2 way or 1 way slab design 1

[haynewp]

Is there an ACI Code provision that states you have to treat a slab as 2 way if it meets the 2 way dimension criteria? For instance if you have a slab that is supported on beams all around and it is a square in plan. Why not just design it as a 1 way slab?

What problems could this bring about by doing this? Excess cracking under service loads in the weak direction? etc...

 

[JAE]

If you have a square-ish slab and treat it as one-way only, then your weak direction rebar will most likely just be temp/shrinkage and yet the slab will initially (pre-cracking) behave as a two-way slab simply due to its geometry.

After it cracks, under significant load, then it will allow further crack openings in the weak direction as the temp steel stretches and ultimately yields...but the competence of the slab will remain in the one-way direction - i.e. safe.

 

[haynewp]

I think the one way treatment be more economical in some cases since you don't have to design for 100% of the load in each direction.

For instance, what about a slab that is supported only on columns. Why not design the slab using the column and middle strips with 100% of the load in only one direction instead of two, if greater cracking can be tolerated?

 

[JAE]

haynewp - there have been lots of projects where I have a one-way slab "system" throughout - perhaps supported by repetative beams where there are occasions when a small portion of the slab gets to be square - and two way action is going to happen. I still design for one way in these situations as I don't like to mix one-way and two-way systems in the same project unless it is a significant area involved.

With just slab and columns - I think I'd stay with the two-way throughout. Even with a more rectangular bay layout - you'd still get two-way action happening and I wouldn't be comfortable with a one-way slab-column system as I've never seen documentation, examples, or code that covered it.

 

[UcfSE]

Another point, if you design the slab as one-way you will have to meet one-way deflection criteria. You're slab will be much thicker than need be. It's easier to design but much less efficient with material. With that said, I agree with JAE 100% about cracking problems. I might add some bars at the perimeter to help with the cracking when it transitions from two-way to one-way action.

 

[rapt]

haynewp,

Are you suggesting that in a two way slab supported by columns that 100% be carried in one direction and 0% in the other direction?

If so you had better go gack to school. You have to reinforce to carry the total load in each direction spread between column and middle strips for STRENGTH, not just crack control. FEM analysis backs this up. If using yield line, you can reduce this slightly for the negative moment area but definitely not to the extent you are suggesting.

An alternative is to reinforce 100 averaged over the total width in one direction and 100% concentrated over the column line in the other direction (reinfroce as a onw way system). AS JAE pointed out above,
1 crack control is much worse and reinforcement is required in the other areas anyway. Even if crack widths are not considered to be a problem, completely uncontrolled cracking as you would get in this case is always a problem and reinforcement must be added to control it in unreinforced araes. If not enough reinforcement is placed in these areas, the reinforcement provided will yield and will be useless anyway so extra calculation is required as is extra reinforcement.
2 deflection is much worse as severe redistribution of elastic actions is occuring at service loadings. This is not a good thing to have happening in a concrete slab.
3 the design methods required to allow for 1 and 2 above will be extremely onerous and would increase the design time erormously. You could not use normal deflection calculation methods to calculate the deflections or the crack control reinforcement. Basically an elastic/plastic analysis allowing for long term effects would be needed.
4 it is not as economical wrt materials.

 

[haynewp]

That sounds like an overly conservative Australian way of thinking to me.

 

[haynewp]

Just joking with you...

I now realize how stupid my idea of a column supported slab acting as I had suggested really is. I just didn't think about it before I posted it.

 

[rapt]

haynewp,

Great to get that sorted out. You had me worried with that suggestion. I did not think that we Australians were that conservative.

Getting back to your original question of a square slab supported on 4 sides by beams, you only have to design for 50% of the load in each direction. This will give the most economical design.

Theoretically you can design for all of the load in one direction but
1 crack control in the other direction will be very bad
2 deflections will be worse as the slab is being expected to carry the full load in one direction (assuming you have not reinforced the other direction for crack control)
3 what are you going to design the beams for? In fact, you can leave out the beams in one direction as the beams in the other direction are carrying the full load.

As a result of all of this, you will add reinforcement in the slab in the other direction for crack control. Once you add some you must have enough to ensure the reinforcement does not yield. You may as well design it as two way with 50% of the load carried in each direction and get the most economical result and the best serviceability for the support layout you have provided.

 

[haynewp]

I could just tell from your handle and your writing style who you were and knew you were from Australia. I have emailed a couple of times with you in the past regarding PT slabs.

(Thinking about coriolis effect.....)