INTERNATIONAL SOCIETY FOR

SOIL MECHANICS AND

GEOTECHNICAL ENGINEERING

This paper was downloaded from the Online Library of

the International Society for Soil Mechanics and
Geotechnical Engineering (ISSMGE). The library is
available here:

https://www.issmge.org/publications/online-library

This is an open-access database that archives thousands

of papers published under the Auspices of the ISSMGE and
maintained by the Innovation and Development
Committee of ISSMGE.

The paper was published in the proceedings of the 6th

International Conference on Geotechnical and Geophysical
Site Characterization and was edited by Tamás Huszák,
András Mahler and Edina Koch. The conference was
originally scheduled to be held in Budapest, Hungary in
2020, but due to the COVID-19 pandemic, it was held
online from September 26th to September 29th 2021.
 Evaluation of bearing capacity and in situ shear strength
using the screw plate load test in clay and silt
Ø. Blaker
Norwegian Geotechnical Institute, Oslo, Norway, [email protected]
D. J. DeGroot
University of Massachusetts, Amherst, MA, USA, [email protected]
J. T. DeJong
University of California, Davis, CA, USA, [email protected]

ABSTRACT: Recent studies show that silts are sensitive to sampling disturbance, and that the effects of sampling can
be adverse and opposite of those typically observed for clays. Silts often exhibit a tendency for dilative behavior upon
undrained triaxial shear. As a result, the interpreted shear strength is highly dependent on which failure criterion is selected
but there is limited guidance or consensus on what criterion represents the relevant in situ shear strength for design appli-
cations. To this end, in situ Screw Plate Load Tests (SPLT) have been conducted at Halden, Norway, to investigate the
bearing capacity and behavior of the silt and clay deposits under field loading, and uncertainties associated with un-
drained/drained/partially-drained conditions. Normalized penetration velocity indicates that the SPLTs were likely par-
tially-drained in the silt unit and undrained in the clay unit. This information was used to back-calculate estimates of the
in situ strengths for comparison with laboratory tests conducted on undisturbed specimens from both soil units.

Keywords: silt; clay; triaxial test; screw plate load test; bearing capacity.

where v = rate of penetration; D = penetrometer diameter;

1. Introduction and ch = coefficient of horizontal consolidation. V > 10
– 100 have been suggested to be indicative of fully un-
An increasing number of geotechnical projects involv- drained conditions, while fully drained conditions typi-
ing silt has sparked a series of research efforts to better cally occurs for V < 0.05 – 0.01 [1-3]. Penetrometer
understand the fundamentals of this intermediate soil, the measurements conducted under V = 0.05 - 10 may there-
effects of sampling disturbance and uncertainties associ- fore be affected by partial drainage.
ated with undrained/drained/partially-drained conditions. Furthermore, recent studies demonstrated that silts are
For sands and clays, deformation and strength parameters particularly sensitive to sampling disturbance, and that
can be evaluated in situ through well-established correla- the effects of tube sampling on engineering properties
tions with measured or derived parameters from cone can be adverse and opposite of those typically observed
penetration tests with pore pressure measurements for clays [4]. Tube samples of silt often exhibit a ten-
(CPTU), dilatometer tests (DMT), self-boring pressure- dency for dilative behavior and strain hardening upon un-
meter tests (SBP), or back-calculated and interpreted drained triaxial shear in compression and, as a result, the
from plate load tests (PLT). The CPTU, for example, can undrained shear strength of this material cannot be read-
be used to estimate undrained shear strength (su), effec- ily interpreted at the conventional peak shear stress as for
tive stress friction angle ('), constrained modulus (M) soft structured clays [4-8]. The shear strength of the ma-
and small strain shear modulus (Gmax) of a soil with terial depends on the criterion selected for interpretation
depth, and to estimate axial pile capacity (Qult) from the and there is limited guidance or consensus on what crite-
cone resistance (qc). rion most accurately represents the relevant in situ shear
Methods for interpretation of laboratory and in situ strength for design.
tests in silt have not seen the same developments or con- As sampling of silt has traditionally been considered
clusive research as for clays and sands, and there are still challenging, and quantitative assessment of sample qual-
large uncertainties associated with in situ behavior and ity using clay-based criteria is highly questionable in this
appropriate geotechnical parameters for practical engi- soil type, in situ loading tests were considered attractive
neering design in this soil type. Partial drainage effects for evaluation of bearing capacity and shear strength.
may have a significant effect on sample quality, the in- Marsland [9] used PLT data to back-calculate undrained
terpreted soil behavior type and soil properties from in shear strength of stiff, fissured London clay, showing that
situ and laboratory testing. For example, results from the large-scale undrained shear strength was significantly
"twitch" testing at variable penetration rates have demon- lower than that measured in small undrained triaxial com-
strated how CPTU measurements change with normal- pression test (CAUC) specimens. A variation of the PLT,
ized penetration velocity (V), expressed as: the SPLT uses a single flight helical screw to advance
from ground level without the need for a pre-augered
𝑉 = 𝑣𝐷 ⁄𝑐ℎ (1) borehole, thus retaining the overburden stress [10]. This
configuration was adopted and used to evaluate com-
pressibility of different sands and the influence of
preconsolidation stress on sand deformability by pressure was set equal to the in situ vertical effective
Schmertmann [11] and Dahlberg [12], respectively. The stress and the equipment was allowed to rest for about 15
device has also been successfully used in a number of min to allow equalization of installation pore pressures
different clays [13-17], but only a few results have been near the screw plate. Displacement gauges were zeroed,
conducted in silt. Janbu and Senneset [18] and Sandven and continuous rate of deformation testing was con-
[19] report incremental loading SPLTs (i.e., fully drained ducted using the GDS pump. A GDS flow rate of about
conditions) conducted at a silt site in Stjørdal, Norway 40 mm3/s was typically used, providing a displacement
for evaluation of in situ compressibility of the deposit. rate of about 1.33 mm/min (0.5D/hr). Readings of cylin-
This paper presents results of three SPLTs conducted der pressure and plate displacement (s) at fixed time in-
at the National GeoTest Site for silt in Halden, Norway. tervals (t) were recorded to a displacement of about s =
It investigates load-deformation behavior in the clayey 0.2D. After completion of a test, the reference beam and
silt and underlying clay units, interpretation of engineer- deformation indicators were dismounted, and the system
ing parameters and compares the measured bearing ca- carefully vented to atmospheric pressure. The oil reser-
pacities with calculated base unit resistance for an equiv- voir was vented and the hydraulic cylinder, typically
alent diameter closed end pile. fully extended after testing, was reset to its original posi-
tion using the drill rig. Finally, the pumps were discon-
2. Methods nected, and the screw plate advanced to the next test
depth.
2.1. Sampling
Soil samples were collected at the Halden, Norway re-
search site [8] using the Sherbrooke block sampler [20]
in location HALB04, the NGI 54 mm inner diameter (ID)
composite piston sampler [21] in location HALB03 and
the Gregory Undisturbed Sampler (GUS), a hydraulic
fixed piston sampler, manufactured by Acker Drill Com-
pany, PA, USA in location HALB07. All locations are
presented on the map in Figure 1.

2.2. Field equipment

The screw plate equipment consisted of a single helix
flight auger (Figure 2) with D = 160 mm (Area, A = 200
cm2) and a 45 mm pitch. The plate was founded in ductile
cast iron (EN-GJS-500) by Ulefoss Foundry, Norway
based on a model by Strout [22]. The screw plate was
positioned directly in front of a custom-made down-hole
hydraulic jack and double-rod configuration described by
Janbu and Senneset [18]. The outer 42.5 mm outer diam- Figure 1. In situ testing and sampling locations at Halden. SPLTs
were conducted in HALSP01 and HALSP02.
eter (OD) steel rods provided torque during installation
and reaction from the jack to the drill tower of the
Georigg 607 (Geotech AB, Sweden) drill rig during static
loading. A simple load frame was positioned between the
outer rod and drill rig and allowed access to the top of the
27 mm OD center rods. The unloaded center rods pro-
vided direct measurement of the plate displacement using
two Mitutoyo Digimatic ID-C 0.001/50.8 mm defor-
mation indicators mounted on an independent reference
beam. An Enerpac P392 hand pump and a 64 MPa GDS
high pressure volume controller provided hydraulic pres-
sure to the closed system through a 400 MPa capacity hy-
draulic hose connected to the jack positioned directly be-
hind the screw plate. Hydraulic cylinder pressure to plate
stress (qp) conversions were calibrated in the laboratory
using an Interface (Interface Inc., Scottsdale, AZ, USA)
250 kN load cell.
The screw plate was carefully installed by rotation
from ground level to target depth (z) by the drill rig. The
rate of penetration during installation was adjusted to
equal the pitch of the screw plate (i.e. about 45 mm per
360° rotation) in order to minimize disturbance to the sur-
Figure 2. Screw plate with diameter D = 160 mm and pitch of 45 mm
rounding soil. The Enerpac pump and GDS volume con- used at the Halden GeoTest Site.
troller were connected to the hydraulic hose, the plate
2.3. Triaxial testing
where 𝑞𝑏,𝑢𝑙𝑡 = the ultimate unit base resistance and 𝐴𝑏 =
Triaxial specimens were prepared by trimming of area of the pile base. The ultimate unit base resistance of
Sherbrooke block and GUS specimens using the proce- a pile tip equivalent to that of the screw plate (D = 160
dures described by Lacasse and Berre [23] and Ladd and mm) was assessed using a number of methods, including:
DeGroot [24]. During back pressure saturation the test
specimens were first subjected to an isotropic stress (cell • the classical bearing capacity equation (disre-
pressure) equal to the estimated value of the initial nega- garding the 0.5𝛾′𝐷𝑁𝛾∗ term due to its small
tive pore pressure (suction) within the specimen. The po-
relative contribution), i.e.:
rous filter stones were initially dry. At the initial isotropic
stress, de-aired water was flushed through the porous
𝑞𝑏,𝑢𝑙𝑡 = 𝑁𝑐∗ 𝑠𝑢 + 𝑁𝑞∗ 𝜎𝑣0
′
(3)
stones and any tendency for volume change was pre-
vented by adjusting the cell pressure until a stable condi-
tion was reached. Following this stage, backpressure was where 𝑁𝑐∗ , 𝑁𝑞∗ , 𝑁𝛾∗ = dimensionless bearing ca-
applied and all B values, which were measured at the end pacity factors for deep foundations, including
of the consolidation phase, were ≥ 97%. All specimens necessary shape and depth factors; 𝑠𝑢 = un-
were anisotropically consolidated to the best estimate in drained shear strength; and 𝛾′ = effective unit
situ vertical effective stress, 'v0 and horizontal effective weight of soil [31-33].
stress  'h0 using an assumed K0 = 0.5 [8]. All specimens • CPTU-based methods, including:
were allowed to creep for 12 to 24 hours prior to un- - Purdue-CPT [27],
drained shear testing performed at a strain rate of 0.5 – - NGI-05 [34, 35],
1.4 %/hr. The total radial stress was kept constant while - ICP-05/MTD-1996 [36, 37], and
the total axial stress was increased in compression - UWA-05/UWA-13 [38, 39].
(CAUC). All stress measurements were corrected for
membrane resistance and changes in specimen area [25]. All CPTU-based design methods are summarized by
Han, et al. [40].
2.4. Analysis
2.4.3. Shear strength
2.4.1. Ultimate bearing capacity from SPLT Undrained shear strength from CAUC tests on clay
specimens were assessed at peak shear stress, i.e. 𝑠𝑢𝐶 =
Three methods were used to assess the ultimate bear- 0.5(𝜎1 − 𝜎3 )𝑚𝑎𝑥 . For silt specimens displaying dilative
ing capacity, qult from the SPLT stress-displacement type behavior during undrained shear, and thus, no peak
curves: shear stress, 𝑠𝑢𝐶 was evaluated using the following
• 0.1B method – ultimate bearing stress limited strength criteria [41]:
by a relative displacement, typically 10% of
the footing width or pile diameter, B [26, 27]. ▪ maximum deviator stress, (𝜎1 − 𝜎3 )𝑚𝑎𝑥 ;
In this case, 10% of the screw plate diameter, ▪ an assigned limiting vertical strain, 𝜀𝑣𝑓 ;
D, i.e. 𝑞𝑢𝑙𝑡 = 𝑞0.1𝐷 . ▪ state of zero excess shear induced pore pres-
• Tangent intersect – bearing stress correspond- sure at failure ∆𝑢𝑓 = 0, which is equivalent
ing to a distinct change in plate displacement, to Skempton's A parameter at failure equal to
i.e. intersection of initial and final tangent
zero, 𝐴𝑓 = 0;
slope of stress - settlement plot [28], i.e. 𝑞𝑢𝑙𝑡 =
▪ point at which the effective stress path first
𝑞𝑇𝐼 .
reaches the failure envelope, defined by the
• Curve fitting – ultimate bearing capacity ex-
𝐾𝑓 line;
trapolated using an exponential curve inter-
secting the bearing stress, qx and qy at 0.015D ▪ maximum obliquity, (𝜎1′ ⁄𝜎3′ )𝑚𝑎𝑥 ;
and 0.02D, respectively [15], i.e. 𝑞𝑢𝑙𝑡 = 𝑞𝐾𝑃 . ▪ maximum shear induced pore pressure,
𝑢𝑚𝑎𝑥 .
Other methods are available, e.g. the Log-Log method
[29], but were considered inappropriate for the interpre- Undrained shear strength assessed from the screw
tation of the load tests described in this paper. For all plate load tests were back-calculated using:
methods listed above the displacement at failure (𝑠𝑓 )
𝑠𝑢 = 𝑞𝑢𝑙𝑡 ⁄𝑁𝑐∗ (4)
were taken as the displacement corresponding to 𝑞𝑢𝑙𝑡 .
3. Test program and site description
2.4.2. Pile ultimate unit base resistance
Three SPLTs were performed, one at a depth of 11.3
A deeply embedded screw plate (z/D > 33) may be
m in borehole HALSP01, and one each at 11.3 and 17.8
compared to the base of a circular closed end pile (CEP)
m depths in borehole HALSP02 (Figure 2) at the Norwe-
with equivalent diameter and area. The ultimate base re-
gian GeoTest Site (NGTS) for silt. The site is located in
sistance of a pile is expressed as [30]: Halden, Norway, approximately 120 km south of Oslo
and has been well characterized [see 8] by combining the
𝑄𝑏,𝑢𝑙𝑡 = 𝑞𝑏,𝑢𝑙𝑡 𝐴𝑏 (2)
 Figure 3. Classification and CPTU characteristics of the Halden research site; (a) Soil Units, (b) natural water content and Atterberg limits, (c) total
unit weight, and (d) clay particle and fines content, (e) corrected cone resistance, qt, (f) pore pressure, u2, and (g) sleeve friction, fs. Modified
from Blaker, et al. [8].

results of a number of geological, geophysical and ge-

otechnical site investigation tools; including sampling, 4. Results
CPTU, CPTU pore pressure dissipation tests and field
vane tests (FVT).
A silty, clayey sand constitutes the top soil (Unit I) and 4.1. Triaxial testing
extends down to about 4.5 to 5 m depth. The geologically The CAUC clay specimen from 18.6 m depth had a vol-
normally consolidated clayey silt below (Units II and III) umetric recompression strain of vol = 2.7%, correspond-
extend down to about 15 to 16 m depth, with soil behav-
ing to e/e0 = 0.054, thus giving it a "good to fair" sample
ior type index (Ic) generally plotting between 2.6 and
quality rating [42]. During shear the specimen showed a
2.95. Normalized cone resistance (Qt) and pore pressure
peak shear stress and exhibited strain softening thereaf-
ratio (Bq) in these soil units are generally in the order of
ter. The undrained shear strength indicated from this test
7.5 and 0.1 – 0.3, respectively. The silt is uniform and
structureless to mottled, with primary bedding and lami- was suC = 82 kPa at a vertical strain of vf = 0.8%. The
nations almost absent due to bioturbation. Both Units II pore pressure at peak shear stress was 35 kPa correspond-
ing to a Skempton's pore pressure parameter Af = 0.59 at
and III contain similar amounts of quartz (40%), plagio-
failure (Figure 4). Interestingly, the effective stress path
clase (30%), feldspar (12%), clay minerals and mafic
tags the failure envelope defined by the Kf line of the
minerals (amphibole). Clay minerals are illite and chlo-
CAUC tests conducted in the silt units, indicated by the
rite, and the presence of expanding clay minerals are low
or absent. Unit IV, a low to medium strength clay has a maximum obliquity friction angle 'mo = 35.8° [8].
slightly laminated structure, with occasional shell frag- The CAUC silt specimen from 8.4 m, 11.5 m and 12.6
ments and drop stones. Qt and Bq are generally in the or- m depth [8] had recompression metrics of vol = 1.3%,
der of 4 and 0.8 – 1.0, respectively. Depth to bedrock dips 1.0% and 1.1% for volumetric strain and e/e0 = 0.029,
sharply from the northeast to southwest but is typically 0.023 and 0.026, respectively. By the clay-based sample
identified at 21 m depth in the southern part of the site. quality framework these low values of e/e0 would rate
Table 1 and Figure 3 summarizes typical soil properties the specimens as "good to excellent" sample quality [42,
and CPTU characteristics of the silt at 11.3 m and clay at 43]. However, the clay-based sample quality criteria have
17.8 m depth. been shown to be misleading for low plasticity silts [4,
44]. Figure 4 shows that, except for the initial contractive
Table 1. Typical soil properties at Halden Research site, 11.3 m and
17.8 m depth. type behavior, the specimens develop net negative pore
z w ei 1) wL 2) Ip Fines Clay cv0
pressure changes, and thus, show a strong tendency to-
3) 3) 4)
wards dilative behavior. The test results show a distinct
initial S-shape behavior in stress-path space, particularly
[m] [m] [-] [%] [%] [%] [%] [m2/yr] for the specimens sampled at 8.4 m and 11.5 m depth.
11.5 27 0.73 23 1 89 9 221 Phase transformation points (PTP), i.e., the point at
18.6 34 0.96 27 9 87 28 10 which the soil transitions from contractive type behavior
1)
ei = initial void ratio. Note also emax = 1.51 and emin = 0.60, giving to dilative type behavior, are located at an angle of ap-
an estimated relative density, Dr = 86% for z = 11.3 m;
2)
proximately 'PTP = 33°. The stress-path results generally
Liquid limit determined by the Casagrande cup;
3)
Fines < 0.063 mm, clay < 0.002 mm; track the Kf line at a maximum obliquity friction angle
4)
Coefficient of vertical consolidation at 'v0. 'mo = 35.8° [8] to the end of the test. Due to this strain
hardening behavior interpretation of the undrained shear
strength from these CAUC tests is complex and the
 displacements for small changes in load thereafter. The
results of the two silt tests show more gradual increase in
deformation with load. The tests exhibit a significantly
more pronounced strain-hardening relative to the clay
test - similar to the triaxial test results described above.
There is reasonable agreement between the two tests con-
ducted at 11.3 m depth in boreholes HALSP01 and
HALSP02, although some variability is evident. All tests
were stopped at a displacement corresponding to about
0.2D.
Although traditionally calculated for CPTU twitch tests
[1], an assessment of normalized penetration velocity
gives V = 10 for the SPLT in the clay unit (assuming ch =
cv0, Table 1) indicating that undrained conditions pre-
vailed, as expected. From the load tests in the silt unit,
the normalized penetration velocity is about V = 0.5, and
thus, suggests partially drained conditions during load-
ing. These conditions cause complex pore pressure fields
surrounding the screw plate, with large gradients in the
vertical direction. Locally near the plate the soil shear re-
sistance is fully mobilized and likely developed negative
pore pressure changes combined with some dilation due
to partial drainage. Whereas at some distance below the
plate (and radially), soil elements may have experienced
positive pore pressure changes combined with some con-
traction due to the increase in compression stresses being
greater than the mobilized shear stresses (resulting in the
soil remaining well below the failure envelope). Glob-
ally, however, the load-displacement behavior of the silt
tests suggests a dilative type of behavior, with stresses
acting on the screw plate increasing at a significantly
larger rate relative to the test in clay.

4.2.2. Bearing capacity

Ultimate bearing capacities from the SPLTs were as-
sessed using three different criteria as detailed above and
illustrated for each individual SPLT on Figure 5. The
bearing capacity interpreted at a displacement equal to
10% of the plate diameter, gave consistently higher val-
ues, i.e. q0.1D > qTI, qKP, relative to the other two criteria.
The tangent intersect and Kay and Parry [15] interpreta-
tion methods gave similar values of qult in both the silt
and clay units (Table 3).

4.2.3. Undrained shear strength

The back-calculated undrained shear strength in the
Figure 4. Undrained triaxial test results (CAUC) from the Halden clay
clay from q0.1D, qTI and qKP (Eq. (4)) gave values of su =
and silt units. 54 kPa, 46 kPa and 47 kPa, respectively (Table 2) when
applying a bearing capacity factor of Nc* = 9. These val-
results provide no unique (peak) undrained shear ues are considered "average" or "mobilized" undrained
strength. Undrained shear strength evaluated at different shear strengths for the soil at the screw plate embedment
criteria [41] are presented in Table 2. depth, thus approximately equivalent to the direct simple
shear (DSS) undrained shear strength (suD) of the same
4.2. Screw plate load testing soil. The DSS and CAUE undrained shear strengths of
the Halden clay can be estimated as suD = 57 kPa and suE
4.2.1. Load-displacement behavior = 34 kPa, respectively, based on the strength anisotropy
factors suD/suC = 0.69 and suE/suC = 0.42 reported by
Typical stress - displacement curves from the silt (11.3 Lunne, et al. [42] for similar clays from the Oslo, Norway
m) and clay (17.8 m) tests are presented in Figure 5. The area. Thus, the undrained shear strength back-calculated
SPLT results from the clay shows a distinct change in from q0.1D provides excellent agreement (within 5%) with
displacement around qp = 400 kPa and relatively large the laboratory test and strength anisotropy of the region,
Table 2. Apparent and measured undrained shear strength from screw
were likely partially drained (V = 0.5). However, back-
plate load tests, field vane and triaxial tests at Halden. calculation of in situ strength parameters using conven-
z [m] 11.3 – 11.5 17.8 – 18.5 tional methods requires an assumption of the prevailing
Type (silt) (clay)
conditions as either drained or undrained during loading.
By assuming undrained conditions su of the silt was back-
[-] [kPa]
calculated using Eq. (4). Table 2 presents the results from
Laboratory CAUC these back-calculations, in terms of average su,TI, su,0.1D
su,C [(1 − 3)max] 94 82 and su,KP representing the undrained shear strength calcu-
su,C (umax) 50 - lated from qTI, q0.1D and qKP, respectively. Interestingly,
su,C (vf = 2%) 57 - the TI and KP results (72 kPa and 64 kPa) show agree-
su,C (Af = 0) 84 -
ment with the CAUC test at the same depth level for suC
interpreted using the shear stress at the Kf line and at max-
su,C [(1/3)max] 70 -
imum obliquity criteria (70 kPa). It is hypothesized that
su,C (Kf) 70 - the SPLT tests in the silt do generate negative pore pres-
su,C (vf = 10%) 84 - sures changes, and that the TI ( KP) failure criteria rep-
In situ tests resent the point at which the soil elements involved in the
su,TI 72 1) 46 global failure mechanism below the plate start becoming
su,0.1D 92 1) 54
fully mobilized. Furthermore, the undrained shear
1)
strengths back-calculated from q0.1D (su,0.1D = 92 kPa)
su,KP 64 47
show similarities with the undrained strength interpreted
su,FVT 45 41 at (1 − 3)max of the companion CAUC test (suC = 94
su,SBP 51 - kPa). This implies that the shear stress obtained from
Note: 1)
Average of two tests CAUC tests on silt block sample specimens at large
strains can be used to reliably estimate the bearing capac-
ity at 0.1D for short term loading and that the strain hard-
ening effect can be relied upon. This, however, requires
high quality samples with minimum of sample disturb-
ance from sampling, transportation and handling. Recent
studies have shown that effects of disturbance on silt
samples can have opposite effects of that often seen for
structured clays, i.e., larger interpreted strength and stiff-
ness properties with increasing disturbance [4, 44].

4.2.4. Effective stress friction angle

For back-calculation of the effective stress friction an-
gle of the silt using conventional methods drained condi-
tions are required. By assuming fully drained conditions
during SPLT loading ' were estimated using the stress-
displacement curve and Eq. (3). The largest uncertainty
in this back-calculation is the bearing capacity factor,
Nq*, which varies significantly in the literature [45, 46]
(Figure 6). The bearing capacity factor computed from
Figure 5. Typical stress-displacement curves from screw plate load the SPLTs at 11.3 m depth are failure criteria dependent,
tests in the silt (11.3 m) and clay (17.8 m) units at Halden. Ulti-
mate bearing stress assessed using the 0.1D, (q0.1D), tangent inter- but range between Nq* = 4.3 and 6.5, resulting in corre-
sect (qTI) and Kay and Parry (qKP) methods. sponding values of ' = 12° – 24° using the curves in Fig-
Table 3. Bearing capacity interpreted from screw plate load tests at ure 6. Effective stress friction angles in this range are
11.3 m and 17.8 m depth. considered unrealistically low compared to results from
Borehole Depth TI 0.1D KP triaxial tests conducted specimens of Halden silt and
[-] [m] [kPa] [kPa] [kPa]
other international silts reported in literature [7, 19, 41,
47-49]. This implies that the SPLTs were not fully
HALSP01 11.3 634 810 566
drained during loading, i.e. partial drainage prevailed as
HALSP02 11.3 660 842 581 suggested by V = 0.5, and that the measured bearing ca-
HALSP02 17.8 410 487 422 pacities cannot be used to reliably back-calculate the fric-
tion angle. The bearing capacity factor appear highly un-
and validates both the SPLT stress-displacement results certain in silts. Helical Anchors Inc. [50] suggests Nq* =
and the equipment as an effective tool for evaluation of 28 for compression loading of a helical pile in a cohe-
undrained shear strength in soft clay. FVT results at the sionless soil and hence with an effective stress friction
same depth [8] resulted in su,FVT = 41 kPa (i.e. su,FVT/suC = angle of 'mo = 35.8°, overestimates the ultimate bearing
0.5), and thus show better agreement with the back-cal- capacity (unfactored) of the screw plate load tests at 11.3
culated undrained shear strength using qTI. m depth by factors of 3.5 to 5.5. Using the constant vol-
Drainage conditions during the SPLTs in the silt unit ume friction angle (approximately equal to 'PTP) of 'cv
are complex and uncertain, but as noted above the tests = 33° reduces the corresponding value of Nq* to about 19.
The Canadian foundation engineering manual [33] pre-
sents typical bearing capacity factors for deep founda-
tions in silt as 10 – 30 (cast-in-place piles) and 20 – 40
(driven piles), and as a result, also overpredicts qult. For
offshore piles in cohesionless soils API RP2A [31] sug-
gests Nq* in the range of 8 – 12 for medium dense to dense
silts, giving better agreement with the SPLT bearing ca-
pacity results. However, predictions of axial capacity of
piles driven into cohesionless soil using API RP2A have
been noted to be inaccurate [36, 51] and more recent
guidelines [e.g. 52] recommend CPTU-based methods to
assess bearing capacity in these soils.

5. Measured and calculated capacity

Figure 7 presents the measured SPLT bearing capacity
at s = 0.1D displacement (q0.1D) plotted with ultimate unit
base resistance (qb,ult) of an equivalent diameter closed
end pile using clay methods at 17.8 m depth and cohe-
sionless soil (sand) methods for the silt at 11.3 m depth. Figure 6. Bearing capacity factors, Nq* for deep foundations in cohe-
In the clay the measured SPLT result shows excellent sionless soil as function of effective stress friction angle.
agreement with the calculated bearing capacity using Nc*
= 9 and a DSS undrained shear strength, suD, as noted in
Section 4.2.3 above. The API RP2A and NGI-05 meth-
ods use the unconsolidated undrained shear strength
(suUU), in this case assumed equal to suC, and appear to
overestimate the capacity by about 50%. The ICP/MTD-
1996 and UWA-13 methods (using corrected cone re-
sistance, qt) also overestimate the capacity in the clay, by
a factor of 1.33. Helical Anchors Inc. [50] do not state
what undrained shear strength to use for design but for
illustration purposes suD was used in Figure 7 for calcu-
lation of qb,ult. CGS [33] suggests the minimum undrained
shear strength (i.e., suE) for capacity assessment, and as a
result qb,ult is underestimated relative to q0.1D. In sum-
mary, the best agreement with the measured bearing ca-
pacity of the SPLT at 0.1D in the Halden clay was ob-
tained by using suD and a bearing capacity factor equal to
9.
Relative to the measured values of q0.1D, the classic
drained bearing capacity equation for deep foundations Figure 7. Measured SPLT bearing capacity (q0.1D) versus calculated
base unit resistance of an equivalent closed end pile (qb,ult).
in cohesionless soil typically over estimates the unit base
resistance at Halden by a factor of up to 4.5, but the val-
ues of qb,ult are highly dependent on the selected bearing 6. Summary and conclusions
capacity factor, Nq*. For example, API RP2A using Nq* =
The screw plate load test (SPLT) was considered an
8 shows fair agreement with the measured values from
attractive tool for investigation of the in situ soil behavior
the SPLTs. The CPTU-based methods all underestimate
of the silt deposit at Halden, Norway described by
the unit base resistance at 10% vertical displacement. It
Blaker, et al. [8], which displays dilative type behavior
should be noted, however, that these methods were de-
during undrained shear in the laboratory CAUC tests and
veloped for sands with significantly higher cone re-
a maximum obliquity friction angle of 'mo = 35.8°. One
sistances and that CPTU qc at 11.3 m depth at Halden
test was conducted in the clay unit below 16 m depth and
were measured using the conventional penetration rate of
two companion tests were performed in the silt at 11.3 m
20 mm/s, giving normalized velocities of about V = 180
depth. The main findings were:
[53], i.e., fully undrained conditions. Furthermore, rela-
▪ The SPLT in the clay were conducted with a nor-
tive density (Dr) derived from qc and estimated effective
malized velocity of about V = 10, indicating un-
horizontal stresses, 'h [54], were developed for clean
drained conditions during loading. The soil dis-
sands. Dr estimates at Halden (80% - 86%) were based
played a distinct break in the stress - displacement
on measured initial void ratios (ei) of seven triaxial spec-
curve during loading.
imens trimmed from a block sample collected at 11.5m
▪ Interpretation of the clay test confirmed (within
depth and maximum and minimum void ratios measured
5%) the theoretical bearing capacity estimated us-
on air dried silt from the same block sample (Table 1).
ing the direct simple shear (DSS) undrained shear
Values of qc and Dr used in the CPTU-based methods for
strength of the same soil, thus validating the
calculation of qb,ult are therefore somewhat uncertain.
 stress-displacement curve and the equipment as [2] DeJong, J.T. and M. Randolph, "Influence of Partial
Consolidation during Cone Penetration on Estimated Soil
an effective tool for evaluation of undrained shear
Behavior Type and Pore Pressure Dissipation Measurements".
strength in soft clay. Journal of Geotechnical and Geoenvironmental Engineering,
▪ The two SPLTs performed in silt showed good re- 2012. 138(7): p. 777-788
peatability and a normalized velocity of about 0.5. https://doi.org/10.1061/(ASCE)GT.1943-5606.0000646.
[3] Kim, K., M. Prezzi, R. Salgado, and W. Lee, "Effect of Penetration
Normalized velocities in the range 10 > V > 0.05
Rate on Cone Penetration Resistance in Saturated Clayey Soils".
have been suggested to be indicative of partially Journal of Geotechnical and Geoenvironmental Engineering,
drained conditions. Thus, the rate of loading used 2008. 134(8): p. 1142-1153 https://doi.org/10.1061/(ASCE)1090-
at Halden likely caused complex pore pressure 0241(2008)134:8(1142).
[4] Carroll, R. and M. Long, "Sample Disturbance Effects in Silt".
fields surrounding the screw plate.
Journal of Geotechnical and Geoenvironmental Engineering,
▪ Both silt tests displayed a significantly more pro- 2017. 143(9): p. 04017061
nounced strain-hardening behavior relative to the https://doi.org/10.1061/(ASCE)GT.1943-5606.0001749.
clay SPLT. This behavior confirmed the observa- [5] Fleming, L.N. and J.M. Duncan, "Stress-Deformation
Characteristics of Alaskan Silt". Journal of Geotechnical
tions from the stress-strain and stress-path devel-
Engineering, 1990. 116(3): p. 377-393
opment during undrained triaxial shearing https://doi.org/10.1061/(asce)0733-9410(1990)116:3(377).
(CAUC) of the bock sample from the same depth. [6] Høeg, K., R. Dyvik, and G. Sandbækken, "Strength of undisturbed
▪ Due to the strain hardening effect the bearing ca- versus reconstituted silt and silty sand specimens". Journal of
Geotechnical and Geoenvironmental Engineering, 2000. 126(7):
pacities at a displacement equal to 0.1D gave con-
p. 606-617 https://doi.org/10.1061/(asce)1090-
sistently higher values relative to the tangent in- 0241(2000)126:7(606).
tersect and Kay and Parry [15] methods. [7] Long, M., G. Gudjonsson, S. Donohue, and K. Hagberg,
▪ It is suggested that the SPLT generated negative "Engineering characterisation of Norwegian glaciomarine silt".
Engineering Geology, 2010. 110(3): p. 51-65
pore pressure changes in the silt immediately be-
https://doi.org/10.1016/j.enggeo.2009.11.002.
low the plate, and that qult for the tangent intersect [8] Blaker, Ø., R. Carroll, P. Paniagua, D.J. DeGroot, and J.-S.
criteria represents the start of a fully mobilized L'Heureux, "Halden research site: geotechnical characterization of
shear stress state below the screw plate, equiva- a post glacial silt". AIMS Geosciences, 2019. 5(2): p. 184-234
https://doi.org/10.3934/geosci.2019.2.184.
lent to the Kf and maximum obliquity failure cri-
[9] Marsland, A., "Clays subjected to in situ plate tests". Ground
teria used for assessment of su from CAUC tests. Engineering, 1972. 5(6): p. 24-31.
▪ The negative shear induced pore pressures and [10] Kummeneje, O., "Fundamentering av oljetank i Drammen
undrained shear strength at large strains observed (Foundation of an oil tank in Drammen)". Norwegian
Geotechnical Institute Publication (In [Norwegian]), 1956. 12: p.
from CAUC testing on the silt block sample can
1-6.
likely be relied upon for short term loading in the [11] Schmertmann, J.H., "Static Cone to Compute Static Settlement
field. For extrapolation to other silt sites one must over Sand". Journal of the Soil Mechanics and Foundations
ensure high quality samples for laboratory testing Division, ASCE, 1970. 96(SM3): p. 1011-1043.
[12] Dahlberg, R., "Settlement characteristics of preconsolidated
and that the effects of disturbance on the engi-
natural sands: In-situ screw-plate, pressurementer and penetration
neering design parameter are properly evaluated. tests". 1975, National Swedish Institute for Building Research:
▪ Fully drained bearing capacities were likely not Stockholm, Sweden. p. 315.
measured during the SPLTs at Halden. The bear- [13] Bergado, D. and N. Huan, "Undrained Deformability and Strength
Characteristics of Soft Bangkok Clay By the Screw Plate Test".
ing capacity factor is a function of effective stress
Geotechnical Testing Journal, 1987. 10(3): p. 113-122
friction angle and, as a result Eq. (3) typically https://doi.org/10.1520/GTJ10943J.
over predict qult at Halden. Similarly, as an effect [14] Bergado, D.T., K.C. Chong, P.A.M. Daria, and M.C. Alfaro,
of the undrained response and relatively low val- "Deformability and consolidation characteristics of soft Bangkok
clay using screw plate tests". Canadian Geotechnical Journal,
ues of cone resistance the CPTU-based methods
1990. 27(5): p. 531-545 https://doi.org/10.1139/t90-069.
for estimation of qult under predict the bearing ca- [15] Kay, J.N. and R.H.G. Parry, "Screw plate tests in a stiff clay".
pacity. Ground Engineering, 1982. 15(6): p. 22-27.
[16] Selvadurai, A.P.S., G.E. Bauer, and T.J. Nicholas, "Screw plate
testing of a soft clay". Canadian Geotechnical Journal, 1980.
Acknowledgements 17(4): p. 465-472 https://doi.org/10.1139/t80-055.
[17] Kay, J.N. and D.L. Avalle, "Application of Screw Plate to Stiff
This study has primarily been financed by the Norwe- Clays". Journal of the Geotechnical Engineering Division, ASCE,
gian Geotechnical Institute (NGI) and the Research 1982. 108(GT1): p. 145-154.
Council of Norway (RCN) under Grant No. 245650. The [18] Janbu, N. and K. Senneset, "Field compressometer - principles and
applications", in Eighth International Conference on Soil
field vane tests were conducted with support from the US Mechanics and Foundation Engineering, N.A. TsytovichN and S.
National Science Foundation (NSF) under Grant Nos. Chetyrkin, (eds.). 1973. p. 191-198.
CMMI-1436793 and CMMI-1436617. Any opinions, [19] Sandven, R., "Geotechnical properties of a natural silt deposit
findings, and conclusions or recommendations expressed obtained from field and laboratory tests", in Characterisation and
Engineering Properties of Natural Soils, T.S. Tan, et al., (eds.).
in this material are those of the author(s) and do not nec- 2003, A.A. Balkema: Lisse. p. 1121–1148.
essarily reflect the views of NGI, RCN or NSF. This sup- [20] Lefebvre, G. and C. Poulin, "A new method of sampling in
port is gratefully acknowledged. sensitive clay". Canadian Geotechnical Journal, 1979. 16(1): p.
226-233 https://doi.org/10.1139/t79-019.
[21] Andresen, A. and P. Kolstad, "The NGI 54-mm samplers for
References undisturbed sampling of clays and representative sampling of
coarser materials", in Proc., Int. Symp. of Soil Sampling, State of
[1] Randolph, M.F., "Characterization of Soft Sediments for Offshore the Art on Current Practice of Soil Sampling. 1979, Japanese
Applications", in Geotechnical and Geophysical Site Society of Soil Mechanics and Foundation Engineering: Tokyo,
Characterization, A. Viana da Fonseca and P.W. Mayne, (eds.). Japan. p. 13-21.
2004, Millpress: Rotterdam, The Netherlands. p. 209–232.
[22] Strout, J.M., "Evaluation of the field compressometer test in Canadian Geotechnical Journal, 2006. 43(7): p. 726-750
sand", Ph.D., The Norwegian University of Science and https://doi.org/10.1139/t06-040.
Technology, 1998. [43] Lunne, T., P.K. Robertson, and J.J.M. Powell, "Cone penetration
[23] Lacasse, S. and T. Berre, "State-of-the-Art: Triaxial testing testing in geotechnical practice". 1997, London: Blackie
methods for soils", in Advanced Triaxial Testing of Soil and Rock, Academic & Professional.
ASTM STP 977, R. Donaghe, R. Chaney, and M. M. Silver, (eds.). [44] Lukas, W.G., D.J. DeGroot, J.T. DeJong, C.P. Krage, and G.
1988, ASTM International: West Conshohocken, PA. p. 264-289. Zhang, "Undrained Shear Behavior of Low-Plasticity
[24] Ladd, C.C. and D.J. DeGroot, "Recommended practice for soft Intermediate Soils Subjected to Simulated Tube-Sampling
ground site characterization: Arthur Casagrande lecture", in Proc., Disturbance". Journal of Geotechnical and Geoenvironmental
12th Panamerican Conf. on Soil Mech. and Geotech. Eng., P.J. Engineering, 2019. 145(1): p. 04018098
Culligan, H.H. Einstein, and A.J. Whittle, (eds.). 2003, Verlag https://doi.org/10.1061/(ASCE)GT.1943-5606.0001967.
Glückauf: Essen, Germany. p. 1-55. [45] Meyerhof, G.G., "Bearing Capacity and Settlement of Pile
[25] Berre, T., "Triaxial Testing at the Norwegian Geotechnical Foundations". Journal of the Geotechnical Engineering Division,
Institute". Geotechnical Testing Journal, 1982. 5(1/2): p. 3-17 ASCE, 1976. 102(GT3): p. 195-228.
https://doi.org/10.1520/GTJ10794J. [46] Vesic, A.S., "Design of pile foundations". 1977, National
[26] Briaud, J.-L. and P. Jeanjean, "Load settlement curve method for Cooperative Highway Research Program Synthesis of Highway
spread footings on sand", in Vertical and Horizontal Deformations Practice No. 42: Washington, D.C., USA.
of Foundations and Embankments, GSP 40, A.T. Yeung and G.Y. [47] Cola, S. and P. Simonini, "Mechanical behavior of silty soils of
Félio, (eds.). 1994, ASCE: New York, NY. p. 1774-1804. the Venice lagoon as a function of their grading characteristics".
[27] Salgado, R., S.I. Woo, and D. Kim, "Development of Load and Canadian Geotechnical Journal, 2002. 39(4): p. 879–893
Resistance Factor Design for Ultimate and Serviceability Limit https://doi.org/10.1139/t02-037.
States of Transportation Structure Foundations". 2011, Joint [48] Nocilla, A., M.R. Coop, and F. Colleselli, "The mechanics of an
Transportation Research Program, Indiana Department of Italian silt: an example of ‘transitional’ behaviour". Géotechnique,
Transportation and Purdue University: West Lafayette, IN, USA. 2006. 56(4): p. 261-271
[28] Trautmann, C.H. and F.H. Kulhawy, "Uplift Load‐Displacement https://doi.org/10.1680/geot.2006.56.4.261.
Behavior of Spread Foundations". Journal of Geotechnical [49] Long, M., "Engineering characterization of estuarine silts".
Engineering, 1988. 114(2): p. 168-184 Quarterly Journal of Engineering Geology and Hydrogeology,
https://doi.org/10.1061/(ASCE)0733-9410(1988)114:2(168). 2007. 40: p. 147-161 https://doi.org/10.1144/1470-9236/05-061.
[29] De Beer, E.E., "Experimental Determination of the Shape Factors [50] Helical Anchors Inc., "Engineering Design Manual, Rev 02".
and the Bearing Capacity Factors of Sand". Géotechnique, 1970. 2014, Helical Anchors, Inc.: Minneapolis, MN, USA.
20(4): p. 387-411 https://doi.org/10.1680/geot.1970.20.4.387. [51] Randolph, M.F., R. Dolwin, and R. Beck, "Design of driven piles
[30] Salgado, R., "The Engineering of Foundations". 2008, Boston, in sand". Géotechnique, 1994. 44(3): p. 427-448
MA, USA: McGraw Hill. https://doi.org/10.1680/geot.1994.44.3.427.
[31] API, "Recommended practice for planning, designing and [52] API, "Geotechnical and Foundation Design Considerations.
constructing fixed offshore platforms – working stress design. API ANSI/API Recommended Practice 2GEO 1st edition, April 2011.
RP 2A-WSD, 20th Edition". 1993, American Petroleum Institute: ISO 19901-4:2003 (Modified), Petroleum and natural gas
Washington, DC, USA. industries - Specific requirements for offshore structures, Part 4 -
[32] Bowles, J.E., "Foundation Analysis and Design". 1996, New Geotechnical and foundation design considerations". 2011,
York: McGraw-Hill. American Petroleum Institute: Washington, DC, USA.
[33] CGS, "Canadian Foundation Engineering Manual, 4th edition". [53] Carroll, R. and A.P. Paniagua López, "Variable rate of penetration
2007, Canadian Geotechnical Society: Richmond, B.C., Canada. and dissipation test results in a natural silty soil", in Cone
[34] Clausen, C.J.F., P.M. Aas, and K. Karlsrud, "Bearing Capacity of Penetration Testing 2018: Proc. of the 4th Int. Symp. on Cone
Driven Piles in Sand, the NGI Approach", in Proc. of the 1st Int. Penetration Testing, M.A. Hicks, F. Pisanò, and J. Peuchen, (eds.).
Symp. on Frontiers in Offshore Geotech., S. Gourvenec and M. 2018, CRC Press: London.
Cassidy, (eds.). 2005, Taylor & Francis: London. p. 677-682. [54] Jamiolkowski, M., D.C.F.L. Presti, and M. Manassero,
[35] Karlsrud, K., C.J.F. Clausen, and P.M. Aas, "Bearing Capacity of "Evaluation of Relative Density and Shear Strength of Sands from
Driven Piles in Clay, the NGI Approach", in Proc. of the 1st Int. CPT and DMT", in Geotechnical Special Publication No. 119:
Symp. on Frontiers in Offshore Geotech., S. Gourvenec and M. Soil Behavior and Soft Ground Construction, J.T. Germaine, T.C.
Cassidy, (eds.). 2005, Taylor & Francis: London. p. 775-782. Sheahan, and R.V. Whitman, (eds.). 2003, American Society of
[36] Jardine, R.J. and F.C. Chow, "New Design Methods for Offshore Civil Engineers: Reston, VA. p. 201-238.
Piles". MTD Publication. Vol. 96/103. 1996, London: Marine
Technology Directorate. 48.
[37] Jardine, R.J., F.C. Chow, R. Overy, and J. Standing, "ICP Design
Methods for Driven Piles in Sand and Clay". 2005, London:
Thomas Telford Publishing.
[38] Lehane, B.M., J.A. Schneider, and X. Xu, "The UWA-05 Method
for Prediction of Axial Capacity of Driven Piles in Sand", in Proc.
of the 1st Int. Symp. on Frontiers in Offshore Geotech., S.
Gourvenec and M. Cassidy, (eds.). 2005, Taylor & Francis:
London. p. 683-689.
[39] Lehane, B.M., Y. Li, and R. Williams, "Shaft Capacity of
Displacement Piles in Clay Using the Cone Penetration Test".
Journal of Geotechnical and Geoenvironmental Engineering,
2013. 139(2): p. 253-266
https://doi.org/10.1061/(ASCE)GT.1943-5606.0000749.
[40] Han, F., M. Prezzi, R. Salgado, and M. Zaheer, "Axial Resistance
of Closed-Ended Steel-Pipe Piles Driven in Multilayered Soil".
Journal of Geotechnical and Geoenvironmental Engineering,
2017. 143(3): p. 04016102
https://doi.org/10.1061/(ASCE)GT.1943-5606.0001589.
[41] Brandon, T.L., A.T. Rose, and J.M. Duncan, "Drained and
undrained strength interpretation for low-plasticity silts". Journal
of Geotechnical and Geoenvironmental Engineering, 2006.
132(2): p. 250-257 https://doi.org/10.1016/(asce)1090-
0241(2006)132:2(250).
[42] Lunne, T., T. Berre, K.H. Andersen, S. Strandvik, and M. Sjursen,
"Effects of sample disturbance and consolidation procedures on
measured shear strength of soft marine Norwegian clays".