Professional Documents
Culture Documents
Introduction
The focus of this paper is to investigate the material, which has come to be known as
Reese’s Stiff Clay below Water Table. Originally, this type of soil was a major concern
to the oil industry since several off shore structures encounter similar conditions on the
ocean floor. Therefore, studies and tests were conducted in order to understand the
properties of the material and come up with a numerical model that could be useful for
analysis. The tests consisted of both static and cyclic load case. The cyclic case was
intended to represent wave action encountered by offshore structures.
This paper will summarize the tests conducted and their results, as well as, discuss the
numerical model Reese created – how it compares to field test results and the practicality
of its construction. Lastly, the computer program Florida Pier, and its incorporation of
Reese’s model for stiff clay below the water table, will be examined.
Test Summary
The test involved two nominally 24” diameter piles and one nominal 6” diameter pile,
driven into stiff clay and subjected to lateral loading under both static and cyclic loading.
The surface soils consisted of stiff, preconsolidated clays of marine origins, while the
water table for the tests was above the ground surface by a few inches to resemble
conditions for an offshore platform. The results obtained from loading were then
analyzed to obtain the corresponding p-y curves from which a procedure was used to
predict p-y curves for stiff clays under the water table.
Testing Procedure
A pit 45ft wide by 50ft long by 3ft deep was excavated as shown in Figure 1. This pit
was then flooded with water to ensure saturation of the near surface clays to emulate
actual conditions of clay on the ocean floor. This flooding was done 5 months before
installation of the piles and 6 months before the actual testing. After borings were
conducted, an additional 2.5 foot was excavated and after installation of the piles, the top
6” of the soil was removed. This meant that the bottom of the pit was around 6ft below
the original ground surface. Borings logs, triaxial tests and stress-strain curves of the
soils samples (refer to Figure 2) taken were conducted and prepared to obtain the soil
properties and parameters to be used in the analysis.
The 24” diameter piles had a total length of 60ft and had two 5/8” semicircular thick
wrappers installed around the test pile by circumferential welds at the flange and at the
bottom of the wrapper section in addition to two longitudinal welds, since the existing
3/8” wall thickness of the piles was not sufficient for installation in stiff clay. It was
driven using a Delmag D-12 diesel hammer.
The 6” diameter pile consisted of an upper 28 foot long instrumented section and a 15
foot long un-instrumented lower section. They had wall thicknesses of 0.718” and 0.475”
respectively. The instrumentation on all of the piles included stain gages.
The two 24” piles were subjected to a total of six loading series each using a hydraulic
ram and a load cell in series (refer to Figure 4). Three of the loading series were to
relieve stress concentrations due to the addition of the 5/8” wrappers while the other three
were for the purpose of recording data. A maximum load of 50,000 lb was applied which
corresponded to a maximum stress of 15,000 psi.
The 6” pile was subjected to a total twelve loadings (six with a free head and six with a
rotational restraint at the top of represent rotational restraint for piles in jackets of
offshore platforms) (refer to Figure 5). The maximum load applied was 5000 lb which
corresponded to a maximum stress of 20,000 psi.
Fig. 1 Test layout
d4y
EI =p (1)
dx 4
p = − Es y (2)
To obtain the parameters that would be used for these equations and to obtain the p-y
curves, and knowing the boundary conditions at the top and bottom of the pile, the
experimental bending moment curves are used to obtain p and y using the following
equations:
M ( x)
y= (3)
EI
d2
p= M ( x) (4)
dx 2
Equation (3) is solved numerically, while to solve Equation (4), they assumed that the
soil modulus for a particular moment curves could be described using the following
equation:
E s = kx n (5)
Solving these equations, p-y curves were developed as shown in Figures 6, 7, 8 and 9.
It is clear that there is redundancy in the equations e.g. y could be obtained from Equation
(1) and (3). This redundancy was utilized to obtain the best possible curve fitting.
Fig. 6 p-y Curves for 24” diameter Piles (Static Case)
pi
E si = (6)
yi
E si = kx (7)
Values of k were computed based on experimental results. For other values, Reese
proposed using the values suggested by Terzaghi for the static and cyclic cases, where the
values for the cyclic casces were 40% of the static cases.
By careful observation of the experimental p-y curves, it was clear that there is a well
defined value of the ultimate soil resistance that increases with depth. The theoretical
value of the ultimate soil resistance at depth H, based on the model utilizing a wedge of
clay moving up and out from a pile, could be calculated as shown in Equation 8:
where pc1 is the ultimate soil resistance at depth H, ca is the average undrained shear
strength, b is the diameter of the pile and γ’ is the submerged unit weight of the soil.
The theoretical value of the ultimate soil resistance assuming the soil fails by flowing
horizontally could be computed using Equation 9:
pc 2 = 11cb (9)
where c is the undrained shear strength of the clay at the depth for the p-y curve. The
values calculated based on theory from the last two equations was higher than the
experimental results. As a result, a constant A and B was computed as shown in Figure
2.12, to adjust the theoretical values to the actual values where A and B are the constants
for the static and cyclic cases respectively.
( pu ) s
A= (10)
pc
( pu )c
B= (11)
pc
where (pu)s and (pu)c are the ultimate resistances from the experimental results and pc is
the ultimate resistance from theory (the lowest of equations 8 and 9 above).
Regarding the rest of the p-y curve, εc which corresponds to a stress of 50% of the
ultimate stress, is used. yc is then defined as follows:
yc = ε c b (12)
This parameter is then used for the equation of the parabola shown in Figure 10, as
follows:
0.5
y
p = 0.5 pc (13)
yc
The parabolic shape of the p-y curve in Figure 10, is assumed to be from the
intersection of the straight line with the parabola corresponding to Equation 13. This
continues until Ayc, where A is as defined above. An offset is constructed at Ayc of the
following value:
1.25
y − Ayc
poffset = 0.055 pc (14)
Ayc
0.0625 pc
E ss = − (15)
yc
This straight line ends at a deflection of 18Ayc where it becomes a horizontal line.
For the cyclic case, the following definition is made:
y p = 4.1Ayc (16)
y − 0.5 y p
2.5
p = Bpc 1 − (17)
0.45 y p
Again, like the static case, this parabola begins at the intersection with the straight line
defined in the very beginning, but continues only until 0.6yp, where it changes to a
straight line with the following slope:
0.085 pc
E sc = − (18)
yc
Fig. 16 Comparison Between Measured and Computed Values for Deflection for Static
Loading (24” Diameter Piles)
Fig. 17 Comparison Between Measured and Computed Values of Deflection for 6”
Diameter Pile
Fig. 18 Comparison Between Measured and Computed Maximum Bending Moment for
Cyclic Loading (24” Diameter Pile)
Fig. 19 Comparison Between Measured and Computed Values of Deflection for Cyclic
Loading (24” Diameter Piles)
Fig. 20 Comparison Between Measured and Computed Bending Moment Curves for
Cyclic Loading
Examples
A set of p-y curves was constructed for stiff clay for a pile with a diameter of 0.61m.
The soil profile is shown in figure 21. The submerged unit weight of the soil was
assumed to be 7.9 KN/m3 for the entire depth.
Fig. 21 Soil profile used for example p-y curves for stiff clay
Florida Pier allows one to view the p-y curve for the modeled soil. Fig. 28 and Fig. 29
compare theoretical p-y curves of Reese’s model for stiff clay below the water table with
the curves generated in Florida Pier. Fig. 28 is constructed for the values above, while
Fig. 29 is computed for the same values with double the diameter of the pile.
Fig. 28
Fig. 29
Fig.30, 31, 32 compare the resultant displacements, soil resistance, and moments
respectively for the standard steel pile, double the diameter, and double the undrained
shear strength. These pile were subjected to a lateral force of 600 kN. Also included in
the figures, is the standard steel pile subjected to a lateral force of 44.5kN. From Fig 30
it is clear that despite the varying conditions, the overall behavior of the pile is the same.
The majority of the displacement takes place in the upper 25% of the pile length.
Doubling the Diameter and doubling the undrained shear strength of the soil result in
similar deflections, which expectantly are less than the standard pile.
Fig 31 shows that by doubling the diameter of the pile, the soil resistance was reduced.
This is due to the increased area of soil affected. Fig 32 shows that by doubling the
undrained shear strength; the moment induced in the pile is significantly reduced. It is
not clear why this occurs; perhaps the stiffer soil limits the pile’s curvature and therefore
reduces the induced moment.
Displacement
meters
-1.00E-02 0.00E+00 1.00E-02 2.00E-02 3.00E-02 4.00E-02
0
0.2
0.4
45kN
600kN
0.6
x/l
Dbl. Dia.
Dbl. Su
0.8
1.2
Fig. 30
Soil Resistance
kN
0.2
0.4
45kN
600kN
0.6
x/l
Dbl. Dia.
Dbl. Su
0.8
1.2
Fig. 31
Moment
kN*m
-5.00E+01 5.00E+01 1.50E+02 2.50E+02 3.50E+02 4.50E+02 5.50E+02 6.50E+02
0
0.2
0.4
45kN
0.6 600kN
x/l
Dbl. Dia.
Dbl. Su
0.8
1.2
Fig .32
Fig. 33 compares the displacements of the standard steel pile (12 m) with the same pile of
a length equal to 3 m. This causes the deflection of the pile to be distributed throughout
the entire length of the pile rather than the top 25%.
Displacement
meters
-1.00E-02 0.00E+00 1.00E-02 2.00E-02 3.00E-02 4.00E-02
0
0.2
0.4
3m
0.6
x/l
12m
0.8
1.2
Fig. 33
Fig. 34 compares the standard steel pile to the pre-stress pile. Both piles are loaded with
the largest lateral load Florida Pier will converge for. Assuming that Florida Pier fails to
converge when the soil fails, these piles are loaded to the ultimate capacity of the soil
(Steel = 600kN, Pre-Stress = 400kN). Therefore, Fig. 34 shows that the standard steel
pile is more ductile than the pre-stress pile and the soil is able to accommodate larger
deflections for the steel pile.
Displacement
meters
-1.00E-02 0.00E+00 1.00E-02 2.00E-02 3.00E-02 4.00E-02
0
0.2
0.4
Pre-Stress
0.6
x/l
Steel
0.8
1.2
Fig. 34
5.0 2x2 Pile Group
A 2x2 pile group was studied using FB-Pier for both the fixed and pinned head cases.
The following data was used in the analysis which was the same as the data used for the
single pile case:
Pile = 0.61m steel circular tube (12m long)
Pile Cap Thickness = 1.5m
c = 200kPa
γ = 11 kN/m3
Ks = 135 MN/m3
E(50) = 0.005
ca = 250kPa
Loading applied = 1440 kN
The program utilized p-y multipliers which were 0.8 for the leading piles and 0.4 for the
trailing piles. Figure 5.1 below shows a sketch of the 2x2 pile group.
Pile 3 Pile 4
1440kN
Pile 1 Pile 2
Fixed Head
Length Piles 1&3 Piles 2&4 Piles 1&3 Piles 2&4 Piles 1&3 Piles 2&4
Length Piles 1&3 Piles 2&4 Piles 1&3 Piles 2&4 Piles 1&3 Piles 2&4
The data above were then plotted and Figures 5.2, 5.3 and 5.4 were obtained as shown
below.
Pile Displacement vs. Normalized Length Along Pile for 2x2 Pile Group
Pile Displacement (m)
-2.00E-03 0.00E+00 2.00E-03 4.00E-03 6.00E-03 8.00E-03 1.00E-02 1.20E-02 1.40E-02 1.60E-02
0
0.1
0.2
0.4
0.7
0.8
0.9
0.1
0.2
0.3
Normalized Length Along Pile
0.4
Fixed Piles 1&3
0.5 Fixed Piles 2&4
Pinned Piles 1&3
Pinned Piles 2&4
0.6
0.7
0.8
0.9
0.1
0.2
0.3
Normalized Length Along Pile
0.4
0.7
0.8
0.9