Superstructure-Pile Interaction considerations,any logic? 6

[IJR]

The routine procedure in modelling superstructures is to start at foundation level, not base of foundation even if long piles and weak underlying soil exist.

There is usually a little suspicion in utilizing this procedure when long piles are involved and a good mass of soil below superstructure is known to be weak.

Should one model the whole thing then? I noticed that even in this situation a thick pile cap is usually used connecting a large number of piles together. In my opinion unless under severe quake loading will this pile sway significantly and may not even sway given a good damping in the even weak soil

Does any one want to extend this discussion?

respects
IJR

 

[ocean13]

well....i'll model the substructure and superstructure separately.

depending on the fixity of your pile ends, the moment in the pilecap can be quite significant. will end up with a heavy reinforced pilecap. this obviuosly is directly linked to whatever load you put on top (from the superstructure.

 

[Ron]

IJR...the structure / pile interaction is most important for conditions other than compression loading of the pile, i.e., for seismic or wind events. Most codes require consideration of overturning under wind loadings and mass-lateral interaction for seismic events, therefore tension and shear interaction are significantly important.

Should they be modeled together? It doesn't really matter a great deal. I model them separately, just because that's a habit of my early training. To model them together requires a compromise of the modeling parameters as one modeling technique or program does not typically handle all the variables involved very well. As an example, the soil shear resistance for uplift computations is not conveniently modeled by the same program you would use to determine the stress state of the superstructure. You have to "fudge" the parameters based on experience and good guessing, whereas if you use the resultant reactions from one specific modeling program to feed another, I think your results are more accurate.

 

[emmgjld]

Ron directly hits the issue.

The structure is rigid, with planned/designed flexibility built into the hopefully correctly constructed superstructure. You 'know' what it is.

The foundation is constructed in a mass which is somewhat flexible and changable, due to weathering, ongoing settlement/expansion, changing water conditions etc.. Too many characteristics must be estimated, often in a gross manner(the pun is partually intended).

Calculate/model what you can, in the most accurate manner, then combine. No sense in estimating (fudging) what is possibly/probably already an estimate.

 

[Hariharan]

Depends on what structure it is, and what is the sensitivity of
the results to foundation behaviour. What is the structure you
have in mind? What analysis are you planning to perform?
Static, dynamic, seismic ?

Hariharan

 

[IJR]

Hariharan

I am inclined towards doing separate analyses. But here is a response to your request for clarification

I am going to put a stadium wing on a weak soil. The wing is a RC frame structure and is large capacity(a crowd of not less than 15000 per wing). And being so huge, its overall height goes up to 120ft(30m). There will be 5 floors under spectators.

The soil is very weak (SPT reports N=2 for a couple of meters)and thick piles will be used. Pile diameters 3ft(1m), over depth of 36ft(12m) and connected to a slab 4ft(1.2m) all under a wing.

Seismic zone 4 to UBC(That is the highest in seismic zoning here). The wing is split in 3 blocks each roughly 40m by 40m in plan(120ft*120ft).

Thanks for your interest and any further comment is most welcome.

Respects
IJR

 

[Hariharan]

IJR,

This should be an interesting project!

1. Being in Zone 4, I expect you will have shear walls to resist
lateral loading, so as not to stress the columns in bending.
This would put rigidity into the superstructure whose
natural period would be quite low, and the response spectrum
ordinate would be high.

2. Now, if you consider the flexibility of the foundation system,
the period would be higher, and you need to design for a
lower lateral loading. For this purpose, it is not necessary to
fully model the soil-pile-structure interaction in the analysis. It
is adequate if you compute the stiffness of the pile in lateral
loading and incorporate it as a boundary condition in your
superstructure model.

3. If the soil is very weak at the top, I presume it is much
stronger at the bottom so that you can terminate the piles at a
relatively shallow penetration of 12metres. The piles could be
critical in tension, unless the dead loads are very large.
The pile is a short pile (L/D = 12) and its
lateral capacity could also be quite small (in spite of the large
diameter!). I would recommend a lateral load analysis of the pile
to be performed first to determine the pile spring constant and
preliminary design; use the derived info. for further analysis.
This would reduce the number of iterations to arrive at the final
design.

4. Do you need such a thick slab? Can you excavate the top soil
and place the slab below, so that the pile is not subject to large
bending stresses? The economics need to be worked out,
of course.

Good luck,

M. Hariharan

 

[IJR]

Thanks Hariharan for keeping the faith and for the concentration on the subject.

I agree with you on all aspects.

1) I think I missed your point when you said "lower down the slab to cut down pile moments". Currently the piles are attached rigidly to the slab, wont that help?

2)Actually pile and slab is not part of my assignment, and I considered using pile flexibility to advantage in cutting down seismic energy intake of my superstructure, but the geotech info is not so clear to me, and may not be complete given a very large stadium area. For example, is it correct to model all piles with the same free length?. Is underlying soil so uniform to allow this?

3) Given such uncertainties, how much safety is there if I model my superstructure with a diaphragm constraint and lateral spring supports at the base?.

4) If I do so, what level should be considered as ground excitation level?

5) The nature of the stadium superstructure is such that sufficient stiffness should be built into it, given large crowd(this is a stadium for the best soccer team in here)

Somewhere sometime I will probably be involved in such detailed analysis and your every word will be taken seriously

respects
IJR

 

[dik]

Something I've done... by thick piles at the top; is it possible to use the auger to 'nibble' the top of the pile to flare it in the direction of the shear wall to create a pile that may be 2' in diameter and for the last 6 or 8' going from 2' dia to 2'x6'?

 

[Hariharan]

IJR,

Further deliberations:

1. Piles should be mandatorily attached rigidly to the slab.
i.e. the connection should be a moment connection. I am not
very comfortable with pin connections in concrete!
The point I had made was that if the top soil is quite weak, it
does not offer much lateral resistance to the pile. The pile
therefore has a longer length to point of fixity and hence
higher BM due to lateral loads. If the poor soil were excavated
and the slab cast at the lower level, the length of fixity would
reduce and so would the BM.

In a pile-founded structure, all loads will go to the piles and the
slab will really not act as a foundation. This will be more true
if the bearing layer is weak. The pile could experience negative
skin friction which would reduce its load bearing capacity.

If the poor soil is removed, these problems would not exist. The
slab would be cast over the better soil.This could give you a
basement type additional space for other utilities as a bonus!

The negative aspect is that the system would be more rigid and
the lateral loads would be higher. That is why an economic
analysis would be advantageous.

2&3. Whatever geotech info you will get will be an 'average' or
sometimes 'lower bound' value over the area. It is absolutely OK
to model all the piles with an average, uniform property. For
seismic analysis, however, I would consider an upper bound and
a lower bound for the lateral stiffness and design for the more
conservative case. No geotech data will be "accurate", and the
upper & lower bound approach will give you a good level of
confidence in the design.

4. Ground excitation level should be the slab level in this case?

5. Noted. Hence the need for incorporating the pile flexibility.

Hariharan

 

[Zulak]

It seems odd that your geotech report is that vague on such a significant structure in a seismically volitale area. In all the projects that I've done, I haven't found a site yet that has uniform soils across 50 feet of the site. But, unfortunately, you can only design with what was provided, so you must design based on worst case conditions. Or you can recommend that further soil investigations be performed to more clearly define the limits of the poor soils. It is always better to be safe than sorry.

 

[IJR]

4th July greetings to all pals in States

DAMPING!!

So far all friends above who seem to suggest pile flexibility be included in analysis have seismic responses in mind.

now

However bad the soil may be, and given that hundreds of piles are connected to one another by a thick slab and thus forced to move in unison, wont damping turn off vibration in a matter of milliseconds?

How reliable is soil damping information?

Regards
IJR

 

[kme]

Could you elimite the lateral issues on your piles with some batter piles? This may bring you a little more closer to a "rigid" foundation system (i.e. high natural frequency) where you may eliminate or greatly reduce any amplification effects from you pile system.

 

[ishvaaag]

Just some points on what read.

Battered, er, not in seismic area 4. Heads can shatter.

Where is level 0? if no soil interaction model? Start from the slab, yes.

IF the partial purpose of this deliberation is to allow for lower energy intake for the design, for the final studies, if you can, do a complete and the best soil-interaction model to your avail. It is only logical that if the purpose is more economical works more dear design concurrs.

Substituting horizontal springs at foundation level as substitute of what in piles happens? In my mind not for the purpose above, but could be well for others. By the way, the spring not as much "the stiffness of the piles" but of the piles embedded in the ground.

In general the concepts seem quite correct; a rigid box structure that through piles maybe takes more energy than if not, but precisely the piles soon will damp through moving within the ground and their deforming in their length, specially their heads. I think is a sound stiff cell concept. In any case, having a rigid slab, maybe the effect can be gained at lower energy and money cost only form a slab, but for some reason the piles will have been selected.

 

[ddaacc]

ishvaaag,
at what type of ground acceleration and batter angle wpoud you be concerned abbout the heads shattering?Thanks.

 

[ishvaaag]

Construction of Prestressed COncrete Structures 2nd ed. by Ben E. Gerwick Jr, in more than showing a failed battered pile head photo under seismic event (p.270) after signaling that the use of raked piles is not appropriate in areas of high seismicity simply pass to state that in such areas "Only vertical piles are used", reason being given the whole seismic shear concentrating on heads. It gives additional guide for the vertical piles for such areas.

No guide is given on inclination.

Given that the reason is the concentration of the shear at the head one must guess the problem starts soon, (5 to 10 deg?) but one might venture 5 deg from vertical not being much critical since this degree of tolerance is many times accepted for the piles behave as vertical.