Concrete Slab - Compression Reinforcing & Long-Term Deflection (ACI 318)

[jtse]

Hello all,

Does anyone know of any technical references regarding the use of slab reinforcing (top and bottom mats) as being effective in compression (due to moment, not axial loads)? I'm trying to determine if I can reduce long-term deflections per ACI 318-11, Section 9.5.2.5, by considering the bars as effective in compression. ACI 318-11 Section 7.11.1 requires compression reinforcement for beams to be tied, so I will follow this requirement (and disregard the steel, since tying it isn't practical) unless I can find something that allows its inclusion in slabs.

For background, I'm working on a unique 2-way reinforced concrete slab structure with a few challenges:
- Irregular bays and skewed slab edges
- Large cantilevers past exterior columns (8' to 10' from grid line)
- Very heavy sustained loads on cantilever (1,500-2,000 lb/ft along edge at worst case)
- The architect wants a flat panel appearance (no beams or drop caps); we're trying to limit total slab thickness to 12"

Any references or engineering judgments/opinions will be appreciated! Thanks!

 

[TehMightyEngineer]

Welded wire reinforcement might get you some credit for "tied" in compression as the transverse wires will laterally support the longitudinal wires.

Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

http://www.americanconcrete.com

 

[JAE]

While section 7.11.1 requires compression reinforcement in BEAMS to be tied, it is silent on the slabs.
However, the use of compression reinforcement in two-way slay systems is not really implied or anticipated by ACI.



Check out Eng-Tips Forum's Policies here:
faq731-376

 

[SethGuthrie]

A colleague of mine wrote a paper on "Accurate Long-Term Deflection Prediction in Flat Slabs" that you can download here:

https://communities.bentley.com/oth...s/member_folders/m/jason_colemans_files/87861

(you will need to be logged into Bentley Communities to download a copy).

In the method, which also happens to be used in Ram Concept, compression reinforcement is always considered, whether the cross section is actually cracked or not. Uncracked transformed properties are used where the cross section is not cracked.

 

[rapt]

Tied in compression requires ties from the compression face back into the tension face. WWF will not do that for you. Whether or not tieing is required would depend on the level of compression stress in the compression face steel. In a slab it may not be that high depending on the reinforcement depth relative to the neutral axis depth. So it may not need to be tied.

The effect of compression steel on long term deflection can be calculated. But to get reasonable answers, you need to do proper cracking, tension stiffening, shrinkage and creep calculations. The code simply factors are no use for it.

It depends on the distance of the reinforcement from the neutral axis. If it is near the neutral axis, its effect will be zero. If it is fairly highly stressed, it can have a significant effect. The effect does not completely negate long term deflection effects.

Its effect on creep depends on the ratio of (kd - dc) / kd (ratio of distance of the steel from the neutral axis top the neutral axis depth )
its effect on shrinkage depends on (kd - dc) / (d - kd) (ratio of relative distances of the compression and tension steels from the neutral axis)


Just letting us know the cantilever length does not help. The length of the first internal span is very important for cantilever deflections. But 12" deep sounds very tight.

 

[jtse]

Thanks for the replies, everyone.

Seth, that paper is very interesting. I wish we had a license of RAM Concept. If we see more buildings like this we may have to get that module added to our subscription. I'm currently running RISA Floor ES, which (if I understand correctly) runs a linear-elastic analysis and applies a cracked section factor to the entire slab: 1.43 x 0.25 = 0.358 Ig (per ACI 318-11 Section 10.10.4.1 and Commentary). It currently doesn't include creep/shrinkage/long-term deflections in the calculations, so I'm left to export the dead load deflections and manually increase them, then check against Code requirements. It's pretty rudimentary compared to the paper you referenced.

rapt, the structure has 1 backspan (varying from 20' to 24') with an 8' max cantilever (in 2 directions at a corner). I agree that the behavior is quite complex, and I'm looking for a simplified (quick, for investigating multiple design options) approach to approximate the deflection. ACI 318 Eqn 9-11 is intended as such, and my original thought was to see if there is justification for applying this to a slab with untied reinforcement. Based on my judgment and the responses above, I don't think so, especially given the low strains in the slab steel due to its shallow depth (when compared to a beam).