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.
Monotonic and cyclic behavior of thickened copper tailings
Comportement monotone et cyclique de résidus de cuivre épaissis

Urbano, Sebastián; Puma, Daniela; Ovalle, Carlos, Sáez, Esteban

Department of Structural and Geotechnical Engineering and Center for Integrated Natural Disaster Management
CONICYT/FONDAP/15110017, Pontificia Universidad Católica de Chile, Chile, [email protected]

ABSTRACT: This paper presents the results of a series of drained and undrained monotonic and cyclic triaxial tests, as well as
resonant column and cyclic torsion tests carried out on samples of thickened tailings from a copper mine in operation in northern
Chile. Samples were remoulded at the in-situ density, with the aim of characterizing the mechanical behavior and to obtain the
degradation curves of shear modulus and damping in a wide range of distortions. This information will allow for studying the seismic
behavior of the deposit in advanced numerical models. It is observed that the cyclic resistance of the material and post- liquefaction
resistance are higher than reported data for other thickened tailings, and the degradation shear modulus and damping increase curves
(Silty Clay) are similar to those reported for natural non-plastic fines and sands.

KEYWORDS: thickened tailings; liquefaction, cyclic behavior.

RÉSUMÉ: Cet article présente les résultats d’une série d’essais triaxiaux monotones et cycliques drainés et non drainés, ainsi que des
essais de colonne résonante et de torsion cyclique effectués sur des échantillons de résidus miniers épaissis provenant d’une mine de
cuivre en exploitation dans le nord du Chili. Les échantillons ont été remaniés à la densité in-situ, dans le but de caractériser le
comportement mécanique et d'obtenir les courbes de dégradation du module de cisaillement et d'amortissement dans une large gamme de
distorsion angulaire. Cette information permettra d'étudier le comportement sismique du dépôt de résidus à l’aide des modèles
numériques avancés. On observe que la résistance cyclique du matériau et la résistance à la liquéfaction sont plus élevées que les données
rapportées par d'autres auteurs, et que le module de cisaillement de dégradation et les courbes d'incrément d'amortissement sont
semblables à celles rapportés pour des sols fins non-plastiques et pour des sables.

MOTS-CLÉS : résidus épaissis, liquéfaction, comportement cyclique.

sedimentation in a mechanical equipment. The thickened slurry

1 INTRODUCTION is then hydraulically deposited and forms cones of soft slopes of
Over the last two decades, the world-wide demand for copper around 2 to 4% (Robinsky 1999). Thus, the gain in process water
has been rising steadily. Hence, some of the large copper recovery is made at the thickening plant, before tailings
recovery operations (>100,000 tonnes per day) in Chile, mainly discharge. Since free water is almost null in the deposit, there is
located in the high Andes Mountains of the Atacama Desert, are almost no sedimentation neither segregation, so the water
looking to expand their production. Although rich in minerals, discharged is hardly recoverable and water is lost by evaporation,
the Atacama Desert is extremely dry. seepage and retention into the tailings voids.
Typically, after the mineral recovering process 99% of the Advantages of this method could be: improvement of slope
rock extracted from the mine is waste in the form of slurry stability, reduction of the confining dam height, deposited
tailings containing about 50% of water in weight. Copper volume, and liquefaction potential (Robinsky 1999). However,
tailings are typically a mix of sand and non-plastic silt. The most even at the shrinkage limit, typical sandy-silt copper tailings
common tailings management method is the hydraulic discharge could stay quite loose and contractive (Ishihara et al. 1980,
of the slurry into a storage facility confined by dams, containing Cifuentes & Verdugo 2007, Santos 2011). Therefore, to ensure
loose and saturated tailings. Under these conditions, earthquakes seismic stability it is necessary to evaluate case by case the
could trigger tailings liquefaction and there is a well-documented vulnerability of flow failure and/or cyclic mobility.
history of earthquake-induced tailings dam failures (Dobry and TTD projects in the world have been mainly developed in low
Alvarez 1967, Villavicencio et al. 2013). Liquefaction is the to medium tailings rate production mines located in regions with
second cause of tailings dams failures, 90% occurs in active low seismic hazards (Jewell et al. 2002). However, during the
mines, United States and Chile lead the ranking of reported cases past 10 years, large mines in Chile have been interested in TTD
(Rico et al. 2008). in order to save water and today some large projects are starting
The optimization of tailings dams operation deals mainly operation. Hence, there is still no reported cases of the seismic
with maximizing the storage volume, minimize the dam high and response and experimental data related to the mechanical
volume, and maximize water recovery from the pond in order to properties of the in-situ material is quite limited. Moreover,
minimize the amount of fresh water required. However, water predictive methods for liquefaction potential evaluation and
storage in open ponds is subjected to seepage and high engineering designs are in general quite simplified (Poulos et al.
evaporation rates. To avoid high water losses in the deposit, an 1985, Palma et al. 2007). Consequently, there is a need of a more
alternative is to recover more water at the process plant by sophisticated analysis to predict the seismic behavior, including
thickening the tailings up to a dense slurry with almost no free undrained and dynamic behavior of the tailings.
water, but still able to flow into the deposit. This method is well- This paper describes an accurate experimental program for
known as the Thickening Tailings Deposition (TTD) and has material characterization, in order to obtain monotonic and
been originally proposed by Robinsky (1982) in Canada. cyclic properties. These properties are used for constitutive
Thickening is assessed by flocculation and accelerated model calibration implemented in an advanced numerical FEM

- 1585 -
 Proceedings of the 19th International Conference on Soil Mechanics and Geotechnical Engineering, Seoul 2017

model that considers cyclic behavior. The results intend to be

used in future analysis of seismic stability related to large
thickened tailings deposits.

2 EXPERIMENTAL PROGRAM
The material was taken from a thickened tailing operation
located in Atacama Desert. In particular, the material extraction
zone was near to discharge. A total of 8 bucket with 25 liters of
storage each one were collected from that point. Tailings on this
area are discharged with a 65% of solids (weight of solids / total
weight), in situ data of surface tailings shows an average dry
density of 15.2 kN/m3, obtained from 20 sand cone tests (ASTM
D1556).
Experimental laboratory program started carrying out grain
size distribution test to four different samples. Figure 1 shows
the results, according to ASTM D2487 standard, the tailings
classify as clayey silt (CL-ML), Plastic Index of 6, Liquid Limit
of 24, fines content of 65 to 75% (finer than 0.074 mm) and
specific gravity (Gs) of 2.76. From Standard Proctor compaction
curve, the maximum dry density (γdmax) was 17.3 kN/m3 and an Figure 2. Monotonic triaxial compression tests and critical state analysis
optimum water content of 16%.

100

80
% Passing by mass

60

40 M-1
M-2
20 M-3
M-4
0
0.001 0.01 0.1 1
Figure 3. Excess pore pressure in monotonic undrained triaxial tests
Particle size (mm)

Figure 1. Grain size distribution

Sample preparation for cilyndrical samples preparation was

achieved by wet-tamping at a dry density of 15.2 kN/m3,
corresponding to the shrinkage limit and a medium loose state of
88% of the maximum dry density from Standard Proctor.
Cylindrical samples were 100 mm height and 50 mm in
diameter for monotonic tests, and 140 mm height and 70 mm in
diameter for cyclic tests. Each sample was prepared by five
compacted layers with a water content of 18% in a rigid mold
covered by a latex membrane. Samples were saturated first with
CO2 gas and later with de-aired water up to a Skempton
parameter (B) of at least 95%. Figure 4. Undrained shear strength in triaxial tests

2.1 Monotonic tests 2.2 Cyclic tests

Strain controlled isotropically consolidated drained (TMD) and
undrained (TMU) triaxial monotonic compression tests at 2.2.1 Cyclic liquefaction triaxial test
effective confining pressures (σ’3) of 100, 300 and 500 kPa were Cyclic extension-compression undrained triaxial tests (TCU)
performed using a vertical displacement rate of 0.05 mm/min. were carried out controlling the deviatoric stress amplitude at
Deviatoric (q) and effective mean (p’) stresses are defined as: constant load frequency of 0.05 Hz. The imposed cyclic load was
q=σ’1-σ’3 and p’=(σ’1+2σ’3)/3. Figure 2 presents effective stress chosen for a range of Cyclic Shear Ratio (CSR=q/(2σ’3)), in
and strain paths, and the common critical state line (CSL) for order to obtain data allowing to represent the response in a
drained and undrained tests. As expected, medium loose tailings typical range of number of cycles for liquefaction between 10
behaved with excess pore pressure in the range of 20 to 30% of and 100. Figure 5 shows a typical result of a TCU test up to
the total mean stress (Figure 3). Figure 4 shows the relation liquefaction at cycle number 24, for a liquefaction criteria based
between the undrained shear strength (Su=q/2) and σ’3, with a on 10% of axial strain in double amplitude. Figure 6 presents the
unique ratio Su/σ’3=0.56 at critical state. results for all TCU tests. Characteristic CSR values associated to
20 cycles for liquefaction typically used a simplified method
(Seed, 1987) is 0.3.

- 1586 -
 Technical Committee 203 / Comité technique 203

0.4

0.2
CSR

0
0 5 10 15 20 25
-0.2

-0.4

120

90
u (kPa)

60

30

0
0 5 10 15 20 25
-30

20

10
a (% )

0
0 5 10 15 20 25
-10

-20
Number of Cycle
Figure 5. Cyclic undrained test result for σ’3=100 kPa and CSR=0.28

Figure 7. Normalized shear modulus degradation and damping

3 NUMERICAL MODELING
The Ecole Centrale Paris (Hujeux, 1985) elasto-plastic multi-
mechanism model was used because its ability to reproduce large
range of deformation and cyclic behavior. This effective-stress
model can take into account the soil behavior in a large range of
deformations. The representation of all irreversible phenomena is
made by four coupled elementary plastic mechanisms: three
plane-strain deviatoric plastic deformation mechanisms in three
orthogonal planes, and an isotropic one. The model uses a
Coulomb-type failure criterion and the critical-state concept. The
Figure 6. Undrained cyclic resistance in TCU tests evolution of hardening is based on the plastic strain (deviatoric
and volumetric strains for the deviatoric mechanisms, and
2.2.2 Cyclic degradation tests volumetric strain for the isotropic one). A kinematical hardening,
based on the state variables at the last load reverse, is used to
Cyclic stiffness degradation and hysteretic damping of the take into account the cyclic behavior. The soil behavior is
thickened tailings samples were measured in cyclic shear tests at decomposed into pseudo-elastic, hysteretic, and mobilized
different confinement (100 and 300 kPa) and very low to large domains. Calibrated material parameters are displayed in Table 1.
shear strains. Tests performed include resonant column (RC), Figures 6, 7 and 8 present the calibration of different tests at 100
torsional shear (TS) and drained cyclic triaxial (TCD). Shear and 300 kPa of initial confinement.
strain ranges used were 10-4 to 10-2 for RC, 10-2 to 10-1 for TS
and larger than 10-1 for TCD. According to ASTM D3999 and Table 1: Calibrated parameters of the ECP model.
D4015, shear stiffness (G) and shear strain (γ) were obtained Parameter name Value Meaning
after 40 cycles. At low strain (γ=10-3) the maximum shear
(MPa) 192.5 Reference shear modulus
stiffness (Go) takes values of 130 and 207 MPa at confinements
of 100 and 300 kPa, respectively. These values are typical for (MPa) 280.6 Reference bulk modulus
medium loose silty sands. Normalized stiffness (G/Go) and Degree of non-linearity for
damping (D) values are presented in Figure 7, where both results 0.43 elastic modulus evolution
are compared with the data published by Seed and Idriss (1970) with confinement
for sands, and Vucetic and Dobry (1991) for non-plastic fine (MPa) Reference confinement for
1.0
soils. There is a good agreement between both curves and results non-linear elasticity
obtained. 30.5 Friction angle at critical state
30 Characteristic angle (limit

- 1587 -
 Proceedings of the 19th International Conference on Soil Mechanics and Geotechnical Engineering, Seoul 2017

between contractive and Undrained strength ratio Su/σ’3 resulted of 0.56 and liquefaction
dilative behavior) resistance of 0.3 using Seed (1987) method. Both measures are
Plastic compressibility higher if are compared with others tailings tests densified by
30.8 (related to density’s drying (Cifuentes and Verdugo 2007, Santos 2011, Osorio 2009).
hardening) This, could be explained by the high amount of fines (Puma et al
Distance between virgin 2016), which would allow high density by shrinkage drying.
isotropic consolidation line In relation to degradation curves, due to its low plastic index, the
2.1
(ICL) and critical state line material behaves similar to classical reported clean sands and
(CSL) non-plastic fines along the whole shear strain range.
Yield surface shape factor The calibration of the ECP model shows a satisfactorily
( is Mohr-Coulomb adjustment for both, monotonic and cyclic paths. Because of this,
0.45 it is expected to be used for future researches involving seismic
type and is Cam-
Clay type) simulation to assess possible failure triggered by cyclic mobility
Projection of initial void or liquefaction.
(kPa) 100
ratio on CSL
Hardening coefficients for 5 REFERENCES
2.4×10-3
shear mechanisms ASTM D1556 – 15. “Standard Test Method for Density and Unit
1.1×10-4 (monotonous and cyclic) Weight of Soil in Place by Sand-Cone Method”
Hardening coefficients for ASTM D2487 – 11. “Standard Practice for Classification of Soils f
1×10-3 or Engineering Purposes (Unified Soil Classification System)”
isotropic mechanism
5×10-3 (monotonous and cyclic) ASTM D3999-03. Standard Test Methods for the Determination of
the Modulus and Damping Properties of Soils Using the Cyc
lic Triaxial Apparatus.
ASTM D4015-00. Standard Test Methods for Modulus and Dampin
g of Soils by the Resonant-Column Method.
Cifuentes, L., Verdugo, R. (2007). Resistencia estática y cíclica de
relaves espesados. VI Congreso Chileno de Ingeniería Geotéc
nica, 28-30 de noviembre de 2007, Valparaíso, Chile.
Dobry, R., Alvarez, L. (1967). “Seismic failures of Chilean tailings
dams”. Journal of the Soil Mechanics and Foundations Divi
sion, Vol. 93, No. 6, pp. 237-260
Hujeux JC. (1985). Une loi de comportement pour le chargement c
yclique des sols. Génie Parasismique ,Presse ENPC 1985:287
–302.
Ishihara, K., Troncoso, J., Kawase, Y., Takahashi, Y. (1980). “Cyclic
strength characteristics of tailings materials”. Soils and
Foundations Vol. 20, No.4, pp. 127-142
Jewell, R.J., Fourie A.B., Lord E.R. (2002). “Paste and Thickened
Tailings-A Guide”. Australian Center for Geomechanics (ACG)
Osorio, B. (2009). Resistencia estática y cíclica de relaves integrales.
memoria para optar al título de ingeniero civil, Universidad de
Chile
Palma, C., Vardugo, R., Moreno, J. (2007). “Comportamiento sísmico de
un depósito de relaves espesados frente a un sismo severo”. VI
Congreso Chileno de Ingeniería Geotécnica. Valparaíso, Chile.
Poulos, S., Robinsky, E., and Keller, T. (1985). ”Liquefaction Resistance
of Thickened Tailings.” J. Geotech. Engrg., 111(12), 1380–1394
Puma, C., Ovalle, C., Nazar, R. (2016): Efectos del método de
preparación de muestras en el comportamiento monótono y cíclico
de relaves integrales espesados. IX Congreso Chileno de
Ingeniería Geotécnica, 5 al 7 de Diciembre 2016, Valdivia, Chile.
Rico, M., Benito, G., Salgueiro, A., Diéz-Herrero, A., Pereira, H. (2008).
“Reported tailings dam failures. A review of the European
incidents in the worldwide context”. Journal of Hazardous
Figure 8. Monotonic tests simulations
Materials 152 (2008) 846–852
Robinsky E.I. (1982). "Thickened Tailing Disposal in any Topography",
In general terms, monotonic and drained cyclic test at 100 and Light Metals 1982, Proceedings of 111th AIME annual Meeting,
300 kPa are satisfactorily reproduced by parameters presented in February 1982, Dallas, Texas, pp. 239-248.
Table 1. Nevertheless, undrained cyclic behavior shown in Robinsky E.I. (1999). “Thickened Tailings Disposal in the Mining
Figure 6 is only reasonably reproduced at 100 kPa. Indeed, the Industry”. E.I. Robinsky Associates Limited, Toronto, Canada
model tends to overestimate the effect of confinement in the Santos, E. (2011). “Comportamiento monótono y cíclico no drenado de
number of cycles to reach liquefaction. A different set of arenas de relaves integrales”. Tesis de Magíster en Ingeniera Civil,
Universidad de Chile, Santiago, Chile.
parameters at 300 kPa is being calibrated to correct this effect.
Seed, H. (1987). “Design Problems in Soil Liquefaction". Journal of
ASCE 113 (GT8): 827-845
4 CONCLUSIONS Seed, H., Idriss, I. (1970). “Soil moduli and damping factors for dynamic
Experimental program performed allowed to characterize the response analyses”. Report EERC 70-10, Earthquake Engineering
Research Center, U.S.A.
monotonic and cyclic thickened tailings behavior at the in situ Villavicencio, G., Espinace, R., Palma, J., Fourie, A., Valenzuela, P.
characteristic dry density conditions (surface deposited tailings). (2013). “Failures of sand tailings dams in a highly seismic
The material presented a critical state characterized by a straight country”. Canadian Geotechnical Journal, 2014, 51(4): 449-464
line in e-log(p’) plane and a Mohr-Coulomb residual strength Vucetic, M., Dobry, R. (1991). “Effect of soil plasticity on cyclic
with a cohesion of 10 kPa and friction angle of 30 degrees. response”. J. Geotech. Engrg. ASCE 117(1): 89-107

- 1588 -