Consolidation settlement of stiff, unsaturated calcareous clays 1

[sgsibob]

I read the old thread on the expected consolidation of clays compacted wet vs dry of optimum and found it interesting but not determinative of a situation we have encountered.

How would the expected consolidation parameters differ for an undersaturated stiff clay versus a saturated stiff clay, all other things being equal? The clays I am thinking about are calcareous old lake bed sediments, lean to fat, with saturations ranging between 80% and 95% (above a deep water table) and 100% (below the water table). Dry densities in the 80-90 pcf range are not uncommon. Despite the low densities, these clays are stiff to hard in terms of drilling behavior (in fact, they contain carbonate nodules at places)and SPT values are not uncommonly 25-50 blows/ft.

The OCR ranges up to 2, but we still see consolidation behavior in tests.

Our approach to testing has been to run conventional 1-D consolidation tests, saturating the samples (well, at least adding water...) at normal stresses ranging between 1000 and 4000 psf to try to see what is happening in the elastic and unsaturated regions. We typically get 2% to over 8% instantaneous deformation upon adding water. The shape of the load vs deformation curves above and below the saturation point seem identical so our assumption has been that the expulsion of pore fluid is what is important in terms of consolidation magnitude, and it makes little difference in terms of magnitude if what is being expelled is water or air. Differences in air vs water permeability might affect time-rate calculations, although even then, the overall time to an advanced state of consolidation should be the same because one has to get rid of the water sooner or later. From this logic, we have obtained consolidation parameters and used these in consolidation models conventionally.

I am seeking comment on whether the use of conventional consolidation curve analysis and conventional settlement calculations is appropriate in this case.

By the way, we don't know to what degree our consolidation coefficients are being affected by sampling disturbance. We suspect some, but perhaps not all that much. Because the material is so stiff and nodular we have not attempted to drive Shelby tubes; we use a 3-inch drive sampler. This probably introduces some disturbance, although the elastic tests we have run don't indicate much cracking or fissuring. We decided there is no practical way to quantify the effects of disturbance and have pretty much presumed that the load-deformation behavior we see in our tests is reflective of the in situ material, or at least that the assumption will be conservative as to actual settlement magnitudes.

We recognize that the undrained shear strength of these clays will reduce settlement when loaded by embankment, asserting that at least initially they would consolidate fully but for the rigidity of a calcareous soil structure. However, we know from our testing those bonds weaken or are lost entirely when the material is wetted, which we have to assume will eventually occur below new embankment to one degree or another.

 

[TDAA]

You seem to be describing hydro-compactive soils that are compressing under load, when wetting. This is a much different issue than consolidation over time. Most of the consolidation over time is due to the slow dissipation of pore water pressure.

 

[sgsibob]

Yes, that may be part of the behavior, but the rate of consolidation (when it does occur) and the consolidation magnitudes are similar to that of ordinary pore-water dissipation. Comparing the consolidation signatures of the clays from below and above the water table demonstrates this.

So my question does not relate to collapsing clays that are dry, they just aren't fully saturated. The movement that we see upon wetting is sudden, but is a small proportion (like 5%) of the total deflection that results from the consolidation test.

Whether it is hydrocompactive behavior or consolidation, the result in terms of long-term settlement after loading is the same, is it not?

 

[fattdad]

the only thing I'll add is that consolidation of 90 the percent saturation sample (i.e., as you describe above the water table) may still be driven by the permeability of the soil. There is a point where the air is isolated within the sample that for volume change to occur requires pore water flow. I would not be surprised if the Mv values for both samples are similar.

On a separate note: Just because you are below the water table is no guarantee that there is 100 percent saturation. In many clays, the isolation of air within the soil matrix requires sufficient water pressure (i.e., depth below the water table) to attain 100 percent saturation.

On the matter of collapsing, I really don't know. Interesting, however. . .

f-d

¡papá gordo ain’t no madre flaca!

 

[aeoliantexan]

The behavior you describe sounds similar to that of low to moderate density unsaturated loess. I don't know why lake deposits should be collapse-susceptible, but I guess it is possible.

Experiments on loess by Milovic of Yugoslavia show different consolidation curves for apparently identical samples with different moisture contents. His results can be replotted to create a family of e-log p curves with apparent preconsolidation pressures that increase with decreasing water content. Saturation from any lower water content at any pressure causes collapse to the lowest curve. More interestingly, a moisture content increase less than saturation can cause partial collapse from one curve to a lower one. After the moisture increase and collapse occcur, additional pressure will cause consolidation along the curve for that water content.

It may be excessively conservative to assume the soil will eventually become saturated under the embankment, but even a modest increase in moisture content may cause additional settlement. We see this occurring in both highways and subdivisions where filled areas begin to settle several years after construction. This delayed settlement causes dips in the pavement of highways and damaging differential settlement of buildings, expecially those built on hillsides and supported partly in cut and partly on fill.

Preloading can reduce these delayed settlements if the water contents are high enough -- say several percent above the plastic limit. If the loess is too dry, nothing happens under the surcharge, but delayed settlement may still occur.

Still referring to loess, in my experience, primary consolidation settlement of unsaturated soil is nearly instantaneous, thanks to the compressibility of the air, but secondary compression may be quite significant. Saturated loess behaves more like conventional clay, but consolidation may take place relatively quickly due to the high vertical permeability cause by tiny root holes.

Pushing or driving samplers in unsaturated soil can cause a great deal of compaction to higher densities, leading to higher measured strength and apparent preconsolidation pressure. I favor 4.5-inch or larger shelby tubes pushed twice - first about 4 inches to cut out the compacted zone left by the auger, then about 12 inches to get the sample. If you can't push the sample out with your hand, it probably is compacted. Hand-carved samples from pits are best.

Hope this is helpful

 

[sgsibob]

That's the kind of response that I was looking for.

Given your experience, I think we will give Shelby tubes a try for the next round of drilling, but keep the ring sampler on standby.

You mentioned that preconsolidation may not completely happen. I sincerely hope that will not be the case, but there will probably be some effects of the cementation in situ under the preloading that may take a long time to break down as moisture accumulates.

 

[Doc08]

Are you sure you can push a shelby in a soil with SPT 25-50 without damaging the tube and potentially the sample as well ? You might want to consider something a bit more suitable for such a soil, (if memory serves me right) perhaps something like the Denison or the GUS samplers would be more appropriate. Such samplers may be equiped with a clay bit on the outside and collect the sample in a plastic tube (much like a double core barrel with an inner tube) inside with fairly little disturbance.

 

[aeoliantexan]

The Denison, or Acker-Denison sampler may be a good idea. I think the Gus sampler is for soft soils. Other options might include the Pitcher sampler or one of the continuous sampling systems used with hollow-stem augers and built by CME or Mobile.

sgsibob, can you tell us roughly where the project is located?