Artículo EA1
Artículo EA1
Artículo EA1
https://doi.org/10.1007/s40098-024-00942-5
ORIGINAL PAPER
13
Vol.:(0123456789)
Indian Geotech J
stiffness. Cook et al. [13] demonstrated that the operative readings, using a database for saturated clay, silt and sand
shear modulus (G) is the most critical factor in determining from well-documented experimental sites [3]. The following
the Q–w performance of a loaded pile. In the pile Q–w anal- two correlations have been proposed for soils dominated by
ysis, the shear modulus (G) is preferred over the Young’s sandy silt. Trevor et al. [4] proposed a correlation for sandy
modulus, E, for two reasons: (1) the shear modulus is often silt soils using corrected cone resistance qt and fs values.
unaffected by whether the loading conditions are undrained The correlation proposed by Holmsgaard et al. [2] uses the
or drained and (2) the primary deformation of the soil occurs value of qt and is suitable for sandy silt with clay stripes.
in shear along the pile shaft. The small strain shear modulus In addition, this paper proposes three correlations for silty
(Gmax or G0) is generally calculated based on the measured soils. Four correlations from previous studies are presented
shear wave velocity (Vs). in Table 1.
It is crucial to comprehend the potential presented by
using SCPTu data for the analysis of a pile. This study used
analytical elastic continuum solutions to achieve a deeper Use of SCPTu Data to Predict the Load–
understanding of the correlations of the test data. Four Displacement Capacity of a Pile
known empirical correlations between Vs and CPTu were
analysed, and three new correlations for silty soils were pro- SCPTu was used to determine the penetrometer readings and
posed. An elastic solution proposed by Randolph and Wroth in situ shear waves (S). According to shear wave theory, Vs
[14, 15] was applied to determine the Q–w relationship of can be measured. For this purpose, a geophone is integrated
piles based on Niazi and Mayne’s [11] methodology. The into the cone. Gmax or G0 is calculated based on the Vs value.
exponent parameter g [16] for the screw piles in silty soils Most commonly, shear waves are measured at every metre
was analysed. Correlations between the normalised opera- in the borehole with the down-hole SCPTu. For this reason,
tive shear stiffness (G/Gmax) and normalised pseudo-strain both Vs readings and G0 values represent the average value
(γp/γp-ref) were developed based on the back-calculations of of a one-metre interval.
the static pile load test results. In normalised pseudo-strain, The initial value of soil stiffness (Gmax or G0) gradually
γp = wt/d and γp-ref = 0.01, where wt represents the settlement decreases to G as the strain increases. The decrease in the
at the pile top and d is the pile diameter. The results of the modulus can be described with a modified hyperbola. Fahey
back-calculations of the static pile load tests were compared and Carter [16] proposed a modified hyperbola form for the
with those obtained by Niazi and Mayne [11]. Because the monotonic torsional shearing of normally consolidated sands
piles were not loaded to their ultimate capacity during the (1):
static pile load test, extrapolation was used. Since most of ( )g
the tested piles had significant settlements, potential biases G
=1−f
𝜏
. (1)
resulting from these extrapolations were found to have no Gmax 𝜏max
significant effect on the results.
Here, f and g are fitting parameters, and G is a shear mod-
ulus calculated as G = E/[2(1 + ν)], where ν` = 0.2 (drained
case) is the approximate value of the Poisson’s ratio of geo-
CPT– Vs Correlations
materials at small strains. The value of τ/τmax can be treated
as the reciprocal of the factor of safety (FS) or 1/FS. There-
Four empirical correlations that connected CPTu readings
fore, it can also be considered a level of mobilised load, Q/
with Vs (m/s) were included in the analyses. The first cor-
Qult. At small strains, the material’s stiffness is finite and can
relation uses a database that includes sand, silt and clay, as
be indicated by the low-strain shear modulus:
well as mixed soils, as derived by Hegazy and Mayne [1].
It uses both cone (qc) and shaft resistance (fs). The second Gmax = G0 = 𝜌Vs2 , (2)
correlation equation derives the value of Vs only from the fs
qt, cone resistance corrected for the pore pressure effects; qc, cone resistance; fs, sleeve friction; and Vs,
shear wave velocity
13
Indian Geotech J
where ρ indicates the total mass density of the material. As differences in pile types and pile installation methods. For an
the strains of G0 are too small to cause excess pore water accurate analysis of different pile types and their load–dis-
pressure, Eq. 2 applies to both drained and undrained condi- placement relationships, the solution must first be calibrated
tions [17]. In addition, previous studies have shown that the with the static pile load test data. An important variable in
value of G0 is the same under both static (monotonic) and this analysis is the parameter g in Eq. 1, which is assigned
dynamic loading conditions [18, 19]. values of 0.2, 0.3 and 0.4 in the analysis.
The first-order evaluation values of f = 1 and g = 0.3 pro- To determine the value of G0, it is essential to know the
vide normally reasonable approximations and have been total mass density ρ values. The total mass density of the
confirmed by laboratory torsional, triaxial and simple shear soils can be determined in the laboratory. However, collect-
tests for the monotonic loading of unaged and uncemented ing soil samples from deep layers below the water table is
quartz sands and insensitive and unsaturated clays [20, 21]. a complex and time-consuming process. Furthermore, it is
The exponent parameter, g ≈ 0.3 ± 0.1, is suitable for many practically impossible to obtain undisturbed samples from
uncemented and nonstructured soils [22, 23]. In our analysis, silty soils below the water table. Thus, an expression that
the values of f = 1, g = 0.2, g = 0.3 and g = 0.4 were adopted uses all three CPTu readings (fs, qt and u2) is included to
in Eq. 1. determine the soil unit weight [24]. The mass density, ρ, can
In addition to the load-bearing capacity of a pile, it is be obtained from the unit weight as ρ = γt / 9.8. Equation (4)
necessary to determine the Q–w relationship to design an is based on a variety of clays, silts and sands.
economical foundation solution. A suitable approach is to ( ) ( )
use the approximate analytical elastic solution for pile–soil 𝛾t = 11.46 + 0.33 ⋅ log(z) + 3.1 ⋅ log fs + 0.7 ⋅ log qt .
interaction, provided by Randolph and Wroth [14, 15]. This (4)
solution accounts for piles in Gibson-type soils (i.e. linearly
increasing soil stiffness with depth) and in homogenous two-
layered soils (i.e. constant soil stiffness with depth). The Test Sites and Tested Piles
settlement of pile top (wt) for a compressible pile can be
calculated as follows [14, 15]: All three test sites are located in Tallinn, northern Estonia,
[ ] and are found in old valleys buried under quaternary sedi-
4η tanh
Qt 1 + 𝜋𝜆 1−𝜈 (𝜇L)L ments. Marine, lacustrine and alluvial deposits incorporate
( s )𝜉(𝜇L)r0
wt = [ ], (3) sandy silts, silty sands, clay and sand at the Ahtri site and
GL r0 1−ν 𝜉 + 2𝜋𝜌E𝜁 (𝜇L)r
4𝜂 tanh (𝜇L)L
silty sand, sandy clay, clay and sand at the Paldiski mnt site
( s) 0
[25]. At the Soodi site, the soil deposits comprise varying
where wt represents the settlement at the pile top; Qt is the layers of silty clay, silty sand, clay and sand [26].
load applied at the pile top; GL is the operative soil shear At all three sites, static loading tests were conducted. At
modulus at the pile base; η is the factor for underreamed the Ahtri site, one Fundex pile (A3) was tested; at the Pald-
piles that take greater loads at the pile base, calculated as iski mnt site, two Fundex piles (P-1 and P-3) were tested [25]
η = rb / ro; L is the pile length; ro is the radius of the pile and at the Soodi site, two displacement piles (S-1 and S-2)
shaft; rb is the pile base radius for underreamed piles; μL is and two Bauer full displacement piles (S-3 and S-4) were
the measure of pile compressibility, calculated as μL = 2·[2/ tested [26]. All piles were tested in accordance with EVS-EN
(ζ λ)]0.5·(L/ds); ζ is the measure of influence radius, calcu- 1997–1:2006 based on EN 1997–1:2004 [27]. Because the
lated as ζ = ln(rm/ro); rm is the maximum radius of influence, piles were not loaded to their ultimate capacity, Chin’s [28]
calculated as rm = L{0.25 + ξ·[2.5ρE·(1—νs) -0.25]}; λ is the extrapolation was used to complete the Q–w curves. This
pile–soil stiffness ratio, calculated as Ep/GL; Ep is the pile method was chosen because it was also used by Maertens
modulus; ξ is the factor for end-bearing piles resting on a and Huybrecht [29] in silty soils with the same type of piles.
stiffer stratum (Gb > > GL), calculated as GL/Gb; Gb is the In a previous study [25] using Fundex piles, Chin’s method
soil shear modulus below the pile base; ρE is the modulus also demonstrated favourable results in silty soils. Therefore,
variation factor, calculated as GM/GL; GM is the operative it was assessed as the most reliable method. Table 2 shows
soil shear modulus at the midpoint of the pile embedment the pile types and lengths, with the maximum testing loads
depth, calculated as GM = (Go + GL)/2; Go is the operative and the respective settlements. The diameter specifications
shear modulus at the pile top (Z = 0); and νs is the Poisson of Fundex piles in Table 2 are denoted by two numerical
ratio of soil. values. The first value corresponds to the diameter of the
The solution is suitable for both constant and linearly shaft, while the second value represents the diameter of the
increasing soil stiffness with depth (Gibson-type model) pile tip. This particular study focused on Fundex piles with
conditions; it incorporates end-bearing and floating piles. two different diameters, where the pile tip diameters were
However, the solution does not directly account for the measured at 0.45 and 0.56 m. The respective diameters of
13
Indian Geotech J
Pile type Fundex 450/560 Fundex 350/450 Fundex 350/450 DSP 406/560 DSP 406/560 FDP 440 FDP440
Pile length (m) 23.2 15.5 15.0 12.69 11.34 12.39 12.5
Measured max 3600 1200 1200 1870 1700 1870 1870
load from pile
load test (kN)
Max settlement 22.0 6.3 15.0 35.3 22.8 17.0 22.7
from pile load
test (mm)
the pile shaft were 0.35 and 0.45 m. Notably, both the FDP presented with in situ pore pressure (u0) profiles. The types
pile base and shaft had a consistent diameter of 0.44 m. of soil layers, presented on the left of the sounding figures,
were defined using CPTu and the soil behaviour type (Rob-
ertson 2010 [30]. At all three sites, silty soils predominated.
SCPTu Soundings The Geotech AB Nova cone, mounted on a lightweight
truck, was used to perform all soundings. To ensure a higher
Six SCPTu soundings were conducted at all three sites. At compression force, the truck was secured in the soil using
the Ahtri site, three SCPTu soundings, with depths rang- two screw anchors. To penetrate the covering fill layers,
ing between 20.0 and 23.5 m, were conducted (Fig. 1). At predrilling was conducted. The Nova cone, in accordance
the Paldiski mnt site, two SCPTu soundings, with depths with ASTM D-5778 [31] and EN ISO 22476–1:2012 [32],
ranging between 19.0 and 19.8 m, were conducted (Fig. 2). had a project area of 1000 mm2 and a sleeve surface area of
At the Soodi site, one SCPTu sounding, with a depth of 15,000 mm2.
20.0 m, was performed (Fig. 3). The corrected cone resist- The shear wave velocity was determined through an
ance (qt) and sleeve friction (fs) values were reported for all SCPTu down-hole test, where the energy source was posi-
soundings. At all sounding points, the ground water level tioned on the ground and the receiver was placed in the
was at a depth of almost one metre above the ground surface. cone. Seismic tests were conducted at one-metre inter-
Pore water pressure was measured behind the cone (u2) and vals throughout the borehole, resulting in consistent V s
Soil descripon
Drill out
Clean sand to silty sand
Fig. 1 Results of the piezocone test at the Ahtri site. qt, cone resistance corrected for the pore pressure effects; fs, sleeve friction; u2, pore pres-
sure; and Vs, shear wave velocity
13
Indian Geotech J
Soil descripon
Drill out
Sand & silty sand
Fig. 2 Soil description and the results of SCPTu tests P1 (blue figures) and P2 (red figures) at the Paldiski mnt site. qt, cone resistance corrected
for the pore pressure effects; fs, sleeve friction; u2, pore pressure; and Vs, shear wave velocity
Soil descripon
Drill out
Fig. 3 Results of piezocone tests at the Soodi site. qt, cone resistance corrected for the pore pressure effects; fs, sleeve friction; u2, pore pressure;
and Vs, shear wave velocity
and Gmax values (refer to Figs. 1, 2, 3). Table 3 displays Additionally, the table presents the difference between
the minimum and maximum V s values obtained from the maximum and minimum Vs values (ΔVs). As for the qt
all six soundings, which ranged from 37 to 352 m/s. and fs values, measurements were taken every 20 mm, and
13
Indian Geotech J
Table 3 Summary of Vs values Vs -min Vs -max Δ Vs qt-min qt-max Δqt fs -min fs -max Δfs
alongside the minimum and
maximum results of average m/s m/s m/s MPa MPa MPa kPa kPa kPa
qt and fs values for 1-m-thick
SCPTu-A1 46 259 213 0.5 4.3 3.7 7.9 76.9 69.0
layers
SCPTu-A3 44 203 159 0.5 8.0 7.5 3.4 107.5 104.1
SCPTu-A4 58 352 294 0.6 4.8 4.2 2.3 91.1 88.8
SCPTu-P1 37 233 196 0.4 17.7 17.3 0.6 157.3 156.6
SCPTu-P2 57 270 213 0.3 25.6 25.3 2.8 144.4 141.5
SCPTu-S1 76 248 172 0.4 18.5 18.1 0.5 137.1 136.6
the average values of one-metre-thick layers were used Estimated Versus Measured Vs Values
for the analysis. Table 3 also includes the minimum and
maximum qt and fs values for all six soundings. Alongside the measured Vs results, the calculated values
The measured Vs values were analysed to improve the are based on four empirical regression equations from
database quality by removing outliers based on statistical the literature for all six SCPTu soundings, as presented in
considerations. Outliers were identified through the ‘2σ’ Figs. 4 and 5. The details of these correlations are presented
statistical criteria, where σ is the standard deviation of in Table 1. Because the existing correlations did not pro-
the variable Vs. This resulted in 95% confidence interval vide overwhelming fitness to the dataset used in this study,
criteria for the data. Two Vs values (37 and 352 m/s) out three new correlations were introduced. The first relation-
of 108 were removed from the entire database and not ship (5) proposed in this work uses fs values, and the second
included in the analysis. (6) uses qc and u2 values. Both correlations can be used in
parallel to compare the calculated Vs values. In addition, a
correlation is proposed that exploits both qt and fs values
simultaneously. The values of Vs calculated using Eqs. 5–7
are also included in Figs. 4 and 5. The Vs values obtained
Fig. 4 Comparison of Vs as predicted from CPT–Vs correlations and as measured by SCPTu at the Ahtri site
13
Indian Geotech J
Fig. 5 Comparison of Vs as predicted from CPT–Vs correlations and measured using SCPTu at the Paldiski and Soodi sites
from the correlations are calculated from CPTu readings, offered the best results for the Paldiski and Soodi sites. Of
measured at intervals of 20 mm and rounded to 1-m-thick the existing functions, the correlation proposed by Tre-
layers. The rounded layer interval is the same as that in the vor et al. [4] showed the best results based on Paldiski
Vs measurements. and Soodi site data, according to R2 and the best-fit line.
The correlation offered the R2 value of 0.70. Based on
Vs = 95.7 ⋅ fs0.155 , (5) the R2 and best-fit lines, the correlations presented in this
paper provide equivalent results for the Paldiski and Soodi
Vs = 128.4 ⋅ q0.169 , (6) sites. The R 2 value of function (5) is 0.55, function (6)
t
is 0.69 and function (7) is 0.61. However, based on the
results from all three sites, all three correlations proposed
Vs = 103.9 ⋅ q0.058
t
⋅ fs0.107 . (7) in this paper offer the best results. Functions (5) and (7)
also obtain the highest R2 value of 0.33 for the data from
all three sites. Moreover, functions (6) and (7) provide the
To test the performance of existing and newly derived
lowest COV value of 0.28 for the data from all three sites.
functions, the computed Vs values are plotted against the
The lowest correlation occurred at the Ahtri site, where the
measured Vs values in Fig. 6. Regression analyses were
R2 values of all seven correlations were between 0.10 and
performed on each set of results to achieve the best-fit
0.17. Although the results show great scatter, the correla-
line of the computed/measured ratios of Vs. The linear
tion proposed in this paper offers close results compared
regression function and corresponding coefficient of deter-
to the measured V s values at all three sites, as shown in
mination (R2) were then obtained. The closer the best-fit
Fig. 4. A significant difference occurred only in CPTu-A1,
line to the perfect-fit line, the higher the R2 value and the
CPTu-A3 and CPTu-A4 soundings deeper than 14 m at
better the correlation function. The R 2 values are pre-
the Ahtri site. The correlations proposed by Hegazy and
sented in Table 4, and the equations of the best-fit lines
Mayne [1] and Mayne [3] showed significant flaps at all
are presented in Table 5. Because the results of the Ahtri
three study points. However, in the remaining third of the
site showed the lowest correlation, the data of this site
study points, these correlations did not show a good flap.
were analysed separately; the data from the Paldiski and
The correlation proposed by Holmsgaard et al. [2] showed
Soodi sites were analysed together. All seven correlations
a significant flap with only a few study points.
13
Indian Geotech J
Fig. 6 Evaluation of correlations between computed and measured shear wave velocity Vs for silty soils
13
Indian Geotech J
Ahtri + Paldiski + Soodi 0.32 0.31 0.29 0.30 0.33 0.31 0.33
(n = 106)
Paldiski + Soodi (n = 48) 0.55 0.51 0.70 0.71 0.55 0.69 0.61
Ahtri (n = 58) 0.16 0.17 0.10 0.11 0.17 0.11 0.16
Table 5 Coefficients and exponents for (G/Gmax) Vs. normalised For most analysed piles, the extrapolated results closest
pseudo-strain (γp/γp-ref) formulation (Eq. 8) to the measured values were taken at g = 0.4. For the DSP
Pile classification α1 β1 piles (S-1 and S-2), the value g = 0.4 provided a satisfactory
alignment with the measured values. For FDP piles (S-3 and
(type/installation method) S-4), g = 0.4 tended to underestimate the bearing capacity
Bored cast in situ 1.912 0.97
of the pile. On two out of three Fundex piles (A-3, P-1 and
Auger 1.176 1.01
P-3), the Q–w curves exhibited an adequate alignment with
Driven 0.84 1.07
the g = 0.4 line. Therefore, it is reasonable to use g = 0.4 to
Jacked 0.65 1.25
derive the Fundex pile Q–w curve in silty soils.
Screw pile (current study) 0.305 1.68
Discussion
Estimated Versus Measured Load–Displacement
Capacity of the Pile In this study, the Vs values ranged between 37 and 352 m/s
(Table 3). Poulos [33] proposed that the typical value of
Using the Vs values measured based on soundings SCPTu- Vs for rough estimation is in the range of 85–105 m/s for
A4, SCPTu-S1, SCPTu-P1 and SCPTu-P2, the Gmax val- very soft soils and 276–365 m/s for very stiff soils. For silty
ues were calculated alongside the pile and beneath the soils, Holmsgaard et al. [2] recommended a Vs range of
pile base with Eq. 2. To determine the mass density of 150–250 m/s. Based on several previous studies, Hussien
the soil, CPTu readings were applied with Eq. 4. The soil and Karray [34] presented Vs values for sandy soils in the
layers adjacent to the pile were counted as a single soil range of 127–327 m/s for a stress state of K = 1.0. As shown
layer in which the increase in the value of Gmax is a func- in Figs. 1, 2, 3, Vs values of approximately 100 m/s and as
tion of depth [11]. Note that the depth-Gmax figures were small as 37 m/s were measured in sensitive fine-grained,
compiled from all four soundings and the best-fit line was silty sand and sandy silt layers, which were no deeper than
found (Fig. 7). The soil around the pile base was treated 10 m. In deeper layers, the measured Vs values ranging
as a second soil layer. Gmax values were found on the basis from 150 to 250 m/s agreed with the range proposed by
of the Vs values measured at the depth of the pile base. Holmsgaard et al. [2], with small exceptions. Additionally,
Equation 3 was applied to determine the Q–w relation- these measured values fell well within the range presented
ship of seven screw piles in silty soils. The values f = 1 by Hussien and Karray [34] and Poulos [33]. In the current
and g = 0.2, 0.3 and 0.4 were used in this analysis. The study, most measured Vs values in depths deeper than 10 m
pile types and dimensions are given in Table 2. Piles A-3, were approximately 200 m/s.
P-1 and P-3 were Fundex piles; piles S-1 and S-2 were For the analysis of CPTu and Vs correlations, the results
displacement (DSP) piles; and piles S-3 and S-4 were full obtained at the Ahtri site were treated separately. In the
displacement (FDP) piles. Ahtri site, the soil layers below 14–16 m depth appeared
Because static pile load tests were not carried to the to be more silty than in the two other sites. The results in
pile failure, the Q–w curves were extended using Chin’s Fig. 4 illustrate this as, at depths of 14–16 m, the corre-
[28] extrapolation. The Q–w curves of all seven piles are lations proposed in the present work interlace clearly for
presented in Fig. 8. Only the last two to three pile load test all three study points. Other investigated correlations pro-
values were extrapolated at the Soodi site; most of them vided satisfying flaps in the range of 14–16-m depths. In
were extrapolated from the pile load test of P-1. In piles the deeper layers, only the correlations from Hegazy and
A-3 and P-1, the last third of the Q–w curve values were Mayne [1] and Mayne [3] provided an approximate flap. In
extrapolated. deeper layers, the soils of complex silt and sand mixtures
13
Indian Geotech J
Fig. 7 Depth-Gmax figures for all four soundings with best-fit lines
rested at the Ahtri site. The beginning of the silt and sand As shown in Figs. 4 and 5, the Vs values resulting from
mixtures was well distinguished from the u2 profile (Fig. 1) the correlations provided by Hegazy and Mayne [1] and
between 13 and 16 m, where the value of pore water pressure Mayne [3] approximated each other at all study points. The
dropped sharply. For such soils, all correlations reviewed in correlation provided by Hegazy and Mayne [1] included all
this paper significantly underestimated the Vs values, apart CPTu readings, whereas the one provided by Mayne [3] used
from the correlations provided by Hegazy and Mayne [1] only fs values. In addition, Eq. 7 does not offer better results
and Mayne [3]. than Eqs. 5 and 6. Therefore, correlations with a single input
The correlations proposed using Eqs. 5, 6 and 7 offered value may provide good results. This is clear evidence that
the best flap (Fig. 5) for the Paldiski and Soodi sites. The correlations based on the qt or fs values should be used sepa-
correlation provided by Trevor et al. [4] also showed excel- rately. Through this action, it becomes feasible to compare
lent results. The other three correlations substantially over- correlation findings derived from separate analyses of the
estimated the Vs value in most soil layers at study points same CPTu soundings.
CPTu-P1 and CPTu-P3 at the Paldiski site. At study point Measuring the V s value at the site provides a good
CPTu-S1, the same three correlations significantly overesti- opportunity to derive the pile Q–w relationship based on
mated the Vs value in several layers and significantly under- Eqs. 1–3. One of the key constant in the analysis is the
estimated the Vs value in some layers. This clearly indicates exponent parameter g. In this study, the value of g = 0.4
that creating CPTu and Vs correlations for mixed soils is was found to be the most suitable for screw piles in silty
complicated. In mixed soils, universal correlations may not soils. Based on the results of the seven static pile load tests,
yield good results; hence, a correlation based on a specific back-calculations were performed, resulting in a correlation
soil is necessary. Such mixed soils need further investigation (8) between normalised operative shear stiffness (G/Gmax)
in the future. and normalised pseudo-strain (γp/𝛾 p-ref) with a coefficient
13
Indian Geotech J
Fig. 8 Estimated pile Q–w relationships found using Eq. 2 with g = 0.2, 0.3 and 0.4 for all seven piles. The results are presented alongside those
of the static pile load test curves. The last parts of the pile load test curves were extrapolated using Chin’s method
13
Indian Geotech J
(9)
A form of the correlation analysed in the current study
was presented by Niazi and Mayne [11], who analysed 299
pile load tests at 61 sites. In their study, a modified hyper-
bola was introduced based on the curves of Vardanega and Conclusions
Bolton [35]. Niazi and Mayne [11] proposed the following
correlation (9), based on stiffness reduction from 299 pile This study examines the potential of SCPTu in evaluating
load tests obtained through back-analysis. Coefficient α1 the load-bearing capacity of a pile in silty soils. Since one of
and exponent β1 are parameters that identify the pile type the key values of the entire study is Vs reading, the research
and installation methods, as presented in Table 5. Piles are also analyses the correlations between CPTu readings and
divided into four groups: driven, jacked, auger and bored Vs. This study analysed four correlations between CPTu and
cast in situ. The values of α1 and β1 obtained in the current Vs, as presented in the literature. These correlations were
study are presented in the last line of Table 5. These numbers denied to a variety of soils [1, 3], particularly silty soils [2,
differ significantly from the values proposed by Niazi and 4]. The best flaps for the selected correlations at all three
Mayne [11]. It is evident from Fig. 9 that the outcomes of sites were determined based on Trevor et al.’s [4] correla-
the Niazi and Mayne [11] correlations for various pile cat- tion. The other three correlations demonstrated significant
egories significantly deviate from the correlation suggested deviations from the measured results at several study points.
in this study for screw piles in silty soils. Screw piles exhibit Significant flaps for the Vs values for most soil layers were
the most gradual decrease in shear stiffness within the ini- found in all three site specific correlations proposed in the
tial range of per cent γp (< 0.3), becoming steeper for high paper. At the Ahtri site, the only approximating flaps with
values. Furthermore, it is evident from Fig. 9 that the screw complex silt and sand soil mixtures were found based on
pile exhibits superior load-bearing capacity in comparison to the correlations from Hegazy and Mayne [1] and Mayne
the other referenced pile types, particularly at small strains. [3]. This indicates the need to study more mixed soils to
This indicates the need for similar correlations to comprise produce suitable correlations. Correlations based on qt of
larger numbers of pile types in different soils. fs values should be preferred for comparisons of correla-
tion results obtained from independent readings of the same
CPTu sounding.
The shear wave velocity Vs values ranged from 37 to
352 m/s in the soils examined in this study. The minimum
Vs values and values around 100 m/s were measured in
sensitive fine-grained layers, as well as in silty sand and
sandy silt layers up to a depth of 10 m. Deeper than 10 m,
Vs values of approximately 200 m/s dominated. The range of
Vs values measured in this study mostly coincided with the
range proposed by Holmsgaard et al. [2] for silty soils and
Hussien and Karray [34] for sandy soils. Additionally, the
ranges of Vs values measured in this study generally agree
with those proposed by Poulos [33] for very soft soils and
very stiff soils.
The present study analysed the static load test of seven
screw piles in silty soils at three sites. Three Fundex piles,
two DSP piles and two FDP piles were included in the study.
Because the piles were not tested for their ultimate bear-
Fig. 9 Correlation between normalised operative shear stiffness (G/ ing capacity, Chin’s extrapolation was used to complete the
Gmax) and normalised pseudo-strain (γp/𝛾 p-ref), where γp = wt/d and Q–w curve. In addition, seven SCPTu soundings were con-
γp-ref = 0.01 ducted at the three sites. The Vs values obtained from the
13
Indian Geotech J
13
Indian Geotech J
19. Tatsuoka, F., Jardine, R.J., LoPresti, D.C.F., DiBenedetto, H., & 27. EN 1997–1:2004. Eurocode 7: Geotechnical Design—Part 1: Gen-
Kodaka, T. (1997). Theme lecture: characterizing the pre-failure eral Rules (p. 171), Brussels, Belgium, (2004).
deformation properties of geomaterials. In Proceeedings, 14th 28. Chin, F.K. (1970). Estimation of the ultimate load of piles from
International Conf. on Soil Mechanics & Foundation Engineer- test not carried to failure. In Proceedings, Second Southeast Asia
ing (Vol 4, p 35). Conference on Soil Engineering (pp 81–90), Singapore.
20. Burns SE, Mayne PW (1996) Small- and high-strain measure- 29. Maertens, J., & Huybrechts, N. (2003). Results of the static pile
ments of in situ soil properties using the seismic piezocone pen- load tests at the Limelette test site. In Proceedings, 2nd Sympo-
etrometer. Transp Res Rec 1548:81–88 sium on Screw Piles—Belgian Screw Pile Technology—Design
21. Mayne, P.W. (1995). Application of G/Gmax modulus degradation and Recent Developments (pp 167–214), Brussels, Balkema,
to foundation settlement analyses. In Proceedings, U.S.-Taiwan ISBN: 9058095789.
Workshop on Geotechnical Collaboration (pp 136–148), National 30. Robertson, P.K. (2010). Soil behaviour type from the CPT: An
Science Foundation/Washington D.C. and National Science update. In 2nd International Symposium on Cone Penetration Test-
Council/Taipei. ing, CPT’10, Huntington Beach, CA, USA. www.cpt10.com
22. Mayne, P.W. (2005). Integrated ground behaviour: In-situ and 31. American Society for Testing & Materials (2000) ASTM Book
laboratory tests. Deformation Characteristics of Geomaterials of Standards (vol. 4, Sections 08 and 09), Construction Materials.
(2) (155–177.C), Taylor & Francis, UK Soils Rocks, Philadelphia, PA.
23. Mayne, P.W. (2007). In-situ test calibrations for evaluating soil 32. EN ISO 22476–1:2012 Geotechnical investigation and testing.
parameters. In Characterization & Engineering Properties of Field testing. Part 1: Electrical cone and piezocone penetration
Natural Soils (Vol 3, pp 1602–1652), Taylor & Francis Group, test. ISO 2012.
London. 33. Poulos HG (2022) Use of shear wave velocity for foundation
24. Mayne, P.W., Peuchen, J., & Bouwmeester, D. (2010). Estimation design. Geotech Geol Eng 40:1921–1938. https://d oi.o rg/1 0.1 007/
of soil unit weight from CPTs. In Proceedings, 2nd International s10706-021-02000-w
Symposium on Cone Penetration Testing (CPT’10) (CA, Vol 2, pp 34. Hussien MN, Karray M (2016) Shear wave velocity as a geotech-
169-176), Huntington Beach. nical parameter: An overview. Can Geotech J 53(2):252
25. Leetsaar L, Korkiala-Tanttu L, Kurnitski J (2022) CPT, CPTu and 35. Vardanega PJ, Bolton MD (2013) Stiffness of clays and silts: Nor-
DCPT methods for predicting the ultimate bearing capacity of cast malizing shear modulus and shear strain. Journal of Geotechnical
in situ displacement piles in silty soils. Geotech Geol Eng. https:// and Geoenvironmental Engineering 139(9):1575–1589
doi.org/10.1007/s10706-022-02292-6
26. Leetsaar, L., & Korkiala-Tanttu, L. (2023). Deterministic and Publisher’s Note Springer Nature remains neutral with regard to
probabilistic analyses of the bearing capacity of screw cast in situ jurisdictional claims in published maps and institutional affiliations.
displacement piles in silty soils as measured by CPT and SDT.
The Baltic Journal of Road and Bridge Engineering. https://b jrbe-
journals.rtu.lv/article/view/bjrbe.2023-18.600
13