lateral spreading analysis

[DougHole]

i am doing a newmark sliding block analysis ("Jibson 2007") for a slope to estimate the lateral displacement under seismic conditions. To do this, i need to determine the yield acceleration. This is usually done by tweaking the acceleration coefficient in Slope/W until the FoS is 1.0. i have some liquefiable sand.

For this analysis, should i be using liquefied parameters for the sand (Su approx 10kPa) or conventional parameters (phi'34 deg)?

 

[EireChch]

Is this a lateral spreading issue? (i.e sloping ground or free face underlaid by potentially liquefiable soils? If so you need to do a liquefaction analysis. This will tell you depth and thickness of liquefiable layers. You would then use this information as some of the inputs in a lateral spreading analysis.

When you say critical acceleration and sliding block method I t sounds like a shaking induced slope displacemt displacement. (Different to lateral spreading from liquefiable ground).

I have used the Ambreseys and Menu method to assese seismic induced slope displacemt. Modelled the slope under worst case ground water conditions and using static parameters. Determined yield acceleration as you have discussed. The read it off a chart. A simplistic approach. I think Jonathan Bray has a new approach which is good. I think he even gives out a free spread sheet if you ask nicely!!

You can of course have a combination of the two mechanisms but this is getting a bit complicated and may need a finite element analysis like plaxis?
Hope this helps.

 

[DougHole]

Hi EireChch

What is the difference between "shaking induced slope displacement" and "lateral spreading from liquefiable ground"?

with my situation, it is a slope into a river which i guess is a "free face". My liq analyses, indicated that i had liquefiable soils between 6-8 m below ground level. No other soils above or below this are liquefiable. i think there should be 3 analyses to assess "seismic stability" (that's my 'catch all' term for lateral spreading/seismic slope stability stuff!")

1. Pseudo static but using all static parameters (in Slope/W) i.e. apply the PGA as the seismic coefficient in Slope/W. Output will be a FoS and i'm shooting for 1.1. is this what you mean an "Shaking induced slope displacement"?
2. Post seismic analysis i.e. do the slope/W analysis using the liquefied parameters for the liquefiable soils (6-8m depth) but no "shaking" i.e. 0g as the seismic coefficient in Slope/W. the output will be a FoS and i'm again shooting for FoS>1.1.
3. A displacement analysis. For this, i will use Jibson 2007 (or i could use Abraseys &Menu) i think these methods are similar. As mentioned before this involves playing with the seismic coefficient in slope/w until you a get a Fos of 1.0. This is the "yield acceleration". you then plug this into the Jibson formula or look up the Ambraseys chart.

Eirechch - does this sound right? or should i be doing something else?? (Youd 2002?)

 

[EireChch]

In my opinion shaking induced displacement typicly occurs on a hillside or steeper slopes, normally abovethe ground water table so liquefaction is out. The soils are overloaded from the additional seismic loading forcing them to "shuffle" down the slope.

Liquefaction induced spreading is when liquefiable soils are located near surface beneath gently sloping ground or near a free face. This typically occurs near rivers, changes in topography.

There are of course grey areas, like at the foot of a hill which are underlain by both alluvual/loess soils and the grousnd water level is near surface. These are quite tricky to asses in my experience.

You seem to be covering all basis by I'm not sure if you need to do that much.

Could you upload a typical cross section showibg slope angle to the river back, height of the bank, and also include some geotechnical information. I could be be to give a better answer then.

I would nearly guarantee that you will have virtually no lateral spreading if you do a lateral spreading analysis. A 6m thick layer of non liquefiable should not spread.

Also if the slope angle is relatively low (<10-12 degrees, Ambreseys and menu or the likes might not give any displacemt.