Flat Slab design-column moments 2

[narendranath n s]

Hi i am narendranath,columns can be designed only for the axial loads? in the flat slab design by neglecting the slab negative moments and the slab for the simply support condition.

 

[BAretired]

Don't know what you mean by "the simply support condition".

Columns should be designed for axial load combined with moment. For interior columns, the gravity load moment may be small but the columns should be designed for a minimum eccentricity as defined in your code. For exterior columns, the moment may be considerable and cannot be ignored.

For lateral loads, all columns may have significant moments.

BA

 

[dik]

If multi-storey, after the first couple of floors, the effect of moment on columns is generally very minor. The first supported slab can apply significant flexural moments into columns depending on the column location, loading and bay spacing.

Dik

 

[KootK]

1) In general, I would say that no, column moments cannot be ignored in column design. Unless special measures are taken to prevent it, there is a risk that your columns may be compression controlled (P > P-balanced) and would thus fail via brittle concrete crushing before the moments were able to be redistributed to the slab.

2) If the axial load on a column is kept below P-balanced on the interaction diagram (tension controlled), then it may indeed be safe to design the column for axial load only as the column should be able to yield flexurally and redistribute its moment to the floor structure. I've done this a couple of times with beam and slab systems but never with flat slabs. More on that later. In my opinion, some extra conservatism is appropriate with this strategy. I've used 1.25 Fy instead of 0.9 Fy in my check to ensure that my axial load is safely below the balanced point. I feel that's consistent with the probable strengths of concrete members used in seismic design. One of the problems with all this is that it leads to pretty big columns in a lot of cases. Other than saving computational effort, it's really only beneficial in columns with high Moment/Axial load ratios. Maybe roof level columns.

3) There are two major non-column problems with the strategy of ignoring the column moments in a flat slab. Firstly, you may see a lot of undesirable cracking in the slab over the columns if you're designing for moments a lot smaller than the calculated elastic moments. Secondly, if you underestimate the demand for punching shear moment capacity at the slab/column joint, you could end up with a brittle, catastrophic punching shear failure causing your slabs to pancake. See the sketch below intended to frighten you. Although that one's surely got more to do with axial load overstress rather than moments.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[Ingenuity]

See the sketch below intended to frighten you. Although that one's surely got more to do with axial load overstress rather than moments.

...coupled with the fact that is was of lift slab construction, with long-term history of deterioration and attempted but failed repairs over several column/slab areas. Pipers Row Car Park in the UK. Collapsed in March 1997.

 

[KootK]

Going for dramatic effect, not historical accuracy.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[Ingenuity]

Going for dramatic effect

I think it worked. These types of photos always give me the cold-sweat thoughts of "damn, did I check x, y and z".

 

[rapt]

Always design the columns and especially punching shear for the moments the columns can attract.

Yes, if you can show that the column will be cracked and the floor slab uncracked then you can reduce the column stiffness . But how often does this happen?

In most cases, RC slabs and beams will be cracked under vertical load. For low rise buildings, end columns will often be cracked as well. So you can reduce the column stiffness but you also need to reduce the slab/beam stiffness. But the column carrying both axial and moment will be cracked less than the slab/beam, so the columns will actually attract more moment, not less than would happen with the assumption of Igross for everything.

For Highrise under vertical load, normally the column will be uncracked and the slabs/beams cracked, so the columns will actually attract more moment, not less.

The only time you could possibly expect the columns to be more cracked (less stiff relative to slabs/beams) would be for lowrise PT structures at the end columns.

So basically, use gross stiffness as design codes suggest!

And always check punching shear assuming full column connection!

PS This message is especially relevant in SE Asia and West Asia where engineers have been ignoring columns in their slab/beam design for years!

 

[narendranath n s]

@Kootk,Thanks for your reply.for one of our project the consultant proposing the reduced the column stiffness at roof level because of the columns are not continuing above roof level and not possible to get the 100% fixity condition for the column design.the column stiffness considered at roof level is only 20%.thereby the flat slab is designed with 80% bottom moments.
Is this condition of column is acceptable? and any special detail required at column and slab junction to achieve the reduced stiffness.
If the fixity condition considered 100% we are getting more moments and less axial loads for these forces columns are working.Please advice.

 

[rapt]

Depends on what you call 20% fixity? Do you mean 20% column stiffness or 20% of full fixity?

People throw this "fixity" term around all of the time. In a frame analysis program, you are not assuming full fixity. Full fixity means wl^2/12 for the moment at the end. Normally you are analysing for full connection, using the Ig of each member. An FE analysis will actually effectively give less connectivity than this due to torsional rotation of the slab/beam away from the column. RAPT will do this also by using the Equivalent Column logic to calculate the column I for the frame calculations. Presumably some other 2D software will do this also.

Just analyse it for the actual gross stiffness of the column and of the beam/slab with full connection. That will tell you the moment the column is trying to attract.

Even with only minimum reinforcement, fully cracked column stiffness will be more than 20% of gross. And because a column has axial load, the real cracked stiffness will be significantly more than this.

 

[KootK]

Your most welcome Narendranath. It's true that it's difficult to get a full moment connection between the slab and column at the roof. And that's a fine reason to perhaps add some bottom steel to the slab. That doesn't necessarily make it a safe decision to assume reduced moment transfer for designing the column and punching shear however. What if the beam to slab connection ends up performing better than you expect?

That said, there's been a long standing tradition in my market to reduce the stiffness of perimeter columns by 50% to account for their relatively high moment to axial load ratio and the likely section cracking that would imply. It's not very scientific though so I prefer to design using Igross and full moment transfer at the slab joint via the equivalent column procedure as rapt has suggested.


I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.