Locat Demers 1988
Locat Demers 1988
Locat Demers 1988
sensitive clays
JACQUES
LOCAT
Groupe de recherche en ge'ologie cle l'inge'nieur, De'pnrtement de ge'ologie, Utziversite' Lnval, Sainte-Fay, Que.,
Canada GI K 7P4
AND
DENISDEMERS
Service des sols et chausse'es, MinistPre des Transports du Que'bec, 200 Dorchester Street South, Que'bec, Que.,
Canada G1K 521
Received March 21, 1988
Accepted June 16, 1988
The rheological behavior of some sensitive clays has shown that there are positive relationships between plastic viscosity,
yield stress, remolded shear strength, and liquidity index. Based on the various correlations, remolded strength values can be
predicted for soils with a liquidity index as high as 6. For a liquidity index varying from 2 to 5, the remolded undrained shear
strength and plastic viscosity vary from about 90 to 5 Pa and from 200 to 7 mPa . s respectively. It has been observed that most
sensitive clays behave either as a Bingham or a Casson fluid, the latter behavior being related to less sensitive clays of higher
pore-water salinities. Increasing the salt content from 0.5 to 30 g/L modified the rheological behavior of a single test soil from
that of a Bingham to that of a Casson type of fluid. Such modification in the pore-water salinity illustrates how soils can have
similar plastic viscosity and similar remolded strength but different yield stresses. The range of viscosity measured for
sensitive clays is very different from that of values obtained from back analysis of subaerial or submarine slides where viscous
flow can be expected.
Key words: sensitivity, clays, viscosity, yield stress, shear strength, salinity, liquidity index, submarine slide, landslide.
Le comportement rhCologique de quelques argiles sensibles a dCmontrC une relation positive entre les valeurs de viscositC
plastique, de limite liquide, de rksistance au cisaillement remaniCe et d'indice de liquiditC. En partant de diverses corrClations,
les valeurs de rksistance au cisaillement peuvent &trepredites pour des sols ayant des indices de liquidit6 aussi Clevis que 6.
Pour un indice de liquidit6 entre 2 et 5, la rksistance au cisaillement non drain6 remaniCe et la viscositC plastique varient
respectivement d'environ 90 B 5 Pa et de 200 B 7 mPa. s. I1 a CtC observC que la plupart des argiles sensibles se comportent
comme un fluide de Bingham ou de Casson, ce demier comportement correspondant aux agiles moins sensibles avec des
salinitCs d'eau interstitielle plus ClevCes. L'accroissement de la teneur en sel de 0,5 B 30 g/L a modifiC le comportement au
cours d'un essai, passant d'un type de fluide de Bingham B un fluide de Casson. Une telle modification dans la salinitC de l'eau
interstitielle illustre comment des sols peuvent avoir des viscositCs plastiques et des rCsistances au cisaillement semblables mais
des limites Clastiques diffkrentes. Le domaine de viscositC mesure pour les argiles sensibles est trks different de celui donne par
des valeurs obtenues d'analyse B rebours de glissements terrestres et marins oh l'on s'attend B des Ccoulements visqueux.
Mots cle's : sensibilitC, argiles, viscositC, limite Clastique, risistance au cisaillement, salinitC, indice de liquidit&,glissement
marin, glissement de terrain.
[Traduit par la revue]
Can. Geotech. J. 25, 799-806 (1988)
Introduction
Spectacular landslides like the Saint-Jean-Vianney (Chagnon
1968; Tavenas et al. 1971; Locat and Leroueil 1988), South
Nation River (Eden et al. 1971), and Rissa (Gregerson 1981)
have illustrated the great mobility of remolded sensitive clays.
The long run-out distances observed for these landslides were
related to very low remolded shear strength (or viscosity) of
the soil after failure. Very low remolded shear strengths are
also common for thawed permafrost soils (Watson et al. 1973;
Hutchinson 1974; Johnston 1981). In all cases, the behavior of
the remolded soil mass must be evaluated using flow properties
and parameters. Edgers and Karlsrund (1982) strongly pointed
out the critical role of the viscosity of the soil mass in submarine and subaerial slide dynamics. Hence, flow behavior
can be quite complex and various types of flow behavior can
exist depending on the soil type, pore-water salinity, mineralogy, and water content. The main types of flow are shown in
Fig. 1 (Couarraze and Grossiord 1983; Reiner and Scott Blair
1967), where viscosity corresponds to the slope of these
curves. Thickening liquids (curve 2, Fig. 1) are those for
Prinvd in Canada / IrnprirnC au Canada
800
MV-I
MV-I1
MV-I11
1198
0.023
46 1
0.009
225
0.004
4.2
2.7
14.0
8.0
Shear stress
Max. (Pa)
Min. (Pa)
Shear rate
Max. (s-I)
Min. (s-I)
Viscosity
Max. (rnPa.s) X 10'
Min. (mPa .s) x lo-'
1.4
0.89
Shear rate
FIG. 1. Major types of fluids: (1) Newtonian; (2) thickening;
(3) fluidizing; (4) Casson; (5) Bingham.
given water content, it became possible to use the fall cone for
that purpose (Garneau and LeBihan 1977). However, the
lower strength limit of the fall cone is at 73 Pa (equivalent to a
liquidity index of about 3); this is too high for many sensitive
clays having liquidity indexes between 3 and 6, or more.
Recently, Demers and Locat (1985, 1986), Torrance (1987),
and Locat et al. (1988) started using a viscometer on a more
routine basis and have presented positive relationships between
yield stress and remolded undrained shear strength (measured
by the fall cone).
The liquid limit of a soil varies with salinity (Torrance
1975). Kerr and Drew (1965, 1968), Bentley (1976, 1979),
Dixon (1982), Torrance (1984), and Torrance and Pirnat
(1984) have discussed the effect of salinity on the flow behavior of clay slurries. They observed that increasing the salt content of the pore water from low (less than 2 g/L) to high (above
5 g/L) values resulted in an increase in the yield stress.
Investigating the microfabric of clay slurries, Osipov et al.
(1984) have shown that the yield stress can decrease with
increasing shear rate, as the interaggregate links are broken
near the zone of shearing. Moore (1965) has indicated that, for
partially deflocculated suspensions, the hysterisis loop is very
large, as the restructuration rate is very slow. As for Bingham
fluids, these soils show linear shear stress - shear rate relationships at values above their yield stress.
In this paper, efforts are directed towards the understanding
of the flow characteristics of remolded soils to relate viscosity,
yield stress, remolded shear strength, and liquidity index in
order to provide a simple first approximation of the rheological
parameters of clayey soil.
Methodology
The viscometer used in this study is a rotational rheometer
(Haake-Rotovisco, model RV-12), which is run in a steady
state regime. Sensors are composed of two coaxial cylinders.
The inner one, the rotor, is the mobile part, which is linked to a
gauge measuring the torque applied on the fluid. The outer
cylinder is fixed and insures the temperature control by means
of a liquid cooling system. The main specifications of the
apparatus are given in Table 1. The procedures followed for
viscometric measurements are described in Bentley (1976) and
Torrance (1987), and included three types of tests: (1) dynamic
response, (2) constant shear rate, and (3) hysteresis (Fig. 2).
The hysteresis test is the last one to be carried out. At the
(I)
(I)
a,
C
(I)
a,
r
(I)
-2
-
200
100-
(e)
(f)
??
(I)
a,
(I)
Oo
260
460
Oo
10
20
Time ( m i n )
*
Oo
200
460
Shear r a t e (s-')
FIG.2. Flow characteristics of the two types of behavior observed determined using the dynamic response constant shear rate, and hysteresis
test for (a-c) the Saint-Alban-1soil (I, = 3.0, w = 82%, cur = 88 Pa) and (d-f) the QuCbec soil (I, = 2.4, w = 90%, cur= 190 Pa) (T is yield
stress).
slippage plane would give inconsistent results and yield lower
viscosity values.
Results
Over 70 viscometric tests were carried out. An example of
shear rate - shear stress curves is presented in Fig. 3 to show
the gradual decrease in the yield stress as the water content, or
liquidity index, increases. These curves are used to compute
both viscosity and yield stress. By inspection of the various
curves, two types of fluids could be observed: Bingham and
Casson. The Berthierville and Saint-Alban-1 soils presented a
behavior closer to Bingham type of fluid (Fig. 2a). These soils
exhibited a pronounced thixotropic behavior noted by a continuous decrease of the shear stress at a constant shear rate
(Fig. 2b) and a large hysteresis loop (Fig. 2c). Soils of
QuCbec, Saint-Hyacinthe, and Saint-Alban-2 behaved more
like a Casson-type fluid, were only slightly thixotropic
(Fig. 2f), and could easily reach a stable torque value
(Fig. 2e). The Saint-Wenceslas soil, much coarser than the
others (Table 2), showed some shear thickening at a liquidity
index above 3 and at a shear rate above 200 rpm. In return, it
made viscosity computation less accurate.
Plotting the viscosity versus the yield stress (Fig. 4) shows
that at low values of yield stress, two tendencies exist, and
again, are related to the type of fluid behavior mentioned
above. Artificially increasing the salt content to 30.2 g/L transformed the rheological behavior of Saint-Alban-1 soil from
that of Bingham to that of a Casson-like fluid (Saint-Alban-2),
a feature also observed for other sensitive clays to Bentley
(1979) and Dixon (1982). The same distinction can be made
when the yield stress is correlated to the liquidity index
(Fig. 5); this is particularly true for a liquidity index greater
Shear r a t e (s-'1
FIG.3. A series of shear rate - yield stress curves obtained for the
Saint-Alban-1 soil for various water contents and at a salt content of
0.2 g/L (T is yield stress).
than 3. Here, for a given soil and salinity, the correlation is
quite good.
However, when correlating liquidity index (IL) and viscosity
(v), the distinction between the two type of fluids cannot be
made (Fig. 6). The scatter of the results, particularly for the
coarser grained Saint-Wenceslas soil, is larger than it is for the
other correlations presented. As observed before, the relation-
TABLE
2. Soil characteristics
Depth
(m)
Site
Berthierville
Quebec
Saint-Alban- 1
Saint-Alban-2
Saint-Hyacinthe
Saint-Wenceslas
6.1
-
2.3
2.3
4.3
3.6
( 2 ~ )St
w
(%)
1.38
2.04
0.20
39
47
41
71
0.77
0.52
8
82
82
35
25
-
w,
I,
(%)
(%)
SS
(m2ig)
CF
(%)
S
(g/L)
44.2
52.4
42.2
55.0
62.8
27.4
21
28
20
32
37
11
49
62
49
59
96
22
36
60
49
59
75
19
0.3
5.6
0.5
30.2
1.6
0.3
NOTE:SS stands for specific surface area; CF, clay-size faction; and S, salinity
!0' 10
0
2:
0
o*o
4+0
ccn
." ,+
+
A * Q
0 1 0 -1
cn
5
'=O
o 0
0
A
y*
--
omO
8-0
0
It
10
---
ST-ALBAN-1
ST-ALBAN-2
ST-HYACINTHE
ST-WENCESLAS
--
10
10
~1
ST-HYACINTHE
+ ST-WENCESLAS
It
0 dl
1
10
10
YIELD STRESS (Pa)
FIG. 5. Relationship between liquidity index and yield stress.
10
6-
--
5-
ir:
3
Cr
ST-HYACINTHE
z 45ii
- 35
7 2-
1 --
1 1 1 1
10
10
VISCOSITY (mPa-s)
'
ST-ALBAN-2
ST-HYACINTHE
n
w
10
10
10
10
Discussion
Viscosity measurements
There is very little information in the literature about the viscosity of remolded sensitive clays. Eden and Kubota (1961)
reported a viscosity of about 76.0 mPa. s for an Ottawa soil
having a liquidity index of about 2.3. This is in the range of
what could be predicted from [I]. More recent data on liquidity
index and yield stress are plotted in Fig. 9, along with the
range of data presented here. Included are results of Bentley
(1979), Dixon (1982), and Torrance and Pirnat (1984), which
804
10
'
10"
1
1
I
I
1
I
increase in the salt content from 0.5 to 30.2 g/L changed the
rheological properties from those of a Bingham type of fluid to
those of a Casson type of fluid. This test also illustrates (Fig. 5)
that, at the same liquidity index, leaching would reduce the
yield stress. Figure 10 presents a conceptual model that
describes the changes in the rheological behavior observed for
the Saint-Alban soil. This soil, at the same liquidity index, can
have two values of yield stress. As the salinity is increased
above the coagulation point (van Olphen 1977; Locat and
Lefebvre 1985) the flow behavior evolves from that of a Bingham to that of a Casson type of fluid. In nature, the leaching
process would follow a path in the opposite direction. In the
present case of Saint-Alban, it can be shown that for the liquidity index - remolded shear strength relationship, an increase
805
Conclusions
Acknowledgements
This research was made possible by the financial support
of the Natural Sciences and Engineering Research Council
of Canada and the cooperation of the technicians and engineers
of the Groupe de recherche en gCologie de 1'ingCnieur of
the Department of Geology at Laval. In particular, we would
like to thank M.-A. BCrubC, J.-Y. Chagnon, P. GClinas, and
S. Leroueil for commenting on the manuscript and A. Masson
for editing.
Yield stress
FIG. 10. Conceptual model representing the behavior of the SaintAlban soil as the salinity is increased from 0.5 to 30.2 g/L (see the
text for explanation).
in the salinity results in an increase in the remolded strength as
along segment AB (Fig. 10a). It is more complex for the
liquidity index - yield stress relationship. If the salinity is initially below the critical coagulation point and then increases
above it, at constant water content, it will shift from point C on
the "Bingham" curve to point D on the "Casson" curve and
with continued salinity increase, the yield stress - liquidity
index relationship will follow the Casson curve. Such modifications of the rheological properties with increasing salinity
would result in a change in the soil microfabric from a dispersed to a flocculated state (Moore 1965; van Olphen 1977;
Bentley 1979; Sacks 1985). Such a test was not performed on
the other soils; this conceptual model will therefore need wider
verification.
Applying these results to landslides indicates that for the
Saint-Alban soil, at the same liquidity index or viscosity, the
continuation of the flow, after failure, will be more easily done
in a dispersed state than in a flocculated state, as the yield
stress is lower. This is an important consideration in the analysis, or prediction, of run-out distances in sensitive clays
(Lebuis et al. 1983) or for submarine slides (Edgers and
Karlsrud 1982). For subaerial and submarine flow slides, both
the viscosity and yield stress must be known if flow dynamic
behavior is to be understood. Such correlation is still limited to
fine-grained soils, so its application cannot be made for all
BENTLEY,
S. P. 1976. The contribution of primary mineral particles
to the properties of sensitive clays from eastern Canada. Ph.D.
thesis, Leeds University, Leeds, England.
1979. Viscometric assesment of remoulded clays. Canadian
Geotechnical Journal, 16: 414 -419.
BJERRUM,
L. 1954. Geotechnical properties of Norwegian marine
clays. GCotechnique, 4: 49 -65.
BJERRUM,
L., and ROSENQVIST,
T. 1956. Some experiments with
artificially sedimented clays. Gkotechnique, 6: 124- 136.
CHAGNON,
J.-Y. 1968. Les coulCes d'argiles dans la Province de
QuCbec. Naturaliste canadien (Quebec), 95: 1327- 1343.
COUARRAZE,
G., and GROSSIORD,
J. L. 1983. Initiation ?
lai rhCologie.
Lavoisier, Paris, France.
DEMERS,
D., and LOCAT,J. 1985. L'Cvaluation de la rksistance
remaniCe (c,,) et de la thixotropie d'argiles extrasensibles du
Quebec par l'utilisation du viscosimbtre. Proceedings, 53rd Association canadienne fran~aisepour I'avancement de science Congress, Chicoutimi, Que., p. 144.
1986. CorrClation entre certaines propriCtCs viscosimCtriques
et gkotechniques d'argiles marines remaniCes. Department of Geology, Laval University, Report GGL-86-10.
DIXON,
D. A. 1982. Geotechnical influence on the rheological properties of four post-glacial marine clays. M.Sc. thesis, Carleton
University, Ottawa, Ont.
EDEN,W. J., and KUBOTA,
J. K. 1961. Some observations on the
measurement of sensitivity of clays. Proceedings, American Society for Testing and Materials, 61: 1239- 1249.
EDEN,W. J., FLETCHER,
E. B., and MITCHELL,
R. J. 1971. South
Nation River landslide, 16 May 1971. Canadian Geotechnical Journal, 8: 446-451.
EDGERS,
L., and KARLSRUD,
K. 1982. Soil flows generated by submarine slides-Case studies and consequences. Norwegian Geotechnical Institute, Publication 143, pp. 1- 10.
GARNEAU,
K., and LEBIHAN,
J. P. 1977. Estimation of some properties of Champlain clays with the Swedish fall cone. Canadian Geotechnical Journal, 14: 571 -581.
806
GREGERSON,
0 . 1981. The quick clay landslide in Rissa, Norway.
Norwegian Geotechnical Institute, Publication 135.
HAJELA,R. B., and BHATNAGAR,
J. M. 1972. Applications of
rheological measurements to determine liquid limit of soils. Soil
Science, 114: 122- 130.
HUTCHINSON,
J. N. 1974. Periglacial solifluxions: an approximate
mechanism for clayey soils. GCotechnique, 24: 438 -443.
JOHNSON,
A. M., and RODINE,J. R. 1984. Debris flow. In Slope
instability. Edited by D. Brunsden and D. B. Prior. Wiley, New
York, NY, pp.257-362.
JOHNSTON,
G. H., editor. 1981 Engineering design and construction.
Associate Committee on Geotechnical Research, National Research
Council of Canada.
KERR, P. F., and DREW, I. M. 1965, Quick clay movements,
Anchorage, Alaska. Air Force Cambridge Research Laboratories,
Scientific Report 5.
1968. Quick-clay slides in the U.S.A. Engineering Geology
(Amsterdam), 2: 215 -238.
LEBUIS,J., ROBERT,J. M., and RISSMANN,
P. 1983. Regional
mapping of landslide hazards in Quebec. Symposium on Slopes on
Soft Clays. Swedish Geotechnical Institute, Report 17, pp.
205 -262.
LEROUEIL,
S., TAVENAS,
F., and LEBIHAN,J. P. 1983. PropriCtCs
caractCristiques des argiles de l'est du Canada. Canadian Geotechnical Journal, 20: 681 -705.
LOCAT,J. 1982. Contribution i 1'Ctude de l'origine de la structuration
des argiles sensibles de l'est du Canada. Ph.D. thesis, Department
of Civil Engineering, University of Sherbrooke, Sherbrooke, Que.
LOCAT,J., and LEFEBVRE,
G. 1985. The compressibility and sensitivity of an artificially sedimented clay soil: The Grande-Baleine
marine clay, Quebec, Canada. Marine Geotechnology, 6: 1-28.
1986. The origin of structuration of the Grande-Baleine
marine sediments, Qutbec, Canada. Quarterly Journal of Engineering Geology, 19(4): 365 -374.
LOCAT,J., and LEROUEIL,
S. 1988. Physicochemical and geotechnical characteristics of Saguenay Fjord sediments. Canadian Geotechnical Journal, 25: 382-388.
LOCAT,J., KURFURST,
P., BERUBE,M.-A., CHAGNON,
J.-Y., and
GELINAS,
P. 1988. Viscometric properties of a Beaufort Sea sediment. Geological Survey of Canada, Open File 1708.