Minimum Design Moment: Column suppoting beam

[sybie99]

Good evening,

I have previously posted on minimum moment a column supporting a slab should be designed for. In the case of a column supporting a long span beam (16m) this moment becomes incredibly large if taken from a finite element analysis.

Rapt, one of the members, previously posted that as long as one designs the column to take its cracking moment (considering axial stresses and bending stresses) that is sufficient.

In my case I have a continuous spanning beam of 1400mm deep and 1350mm wide, with two 15m spans.

This beam is supported by three 1350 x 345mm columns, about 10m high. The columns are orientated so that the weak axis will pick up bending from the beam.

If I calculate the cracking moment it is 320kNm, which the column can easily take.

The FEM analysis gives bending moment of 2200kNm, in which case it would be impossible to sufficiently reinforce the column.

I do not need the column to resist any moment in terms of assisting in frame action for stability, or to reduce midspan bending in the beam. The beam supports a slab to which it id tied into, and this slab is connected to shear cores, so in reality I need no connection between the column and beam, only plain bearing. The beam may as well be precast.

If I were to extend the column bars by say 100mm into the beam, and then also wrap these so there is little to no bond, no moment transfer can happen due to stresses transferred via bars. I actually need no shear connection between beam and column (not seismic region) so actually I don't have to extend any column bars into the beam, but would at least have some bars extending so that the column is laterally restrained at the top.

In such a case it would seem unrealistic to design for the extremely large moments I get from an FEM analysis.

So, my question is, what would be the minimum moment to design the column for?

Thanks,

 

[sybie99]

I was just thinking, seeing that I do not need any continuity between the beam and column, only a shear connection so that column has restraint at the top, could I not finish the column top nice and smooth, have say 6 25mm bars protruding 100mm into the beam, wrapped in a membrane to minimise bond, and the also apply a slip membrane (malthoid) on top of column.

This way I cannot see how any significant moment transfer would be possible.

 

[CELinOttawa]

I agree with your approach. It is all about the load path, and so long as you've debonded the section, you should be fine.

Remember that bond breaking tape alone is enought for plain bar (which I doubt you're using!) and you must use a smooth sleeve to prevent bond transfer mechanically in the case of deformed bars.

PS: I have been responsible for the long frames and cross frames of a Rugby stadium, so I am well familiar with reinforced concrete design... HOWEVER the vast majority of my work has been in precast. I have heard of, but not implemented, a cracking moment solution. I see no reason for RAPT's approach to be in error, other than that this approach fundamentally violates SLS without very careful consideration. Further to this, if it is *possible* for load to be taken into the column, I would oppose the detail as being a poor choice when you could prevent this with another detail.

I'll be very interested to hear what some of our US friends think.... The north is still a pretty site-cast place.

 

[rapt]

Firstly, I never said that it only needs to be designed for its cracking moment. I said that the cracking could be taken into account in determining the stiffness of the column. So the I of the column could possibly be reduced. If you do this, you should also consider the cracking in the beam and see if it also reduces the stiffness of the beam. Normally in RC structures, the beam will be cracked and the column may be cracked depending on the axial load. If reduced stiffness is to be considered for the columns, all members should checked to see if they should be reduced. You cannot do one without the other.

If you create a pin connection at the joint, that is another matter. There is nothing wrong with this as long as you provide for all of the design actions that are happening. Even with a pin designed, this is not the same as a precast connection. The concrete is monolithic and rotations will occur between the beam and the column. How that affects the connection at the bottom, I do not know. It is not your basic shear design for the beam any longer as there is no top tension cord. So you cannot use the ACI beam shear provisions for the beam design. You cannot use shear friction as there is no top tension cord. No design code that I know of covers this.

If you are going to play games like this with your design, I hope you are going to build some test models first, and test them fully, especially under reversal loadings. You might not have earthquakes, but you do have sway and wind cases that cause similar problems at connections.

 

[sybie99]

Rapt, you say the concrete is monolithic, but I will de bond the beam and column using a malthoid layer. So in effect no tensile stress can transfer between beam and column, except some minimal transfer between the debonded rebar I protrude into beam to resist the shear I require in order to laterally restrain the column.

The dimensions of my column, 1350x345, only just classify it as a column as per BS8110. If it were 1400 x 345 it would be classified as a wall and the following would apply, taken from BS8110 for wall design:

3.9.3.3 Design transverse moments
Design transverse moments, when derived from beams or other construction designed to frame
monolithically at right angles into the walls, should be calculated using elastic analysis. When construction is designed to be simply supported by the wall, the eccentricity may be assessed as for plain walls (see 3.9.4)and the resultant moment calculated. Except for short braced walls loaded almost symmetrically, the eccentricity in the direction at right angles to a wall should be taken as not less than h/20, or 20 mm if less, where h is the thickness of the wall.
-----------------------------------

So, in my case construction is designed to be simply supported by wall, and I can calculate moment using eccentricity as for plain walls?

 

[CELinOttawa]

That code clause lists a minimum.... You must still design to the actual conditions of your geometry.

Or am I missing how you think this helps?

 

[hokie66]

The calculation based on uncracked stiffnesses just establishes that the column will crack. I wouldn't even attempt anything like a pinned connection with bond broken and that sort of thing. I see continuity in the connection as advantageous. However, I would be concerned that this column is quite slender.

 

[sybie99]

Hokie,

Can I ask what your concern would be in creating a pinned connection the way I propose? The beam is tied to a very large slab (550mm thick waffle) which is connected to lateral bracing system in the form of shear walls.

I do not understand what the concern would be, there is no load reversal, no seismic, no sway.

The column is 10m high from base to beam soffit. So slenderness ratio is 28 with effective length of 1. If I model it as fixed at base (base can take moment) and pinned at the top, un braced, this would be worst case and gives le = 2 x 10 = 20, 20/0.345 = 58 < 60 prescribed by BS8110.

 

[JAE]

When the slab deflects doesn't the load reaction on top of a "pinned" column get pushed over close to the edge of the column - thus creating moment on the column? In other words - you would have an eccentrically loaded column.

 

[johnbridge231]

Exactly. Exentically loaded column can lead to biaxial bending...

Analysis and Design of arbitrary cross sections
Reinforcement design to all major codes
Moment Curvature analysis

http://www.engissol.com/cross-section-analysis-design.html

 

[sybie99]

JAE, I agree that can happen, which would result in moment = to N x h/2, where h = column depth.

This is still a much smaller moment than would be the case if the column and beam are fully tied in.

 

[rapt]

What is the transverse spacing of your columns and slab loading? I assume there are no columns above, only below.

2200KNm sounds excessive for this column/beam arrangement unless the loads are very high.

Why isn't the beam prestressed. 1400 deep spanning 15m sounds uneonomical!

 

[sybie99]

Hi rapt, I think that when using FEM there was an error with the slab meshing into the column giving very high moments. If I do a simple frame analysis I get lower moments, about 500kNm.

The layout is a continuous beam of 1400 x 1350 wide, 3 spans, 16m,5m,17m. 10m tranverse spacings. Dead load is 80kN/m from slab SW and an additional dead load of 40kN/m. A live or imposed load of 40kN/m.

We looked at PT but in this case there wasn't much cost difference and the client did not want to get them involved but wanted the main contractor to do all concrete works.

 

[sybie99]

columns are 345x1350, 10m high, rotated such that weak axis in bending

 

[rapt]

That was my thinking. 2200KNm sounded awfully large for what you had described. For normal loadings I was getting about 600KNm. So your column can take that and you no longer have a problem with the column design!

If RC is as cheap for these beams, there is something wrong with the PT industry in your area! PT should win this one by a mile if there are a reasonable number of beams!