Foundation Design Approach of Past, Present and Future 1

Colleagues:

Ran into an interesting paper the other day - on the net but protected. Please refer to Geo-Strata Magazine, November/December 2008 issue. The paper of the Subject title was written by Dr. Fellenius.

He puts forth the idea that many of us have - that capacity, except in few instances of very soft to soft clays is not the relevant design characteristic to consider - but that design should be based on serviceability issues, i.e., permissible settlements, rotations, etc. Many of us have basically said the same thing in many of the threads. This leads one to wondering why design codes put so much emphasis on capacity determinations and so little on serviceability - i.e., AASHTO LRFD says to use Hough's equations for settlement (yet in some 15 text references I have with me in Indonesia, none of them mention Hough (I know, OldestGuy, you like Hough - and perhaps rightly so)) So few codes, that I have seen have given any more than a scant sentence to interaction of adjacent footings - concentrating only on single footings unencumbered by anything nearby.

Let us know your thoughts on Dr. Fellenius' article!

 

[moe333]

FYI, you can download this paper from Fellenius website:

http://www.fellenius.net/papers.html

I have not read this paper yet but I have read a lot of his other publications and have always found them to be very insightful and right on target.

 

[jheidt2543]

BigH:

I found the paper at this link:

http://www.fellenius.net/papers/284 Past-Present-Future.pdf

 

[escrowe]

Yes, expected settlement and related issues should control foundation design. Terzaghi (1938) and quite a few others have commented on this issue. 'Settlement governs' is a common phrasing in the design texts.

But how to change codes to reflect this?

As a practical matter, I think we'll have to continue to address foundation design in terms of allowable bearing capacity until the building codes no longer reference it.

 

[Mccoy]

Dr. Bengt Fellenius, since I know him, has always been insisting on such an idea, in part pretty much sensible, in part pretty darn one-sided.
Going to read his paper and I'll be back later

 

[Qshake]

BigH - Perhaps your question is one of rhetoric but if not I will offer my humble opinion from the dark art of structural engineering.

First let me say that I agree with the serviceability approach for certain soil types.

Our fascination with load based designs I believe are based on the structural engineer or architect needs. Since our main concern is transmitting load, we, as an industry, have come to speak directly in those terms.

When serviceability issues arise, again, the SEs or AEs wanted an answer in terms of settlement for a given load.

Naturally, if you do this long enough someone is going to write a code with it as the main crux of the matter.

With the many years I have I have also learned the codes are, believe it or not, geared to simplicity so as not to overwhelm engineers with cumbersome, highly technical content. In this light, code writers examined the single variable and try to write all equations and or code material with that in mind.

Regards,
Q

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[BigH]

Thanks Q - any relation to Mr Gadget in James Bond? Mmmm

One of the points that seem out of line is forcing one to perform calculations for 4 or 5 load cases where, when compared to experience/judgment and subsequent computations, they are not critical since serviceability governs. Seems like a lot of wasted time. For me, in most cases, I use 2*Su for clays and seldom bother for granular. Of course, I hope I have enough experience and judgment to know when I MUST delve into the bearing capacity in a more depth.

 

[theCorkster]

In practice the question that always has to be answered is what governs the design, and that is serviceability; then define serviceability. Bearing capacity and settlement analysis go hand-in-hand, together. Experience is what guides you to know when either bearing capacity or settlement governs, and when you need to evaluate both. And simply, I do not believe that this is something that can be or should be codified.

As an example, the governing factor for the foundation design for a 200-foot tall stack was based on differential settlement - the differential settlement of the supporting mat that would result in the public's perception that the stack was "leaning too much." The same is true for earth dams with slopes constructed flatter than necessary because the visual appearance of the steeper slope is worrisome to the casual observing public.

Define serviceability, then analyze the problem within that context.














. enforcement personnel in many instances have neither the experience of interest in delving into the background of the code to understand when the coreviewr's

 

[oldestguy]

OK Big-H, you may recall Hough was my prof in undergrad and grad school. He only wrote a couple of ASCE and ASTM papers and a text that was quite practical and spent his time working, not writing. A very practical guy however.

Who does the most writing and therefore gets recognized as "engineer of the year" etc, due to this "professional" activity and giving seminars. Therefore it is only natural for code writers to assume they are experts and should be relied on.

In my experience I have been to many a seminar by these so called experts. In most cases I have walked away wondering why I went, having learned nothing new.

The moist I have learned from others came from practicing engineers like Rutledge, etc. However which of them, besides Osterberg, are given credit for their achievments in codes?

It's the old professors with their many papers and books. Hough had a term for this: "diarrhea of the pen".

 

[BigH]

oldestguy - there was no disparaging about Hough - how could/can one be for a fellow "Big Red" - and, in fact, you are right. One can learn so much more from practicing engineers than many professors - for me, it was Matich - for others it would have been Golder or Milligan. The one point that I was trying to make, though, is when codes forces one to use a specific method rathe than allowing the engineer to assess what is most suitable for his particular case. (probably haven't said it that well . . .) We did okay against Mizzu, eh? - for the first half . . .

 

[Mccoy]

There is a previous paper on the subject by dr. Fellenius with some more detail:

http://www.fellenius.net/papers/219 Delusion of Bearing Capacity.pdf

Besides, everywhere he doesn't loose the chance to expose his 'heretic' ideas about bearing capacity and the fallacy of modern codes such as LRFD and the Eurocodes.

I'd say he's not wrong on one side, on the other side:

[ul square]
[li]what about the velocity issue? I don't know if I'm saying some foolishness, but we might as well call a very fast settlement a bwearing capacity failure since the structure has no time to accomodate itself to the movement. The velocity variable is not contemplated in the paper figures [/li]
[li]While we have miriads of lab tests where a clear failure surface is visible (I like the Hill's surface pictured in the Lamb & Withman book) we have on 2-3 real size tests to show that there is not a phenomen which can be called failure. A tad to little to draw drastic conclusions as those illustrated by dr. Fellenius[/li]
[li] We all know that there are some parameters like fooundation width which govern the limit state which governs design; in example, in narrow footings bearing capacity calculations often govern whereas in wide footing it's the other way around.[/li]
[li]There are some foundations types where a failure by bearing capacity would be almost unheard of, like in big slabs, hi-coverage foundation grids and similar. However, since bearing failure entail potential loss of human life and substantial economic damage, codes impose the relevant calcs. Rationale is that probability of failure must be lamost nihil, and we have to show in our calcs that it is indeed like that[/li]
[li]what about design in seismic conditions, more parameter, sometimes with complex interactions show up so it's not so simple to just say: settlements govern[/li]
[/ul]

Speaking of deep foundations, actually there is indeed this accepted, maybe warped logic where an index of bearing capacity failure is given by a settlement ratio (in relation to base dimensions). Again, maybe velocity of settlemnt might be the key factor of separation here, please tell me if what I'm talking nonsense.

 

[Mccoy]

One add-up to my previous post:

it is not totally true that the geotech community is still standing at the traditional, 'outdated' 3-terms formula.

Since about 10 years has been available an interesting solution, which also appears as an attachment to Eurocode8 - 5 (the seismic part of the European building code), known as the Paolucci-Pecker formula (I can see BigH smiling):

http://books.google.it/books?id=2No...X&oi=book_result&ct=result&resnum=1#PPA265,M1

also please see Gazetas on seismic bearing capacity issues

http://www.ecees.org/Paper/paper4007.pdf

Here, again failure is defined as 'unacceptable permanent displacements' so 2 conditions are required:

1- deformation must not be elastic, i.e.: irreversible
2- deformation must not be small compared to structural capability to withstand settlements

 

[mudman54]

let's keep the triple N formula, but it is only step 1 in the evaluation. it is always instructive to see if the building code presumptive bearing capacity can be derived from the formula to see how your site material properties are working. wouldn't it be great if we had more published O-cell data so we could tackle the settlement problem!