You are on page 1of 13

Engineering Geology 275 (2020) 105729

Contents lists available at ScienceDirect

Engineering Geology
journal homepage: www.elsevier.com/locate/enggeo

Influence of non-plastic fine on static iquefaction and undrained monotonic T


behavior of sandy gravel

Hamidreza Rahmani, Seyed Abolhasan Naeini
Civil Engineering Department, Imam Khomeini International University, Qazvin, Iran

A R T I C LE I N FO A B S T R A C T

Keywords: Many geological hazards such as landslide, liquefaction, slope failure and rockfall occur in mixed gravelly al-
Static liquefaction luvium or colluvium deposits, highlighting the importance of studying gravel–sand–silt mixtures. Although
Steady state line many natural strata are built using these mixtures, they are not entirely understood. A fundamental experimental
Intergranular void ratio study was designed to explore the influence of non-plastic fine content on the undrained behavior of gravel-sand-
Residual strength
silt mixtures as common geomaterials in nature. Undrained monotonic triaxial tests were carried out. Three
Sandy gravel
Silt
types of sandy gravels with different sand-to-gravel ratios were selected as base gravels. The impacts of the non-
plastic fine content on the stress–strain behavior, static liquefaction potential, steady sate line, residual strength
and critical state friction angle were studied. The laboratory results show that both sand-to-gravel ratio and non-
plastic fine content have significant impacts on the undrained response of gravelly mixtures. It was observed that
mixtures with zero non-plastic fines content exhibit the non-flow liquefaction behavior. However, adding silt
content up to 45% shows the limited liquefaction or flow behavior, depending on the base gravelly soils and fine
content percentage. A threshold for non-plastic fine content was found to be 35%–45%. Below this threshold, the
steady state line shifts downward as silt content increases, and then shifts upward with further increase in the silt
contents. The identical trend was also noted for residual strength and critical state friction angle. For silt contents
below the threshold, the equivalent intergranular void ratio concept (modified with a new logical assumption)
was used to analyze the behavior of gravelly mixtures. The results clearly indicate that the accurate value of
equivalent intergranular void ratio can fully explain the behavior of sandy gravel-fine mixtures, similar to sand-
fine mixtures. From a microstructure perspective, both sand and silt content act as fine-grains in gravelly
mixtures and should be considered when studying the mechanical properties of theses mixtures.

1. Introduction undrained conditions, and usually liquefaction has been considered


only for clean sands or silty sands. But today, many researchers believe
Gravelly mixtures (gravel-sand-silt mixtures) are found abundantly that if the drainage conditions are limited (i.e. filling the voids of the
in natural strata such as colluvium deposits, rockfills, alluvial deposits gravelly particles with finer soil), this may generate the significant pore
and glacial tills, among others. Slopes formed by colluvium deposits, water pressure and reduce the effective stress, causing initial liquefac-
with loose state and low shear strength parameter, have high potential tion. A number of case studies are reported in literature over the last
of instability when exposed to rainfall or earthquakes (Chang and five decades, wherein liquefaction occurred for gravelly soils (Andrus,
Phantachang, 2016). Many natural deposits, wherein landslide occurs, 1986; Bezerra et al., 2005; Cao et al., 2011; Do et al., 2017; Evans and
are composed of coarse and fine mixtures (Doi et al., 2019; Jiang et al., Zhou, 1995; Hatanaka et al., 2011; Koester et al., 2007; Kokusho et al.,
2015; Wen et al., 2007). Similarly, in many man-made constructions, 2011; Lin and Chang, 2002; Sirovich, 2011; Yegian et al., 1994).
including embankments, earthfill dams and roadbeds, different per- Several experimental and field studies were performed to examine
centages of gravel may be used. Therefore, the evaluation of gravel- the susceptibility in gravelly soils. Evans and Zhou (1995) were among
sand-silt behavior as a common geomaterials mixture in natural de- the first to determine the effect of gravel contents on liquefaction re-
posits can be of interest. sistance in sandy gravel mixtures. They conducted undrained cyclic
In the past, gravelly soils were considered as non-liquefiable soils triaxial tests for gravel contents varying from 0% to 100%. Results in-
because of their properties to dissipate excess pore water pressure in dicated that the liquefaction resistances in sand-gravel mixtures


Corresponding author.
E-mail address: Naeini_h@ikiu.ac.ir (S. Abolhasan Naeini).

https://doi.org/10.1016/j.enggeo.2020.105729
Received 17 February 2019; Received in revised form 10 November 2019; Accepted 14 June 2020
Available online 30 June 2020
0013-7952/ © 2020 Elsevier B.V. All rights reserved.
H. Rahmani and S. Abolhasan Naeini Engineering Geology 275 (2020) 105729

increase rapidly with gravel content. For example, a composite of sandy and Chang, 2002; Wang and Wang, 2017; Xenaki and Athanasopoulos,
gravel with 40% relative density and gravel content of 40% behaves 2008). The occurrence of liquefaction after the 1988 Armenia earth-
similar to sand with 65% relative density. Amini and Sama (1999) quake in a sandy gravel alluvium deposit with almost 25% silt content
carried out a series of stress-controlled undrained cyclic triaxial tests to is an actual case history, confirming this issue (Yegian et al., 1994).
examine the behavior of uniform and stratified sand–silt–gravel mix- Yegian et al. (1994) studied the geological characteristics and field
tures. They revealed that despite the difference between soil fabrics in inspection of three sandy gravel alluvial sites, with different silt con-
stratified and homogeneous specimens, there is no considerable dif- tents, located near the fault. They found that if the drainage conditions
ference in the liquefaction resistances. Additionally, the liquefaction in gravelly soils are limited either by fine grains or impermeable layers,
resistances for both stratified and homogeneous mixtures increase with liquefaction is more likely to happen. Gravelly deposits, with these
the silt content. Lin and Chang (2002) studied the effects of gravel conditions, are more probable to exist in nature (particularly, saturated
content and relative density on sandy gravel mixtures, and concluded deposits are located in the vicinity of faults or on natural slopes), and
that increasing the gravel content and relative density creates the therefore in the present study, the influence of non-plastic fine content
higher liquefaction resistance. Kokusho et al. (2004) reported that de- on gravelly soils is investigated. A series of undrained monotonic
spite the differences between particle gradations in sand and gravel triaxial tests were performed on three types of sandy gravel soils. The
mixtures with identical relative densities, only a slight difference was non-plastic fine content percentage was varied in each type of the soil.
observed in undrained cyclic shear strength. However, the undrained The influences of the fine content variations on undrained monotonic
monotonic shear strength for well-graded soils is greater than that for behavior, residual strength, steady state line and critical state friction
poorly-graded soils. Hence, it is very complex to predicate the lique- angle were obtained.
faction behavior of gravelly soils solely based on the soil gradations. Lin
et al. (2004) implemented the in situ large hammer penetration tests 2. Materials and methods
(LPT) and shear wave velocity measurements. They verified these
methods with large scale cyclic triaxial laboratory tests, and concluded 2.1. Tested soil
that both methods are suitable for liquefaction resistance evaluation in
gravelly soils. Ghionna and Porcino (2006) studied the effect of sample Different mixtures of gravel, sand and silt were used for the la-
preparation methods, such as Water Sedimentation (WS) and Air Plu- boratory tests. The gravel was obtained from natural resources, having
viation (AP), on cyclic liquefaction resistance. They compared the rounded or subrounded shapes and particle size between 4.75 and
cyclic liquefaction resistances for undisturbed and reconstituted sam- 9.5 mm (GP according to the Unified Soil Classification System or
ples in both isotropically and anisotropically consolidation conditions. USCS), specific gravity (Gs) of 2.69, mean particle size (D50) of 6.72 and
The results showed that, for both different consolidation conditions, the D10 = 5.09 mm. The uniform clean sand with grain size of 0.15–2 mm
liquefaction resistance is estimated more conservatively in AP method (SP, according to USCS), D50 = 0.72 mm and specific gravity of 2.72
compared to WS method. Also, behavior of samples in anisotropically was obtained. The silt (ML, according to USCS) was used as a non-
consolidation greatly depends on loading conditions. Xenaki and plastic fine content (F.C) with specific gravity of 2.67 and D50 = 0.027.
Athanasopoulos (2008) carried out cyclic triaxial and resonant column As stated, the main objective of this study is to investigate the non-
tests on sandy–silty clays and sand–gravel mixtures, used as earthfill plastic fine content influence on undrained behavior of gravelly soils
dam materials. Results showed that liquefaction resistance of sandy (GP or GM according to USCS). However, since gravel and silt mixtures
gravels is greatly dependent on the relative density. Loose specimen without sand particles are less common in nature, three types of sandy
with relative density of 20% liquefies, unlike the dense specimen with gravels were selected as base gravelly soils, categorized below:
relative density of 90%. Cao et al. (2011) performed in situ investiga- “Type-A” with sand-to-gravel content ratio (S/G) of 0.25, “Type-B”
tions on 118 identified liquefaction sites after 2008 Wenchuan earth- with S/G = 0.43 and “Type-C” with S/G = 0.67.
quake in China. They used the Dynamic Penetration Test (DPT) to es- In order to isolate the impacts of the fine content, sand-to-gravel
timate the liquefaction resistance of gravelly soils. Chang et al. (2014) ratio was kept constant in each soil type. Various percentages of silt
examined the undrained behavior of gap-graded gravelly soils (cate- (0 ~ 45%) were carefully added to the base soil. 18 mixtures of sandy
gorized in three groups as sand-like, gravel-like and in-transitions soils). gravels with different silt contents were prepared (see Table 1).
They concluded that for sand-like soils, gravel content has a small or Although the main objective of this research is to study the prop-
negligible influence on the liquefaction resistance; however, for gravel- erties of silty gravel (GP or GM), since majorities of the liquefied
like specimens, the sand grains play a significant role on the liquefac- gravelly sites consists of sand and gravel mixtures, and the pure gravel
tion resistance for samples with sand content percentages higher than is less reported (Evans and Zhou, 1995), the sandy gravel (instead of
7%. Do et al. (2017) conducted a series of undrained monotonic triaxial pure gravel) was selected as the base soil. Furthermore, the gravel
tests on gravelly soils with various gravel contents of 0%, 20%, 40% content in liquefied sites has been reported even up to 80% (Guoxing
and 60%, and with different initial densities and confinements. They et al., 2018). Therefore, three types of sandy gravel, according to
showed that when gravel content of gravelly soil increases, the lique- Table 1, were selected in a way to ensure that with increasing the silt
faction resistance increases. Wang and Wang (2017) studied the li- content, the soil type still remains GM or GP. In fact, the sand-to-gravel
quefaction characteristic of gravelly soils using the undrained cyclic ratio of 0.25 represents a high percentage of gravel (80% gravel) and
triaxial test under strain control conditions. Similarly, they reported the sand-to-gravel ratio of 0.67 represents high amount of sand in
that in gravelly soils, the liquefaction resistance is proportional to the gravel-sand-silt mixtures, where increasing the silt content does not
gravel contents. change the soil type.
Reviewing the technical literature available for the liquefaction To summarize, by selecting these three different ratios, one is able
susceptibility of different sand-fine soils (Belkhatir et al., 2011; Chang to cover the effects of sand contents in a wide range (from 20% sand
et al., 2011; Cherif Taiba et al., 2016; Ecemis and Karaman, 2014; content to 40%) as well as to study the impacts of the silts on the un-
Lashkari, 2016; Md. Mizanur and Lo, 2012; Missoum et al., 2013; drained behavior of gravelly mixtures.
Monkul et al., 2016; Naeini and Baziar, 2004; Porcino and Diano, Particle size distribution curves for the Type-A, B and C soils are
2017), one may find it very interesting to study the impacts of fine plotted in Fig. 1.
content on the liquefaction susceptibility for gravelly soils. In some The maximum and minimum void ratios were obtained by following
studies on liquefaction of gravelly soils, tested soils contained different the procedures specified in ASTM test methods (ASTM D4253, 2016;
percentages of gravel, sand and fine, as these combinations are found ASTM D4254, 2016), and then the desired relative density (Dr) for each
abundantly in nature (Amini and Sama, 1999; Kokusho et al., 2004; Lin mixture was assessed. Referring to ASTM standards, there is a limitation

2
H. Rahmani and S. Abolhasan Naeini Engineering Geology 275 (2020) 105729

Table 1
Specifications of different mixtures.
Soil type Sample S/G1 G.C2 (%) S.C3 (%) F.C4 (%) Gs USCS
identifier

Type-A A0 0.25 80 20 0 2.70 GP


A10 0.25 72 18 10 2.69 GP-GM
A20 0.25 64 16 20 2.69 GM
A30 0.25 56 14 30 2.69 GM
A40 0.25 48 12 40 2.69 GM
A45 0.25 44 11 45 2.68 GM

Type-B B0 o.43 70 30 0 2.70 GP


B10 o.43 63 27 10 2.70 GP-GM
B20 o.43 56 24 20 2.69 GM
B30 o.43 49 21 30 2.69 GM
B40 o.43 42 18 40 2.69 GM
B45 o.43 38.5 16.5 45 2.69 GM

Type-C C0 0.67 60 40 0 2.70 GP


C10 0.67 e 36 10 2.70 GP-GM
C20 0.67 48 32 20 2.70 GM
C30 0.67 42 28 30 2.69 GM
C35 0.67 39 26 35 2.69 GM
C40 0.67 36 24 40 2.69 GM

1
Sand to gravel ratio, 2Gravel content, 3Sand content, 4Fine content.

on the maximum fine content which could be present in the cohe-


sionless soils. Therefore, for mixtures with fine content higher than
20%, the graduated cylindrical method (described by Lade et al.
(1998)) was used to determine the minimum void ratio. Fig. 2 shows
the minimum void ratio (emin) and maximum void ratio (emax) as a
function of non-plastic fine content for all three types of mixtures. It is
observed that void ratios have an almost V-shape for different mixtures,
and the slight difference in the trends is due to the dependency of these
curves on the particles size and the coarse and fine grains contents
(Lade et al., 1998).

2.2. Sample preparation and experiment procedures

Undrained monotonic triaxial tests were carried out on cylindrical


samples with diameter of 69 mm and height of 140 mm. Wet tamping
method was used due to its advantages for making loose and homo-
geneous samples without creating any segregation of grains.
Additionally, due to the significance of specimen preparation method in
triaxial tests, all samples were made with high accuracy under exactly
identical conditions. First, the weight of each component of the mix-
tures (as described in Table 1) was calculated based on the minimum
and maximum void ratios to achieve the desired initial relative density.
Then the components were carefully mixed together in dry conditions
to produce the uniform gravel-sand-silt mixtures. To keep the densities
of samples constant while varying the fine content percentages, the
initial relative density (Dr before consolidation) was selected in a way
to ensure that the target relative density of samples (Dr after con-
solidation) was approximately 35% (i.e., Dr = 32%–38%). The appro-
priate initial relative densities for each mixture and for various effective
confining stresses were obtained using iterative efforts. After dry
mixing, specimens were tamped in a split mold in seven equal layers
with the moisture content of 3%–5%. The height of layers was exactly
Fig. 1. Grain size distribution curves for mixtures of (a) Type-A; (b) Type-B; (c)
monitored. The surface of each layer was scarified to maximize the Type-C.
interaction between layers. After applying the partial vacuum to
maintain the specimen, the triaxial chamber was assembled and filled
with water to remove all air spaces in the chamber. Then, with applying pressure of 200 kPa was applied with increment of 35 kPa, and the
the effective confining stress of 20 kPa, CO2 was passed from bottom of degree of saturation was evaluated using pore pressure parameter (B).
the sample to the top for at least 40 min to obtain fully saturation. For B values equal or greater than 0.95, the sample was assumed sa-
Subsequently, the deaired water was passed through the sample with a turated and the consolidation process began. Considering the higher
gravity pressure of less than 10 kPa to avoid creating the pre-stressed susceptibility of liquefaction at low stress range, three effective con-
specimens. The deaired water flow continued until the volume of water fining stresses of 50, 100 and 150 kPa (corresponding to the surface
flowing out was approximately two times the sample volume. A back depths) were selected. The variations in height and volume of the

3
H. Rahmani and S. Abolhasan Naeini Engineering Geology 275 (2020) 105729

stress among all sandy gravel mixtures, implying that as sand content
increases in the pure gravel, the peak deviator stress of mixtures re-
duces. That means with increasing the sand content, the strong contact
between gravel particles decreases and therefore the structure of force
chain in the mixtures becomes weaker.
The third trend is, as silt content increases (up to 45%), undrained
shear strength of Type-A soil decreases constantly. This could be ex-
plained by considering the positions of non-plastic contents. The non-
plastic fine contents are positioned between the gravel particles, acting
as a lubricant and reducing the contacts of gravel particles. But in Type-
B and C soils, when silt content increases up to 40% and 35% respec-
tively, undrained shear strength decreases; while with further increase
in the silt content, undrained shear strength slightly increases.
Therefore, there is a threshold for silt content in Type-B and C soils after
which the contact of gravel particles has less impact on the soil re-
sistance. In fact, the gravel particles are dispersed in the fine grain
matrix. The similar thresholds for fine content were observed by many
researchers in sand-fine mixtures (Chang and Hong, 2008; Missoum
Fig. 2. Variation of global void ratio vs. fine content. et al., 2013; Monkul et al., 2017; Naeini and Baziar, 2004).
Fig. 4 shows the undrained peak deviator stress as a function of the
non-plastic fine content for all mixtures and for different effective
specimens after saturation and consolidation progression were recorded
confining pressures. Soils types and effective confining pressures were
accurately. After specimens were consolidated isotropically, the un-
noted in the legend of Fig. 4. For silt contents within 0–20%, the peak
drained monotonic triaxial shearing was applied.
deviator stress decreases in all three types of soils. Type-A and C soils
have the highest and the least peak deviator stress, respectively, for all
3. Test results and discussions effective confining pressures. At F.C = 30%, the peak deviator stress of
different specimens become closer, however the above statement is still
3.1. Monotonic undrained behavior valid; that is the peak deviator stress is inversely proportional to the
sand-to-gravel ratio. Another remarkable observation in Fig. 4 is that in
Undrained monotonic triaxial tests were conducted on three types of all confining pressures, as silt content increases form F.C = 30–40%,
sandy gravels (Type A, B and C) with varying non-plastic fine content. the peak deviator stress of Type-A and B soils decrease; while, the peak
Fig. 3 illustrates the typical stress–strain behaviors and excess pore deviator stress of Type-C soil first decreases (up to F.C = 35%) and then
water pressure versus axial strain for pure gravel and various mixtures. increases afterwards. Fig. 4 also indicates that the peak deviator stress
The mixtures were subjected to a confining pressure ofσc′ = 150 kPa of Type-A soil has the lowest value at F.C = 45%, unlike Type-B and C
with an almost similar post-consolidation relative density of 35%. The soils. The peak deviator stress of Type-A, B and C soils reach the lowest
results of pore pressures were corrected for membrane compliance ef- value at F.C = 45%, F.C = 40% and F.C = 35%, respectively. Thus, the
fect by measuring the axial strain in consolidation process and as- sand-to-gravel ratio is an important parameter to identify the fine
suming the isotropic compression behavior (it was assumed that ske- content percentage at which the mixtures resistance behavior changes.
letal volume change is approximately three times the axial strain), The different behavior is caused by the interaction between individual
according to Ansal and Erken (Ansal and Erken, 1996). Fig. 3 clearly particles, discussed later in detail.
indicates that in all three types of sandy gravels, the non-plastic fine has
a significant role in undrained shear strength. Some trends are found in
Fig. 3. The first one is all gravelly base soils with zero non-plastic fine 3.2. Static liquefaction
(silt) content have a strain-hardening behavior (A0, B0 and C0 in
Fig. 3(a), (c) and (e)) or dilative tendency (A0, B0 and C0 in Fig. 3(b), Due to growing concerns of flow (static) liquefaction and strain
(d) and (f)). The similar trend was observed in the pure gravel. Further softening of soils in geotechnical applications, the influences of non-
increase in the silt content causes transition behavior from dilative to plastic fine content and sand-to-gravel ratio on liquefaction were stu-
contractive tendency (i.e., A30, B20 and C10) and eventually, the died. The stress path was used in the q-p' plane, where q is the deviator
σ ′ + 2σ ′
mixtures with high non-plastic fine content exhibit a strain-softening or stress and p′ = 1 3 3 is the mean effective stress. Depending on the
contractive behavior (the mixtures with the silt content higher than undrained shear behavior, the granular soils are categorized into non-
30%, 20% and 10% for Type-A, B and C soils, respectively). For ex- flow liquefaction (NF), limited flow liquefaction (LF) and flow lique-
ample, the specimen A0 shows a strong dilative response without losing faction (F) types (Lashkari, 2016). The non-flow types exhibit a strain
strength at various axial strains; while the specimen A45 is extremely hardening and dilation behavior in undrained shearing until it reaches
contractive, exhibiting a considerable loss of strength after reaching the the steady state at large axial strains. Fig. 5(a) presents the q-p' diagram
initial peak strength at a very small axial strain (less than 1%). Varia- of specimen A10 at three confining pressures, showing the non-flow
tions in the undrained behavior of mixtures occur at different fine behavior. Specimen B20 shows the limited flow type in which the stress
contents. In other words, the influence of non-plastic fine content on path exhibits contractive behavior at low axial strains even after initial
stress–strain behavior of mixed soils depends on the grain size dis- peak, and then exhibits dilation behavior for sufficiently high strains,
tribution of base sandy gravels. and eventually follows the critical state line (Fig. 5(b)). Static or flow
The next trend was found by reviewing the mixtures with zero silt liquefaction is defined when soils lose the shear strength dramatically
content (A0, B0, C0 and pure gravel). The pure gravel has the greatest after reaching the peak strength at a small strain. An example of this
peak deviator stress among all the samples. When sand is added to the behavior is presented in Fig. 5(c) for specimen C40.
samples, the specimen A0 (S/G = 0.25) has the greatest peak deviator To show how static liquefaction susceptibility varies with fine
stress among all sandy gravel mixtures (Fig. 3 (a)), compared to the contents, Fig. 6 presents the typical q-p' diagrams for all three types of
deviator stress of B0 with S/G = 0.43 (Fig. 3 (c)) and C0 with S/ soils at the effective confining pressure of 150 kPa. It clearly indicates
G = 0.67 (Fig. 3(e)). The specimen C0 has the smallest peak deviator that the flow liquefaction is impacted by the non-plastic fine content

4
H. Rahmani and S. Abolhasan Naeini Engineering Geology 275 (2020) 105729

Fig. 3. Typical deviator stress and excess pore water pressure vs. axial strain at effective confining pressure of 150 kPa for Type-A, B and C mixtures.

5
H. Rahmani and S. Abolhasan Naeini Engineering Geology 275 (2020) 105729

Fig. 4. Influence of silt content on peak strengths of mixtures for various effective confining pressures.

and base gravelly soils characteristics. For silt contents within 0–10%, obtained from undrained triaxial tests for three types of mixtures. Fig. 7
all three types of soils, except C10 (which has a limited flow behavior), clearly highlights the significant influences of non-plastic fines contents
have a non-flow behavior (A0, B0, C0, A10 and B10). For silt content of on the steady state lines. For Type-A gravelly soils (Fig. 7(a)), it is
20%, A20 shows a stable response, however B20 shows the limited flow concluded that SSL moves downward in the ec–log p' plane as silt
behavior (or temporary liquefaction), and C20 shows the flow behavior. content increases. That means the fine contents are located between
For silt content of 30%, A30 shows the limited flow behavior, while B30 coarse particles and, therefore, they may reduce the interactions be-
and C30 both show the flow behavior. When silt content is 40%, all tween gravel particles. Fig. 7(b) indicates that steady state lines for
three types of soils show the flow behavior (A40, B40 and C40). It is Type-B soils move downward as silt content increases to 40% and then
worth noting that the liquefaction resistance for C40 is greater than that moves upward for silt content of 45%. When the fine content becomes
for C35, because the specimen C40 has a greater peak point in q–p′ sufficiently large, the behavior of mixed soils is controlled by fine
space (considered as onset of flow liquefaction (Md. Mizanur and Lo, grains, and the contact of gravel particles is less significant. The similar
2012)). In other words, the liquefaction susceptibility of Type-C soil trend was also observed for Type-C soils, confirming as silt content
increases as silt content increases up to 35%, and then decreases when increases to 35%, the steady state line moves downward, but for spe-
silt content is 40% (Fig. 6(c)). The similar trend exists for Type-B soil cimens with 40% silt content, this line moves up (Fig. 7(c)). The silt
for which the liquefaction resistance decreases for silts content below content threshold for Type-C soils (with higher sand content) is lower
40%, and then increases for silt content of 45% (Fig. 6(b)). However, than Type-B soils, and it is concluded that in gravelly mixtures, this
for Type-A soil, the liquefaction resistance reduces continuously for all threshold is also a function of sand content. It seems that Type-B and C
silt contents considered, up to 45% (Fig. 6(a)). It is concluded that the soils with higher sand-to-gravel ratio relative to the Type-A soil, act
soil behavior is greatly dependent on the sand-to-gravel ratio of base almost similar to the sandy soils. This conclusion was previously re-
gravel and the non-plastic fine content percentage. As expected, the ported in technical literature, where the influence of fine content on
threshold for non-plastic fine content of mixed gravelly soils (combi- sandy soils was studied (Monkul et al., 2017; Naeini and Baziar, 2004).
nation of gravel, sand and silt) decreases as the sand-to-gravel ratio of
base gravel increases. In fact, the mixtures start to behave similar to the 3.4. Equivalent intergranular void ratio
sandy soils when the sand-to-gravel ratio is sufficiently high.
In past decades, the evaluation of how fine content may impact the
behavior of granular soils has been one of the interesting topics in
3.3. Steady state line (SSL) geotechnical researches. Due to inability of global void ratio (e) to fully
evaluate the real influences of fine content, the equivalent intergranular
The steady state (SS) is defined as a state at which the soil deforms void ratio (eg∗) theories (suggested by Thevanayagam et al (2002)) were
continuously at a constant shear stress and a constant volume (Roscoe used instead as valuable indices to characterize the mixed granular
et al., 1958). The steady state is commonly used as a significant prop- (mixed sandy or gravelly soils) responses (Chang and Phantachang,
erty in practical applications and fundamental studies of laboratory 2016; Monkul and Ozden, 2007). Referring to this theory, for mixtures
tests (Been and Jefferies, 1985). In cases of limited flow liquefaction in which the fine content (F.C) is less than the certain fine content
behavior, the deviator stress reduces to a transient minimum value, threshold (F.Cth), coarse grain contact and fine grain contact play the
denoted as Quasi-Steady State (QSS). This temporary drop in shear primary and secondary roles, respectively. In fact, the behavior of
stress occurs after reaching the initial peak at the beginning of un- mixture is dominated by the coarse particles. On the other hand, for
drained loading. Deviator stress then gains strength towards the steady mixtures with F.C greater than F.Cth, fine grain friction has a significant
state at the greater axial strains (Ishihara, 1993). Usually, the re- effect on mixed soils responses, and intergranular contact friction plays
lationship between the consolidated void ratio (ec) and the mean ef- the secondary role. Thus, the behavior of mixture is dominated by the
fective stress (P') has a linear trend in semi-log plane ec-logp' (Been and fine grains (Thevanayagam et al., 2002). For mixtures with F.C <
Jefferies, 1985). Fig. 7 shows the steady state lines, which were F.Cth, the equivalent intergranular void ratio is defined by

6
H. Rahmani and S. Abolhasan Naeini Engineering Geology 275 (2020) 105729

Fig. 5. Typical effective stress paths for (a) non-flow behavior; (b) limited li-
quefaction behavior; (c) flow liquefaction behavior.
Fig. 6. Typical effective stress paths at confining pressure of 150 kPa for mix-
tures of (a) Type-A; (b) Type-B; (c) Type-C.
Thevanayagam et al. (2002):

e + (1 − b) F . C The parameter b in Eq. (1) varies from 0 to 1.0. However, choosing


eg∗ =
1 − (1 − b) F . C (1) the appropriate value for b may not be very straightforward as many
effective factors, including fine content percentage and grading char-
where parameter b refers to the fine grains contribution in active in- acteristics of mixtures, can change that (Md. Mizanur and Lo, 2012;
tergrain contacts, i.e. the greater the value of b, the higher participation Mohammadi and Qadimi, 2015; Rahman et al., 2011; Rahman and Lo,
of fine grains in active particle contacts. 2008). Rahman and Lo (2008) introduced the semi-empirical equation

7
H. Rahmani and S. Abolhasan Naeini Engineering Geology 275 (2020) 105729

based on F.C, F.Cth and grain size ratio (χ = D10/d50) as:


b = {1 − exp[−μ (F . C / F . Cth/ k )n]}(r ∗ F . C / F . Cth)r (2)

where r = χ −1, D10 is the size of coarse particles at 10% finer,


k = 1 − r 0.25 , μ and n are constant fitting parameters.
Reviewing the gravel, sand and silt fractions in each mixture
(Table 1), it is understood that the behavior of soils is not solely dic-
tated by the silt content. Instead, the combination of sand and silt
content together is the key factor in determining the behavior of mixed
gravelly soils. For example, samples A20 (G.C: 64% and S.C: 16%), B20
(G.C: 56% and S.C: 24%) and C20 (G.C: 48% and S.C: 32%) have non-
flow, limited flow and flow liquefaction behaviors, respectively, al-
though the amount of silt is identical in each of those three specimens.
Additionally, A30 with the total sand and silt content of 44% shows
limited flow liquefaction, similar to B20 with the total sand and silt
content of 44%. In a similar argument, B30 with the total sand and silt
content of 51% has a flow liquefaction behavior almost like C20 with
the total sand and silt content of 52%. Additionally the sizes of sand
(d50 = 0.72 mm) and silt (d50 = 0.027 mm) grains are very different
from gravel particles (d50 = 6.72 mm). Therefore, the combination of
sand and silt content is regarded as an effective fine content (F.C*) and
gravel particles are regarded as coarse grains in this study. We use the
intergrain concept to explain the results. Due to differences between
mean particle sizes in sand and silt, the weighted average of mean grain

size (d50 ) for the fine portion, used in Eq. (2), is defined as:

∗ d50(sand) ∗ (S. C ) + d50(silt ) ∗ (M . C )


d50 =
(S. C + M . C ) (3)

where S.C and M.C refer to sand content and silt content, respectively.
As mentioned before, selecting the appropriate value for b is one of the
crucial factors when using this concept. Theoretically, the equivalent
intergranular critical state line for mixtures with fine contents less than
F.Cth (CSLg∗) in eg∗ − log p′ plane, matches with the critical state line of
clean coarse particles (CSL0 corresponds to F.C*=0%). Because, the
values of eg∗ and e (global void ratio) are equal in mixtures with zero fine
content (F.C*=0%). Therefore, CSL0 can be used as a reference line to
estimate the appropriate value for b (Mohammadi and Qadimi, 2015).
There are many different methods to determine the fine content
thresholds (transitional) in coarse and fine mixtures. One approach is to
use the index laboratory test results: the fine content percentages, at
which emin and emax are minimum, are the fine content thresholds. In
another approach, the fine content percentages, where the value for
steady state line in monotonic tests is minimum, are taken as the
thresholds (Zuo and Baudet, 2015). Therefore, considering Figs. 2 and
7, it is clear that the fine content thresholds for Type-A, B and C soils are
respectively 45%, 40% and 35%. In this study, the best value of b for
each mixture has been approximated by matching the CSLg∗ (obtained
from exponential trend-line of eg∗ data in eg∗ − log p′ plane) with the CSL0
(critical state line of pure gravel) (Mohammadi and Qadimi, 2015).

Knowing F.Cth and using F.C* and d50 instead of F.C and d50 in Eq. (2),
the values for n and μ were chosen in a way to ensure that CSLg∗ has the
closest coincidence with CSL0. Fig. 8 presents the diagrams of CSLg∗ and
CSL0 for each type of soil. The values for constants μ and n for three
types of gravelly soils were obtained through using a trial and error
process, and they are shown in the legends of Fig. 8. The minor dif-
ference in the values is due to dependency of μ and n on the fine and
coarse particles characteristic (Mohammadi and Qadimi, 2015; Rahman
and Lo, 2008). The summary of results for three effective confining
pressures of 50, 100 and 150 kPa is given in Table 2. It is clear that b
increases with the fine content (Table 2), but the increasing rate is slow
because of the high grain size ratio (D∕d, where D and d refer to the
coarse and fine grain sizes, respectively). Mizanur and Lo (2012) also
reported the similar finding for D/d > 7. Table 2 clearly indicates that
Fig. 7. Influence of silt content on steady state line for mixtures of (a) Type-A; for Type-A soil, as the silt content increases up to 45% (threshold
(b) Type-B; (c) Type-C.
value), the equivalent intergranular void ratio increases, unlike the
global consolidated void ratio which is decreasing. It means that for

8
H. Rahmani and S. Abolhasan Naeini Engineering Geology 275 (2020) 105729

Table 2
Results of equivalent intergranular void ratio analysis.
soil Type Sample S/G1 G.C2 (%) S.C3 (%) F.C4 (%) Gs USCS
Identifier

Type-A A0 0.25 80 20 0 2.70 GP


A10 0.25 72 18 10 2.69 GP-GM
A20 0.25 64 16 20 2.69 GM
A30 0.25 56 14 30 2.69 GM
A40 0.25 48 12 40 2.69 GM
A45 0.25 44 11 45 2.68 GM

Type-B B0 o.43 70 30 0 2.70 GP


B10 o.43 63 27 10 2.70 GP-GM
B20 o.43 56 24 20 2.69 GM
B30 o.43 49 21 30 2.69 GM
B40 o.43 42 18 40 2.69 GM
B45 o.43 38.5 16.5 45 2.69 GM

Type-C C0 0.67 60 40 0 2.70 GP


C10 0.67 54 36 10 2.70 GP-GM
C20 0.67 48 32 20 2.70 GM
C30 0.67 42 28 30 2.69 GM
C35 0.67 39 26 35 2.69 GM
C40 0.67 36 24 40 2.69 GM

1: Sand to Gravel ratio, 2: Gravel Content, 3: Sand Content, 4: Fine Content.

higher silt content, the active contacts between coarse particles de-
crease, and this is compatible with the behavior of Type-A gravelly
soils, which was discussed in the previous sections. The similar trend is
observed (Table 2) for Type-B and C mixtures with F.C less than F.Cth,
which could be confirmed by looking into the undrained monotonic
results.
Given the modified intergranualr void ratio theory is in good con-
sistency with the experimental results, these behaviors can be explained
from a microstructural perspective and by considering the active con-
tact between the coarse-particles and the position of fine-grains in the
mixtures.The results for samples with zero silt content (A0, B0, C0 and
pure gravel) shows that as the sand-to-gravel ratio increases from pure
gravel to C0, the mixture structure becomes weaker. In fact, by locating
sand grains between the gravel particles, the active contacts of mixture
reduce. Increasing the silt content also showed similar behavior in all
three types of sandy gravel mixtures. This implies, given the huge dif-
ferences in the sizes of sand and silt grains to that of the gravel particles,
sand and silt act as fine-grains in gravelly mixtures. In the framework of
the intergranualr void ratio theory, some of the fine-grains are entirely
located in the void space between the coarser-grains and only play the
filler role. However, other parts of the fine-grains act as a lubricant
when they are located between coarse-grains, and consequently reduce
strong contacts between them.
In fact, up to a certain limit of fine-grains content (fine content
threshold) in gravelly mixtures, the mechanical behavior is affected
primarily by the coarser-particle contacts and fine-grains have a sec-
ondary role in the transfer of inter-grain forces. But with further in-
crease in the fine content of mixture, the role of the fine-grains becomes
more significant. After reaching thresholds for fine content (corre-
sponding to 45%, 40% and 35% of silt content for Type-A, B and C
sandy gravels, respectively), fine-grains start to play a primary role in
the behavior of mixtures, while the role of coarser-particles starts to
diminish. In other words, in this situation, the gravel particles disperse
in the fine-grain matrix and will not play a significant role in the
transfer of inter-grain forces.
For this reason, in the mixtures with a higher sand-to-gravel ratio,
the change of the mixture response from stable to limited flow lique-
Fig. 8. Compatibility of CSL*g with CSL0 in various mixtures, (a) Type-A; (b) faction occurs at lower silt content. This trend is also observed when the
Type-B; (c) Type-C.
response of mixtures changes from the limited flow liquefaction to the
flow liquefaction. Therefore, it will be crucial to consider both sand and
silt contents in the mixtures if one desires to study the liquefaction
susceptibility in gravel-sand-silt mixtures. Given the varieties of soil

9
H. Rahmani and S. Abolhasan Naeini Engineering Geology 275 (2020) 105729

Fig. 9. Influence of silt content on critical state friction angle of Type-A, B and C Fig. 10. Critical state friction angle vs. equivalent intergranular void ratio for
mixtures. Type-A, B and C mixtures.

combinations in nature, this conclusion could be of great importance the particles contacts.
and very applicable. Because it suggests even mixtures with low per-
centages of non-plastic silt content may also have the static liquefaction 3.6. Residual strength (Sus)
potential.
The residual strength or steady-state strength of liquefied soils is
3.5. Critical state friction angle another important variable in geotechnical applications. It is defined as
the deviator stress of granular soils mobilized at the steady state
The friction angle of granular materials at critical state is known as (Ishihara, 1993). But in some cases, samples experience the Quasi-
critical state friction angle (Φcs). This friction angle is used as a sig- Steady State (QSS) for approximately moderate range of strains prior to
nificant characteristic of soils in geotechnical applications and con- reach the steady sate. In this state, the shear stress is less than the de-
stitutive modeling (Yang and Luo, 2018). Using the compression viator stress for steady state. Therefore, the true identification of
triaxial test results, the critical state friction angle is computed as: minimum shear stress (usually known as Sus) is very important. The
residual strength is defined as (Ishihara, 1993):
3M
sin Φcs = qs M
6+M (4) sus = cos φcs = cos φcs × pS′
2 2 (5)
q
here M is the critical stress ratio ( P′ at critical state). Fig. 9 shows the
influence of silt content on the critical state friction angle for all three where qs and ps′ refer to the deviator stress and the effective mean stress,
types of soils. When silt content is zero, the critical state friction angle respectively, at the QSS (state with the minimum shear stress), which
decrease as sand-to-gravel ratio increases. This conclusion is in agree- can be obtained from the q-p' diagram. Ishihara (Ishihara, 1993) de-
fined the residual strength, normalized by the effective isotropic con-
ment with the results of Simoni and Houlsby (2006), who reported that
as gravel content increases in the sand-gravel mixtures, the critical state fining pressure (σc′):
friction angle increases as well. This could also be confirmed by com- sus M p′
paring the equivalent intergranular void ratios of A0, B0 and C0 (ac- = cos φcs × s
σ ′c 2 p′c (6)
cording to Table 2, the average values of eg∗ for A0, B0 and C0 samples
are respectively 0.779, 0.804 and 0.813). According to Fig. 9, the cri- where pc′ in the q-p' space is used instead ofσc′. Fig. 11 shows the residual
tical state friction angle varies from 29 to 47 degree. To compare, this strengths of mixtures, exhibiting limited or flow liquefaction behavior,
value is within 21 to 38 degree for sandy soils with the almost identical as function of silt contents at three effective confining pressures of 50,
relative density (Negussey et al., 1988; Sadrekarimi and Olson, 2011; 100 and 150 kPa. In Fig. 11, it is clear that the residual strengths of
Yang and Luo, 2018). The difference could be due to strong contact Type-A mixtures decrease with adding the silt content up to 45% for all
between gravel particles, relative to the sand grains, and because the confining pressures. However, for Type-B mixtures, as the silt content
critical state friction angle is highly dependent on the particles shape. increases up to 40%, the Sus decreases, and with further increase in silt
Fig. 10 presents the critical state friction angle as a function of content (45%), the residual strength increases. The similar trend was
equivalent intergranular void ratio for mixtures with F.C < F.Cth. also observed for Type-C mixtures, however the fine content threshold
Figs. 9 and 10 clearly demonstrate that for silt contents below the fine was 35% this time, as expected. One of the remarkable points in Fig. 11
content thresholds (45%, 40% and 35% for Type-A, B and C soils), the is that the ratio of Sus/σc′ is reasonably constant for each specimen. The
critical state friction angle decreases when the equivalent intergranular similar trend is reported in previous studies on sandy soils (Baziar and
void ratio increases. This trend is completely consistent with the Dobry, 1995; Naeini and Baziar, 2004). This is particularly of im-
equivalent intergranular void ratio results, unlike the global con- portance as it allows to estimate the Sus in the field without measuring
solidated void ratio results. Then, with further increase in silt content the in situ density. The influence of silt content on Sus/σc′ for all types of
and after passing the threshold, the critical state friction angle in- liquefied soils is compared in Fig. 12. Similar to the results of previous
creases. Therefore, knowing the accurate value of the equivalent in- sections, this parameter is also affected by both sand-to-gravel ratio and
tergranular void ratio is very important in order to fully explain the silt content. Sus/σc′ for Type-A mixtures continuous to decreases as slit
results of the critical state friction angle, which is strongly affected by content increases up to 45%, but Sus/σc′ reduces for Type-B and C

10
H. Rahmani and S. Abolhasan Naeini Engineering Geology 275 (2020) 105729

Fig. 12. Influence of silt content on normalized residual strength of Type-A, B


and C mixtures with limited or flow liquefaction behavior.

Fig. 13. Normalized residual strength vs. equivalent intergranular void ratio for
Type-A, B and C mixtures with limited or flow liquefaction behavior.

4. Conclusion

Influences of silt content on the behavior of gravelly soils were


studied using undrained monotonic triaxial tests. Considering the
Fig. 11. Residual strength vs. initial confining pressure for specimens with combinations of natural soils, three types of sandy gravels with dif-
limited or flow liquefaction behavior; (a) Type-A; (b) Type-B; (c) Type-C. ferent sand-to-gravel ratios (S/G) were selected as base soils (Type-A
with S/G = 0.25, Type-B with S/G = 0.43 and Type-C with S/
G = 0.67). Various percentages of silt as a non-plastic fine content
mixtures up to 40% and 45% silt content respectively, and then in-
(0–45%) were added to the base soils to produce different mixtures
creases again with adding further silt content.
(Table 1). All specimens were consolidated at three effective confining
Comparing the results of Table 2 and Fig. 12, it is seen that although
pressures. The relative densities of the samples were carefully kept
the global consolidated void ratio (ec) decreases with fine content, the
constant to avoid the density effect on the responses. The following
normalized residual strength reduces as well. Therefore, the global void
conclusions were made:
ratio cannot be a clear indication to predict the normalized residual
strength. However, if the equivalent intergranular void ratio is con-
1) When silt content is zero, all types of base soils show the dilative
sidered instead, the consistency is observed; that is as silt content in-
behavior, but when non-plastic fine content is added, they start to
creases up to fine content thresholds (45%, 40% and 35% silt content
show the transition behavior first and then, with further increase in
for Type-A, B and C mixtures, respectively), the Sus/σc′ ratio decreases
the silt content, mixtures show the contractive or strain-softening
and the equivalent intergranular void ratio increases as shown in
behavior. However, the levels of silt content, at which behavior
Fig. 13.
changes, are different for various types of sandy gravels.

11
H. Rahmani and S. Abolhasan Naeini Engineering Geology 275 (2020) 105729

2) The peak deviator stresses of sandy gravel mixtures decrease as of soils using a vibratory table. Am. Soc. Test. Mater. https://doi.org/10.1520/
sand-to-gravel ratio increases. Increasing the non-plastic fine con- D4253-14.
ASTM D4254, 2016. Standard test methods for minimum index density and unit weight of
tent (up to 45%) reduces the undrained shear strength for Type-A soils and calculation of relative density. Am. Soc. Test. Mater. https://doi.org/10.
mixtures. However, for Type-B and C mixtures, the peak deviator 1520/D4254-00R06E01.1.3.
stress reduces first when silt content is below the thresholds (40% Baziar, M.H., Dobry, R., 1995. Residual strength and large-deformation potential of loose
silty sands. J. Geotech. Eng. https://doi.org/10.1061/(asce)0733-9410(1995)
and 35% for Type-B and C mixtures, respectively) and with further 121:12(896).
increases in the silt content, the peak deviator stress increases again. Been, K., Jefferies, M.G., 1985. A state parameter for sands. Géotechnique 35, 99–112.
3) All specimens with zero fine content exhibit the non-flow behavior. https://doi.org/10.1680/geot.1985.35.2.99.
Belkhatir, M., Arab, A., Schanz, T., Missoum, H., Della, N., 2011. Laboratory study on the
When silt content is added, specimens show the limited liquefaction liquefaction resistance of sand-silt mixtures: Effect of grading characteristics. Granul.
behavior, and ultimately with further increase in the fine content, Matter 13, 599–609. https://doi.org/10.1007/s10035-011-0269-0.
they show the contractive tendency and flow liquefaction behavior. Bezerra, F.H.R., da Fonseca, V.P., Vita-Finzi, C., Lima-Filho, F.P., Saadi, A., 2005.
Liquefaction-induced structures in quaternary alluvial gravels and gravelly sedi-
The influence of the non-plastic fine content on the static liquefac-
ments, NE Brazil. Eng. Geol. 76, 191–208. https://doi.org/10.1016/j.enggeo.2004.
tion susceptibility is greatly affected by the portion of sand content 07.007.
present in the base gravelly soil. For higher sand-to-gravel ratios, the Cao, Z., Leslie Youd, T., Yuan, X., 2011. Gravelly soils that liquefied during 2008
static liquefaction occurs at lower fine content. In fact, liquefaction Wenchuan, China earthquake, Ms=8.0. Soil Dyn. Earthq. Eng. 31, 1132–1143.
https://doi.org/10.1016/j.soildyn.2011.04.001.
susceptibility increases with sand-to-gravel ratio. However, after Chang, W.-J., Hong, M.-L., 2008. Effects of clay content on liquefaction characteristics of
reaching the fine content threshold, the liquefaction resistance gap-graded clayey sands. Soils Found 48, 101–114. https://doi.org/10.3208/sandf.
grows up again. 48.101.
Chang, W.J., Chang, C.W., Zeng, J.K., 2014. Liquefaction characteristics of gap-graded
4) The steady state line is also impacted by both non-plastic fine con- gravelly soils in K0 condition. Soil Dyn. Earthq. Eng. 56, 74–85. https://doi.org/10.
tent and sand-to-gravel ratio of gravelly base. The addition of silt 1016/j.soildyn.2013.10.005.
content, up to the threshold values, shifts the SSL downward but as Chang, W.J., Ni, S.H., Huang, A. Bin, Huang, Y.H., Yang, Y.Z., 2011. Geotechnical re-
connaissance and liquefaction analyses of a liquefaction site with silty fine sand in
silt content increases further (45% for Type-B specimens and 40% Southern Taiwan. Eng. Geol. 123, 235–245. https://doi.org/10.1016/j.enggeo.2011.
for Type-C specimens) the SSL moves up. In other words, the silt 09.003.
percentages of 45%, 40% and 35% represent the lowest locations of Chang, W.J., Phantachang, T., 2016. Effects of gravel content on shear resistance of
gravelly soils. Eng. Geol. 207, 78–90. https://doi.org/10.1016/j.enggeo.2016.04.
SSL for Type-A, B and C mixtures, respectively, which are basically 015.
the fine content thresholds for each type. Cherif Taiba, A., Belkhatir, M., Kadri, A., Mahmoudi, Y., Schanz, T., 2016. Insight into the
5) The identical trend was also observed for the residual strength (and effect of granulometric characteristics on the static liquefaction susceptibility of silty
sand soils. Geotech. Geol. Eng. 34, 367–382. https://doi.org/10.1007/s10706-015-
also normalized residual strength) and critical state friction angle of
9951-z.
gravelly mixtures. These variables decrease with silt content up to Do, J., Heo, S.B., Yoon, Y.W., Chang, I., 2017. Evaluating the liquefaction potential of
the threshold values, and then increase with additional increase in gravel soils with static experiments and steady state approaches. KSCE J. Civ. Eng.
silt content. Also for each specimen, the ratio of residual strength to 21, 642–651. https://doi.org/10.1007/s12205-016-1365-9.
Doi, I., Kamai, T., Azuma, R., Wang, G., 2019. A landslide induced by the 2016
the confining pressure is almost constant, which is an important Kumamoto Earthquake adjacent to tectonic displacement – Generation mechanism
conclusion needed to estimate the residual strength in field. and long-term monitoring. Eng. Geol. 248, 80–88. https://doi.org/10.1016/j.enggeo.
6) Unlike the global void ratio (even the consolidated void ratio), the 2018.11.012.
Ecemis, N., Karaman, M., 2014. Influence of non-/low plastic fines on cone penetration
equivalent intergranular void ratio theory(eg∗) , if selected appro- and liquefaction resistance. Eng. Geol. 181, 48–57. https://doi.org/10.1016/j.
priately, can strongly predict the behaviors of gravelly mixtures enggeo.2014.08.012.
with fine contents below the thresholds. Given the consistency of Evans, M.D., Zhou, S., 1995. Liquefaction behavior of sand-gravel composites. J. Geotech.
Eng. 121, 287–298. https://doi.org/10.1061/(ASCE)0733-9410(1995)121:3(287).
laboratory results with this theory, it is concluded that both sand
Ghionna, V.N., Porcino, D., 2006. Liquefaction resistance of undisturbed and recon-
and silt grains act as fine grains in gravelly mixtures. In the mixtures stituted samples of a natural coarse sand from undrained cyclic triaxial tests. J.
with fine content less than threshold, the mechanical behavior is Geotech. Geoenviron. Eng. 132, 194–202. https://doi.org/10.1061/(asce)1090-
0241(2006)132:2(194).
primarily affected by the coarser particles contact, and fine grains
Guoxing, C., Qi, W., Tian, S., Kai, Z., Enquan, Z., Lingyu, X., Yanguo, Z., 2018. Cyclic
have a secondary role in the transfer of inter-grain forces. However, behaviors of saturated sand-gravel mixtures under undrained cyclic triaxial loading.
once fine content increases beyond the threshold, the behavior of J. Earthq. Eng. 1–34. https://doi.org/10.1080/13632469.2018.1540370.
mixtures is controlled significantly by fine-grains contacts, and the Hatanaka, M., Uchida, A., Ohara, J., 2011. Liquefaction characteristics of a gravelly fill
liquefied during the 1995 Hyogo-ken Nanbu Earthquake. Soils Found. 37, 107–115.
role of gravel particles becomes less important and therefore the https://doi.org/10.3208/sandf.37.3_107.
opposite trend was observed. Given the variety of soil combinations Ishihara, K., 1993. Liquefaction and flow failure during earthquakes. Géotechnique 43,
in nature, this conclusion could be very important and applicable, 351–451. https://doi.org/10.1680/geot.1993.43.3.351.
Jiang, X., Cui, P., Ge, Y., 2015. Effects of fines on the strength characteristics of mixtures.
noting that both sand-to-gravel ratio and silt content should be Eng. Geol. https://doi.org/10.1016/j.enggeo.2015.09.011.
taken into consideration to study the gravelly mixtures behavior. Koester, J.P., Daniel, C., Anderson, M.L., 2007. In Situ investigation of liquefiable gravels.
Transp. Res. Rec. J. Transp. Res. Board 1714, 75–82. https://doi.org/10.3141/
1714-10.
Declaration of Competing Interest Kokusho, T., Hara, T., Hiraoka, R., 2004. Undrained shear strength of granular soils with
different particle gradations. J. Geotech. Geoenviron. Eng. 130, 621–629. https://doi.
org/10.1061/(asce)1090-0241(2004)130:6(621).
The authors declare that they have no known competing financial
Kokusho, T., Tanaka, Y., Kawai, T., Kudo, K., Suzuki, K., Tohda, S., Abe, S., 2011. Case
interests or personal relationships that could have appeared to influ- study of rock debris avalanche gravel liquefied during 1993 Hokkaido-Nansei-Oki
ence the work reported in this paper. earthquake. Soils Found. 35, 83–95. https://doi.org/10.3208/sandf.35.83.
Lade, P.V., Liggio, C.D., Yamamuro, J.A., 1998. Effects of non-plastic fines on minimum
and maximum void ratios of sand. Geotech. Test. J. 21, 336–347. https://doi.org/10.
References 1520/GTJ11373J.
Lashkari, A., 2016. Prediction of flow liquefaction instability of clean and silty sands. Acta
Amini, F., Sama, K.M., 1999. Behavior of stratified sand-silt-gravel composites under Geotech. 11, 987–1014. https://doi.org/10.1007/s11440-015-0413-9.
seismic liquefaction conditions. Soil Dyn. Earthq. Eng. 18, 445–455. https://doi.org/ Lin, P.S., Chang, C.W., 2002. Damage investigation and liquefaction potential analysis of
10.1016/S0267-7261(99)00011-1. gravelly soil. A/Chung-kuo K. Ch’eng Hsuch K’an. J. Chinese Inst. Eng. Trans. Chin.
Andrus, R.D., 1986. Geotechnical evaluation of a liquefaction induced lateral spread, InstEng. 25, 543–554. https://doi.org/10.1080/02533839.2002.9670729.
Thousands Springs Valley Idaho. In: Proc. 22nd Ann. Symp. Eng. Geol. Soils Eng. Lin, P.S., Chang, C.W., Chang, W.J., 2004. Characterization of liquefaction resistance in
Ansal, A.M., Erken, A., 1996. Posttesting correction procedure for membrane compliance gravelly soil: Large hammer penetration test and shear wave velocity approach. Soil
effects on pore pressure. J. Geotech. Eng. 122, 27–38. https://doi.org/10.1061/(asce) Dyn. Earthq. Eng. 24, 675–687. https://doi.org/10.1016/j.soildyn.2004.06.010.
0733-9410(1996)122:1(27). Md. Mizanur, R., Lo, S.R., 2012. Predicting the onset of static liquefaction of loose sand
ASTM D4253, 2016. Standard test methods for maximum index density and unit weight with fines. J. Geotech. Geoenviron. Eng. 138, 1037–1041. https://doi.org/10.1061/
(asce)gt.1943-5606.0000661.

12
H. Rahmani and S. Abolhasan Naeini Engineering Geology 275 (2020) 105729

Missoum, H., Belkhatir, M., Bendani, K., Maliki, M., 2013. Laboratory investigation into Roscoe, K.H., Schofield, A.N., Wroth, C.P., 1958. On the yielding of soils. Géotechnique 8,
the effects of silty fines on liquefaction susceptibility of Chlef (Algeria) Sandy Soils. 22–53. https://doi.org/10.1680/geot.1958.8.1.22.
Geotech. Geol. Eng. 31, 279–296. https://doi.org/10.1007/s10706-012-9590-6. Sadrekarimi, A., Olson, S.M., 2011. Critical state friction angle of sands. Geotechnique 61,
Mohammadi, A., Qadimi, A., 2015. A simple critical state approach to predicting the 771–783. https://doi.org/10.1680/geot.9.P.090.
cyclic and monotonic response of sands with different fines contents using the Simoni, A., Houlsby, G.T., 2006. The direct shear strength and dilatancy of sand-gravel
equivalent intergranular void ratio. Acta Geotech. 10, 587–606. https://doi.org/10. mixtures. Geotech. Geol. Eng. 24, 523–549. https://doi.org/10.1007/s10706-004-
1007/s11440-014-0318-z. 5832-6.
Monkul, M.M., Etminan, E., Şenol, A., 2017. Coupled influence of content, gradation and Sirovich, L., 2011. Repetitive liquefaction at a gravelly site and liquefaction in over-
shape characteristics of silts on static liquefaction of loose silty sands. Soil Dyn. consolidated sands. Soils Found. 36, 23–34. https://doi.org/10.3208/sandf.36.4_23.
Earthq. Eng. 101, 12–26. https://doi.org/10.1016/j.soildyn.2017.06.023. Thevanayagam, S., Shenthan, T., Mohan, S., Liang, J., 2002. Undrained fragility of clean
Monkul, M.M., Etminan, E., Şenol, A., 2016. Influence of coefficient of uniformity and sands, silty sands, and sandy silts. J. Geotech. Geoenvironmental Eng. 128, 849–859.
base sand gradation on static liquefaction of loose sands with silt. Soil Dyn. Earthq. https://doi.org/10.1061/(asce)1090-0241(2002)128:10(849).
Eng. 89, 185–197. https://doi.org/10.1016/j.soildyn.2016.08.001. Wang, Y., Wang, Y.L., 2017. Liquefaction characteristics of gravelly soil under cyclic
Monkul, M.M., Ozden, G., 2007. Compressional behavior of clayey sand and transition loading with constant strain amplitude by experimental and numerical investigations.
fines content. Eng. Geol. 89, 195–205. https://doi.org/10.1016/j.enggeo.2006.10. Soil Dyn. Earthq. Eng. 92, 388–396. https://doi.org/10.1016/j.soildyn.2016.10.029.
001. Wen, B.P., Aydin, A., Duzgoren-Aydin, N.S., Li, Y.R., Chen, H.Y., Xiao, S.D., 2007.
Naeini, S.A., Baziar, M.H., 2004. Effect of fines content on steady-state strength of mixed Residual strength of slip zones of large landslides in the Three Gorges area. China.
and layered samples of a sand. Soil Dyn. Earthq. Eng. 24, 181–187. https://doi.org/ Eng. Geol. 93, 82–98. https://doi.org/10.1016/j.enggeo.2007.05.006.
10.1016/j.soildyn.2003.11.003. Xenaki, V.C., Athanasopoulos, G.A., 2008. Dynamic properties and liquefaction resistance
Negussey, D., Wijewickreme, W.K.D., Vaid, Y.P., 1988. Constant-volume friction angle of of two soil materials in an earthfill dam-Laboratory test results. Soil Dyn. Earthq. Eng.
granular materials. Can. Geotech. J. 25, 50–55. https://doi.org/10.1139/t88-006. 28, 605–620. https://doi.org/10.1016/j.soildyn.2007.10.001.
Porcino, D.D., Diano, V., 2017. The influence of non-plastic fines on pore water pressure Yang, J., Luo, X.D., 2018. The critical state friction angle of granular materials: does it
generation and undrained shear strength of sand-silt mixtures. Soil Dyn. Earthq. Eng. depend on grading? Acta Geotech. 13, 535–547. https://doi.org/10.1007/s11440-
101, 311–321. https://doi.org/10.1016/j.soildyn.2017.07.015. 017-0581-x.
Rahman, M.M., Lo, S.R., 2008. The prediction of equivalent granular steady state line of Yegian, M.K., Ghahraman, V.G., Harutiunyan, R.N., 1994. Liquefaction and embankment
loose sand with fines. Geomech. Geoengin. 3, 179–190. https://doi.org/10.1080/ failure case histories, 1988 Armenia Earthquake. J. Geotech. Eng. 120, 581–596.
17486020802206867. https://doi.org/10.1061/(asce)0733-9410(1994)120:3(581).
Rahman, M.M., Lo, S.R., Baki, M.A.L., 2011. Equivalent granular state parameter and Zuo, L., Baudet, B.A., 2015. Determination of the transitional fines content of sand-non
undrained behaviour of sand-fines mixtures. Acta Geotech. 6, 183–194. https://doi. plastic fines mixtures. Soils Found. 55, 213–219. https://doi.org/10.1016/j.sandf.
org/10.1007/s11440-011-0145-4. 2014.12.017.

13

You might also like