to bearing capacity, the rules about strength in supporting slabs take care of bearing issues.

You are technically correct in all cases, but experienced structural engineers know what assumptions are appropriate and when to delve into specific details.

 

[dik]

Usually design columns for the maximum moment from the flat slab with K=1. As the loading becomes greater, the moment is less of a concern; it's usually only the upper floor, or in rare instances the upper two floors where moment and axial load controls design.

Dik

 

[hokie66]

The most important thing to remember in a structure like this is to design each building element for its own worst case. For column design, a pinned assumption is usually conservative, as that gives you the greatest moment in the body of the column.

 

[Matt_Struc]

Dear all,
Thank you for your great replies. I have come to the conclusion that columns must definitely be designed as fixed ended not pinned, though I have seen people pinning columns even at connecting to beams!
I also understand that engineers choose not to include moments on their transfer plans as it overly complicates the design process, specially that it is handled by a PT contractor. However, I think it would be unconservative to do so in some cases.
I am actually dealing with a four storey super structure that is very slender, that only columns exist on the periphery of the building, so all columns are edge columns and moments are relatively high causing many of the columns to fail.
Hokie66, I think in my case all moments applied on the columns are at the joint only due to unbalanced moments from slab, if I pin the columns, the moment in the body will become zero which wouldn't be conservative at all.
Again many thanks to all of you for your precious replies, and I promise more questions soon :O

 

[hokie66]

I assume now that you are referring to the columns which support the transfer structure, which spans the full width of the structure. Transfer beams need to be very stiff, and the stiffer they are, the less moment is distributed to the supporting columns.

 

[Matt_Struc]

Dear all,
since the columns would be considered to be fixed under gravity dead and live loads. How about analysis for lateral loads wind and earthquakes, if my lateral resisting system is limited ductility shear walls and my columns are not supposed to contribute to my lateral loads. The questions is, would I create a second model in which I pin all the columns as to behave as if plastic hinges have formed in them. And just to clarify this would be done only for earthquake loads, right? other lateral loads such as wind, should we expect a combined system of columns and shear walls to act? or do we just pin the columns in that case as well?

 

[WARose]

since the columns would be considered to be fixed under gravity dead and live loads. How about analysis for lateral loads wind and earthquakes, if my lateral resisting system is limited ductility shear walls and my columns are not supposed to contribute to my lateral loads. The questions is, would I create a second model in which I pin all the columns as to behave as if plastic hinges have formed in them.

I would say (assuming we are talking about the equivalent lateral force procedure and a linear analysis) keep the connection (i.e. column to slab) fixed in all cases.

 

[Shu Jiang PEng SE]

The moment of the column will be not significant at all if the following two condition exist:
1. structure has lateral resistance component, like lift core in your building
2. The roof and floor is stiff and can be classified as rigid diaphragm
The lateral stiffness of the lift core is generally way larger than the sum of all the columns at each level, and the rigid diaphragm will transfer the lateral load to lift core. Therefore, you can check the column strength per k=1 (pin-pin connection) under vertical load only, like dead load and live load.

 

[Matt_Struc]

I would say (assuming we are talking about the equivalent lateral force procedure and a linear analysis) keep the connection (i.e. column to slab) fixed in all cases.

If we pin the columns, then wouldn't this reduce the force transferred to the walls? in that case how can we make sure that shear walls as the sole lateral resisting system is effective?

 

[hokie66]

No, you have it backwards. If the columns are not moment resistant, then all the lateral force goes to the shear walls.

I think you are depending entirely too much on a single (whole of building) model for designing the various structural elements. For the lateral resistance of a building dependent on core shear walls only, forget the columns.

 

[WARose]

If we pin the columns, then wouldn't this reduce the force transferred to the walls? in that case how can we make sure that shear walls as the sole lateral resisting system is effective?

I think you'll find that the shear walls are so stiff....that only a very minimal amount of lateral force will wind up in the "moment frame" of a slab-column. And most of that force will (likely) be from compatibility.

 

[WARose]

By the way, to build on what hokie66 was talking about above, for the shear walls, I'd probably figure the load going to them via a rigid diaphragm analysis (which would disregard the column stiffness). Every time I've done one of these I've wound up doing that anyway because it's hard to back out a FEA model with the correct shear loads (per level) in such a situation. At the very least, it's a good backcheck.

 

[Matt_Struc]

Thank you all for responding, it has been a pleasure hearing great opinions from you guys!