Flat slabs limited "edge column" moment transfer

[SnapSpace]

Hi!
I have a question on this subject.
When calulating this problem by hand, you divide flat slab into 2 strips, one is the column strip and the second is the middle strip.
When calculating negative moments at the "edge column" we had a limitating moment trasfer to column that was expressed as Mt=0,18*b*d^2*fck. This expression limited the slab section of becoming over-reinforced.
Now, if I calculated this using quivalent frame method, I had an option of reducing negative edge moments by not more then 50%, if calculated moments where higher then this, but then I had to increase the postivive momets in the span by the same precentage..
Im now using structural analysis program that uses FEM.
My question is, should I also limit the values of Momet transfer to edge column, and also reduce -moment/increse +momets if necesary or not?
Thank you

 

[Lion06]

what were you using to get that equation (.18bd^2)*(fck)?
I've never seen that before.
ACI's approximate method works well by hand.

 

[SnapSpace]

That the expresion dictated by Eurocode.
DO you have something familiar in ACI?

 

[Lion06]

If that is giving you so high of a moment that your seections is overreinforced for a slab I think you may be calcing something wrong. The moment at the end bay of a flat slab should be low (certainly lower than at the first interior column). If you are getting an over reinforced section at the exterior column and not at the interior columns, I would look at your analysis again.
Is you slab thickness constant?

 

[SnapSpace]

Yes, the thickness is constant.
But you got me wrong.
Does your ACI code limits the value of the moment that can be transfered from slab to column?

 

[Lion06]

It depends a lot on the geometry, but there is no way that the exterior column would ever see more negative moment than an interior column struip. It may be posible with some really funky geometry, but then you would'nt use the approximate method, you would need to use something else.
Can you correct the issue by increasing the slab by 1/2" or maybe even an inch?

 

[frv]

Yes.

Although depending on your shear, you could theoretically transfer all moment into a column.

 

[csd72]

I think that Mt is the limiting expression of how much moment can be transfered to the columns due to punching shear restrictions.

If this is the case, then to increase moment transfer you need to provide spandrel beams along the edge to transfer moment to the column via torsion.

 

[csd72]

Check out:

http://www.eurocode2.info/main.asp?page=1154&pdf=How2_Errata.pdf&return=1154&msg=download

You will need to register to get the downloads, but this should help.

 

[SnapSpace]

csd72 , egsaclty!
Basicly, what this expresion is doing is limiting your neutral axis depth to 0,5d insuring that the stell in your section (slab) would yiled.
If this is not satisfied, your slab would probably fail due punching mechanism.
But my question is, should I check what is the value of Mt, after my software provides me with the results of bendnig moment isocurves, to insure that the punching shear restrictions are satisfied???

eg. look at the picture,

http://img174.imageshack.us/img174/3379/fslabcopyld3.jpg

I have value of Mx=55,75kNm/m in the slab at the end column.

Mt for concrete strenght of 25Mpa and column 400x400mm, and slab thickness 200mm (d=180mm) is:
Mt=0,18*25*180^2*(400+400)= 116,64kNm/m wich is more then 55,75, so its O.K.!
ughhh I compllicated things now :-(

 

[SnapSpace]

I got some strange slab moment results at edge columns :-(

I modeled a simple structure, flats slabs thicknes 20 cm, rectangular column array 6,00 m spans, column 40/40 cm and storey hight 3,00 m.
Loads where: self-weight and 10kN/m^2 on slab deck.
Heres the structure:

http://img206.imageshack.us/img206/3079/s12tp4.jpg

and the rusults:

http://img142.imageshack.us/img142/2566/s32tk6.jpg

As you can see in the picture, Mx at the end column is the biggest one in the slab*!*!=!(&#
Is this normal?
I thought it should be the smallest one?
Can some one use this egsample and recalculate it in thier software.

 

[csd72]

Snapspace,

You need to look at the links and follow the examples step by step, I have not used eurocodes yet(that is a pleasure that I will have soon though!). signing up is free and there is really great guidance there.

a few points I would suggest:
1. check that you are allowed to reduce the end moment by the 50% according to the code, just because the software allows it doesnt mean that it is okay.
2. compare the bending to your Mt.
3. if the bending exceed Mt than you will need to design a spandrel beam as mentioned above to transfer the moment into the column.

 

[SnapSpace]

csd72,
Thanx for the response. I downloaded the "worked egsamples" from the link you gave me, and Im looking at the "flat slab" egsample...and yes they are using the expresion for Mt. that limits the value.
But whats wrong is that they didnt even calculate the moments at the end column, only at the interior column.
They used the Mt expresson and used that value as the "end moment".
I think this isnt such a great idea, expecialy if you look at my bending moment results that I got from computer analysis, wich clearly show that the bending moments at end columns are the greatests.
What do you think?

 

[csd72]

It is typical that the bending moment transfer into the edge columns is an issue, so you are on the right track.

The traditional method is to use beams along the edge of the slab that frame into the columns(called spandrel beams) these are then designed to take the end moment via torsion into the column.

If you have no room for spandrel beams then the thickness of the slab needs to be sufficient to transfer this load, this is often the limiting criteria.

Alternatively you can look at proprietary shear reinforcement, but I dont know how this fits in with either bending moment issues or eurocodes.

 

[SnapSpace]

Can you run the analysis in you software so that I can compare the Mx results, because I have a bad feeling abouth the results I got.
A friend of mine did the same egsample and his values of momets were very different expecialy at the end column.
Please
I described all the values few posts earlier.

 

[csd72]

Sorry, I dont have the time, but as a general rule if the end moment is less than 30% of the simple span moment(wl^2/8) then it is probably correct, around 15% is more common.

 

[SnapSpace]

Well, mine is around 10% greater than the mid span-moment!

 

[csd72]

Then you look like you have a problem with the analysis, or is the column a much stiffer section than the slab?

Check that you have not provided a fixed moment support at the intersection of the columns and the slab.

 

[SnapSpace]

Yes, I did that!
I modeled fixed connection with slab and column.
How elese would you get a bending moment in the column?

 

[csd72]

I mean support, not connection. Where have you modelled your support nodes?.

To help me get a better idea, what are your spans, sections, live loads and moments?