Professional Documents
Culture Documents
net/publication/325612179
CITATIONS READS
2 374
2 authors:
All content following this page was uploaded by Gong Chaosittichai on 03 May 2019.
© ASCE
IFCEE 2018
IFCEE 2018 GSP 294 322
Full-scale load testing serves as a means of paramount importance not only to better
understand the actual pile behavior under lateral loads but also to validate the results of the
aforementioned analytical and numerical methods. In spite of this, a very limited number of
lateral load tests on piles or drilled shafts in Bangkok clay have been performed. The majority of
these are tests on large-diameter bored piles. It is impossible to test such shafts to large
displacements or deflections because excessive lateral forces would be required. As such, full
details of load-deflection behavior of deep foundations in this particular urban soil are not
Downloaded from ascelibrary.org by Pongpipat Anantanasakul on 06/19/18. Copyright ASCE. For personal use only; all rights reserved.
available. The present study was undertaken to fill such a gap. Three field load tests on
instrumented steel piles of moderate size were performed. The piles were loaded to head
displacements up to 125 mm. The load-deflection responses of test piles were investigated, and
the applicability of the Matlock y model for soft clay for prediction of the behavior of
laterally loaded piles in Bangkok clay was primarily assessed.
Full-scale lateral load tests were performed to study the load-deflection responses of driven piles
in Bangkok clay. The site for these field load tests was located in Mahidol University, Nakhon
Pathom, Thailand. The university is located in the western suburb of the Greater Bangkok area
with an average elevation of 2 m above mean sea level (MSL).
A dedicated subsurface exploration program was conducted to characterize the ground
conditions. Two boreholes 20 m deep were drilled by wash boring. One borehole was located at
the center of test site, while the other was located 50 m away. Intact samples were collected
every 1.5 m using Shelby tubes from a depth of 1.5m to 13.5 m. At greater depths, the soils were
relatively stiff, and samples were collected by split-barrel samplers. Standard penetration tests
were performed every 1.5 m from a depth of 13.5 m toward the ends of boreholes. The
groundwater level was observed to be at 1 m below the ground surface, and it did not noticeably
vary within the period of the load tests. Laboratory index properties tests, 1-D consolidation
tests, unconfined compression (UC) tests, unconsolidated-undrained (UU) triaxial tests and one
consolidated-undrained triaxial (CU) test were performed on the collected samples. The results
of exploratory boring and laboratory experiments are presented in Fig. 1 and Table 1.
From the ground surface to a depth of 12 m, the soil deposit consisted mainly of medium
clay of high plasticity. There was, however, a seam of low-plastic clay between 4-5.5 m. The
undrained shear strengths determined by the laboratory UC tests and UU and CU triaxial tests
were in good agreement. They appeared not to change appreciably with depth and consistently
varied between 25-35 kPa. The top clay layer was slightly overconsolidated, and the
overconsolidation ratios (OCR) varied from 2.5 at a depth of 3.25 m to 2 at 7.75 m. The
specimen for CU triaxial test was obtained at 7.75 m below the ground surface. To simulate the
field conditions, the triaxial specimen was consolidated to a maximum isotropic effective stress
of 105 kPa and then unloaded to 51 kPa, thus resulting in an overconsolidation ratio of 2.06 prior
to shear.
© ASCE
IFCEE 2018
IFCEE 2018 GSP 294 323
The top clay layer was underlain by a layer of very stiff clay of high plasticity, 6 m in
thickness (12-18 m). The values of N-SPT determined at various depths within this layer were
almost similar, and these were close to 20. For depths of 18-20 m, a layer of low-plastic clay was
encountered. The soil at these depths was much stiffer, and an N-value of 50 was observed.
Downloaded from ascelibrary.org by Pongpipat Anantanasakul on 06/19/18. Copyright ASCE. For personal use only; all rights reserved.
© ASCE
IFCEE 2018
IFCEE 2018 GSP 294 324
500 5.17
12.0-13.0 CH 19.4-19.5 26-28 59-60 35-37 >100 - - -
(12.25m) (12.25m)
13.0-18.0 CH - 23-26 - - 50-60 33-42 - - - 22-35
A preliminary analysis of piles subject to lateral loads was performed using the
commercial software LPILE. The Matlock models for soft clay (Matlock 1970) and for
stiff clay (Reese et al. 1975) were employed in the analysis. Concrete piles of various cross-
sections and lengths were analyzed. The numerical results suggested that the test pile would
yield in bending at an applied horizontal load of 11 ton, and the use of four square concrete
reaction piles 0.3 m wide and 13 m long, each with a safe lateral load of 3 ton was found to be
optimum for the field load tests.
The test and reaction piles were driven using a 5.2 ton drop hammer. All piles were
installed to a depth of 12 m. As shown in Fig. 2a, seven reaction piles were employed. They were
installed in two perpendicular rows, each consisting of four piles. The center-to-center spacing
among the concrete piles was 0.7 m, thus rendering a width of 2.4 m for the pile group. The test
piles and corresponding reaction piles were located 2.2 m apart to accommodate a horizontal
load-providing assemblage. This assemblage consisted of a hydraulic actuator, load cell and
transfer beams. Its arrangement is shown in Fig. 2b. A pin support was used at the connection
between the test pile and the load providing assemblage. As such, no moment was transferred
among these structural elements.
The load cell was employed to monitor the lateral loads applied to the pile heads at 0.25
m above the ground surface. The head displacements of piles were tracked by three LVDTs
attached at different heights (Fig. 2b). The output signals from the strain gauges, load cell and
LVDTs were recorded and processed using a dedicated data acquisition computer.
An inclinometer was also utilized to monitor pile lateral displacements. Inclinometer
casings 12 m long were attached to the insides of test piles using welded U-bolts. High-strength
epoxy was used to fix the casings and bolts together, thus preventing slippage of the casings
during pile installation and testing. Inclinometer readings were recorded manually every 0.5m for
the entire depth of test piles.
© ASCE
IFCEE 2018
IFCEE 2018 GSP 294 325
Downloaded from ascelibrary.org by Pongpipat Anantanasakul on 06/19/18. Copyright ASCE. For personal use only; all rights reserved.
(a) (b)
Figure 2. Arrangement of test and reaction piles (a) and schematic of lateral load tests (b).
TESTING PROGRAM
Set periods of 30, 51 and 72 days were provided for the three load tests, respectively, after the
piles were installed. The set periods were to allow the deep foundations to gain more realistic
load carrying capacities as the excess pore pressures in the foundation soils, which were created
during pile driving, dissipated with time. Horizontal loads were applied in steps with a load
increment of 1 ton. In each step, the hydraulic actuator was extended at rates close to 1 mm/min
until the target load was achieved. This target rate was identical to that used to determine the
undrained shear strengths of soils from the laboratory UU and UC tests. The tests were
performed to maximum loads of 10 to 11 ton where the hydraulic actuator reached its
displacement limit of 150 mm. This protocol for load application complied with the
specifications of ASTM D3966-07 standard. The lateral load was held constant for 15 minutes
for each step to allow the interaction between pile and foundation soils to reach equilibrium.
Observations made during the course of testing showed no significant change of pile
displacements after this hold period. Inclinometer readings were obtained and used to cross-
check the pile displacements determined from strain gauge results. The field testing conditions
for all tests were carefully kept as identical as possible so the repeatability of experimental
results could be investigated.
RESULTS OF FIELD LOAD TESTS
Analysis of Strain Gauge Results
The strain gauge results were analyzed within the framework of elastic beam on foundation. The
pile under consideration is assumed to be elastic with a constant value of flexural stiffness ( ).
The curvature ( ) is expressed in form of ⁄ where is the longitudinal strain on the
surface of pile and is the pile diameter. The bending moment is computed as . The
shear force ( ) and perpendicular external load or soil reaction ( ) in this application can be
determined by differentiating the bending moment once and twice, i.e.; and
. The pile slope ( ) and deflection ( ), on the other hand, are computed
© ASCE
IFCEE 2018
IFCEE 2018 GSP 294 326
∫ ∑( ) (1)
Downloaded from ascelibrary.org by Pongpipat Anantanasakul on 06/19/18. Copyright ASCE. For personal use only; all rights reserved.
where is the number of strain gauge elevations. The shear forces and soil reactions were
calculated from the bending moments using a two-point finite difference formula:
(2)
It was observed that the experimental maximum bending moments in piles were less than
the yield bending moment of 210 kN-m for lateral loads of 10 ton and smaller. For such
magnitudes of , the relationship between the pile bending moment ( ) and curvature ( ) is
linear; the value of is constant and the calculations of and using the presented
numerical differentiation and integration were performed accordingly.
© ASCE
IFCEE 2018
IFCEE 2018 GSP 294 327
in set periods did not result in any significant change in the load-displacement characteristics of
test piles. This finding supports the testing assumption that the excess pore pressures in the
foundation soils created during pile driving were largely and sufficiently dissipated within the
initial set period of 30 days.
The experimental curves are compared to the numerical predictions of LPILE. The
model for actual foundation soils consisted of eight soft clay layers underlain by one stiff clay
layer. The Matlock y models for soft clay and for stiff clay were used in the LPILE analysis.
It can be observed that LPILE significantly under-predicts the experimental pile displacements at
similar values of . The difference between numerical and experimental results increases with
the magnitude of .
Figure 4. Bending moments, pile deflections and soil reactions for lateral loads of 4,
7 and 10 ton. Numerical predictions of LPILE are plotted for comparison.
Pile Deflection Behavior
The variations of pile bending moment, deflection and soil reaction of load steps of 4, 7 and 10
ton for the first load test are plotted versus depth in Fig. 4. It can be observed that the maximum
bending moments increase and take place at slightly greater depths with increasing . The
locations of maximum bending moments are within a depth of 3.5 m or 13B from the ground
surface. This depth is greater than the upper limit of 8-10B usually reported in the literature
(Ducan et al. 1994). The maximum bending moment of pile for a lateral load of 10 ton is 185
kPa, and this is smaller than the yield bending moment of 210 kPa. This observation indicates
that the pile responses under such lateral loads were indeed elastic. Near the ground surface, the
© ASCE
IFCEE 2018
IFCEE 2018 GSP 294 328
test pile deflects in the direction of lateral loads. The trend, however, reverses and the pile
deflects in the opposite direction at greater depths. The magnitude of pile deflection, not
surprisingly, increases with . The variation of soil reaction with depth generally follows the
trend of bending moment. The locations of maximum soil reactions, however, are at slightly
greater depths.
The numerical predictions of LPILE are also plotted for comparison. It can be observed
that LPILE significantly under-predicts the pile deflections and, thus, predicts unrealistically
Downloaded from ascelibrary.org by Pongpipat Anantanasakul on 06/19/18. Copyright ASCE. For personal use only; all rights reserved.
small bending moments. The deviation between the experimental results and model predictions
generally increases with the magnitude of . The magnitudes of maximum soil reactions for
7 ton can be matched by the numerical results. At greater values of , however, the soil reactions
significantly increase, particularly at pile depths of 10-14B, and LPILE under-predicts the
magnitudes of for this portion of piles (Fig 4c). The predicted locations of maximum soil
reactions are about 6B higher than the actual locations.
© ASCE
IFCEE 2018
IFCEE 2018 GSP 294 329
© ASCE
IFCEE 2018
IFCEE 2018 GSP 294 330
displacements, pile deflections and maximum bending moments were under-estimated. This was
likely because the Matlock model assumed an unrealistic failure mechanism of the soil
surrounding the top portion of pile. This resulted in a prediction of excessive ultimate unit soil
resistance and the stiff relationships. Better predictions of the experimental load-
deflection behavior by LPILE may be achieved by employing values of the input shear strength
65% of the values of actual undrained shear strength with the Matlock y model for soft clay.
REFERENCES
Downloaded from ascelibrary.org by Pongpipat Anantanasakul on 06/19/18. Copyright ASCE. For personal use only; all rights reserved.
ATSM, A 53. “Standard Specification for Pipe, Steel, Black and Hot-Dipped, Zinc-Coated,
Welded and Seamless.” American Society for Testing and Materials, West
Conshohocken, Pennsylvania.
ATSM, D 3966-07. “Standard Test Methods for Deep Foundation Under Lateral Load.”
American Society for Testing and Materials, West Conshohocken, Pennsylvania.
Broms, B. B. (1965). “Design of laterally loaded piles.” ASCE Journal of the Soil Mechanics
and Foundations Division, 91(3), 79-99.
Brown, D. A., Shie, C. F., and Kumar, M. (1989). “p-y curves for laterally loaded piles derived
from three dimensional finite element model.” Proc., 2nd Int. Symp., Numerical Models
in Geomechanics, Niagra Falls, Canada, New York, Elsevier Applied Sciences, 683-690
Cox, W. R., Reese, L. C. and Grubbs, B. R. (1974). “Field testing of laterally loaded piles in
sand.”Proceeding of Offshore Technology Conference, Houston, Texas, Paper No. OTC
2079, 459-472.
Duncan, L. T., Evans, and Ooi, P. O. (1994). “Lateral load analysis of single piles and drilled
shafts.” ASCE Journal of Geotechnical Engineering, 120(6), 1018-1033.
Georgiadis, K. and Georgiadis, M. (2012). “Development of p-y curves for undrained response
of pile near slopes.” Journal of Computer and Geotechnics, 40(1), 53-61.
Hetenyi, M. (1946). Beams on Elastic Foundation. University of Michigan Press., Ann Arbor,
Michigan.
Matlock, H. (1970). “Correlations for design of laterally loaded piles in soft clay.” Proceeding of
the 2nd Annual Offshore Technology Conference, Houston, Texas, Paper No. OTC 1204,
577-594.
Reese, L. C., Cox, W. R. and Koch, F. D. (1975). “Field testing and analysis of laterally loaded
piles in stiff clay.” Proceeding of the 7th Annual Offshore Technology Conference,
Houston, Texas, Paper No. OTC 2132, 672-690.
Reese, L. C., Wang S. T., Isenhower, W. M. (2004). Computer Program LPILE Plus Version 5
Technical Manual, Ensoft, Inc, Austin, Texas.
Trochanis, A. M., Bielak, J. and Christiano, P. (1991). “Three-dimensional nonlinear study of
piles.” ASCE Journal of Geotechnical Engineering, 117(3), 429-447.
Welch, R. C. and Reese, L. C. (1972). “Laterally loaded behavior of drilled shafts.” Research
Report No. 3-5-65-89, Center for Highway Research, University of Texas at Austin.
© ASCE