two-way slab design to ACI318-08
two-way slab design to ACI318-08
(OP)
I recently started to use ACI318 to design two-way slabs (with and without beams). I got two questions and I will be grateful if I can obtain some help.
1) Calculate the total static moments in the long and short directions (clause 13.6.2.2). This will give a smaller value in short direction. Slab in short direction is actually more stiff (wider in cross section) in comparison to slab in long span direction and should it take a bigger bending moment? Since the deflection at the centre must be identical in both directions.
2) Proportion between negative and positive bending moment at interior panel (13.6.3.2), 0.65 and 0.35. The latter times Mo, ql2/8, gives equivalent bending moment at support and mid-span, i.e. 0.08125ql2 and 0.04375ql2. These do not seem to be conservative in some scenarios, e.g. mid panel of 3-span slabs. By the way, I don't think any moment distribution has been taken into account in above coefficients, as stated in R13.6.1.7.
I wish to thank you all in advance.
1) Calculate the total static moments in the long and short directions (clause 13.6.2.2). This will give a smaller value in short direction. Slab in short direction is actually more stiff (wider in cross section) in comparison to slab in long span direction and should it take a bigger bending moment? Since the deflection at the centre must be identical in both directions.
2) Proportion between negative and positive bending moment at interior panel (13.6.3.2), 0.65 and 0.35. The latter times Mo, ql2/8, gives equivalent bending moment at support and mid-span, i.e. 0.08125ql2 and 0.04375ql2. These do not seem to be conservative in some scenarios, e.g. mid panel of 3-span slabs. By the way, I don't think any moment distribution has been taken into account in above coefficients, as stated in R13.6.1.7.
I wish to thank you all in advance.






RE: two-way slab design to ACI318-08
As far as your questions, there's a lot of magic and mystery involved with the two way slab method. If you pick at it too deeply, you'll find all kinds of inconsistencies. Yet it's been used literally thousands of times and the buildings don't fall down.
RE: two-way slab design to ACI318-08
The direct design method accounts for the various continuity factors and re-distributions of moments. Refer to the commentary under R13.6.
The problem with analyzing the method is that it fudges a bit of the math to give a workable method of design. Going back to first principles or using FEA will get you a different design, but one which may well not hold up to scrutiny as well as the Code methods.
This is where the Code actually makes us more productive - allowing us to get to a final design, rather than dwelling on the essence of the problem.
RE: two-way slab design to ACI318-08
The ratio of two spans is well under 2 and the direct design method can be used. But I don't think any moment redistribution is allowed by the direct design method as stated in 13.6.1.7:
"13.6.1.7 Moment redistribution as permitted by 8.4 shall not be applied for slab systems designed by the direct design method."
My questions were coming from the comparison between ACI318 and Eurocode for concrete design. For instance, positive moment at mid-span of an interior panel is 38.4% (when l2/l1=2) and this is about 10% higher than 35% being used in ACI. Moment redistribution is not allowed in the EU code either.
FEA will be an indispensable part of my job as pattern loading (+ irregular spans in both directions) play an important role and this cannot be sufficiently investigated using the direct design method. My FEA peak values at column strips are about 15% higher than the ACI values but the opposite is true at the middle strips. The peak bending moment value obtained at middle strips is about 3 to 5 times greater than that determined using ACI method (65% negative, 35% positive). Has anyone come across such a case in the past?
Oversee
RE: two-way slab design to ACI318-08
As others said, it is intended to simplify things, and it works so long as you meet the limitations.
I have not looked in a Eurocode, but if the numbers using two different methods are within 10% I think that is acceptable given load/phi factors and how sketchy our loading values are.