forcing one way bending in a two way slab

[geopat69]

Hi everyone

In the past, i have been often told that it is possible to design a 2 way concrete slab as one way system -> "just assume ALL the load spans in the shortest direction".

The above idea, at first, seems fine from a strength perspective, clearly if it can handle all that load for strength in one direction then the system is fine!

But I am now questioning this, more from a serviceability view! In particular an idealistic case of a concrete slab simply supported on 4 sides, for example.

If only bottom reinforcing is placed to span in the short direction only, wouldn't the slab want to span in the long direction regardless..AND...then form a crack due to no bottom reo being provided in the long direction???

I guess what I am grappling with is the "idea" to force a slab to span in short direction (by providing reo in that direction" only to have it behave in a different way and "crack" due to it finding a more desirable load path i.e. finding the shorter edges in this case, even though long reo does not go in that direction).

In summary, I understand that the one way analogy is good for conservatively designing for strength (no one cares about cracking for strength calc's!!), but the one way analogy makes me nervous for serviceability!!


Appreciate some feedback on the above.



regard

Geoff

 

[BAretired]

You are absolutely correct. But usually, there is temperature and shrinkage reinforcement placed in the long direction so that cracking is at least partially controlled.

It also depends on the ratio of long to short side. If you are looking at slabs where the ratio is very high, the slab will perform pretty much as a one way slab...but you may want to add a little longitudinal reinforcement at the ends to reduce cracking.

If the panel is nearly square, it is probably a bad strategy to reinforce it as a one way slab.

BA

 

[rapt]

Geoff,

You are correct.

For strength, providing a load path is the important consideration and a plastic or yield line approach can be used to satisfy this. But that does not provide crack control, and , depending on the chosen load path, can also make deflections worse if a lot of redistribution is required.

But the code requires that more than strength is satisfied.

To satisfy serviceability crack control and deflections must be checked, and crack control is dependant on the elastic load path, not the ultimate strength plastic load path that the designer has used. If there are tensile stresses in both directions, crack control reinforcement is needed in both directions, and it cannot jut be satisfied by supplying nominal crack control reinforcement, the amount required is related to the level of stress in each direction.

e.g. if you had a square panel, logically each direction could be designed for strength for wl^2/16. But also wl^2/8 in one direction with nothing in the other direction would work for strength. But for crack control, the original reinforcement required for wl^2/16 would be required in both directions. If this was not done cracking in the lesser reinforced direction would not be acceptable, and deflections would increase as the slab is essentially one way!

 

[KootK]

For the benefit of any future readers of this post, we should point out that one way design is only acceptable for two way slabs that are supported on walls or beams. For two way slabs supported on columns, one way design would be a strength problem in addition to a serviceability problem.

I've seen recommendations that a slab can effectively be considered one way once the aspect ration becomes higher than two to one. While agree with the sentiment, even with an aspect ration of 100:1, you're still going to get very two way elastic behavior at the corners. Most designers will detail the rebar in these areas to reflect two way behavior even if the reinforcement supplied is nominal rather than based on strength computation. Many codes also have provisions for corner top steel which should not be waived even if one way behavior assumed for design.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[asixth]

And as a minimum there should be distribution reo in the secondary direction. It shouldn't be used in strength calculations but it does help the load to distribute itself into the strip that you are designing for. As a minimum I use 12mm dia bars at 300mm centres.

 

[rapt]

KootK,

Correct. I was assuming he was using the definition of a two way slab being a slab with continuous supports, while a Flat slab is a slab supported on discrete column supports (British definition at least).

 

[geopat69]

Hi guys/.. The responses so far are excellent and have really cleared things up - I do thank you for your contributions.!

Geoff