Slab to help resist sliding?

[STRBeth]

I have a building with structural steel columns on spread footings. The lower level is a slab on grade, reinforced with WWF. The columns have isolation joints around them. In the event of an earthquake, can I count on the slab to do anything to resist sliding? I realize that the columns would have to move enough to compress the compressible material in the isolation joints and then engage the slab. If I was comfortable with this movement, are there any references or info that would help me quantify the slab's contribution?

 

[msquared48]

If there is any chance of liquefaction, I would not rely on lateral stability from the slab for sliding.

However, if not, if the slab is large enough, I don't see why the connection could not be detailed to achieve this.

Mike McCann
MMC Engineering

 

[STRBeth]

No chance of liquefaction. Mike, how would you detail it? If I get rid of the isolation joint, then I feel like I'm setting myself up for problems. And during an EQ, I'm not sure I care that the footing moves enough to compress the material and engage the slab.

 

[ishvaaag]

Till static friction is overcome, both footings and unconnected slabs will move with the ground. In this situation most of the weight will be being imparted directly to soil and so scarce collaboration of its weight can be assigned to resist sliding from the main structural body.

Soon after that, the slab will be thrown against the columns, or viceversa. Quite likely the pounding of one system on the other will be a dampening effect for the movement; to which effects both slabs and main structural system must be able to survive such pounding action.

Since pounding action may come to work, I doubt any of the ordinary structural software packages will be able to model the interaction. Just realizing that, one either would produce quite stiff joints able to consider continuity all along movement, or most designers wouldn't account the slab weight as contributing against slippage at least (except when monolithic).

 

[SAIL3]

I personally would not count on friction to resist seismic loads....seismic also has a vert component.

 

[STRBeth]

Sail,
How do you resist seismic then? All through passive pressure? I'm considering the uplift due to seismic, but obviously the weight of the footing counteracts this. So that net downward load is good for resisting sliding by friction, no?

 

[msquared48]

STRbeth:

Just use greased dowels at the joint. The movement to compress the material will not be inhibited, yet the alignment maintined.

Mike McCann
MMC Engineering

 

[ATSE]

I'll take friction over passive. Even with a vertical seismic component, you're not going to have a flying carpet slab.
If you use the slab to resist lateral via sliding friction (or restrain the lateral kick at moment frame column bases, for the case of PEMB), you need to change your slab joint detailing and reinforcing. Also, does your slab have a vapor barrier? That changes things as well.
You will need to treat the slab as a structural diaphragm. Replace the (near worthless) WWF with deformed bars, and make the reiforcing continuous across all joints, at least at bands aligned with the columns. Note that this will probably create other slab cracking problems. It's a trade off.
Isolation joints are okay unless they have compressible material separating the slab and the slab concrete encasing the column.
For a reference, ACI 318 1.1.7 and 21.12.3.4 (and commentary).

 

[racookpe1978]

I'm missing something - can you explain?

Is the "sliding" a vertical movement of the columns past the slab if the soil (under the slab and columns) liquifies?

Or is the "sliding" a (potential) horizontal motion of the columns getting "kicked" sideways at the bottom, but that motion is expected to be resisted by the holes (and fillers) in the slab around each column?

In the first case, you'd need rebar linking the column and slab strong enough to resist the shear (up then down) during the max quake as loads resonate. In the second case, the slab would tend to be pulled apart in tension between columns when (if) the columns were pulled sideways (which the WWF seems to be too little a reinforcement to resist), since the whole slab (or just broken parts of it) would be unlikely to be able slide sideways the same distance as the columns.

 

[InDepth]

STRBeth,
this depends on the amount of load you have to resist. I would avoid feeling comfortable with compressing the gap since this adds to pdelta effect.
1) First check if footing can resist by sliding friction
2) Second check if sliding friction combined with passive on one face and side vertical side friction on two perpendicular faces (trapezoidal dist based on depth).
3) If you still need more capacity, then consider increasing the depth or dimensions of the footing. You could even add a small key beneath the footing like a retaining wall to increase passive.
4) Need more capacity, then add a grade beam connecting to other footings to activate their resistance. This grade beam could be cast into the side of the footing, top of footing, or even poured to be integrally connected to sog.
5) Alternatively, Detail the slab on grade with downturned edges that attach to the top of the footing. This of course will assume that settlement of the footing will be minimal and the sog is "flexible" enough. Treat slab as a diaphram or as a way to transfer load to other footings (like grade beam concept).
6) Need more, geez...ok...so a calc exercise...ok if footing has grade beams on all four sides (or 3 if an edge condition) then use the side friction on the grade beams parallel to the direction of force and then add the passive resistance against the grade beams perpendicular to the footing. ie. Use an effective width of grade beam to add passive resistance to the total footing passive resistance.
7) Need more, ok ok...now you are in pile territory

 

[InDepth]

On slab contribution you will have to analyze your typical detail for sog. I doubt your geotech will give you a friction resistance for the gravel, sand, plastic visqueen over soil detail. Typically the vapor barrier is assumed to break the soil slab interface friction. Even if you got a friction value the weight of your sog will only contribute a small resistance to the total.

 

[Teguci]

ATSE mentioned it but I will highlight. If the slab is designed as a typical slab on grade, it is not required to meet ACI 318 requirements. However, if structural loads are required to be transferred through the slab on grade, then the design requirements for the SOG are elevated to the requirements of ACI 318. If this is an existing condition, I would assume that you can't use the slab to transfer structural loads.

 

[InDepth]

Even if it's existing and the calcs work out...then it works.