Liquefaction Triggering Analysis for a Submarine Condition

[moe333]

I have searched but unable to find guidelines/discussions for evaluating liquefaction triggering (on flat or gently sloping ground) for a submarine condition (lake bottom or ocean). In particular, the Cyclic Stress Ratio (CSR); will the total stress include the water pressure above the soil, and will the stress reduction factor (rd) be affected by the water column? A-max could also be affected by the confinement from the water column but rd would likely account for this.

Any references or comments appreciated.

 

[dgillette]

That's an interesting question. My initial reaction is to say no, the overlying water isn't relevant at all, but then I thought that if the velocity at the soil surface is high enough the viscosity of the water might have some effect, perhaps increasing CSR a little because some of the water gets dragged along by the soil? I sure don't know how I would quantify it without a research project, but I think the effect would be small, based on no actual data or analysis. (Data without theory are trivia; theory without data is bull$#!+.)

I would start out with the simplifying assumption that the water is "soft" - so the cyclic stress would come only from the total mass of the solid soil below the mudline. (I say "mass" because total overburden stress is used in the simplified equation as a proxy for mass in f=ma, so we don't need to figure out slugs and pound forces and all that. Maybe if HBS had been working in a metric country, it would have come out looking different. rd is sometimes called the mass participation factor - how much of the mass of overlying soil participates in f=ma?)

Strictly speaking, the total overburden stress includes the weight of the free water above the mud line (sigma=sigma'+u), but the simplified equation only makes sense if you consider the total overburden stress to represent the mass that gets accelerated by the cyclic stress on the base of a layer.

Any other ideas?

Cheers!
DRG
PS: Does Firestone ever make it down your way from Paso Robles? A friend passing through CO left me with a sampler case that included their summer ale (which I liked), and their Double Barrel Ale, which I really liked. Their IPA was a bit too hoppy for my liking, so I'll save it for you if you'll be here soon.

 

[dgillette]

About 5:30 this morning, while lying awake cursing daylight savings time and the forces that wouldn't let me sleep until the alarm goes off at 6:15, it occurred to me that the peaks of horizontal acceleration and the peaks of surface velocity would be out of phase. If the viscous force is proportional to the shear strain rate in the free water, it would be highest when the soil surface velocity is highest, which is when the ground acceleration is lowest. Therefore, the effect of the water viscosity on peak cyclic shear stress should be very small.

What?

 

[moe333]

My best thoughts usually come in the middle of the night. I've been OK with the time change but my 6 year old is haunting me by appearing at my bedside at 5:30 because he still thinks it's 6:30.

My initial thought was Rd and therefore CSR would increase because the weight of water would make the soil profile stiffer. But I guess if Rd is proportional to effective stress this wouldn't be the case.

Never thought about the viscous force, what you say makes sense. I'm surprised I couldn't find any references on this..particularly for evaluation of bridge piers/bents.

I have tried the Double Barrel Ale before...can't recall how I liked it but I'm sure I finished the bottle. It's beer week down here..maybe I'll try it again. No trips to CO in the works.

Thanks, and Cheers!

 

[dgillette]

It's not that rd is proportional to eff stress, it's that the shear stress has to be proportional to the mass that gets accelerated, the m in f=ma. m is equal to the total mass of the overlying soil and its pore water, but not the free water above it (if you assume it to be "soft" so it just sits there while the ground vibrates horizontally beneath it). On land, m per unit area above some depth is conveniently equal to sigma-total/g, in slugs/ft2 or kg/m2, and that's how it shows up in the simlified eqn. If not for viscosity of water, rd under water should be identical to what it is on land, with depth measured from the soil surface. Maybe a little different with viscosity of water included. (rd isn't 1.0 at depth because the soil isn't rigid. It depends on freq content and SWV profile.)

Know a numerical modeler who needs a master's report topic? (or maybe a thesis topic?) This could be it for someone who can figure out how to estimate the viscous drag (which I'm sure has been done already by someone).