Liquefiable Soils - Where Is The "Base" of the Structure?

[WAstruc10]

I am designing a large but lightweight one-story steel building, similar to a PEMB. The foundation system is augercast piles due to liquefiable (seismic) soil conditions. The slab-on-grade is considered sacrificial and non-structural (so if the subgrade settles 7 inches during liquefaction, the slab will crack and settle along with it). But the interesting part is that this slab on grade over half the building footprint is 4-ft above the surrounding grade level, so there are stemwalls acting as retaining walls around the half-building perimeter to restrain that 4-ft height of backfill beneath the slab. The perimeter pilecaps are below the lower site grade, and there are concrete pilasters built into that stem/retaining wall that in turn support the steel columns & braced frames. My questions:
1. With the liquefiable soils, do we need to include the weight of the concrete foundation system in the seismic analysis?
2. If so, I imagine it would be a 2-stage analysis with the steel structure analyzed separately, but what "R" value would be applicable to the concrete system? I'd like to avoid detailing everything as special concrete moment frames for obvious reasons.
3. If we have to include the concrete weight in the lateral analysis, would we also need to include the weight of the 4-ft of soil retained BETWEEN the perimeter stemwalls? That would be an insane amount of weight.

Of course in typical soils all this would be moot, but the liquefiable issue has my brain turned sideways. Thanks!

 

[WARose]

1. With the liquefiable soils, do we need to include the weight of the concrete foundation system in the seismic analysis?

I would. I'm assuming here the stemwalls & slab on grade have no ties to the LFRS [Lateral Force Resisting System].....so I wouldn't include them in seismic dead load. But the pilecap and some of the piles? Yes.

You can mitigate some of this by Chapter 19 in ASCE 7-10 that can give you some reduction in the seismic load.

2. If so, I imagine it would be a 2-stage analysis with the steel structure analyzed separately,....

ASCE 7 has such a procedure in Sect. 12.2.3.2.

3. If we have to include the concrete weight in the lateral analysis, would we also need to include the weight of the 4-ft of soil retained BETWEEN the perimeter stemwalls? That would be an insane amount of weight.

No. (Assuming I am visualizing this correctly.)

 

[WAstruc10]

Thanks. So I think you agree that a 1-stage analysis would not be appropriate (shifting some of the concrete weight up to the steel roof system, in effect). I agree. But if a two stage analysis is appropriate, what R would you base that on? Doesn't make much sense to me to assign an R value and calculate base shear for a pure foundation system by itself (plus reactions from the first stage analysis of the steel structure above). And what is your reasoning for not having to include the weight of the soil that is sandwiched between the above grade perimeter stemwalls? Again, that soil supporting the slab on grade is 4-ft above the lower grade outside the bldg footprint.

 

[WARose]

But if a two stage analysis is appropriate, what R would you base that on?

See Sect. 12.2.3.2 in ASCE 7. You'd use the R-Value normally used for the steel frames above. Do the ELF for that portion.....then do the same thing below with the appropriate R-Value for the RC frames. In the second stage, the reactions (from above) would be amplified as per 12.2.3.2.d.

And what is your reasoning for not having to include the weight of the soil that is sandwiched between the above grade perimeter stemwalls? Again, that soil supporting the slab on grade is 4-ft above the lower grade outside the bldg footprint.

Is the LFRS supporting it? If the answer to that is "no" than I don't see how it is seismic dead load. If anything, I would think it would dampen the motion......that is assuming the stem walls hold up in such a event.

All this is hard to say without a pic.