Foundation failure mode

[FSS]

I am trying to visualize possible failure modes for foundations of elevated structures supported by piers, caissons, or even piles. Examples include communication towers, stair towers, or any other structures with multiple legs. To keep things simple consider a four legged frame (with four individual piers) with lateral loads applied at 45 degree angle. Assuming that two opposite corners are along the resultant neutral axis, and each of the other two corners are either in tension or compression, how will foundation failure occur? If tensile capacity controls pier design, how can upward movement occur without associated compressive failure, which is usually a higher limiting value? So does it really "fail" under tension?

Your comments and any technical references would be appreciated.

 

[ocean13]

you need a structural analysis approach to consider your problem.

well, depending on the resultant forces and moments from your "structural frame" and what type of foundation u r considering; possible failure mode:
bearing capacity failure, pile structural failure,
overturning failure, sliding failure....etc.

any structural member in tension are generally weaker than in compression.

 

[FSS]

Perhaps my question was a little too wordy. Based on analysis assume a simple case where one drilled pier has an uplift load and one has a compressive load. Although tension usually controls design of individual shaft, how can the shaft fail in tension if the opposite shaft does not fail in compression (since it has a higher capacity)?

 

[ocean13]

well, imagine a one-legged monster balancing and crumpling down anytime

 

[Hariharan]

FSS,

Your question is perfectly valid. An individual element, when
stressed to the limits, will behave differently (inelastic, plastic,
in contrast to elastic behaviour assumed in analysis). The
structural system will not fail until the structure is unable to
carry any additional load, i.e., the structure behaves like a
'mechanism'. This behaviour can be analysed using nonlinear
analysis FEM packages. The analysis is called "pushover
analysis" and is used to determine the overload factor available
prior to collapse of the structure.

In your specific situation, you are right that the structure will
not fail if only a tension pile reaches the limits. Strictly, the load-
deformation characteristic of axial load transfer between pile
and soil is nonlinear. As the load nears the limiting value, some
of the loads
will be redistributed among the other 3 piles until a structural
or soil failure takes place. This behaviour can be studied using
the pushover analysis. However, in this kind of behaviour, the
deformations resulting from nonlinear behaviour will distort
the original results you got from linear analysis and stress
other connected elements more. The benefits of considering the
nonlinear behaviour are not substantial for general structures
and should be ignored, considering the uncertainties in soil
properties and foundation design.

This is the theoretical aspect. As a practical designer, I wouldn't
take any benefits from such behaviour unless it is fully
justified (e.g. in some cases where we have to evaluate the safety
of an existing structure). The situation you have described is
usually encountered in piled jetty structures.

We recently tried to verify the effects of higher pile settlement
on the safety of a jetty structure. The pile was quite safe, but the
connecting beam elements at the superstructure were getting
overstressed.

Hope this helps.

M. Hariharan

 

[dimitrios]

hello FSS,

I couldn't help myself from writting down what I understood from your question and description, although there is a strong danger that I might show myself to be a little naive.

In your initial description I understand that in order for a moment to produce compression and tension stresses in the opposite piles there must be a certain resistance from the piles to the moment acting on the pile structure.

I mean that if the piles are in a extremeley soft soil then the whole structure would simply pivot and there would be no tension stresses in the pile. Tension stresses present themselves in one of the piles because this pile is not allowed (or at least partly resisting) the uplift movement. So a pile that due to its embedment length and lateral friction, refuses to move upwards will be subjected to tension and perhaps fail.
Now the other pile that is subkected in compression does not have to break in compression (necessarily) since the whole structure can pivot over this compressed pile (I think with a relevant displacement of the position of the neutral axis) pulling at the same time on the pile in tension (which continues to resist), which may lead it to its failure and therefore continue the movement if the acting moment is not influenced by displacement.

This is actually my understanding of the problem.

Regards

 

[bogard]

A simple solution would be to consider the piles in the direction of lateral load on the structure as providing a resisting couple, with tensile (downward reactions) from the trailing piles and compressive (upward reactions) from the leading piles. The resisting moment due to axial loading of the piles is limited by the tensile capacity of the trailing piles.

Additional lateral load and moment can be resisted by shear and moment in the piles themselves. Failure usually occurs when the piles on the compression side fail in bending. The structure initially rotates about the centroid of the reactions, with some horizontal translation. After the trailing piles fail in tension, further rotation occurs about the connection between the leading piles and the superstructure. When the leading piles fail in bending, either by forming a plastic hinge at the structural connection or by crimping, the structure will collapse.

The piles need not fail in compression for collapse to occur. Any structure can overturn with no compressive failure in the soil or the foundation.