Settlement calculation - layer thicknesses 1

[danyul]

hello
this is rather basic but i have seen several variations and was wondering what you have to say...

for consolidation settlement calculations, assuming a homogeneous soil, square ftg, how many layers and what thicknesses are appropriate?

examples ive seen:
3 layers - 0.5B, B, 2B seems conservative
2 layers - B, B
every half foot until %reduction is 5%. conservative
thanks

 

[fattdad]

Depends on the thickness of fill for grading/building pad too. I'm not sure it truely relates to the footing size. Also, need to consider the potential for live loading if it's a warehouse with rack loads.

f-d

¡papá gordo ain’t no madre flaca!

 

[msucog]

i believe either usace or navfac/ufc have suggestions in terms of B.

for homogeneous soil, i believe it says something like .5B, B, 1.5B,2B

for layered system, maybe .25B,.5B,.75B,B,1.5B,2B

obvious caveats would be to use changes in strata, consistency, etc as well as your judgment. and for consolidation settlement calcs, i'm not sure what the true definition of "conservative" would be unless you've got lots of field data versus calcs to back it up.

 

[oldestguy]

A useful rule to use for these computations is to run your calcs down until the added pressure is 10 percent of the existing pressure. So, for a lightly loaded footing in the bottom of a deep excavation, you might find that depth of calc pretty shallow.

On the other hand for a footing placed on a compacted fill, it gets more complicated, since the rules you state could not do so well. Timing of the fill is important also.

Another situation that I like to play with is a wall footing for a basement, with fill placed later alongside the wall.

I guess I didn't help much, but keep that 10 percent rule in mind on these and it takes care of many situations.

Oh, use what ever layer thickness you want, recognizing the approximations involved with the whole process.

 

[msucog]

i too agree with the 10% that oldestguy mentions as well as the special situations and other comments.

"geotechnical engineering is an art"--very true statement

 

[danyul]

when you say "10% of the existing pressure" does that mean where the pressure is reduced to 10% of the original stress in the ground (gamma*depth) or where the pressure is reduced to 10% of the original load?

im guessing the second...

thanks

 

[oldestguy]

hi danyul: The wording is "10 percent of existing pressure",, meaning that the ratio of delta p divided by the intergranular existing vertical pressure becomes under 1/10. The pressure distribution that you may get from your footing likely decreases rapidly with depth below the footing (I simply divide the area of the spread out load with a 2:1 side slope from footing edge into the footing load, ignoring weight of footing to calculate that added pressure) It is that reduced vertical pressure that then is compared to the existing vertical intergranular pressure that is there before any work is done. I take into account submerged unit weight of all soil involved. Drawing a chart of pressures versus depth is an easy way to do it.

Using this approach, you can also take into account what depth you need to calculate to when you are figuring what happens when a surcharge is placed. In that case the decrease of added pressure with depth may not decrease much with depth and so you are then using many layers down below that original ground surface.

 

[Mccoy]

danyul,
some codes like the European geotechnical code specify 0.2 sigma_v, that is you can stop at the depth where delta_P <= 0.2 sigma_v, a condition a little less conservative than the one cited by oldestguy.

As to the layer thickness, I'd choose it as little as practically possible, of course according to the type of tests which have been carried out.

 

[danyul]

for a continuous ftg... would you still go to 10%...

seems like it is very deep before it gets there.

 

[oldestguy]

Yes and you are right, the effect is deeper for the same footing width. But, most continuous footings are narrower and so that is an advantage in getting a rapid diminshment of pressures with depth.

I have spreadsheet programs set up for each kind. There are probably a zillion softwares out there set up for these comps, but a spreadsheet gives you the opportunity to really play with the variables that a dedicated software program may not allow. If you feel like it, let the spreadsheet go deeper than 10 percent, but notice the minor effect it has on the result

The main thought I'd like to leave is where you can actually monitor the performance of a site under loads, it is much more comfoftable than trying to predict settlements by use of computations alone. Think about how you would do this for a footing.....not easy, but doable at times. What does a fully loaded scraper weigh?

One more item: Take a site where you wish to preload the site before placing spread footings. Some overload of the shallow soils (beyond the surcharge loading) from the spread footings seems justifiable, since the depth of "not fully compressed soil" below that footings is quite shallow. This can be avoided pretty well by undercut and compacted fill placement if it appears important.

 

[BigH]

You could use a single layer - but the "characteristic" point (i.e., existing OB pressure locations) would be based on the average "center of gravity" for the load. - typically in the order of 0.42 or so of the depth of influence if I remember correctly (and not "half" depth). McCoy - I see that your namesake put out another CD this year!

 

[Mccoy]

How are you, bigH, lately I've just been browsing quickly the forum discussions. Gotta linger around more.

Do you mean 'guitars'? I'm not sure it hit the Italian music stores yet, I'll have to make sure. Rythm section is exceptional: Mccoy, Jack DeJohnette and Ron Carter, plus various guitarists. I really look forward to listening to it!

 

[Delijosi]

I have done similar calculations before and I will suggest that you use 1m laer thickness in the top meters where you would expect more settlement. You can then increase the layer thickness to 2m and so on. Genrally, using thinner layer thickness may increase the accuracy of the settlement values.

Eurocode 7 and probably BS 8004 recommend that you stop settlement calculations at the point that the net applied vertival pressure equals the vertical overburden pressure at that point.

 

[BigH]

Delijosi - increasing the "accuracy" of your settlement calculation is a bit of a stretch in my view. It is a well known fact that if you can predict consolidation settlements within 25 to 30%, you've had a good day . . . (the Luck of the Irish are with you). As for your last paragraph, this is at odds with reality - theoretically it will "never" equal the o/b pressure - which is why most use 10% as a stopping limit.
Further, one cannot know unit weights to an accuracy as would be implied - and the variation of unit weights in a particular stratum (this is one reason that statistics are coming into the mix).
Pressure distributions, settlement calculations are still, in my book, an art (and luck) and not a science. How does one calculate pressure increases due to footing load anyway? Elastic theory - as load on a semi-infinite half space? . . . or on the theory of particulate matter? or using plasto-elastic strains . . . Given that actual pressure increases are inaccurate (used loosely), how can one increase accuracy of settlement computations by varying layer thicknesses. If same soil, doesn't seem necessary - unless it is to minimize difference of curved effects of pressure increase rather than in linear fashion.

 

[iandig]

As an addition to the above and for those of you who are familiar with 'Foundation Design & Construction' by Tomlinson, he actually has a similar approach although he suggests a value closer to 20%. My version of Tomlinson is the 6th edition and it is shown in Figure 2.32 which suggests that the 'lowest level considered for settlement analysis' is where the ?z is = to 0.2??’vo. In practice for small pad foundations this returns a value of around 3 x foundation width [which I have seen used as a rule of thumb by some Engineers], however for bigger pads/slabs this factor reduces. As it’s a simple equation, its easy to factor it into spreadsheets and you can reference it to a well known publication as justification for the assumption.

 

[Delijosi]

First let me say that there was an ommision in my second paragraph. I meant to say 'settlement calculations at the point that the net applied vertival pressure equals the 20% vertical overburden pressure at that point'.

iandid explains it well.