Effective Stress vs Total Stress in Pile Design

[VAD]

What is your opinion on the use of effective stress versus total stress approach in the determination of shaft resistance for the design of driven and bored pile foundations in soft and stiff soils. What are the limitations you percieve in your practice in relation to obtaining suitable parameters for these soil types?. What approach/approaches do you use for design, are you comfortable with it/them and why?.

 

[BigHarvey]

In France the design of piles is mostly based upon in-situ testing ( pressuremeter and CPT ) although our codes ( public works or private works )contemplate the two approaches ( lab and in-situ testing ).
More and more site investigation reports do not contain lab results ( cost and time ) but only in-situ testing results ( mainly pressuremeter ).
Our codes are very directive in terms of design and execution.We can only feel confident ! It is true that there is almost no patholgy on pile foundations in france but i believe we are too conservative also!

 

[MRM]

For clays, I use the total stress approach and for sands I tend to use an effective stress approach for a couple of reasons. The first reason is that when looking at clayey soil samples coming up, I feel more comfortable in estimating Su than I do estimating effective stress strength parameters. We never have the opportunity to run the appropriate tests to get the effective stress parameters anyways. For clean sands, It'sfairly straight forward to estimate an effective friction angle.

Limitations in getting those total stress parameters are many. The most important limitation regarding clayey soils to me is that I'm examining SPT samples that are already disturbed. The only thing you can do is to evaluate the samples before any major moisture loss occurs, and evaluate the soil itself to be sure that the Su parameter would even make sense. One other check I like to do is compare what you believe the Su to be with the ol' Su/sigma v=0.23+/- rule. That has to be one of the most useful relationships ever, in my opinion.

I've had relatively good results using total stress parameters for clayey soils. Most recently, we load tested a 60 foot long, 12 inch wide steel pipe pile. The pile was installed using a vibratory hammer into what I would call "textbook" nc to very slightly oc silty clay. After estimating Su, pile set-up, and running an analysis, I predicted an ultimate capacity of 20 kips. After about 14 days of set-up time, the pile failed between 17 and 18 kips. One of the few times I won a bet!

 

[GeoPaveTraffic]

For sands I always use effective stress.

For clays I have almost (aways?) used total stress. This is due to the area where I practice. All of the clays are between normally to slightly over consolidated. So the effective stress strength is much greater than the total stress strength.

 

[BigH]

VAD - good question and one I've pondered about but not for purposes of skin friction. There is no question that in granular (sandy, silty sand with very low plasticity - or coarser) you would use effective stresses. The driving or installation would have expected pore water pressure increases dissapate quite quickly. I think this is well established - Nordland method or others.

For clayey soils, traditionally the total stress concept has been used and for adhesion, the use of the Tomlinson or Focht factors would be used. For the stiffer clays, they values are quite low (nearly 0.4) - [red]but you know this[/red]. The use of effective stress analysis would be a step forward, but the problem as is well known, too, is that it is likely near impossible to determine the pore water pressure increases due to the driving or boring. The postive pore water pressure increases would have to be subtracted from the in situ stresses to get your effective stress at a point against which you would make your estimate of the skin friction. Since this is very difficult, as in slope stability, we would use total stress analysis.

I've always wondered, though, if we shouldn't use effective stress parameters for long term end bearing in heavily overconsolidated clays (given, of course, that the bulk of the load has been transferred, in fact, to end-bearing. If the loads are high enough, the negative pore water pressures could dissapate (negatively) to the failure line and, in fact, you would have reducing factors of safety with time. But, in almost all problems, though, only 20% or so of the load at most would ever get to the base in normal circumstances so that the base is, in fact, the factor of safety.

I would love to use lots of in situ testing - but the jobs would have to be able to spring for the added costs - and there will be more costs than in the normal run of things - at least for now. Hopefully, things will change.
Again - good question -

 

[MRM]

Regarding stiff clayey soils, I agree with BigH. I would think that the long term conditions would control in that case. The same situation as bearing pressure of a shallow footing established on stiff clays in the case of bearing capacity. Because the long term condition controls, effective stress parameters would be more appropriate.

Ah hah! This brings up a question of my own that I've always found interesting and important, but one that has many opinions. Please check out my other new thread...

 

[dmoler]

This may seem slightly off topic at first, but I hope some agree that it is not. This thread touches on something I have a problem determining for several reasons, and that is the long term conditions. What exactly is "long term", and do we really design for them?. Are the "in-situ" conditions really worth getting extremely exact about for design of long term structures?

For example, I just completed a 60-story building geotech report for foundation recommendations, and several 40-story, and have some trouble with predicting the long term performance of the building with any real accuracy. I have defined the long-term conditions as 150 years for convience. While I will be dead and gone by the time "long term" conditions happen, there is really little to go on in my area (Florida) as an example.

There are no buildings here of that magnitude with a history greater than 30 years. Many settlement monitors were installed for the taller buildings in the 80's, and some data is available for those, but not long term. These are not normal structures, but have very concentrated loads within a small area. Of course, the owners want them to last very long, as they are very expensive, and quite beautiful, actually, at night, on the Atlantic Ocean beachfront and all.

In Florida, load tests are required for any pile over 40 tons, so I have done literally hundreds of load tests in sand/cemented sand/soft limestone. The FHWA drilled shaft manual (Beta method) is really accurate for design of shafts (and augercast piles) in sand/cemented sands/soft limestone. Pile capacities are fairly to very predictable with this method in sands, since the drilling does seem to remould the shaft/sand interface to a predictable resistance. But again, these are short to medium term design times, not really long term.

Settlement estimates give really high values for settlement using correlations from SPT borings with the Schmertmann, Kuhn, and Poulous methods for a long term time frame. The settlements are actually scary for a long term time frame, and only time will tell what happens with these methods.

So, would a clay friction angle of 0 for design purposes really be crazy if the lab test indicate that there is a friction angle? Would the clay slowly creep under real long term loading, and the friction angle reduce to 0 in the long term? Would foundatons in sand with concentrated loading last for the short to medium term, with the long term settlements eventually destroying them? I have a hard time actually saying that current design, especially for commercial development, according to accepted industry practice, is ok for a real long term design. But that is because long term design vs. performance is really not know in some cases.

 

[BigH]

dmoler - for me, there is no comment that is off topic in our group - for in geotechnical engineering almost every comment has some applicability to some problem or another in other areas. Secondly, the whole purpose of our group is to be able to be more "free" about what we write, say, etc. I am hoping that we can get more case histories - with some real data, etc.; also case antedotes.

For me, the long term is judged to be at "nearly" 100% primary consolidation - knowing that secondary consolidation has already started for some zones while primary is still going on. In stress path, that is when the drained condition is finally reached - either from the normally consolidated left side or the overconsolidated right side of the drained stress path. This governs the final strength of the material.

As for settlements, you either have creep settlements (see Schmertmann in sands) or secondary consolidation - but in the later, unless you have clays with substantial secondary properties, the continued settlement would be small - excluding, of course, peats, organic silts and clays, etc. I don't think that you would need to consider 150 years in your analysis; most projects have a projected project life (50 or 80 years). Also, it is very difficult to predict settlements that accurately - your secondary (very long term effects) would/could very well be in the inaccuracy range of your initial estimate. For clays, +-30% would, I believe, still be considered a good estimate; for sands, a paper years ago in Ground Engineering implied to pick three or four of the available analyses (he reviewed something like 15), then take the average of them - arithmetic average? Nominal average; geometric average? - I'll let that be as per another thread in the main forums!

 

[MRM]

dmoler,
Regarding the friction angle for clay you mentioned, please specify if you are talking in terms of the effective stress world (i.e. drained conditions), or total stress world (i.e. undrained condtions). What type of test is giving you the numbers? I'm not sure I follow...

 

[dmoler]

Hi all,
I was away to the Dry Tortugas for a few days for an island project.

Anyway, I promise not to write anything in the future I have not fully proof read. One of the things I have now learned is, once something is posted, it cannot be corrected.

Above, I meant the cohesion, not the friction angle (overworked, underpaid = mistakes sometimes). I have found that the cohesion=0 concept is resisted for long term design of slopes in some areas of the US.

Regarding drained vs. undrained conditions. This ties in to what I was getting at above. I have always used drained vs. undrained based on the water table at the time. But due to the ever increasing population boom and construction boom, the water table keeps lowering, at least in Florida anyway. So, drained vs. undrained for long term design? I am not really sure.