See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/335529384

Impact of methods used to reconstitute tailings specimens on the liquefaction

potential assessment of tailings dams

Article · September 2019

DOI: 10.1590/0370-44672018720164

CITATIONS READS

0 35

2 authors, including:

Waldyr oliveira filho

Universidade Federal de Ouro Preto
18 PUBLICATIONS   37 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Waldyr oliveira filho on 13 November 2019.

The user has requested enhancement of the downloaded file.

 Mariana Martins Corrêa and Waldyr Lopes Oliveira Filho

Mining
Mineração
Impact of methods used to
reconstitute tailings specimens
on the liquefaction potential
http://dx.doi.org/10.1590/0370-44672018720164 assessment of tailings dams
Mariana Martins Corrêa1,2 Abstract
https://orcid.org/0000-0001-8217-2671
Waldyr Lopes Oliveira Filho1,3 The aim of this research is to investigate the liquefaction susceptibility of silt
https://orcid.org/0000-0003-0857-1206 sandy mining tailings by experimental laboratory techniques. The main aspect ana-
lyzed is how techniques of sample reconstitution impact the results obtained in static
1
Universidade Federal de Ouro Preto – UFOP, undrained triaxial tests. Different methods of sample preparation are reviewed, such
Escola de Minas, Departamento de Engenharia as moist tamping (MT), air and water pluviation, and a newly developed one called the
de Minas, Ouro Preto - Minas Gerais – Brasil. Slurry Deposition (SD) method. This research highlights the importance of the "fab-
ric" or particle structural arrangement associated with the various specimen prepara-
E-mails: [email protected], tion techniques when liquefaction potential assessment is of concern. Two series of
3
[email protected] undrained static triaxial tests were performed on specimens prepared according to
MT and with SD techniques on specimens in the loose and very loose state. Results
have demonstrated that MT specimens have shown the whole spectrum of liquefaction
resistance (total liquefaction, limited liquefaction, and no liquefaction) on increasing
density, while the SD campaign has shown only liquefaction resistance even in the
overlapping intermediate densities with the MT series, where the latter has shown liq-
uefaction, although limited. This scientific study critically discusses the risk of taking
laboratory results of replicated soil elements that may not correspond to the reality.

keywords: static liquefaction, fine silt sand mining tailings, specimen preparation tech-
niques, triaxial test, undrained shear strength behavior.

REM, Int. Eng. J., Ouro Preto, 72(3), 507-513, jul. sep. | 2019 507
Impact of methods used to reconstitute tailings specimens on the liquefaction potential assessment of tailings dams

1. Introduction

Laboratory tests play an important field samples; preparation method renders a specific
role in understanding the mechani- b) need to test essentially identical fabric, which may result in distinct stress-
cal response of geomaterials. For this "homogeneous" test specimens (with- strain characteristics and responses (Vaid
purpose, testing undisturbed samples out variability commonly found in field and sivathayalan, 2000; Carraro and
obtained from field investigations is samples); and Prezzi 2007).
considered more appropriate for study- c) demand to characterize man- Sample reconstitution methods, such
ing natural soil behavior, since it allows made engineering materials. as moist tamping (MT), air pluviation (AP)
the structure of soil particles in situ, The use of reconstituted samples and water pluviation (WP) are the most
stress field, depositional history, etc. for liquefaction potential assessment widespread methods in laboratory rou-
to be represented. However, the use of of tailings deposits as proposed by this tines. In addition, Slurry Deposit (SD) is
reconstituted samples of granular soils research can be easily justified by all a newly developed preparation technique,
for laboratory testing has gained wide- three of these reasons. In addition, most derived from WP, which has been consid-
spread acceptance in the past for a num- of our fundamental understandings of ered to yield the most realistic results for
ber of reasons, including (Khalili and the liquefaction phenomena in sands has natural alluvial soils, and tailings deposits
Wijewickreme, 2008): been derived from controlled laboratory formed by disposed slurries. Table 1 sum-
a) difficulties in obtaining high studies. However, previous articles on marizes the methods, their characteristics
quality undisturbed non-cohesive this issue have shown that each sample and applications.

Brief Procedure
Method Type of Soil Characteristics References
Description
incorporates some
desirable charac-
sand, silty sands,
teristics from water
silt, clay, silt / clay Deposition of
pluviation technique,
Slurry / sand mixtures slurry prepared Carraro and
but less conducive to
Deposition and materials within an acrylic Prezzi (2007)
material segregation
with high fines tube
and most adequate for
content
applications of tailings
reconstitution

preferential fabric
which mimics natural
Air Non-cohesive Gravity particle alluvial deposits but Vaid and
Pluviation soils (sand) deposition in air fails when used with Negussey (1988);
fine-grained sands,
that causes segregation

greater homogeneity, Vaid and

Gravity particle
Water Non-cohesive but also conducive to Negussey (1988);
deposition in
Pluviation soils (sand) segregation for fine- Miura and Toki
water
grained sands (1982)

large range of speci-

Non-cohesive Deposition of
men densities, produc-
soils (sand), silty moist layered Castro (1969);
Moist es non-homogeneous
sands, silt, clay, soil, which is Casagrande
Tamping specimens and does
silt / clay / sand compacted after (1979)
not simulate fabric of Table 1
mixtures deposition.
alluvial deposits Most common specimen
reconstitution techniques.

2. Liquefaction assessment by laboratory tests

Soil liquefaction is one of the has been directed to the phenomenon recently received more interest espe-
geotechnical engineering phenomena occurrence under dynamic conditions, cially in the context of tailings dam
that has been researched for decades. performed mainly in countries with collapse disasters, such as Fundao
Its consequences can be catastrophic, constant records of seismic events. TSF, in 2015, in Southeastern Brazil.
whether caused by dynamic or static However, static stress-strain analy- This article deals exclusively with
loading. To this day, most research ses of this mechanism of failure has static liquefaction.
508 REM, Int. Eng. J., Ouro Preto, 72(3), 507-513, jul. sep. | 2019
 Mariana Martins Corrêa and Waldyr Lopes Oliveira Filho

2.1 Influence factors

Silty sand liquefaction is known to preparation method, and (dry) density main research findings on liquefac-
be affected by the following factor: fines (Monkul and Yamamuro 2011). tion resistance while investigating
content, confining stress, test specimen Table 2 presents a summary of those factors.

Variation Resistance Reference

fines content increases increases Monkul and Yamamuro (2011)

Zlatovic and Ishihara (1997);

fines content increases decreases
Monkul and Yamamuro (2011)

relative density increases increases Suhindra et al (1989)

confining stress increases increases Yamamuro and Lade (1997);

reconstituting specimen strength parameters not Zlatovic and Ishihara (1997);

method affected Monkul and Yamamuro (2011)
Table 2 loose deposits more contractive Kuerbis (1989)
Influence factors on liquefaction.

3. Materials and methods

Experimental work comprised tests (CIU) on specimens prepared using ratio), at 50 kPa and 300 kPa confining
of performing two series of isotropic either a moist tamping or slurry deposi- effective stress.
undrained consolidated static triaxial tion technique, variable density (or void

3.1 Sample Characterization

Material used in the study was silt- tests on this material rendered a mass density identified quartz, hematite and goethite. From
sandy tailings with non-plastic fines, from of 2.97 g/cm3, minimum dry density of 1.55 the grain size distribution curve, high fines
iron ore flotation processes, collected at a g/cm3 (according to ABNT MB-3324), and content (% passing #200) above 50% (58%
mine site in the Quadrilatero Ferrifero, in maximum dry density of 1.73 g/cm3 (ABNT exactly) were detected, and material could
the State of Minas Gerais. Indexed property MB-3388). Microscopic mineral analysis be characterized as a fine sandy silt material.

3.2 Moist Tamping Method (MT)

Soil elements were recreated in the lab (Ø 62 mm x 30 mm). saturation was pursued by backpressure in-
for the triaxial tests, firstly using the moist Triaxial tests performed with speci- cremental steps with the triaxial cell already
tamping technique (MT) with undercompac- mens prepared according to the MT tech- assembled. This time the procedure was ef-
tion, as suggested by Ladd (1978), in order nique initially followed a two-fold saturation fective, although extremely slow, taking days
to improve specimen uniformity. The main procedure. First of all, they were submitted to finish, and requiring high back pressure
materials and equipment used in this tech- to a low hydraulic gradient flux outside the values, approximately 700 kPa. The speci-
nique were: a wood socket; latex membrane triaxial chamber using differential elevation mens were then isotropically consolidated
(Ø 71 mm x 255 mm and thickness heads between a water source and the speci- under a confining effective stress of 50 kPa
0.3048 mm); two porous stone discs men. This external saturation, although nec- or 300 kPa. For the undrained shearing
(Ø 71.5 mm x 10 mm thick); a triaxi- essary and common in laboratory routines, phase (CIU), a shear rate of 0.16 %/min was
al split mold (Ø 72.5 mm x 161 mm); was not very effective, since in the end, the determined based on consolidation data and
760 mm/Hg vacuum pump; a triaxial cell Skempton B parameter was below 0.1 for tests lasting until axial deformation reached
with base (Ø 62 mm x 30 mm) and a topcap all MT test specimens. Secondly, specimen 20%, which means a two-hour test.

3.3 Slurry Deposition Method (SD)

Field soil elements were also recre- stoppers (70.0 x 55.0 x 67.0 mm), one with specimens for triaxial tests were also
ated using the reconstitution technique an attached valve on its center; a latex saturated following similar laboratory
for sandy soils with slurry deposition fines membrane (Ø 71 x 255 mm and thickness techniques as described for MT samples.
(SD), as suggested by Carraro and Prezzi 0.3048 mm); two porous stones (Ø 71.5 In this case, as expected, the initial perco-
(2007), with some modifications and mm x 10 mm thick); a triaxial split mold lation at low gradient rendered Skempton
additions of Wang et al. (2011), Abreu et (Ø 72.5 mm x 161 mm); an aluminum B values approximately 0.6, which is much
al. (2016) and others developed in this re- collar; an aluminum tray; a 760 mm / Hg higher than the ones obtained for MT
search. The main materials and equipment vacuum pump; a triaxial cell with pedestal specimens. Saturation was then completed
used in this preparation were: a cylindrical (Ø 62 mm x 30 mm) and a top cap (Ø 62 by applying backpressure steps. The de-
acrylic tube (70.0 outer diameter and 60.0 mm x 30 mm). sired saturation, that is, B values around
inner diameter x 1000 mm height); 2 latex After specimen preparation, SD 0.95, was reached at backpressure values
REM, Int. Eng. J., Ouro Preto, 72(3), 507-513, jul. sep. | 2019 509
Impact of methods used to reconstitute tailings specimens on the liquefaction potential assessment of tailings dams

around 350/450 kPa, quite low as com- isotropically consolidated under 50 kPa or the consolidation data, until the sample
pared with the MT series, and much less 300 kPa stress. The undrained test shear reached an axial deformation of 20%,
time consuming. The specimens were then rate was set at 0.16%/min (CIU), based on which took approximately two hours.

4. Results

The main results of this research whose specimens were prepared accord- to attempt to verify the impact of these
are related to the analysis of CIU triaxial ing to the moist tamping (MT) and slurry specimen reconstituting techniques on
tests performed on the silt sandy tailings, deposition (SD) methods. The goal was liquefaction susceptibility.

4.1 Moist Tamping Triaxial (MT)

Figures 1 and 2 show CIU tests ing to the MIT convention ranging densities (1.59 g/cm3). Figure 3 replotted
results for MT test specimens plotted from low (1.48 g/cm3 and 1.51 g/cm3), all 50 kPa tests together, and also one
in terms of stress path (p' x q) accord- intermediate (1.56 g/cm3) and higher 300 kPa test.

Figure 1
CIU triaxial test results stress path
(MIT) for MT specimens with densities
of 1.48 g/cm3, 1.51 g/cm3 and 1.56 g/cm3.

Figure 2
CIU triaxial test
results stress path (MIT) for MT
specimens with densities of 1.59 g/cm3.

Figure 3
CIU triaxial test results stress
path (MIT) for all MT test specimens.

At low densities (Figure 1), a complete density produces a change from contractive, it is performed on a low density (1.52 g/cm3)
liquefied state is reached, differently from the softening (liquefiable) behavior, to a dilat- test specimen, softening behavior occurs,
intermediate range where a limited liquefac- ing and hardening (liquefaction resistant) and general liquefaction does not occur. All
tion state is observed (also in Figure 1, 1.56 pattern, including an intermediate state of of this corroborates previous studies, with
g/cm3). For the test with the highest specimen slight softening (limited liquefaction). It is influence factors density and confining stress
density (Figure 2), high liquefaction resis- further noted that for the 300 kPa confin- playing important roles on liquefaction resis-
tance is configured. Briefly, the increase in ing effective stress test (Figure 3), although tance (Yamamuro and Lade, 1997).
510 REM, Int. Eng. J., Ouro Preto, 72(3), 507-513, jul. sep. | 2019
 Mariana Martins Corrêa and Waldyr Lopes Oliveira Filho

4.2 Slurry Deposition Triaxial (SD)

Figure 4 shows CIU tests results for SD test specimens with densities under 50 kPa and 300 kPa confining
plotted in terms of stress path (p' x q) between 1.54 g/cm 3 to 1.61 g/cm 3 effective stress.

Figure 4
CIU triaxial test results stress
path (MIT) for all SD test specimens.

The same behavior was observed pressure, and then, as the deformation hardening behavior. It is interesting to
in all four tests, regardless of the density increases, the trajectory turns to the right, note that at large deformations, the stress
difference or the confining stress. Firstly, climbing the same slope, generating large paths of all tests overlap, characterizing
stress paths rise, bending slightly to the values of negative excess pore pressure a limit state of resistance (critical state
left, generating low positive excess pore during shear, and revealing a dilating and strength parameters).

4.3 Triaxial result comparison between the two techniques

4.3.1 Comparison between low and medium densities
Observing Figure 5, there are ap- specimens with liquefaction behavior with lower density (1.54 g/cm3) shows
parently different and conflicting behav- by the MT method for low density liquefaction resistance contrasting with
iors for the same soil element recreated (1.51 g/cm3). For intermediate densities the limited liquefaction behavior of a
according to different methods, MT (1.54 and 1.56 g cm3), the two methods denser specimen (1.56 g/cm3) formed
and SD. First, it is possible to obtain produce very different patterns. The SD, by MT.

4.3.2 Comparison between higher densities

In Figure 6, it can be seen that the and are dilating. These characteristics regardless of the reconstituting technique,
two specimens show resistance to liquefac- are more noticeable with the specimen the same behavior is observed with in-
tion, have hardening with deformation, prepared by MT. It can be concluded that creased density.

Figure 5
CIU triaxial test results
stress path (MIT) for low
and medium MT specimen densities
and intermediate SD specimen density.

Figure 6
CIU triaxial test results
stress path (MIT) for higher
densities using MT and SD specimens.
REM, Int. Eng. J., Ouro Preto, 72(3), 507-513, jul. sep. | 2019 511
Impact of methods used to reconstitute tailings specimens on the liquefaction potential assessment of tailings dams

5. Further discussions

Until the present, discussions on relative densities in the range of loose, very it endorses the previously mentioned idea
liquefaction susceptibility have focused on loose and beyond. One can consider that that the static liquefaction condition may
reconstitution techniques and specimen sand soils with high fines content may need be achieved simply by bringing the material
densities without referring the latter to different limit standards in order to make into a non-stable state (high pore pressure
material limits (item 3.1). Results in Figure comparisons with the copious literature on for example) in which its resistive force is re-
7 suggest several things, for example, that it liquefaction susceptibility that relates it to duced sufficiently to allow static conditions
is possible to achieve higher void ratios than relative density. to produce static liquefaction. Such type of
the maximum considered. Another interest- Even having resolved this issue, there pre-triggering mechanism can be caused
ing matter is that all liquefaction, limited still remains the question why SD speci- by, e.g., continuous vibrations generated by
liquefaction or liquefaction resistance be- mens have not shown liquefaction. This traffic equipment in the mining area setting
haviors happened for specimens presenting research does not examine that aspect, but (Penna and Oliveira-Filho, 2012).

Figure 7
Voids ratio, dry density and relative
density spectrum of all specimens data.

6. Conclusion

The present research established an tions, the Slurry Deposition (SD) specimen tions in the field; however, the SD tech-
experimental program to assess the min- method seems to render the best results. nique may offer an appropriate method for
ing tailings liquefaction potential through • It has been observed that material studying fluvial or hydraulic fill deposits
isotropically consolidated undrained tri- density greatly influenced the results for in the future.
axial tests. It was driven by the challenge very loose soil specimens, since a small • All specimens tested were prepared
of searching for the most adequate speci- difference in this index property was with relative densities in the range of loose
men reconstitution method that could enough to manifest different behaviors and very loose and even beyond categories
make it possible to produce soil elements in the MT campaign, as total liquefac- according to the Brazilian ABNT stan-
with a similar structure, as in tailings dam tion (lower density) and non-liquefaction dards for dry densities limits. This finding
deposits, such as those found in mines (higher density). On the other hand, the generates several questions, for example, if
in the Quadrilatero Ferrifero of Minas SD technique showed dilating behavior, a the maximum and minimum void indexes
Gerais, Brazil. characteristic that corresponds to liquefac- found are true for this material, which does
The principal conclusions of the tion resistance, in all possible experimen- not seem to be the case. If they are not,
study are the following: tal densities, even in the overlapping range what would actually be the range of rela-
• The literature review carried out with the MT series. tive density for the tested specimens. In ad-
indicated that the Moisture Tamping • This research has shown how dition, how logically and consistently the
(MT) specimen reconstitution method is fabric, resulting from different specimen relative density (compactness) would relate
not the appropriate way to replicate soil preparation, can affect liquefaction assess- to the mechanical response of these soils.
elements formed under an alluvial envi- ment. Unfortunately, scientific research • It is fair to say that laboratory
ronment and hydraulic fills, as is the case based on laboratory tests cannot exactly testing of representative samples is a valid
of mining tailings deposits. In such condi- replicate all possible and existing situa- resource. Research has been used to un-
512 REM, Int. Eng. J., Ouro Preto, 72(3), 507-513, jul. sep. | 2019
 Mariana Martins Corrêa and Waldyr Lopes Oliveira Filho

derstand soil liquefaction susceptibility, limit its conclusiveness. This research has ent specimen preparation can affect that
although some intangible factors may shown how fabric resulting from differ- phenomenon assessment.

Acknowledgments

We thank the Graduate Program on gineering Department undergraduate stu- government agency for higher education
Mineral Engineering of UFOP (Federal dents Stephanie Oliveira Moura e Sá and advanced training and FAPEMIG, Minas
University of Ouro Preto) for their general Otávio Pataro for their help with sample Gerais Research Support Foundation, for
support. Also, many thanks to Mining En- characterization, CAPES, Brazilian federal providing a scholarship to the first author.

References

ABNT. Solo – Determinação do índice de vazios mínimo de solos não-coesivos.

MB 3384. São Paulo: Associação Brasileira de Normas Técnicas, 1991. 14 p.
ABNT. Solo – Determinação do índice de vazios máximo de solos não-coesivos.
MB 3324. São Paulo: Associação Brasileira de Normas Técnicas, 1990. 6 p.
ABREU, I. C., OLIVEIRA, S., FILHO, W. L. O., FOURIE, A. Slurry deposi-
tion method for reconstituting specimens for triaxial and consolidation tests.
COBRAMSEG, 2016.
CARRARO, J. A., PREZZI, M. A new slurry-based method of preparation of speci-
mens of sand containing fines. ASTM. Geotechnical Testing Journal, v . 31, n. 1.,
2007.
CASAGRANDE, A. Liquefaction and cyclic deformation of sands: a critical review.
In: AMERICAN CONFERENCE ON SOIL MECHANICS AND FOUNDA-
TION ENGINEERING, 5. Proceedings... Buenos Aires, v. 5, p. 79-133, 1979.
CASTRO, G. Liquefaction of Sands, Cambridge, MA: Harvard University, 1969.
(Ph.D. Thesis).
KHALILI, A., WIJEWICKREME, D. New slurry displacement method for re-
constitution of highly gap-graded specimens for laboratory element testing.
Geotechnical Testing Journal, v. 31, n. 5, p. 424-432, 2008.
KUERBIS, R. The effect of gradation and fines content on the undrained loading res-
ponse of sand. Vancouver, Canada: University of British Columbia, 1989. (Master
of Applied Science).
LADD, R. S. Preparing test specimens using undercompaction. Geotech. Test. J., v. 1,
n. 1, p. 16–23, 1978.
MIURA, S., TOKI, S. A sample preparation method and its effect on static and cyclic
deformation-strength properties of sand. Soils Found, v. 22, n. 1, 1982.
MONKUL, M. M., YAMAMURO, J. A. Influence of silt size and content on lique-
faction behavior of sands. Canadian Geotechnical Journal, v. 48, n. 6, p. 931-942,
2011. SCI.
PENNA, L. R., OLIVEIRA FILHO, W. L. de. Monitoramento de poropressões cons-
trutivas. Revista Escola de Minas, v. 65, p. 443-448, 2012.
SUHINDRA, C., KUBERAN, R., KUMAR, B., AGGARWAL, P. K. Liquefaction
of sand under cyclic triaxial condition. In: INDIAN GEOTECHNICAL CON-
FERENCE, GEOTECHNIQUES OF PROBLEMATIC SOILS AND ROCKS.
Proceedings... Visakhapatnam, 14-16, v. 1, p. 249-253, 1989.
VAID, Y. P., NEGUSSEY, D. Preparation of reconstituted sand specimens, advanced
triaxial testing of soil and rock, ASTM STP 977, 1988.
VAID, Y. P., SIVATHAYALAN, S. Fundamental factors affecting liquefaction
susceptibility of sands. Can. Geotech. J., v. 37, p. 592–606, 2000.
WANG, S., LUNA, R., STEPHENSON, R. W. A slurry consolidation approach to re-
constitute low-plasticity silt specimens for laboratory triaxial testing. Geotechnical
Testing Journal, v. 34, n. 4), p. 1-9, 2011.
YAMAMURO, J. A., LADE, P. V. Static liquefaction of very loose sands. Canadian
Geotechnical Journal, v. 34, p. 905-917. 1997.
ZLATOVIC, S., ISHIHARA, K. Normalized behavior of very loose non-plastic soils:
effects of fabric. Soils Found., v. 37, n. 4, p. 47–56, 1997.

Received: 12 September 2018 - Accepted: 04 February 2019.

All content of the journal, except where identified, is licensed under a Creative Commons attribution-type BY.

REM, Int. Eng. J., Ouro Preto, 72(3), 507-513, jul. sep. | 2019 513

View publication stats