This has been discussed in another thread(s). One way of figuring it is estimating Su by ratio of Su/effective O/B pressure = 0.22 to 0.25 typically. You can then estimate the strength gain due to "full consolidation" of a stage.  Partial consolidation can be handled by %consolidation, approximately.  If in stage loading, you will want to use pvd wick drains (or other means) to help speed up the consolidation to a reasonable time frame.  Was involved with 10m high RE Walls on soft (Su = 400 psf) clays - built in stages with pvd. Walls underwent more than 800mm (2.6 ft) of settlement or more. Try search of other threads for more detailed discussion.

You also need to examine the stability of the completed embankment under fully drained loading.  The factors of safety could be lower, depending on the long term strength of the materials, than the condition you are currently looking at.

My experience has been with MSE Walls, so I may be offbase, but the heights you mention appear very high compared to the strengths you measured. I doubt that you build a slope to 40 ft on that material, depending on the depth/thickness of the soft clay.  Here are some references.  I didn't add the full reference, but there is enough that you should get results from and internet search.
Holtz and Kovacs, An Introduction to Geotechnical Engineering
Ladd and Foot, NEw Design Procedure for Stability of Soft Clays, ASCE Getech Journal, 1974
Ladd, stability Evaluation During Staged Construction, ASCE Geotech Journal, April 1991,

A comment/clarification: I doubt you could build the embankment to a height of 40 ft in a single stage of undrained loading.  

Clarification:

First stage embankment height is 30 feet.  I typed the wrong height in the second sentence second paragraph.  I am using UTEXAS3 for the analysis.

I am/will look at the undrained analysis.

In my literature search I found the Ladd, 1991 paper, but can not find a copy.  Any help?

Thanks for the responses.

If you can't find the 1991 ASCE geotech journal at a local library (USA university libraries with civil engineering programs usually have it)  You can request the document through the Linda Hall Library. There is a link on this page called "Purchase Information" that takes you to their website.
http://pubs.asce.org/WWWdisplay.cgi?9100971

I doubt that long term (Drained Conditions) would be controlling or a soft to firm soil (normally consolidated as was mentioned at the opening thread.  Still, if you have stiff to very stiff clays, there is a point to be made for the possibly HOC condition and long term strength may very well govern.  

I tend to think that 30 ft is rather on the high side of what you will be able to construct unless you are dealing with the higher end of the range.  If I remember correctly, Tchebotarioff in his 1973 book had a chart of embankments vs undrained shear strengths showing typical heights at which local yielding would occur - and ultimate.

I agree that the drained case could not possibly be the critical case (because the material sounds to be normally consolidated), but I still think it is a good check. It gives you and idea of the limiting conditions you could reach, if you could imagine the construction being drained. In other words, if the drained case has an an unacceptably low FS, then forget constructing it without ground improvement. On the other hand if the factors of safety from the drained analysis are acceptable, it only means that you now need to concentrate on the undrained conditions.
I agree that 30 feet seems high for undrained construction. If the soft layer is not very thick, the stability analysis, may indicating that the factors of safety are acceptable, if you only check circular mode. If this is the case, you need to check the wedge failure mode, and check it with Spencer method or Morgenstern Price.

More information:

The soft layers are thin.  There are two low strength layers in the foundations.  Silt and Clay (C=700 psf) layer about 2.5 feet thick underlain by a Silt (C=200 psf) layer about 2.5 feet thick.  Silty Sand is above the Silt and Clay layer and Silty Sand is below the Silt layer.  I have assigned a phi of 28 degreed for the Silty Sand layers.

Wedge Failure
I have only used a wedge analysis when I have geometry that would dictate that the failure surface is a wedge failure, usually on side hill fills, etc.  How do you select the failure surface?  I would seem arbitrary for an embankment fill.

Thanks,

First, since your "soft" layers are silty, are you sure that you have an accurate measure of their undrained strength?  How was this strength determined?

Second, if the "soft" layers are thin (how thin?) and the surrounding soils are sandy; drainage should occur very quickly.  Have you estimated the time for these layers to consolidate?  Can the embankment be built quickly enough to generate significant pore pressures?

Lastly, with respect to wedge failures and "picking" the failure surface.  You should run many many trial failure surfaces.  The search routines in UTEXAS are very good for circular failures, but not so good for wedge failures.  What this means is that you need to select several different starting wedge locations and run the analysis.  Then take the critical results from those runs and use them as input starting wedges for even more runs.  During this process you will need to adjust the resolution input parameter for the wedge analysis, I think it calls it the starting shift or some such thing.

If you have not done this type of analysis, you really should get someone to check you results.

With the dessicated crustal clayey silt and underlying clayey silt being only 5 ft in total depth, assuming that you have compentent material underneath, I would simply remove it and replace.  Much more positive a solution.

I will perform the wedge analysis with many trials.

Removing the 5 ft of soft material will be an option that will be provided.  I would like to have an accurate determination of the "actual" factor of safety so the cost of overexcavation is justified.

Thanks to all for the responses.

A.  The Ladd paper is probably the best available reference on the topic.

B.  Big H is quite likely correct about removing the foundation material.  However, if you can tolerate consolidation settlements and sequence the fill operations properly (so you don't get bearing failures within the fill area), you may be able to leave some of the material in place by excavating a trench around the perimeter and backfilling it with good stout compacted fill.  Maybe 15 m wide at the invert, and put it up under the fill slope a way so it has some overburden to give it confining stress.

C.  Gosh, GeoPaveTraffic, I always liked the search routines in the UTEXASn family.  

I now have a copy of the Ladd paper.  Very long paper.

dgillette:

I like the select overexcavation idea.  The embankment is quite large.  This would save a lot of excavation.  Thanks.

I have never felt confident in the UTEXAS search program either.

Thanks again.

Make sure you do not apply the strength increase due to increased overburden (Su/P ratio) incorrectly.  The stress increase varies across the embankment, from the greatest under the center of the embankment to effectively zero at the toe.  One big lesson learned from the levee embankment failures during Hurricane Katrina is that the increased strength was assumed to occur along the entire clay layer EVEN OUTSIDE THE EMBANKMENT.  

I also cringe at the thought of building 30-ft in one fell swoop - BUT it won't be built instantaneously.  NACFAC DM7-1 has guidance on estimating consolidation for gradual load applications.  With silty, rather than clayey soils, you may get the consolidation you need, but your Su/P ratios are typically much lower (0.1 - 0.15) than for clays.  You may not get the strength you need anyway, without ground improvement.

I agree that excavate and replace may be the way to go and I have used the method of targeted over excavation to create a "key" of good soil through the soft subgrade.

See Poulos and Davis for confirmation of rochplayer's comments on stresses beneath the embankment. (Elastic Solutions to Soil and Rock Mechnanics)  I have used selective removal - i.e., providing toe keys or specified widths before - but make sure that you can live with some waviness in the pavement structure due to leaving in portions of the fill beneath. Can fix by holding off on the wearing course for a year or so. If you are building such a large embankment, the cost of the removal of 5 ft for the whole width would be peanuts comparatively - and, it is nice to sleep at night.

Are these soft soils saturated?  Have you done a consolidation test on the soil layers?  What are your Cv values?  No doubt that with the thick fill, you will get on the virgin consolidation curve, the question is how long will it take to dissipate the excess pore pressure. When the pore pressures, I'd agree that the Su/P ratio of 0.2 (or somewhat better) would be appropriate.

Considering that these are thin layers (with respect to the aspect of the fill), rotational failure is not likely the critical concern - wedge failure would be.  Have you done any direct shear testing?

I'd think some of these quesitons would be important for this group to provide you with any meaningful feedback.  To answer your question, however, I'd use 0.2 (you can also refer to Jamilkowski's (sp) research on this topic.

Good luck. . . .

f-d

¡papá gordo ain’t no madre flaca!

If you could drill at the site, the best option is vane shear tests.

If the project is worth it consider doing some CPTs,

good luck!

geohec:

In January, my firm decided that our field vane shear equipment was not being utilized when it should/could be.  It was decided to try to use this testing method more often, when applicable.  Had the equipment by just did not use it.

It is a test we can do ourselves and will most likely will be one of the ways we will verify when stage 2 construction can begin.  Nothing like measuring instead of estimating.

Thanks to all.

Can't believe that you are leaving 5 ft of clayey material in place under a 50 to 85 ft high embankment.  Removal of it - in total - and a large toe key seems to be more reasonable; but . . . it is your baby, not ours.