Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Register Log in
  • Congratulations IDS on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Clay sensitivity - strain to reach remolded strength?

Status
Not open for further replies.
dgillette

dgillette

Geotechnical
May 5, 2005
1,027
Got an interesting problem, for which I have not found a clear answer.

Have to analyze seismic deformation of an embankment on a clay foundation. The clay is somewhat sensitive (mostly 3 to 4 by VST and CPT sleeve). Unfortunately, the embankment happens to be located near a subduction zone capable of producing fairly high PGA going on and on and on for almost 3 minutes (M~8.5-9). [shocked]

I am trying to figure out how much strain/deformation is required for the clay to be remolded to the point that its strength is much less than its peak strength. With more typical earthquakes, I might not get real concerned, but this one has many more cycles of load in it. The amount of loss of strength affects whether the embankment is just deformed (a la Newmark) or whether it becomes unstable and slides away under gravity loads even after shaking ends.

In the Feb '98 ASCE JGGE, Stark and Contreras analyzed a slide in Anchorage, 1964, in addition to a lot of lab ring-shear data. (The geologic origin of the clay there may be different from my site, and the stresses are very different because mine has a big embankment built on it.) They conclude "...the slide blocks that moved less than 0.15 m mobilized at least 80% of the undrained peak shear strength. Slide blocks that moved between 0.15 and 2.5 m mobilized an undrained shear strength between the peak and residual shear strengths. Slide blocks that moved more than 2.5 m mobilized the undrained residual strength." Analyzing the same slide, Idriss (1985) had fairly similar conclusions about displacement vs remolding. Stark and Contreras' ring-shear data are helpful, but not in a quantitative way because the shearing is concentrated in a very thin layer (a few mm). In a real slide, it would probably be concentrated similarly, but in a thickness of cm or 10s of cm, not just a few mm, so the two are not directly comparable.

What else has been published? Doesn't seem to be a whole lot out there. Talked to Tim Stark (U of Illinois) about it a couple years ago, and he wasn't aware of a whole lot more in the literature.

Thanks,
DRG
 
Replies continue below

Recommended for you

Sort by date Sort by votes
I take advantage of your questions to see if I learn something...




















also nice...

















Hope some of them are of help.
 
Upvote 0 Downvote
Thanks for looking all of those up! It's quite a list.

Looks like you got about the same result I did, i.e., that there just isn't a whole lot published that directly relates to the question.

The two of greatest interest are Quinn et al (Canadian landslides) and Pekcan et al (Adapazari clays and silts), but the latter deals with relatively small strains (5% shear strain, not remolding). Many of the others (like Boulanger and Idriss) are for near-level-ground liquefaction of low-PI clays (and other materials), which is a different phenomenon.* The ones on ring-shear are good for showing what could happen to the strength at larger strains, but they don't really say much about the amount of displacement my foundation would undergo before the strength is seriously reduced (because of the different sheared thickness, mentioned in the original post).

*One might legitimately question whether there is a real difference between a liqufied soil and a remolded somewhat-sensitive clay, but the strains to cause low strength could be very different, particularly if the strain in the latter is primarily monotonic and not primarily cyclic.

DRG
 
Upvote 0 Downvote
At least add a link from wikipedia



Other comments:

I am no match of any geotech engineer and less any with experience in some field. However I will adventure opinion on something seen here and there...

On your quote of Stark and Contreras... Obviously blocks that have moved 2.5 m have mobilized the strength. .14 m is another matter, even our buildings allow in akin scale such displacements at some inelasticity.

On the nature of these soils: we are always talking of porous materials with some fine bridging structure within. Unfortunately for the more highly porous materials the levity of the bridges means that the material will behave mostly in some fragile behaviour, this meaning there won't be much difference in static and dynamic strains for any given attainable stress level since inelasticity won't be present. Well greased clays would be another matter, and maybe, relative to their final strength, the softer non sensitive clays would tolerate bigger displacements, in the same manner that ordinary structural steel allows for more inelasticity than HS steels. A stiff clay will be harder and more fragile and more closer to slate behaviour that uniform softer sound clay.

On figure 3.44 of the text in the first link in this post: The tests show more or less linear relationships for both static and dynamic tests, corrected and not for friction. When friction is discounted, dynamic remoulded shear strength halves that observed static. Since this is consistent with what said when treating of elastic materials, from an energy viewpoint the ground being tested is both homogeneous and not much distinguishable from one elastic behaviour. Hence since what you want is safety against the dynamic effects, and judging from the set of tests in 3.44 the first measure is to halve any allowable shear strength derived trough a proper safety factor from tests where the forces are applied slowly and of course not dynamically.
 
Upvote 0 Downvote
Again, thanks for the link, but I think there may be some misunderstanding in your reading of Figure 3.44 (and I am no closer to a quantitative answer).

First of all, 0.14 m is a lot of displacement on a thin sheared layer. If the layer is 0.1 m thick, the shear strain is 140 percent. If there is a distinct rupture surface, the strain is effectively infinite. In the ring shear tests, S&C found shearing resistance reduced by half at 0.02-0.03 m of displacement on a very thin layer. (Strain cannot be measured in that test.) You are correct that more sensitive clays reach post-peak or remolded conditions at smaller strains. That is well established from various types of testing. However, it is quite difficult to translate lab tests with measurable strains and boundary conditions to an uncontrolled sheared thickness and predicted strain in the field, because the clay in the landslide shears where it wants to and not where we force it to with a test apparatus.

Second, the results of the static and rotary soundings are actually fairly similar - note different scales in 3.44a and 3.44b. For example, with a remolded strength of 5 kPa, both show rates of increase in crowd force of about 5 kN/m, for each additional meter of rod. Apparently, the larger rod diameter in the rotary test (36 vs 25) is offset by greater remolding in the rotary test and "setup" during stops to add rods in the static test. "...The measured value of the rod friction when resuming sounding may be highly erroneous and the stroke of the slip coupling is far too small to enable these effects to be eliminated."

For CPT rod friction, the slope would be about 8 kN/m. Same rod diameter as the rotational test, but the remolding is not as complete (because it's a straight-sided rod), so higher shearing resistance is measured. (On this project, however, our VST measurements after 1800 degrees of rotation and CPT sleeve resistance were in reasonably good agreement.)

"...the first measure is to halve any allowable shear strength derived trough a proper safety factor from tests where the forces are applied slowly and of course not dynamically..." Actually, no; I don't agree with this. There are some significant increases in shearing resistance with higher strain rates. See, for example, Mitchell's "Fundamentals of Soil Behavior." It is not common to take advantage of that effect however (although Mitchell suggested that I consider it on a somewhat similar project with less-sensitive clay and shorter-duration earthquakes).

DRG
 
Upvote 0 Downvote
I for sure must abide and abide with everything you say, perhaps with following the observations...

that of course .14 m on a proportionally thin layer is infinite, and precisely when referring of the same order of magnitude or scale I was thinking on a layer of say 18 m of sheared soil and a 18 m tall building; then my numbers make more sense, and not knowing the thicknesses without the paper I only was pointing to that.

On the misreading you surely are entirely right since I only gave it I quick look and that is what happens when one is more willy than thoughtful. Thanks for your clear explanation, and will read with time the whole text to learn even more.

And on the gain on strength with strain, surely there must be specially for the clays with lower water content, this meaning, again, for those that have the gain at some tolerable strains. I practice almost 100% in building projects and certainly freeing some water starting from 80% content in water wouldn't result in tolerable movements; so I would be urged to not use most of the nonlinear arm (meaning at service level) -again, that not based in the erroneous deduction I made when looking at the 2.44 charts, just the nonlinear stress-strain charts. There can be many other situations (mining mounds, highways on remote areas) where some less controlled behavior is reasonably tolerable and why not, those promoting the project and controlling their technical aspects may give the go ahead.

Again, thanks for your opportunity to learn something, that I have.
 
Upvote 0 Downvote
One big difference between my situation and what you are accustomed to is that, with a large earthquake, an embankment WILL deform. Typically, the yield acceleration for a dam embankment on a good foundation of rock or dense soil is around 0.2 g; we commonly deal with peak ground accelerations 3 or 4 times that. On a clay foundation, the yield acceleration can be a lot less than 0.2. We don't even try to prevent yield and deformation. We simply need to keep it to some tolerable limit, and make sure the consequences of deformation are accounted for. If the embankment in question is a dam, we need to be sure that there isn't potential for transverse cracks to allow a breach to be eroded. (If it's a highway embankment, you just repair it afterwards.)

That Swedish report is interesting reading. Their methods are not common in the US, or at least I've never seen them used.
 
Upvote 0 Downvote
Thanks again for the info, and I see the importance of knowing to what strains at real scale some particular clay doesn't form cracks. A thing a bit related to that that we saw in one of the geotechnical forums was the rate of movement as related to the expected standing risk of failure, but no reference to the base nor materials (that I remember) was quoted.
 
Upvote 0 Downvote
Status
Not open for further replies.

Similar threads

Eng_Girl
  • Locked
  • Question
UCS to Shear Strength
Replies
4
Views
357
geotechguy1
geotechguy1
ABTSGeoScience
  • Locked
  • Question
Rebound Associated with Swimming Pool Design in Over-Consolidated Fat Clay (CH)
Replies
1
Views
140
GeoEnvGuy
GeoEnvGuy
JHeisenberg
  • Locked
  • Question
Drained vs. Undrained: Reduce the Load on Clay from Existing Condition
Replies
2
Views
189
geotechguy1
geotechguy1
geotechguy1
  • Locked
  • Question
Applying aging factors to clays
Replies
1
Views
119
GeoEnvGuy
GeoEnvGuy
SKSA
  • Locked
  • Question
Relation Between Drained Strength of MC failure Criteria and Von Mises Stress
Replies
1
Views
157
EireChch
EireChch

Part and Inventory Search

Sponsor

Back
Top Bottom