Soil Dynamics and Earthquake Engineering 133 (2020) 106111

Contents lists available at ScienceDirect

Soil Dynamics and Earthquake Engineering

journal homepage: http://www.elsevier.com/locate/soildyn

Influence of the relative density and K0 effects in the cyclic response of

Ottawa F-65 sand - cyclic Torsional Hollow-Cylinder shear tests for
LEAP-ASIA-2019
Ruben R. Vargas a, Kyohei Ueda b, *, Kazuaki Uemura c
a
Department of Civil and Earth Resources Engineering, Kyoto University, Japan
b
Disaster Prevention Research Institute, Kyoto University, Kyoto, Japan
c
OYO Corporation, Saitama, Japan

A R T I C L E I N F O A B S T R A C T

Keywords: The Liquefaction Experiments and Analysis Project (LEAP) is a joint international project that pursues the
LEAP project verification, validation and uncertainty quantification of numerical liquefaction models. As part of this project,
Torsional shear test twenty-three Hollow Cylinder Cyclic Torsional Shear Tests were conducted in the installations of the Disaster
Liquefaction
Prevention Research Institute at Kyoto University, focusing on the influence of the relative density and the K0
Clean sands
effects in the cyclic response of Ottawa F-65 Sand. The results are intended to compose part of a reliable database
in the development of current and future V&V processes of liquefaction models.
The paper presents the characteristics of the model preparation, test development, and a brief discussion of (1)
the influence of the relative density and the K0 effects, (2) the volumetric strain due to post-cyclic reconsoli­
dation, and (3) the repeatability and consistency of the tests.

1. Introduction
Along with the development of computational power, several efforts
and developments in the computational modeling of geomaterials have
been being developed by researchers all over the world; in the last de­
cades (especially after 1964 Niigata Earthquake), special attention was
placed on the development of constitutive and numerical modeling tools
for the prediction of cyclic and permanent deformations of liquefaction
prone soils [1]. However, despite the efforts, results of numerical sim­
ulations have still important discrepancies with physical models;
therefore, exercises of verification and validation (V&V) are needed to
enhance the reliability of numerical models for liquefaction.
The Liquefaction Experiments and Analysis Project (LEAP) is a joint
international project that pursues the verification, validation and un­
certainty quantification of numerical liquefaction models, based on
high-quality centrifuge models and element tests. From its conception,
this project was envisioned to be developed as a series of asynchronous
validation exercises, each of them focusing on a certain issue related to
liquefaction phenomena [2]. Until 2019, three main exercises were
developed (including the participation of ten centrifuge facilities around
the world), and its results (including the complete data records of the
centrifuge tests) have been being published for its use in current and
future V&V processes [3–5].
The standard sand selected for the four LEAP Exercises was Ottawa F-
65; this can be described as a clean, poorly graded, whole grain silica
sand, with less than 0.5% fines by mass [6]. In order to ensure that the
same sand would be used at all facilities, sandbags from the same sup­
plier (US Silica, Ottawa, Illinois) were provided by UC Davis prior to the
development of the exercises.
Along with the development of the LEAP Exercises, researchers have
developed several efforts to increase the database of laboratory tests in
order to characterize the physical and mechanical properties of this
sand. With regard to the mechanical characterization, Vasko [7] con­
ducted a series of monotonic and cyclic triaxial tests for a certain relative
density, Parra Bastidas [8] performed monotonic and cyclic simple shear
tests for a diverse range of relative densities, El Ghoraiby et al. [9,10]
have conducted cyclic triaxial shear tests for three different relative
densities, and cyclic direct simple shear tests under two different over­
burden stresses (σ0 v ¼ 40 kPa, and σ0 v ¼ 100 kPa) for a certain relative
density.
As pointed out by Ishihara et al. [11,12], for cyclic tests, the triaxial
test apparatus usually produces greater pore pressures on the side of

* Corresponding author.
E-mail address: ueda.kyohei.2v@kyoto-u.ac.jp (K. Ueda).

https://doi.org/10.1016/j.soildyn.2020.106111
Received 19 October 2019; Received in revised form 14 February 2020; Accepted 25 February 2020
Available online 6 March 2020
0267-7261/© 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
R.R. Vargas et al. Soil Dynamics and Earthquake Engineering 133 (2020) 106111

Table 1 the characteristics of the preparation, test development, and a brief

Hollow cylinder dynamic torsional shear tests under isotropic conditions. discussion; so, the results could be used as a reliable database (along
Specimen Initial Dr after Initial mean Shear Shear with the centrifuge models and element tests developed in the LEAP
No. Dr (%) consolidation effective stress stress Exercises) in the development of current and future V&V processes of
(%) stress σ0 m0 τZθ ratio liquefaction models. It is important to mention that an initial set of nine
(kPa) (kPa) (τZθ/
tests (corresponding to the tests developed under isotropic consolidation
σ0 m0)
for Dr 50% and 60%) were developed for a numerical simulation exer­
Dr50-1-1 48.2 50.4 103.0 10.2 0.099 cise organized in the framework of LEAP-ASIA-2019, and the raw data
Dr50-1-2 48.5 50.5 104.5 13.3 0.127
Dr50-1-3 48.6 50.6 103.0 15.3 0.149
series are available at [13]; the raw data corresponding to the other
Dr50-1-4 50.9 53.1 106.1 20.3 0.191 fourteen tests was included as a “supplementary material” of this paper.
Dr60-1-1 58.7 60.7 103.0 12.1 0.117
Dr60-1-2 58.4 60.4 103.0 12.8 0.125 2. Model preparation and loading conditions
Dr60-1-3 59.1 61.0 105.0 15.1 0.144
Dr60-1-4 59.0 60.6 104.0 18.1 0.174 Twenty-three Hollow Cylinder Cyclic Torsional Shear Tests were
Dr60-1-5 61.6 63.6 104.0 20.6 0.199 conducted in order to study the stress-strain behavior of Ottawa F-65
Dr70-1-1 68.7 70.2 102.9 20.5 0.199 Sand under cyclic loadings. Sixteen tests were carried out under
Dr70-1-2 70.8 71.9 106.0 26.2 0.247 isotropic conditions under a wide range of relative densities (Dr ¼ 50%,
Dr70-1-3 73.3 73.9 106.2 26.4 0.248
60%, 70%, and 80%), with the purpose of understanding its effect;
Dr70-1-4 69.2 70.7 104.1 28.3 0.272
additionally, seven tests were developed under initial anisotropic con­
Dr80-1-1 82.3 84.0 103.1 28.3 0.274
ditions, aiming to study the K0 effects. Tables 1 and 2 shows the
Dr80-1-2 81.4 82.9 103.2 30.6 0.296
Dr80-1-3 83.3 84.9 106.5 41.6 0.390
nomenclature and characteristics of each test.
Regarding the estimation of the Dr values, Carey et al. [6] made a
statistical analysis of the maximum and minimum dry densities of
extension than on the side of compression, producing asymmetric Ottawa F-65 Sand, by using samples that were used in the centrifuge
stress-strain curves (i.e. larger strains in the extension side than in the models developed for LEAP; in this paper, the recommended values were
compression side); on the other hand, the development of simple shear used (ρmin ¼ 1490.5 kg/m3, ρmax ¼ 1757.0 kg/m3).
test allows the development of symmetric stress-strain responses, how­ As part of the planning process of the LEAP Exercises, the prepara­
ever, the initial stress conditions (i.e. initial K0 value), usually, cannot be tion method for the ground models in the centrifuge experiments was
measured. In order to contribute to the understanding of the mechanical the “Air pluviation” technique [3,5,14]; in order to obtain a fabric
properties of Ottawa F-65 Sand, a series of high-quality Hollow Cylinder similar to the physical models, the samples were prepared by using the
Cyclic Torsional Shear Tests (which allows the measurement of the stress “Air Pluviation” technique as well; in order to guarantee the homoge­
conditions at every step, and produces a symmetric stress-strain neity of the sample and reduce the variability, the height of the pluvi­
response) were conducted in the installations of the Disaster Preven­ ation tool was increased to keep the dropping-height constant along
tion Research Institute at Kyoto University, focusing on the influence of with the sample. It has been found that very small variations in the
the relative density and the K0 effects. The aim of this paper is to present weight or dimensions of the sample may affect significantly the

Table 2
Hollow cylinder dynamic torsional shear tests under anisotropic conditions.
Specimen Initial Dr Dr after K0 Initial vertical effective Initial radial effective Initial mean effective Shear stress Shear stress ratio
No. (%) consolidation (%) stress σ0 Z0 (kPa) stress σ0 r0 (kPa) stress σ0 m0 (kPa) τ (kPa) (τ/σ0 m0)

Dr50-2-1 48.3 50.3 0.50 152.70 75.70 101.4 16.4 0.162

Dr50-2-2 50.0 50.7 0.50 153.40 76.30 102.0 19.3 0.189
Dr50-2-3 49.1 51.2 0.51 155.00 78.40 103.9 23.3 0.224

Dr70-2-1 69.5 70.9 0.52 155.4 80.5 103.9 26.00 0.250

Dr70-2-2 69.5 71.1 0.51 152.1 78.0 103.9 28.50 0.274
Dr70-2-3 70.2 71.7 0.51 152.4 77.7 103.9 32.80 0.316
Dr70-2-4 71.5 73.6 0.50 151.5 76.2 103.9 36.00 0.346

Fig. 1. Cycling sinusoidal Shear Stress applied to the sample.

2
R.R. Vargas et al. Soil Dynamics and Earthquake Engineering 133 (2020) 106111

Fig. 2. (a) Hollow Cylindrical Specimen and its coordinate system (b) Stress conditions in the element under study c) Torsional Shear Apparatus at the Disaster
Prevention Research Institute – Kyoto University.

3
R.R. Vargas et al. Soil Dynamics and Earthquake Engineering 133 (2020) 106111

Table 3
Sensor list.
Variable Type Capacity Manufacturer Model

Torque Load (MT) Load Cell 20 N m Seishikou KTU-

20Ti
Vertical Load (W) Load Cell 2 kN Seishikou KTU-
20Ti
Angular Deformation Potentiometer �20� Midori CP-
(θ) 2UTX
Volumetric Pressure �3200 Validyne DP15-
Deformation (εv) Sensor psi 28
Pore Pressure (Δu) Pressure 1 Mpa TEAC TP-BR
Sensor
Vertical Strain (εa) Dial Gauge 20 mm Showa Sokki TCL-
20FA

Fig. 4. Liquefaction Resistance curve for Isotropically Consolidated Samples of

Ottawa F-65 Sand for γDA ¼ 3.0%.

Fig. 5. Liquefaction Resistance curve for Isotropically Consolidated Samples of

Ottawa F-65 Sand for γDA ¼ 1.5%.

Fig. 3. (a) Liquefaction Resistance curve for Isotropically Consolidated Samples

of Ottawa F-65 Sand for γDA ¼ 7.5% (b) Relationship between RL10, RL20, RL30
and Dr for γDA ¼ 7.5%.

estimated relative density; in order to ensure a consistency among the

experiments, multiple and precise measurements were taken; also, a
repeatability test was developed, showing good agreement between the
tests (See Section 5).
After finishing the air pluviation process, negative pressure ( 15 kPa
approximately) was applied to the sample to keep its shape while the cell
chamber was placed. After filling the outer and inner cells with water, an
isotropic 20 kPa pre-consolidation was applied to the sample in order to
avoid deformations in the following steps. After that, the saturation
Fig. 6. Liquefaction Resistance curve for Isotropically Consolidated Samples of
Ottawa F-65 Sand for ru ¼ 0.95.
process was conducted by flowing CO2 gas and degassed water for
approximately 40 min; following the saturation process, 200 kPa of back
pressure was applied to the sample. The saturation level was then
checked by measuring the Skempton’s B value; it has been verified that
in every test presented in this paper, the B value was 0.95 or higher, for

4
R.R. Vargas et al.
Fig. 7. (a) Time History of EPWP ratio development - Model Dr50-1-3 (b) Stress path - Model Dr50-1-3 (c) Stress-Strain response - Model Dr50-1-3.
an increment of 20 kPa under isotropic and undrained conditions.
Following the saturation process, for the isotropically consolidated
tests, the specimen was isotropically consolidated to a total pressure of
approximately 300 kPa (i.e. approximately 100 kPa of effective stress).
After the consolidation process, a stress-controlled cycling sinusoidal
torque was applied to the sample in undrained conditions (Fig. 1 shows
the target wave); as specified in the Japanese Standard JGS-0541-2009
[15], the load was applied until achieving at least 7.5% of double
amplitude shear strain. It is important to mention that because the loads
were applied in a stress-controlled fashion, reliable information could
not be obtained at very large strains.
5
Soil Dynamics and Earthquake Engineering 133 (2020) 106111
As for the anisotropically consolidated tests, after measuring the B
value, the specimen was consolidated under drained conditions, by
applying small increments of the vertical, radial and tangential stresses,
in order to keep anisotropic conditions during the consolidation process
(i.e.Δσ’r ¼ Δσ’θ ¼ K0 Δσ’Z ). The small increments were applied until
achieving approximately σz0 ¼ 350 kPa, and σr0 ¼ σθ0 ¼ 275 kPa (i.e.
σ’Z0 ¼ 150 kPa, σ’r0 ¼ σ’θ0 ¼ 75 kPa, and ¼ σ’m0 ¼ 100 kPa). The
loading process was similar as the one described for the isotropically
consolidated tests; however, as pointed out by Ishihara et al. [12] and
Sawada et al. [16], in order to reproduce a correct mode of deformation
taking place in the field under seismic loading, the lateral deformations
R.R. Vargas et al.
Fig. 8. (a) Time History of EPWP ratio development - Model Dr80-1-2 (b) Stress path - Model Dr80-1-2 (c) Stress-Strain response - Model Dr80-1-2.
during the test should be restrained; this was achieved by restraining the
vertical displacements by using a mechanical locking device, and
monitoring that no volumetric strains were generated.
It worth noting that, as shown in Tables 1 and 2, densification be­
tween 1 and 2% was observed at the end of the consolidation process for
both isotropically and anisotropically consolidated tests. Additionally,
in order to measure the induced volumetric strains triggered by lique­
faction, a re-consolidation process was developed until no changes in the
pore water pressure nor volumetric changes were recorded.
It is important to mention that the specimens were prepared
following the Japanese Standards JGS-0550-2009 [17], and
6
Soil Dynamics and Earthquake Engineering 133 (2020) 106111
JGS-0551-2009 [18].
3. Torsional Shear Apparatus
Fig. 2 shows the hollow cylindrical specimen, the stress conditions in
the element under study, and the Hollow-Cylinder Torsional Shear
Apparatus in which the samples were prepared and tested. As seen in the
figure, the hollow cylindrical sample has an inner diameter of 6 cm, an
outer diameter of 10 cm and a height of 10 cm; this sample can be tested
under a diverse range of stress paths for monotonic and cyclic tests. The
apparatus allows the application of the vertical load and torque through
R.R. Vargas et al. Soil Dynamics and Earthquake Engineering 133 (2020) 106111

Fig. 9. Liquefaction Resistance curve for Anisotropically Consolidated Samples Fig. 12. Liquefaction Resistance curve for Anisotropically Consolidated Sam­
(K0 ¼ 0.50) of Ottawa F-65 Sand for γDA ¼ 7.5%. ples (K0 ¼ 0.50) of Ottawa F-65 Sand for ru ¼ 0.95.

result in different axial strains due to the Poisson’s effects), the same
inner and outer pressures were applied at all steps during the tests (i.e.
σ’r ¼ σ’θ ).
The volumetric strains were measured by means of a low-capacity
differential pressure transducer that measured the water changes in a
burette connected to the backpressure (kept constant at all times at
around 200 kPa) and the specimen. Additionally, the rotational and
axial deformations were measured by a potentiometer and a dial gauge
respectively.
Table 3 shows the measured variables and the characteristics of the
sensors used for its measurement.
As stated by Ishihara et al. [12] and Iddriss and Boulanger [20], for
tests under anisotropic consolidation, the Cyclic Stress Ratio (CSR) can
be defined as (1) the applied shear stress normalized by the mean
effective stress (τZθ =σ’m0 ), or (2) the applied shear stress normalized by
the vertical effective stress (τZθ =σ’Z0 ). In order to establish a comparison
between the isotropically and anisotropically consolidated tests, in this
paper, the τZθ =σ’m0 expression will be used as the definition of the CSR.
Fig. 10. Liquefaction Resistance curve for Anisotropically Consolidated Sam­
ples (K0 ¼ 0.50) of Ottawa F-65 Sand for γDA ¼ 3.0%. It is important to mention that, for Cyclic Torsional Shear Tests, the
mean effective stress (σ0 m0) is defined as follows:
σ’Z0 þ σ’r0 þ σ’θ0
σ’m0 ¼ (1)
3

Where σ’Z0 is the initial vertical stress, σ’r0 is the initial radial stress, and
σ’θ0 is the initial tangential stress.

4. Test results

As mentioned in previous sections, twenty-three Hollow Cylinder

Cyclic Torsional Shear Tests were conducted, and the stress-strain
characteristics of each model were estimated based on the measured
variables (see Table 3), and following the Japanese Standard JGS-0551-
2009 [18].

4.1. Isotropically consolidated tests

Fig. 3a shows the liquefaction resistance curve for Isotropically

Consolidated Samples of Ottawa F-65 sand for γDA ¼ 7.5% (i.e. it shows
Fig. 11. Liquefaction Resistance curve for Anisotropically Consolidated Sam­
the number of cycles required to reach a 7.5% double amplitude shear
ples (K0 ¼ 0.50) of Ottawa F-65 Sand for γDA ¼ 1.5%.
strain); a clear trend and consistency among the tests can be observed. It
is important to mention that the dotted lines shown in Fig. 3a correspond
independent pneumatic cylinders; also, as suggested by Koseki et al.
to extrapolations of the observed trends (following an exponential
[19], the load cell to measure the vertical load and torque was set inside
equation), and hence, are presented only as a reference.
the pressure cell in order to eliminate the effects of friction between the
Fig. 3b shows the relation between the RL10, RL20, and RL30 values (i.
loading shaft and the bearing house.
e. the shear stress ratio that would induce γDA ¼ 7.5% in 10, 20, and 30
To avoid non-uniform normal stresses across the sample (which
cycles) and the relative density (Dr). It is observed that a second order

7
R.R. Vargas et al. Soil Dynamics and Earthquake Engineering 133 (2020) 106111

Fig. 13. (a) Time History of EPWP ratio development - Model Dr50-2-2 (b) Stress path - Model Dr50-2-2 (c) Stress-Strain response - Model Dr50-2-2.

8
R.R. Vargas et al. Soil Dynamics and Earthquake Engineering 133 (2020) 106111

Fig. 14. (a) Time History of EPWP ratio development - Model Dr70-2-3 (b) Stress path - Model Dr70-2-3 (c) Stress-Strain response - Model Dr70-2-3.

9
R.R. Vargas et al.
Fig. 15. (a) Time History of the applied Axial Load - Model Dr50-2-1 (b) Time History of the Effective Vertical and Radial stresses (σ ’Z and σ ’r respectively) - Model
Dr50-2-1.
polynomial relationship fits the data points for each RL value; it is worth
mentioning that, as stated by Ishihara [21] and Towhata [22], the CSR
required to cause γDA ¼ 7.5% under 20 load cycles (i.e. RL20) is used in
the routine practice as a factor that quantifies the liquefaction resistance
of a sand.
In a similar way the liquefaction resistance curves for γDA ¼ 3.0%,
γDA ¼ 1.5%, and ru ¼ 0.95 are presented in Figs. 4–6 respectively. It has
been found that, for all the isotropically consolidated tests, irrespective
of the CSR and the density, the development of ru ¼ 0.95 coincide with
the γDA ¼ 3.0% condition (see Figs. 4 and 6).
In order to briefly discuss the effect of the relative density on the
behavior of Ottawa F-65 Sand, samples Dr50-1-3 (τZθ /σ0 m0 ¼ 0.15) and
Dr80-1-2 (τZθ /σ 0 m0 ¼ 0.30) were selected. Fig. 6 shows that despite the
difference in the CSR, both samples require a similar number of cycles
(around 15 cycles) to develop ru ¼ 0.95. Fig. 7(a–b) and 8(a-b) shows the
time history of ru development and the stress path respectively; for the
sample Dr50-1-3, it can be seen that the ru value reach the unity only
half cycle after reaching a 0.95 value, meanwhile the sample Dr50-1-3
requires more than 6 cycles; this feature, along with the rate of strain
generation, as seen in the stress-strain response (Figs. 7(c) and 8 (c)),
keeps consistency with the fact that the state parameter (ξ) increases
with higher Dr values [23].
4.2. Anisotropically consolidated tests
Figs. 9–12 shows the liquefaction resistance curve for Anisotropically
Consolidated Samples of Ottawa F-65 sand (K0 ¼ 0.50) for γDA ¼ 7.5%,
γDA ¼ 3.0%, γDA ¼ 1.5%, and ru ¼ 0.95 respectively. In a similar way, as
found in the isotropically consolidated tests, a clear trend and consis­
tency among the tests were found. Moreover, the dotted lines shown in
Fig. 9 correspond to extrapolations of the observed trends (following an
10
Soil Dynamics and Earthquake Engineering 133 (2020) 106111
exponential equation), and are presented only as a reference.
As in the case of the isotropically consolidated tests, it has been found
that also, in all of the anisotropically consolidated cases (K0 ¼ 0.50),
irrespective of the CSR and the density, the development of ru ¼ 0.95
coincide with the γDA ¼ 3.0% condition (see Figs. 10 and 12).
Figs. 13 and 14 shows the time history of ru development, the stress
path and the stress-strain history of samples Dr50-2-2 (τZθ/σ 0 m0 ¼ 0.19)
and Dr70-2-3 (τZθ/σ0 m0 ¼ 0.32) respectively; as in the case of isotropi­
cally consolidated samples, these two elements were selected because,
despite the difference in the CSR, both samples require a similar number
of cycles (around six cycles) to develop ru ¼ 0.95. A similar behavior has
been found, with respect to the further generation of EPWP and strain
rates, as described for the isotropically consolidated samples.
As pointed out in Section 2, in order to reproduce a correct mode of
deformation taking place in the field under seismic loading, the gener­
ation of vertical strains were not allowed on the tested samples during
the cyclic process; this caused a successive reduction of the applied axial
load (see Fig. 15a). Since the radial effective stress (and hence the
tangential effective stress) decreased at a slower rate than the vertical
effective stress see Fig. 15b, a progressive increment of the anisotropic
ratio (σ ’Z =σ’r ), from K0 ¼ 0.50 in the initial state until achieving a nearly
isotropic state after reaching liquefaction, was observed.
It is important to mention that, as reported by Ishihara et al. [24], the
transition from the K0 value (prior shaking) to isotropic states (after
reaching liquefaction) can be encountered during earthquakes in
level-ground saturated soil deposits.
4.3. Comparison between isotropically consolidated tests and
anisotropically consolidated tests
Fig. 16a shows a comparison between the liquefaction resistance
R.R. Vargas et al. Soil Dynamics and Earthquake Engineering 133 (2020) 106111

Fig. 16. (a) Comparison of the Liquefaction Resistance curves for Isotropically and Anisotropically Consolidated Samples of Ottawa F-65 Sand for γDA ¼ 7.5% (b)
Comparison of the Liquefaction Resistance curves for Isotropically and Anisotropically Consolidated Samples (CSR values multiplied by 1/1.2) of Ottawa F-65 Sand
for γDA ¼ 7.5%.

11
R.R. Vargas et al.
Fig. 17. Comparison of the Liquefaction Resistance curves for Isotropically and
Anisotropically Consolidated Samples of Toyoura Sand for γDA ¼ 5.0% [25].
curves (γDA ¼ 7.5%) of the isotropically consolidated tests (Dr ¼ 50%
and Dr ¼ 70%) and the anisotropically consolidated tests; by comparing
the results, it has been found that the liquefaction resistance of the
anisotropically consolidated tests is approximately 20% higher than the
tests developed under isotropic consolidation, irrespective of the density
and the strain level (for γDA ¼ 1:5 7:5 %). Fig. 16b shows that if the
CSR values of the anisotropically consolidated tests are multiplied by a
factor of 1/1.2, the results lie on the liquefaction resistance curve of the
isotropically consolidated tests. In fact, as seen in Fig. 17, a similar
behavior was found by the Public Works Research Institute of Japan
[25], whom have conducted similar experiments by using Dr ¼ 60%
samples of Toyoura Sand under σ ’m0 � 100kPa. It is worth mentioning
that further efforts need to be developed in order to study the effect of
other key factors (such as different K0 values, initial fabric, different σ ’m0
values, etc.) that might allow a comprehensive understanding of the
mechanism behind the increase in the liquefaction resistance exhibited
by the anisotropically consolidated clean sands.
4.4. Volumetric strain due to post-cyclic reconsolidation
As mentioned in Section 2, in order to measure the induced volu­
metric strains triggered by liquefaction, after applying the cyclic
loading, a re-consolidation process was developed by opening the
drainage line of the test sample in order to dissipate the developed re­
sidual pore water pressure, and the volumetric changes were recorded
until no changes in the pore water pressure were recorded. The same
procedure was used by Refs. [26,27] to develop simplified procedures to
estimate liquefaction-induced settlements. Fig. 18a shows the correla­
tion between the volumetric strain due to post-cyclic reconsolidation
(εv) and the maximum single-amplitude shear strain during undrained
loading (γmax), proposed by Ishihara et al. [26]; despite the scatter
observed in the data, a good agreement was obtained for Dr > 60%. On
the other hand, Sento et al. [27] proposed a correlation between εv and
the cumulative shear strain during undrained loading (γ acm ¼
Z t
j_γðtÞjdt); as seen in Fig. 18b, a good agreement was found as well
0 CRediT authorship contribution statement
between the experimental data and the proposed correlation for Dr >
60% (by considering a ¼ 0:32Dr þ 30:5 and b ¼ 0:50, “a” and “b”
correspond to internal variables in the correlation [27]).
5. Repeatability of test results
In order to check the consistency of the tests, the specimen Dr70-1-3
was prepared and tested following under the same conditions as the
specimen Dr70-1-2, aiming to be a “repeatability test”.
12
Soil Dynamics and Earthquake Engineering 133 (2020) 106111
Fig. 19 shows a comparison between the above-mentioned tests in
terms of the development of Excess Pore Water Pressure ratio (EPWP
ratio), development of Shear Strain, and the corresponding stress path. A
good agreement between the tests can be observed, showing that the
tests are repeatable and consistent among each other.
6. Conclusions
Liquefaction Experiments and Analysis Project (LEAP) is a joint in­
ternational project that pursues the verification, validation and uncer­
tainty quantification of numerical liquefaction models, based on high-
quality centrifuge models and element tests. Until 2019, three main
exercises were developed and its results (including the complete data
records of the centrifuge tests) have been being published for its use in
current and future V&V processes. Since the sand selected for the four
LEAP Exercises was Ottawa F-65 Sand, this paper aims to contribute to
the understanding of its mechanical properties through a series of high-
quality Hollow Cylinder Cyclic Torsional Shear Tests, focusing on the
influence of the relative density and the K0 effects. The characteristics of
preparation, test development, and a brief discussion have been pre­
sented; so, the results could be used as a reliable database (along with
the centrifuge models and element tests developed in the LEAP Exer­
cises) in the development of future V&V processes of liquefaction
models.
� Based on sixteen isotropically consolidated tests, the stress-strain
characteristics of Ottawa F-65 Sand were investigated under four
different densities and a wide range of shear stress ratios. These tests
were complemented through seven anisotropically consolidated tests
(K0 ¼ 0.50), in order to study the K0 effect.
� It has been found that the RL20 values (for γDA ¼ 7.5%) and the
relative density (Dr), can be correlated by means of a second-order
polynomial relationship.
� In order to reproduce a correct mode of deformation taking place in
the field under seismic loading, the lateral deformations, during an
anisotropically consolidated test, should be restrained; this process
caused a successive reduction of the vertical stress and an increase of
the anisotropic ratio (σ ’Z =σ’r ) value from K0 ¼ 0.50 in the initial
state, until achieving a nearly isotropic state after reaching
liquefaction.
� It has been found that the liquefaction resistance of the anisotropi­
cally consolidated tests (K0 ¼ 0.5) is approximately 20% higher than
the tests developed under isotropic consolidation, irrespective of the
initial density and the strain level (for γ DA ¼ 1:5 7:5 %).
� Regarding the volumetric strain due to post-cyclic reconsolidation
(εv), experimental results keep a good agreement with the correla­
tions proposed by Ishihara et al. [26] and Sento et al. [27], for
Dr>60%.
� The good agreement in the repeatability tests shows that the tests are
repeatable and consistent among each other.
Declaration of competing interest
To the best of our knowledge, the named authors have no conflict of
interest, financial or otherwise.
Ruben R. Vargas: Validation, Software, Formal analysis, Investi­
gation, Data curation, Writing - original draft, Visualization. Kyohei
Ueda: Conceptualization, Methodology, Writing - review & editing,
Supervision. Kazuaki Uemura: Methodology, Validation, Investigation,
Writing - review & editing.
R.R. Vargas et al. Soil Dynamics and Earthquake Engineering 133 (2020) 106111

Fig. 18. (a) Comparison of the volumetric strain due to post-cyclic reconsolidation with the correlations proposed by Ishihara et al. [26] (b) Comparison of the
volumetric strain due to post-cyclic reconsolidation with the correlations proposed by Sento et al. [27].

13
R.R. Vargas et al. Soil Dynamics and Earthquake Engineering 133 (2020) 106111

Fig. 19. (a) Development of EPWP – Models Dr70_2 and Dr70_3 (b) Development of Shear Strain – Models Dr70_2 and Dr70_3 (c) Stress path – Models Dr70_2
and Dr70_3.

Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.soildyn.2020.106111.
References
[1] Manzari MT, Kutter BL, Zeghal M, Iai S, Tobita T, Madabhushi SPG, et al. LEAP
projects: concept and challenges” geotechnics for catastrophic flooding events. CRC
Press; 2015. p. 109–16.
[2] Kutter BL, Manzari MT, Zeghal M, Zhou YG, Armstrong RJ. “Proposed outline for
LEAP verification and validation processes” geotechnics for catastrophic flooding
events. CRC Press; 2015. p. 99–108.
[3] Kutter B, Carey T, Hashimoto T, Zeghal M, Abdoun T, Kokalli P, et al. LEAP-GWU-
2015 experiment specifications, results, and comparisons. Soil Dynam Earthq Eng
2018;113.
[4] Kutter B, Carey T, Stone N, Zheng BL, Gavras A, Manzari M, et al. LEAP-UCD-2017
comparison of centrifuge test results” model tests and numerical simulations of
liquefaction and lateral spreading. Cham: Springer; 2019. p. 69–103. https://doi.
org/10.1007/978-3-030-22818-7_4.
[5] Tobita T, Vargas R, Ichii K, Okamura M, Sjafruddin AN, Takemura J, Lyu H, Ueda
K, Uzuoka R, Iai S, Boksmati J, Fusco A, Torres-Garcia S, Haigh S, Madabhushi G,
Manzari M, Escoffier S, Li Z, Kim DS, Manandhar S, Hung WY, Huang JX, Pham
TNP, Zeghal M, Abdoun T, Korre E, Kutter B, Carey T, Stone N, Zhou YG, Liu K, Ma
Q. “LEAP-ASIA-2019: validation of centrifuge experiments and generalized scaling
law on liquefaction-induced lateral spreading” submitted to Soil Dynamics and
Earthquake Engineering [under review].
[6] Carey TJ, Stone N, Kutter BL. “Grain size analysis and maximum and minimum dry
density of Ottawa F-65 sand for LEAP-UCD-2017”, model tests and numerical
simulations of liquefaction and lateral spreading. Cham: Springer; 2019. p. 31–44.
https://doi.org/10.1007/978-3-030-22818-7_2.
[7] Vasko A. An investigation into the behavior of Ottawa sand through monotonic and
cyclic shear tests. MS Thesis. George Washington University; 2015.

14
R.R. Vargas et al.
[8] Parra Bastidas AM. Ottawa F-65 sand characterization. PhD Dissertation. Davis:
University of California; 2016.
[9] El Ghoraiby MA, Park H, Manzari MT. Physical and mechanical properties of
Ottawa F65 sand, model tests and numerical simulations of liquefaction and lateral
spreading. Cham: Springer; 2019. p. 45–67. https://doi.org/10.1007/978-3-030-
22818-7_3.
[10] El Ghoraiby MA, Manzari MT. LEAP-2018 - stress-strain response of Ottawa F65
sand in cyclic simple shear. DesignSafe-CI; 2018. https://doi.org/10.17603/ds2h
x3h.
[11] Ishihara K, Yasuda S. Sand liquefaction in hollow cylinder torsion under irregular
excitation. Soils Found 1975;15(1).
[12] Ishihara K, Yamazaki A, Haga K. Liquefaction of Ko-consolidated sand under cyclic
rotation of principal principal stress direction with lateral constraint. Soils Found
1985;25(4).
[13] Ueda K, Vargas R, Uemura K. LEAP-Asia-2018: stress-strain response of Ottawa
sand in cyclic torsional shear tests. DesignSafe-CI; 2018. https://doi.org/10.1760
3/DS2TH7Q.
[14] Kutter Bruce L, Carey Trevor J, Stone Nicholas, Hajialilue Bonab Masoud,
Manzari Majid, et al. LEAP-UCD-2017 V. 1.01 model specifications”, model tests
and numerical simulations of liquefaction and lateral spreading. Cham: Springer;
2018. p. 3–29. https://doi.org/10.1007/978-3-030-22818-7_1.
[15] Japanese Geotechnical Society Standards. Method for cyclic triaxial test on soils -
JGS-0541-2009, vol. 2. Laboratory Testing Standards of Geomaterials; 2016.
[16] Sawada S, Sakuraba R, Ohmukai N, Mikami T. Effect of K0 on liquefaction strength
of silty sand. In: Fourth international conference on recent advances in
geotechnical earthquake engineering and soil dynamics. Missouri: EEUU; 2001.
15
Soil Dynamics and Earthquake Engineering 133 (2020) 106111
[17] Japanese Geotechnical Society Standards. Practice for preparing hollow cylindrical
specimens of soils for torsional shear test – JGS-0550-2009, vol. 3. Laboratory
Testing Standards of Geomaterials; 2018.
[18] Japanese Geotechnical Society Standards. Method for torsional shear test on
hollow cylindrical specimens of soils - JGS-0551-2009, vol. 3. Laboratory Testing
Standards of Geomaterials; 2018.
[19] Koseki J, Yoshida T, Sato T. Liquefaction properties of Toyoura sand in cyclic
torsional shear tests under low confining stress. Soils Found 2005;vol. 45. 5.
[20] Idriss IM, Boulanger RW. Soil liquefaction during earthquakes. Monograph MNO-
12. Oakland, CA: Earthquake Engineering Research Institute; 2008.
[21] Ishihara K. Soil behaviour in earthquake geotechnics. Oxford University Press;
1996, ISBN 0-19-85-6224-1.
[22] Towhata I. Geotechnical earthquake engineering. Springer-Verlag Berlin
Heidelberg; 2008, ISBN 978-3-540-35783-4.
[23] Been K, Jefferies G. A state parameter for sands” Geotechnique, vol. 15; 1985. Issue
2.
[24] Ishihara K, Iwamoto S, Yasuda S, Takatsu H. Liquefaction of anisotropically
consolidated sand. In: Proc. 9th international conference on soil mechanics and
foundation engineering, vol. 2; 1977. p. 261–4.
[25] Public Works Research Institute - Japan. Soil dynamics group – report No. 70. 1992
[In Japanese].
[26] Ishihara K, Yoshimine M. Evaluation of settlements in sand deposits following
liquefaction during earthquakes. Soils Found 1992;32(1).
[27] Sento N, Kazama M, Uzuoka R. Experimental and idealization of the volumetric
compression characteristics of clean sand after undrained cyclic shear. J Jpn Soc
Civ Eng 2004;764(III-67):307–17 (In Japanese).