liquefaction/cyclic softening of fine-grained soils

[AK92]

Hi,

I'm performing a liquefaction (for sands) /cyclic softening analysis (for clays) based on the Simplified method which uses CSR and CRR (based on the Summary Report from the 1996 NCEER and 1998 NCEER/NSF Workshops on Evaluation of Liquefaction Resistance of Soils and Idriss and Boulanger 2007 paper on Evaluation of Cyclic Softening in Silts and Clays).

My site is in Jakarta and we have lots of high plasticity silty clays/clayey silts with PI generally > 30 and LL > 60, up to around 15-20m before we hit hard material. Generally, the clays are quite bad, with many reported SPT 'N' blow counts less than 4.

Analysis based on the simplified method showed that many of the very soft clays will undergo significant cyclic softening with very low factors of safety below 0.5. However, some of the literature, for e.g. the Chinese criteria, Andrews and Martin 2000, Seed et al. 2003 excludes medium/high plasticity fine-grained soils from any "liquefaction" potential. I'm a bit confused by the terminology here as I'm not too sure if the "liquefaction" in the criteria includes cyclic softening or not. I'm assuming that these medium/high plasticity fine-grained soils would not undergo liquefaction but would still undergo cyclic softening which I would have to check for.

Is that assumption correct? If that is correct, should I downgrade all my clay strengths in this layer to the remoulded values for pile design? What about lateral loads?

By the way, does anyone know where I can get empirical equations to obtain clay sensitivities based on Atterberg Limits? (I have some charts but they are a pain to use with spreadsheets)

Thanks!

 

[nbr1]

Suggest you evaluate the fine grain soils with CPTu testing; and the proposed CPT method attached which uses Ic and Bq parameters.

 

[Mccoy]

As far as I know, the decrease of shear strength in clays under cyclic undrained conditions is called degradation, whereas liquefaction is more appropriate term for sands. This distinction is substantial and not formal, since in post-liquefaction sands settlements may be very large regardless of the presence of foundation loadings.
I would rule out liquefaction phenomena but I would rule in the onset of degradation phenomena with loss of shear strength, that is lesser Su.
About pile design:
Base capacity: In your profile pile bases would lay upon the competent layer below 15-20 m so clay degradation would not be a concern. OC clays would not degrade because of dilatance and ensuing negative pore pressures.
Lateral capacity: In both drilled and driven piles an adhesion factor which degrades Su is usually applied, so how to relate this degradation to cyclic degradation of an intact volume of clay? I would have to reason about that, maybe applying the minimum value resulting from static adhesion degradation or cyclic degradation of intact soil. Just a single random idea though.

 

[AK92]

@nbr1

Unfortunately, the CPTu equipment is expensive and was not done for this project. We have some CPT results from mechanical CPT's but the strengths seem awfully high compared to the UU and SPT results and the quality seems suspect. I too would think that CPT data would be invaluable for evaluating the properties of the fine-grained soils.

@McCoy

Thanks for the input. I too think that only the cyclic degradation (or cyclic softening) would apply to those high plasticity clays/silts and liquefaction would not be possible.

Another concern that I have is the high void ratios (many of them above 2). In an earthquake scenario I would think that it would be very easy for these soils to deform a lot, causing the piles to lose contact with the surrounding soils resulting in a loss of shaft friction as well as lateral support. If this is correct, perhaps it would be prudent to simply discount the shaft contribution of the upper soft silty clays for design? However, solely relying on base capacity would probably mean that pile settlements would be much larger as it requires a lot of base movement to mobilise the base capacity. Probably that would mean installing the piles deep enough so that the skin friction from the competent material will be sufficient to support the working load, but of course this would mean much higher costs of construction. What do you think?

 

[BigH]

Have you ever read any of the papers produced in the late 60s/early 70s on cyclic porewater pressure response? There is a cut-off level where if the cyclic loading is below this line, there is no longer any additional porewater pressure buildup - can't remember exact wording off hand. Castro had one view of this - could also, I believe, google "Dwight Sangrey". Papers were published in the ASCE Journal of Soil Mechanics and Foundation Engineering (SM), later the Geotechnical Journal (GT) and later ???

 

[dik]

If these clays are in/near a marine environment, any chance they can be highly floculated? and maybe their 'reworked/distrubed' strength is markedly decreased?

Dik

 

[AK92]

@BigH

I haven't seen the papers before. If you have a link to them that would be really helpful! Would the cut-off level be easily related to peak ground acceleration and moment magnitudes of the design earthquake? However, as my site is in an area with high seismic risk (the design earthquake has a moment magnitude of 9.0), I'm not too sure if the loading will indeed fall below the cut-off levels.

@dik
The soils are alluvial soils and probably not of marine environment. By running them through a chart correlating the atterberg limits and sensitivity, I'm getting sensitivities of around 3. But I'm not too sure how accurate or applicable these correlations are.

 

[BigH]

I've googled and got this search url. It seems to have quite a bit of potentially good materials for you.

https://www.google.com/search?q=cas...la:en-US:official&client=firefox-a&channel=sb

 

[Mccoy]

AK, a magnitude 9 design earthquake is a real huge monster, cynematic effects, to be added to inertial effects, may be significant.
I used to use the following calcsheet to have a quick check on cyclic strenght loss but I ceased to do that as soon as I realized that, in my place, strenght reduction is negligible. the numbers in the calc sheet may be similar to the properties of your clays. Even by adopting an extreme 0.9 pore pressure ratio the decrease is significant but not overwhelming. Very plastic clays, with PI in excess of 100, may need a 0.6 overpressure to cause significant strenght reduction. In the Erken-Ulker article for example, following BigH links, many plots suggest that a lesser pore pressure ratio might apply in your case. Other concerns might be post cyclic overpressures and settlements and negative friction upon a part of the pile shaft. Cu = Su, from Yasuhara et al., 1992, 1997

 

[Mccoy]

Quoting myself on a previous post:

Lateral capacity: In both drilled and driven piles an adhesion factor which degrades Su is usually applied, so how to relate this degradation to cyclic degradation of an intact volume of clay?

In your case, NC soft clays, contractive behaviour, Su remolded pretty close to Su intact, so cyclic degradation would very likely prevail upon static lack of lateral friction due to soil disturbance.

 

[AK92]

@BigH and McCoy

Thanks for all the valuable links, I'll be trying to digest as much as I can.

@McCoy
It's great for me to hear that the reduction in strength may be much smaller compared to estimations based on empirical clay sensitivities.

I have found a link to the 1992 Yasuhara paper which contains the formula that you have used at

https://www.jstage.jst.go.jp/A_PRed...&noIssue=1&kijiCd=32_1_100&screenID=AF06S010.

In the 1992 paper, the l in the formula is based on the experimental parameter Ao and Cs/Cc (recompression/compression indices) but doesn't give any indication of how to estimate Ao. Your spreadsheet seems to have it as a function of the plasticity index which would be a lot more helpful.

Would you have references to some papers which have further developments of the theory in this paper? I have found some referenced papers through online searches such as Matsui,1992; Yasuhara,1995 and Moses,2003 but I can't seem to be able to find the titles of these papers that would allow me to locate them.

Thanks for your kind help.

 

[AK92]

Edit: I found the names of the papers already, will try to look them up for more information. Thanks for the help so far.

Matsui T, Bahr MA, Abe N(1992). “Estimation of Shear Characteristics Degradation and Stress-strain Relation of Saturated Clays After Cyclic Loading,” Soils and Foundation, Vol 32, No 2, pp 161-172.

Moses GG, Rao SN, Rao PN(2003). “Undrained Strength Behavior of a Cemented Marine Clay Under Monotonic and Cyclic Loading,” Ocean Eng, Vol 30, No 14, pp 1765-1789

Yasuhara K(1995). “Post-cyclic Behaviour of Clay in Direct Shear Tests,” 10th Asian Regional Conf on SMFE, Int Academic Publishers, Vol 1, pp 119-122.

Yasuhara K, Hyde AFL(1997). “Method for Estimating Postcyclic Undrained Secant Modulus of Clays,” J Geotech and Geoenvir Eng, ASCE, Vol 123, No 3, pp 204-211.

 

[Mccoy]

AK, I found the formula in the 'Designer's guide to EN 1998', the European seismic code. The 1997 Yashuara-Hyde article should be a furhter development of the previous 1992 article, where the exponent is related to the plasticity index.
The relationship is l = 0.939-0.002 I[sub]p[/sub]

 

[AK92]

@Mccoy

Thank you for providing the references. So from what I understand so far, an earthquake would produce pore pressures which would reduce the effective stress of the clay which will thus reduce the undrained shear strength similar to unloading. After the earthquake, the pore pressures dissipate and the clay would then return to its original effective stress state, and consolidate along the recompression line. Post earthquake, this would probably bring about an increased strength due to the consolidation of the clay. So my main problem would be to quantify the degree of increase of the pore pressure during the earthquake itself and then use that to estimate the degradation of the undrained shear strength of the clay.

I understand that I can probably get away with assuming a pore pressure ratio of 0.6 or even 0.9, but I would also like to know if there are any developed methodologies to relate the pore pressure ratio to more quantifiable parameters (such as OCR and the earthquake loading magnitude(be it magnitude, PGA, CSR, etc...).

Is there anyone who had unified the work of developing cyclic degradation, cyclic pore pressure response and liquefaction/cyclic softening susceptibility? It seems that the papers I have read so far seem quite disjointed.