Log In

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips Forums!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.

Students Click Here

Settlement in Clay above water table

Settlement in Clay above water table

Settlement in Clay above water table

Calculation of settlement in clays above the water table is something I'm unfortunately still unclear on.  I am currently calculating settlement in clayey soil above the water table.  Total settlement, no matter how quickly it occurs, is of interest to me due to several of reasons.  I have consolidation data for the clay layers above the water table.  My approach has been to calculate the settlement as if it were consolidation settlement, not elastic, and then assume that settlement will occur relatively rapidly.  Is this a reasonable approach in calculating the magnitude of settlement.  

RE: Settlement in Clay above water table

Your question is quite vague: please pose a specific problem, and ask for help.  Provide all relevant geometry, soil profile, GWT data, consolidation data, soil properties (water content, PL, LL, unit weight, etc.)

RE: Settlement in Clay above water table

It is likely a feasible approach provided the compression parameters you are using were obtained using a consolidation specimen what was not inundated during testing (to model the field conditions). Using Terzaghi's 1-D approach is not exactly applicable since it was derived using a saturated sample and the time rate is derived from the model. However, if you are not particularly interested in the time rate (you know it will be reasonable fast based on your consolidation data), your compression parameters still apply since you are merely following the measured response of the specimen under anticipated loading conditions consistent with the problem you are evaluating. This is commonly done with approaches such as DMT and PMT (use of soil modulus values to calculate stress-strain responses). The key is not to inundate your sample for testing.


RE: Settlement in Clay above water table

Hmmm, I'm going to dissent strongly on this one.  Who is to say that the site won't become saturated at some point during the life of the planned structure?  Saturation / inundation will remove most, if not all, negative pore pressures.  Negative pore pressures will reduce the settlement of the test specimen during lab testing (if the sample isn't inundated), but may not always be present during the life of the structure.  Failure to inundate consolidation samples can lead to very serious underestimates of settlement.

While this may be an acceptable procedure under limited circumstances, those circumstances are relatively rare.  In general, this approach is a bad idea.

RE: Settlement in Clay above water table

I agree to the possibility pointed out by Focht3. However, I believe the impact of negative pore pressure in a consolidation sample will typically be low considering the typical way such specimens are collected, extruded, and trimmed into the ring as applied to 99% of all consulting projects. I assert most of the negative pore pressures have dissipated by the time the sample is prepared and ready for the oedometer (just as excess pore pressures are also likely to be dissipated). Focht3's theory is sound, but for my projects, I chose to follow a practical path, unless I am working on a project that such differences matter (which in my experience is less than those I can identify with unsaturated clay conditions).


RE: Settlement in Clay above water table


Well, to each his own.  Did you get this from Mike Duncan?  It doesn't sound like him... Has this procedure been published in any peer-reviewed forum?  If so, where?  Please provide the citation.

Your statement that  the impact of negative pore pressure in a consolidation sample will typically be low considering the typical way such specimens are collected, extruded, and trimmed into the ring as applied to 99% of all consulting projects  may be true where the soils are more or less normally consolidated, but it does not apply to most of the continental U.S. - or the world, for that matter.  (Offshore conditions are a special case - where inundation must be done to imitate the field condition.)  Most clays soils are not saturated and normally consolidated: they are overconsolidated, and typically "dry".  They gain significant strength from the negative pore pressures within the sample, and will not appreciably consolidate unless they are inundated.  Your procedure won't work for these soils.  But don't take my word for it - ask Del Fredlund, Bob Lytton, Mike Duncan...


Be advised that AZdinak's advice is fraught with potential hazards beyond concerns about negative pore pressures.  The effects of soluble salt cementation loss and collapsible soils will not be observed using a "no water" approach.  And sample dessication is likely in most lab environments.  How do you measure and/or interpret that?  The list of potential problems this approach could overlook - or cause - is significant.

His test procedure violates conventional testing procedures - so you will be "bare bottomed" if you have a problem.  And what is gained?  In the big scheme of things, NOTHING.  The loading phase of the test will not run any faster - air and water are being expelled from the sample, not injected!  And the unload cycle results will be meaningless without it.  How will those be interpreted?  Certainly not using Terzaghi's consolidation theory.

Stick to the basic mechanism of the consolidation test.  Vary the sample preparation if you need to address the presence of shell or rock, or to deal with silt pockets and partings that make traditional trimming techniques virtually impossible to use.  Vary the initial load, applied loads, etc. to account for the specific problem you are dealing with.  Delay inundation until the sample is in the test frame and the dial gage is in place if you are interested in the soil's swell potential.  But follow the basic test procedure.

RE: Settlement in Clay above water table

The basic soil procedure is ASTM D2435, which actually does mention our topics of discussion (Sections 5.5 and 11.3). If you are operating an AASHTO accredited or COE certified lab, this is also your standard for 1-D consolidation.


RE: Settlement in Clay above water table

Dear AZdinak
As far as I've understood by your comments, in most of the projects you are dealing with- when being encountered with a non-saturated clay layer- is that you simulate the natural condition for ordering the oedometer test (i.e. no inundation).
The topic of how to deal with unsaturated soils is always a matter of discussion among soil experts.
I appreciate it if you elaborate little bit more your own reasons for not considering the possibility of saturation or inundation of the layer in most cases. (please clear if I have not got your comments well enough).

On the other hand, although after inundation the P'o contributes to the settlement, but however at the same time OCR is enhanced and according to my experience for pads with small breadths (shallow depth of pressure bulb) it is almost the Cs which is dominant and hence, not very high amount of long-term settlement is anticipated.

RE: Settlement in Clay above water table

The question has led to some interesting comments from respondents. The beauty of geotechnical engineering unfolds. I would say that everyone is correct in some way based on his or her experience. Have you ever noticed how diverse we are in predicting settlements etc?. An answer to the difference in values that people present is often attributed to the fact that we are only experts in our turf of practice. Even then, we have shortcomings. There is also the question of the field settlement versus the laboratory settlement values. It is often said that the field is larger than that obtained using lab results. I have seen the opposite as well.

For the question at hand, I would first try to understand the loading and future conditions and to worry about differential settlement if it is a framed structure.

Obtaining settlement values for an isolated foundation is sometimes misleading. The game is one for the foundation engineer. This breed is lacking these days. It is a hybrid of geotech and structural. Most of what I have seen in practice except for companies that have both diciplines in house is the presentation of a geotech report with recommendations which is presented to the structural engineer to do his thing.

We need to merge both for better understanding all around. The subject will then become more challenging and both parties would appreciate the issues.

The above strays somewhat but I once before said in another thread that I have reservations in providing numbers unless the problem is seen from all angles.

Whether you would like to innundate a sample or not depends on the experience of behaviour of foundations in the locale in question. For those who have tracked practice in various parts of the world, for pavements for example,some prefer to test samples at the equilibrium moisture content which is given empirically as 1.3 time the plastic limit I think-British practice.

I have not seen this applied too much in North America. Similary for CBR values, some soak 4 days to simulate the conditions if the area is the subject of severe rainfall. However, today we determine a resient modulus from CBR values without ever thinking most of the times about the site conditions in time. It is easy to use the empirically derived relationship and not do a test. There are serious problems here if the soil is not understood. Anyhow I have met some that do not think the subgrade is important in the whole process. I one asked a question to a pavement designer about the influence of thesubgrade on his design thickness and was told that he works from the premise that the subgrade is compacted according to specifications. Hence, there is no need to recheck the nature of the subgrade after the grade is built despite if the pavement is to be constructed sometime later. You know what, it works as well in the majority of cases. Some jurisdictions make evaluation of the grade mandatory before the pavement is designed.

I am always amazed at how much testing is undertaken for IDB, World Bank etc projects overseas and how many experts are required whan thie asame is not done where the funding originates. Loan money to so called third world countries is another issue that boggles the mind but it makes good work for some that otherwise would not have been given the opportunity to be a marerials engineer or geotechnical engineers for years which would not be the norm in the first world.  
To each is own, providing that you clearly undestand what you are doing. Anyone taking information verbatim runs the risk of accepting someones experience which may not necessarily fit his or her situation. The responses are therefore in my opinion to get the enquirer some information so that he can decide what is best for himself. In many respects this is all one can do as no one knows for sure all the conditions that the enquirer, hopefully understands.

Perhaps, we need to express our thoughts in this fashion rather than debating about who may be right or wrong. We can only do so if we have a universal doctrine. Even Terzaghi figured that what he developed was not to be used as gospel truth.          

RE: Settlement in Clay above water table

closest I could get to someone giving good "snaps" - or claps!

RE: Settlement in Clay above water table

Good post, VAD.  I especially liked the last paragraph.  

You made a good point: there is no known universal doctrine available in soil mechanics for soil behavior under every situation and because of this, it is impossible to have a single "cure all" solution.  In a related point, we have been unable to discover a universal link between classical newtonian physics and quantum physics and because of this, each of those types of problems are treated with a different set of methods...yet there are probably those that still believe there is a universal soil mechanics doctrine...

I think the point of this forum is to, when a question is asked, present our experiences dealing with a related or similar problem.  The question "asker" can then draw their own conclusions based on their experience and the kind, helpful, non-belittling advice of their peers.  

RE: Settlement in Clay above water table

I apologize if my words were interpreted as aggressive or diffinitive (although there was little ambiguity from the aggressive side). As was pointed out by the last set of posts, we can offer our advice, but by no means should it ever be considered "the answer." I have and continue to enjoy the intricate nature of geotechnical engineering. One of the things I have learned is, while there are standard approaches to certain problems, there are no standard problems. And as VAD pointed out, the "beauty of geotechnical engineering unfolds."

I value the opinions of the professionals that post in these forums. Some may express ideas that I agree with and others may not. Are mine always correct? Certainly not (although at times I might like to think so), and I will continue to learn through my experience and the experience of my peers.

As I have done in the past, I will continue to provide my support of questions asked by forum members. I look forward to future discussions.


RE: Settlement in Clay above water table


(To leave no stone unturned)...Most of my experience is indeed with saturated clays, or those that may potentially undergo saturation over the life of a design. I have, however, run tests at the natural moisture content of the soil (at least what we thought the moiture content was) for areas that were dessicated and likely never to be innundated. I have also run consolidation tests on compacted structural fill (which were designed never to be innundated). However, each of these instances was very specifically influenced by the project design and controlling conditions. As you point out, if you do one thing, something happens, if you do the other, something else happens.

In my mind, as long as the problem is clearly defined and you are attempting to produce data to support a specific set of engineering conclusions, your experience is your guide. I think the discussion generated from what may have been a seemingly "uncomplicated" post, exemplifies just how much the jury is out on this matter. I am confident to say, just as there are instances that innundation is the best approach, there are also cases that consolidation tests at the natural moisture are appropriate.

I have also changed my stance on the "negative pore pressure dissipation due to sample collection, extrusion, and trimming." Based on some of the new knowledge I have aquired from this discussion, I now believe that negative pore pressures may still be present in the specimen if run in an non-innundated state, however, I also believe that innudating the sample does not automatically relieve the negative pore pressures (may make free water available to fill pores to eventually saturate the specimen, but that will not occur, based on my knowledge, until the applied load causes the pore pressures to become positive). This is not to say that near surface negative pore pressures are not relieved by innundation (I can believe this as well).


RE: Settlement in Clay above water table

This has been a good discussion thread.  I always enjoy the comments - even though I don't always agree!  To AZdinak: I did not misunderstand the intent of your comments; and I hope that you didn't misunderstand mine!  I would hope that we all feel that we are "right" when we post an answer or comment to someone's question or comment - if we're not confident in our position, why bother?!  I particularly agreed with your first April 10, 2003 post.  It goes for me, too.

I am reminded of a quote from a letter written by Ralph Peck to Karl Terzaghi - you can find it on page 55 of "Judgement in Geotechnical Engineering" by Dunnicliff and Deere

...It is a pleasure to argue with you, because the argument is always instructive and never personal...

I hope that none of you find "arguing" with me as personal, and pray thay it is, at least occasionally, instructive.  For all of us.

RE: Settlement in Clay above water table

Very good discussion!  We are all prisoners of our own experiance and unless we are exposed to new experiances we do not progress.  That is the great thing about these forums, from the posted questions we get a chance to hear, consider and learn from others experiances.  

RE: Settlement in Clay above water table

I think a 1-D compression test is an applicable method for estimating settlement in unsaturated clays.  Unless local experience is already available, I would perform the test both unsaturated and with inundation to evaluate possible effects of inundation.  It should be kept in mind that inundation may not result in saturation.  Free access to a water supply and high to very high back pressures are often needed in a triaxial test to saturate and unsaturated clay sample.

I am not sure that the question has been answered about how settlements in unsaturated clays should be estimated or calculated.  Two references that I know are the Student Manual by Mike Duncan for estimating settlements and the Corps of Engineers manual for settlement analysis.  I think Duncan advocates an elastic approach using relationships between undrained shear strength and modulus of elasticity.  The COE methods are also based on theoretical elasticity relationships.

If you are going to use an elastic approach, where does the modulus of elasticity come from?  I believe a 1-D compression test, is a reasonable method to estimate the modulus of elasticity.  You can directly measure the 1-D modulus at different stress levels and then estimate the modulus of elasticity for different values of Poisson's ratio.

There are published correlations of overconsolidation ratio versus plasticity index for estimating the undrained modulus of elasticity.  You have to be careful in using such relationships because it is often not stated as to whether the relationship is for saturated or unsaturated samples.  Because the modulus values that you obtain are generally quite high and because the relationships use an undrained shear strength, I have always assumed that they were developed for saturated samples.  I would not use these relationships without knowing this condition because the undrained modulus of a saturated clay would be expected to be very high because the clay particles and the water are not compressible and there is no drainage.

Whenever I use a 1-D compression test to estimate the modulus of elasticity of an unsaturated clay I usually obtain much lower modulus of elasticity values than the published correlations discussed above.

In summary, I believe an elasticity based approach is most applicable to unsaturated clays provided you have a reasonable estimate of the modulus of elasticity.

RE: Settlement in Clay above water table

For general clarification, I believe geofred's posting relates primarily to elastic settlement.  The consolidation test results will allow the engineer to calculate consolidation settlement using some type of pressure distribution scheme (Boussinesq, Westergaard, 2V:1H distribution, etc.) to arrive at the incremental pressure increase.  

Elastic settlement is somewhat harder to analyze since most geotechnical engineers don't have any basis for evaluating how "good" an estimate of elastic settlement really is.  I'm with Mike Duncan on this one; use OCR, Su and an "appropriate" correlation to develop an elastic settlement estimate.

geofred opined, [and I commented] If you are going to use an elastic approach, where does the modulus of elasticity come from?  I believe a 1-D compression test, is a reasonable method to estimate the modulus of elasticity.  You can directly measure the 1-D modulus at different stress levels and then estimate the modulus of elasticity for different values of Poisson's ratio.

You can do this, but the estimate is pretty crude.  You can get a better estimate other ways.  And do you use the initial sample compression to estimate soil stiffness?  Settlement to T10?  T50?  T100?  Total compression in 24 hours?  Incremental or total settlement?  How do you evaluate Poisson's ratio? {Note: Greek letters and some scientific notation coming to Eng-Tips...}  Appropriate strain level for your estimate of E or G?

There are published correlations of overconsolidation ratio versus plasticity index for estimating the undrained modulus of elasticity.  You have to be careful in using such relationships because it is often not stated as to whether the relationship is for saturated or unsaturated samples.  Because the modulus values that you obtain are generally quite high and because the relationships use an undrained shear strength, I have always assumed that they were developed for saturated samples.

I think it reasonable to assume saturation; but it isn't necessary.  Most of the correlations use total, not effective, stress.

I would not use these relationships without knowing this condition because the undrained modulus of a saturated clay would be expected to be very high because the clay particles and the water are not compressible and there is no drainage.

Most of the settlement comes from deformation of the mass, not compression of the water or the soil particles.  Fear of using an unfamiliar technique is quite common.  (I'm guilty of it, too.)  But that doesn't make the technique wrong or unreliable.  Just unfamiliar.

RE: Settlement in Clay above water table

Come on Focht3, the original question was how to calculate settlements in unsaturated clay.  Why would someone want to use saturated undrained soil properties for an unsaturated clay?  As I mentioned before, saturated undrained modulus values should be higher upper bound values, i.e., less compressible.

Think about the phases, soil particles, water, and a fair amount of air voids.  Such a mass should respond to load quickly and not involve "consolidation settlements" from squeezing water out of the voids.

I don't understand the questions about T10, T50, etc when we are not addressing consolidation.  The question remains how do you calculate settlements of an unsaturated clay?  It's a really good question and a condition fairly common in many areas.  I still think an elasticity based approach with a best estimate of the modulus of elasticity is a good method.

You also asked about where does Poisson's ratio come from.  I would answer good old engineering judgment, ranging from 0.5 for undrained saturated conditions to 0.0 for a popcorn or plastic foam type material.  For many unsaturated soils, it's somewhere in between, say 0.25 to 0.35, closer to popcorn then less than 0.25, closer to saturated and undrained than greater than 0.35.

RE: Settlement in Clay above water table


Sometimes a message thread just doesn't work as well as a normal conversation - no opportunity to "dynamically challenge" (i.e. interrupt), or shake your head "no" to indicate that someone has misunderstood you.  Let me see if I can break things apart to help reduce the confusion.  Let's talk specifically about settlement due to 1) primary consolidation, and 2) "elastic behavior."  We'll leave the subject of secondary consolidation for another thread.  I suggest we discuss consolidation settlement first since that should be quite straightforward.

Consolidation Settlement
I don't think anyone disagrees that you can use the results of one dimensional oedometer tests to evaluate consolidation settlement, regardless of whether the sample is fully or partially saturated, right?  While the issue of inundation may be disputed, I think that dead horse has been flogged enough.

"Elastic" Settlement
This seems to be the big sticking point.  geofred seems to be aggravated by my previous question, specifically how he proposes to convert consolidation data to elastic parameters.  I don't see the process, so I'd like clarification on this point.  And his previous post suggested - to me, anyway - that he somehow used the consolidation data to pick his Poisson's ratio.  Apparently he did not mean to imply that; so that point has been clarified.

I have a deep interest regarding building settlement, particularly the immediate or "elastic" settlement of structures.  I've co-authored a published a paper on the subject of correlating immediate building settlement to undrained shear strength via a Gibson soil modulus approach.  (It was presented at the 1982 ASCE meeting in New Orleans.)  So I'm not a novice on this subject.  And I know how to "guesstimate" Poisson's ratio - even estimate it from field performance data.  And I certainly have "baggage" of my own: I strongly prefer a kind of elastic approach using shear strength.  But I'm willing to entertain other approaches.

Please provide specifics - thanks!

RE: Settlement in Clay above water table

Interesting, interesting, interesting...We are indeed getting a lot of good discussion out of this one. Just thought I'd throw out some more fodder.

Many individuals subscribe to settlement estimates using DMT data (an undrained modulus value in low permeability soils and a drained modulus in high permeability soils). I am a skeptical subscriber; I have seen DMT estimated settlements a long ways away from what was measured during construction (come to think of it, I have seen conventional approach settlement estimates a long ways away from those measured). Bottom line, a modulus value (presumably one that sound engineering judgement identifies as the most applicable relative to the prevelant conditions) is an effective way to calculate settlements (whether they be elastic or plastic deformations).

However, as I am sure the discussers agree, the choice of the modulus must consider the loading conditions, initial and final loading, timing, and a miryad of other things. I would venture a guess and say for the discussion, if we are talking about picking the "silver bullet" value, I would have low confidence in my results and be conservative. On the other hand, if (depending on various methods) I can bracket the value and look at the problem statistically (there a very simple methods that none of us should be afraid of) and extract a degree of confidence from my analysis with little more effort, I am much happier. I agree with both poster's on various aspects of their comments, but reiterate Focht3's last statement (...I'm willing to entertain other approaches) as being the keystone to move beyond the "what's the right analytical number to use in my calculations" and more towards, how does the variance of the parameters I am evaluating influence the outcome of my results in a practical sense.

Just some thoughts on matters concerning identification of values for analysis.


RE: Settlement in Clay above water table

Two points:

1.  Have anyone involved here ever tried to use the particulate particle theory rather than elastic theory for settlements?  M.E. Harr (Purdue) wrote a book on the subject some years back.  Whilst I was tad "overwhealmed(?)" on the probability theory, etc. there were some good thoughts on the differences between particulate and continuum computations.

2.  I believe that sometimes we all must sit back and really take a look at the "accuracy" expected of calculations - especially involving settlement.  It seems that "papers" all hit the settlement "right on" but that is doubtful.  Most, I would surmise, would agree that if we are within 25% of the actualy settlement in our predictions we had a decent day.  In 2002 there was an elaborate study published in the Canadian Geotechnical Journal regarding the use of electric cones (piezo with two pwp transducers) to estimate the preconsolidation pressure and comparison of this work with Mayne's work.  These were on Quebec clays.  In the end, even with the "sophisticated" cone, the scatter of data was up to 20%.  So - are we sometimes getting "smarter" than we really are??  

A side note on accuracy:  When I was in Indonesia, I saw engineers and university professors taking a standard penetration test N value and determining the capacity of driven piles to the nearest 1/100th of a tonne.  Talk about accuracy - some accuracy . . . . not.

 - - -  I am kind of keen to see your thoughts on the particulate theory - - -  

RE: Settlement in Clay above water table

While I have heard of Harr's approach, I haven't read the book.  I guess I'll have to find a copy...

On the issue of accuracy: I'd call a day with all my settlement predictions within 25% of the actual value as an outstanding day.  To me, within 30% is a good day, and within 50% a decent one.  At least with respect to movements.  I'd go along with BigH's 25% range when talking about ultimate pile capacity and similar load capacity predictions.  For me, movement is much harder to predict.

Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members! Already a Member? Login


Taking Control of Engineering Documents
This ebook covers tips for creating and managing workflows, security best practices and protection of intellectual property, Cloud vs. on-premise software solutions, CAD file management, compliance, and more. Download Now
The Great Project Profitability Debate
A/E firms have a great opportunity to lead the world into the future, but the industry’s greatest asset—real-time data—is sitting wasted in clunky, archaic ERP platforms. Learn how real-time, fully interactive dashboards in a modern ERP allow you to unlock data that will shape the future of the world. Download Now

Close Box

Join Eng-Tips® Today!

Join your peers on the Internet's largest technical engineering professional community.
It's easy to join and it's free.

Here's Why Members Love Eng-Tips Forums:

Register now while it's still free!

Already a member? Close this window and log in.

Join Us             Close