Waffle Slab

[80smoviefan]

Hi all,

I have a couple questions regarding the design of a waffle slab (i.e. perimeter beams with internal beams under load-bearing walls, polystyrene pods between beams and mesh and slab on top).

My first question is say you are designing on liquefiable soil and you have to consider loss of ground support or loss of bearing over a certain radius (to imitate liquefaction). Now if I was to lose support at the support of beam (say at an intersection of 4 beams), would I analyse the beam as simply supported, pinned at each end, despite the loss of bearing at its support?

What I trying to say is, our those intersecting beams still providing support for the beam I am analysing despite the loss of bearing at that support?

My second question is in regards to the load paths of point loads. Say I have a point load applied, lets say where 3 or 4 beams intersect, now can I rely the point load evenly spreading onto each beam - and how much distance of concrete could I rely on for the point load spreading across.

My thoughts were at least the depth of the beam (due to the 45 degree concrete angle concept) plus an additional amount due to shear reinforcement. How much realistically do I have?

Third and final question is if pads are required for areas under high point loads - how are the pads designed to isolated pads - due to the presence of beams connecting to the pad. Currently I am simply dividing the point load and distributed load by the allowable bearing pressure to give me a required minimum width - is this concept still valid?

Sorry for all the writing - just designing my first slab.

Thanks

 

[asixth]

Are you design this to AS2870-Design of Residential Footings. If you can get a copy of AS2870 it has a prescriptive section.

If the soil lacks bearing I would be detailing a piled supported waffle slab but piles, even timber piles, can be really expensive. Other than that you design the ribs the same way that you would design any concrete beam. If the end is not continuous than it's assumed pinned, if it is continuous than it get's designed taking into account some level of continuity based on what happens in the next span.

Point loads get distributed based on the relative stiffness of the two-way system. If span in one direction is shorter than the other than the load will distribute that way. I will put up an old copy of 2870.

https://www.dropbox.com/s/5dil0u9hu7mk8m6/AS2870 Residential Slabs and Footings.pdf?dl=0

 

[KootK]

You may find it helpful to evision the waffle slab as more of a conventional elevated slab (piles) or slab on grade (No piles). While the waffle slab ribs suggest "beams", the grillage assembly taken as a whole really behaves much more like a solid slab or slab on grade. And like a solid slab, load distribution will be imperfect but still great. Much better than a 45 degree load spread.

This is an excellent reference for simple methods to estimate load distributions in grillages: Link

As for your specific questions:

1) You would analyze it like the true two way system that it is (like a solid slab). If it's pile supported, then the loss of soil support is irrelevant. If it's grade supported, then you would analyze it as a slab on a bunch of soils springs with springs missing where you've got the liquifaction. The crossing beams still definitely provide support. I'm sure that there are simplified ways to tackle the design and you'll likely find those methods in the documents that Asixth and I have recommended.

2) In a grillage, as with a solid slab, load distribution capabilities are excellent, albeit a bit difficult to assess. Options include a true grillage analysis or simplified methods discussed in the references above. Certainly, I think that you'd be conservative with the 45 degree load spread assumption.

3) That concept is still valid, if perhaps a bit conservative due to the waffle slabs excellent load distribution capabilities. That's how I'd handle it.



I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[JAE]

Your main question seems to be how the various intersecting beams help out.

Force follow stiffness so any intersecting beam would help to some degree. The difficulty is estimating soil and beam stiffnesses.
This is a highly indeterminate analysis that would requrie software to analyze with any accuracy.

KootK is right about using soil springs to mimic the soil.

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

 

[80smoviefan]

Ok, so how much concrete do you think I could roughly rely for spreading the point loads. Say my beam is 400 mm deep - would 600 mm be conservative?

 

[JAE]

It depends on the relative stiffness between the beam and the underlying soil.
We can't give you an answer more than that without literally guessing.
Your estimate/guess is as good as we would offer I think.


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

 

[KootK]

I need to refine my answer as I may have led you astray. Waffle slabs, again like flat plate slabs, actually kind of suck at handling localized shear concentrations. Hence the punching shear phenomenon and the practice in suspended waffle slabs of filling the waffle holes with extra syrup around the columns.

If your load happens to land at at a rib crossing, I'd say that you've got at most two ribs participating in shear. If your load lands on a rib mid-way between crossing ribs, you may mobilize only a single rib in shear. The fact that you're unlikely to have hanger steel at your rib crossings makes it even less likely that you'll mobilize more ribs in shear.

Sorry for steering you wrong in my first post. Waffle slabs are only great at distributing moment laterally.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[KootK]

With regard to the distribution of bending moments and soil bearing stress, you could simplify the analysis by essentially treating the problem as that of a simple footing. Disregard the top steel and make the extent of your fictional footing the distance over which flexure will still work assuming a uniform resisting soil pressure. Yeah, the assumption of uniform soil bearing stress is bunk but, then, it's also bunk when we design footings and we never seem to be bothered by that.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.