Lateral Pile Loads P-Y Curves LPile/Com 624

[pgyr]

I'm in the the pile design phase for an integral abutment single span (130') bridge. LRFD per AASHTO. The subsurface exploration has been completed. We have slightly to moderately overconsolidated alluvial or lacustrine clays to 45 feet underlain by very dense sand and weathered shale. The clay appears to be reasonably uniform although there are sand "stringers" throughout-SPT results show some stiffer and softer intervals. Water table is about ten feet below abutment elevation.

Driven piles in a single row are the preferred choice of our structural engineer. Since I am confident of the end bearing stratum, my uncertainty lies with lateral load properties of the piles.

Laboratory testing on Shelby tube type samples has consisted of several consolidation tests (to confirm our choice of deep vice shallow spread footings and to evaluate embankment settlement magnitude and two unconfined compression tests. Results Su=1.2 KSF @22 feet and 1.8 KSF @ 32 feet.

Per FHWA NHI-05-042 and the COM624 manual we could use this data to assign K and e50 values for P-Y curve generation and proceed with the analysis. I understand LPILE offers the same approach albeit with an easier interface.
Does this approach provide sufficient accuracy and conservatism in our lateral loading analysis? Would we gain anything by performing three or four CU triaxes to obtain "better" data? My fear here is that non-uniform strength in subsurface will give us data scatter that will not provide p-y curves of any greater accuracy than those assigned based on C value in the literature.

If we proceed with program generated p-y curves based on C values, what type of parameteric exercise should be performed to assure conservatism at both ends of the spectrum?
Thanks in advance.

 

[Panars]

For an integral abutment, you want the piles to be somewhat flexible so that the abutment can move with thermal expansion of the deck. Using the unconfined test data to estimate the k and e50 values is probably sufficient. Ask yourself what is the failure mode for an integral abutment under lateral load. The braking force will also be resisted by the passive earth pressure at one of the abutments. The thermal expansion and contraction may cause problems in the superstructure, but shouldn't fail the piles.

Even if the stratified soil results in non-uniform p-y curves, the resulting shear, moment, and deflection of the pile will be "averaged" over the length of the pile, so the results won't be that much different from using the "average" p-y curves.

 

[Doc08]

If breaking forces / cyclic forces are involved you might want to consider Lpile (Ensoft) as it has an option for cyclic loading but the selction of P-Y curves could be a pain. As to the selection of parameters, depending on the magnitude of the loads applied, depth and diameter of the pile, I would venture to guess that most of the action will be taking place in the upper 20ft or so. So I guess that I would concentrate the additional testing in that zone. You can also avoid the PY curves selection process by using a software that adopts the bearing capacity equations, calibrated to field data, to calculate the soil springs. We own a copy of Lpile (Ensoft) and LatPilePro (Tsoft), the latter has the advantage of simplicity as it requires only easily attainable soil parameters (C,?,unit weight) or (Ks), so it takes away some of the guess work involved in the selction of P-Y curves, and the accuracy compared to filed measured results is not bad at all.

Good luck

 

[Doc08]

Out of curiosity I ran a sample problem using two layers of clay 25ft with a cohesion of 1.2ksf, and the second layer with a cohesionof 1.8. The pile is circular and 1ft in diameter extending 40 ft down. The applied load is 100kips. I am getting about 1in deflection at the top and a maximum moment of 404 k-ft occuring at about 10 ft from the surface. Reducing the cohesions by a half, raises the deflection to 1.5in and the moment to 478 k-ft occuring also at about 10-11ft from the top. I hope this gives a feel for the problem at hand.

 

[pgyr]

Our analysis is also showing the point of fixity in the upper ten feet. I think we will proceed with an additional test or two in the upper soil column.
Thanks