Bulletin of Engineering Geology and the Environment (2022) 81: 261

https://doi.org/10.1007/s10064-022-02739-3

ORIGINAL PAPER

Shear strength of grouted clay: comparison of triaxial tests to direct

shear tests
Zhanchao Yin1 · Qingsong Zhang1 · Xiao Zhang1 · Jiaqi Zhang1 · Xianghui Li1

Received: 16 September 2021 / Accepted: 13 May 2022 / Published online: 7 June 2022
© Springer-Verlag GmbH Germany, part of Springer Nature 2022

Abstract
Slurry–soil interface direct shear (IDS) tests and the triaxial compression (TC) tests are currently used to determine the
shear strength of grouted soil. However, many previous studies focused on sandy soils, and limited attention has been paid
to clayey soil. This study introduces two grouting test methods designed to fabricate the grouted clay specimen suitable
for TC and IDS testing. A group of grouting orthogonal tests is designed, and the shear strength parameters are calculated
and compared based on the Mohr–Coulomb (MC) criterion. Findings showed that outcomes obtained from the IDS test are
unreliable. The test results showed that the shear strength obtained from IDS test yields more serious deviations and lower
MC criterion fitness. The test also showed that shear strength parameters under IDS test are unreliable compared with TC
tests. Additionally, the cohesion measured by the IDS test was always larger than that associated with the TC test, which is
caused by the extension of the shear surface in the shearing process according to the failure pattern. The overestimation of
shear strength in the case of the IDS test decreased the predicted surface settlement, as proved by numerical simulations.
The conclusions indicate that using TC tests to obtain accurate shear strengths of grouted clay is crucial for predicting the
soil deformation and developing rational construction schemes.

Keywords Grouted clay · Interface direct shear test · Shear strength · Triaxial compression test

Notation wP Plastic limit

ρ Density
The following symbols are used in this paper:
ρd Dry density
Ad Shearing area within sample in interface direct
φ Internal friction angle
shear test
φc Internal friction angle in triaxial compression tests
C Cohesion
φd Internal friction angle in interface direct shear test
cc Cohesion in triaxial compression test
φg Internal friction angle of grouted clay
cd Cohesion in interface direct shear test
σc Uniaxial compressive strength
cg Cohesion of grouted clay
τd Peak shear stress obtained by direct shear test
e Void ratio
σd Normal stress in interface direct shear test
P Grouting pressure
σ1 Maximum principal stress in triaxial compression
P0 Normal load in interface direct shear test
test
Pd Peak shear load or shear load at specified horizontal
σ3 Confining pressure in triaxial compression tests
displacement in interface direct shear test
R2 Fitting coefficient
W:C Water–cement ratio
Introduction
w Moisture content
wL Liquid limit
Currently, a growing number of tunnels are being con-
structed in soft ground in China following the enforcement
* Xiao Zhang of the Belt and Road Initiative and the acceleration of urban-
[email protected] ization. Surface settlements and excessive deformations are
frequently observed during soft ground tunnel construction
1
Geotechnical and Structural Engineering Research Center, (Attewell and Farmer 1974; Cooper et al. 2002; Fang et al.
Shandong University, Jinan 250061, China

13
Vol.:(0123456789)
261 Page 2 of 16 Bulletin of Engineering Geology and the Environment (2022) 81: 261

2016), which is often attributed to the insufficient shear triaxial compression (TC) tests to measure the shear strength
strength of natural soils (Schnaid et al. 2001; Divall et al. of grouted soil (Schnaid et al. 2001; Markou and Atmatzidis
2018; Meng et al. 2020). Recently, grouting with cement, 2003; Dano et al. 2004; Ajorloo et al. 2012; Nouri Delavar
a geotechnical technique, has been extensively adopted to and Noorzad 2020). However, only homogenous grouted
improve the mechanical characteristics of soils (Dammyr samples prepared by mixing or permeation were included.
et al. 2017; Park and Oh 2018; Yuan and Han 2020). For Furthermore, similar studies on the mechanical properties
instance, both the London Underground in the UK (Farrell of grout-infilled fractured rocks subjected to triaxial com-
2015; Packer et al. 2018) and the Shanghai Metro Line in pression have been conducted (Zhou et al. 2020). However,
China (Zhang et al. 2018) have adopted grouting methods limited attention has been paid to the mechanical parameters
for this reason. of grouted clay under the TC test. Although both methods
The interest in reinforcement evaluation of grouted soils can obtain the shear strength parameters of grouted soil,
has grown during the past decades, especially as it pertains different loading conditions in the laboratory may lead to
to the improvement of shear soil strength. Many previous diverse estimation results, which are commonly observed
studies on the shear strength of grouted soils focused on in geological material mechanical tests.
sandy soils or completely decomposed granite soil (CDGS) To date, the effect of grouted reinforcement on clayey
(Dano et al. 2004; Saiyouri et al. 2011; Ajorloo et al. 2012; soils remains unclear. Moreover, the mechanisms underlying
Chieregato et al. 2014; Hossain and Yin 2014; Kakavand grouted clayed soils are not yet fully understood. With the
and Dabiri 2018; Todaro 2021). However, very little is cur- novel self-developed setup for clayey soil grouting reinforce-
rently understood about the shear strength of grouted clayey ment setup, this study compared the shear strength indices
soils. The most extensively used material for grouting is a measured by IDS and TC testing and analyzed the underly-
cement suspension. For sandy soils and CDGSs, cement ing mechanism of different testing methods in clayey soils.
slurry can diffuse via infiltration to fill the soil voids. The Based on the test results, a numerical simulation using the
grouted specimens of previous studies, however, were homo- shear strength indices measured by different methods was
geneous media and were prepared by mixing or permeation conducted to study the different grouting effects. Finally, the
grouting. Conversely, when the cement slurry is grouted into computational data were compared with the field data, and
a clayey soil, fracture grouting under high pressure cannot the fracture grouting reinforcement mechanism of clayey
be avoided because the cement particles cannot enter the soil soil was revealed.
voids (Murdoch 1993; Widmann 1996; Li et al. 2017). As a
result, the existence of slurry veins in grouted clay is inevi-
table, thus making the grouted clay different from cemented Methodology
sands or cemented sandy gravel (Ismail et al. 2002; Wang
and Leung 2008; Wang et al. 2019). The existence of slurry In order to prepare grouted specimens suitable for TC test
veins causes a weakness in the grouted clay, which is an and IDS test, two grouting devices are introduced in detail.
inhomogeneous medium. The impact of slurry veins on Additionally, the test material and test program are also pre-
the shear strength of clayey remains unclear. Additionally, sented in this section.
owing to the slurry veins, conclusions on grouted sandy soils
cannot be transferred to grouted clayey soils. Experimental setup
The shear strength of geological materials (e.g., soils and
rocks) is often represented by the Mohr–Coulomb (MC) the- Grouting device for TC test
ory. The determinations of cohesion and internal friction
angle are fundamental for estimating the mechanical proper- Limited by the lack of appropriate devices to test fractur-
ties using the MC strength criterion (Zhao 2000; Kahraman ing reinforcement measures, TC tests for grouted clay are
et al. 2006; Alejano and Carranza-Torres 2011; Gong et al. difficult to perform in the laboratory. A rock specimen can
2020). Typically, direct shear tests and triaxial compression be sampled from a given site, but sampling the grouted soil
tests are employed in the laboratory to measure c and φ. will damage the specimen. El Mohtar and Rugg (Suits et al.
The direct shear test is commonly suitable for homogeneous 2011) introduced a three-way split mold to prepare grouted
media, because the shear surface is defined by the opera- soil specimens in the laboratory for triaxial compression
tors. Because the slurry–soil interface constitutes a weak tests. However, it only targeted permeation grouting. Limited
surface of a grouted sample, many studies have measured by the device strength and sealing performance, the appa-
shear strength by conducting interface direct shear (IDS) ratus could not be used to obtain fracture grouted samples.
tests on artificially fabricated slurry–soil interfaces (Hossain To address these problems, a novel grouting reinforce-
and Yin 2011; Zhang et al. 2015; Moayed et al. 2019; Lee ment testing system was developed as shown in Fig. 1a.
et al. 2020). Additionally, many studies have used traditional The system comprises a grouting reinforcement module,

13
Bulletin of Engineering Geology and the Environment (2022) 81: 261 Page 3 of 16 261

Fig. 1  Grouting device for

triaxial compression (TC) test: a
grouting reinforcement device;
b schematic of reinforcement
module and clay fracturing

air compressor, slurry barrels, dynamic mixer, and digital The pressure-supply module was an air compressor that
regulator. provided a constant grouting pressure. The slurry barrel had
The grouting reinforcement module comprised a sealed a height of 280 mm, an inner diameter of 151 mm, and a
loading box and a grouting pipe. The loading box was a wall thickness of 5.5 mm. The grouting amount during test-
cylinder made of three pieces of steel and sealed by lids ing was controlled by a digital regulator by setting the end
and threaded bars. The three steel sheets were assembled time and interval of the output air pressure. The dynamic
tightly with bulges and grooves on the edge. They were mixer was placed atop the grouting pipe to mix two types of
thick in the middle and thin at both ends with a given slurry when double liquid grouting was employed. Grouting
radius to ensure the compactness of the sealed ring. The reinforcement testing was performed as follows:
dimensions of the loading box had a height of 125 mm,
and an inner diameter of 61.8 mm, which conform to the 1. Clay samples were prepared with a certain moisture
TC test-size requirement (American Society for Testing content using the water-spraying method. The clay
and Materials STM D4767-11 2020). After launching the was filled into the loading box using the layered filling
testing system, as shown in Fig. 1b, the slurries entered the method, and all layers had heights equal to 5 cm;
loading box driven by air pressure. The clay fractured and 2. The testing system was assembled according to the
formed grouting veins when the pressure was sufficiently working principle, and the connectivity and sealing via
high. Several holes with diameters equal to 5 mm were water injection was checked;
drilled into the grouting pipe to reflect the actual grouting 3. After the placement of the slurry with given water-to-
situation. cement ratio (W:C) values into the barrel, the testing sys-

13
261 Page 4 of 16 Bulletin of Engineering Geology and the Environment (2022) 81: 261

tem was started. The grouting amount was regarded as Table 1  Properties of site clay
the end standard, and the grouting amounts of all samples Moisture Density Liquid limit Plastic limit Void ratio
were equal to 200 ml in all the preliminary tests; content ρ (g/cm3) wL (%) wP (%) e
4. The loading box was opened, and the grouting rein- w (%)
forcement was removed 5 h after the end of the test.
18.6 1.91 32.9 18.4 0.81
Subsequently, the consolidated undrained triaxial shear
test was performed on the reinforcements after curing at
20 ± 2 °C for 7 days. 1. The remolded clay was filled into the loading box using
the layered filling method;
Grouting device for IDS test 2. After grouting, the entire grouted clay was pushed out
of the loading box by the lifting jack. The samples were
The grouted clay specimen for the IDS test requires an obvi- obtained at the slurry–clay interface using a ring knife
ous soil–slurry interface as the defined shearing surface, for the IDS test.
and the grouting device should meet the requirements of
grouting parameters to simulate the in situ grouting. Based Test material
on previous research (Hossain and Yin 2011; Zhang et al.
2015), the grouting device and principle of preparing IDS The clay was sampled from the Yuhan Tunnel site in Jinan,
test specimens are shown in Fig. 2. With the necessary pres- Shandong Province. All physical and mechanical tests
sure-supply device and auxiliary equipment, Fig. 2a shows adhered to the procedures described in ASTM standards
the reinforcement module of the grouting system. The rein- (ASTM D6528-17 2017; ASTM D4767-11 2020). The basic
forcement module comprised an upper loading box and a properties of the site clay, including density, natural mois-
lower force-transfer device. The filling height in the loading ture content, liquid limit, plastic limit, and void ratio, are
box was adjusted by a lifting jack to change the sample den- listed in Table 1.
sity. As shown in Fig. 2b, a weak vertical interface was set To prepare the test samples, the clays sampled from the
in filled clay. Consequently, the slurries diffused direction- site were remolded in lab. For comparison with the grouted
ally along the prefabricated interface and formed regular clay, the cohesion and internal friction angle of clay without
grouting veins. The thickness of the directional grouting grouting are listed in Table 2.
veins ranged from 10 to 25 mm, and the sampling interface Locally available Portland cement was used to prepare the
was located precisely at the center of the specimen. Further- grout. The cement was mixed with water at water–cement
more, the position of the specimen in the shearing box was ratios of 0.8, 1, and 1.2. The slurry density and uniaxial
adjusted by shims to ensure that the shearing surface was compressive strength (UCS) of pure cement masses are
located at the slurry–soil interface. Grouting reinforcement listed in Table 3. The 3 d, 7 d and 14 d means the different
testing was performed as follows: curing time.

Fig. 2  Grouting device for interface direct shear (IDS) test: a structure of reinforcement module of grouting device; b diagram of making slurry–
clay interface sample

13
Bulletin of Engineering Geology and the Environment (2022) 81: 261 Page 5 of 16 261

Table 2  Cohesion and internal friction angle of remolded clay was the same as that of the TC test, and each testing condi-
Dry density Moisture Cohesion c (kPa) Internal friction tion was sampled and measured three times.
ρd (g/cm3) content angle φ (°)
w (%) Procedures for determining shear strength
1.4 9 60.24 28.2
parameters
15 20.94 25.4
For the IDS tests, a normal load, P0, was imposed on the
21 10.89 22.1
sample before the shear load, Pd, was applied. The shear
1.6 9 68.76 30.4
stress, τd, and normal stress, σd, acting on the shearing area,
15 34.20 27.9
Ad, are described by Eq. (1).
21 21.00 26.1
1.8 9 74.00 34.0 Pd P
15 49.3 32.0 𝜏d = , 𝜎d = 0 (1)
Ad Ad
21 38.23 30.2
The cohesion and internal friction angle can be obtained
by linear fitting according to the MC criterion, as expressed
by Eq. (2),
Test program
𝜏d =cd +𝜎d tan 𝜑d (2)
Given the variety of in situ geological and grouting condi-
In the cases of the TC tests, the cohesion and internal
tions and considering the test period, a group of orthogo-
friction angle can be obtained by the relationship between
nal tests was designed to conduct the grouting reinforce-
the maximum principal stress, σ1, and confining pressure,
ment. Four parameters were selected as the orthogonal
σ3, as expressed by Eq. (3),
testing factors to evaluate the strengthening effectiveness,
including dry density ρ d, moisture content w, grouting cos 𝜑c 1 + sin 𝜑c
pressure P, and W:C. Each factor set three levels, and 𝜎1 = 2cc + 𝜎 (3)
1 − sin 𝜑c 1 − sin 𝜑c 3
the orthogonal table L9 (34) was selected (Haofeng et al.
2018; Jinpeng et al. 2018; Zhang et al. 2019), as shown Equation (3) is also transformed by the MC criterion,
in Table 4. and two coefficients, A and B, are employed to more clearly
The consolidated undrained TC test was performed illustrate the formula, as expressed by Eq. (4). Accordingly,
using a TSZ triaxial compression apparatus. Isotropic cc and φc can be calculated using Eq. (5):
consolidation was adopted in the consolidation process,
cos 𝜑c 1 + sin 𝜑c
and three levels of confining pressure were applied during A= 2cc ,B= (4)
the compression process, namely, 100, 200, and 300 kPa. 1 − sin 𝜑c 1 − sin 𝜑c
Three specimens were measured at each confining pres-
sure. The corresponding deviator stress (σ1–σ3) was con- A
cc = √ ,𝜑c = arcsin
B−1
sidered as the ultimate strength when the axial strain was B +1 (5)
2 B
15%.
According to this IDS testing design, a direct shear test
was conducted on the slurry–soil interface of the samples.
Three levels of normal loading, 100, 200, and 300 kPa, were Table 4  Orthogonal test arrangement
imposed on the sample during the IDS test. The test process
Testing ID Factor
ρd (g/cm3) w (%) P (MPa) W:C

Table 3  Properties of cement grout D1.4-M9-P1.5-WC0.8 1.4 9 1.5 0.8:1

D1.4-M15-P2-WC1 1.4 15 2.0 1.0:1
Water–cement Density Uniaxial compression strength
D1.4-M21-P2.5-WC1.2 1.4 21 2.5 1.2:1
ratio ρ (g/cm3) (UCS) of pure cement masses
W:C σc (MPa) D1.6-M15-P1.5-WC1.2 1.6 15 1.5 1.2:1
D1.6-M21-P2-WC0.8 1.6 21 2.0 0.8:1
3d 7d 14 d
D1.6-M9-P2.5-WC1 1.6 9 2.5 1.0:1
0.8:1 1.62 3.25 3.82 5.58 D1.8-M21-P1.5-WC1 1.8 21 1.5 1.0:1
1:1 1.50 2.56 3.48 5.10 D1.8-M9-P2-WC1.2 1.8 9 2.0 1.2:1
1.2:1 1.44 2.01 2.85 4.37 D1.8-M15-P2.5-WC0.8 1.8 15 2.5 0.8:1

13
261 Page 6 of 16 Bulletin of Engineering Geology and the Environment (2022) 81: 261

Results or the fitting coefficient (R2). A greater R2 value indicates

smaller data deviation. All test identifiers (IDs) in the TC
The test results under two methods are reported in this sec- test case had a better fitting coefficient (> 0.9) compared
tion. By analyzing the discreteness and MC criterion fitting with those in the IDS test case. The IDS results were scat-
degree of test data, we discuss the difference between two tered in certain test conditions (R2 = 0.658), suggesting that
test methods. the IDS test was less appropriate than the TC test, as shown
in Fig. 4c. The data were the average value of each testing
Shear strength condition.
The coefficient of variation (CV, %) was used to compare
In this section, four conditions in orthogonal tests are the discreteness of the results in the cases if the two differ-
selected to conduct the comparison of results. According to ent methods, as listed in Table 5. CV is equal to variance
Eqs. (2) and (3), linear fitting was performed for the relation- divided by mean. The larger the CV is, the larger the data
ship between the maximum principal stress and confining discreteness is. It can be observed that the CV of TC was
stress (from the TC test) and that between the shear and nor- smaller than obtained in the IDS test case. Conversely, the
mal stresses (from the IDS test). The fitting results are also CV decreased with an increase in the confining stress or
plotted in Figs. 3 and 4. The data deviation in different test normal stress, and the changing range of the TC results was
conditions can be reflected objectively by the goodness-of-fit smaller than the IDS results.

Fig. 3  Fitting lines for the TC test results

13
Bulletin of Engineering Geology and the Environment (2022) 81: 261 Page 7 of 16 261

Fig. 4  Fitting lines for the IDS test results

Cohesion and internal friction angle the repeatability and accuracy, each testing condition was
measured three additional times (six times in total). Con-
In this section, the other five conditions in orthogonal tests sidering the D1.8-M21-P1.5-WC1 as an example, the results
were selected to calculate the values of c and φ. Considering are shown in Fig. 5. It can be observed that the cohesion
obtained by the IDS test was larger than that obtained by the
Table 5  Coefficient of variation of test results
TC test. Another obvious characteristic of the value distribu-
tion is that the cohesion obtained from the TC test exhibited
Testing ID Test condition Confining stress/Normal stress a minor variation at different test numbers, whereas those
100 kPa 200 kPa 300 kPa obtained from the IDS test changed over a wider range. The
large fluctuation of cohesion associated with the IDS test
D1.4-M15- IDS 16 14.4 7.8
P2-WC1
implies that the testing method has a greater impact on the
TC 9.1 5.1 2.4
shear strength parameters of the grouted clay. However, the
D1.6-M15- IDS 20.2 13 8
P1.5-WC1.2 values of the internal friction angle in the cases of the two
TC 0.4 5.7 4.2
methods were both unstable.
D1.6-M9- IDS 17.3 11.9 5.6
P2.5-WC1 The shear strength parameters (the value of average and
TC 7.2 4 3.9
error) of all test IDs are plotted in Fig. 6. The fluctuations
D1.8-M9- IDS 12 7.8 3.1
P2-WC1.2 of results can be reflected by the sizes of the error bar. As
TC 10.8 8.4 6.2
shown, the fluctuation of cohesion obtained by the TC test

13
261 Page 8 of 16 Bulletin of Engineering Geology and the Environment (2022) 81: 261

Fig. 5  Results of shear parameters and comparison of the two measurement methods at different testing numbers: a cohesion and b internal fric-
tion angle

is smaller than that obtained by IDS test at all test IDs. In large differences among different testing IDs, owing to the
the case of the internal friction angle, the results exhibit differences in cohesion with the original clay sample. How-
similar fluctuations in both methods. It can be inferred that ever, the change in the internal friction angle is uncertain,
the internal friction angle of grouted clay is uncertain, even and the internal friction angle of the grouted clay seems
at the same test condition. A group of bar diagrams is used to to have no quantitative relationship with the clay before
present the percent change in cohesion and internal friction grouting.
angle computed respect values reported in Table 2, as shown Tables 6 and 7 summarize the values of c and φ
in Fig. 7. These percentages were calculated by the average obtained from two different testing methods. The devia-
values of cohesion and friction angel obtained from test. toric stress and peak shear stress are the average values,
There was an apparent increase in cohesion in two differ- and c and φ are calculated according to Procedures
ent measurement methods. The percentage increases exhibit for determining shear strength parameters. The change

Fig. 6  Results and error bars of shear parameters between the two measurement methods at different test ID: a cohesion and b internal friction
angle

13
Bulletin of Engineering Geology and the Environment (2022) 81: 261 Page 9 of 16 261

Fig. 7  Percent changes of shear parameters between two measure methods at different test IDs: a cohesion and b internal friction angle

Table 6  Shearing strength indices of grouted clay subjected to triaxial compression (TC) tests
Testing ID Deviatoric stress (σ1–σ3) (kPa) Cohesion Internal friction angle
cc (kPa) φc (°)
σ3 = 100 kPa σ3 = 200 kPa σ3 = 300 kPa

D1.4-M9-P1.5-WC0.8 472 652 845 83.5 (↑39%) 30.2 (↑7%)

D1.4-M15-P2-WC1 565 774 988 99.9 (↑375%) 32.7 (↑29%)
D1.4-M21-P2.5-WC1.2 497 628 797 107.7 (↑888%) 26.3 (↑20%)
D1.6-M15-P1.5-WC1.2 574 820 992 105.3 (↑215%) 32.4 (↑16%)
D1.6-M21-P2-WC0.8 560 765 952 104.6 (↑398%) 31.2 (↑20%)
D1.6-M9-P2.5-WC1 465 614 820 81.4 (↑21%) 29.3 (↓4%)
D1.8-M21-P1.5-WC1 545 743 994 86.5 (↑126%) 33.9 (↑12%)
D1.8-M9-P2-WC1.2 525 709 903 98.3 (↑33%) 30.5 (↓10%)
D1.8-M15-P2.5-WC0.8 521 740 896 101.4 (↑106%) 30.3 (↓5%)

Table 7  Shearing index of grouted clay subjected to interface direct shear (IDS) tests
Testing ID Peak shear stress τd (kPa) Cohesion Internal friction angle
cd (kPa) φd (°)
σd = 100 kPa σd = 200 kPa σd = 300 kPa

D1.4-M9-P1.5-WC0.8 184 204 264 136.6 (↑127%) 22.3 (↓21%)

D1.4-M15-P2-WC1 201 282 331 141.3 (↑573%) 33 (↑30%)
D1.4-M21-P2.5-WC1.2 144 217 275 81 (↑643%) 33.2 (↑50%)
D1.6-M15-P1.5-WC1.2 206 266 357 125.3 (↑266%) 37.1 (↑33%)
D1.6-M21-P2-WC0.8 201 257 336 128 (↑504%) 33.9 (↑30%)
D1.6-M9-P2.5-WC1 228 298 332 182 (↑165%) 27.5 (↓10%)
D1.8-M21-P1.5-WC1 245 286 381 168 (↑340%) 34.7 (↑2%)
D1.8-M9-P2-WC1.2 256 331 437 160.3 (↑123%) 42.1 (↑24%)
D1.8-M15-P2.5-WC0.8 197 260 301 148.7 (↑202%) 27.5 (↓14%)

13
261 Page 10 of 16 Bulletin of Engineering Geology and the Environment (2022) 81: 261

Fig. 8  Mohr circles and envelopes of grouted clay with different identifiers by two shear test methods

in the amplitude percent of cohesion and internal friction multiple tests. Accordingly, it can be concluded that the
angle is also shown in the shearing index table (compared cohesion obtained by the IDS test may be too large.
with Table 2). It can be observed that grouting improves
significantly the cohesion of clay in the two testing meth-
ods. However, compared with the cohesion, the change Comments on results
in the internal friction angle is uncertain before and after
grouting. For the shear strength, significant discreteness of the
Moreover, the MC strength envelope can be usually results was found in the IDS test following the analysis
determined by the tangent of a series of Mohr circles. of the CV. Furthermore, the R 2 of TC test is also better
According to the above shearing indices table, Fig. 8 plots than that of IDS test, as shown in Figs. 3 and 4. Assess-
the Mohr envelopes for grouted clay obtained by the two ments of test result deviations by CV and R2 have different
testing methods and remolded clay. The Mohr circles also emphases. For a certain testing method, R2 reflects the dis-
include the unconfined compressive strength of the grouted creteness of all test data for a specimen type, whereas CV
clay obtained by the uniaxial compression test. As this fig- represents the data deviation of the specimen type tested
ure shows, the shear strength envelopes obtained by the TC at a fixed normal stress or confining stress. The excessive
tests are basically consistent with the Mohr circles, and the uncertainty associated with the IDS test seems to reflect
test results agree well with the MC strength criterion. How- the shortcomings of the test itself.
ever, the test results obtained from the IDS test are not well The Mohr circles and envelopes were plotted by the cal-
matched with the Mohr circles. For the MC strength enve- culated c and φ (Fig. 8). Compared with the remolded clay,
lope, the intercept is cohesion, and the slope indicates the the cohesion had an apparent increase. However, compared
internal friction angle. Apart from D1.4-M21-P2.5-WC1.2 with the cohesion, the change in the internal friction angle
(Fig. 8c), the cohesion obtained by the IDS test was larger was uncertain before and after grouting. In fact, some pre-
than that obtained by the TC test. This exception may be vious studies have shown that the grouting strengthened
related to error caused by uncertainty of grouted clay in only cohesion to the grouted soils and did not obviously

13
Bulletin of Engineering Geology and the Environment (2022) 81: 261 Page 11 of 16 261

alter the internal friction angle (Ata and Vipulanandan shearing area, as plotted in Fig. 10b, and makes the shear
1999; Nouri Delavar and Noorzad 2020). Therefore, it stress measured in the test higher than its actual value. The
can be inferred that the improvement process of grouting actual shearing surface in the IDS test is shown in Fig. 11.
on the shear strength of clay was mainly reflected in the As shown by the above analysis, the actual shearing surface
improvement of cohesion. is not a horizontal plane, and the difference between the soil
After grouting, the resulting material becomes a mix- and slurry seems to make this effect more obvious. As the
ture of denser clay and cement grout veins. In the process normal stress increases, the extension extent of shearing area
of fracture grouting, the clay is compacted by grouting decreases, that is, the measured value at high normal stress
veins formed by high pressure, and the water in the voids values are closer to the actual values. As a result, the cohesion
is squeezed out. This compaction behavior weakens con- is larger, and the internal friction angle is smaller according to
siderably the effects of clay properties on the strength of the MC criterion; this explains why the values of cohesion in
the grouted clay. At the same grouting volume per unit the IDS test are too large. Unfortunately, the obvious smaller
volume, the compaction degree of the injected clay was internal friction angle was not clearly observed in test results.
approximate. Therefore, the cohesions of grouted clay This situation only can be found in Fig. 9a, f, and i. However,
at different test conditions were similar, which can be the excessive shear stress caused by the extension of shear-
observed from the TC test results. ing surface will undoubtedly lead to higher cohesion. As for
Figure 8 also shows that the cohesion obtained by the internal friction angle in IDS tests, it may be more affected
IDS test is larger than that obtained by the TC test. The by the interface properties. However, those discussions are
analysis of the failure patterns of the grouted clay intended beyond the scope of this paper and should be further studied.
to explain this phenomenon. Figure 9 shows the typical In summary, the shear strength parameters of grouted
failure patterns of the grouted clay subjected to the TC clay provided by the IDS test were unreliable compared
test. A significant characteristic different from the original with the TC test. The excessive discreteness and the low MC
soil is that the grouted clay has an obvious shear failure criterion fitting degree of IDS data reflect this unreliabil-
surface under the TC test, such as rock and other hard ity remarkably. From the point-of-failure pattern, it seems
materials. Therefore, it can be inferred that the strength of that the extension of the horizontal shear surface caused by
grouted clay is mainly provided by the dense clay itself, the interface difference in IDS led to unsatisfactory results.
and the different vein distributions have minor influences. Therefore, it can be inferred that the occurrence of the pure
We cannot deny that the existence of slurry veins causes shear surface was a robust indication of the approach that
a weak surface that is reflected in the failure surface that yielded the most accurate or reliable shear strength param-
mainly appears in the direction of the slurry vein exten- eters of grouted clay.
sion, as shown in Fig. 9.
We inferred that the excessive error and deviation in the
IDS test is the existence of a secondary shear surface. When Numerical simulation on the surface
the shearing box moves horizontally, the sample is subjected grouting of a tunnel
to horizontal pressure. This forms two intersecting sliding sur-
faces between the upper and lower boxes. Thus, the secondary The effect of shear parameters of grouted clay on engineer-
shearing surface is shown in Fig. 10. Because of the exten- ing is analyzed in this section by numerical simulation. Sim-
sion of the secondary shearing surface, the shearing surface ulation results show that overoptimistic cohesion obtained
will no longer follow the horizontal line. Instead, it forms a from IDS test underestimate the engineering risk.

Fig. 9  Failure patterns of

grouted clay subjected to TC
tests

13
261 Page 12 of 16 Bulletin of Engineering Geology and the Environment (2022) 81: 261

Fig. 10  Diagram of secondary shearing surface and actual shearing area: a diagram of shearing surface; b actual shearing area

Project overview and calculation model Calculation parameters

To support further the conclusion that the TC test results The soil parameters were calculated using the MC model,
are more reasonable than the IDS test, a numerical simula- and the calculation parameters of clay and grouted clay in
tion of surface grouting was conducted based on the Yuhan this section were defined by testing the sample with the ID:
Road Tunnel project. During excavation, obvious vault sub- D1.6-M9-P2.5-WC1. The initial support and lining were
sidence, excessive ground settlement, and water seepage placed using C15 and C25 concrete, respectively. The con-
occurred in the YK2 + 692–712 section (An et al. 2021). crete differed from the soil, which was given an elastic model
Surface grouting was selected as the method to enhance clay in the calculation. The anchor and steel bar were considered
strength and ensure excavation safety. The tunnel stratum by equivalent conversion to improve the lining parameters.
and outline are shown in Fig. 12. The physical and mechanical parameters of the soils and
Using the YK2 + 692 section as an example, wherein the support in the calculation are summarized in Table 8.
disasters occurred as the modeling prototype, the spacing
between two tunnels was 4 m, and the thickness of soil cov- Analysis of simulation results
erage was 10 m, with a loess thickness of 3.5 m. Considering
the boundary effect, the size of whole calculation model was During the construction process, the left tunnel was exca-
80 m (x-direction), 38.5 m (z-direction), and 20-m (y direc- vated first, followed by the right tunnel. Hence, the calcula-
tion), as shown in Fig. 13. The tunnel was excavated by the tion process is the same as the construction process, and the
double-benching method. According to the testing results tunnels on both sides are excavated in turn. To prevent dam-
measured by TC and IDS tests, the calculation parameters age to the surface and excavation area caused by high-grout-
of grouted clay were defined. Finally, the surface settlement ing pressure, the grouting boundary was determined to be
simulation results were compared with the construction site. approximately 4.5 m below the loess layer. The calculation

Fig. 11  Failure patterns of

grouted clay subjected to IDS
tests

13
Bulletin of Engineering Geology and the Environment (2022) 81: 261 Page 13 of 16 261

Fig. 12  Yuhan Road tunnel

stratum and outline

process was implemented after a certain simplification. The settlement in the second tunnel area. Thus, an asymmetrical
surface settlements of the simulation under two different cal- pressure effect occurs, which can be attributed to the distur-
culation parameters were compared with field data. bance induced by the first tunnel construction.
The central axes of both tunnels are taken as the center- As shown, the simulation results which were obtained
lines, and the surface settlement values on both sides of each with the use of different parameters were close in the first
centerline are shown in Fig. 14. Two curves are obtained tunnel area. The gap gradually increased with the transfer to
from the simulation, and their calculation parameters are the centerline and the second tunnel area. However, the field
defined by the results of the TC and IDS tests. The other data were larger than the simulation results, and the maxi-
curve was obtained by field measurements. The field data mum settlement position was approximately at the center of
were measured every 3 m, which differed from the monitor- the second tunnel. Although the simulation results using the
ing point in the calculation model. The settlement trough two test parameters failed to estimate correctly the maximum
is symmetrically distributed along the centerline (Chakeri settlement position (both at the centerline), the simulation
et al. 2011). However, the shape of the settlement trough in results using TC test parameters reflected the asymmetrical
a closely spaced tunnel differs from that in a normal settle- characteristics of the settlement trough to a certain extent.
ment trough. An obvious difference is reflected in the larger Additionally, the calculated results which were obtained

Fig. 13  Diagram of tunnel calculation model

13
261 Page 14 of 16 Bulletin of Engineering Geology and the Environment (2022) 81: 261

Table 8  Physical and Material Density Elastic modulus Cohesion Friction Tensile strength
mechanical parameters of model (g/cm3) (MPa) (kPa) angle (°) (kPa)

Loess 1.71 11.9 14.9 22.7 –

Clay 1.75 13.5 68.76 30.4 –
Initial support 2.48 22 × ­103 – – 0.91 × ­106
Lining 2.51 30 × ­103 – – 2.67 × ­106

Fig. 14  Surface settlement of simulation results and field data

with the use of IDS test parameters were smaller than those Conclusions
obtained using TC test parameters. From a data perspective,
the calculated parameters obtained from the IDS test overes- It is of utmost importance to measure the shear strength
timated the reinforcement effect of grouted clay. Therefore, of grouted clay to evaluate the surface settlement or other
the shear strength parameters measured by the TC test were soil deformation. In this study, two types of shear tests were
more reasonable. conducted on grouted clay to estimate their shear strength

13
Bulletin of Engineering Geology and the Environment (2022) 81: 261 Page 15 of 16 261

parameters based on the two developed lab grouting devices. Chieregato A, Oñate Salazar CG, Todaro C et al (2014) Laboratory grouting
The results were comprehensively compared and analyzed test for waterproofing and consolidation of granular soils by means of
innovative materials. Geoing Ambient e Mineraria 141:63–68
from the perspective of testing data deviations and speci- Cooper ML, Chapman DN, Rogers CDF, Chan AHC (2002) Move-
men failure characteristics, and some key conclusions were ments in the Piccadilly Line tunnels due to the Heathrow Express
drawn. construction. Geotechnique 52:243–257. https://​doi.​org/​10.​1680/​
The testing deviation indicated that the shear strength geot.​2002.​52.4.​243
Dammyr Ø, Nilsen B, Gollegger J (2017) Feasibility of tunnel boring
parameters of the grouted clay obtained from the IDS test through weakness zones in deep Norwegian subsea tunnels. Tunn
were unreliable compared with the results obtained by the TC Undergr Sp Technol 69:133–146. https://​doi.​org/​10.​1016/j.​tust.​
test. During the IDS test process, the horizontal pure shear 2017.​06.​012
fracturing surface expanded to a curved surface because of the Dano C, Hicher P-Y, Tailliez S (2004) Engineering properties of
grouted sands. J Geotech Geoenvironmental Eng 130:328–338.
complex stress, which may have caused the excessive errors. https://​doi.​org/​10.​1061/​(asce)​1090-​0241(2004)​130:​3(328)
Using a surface grouting simulation, we found that the IDS Divall S, Goodey RJ, Stallebrass SE (2018) Twin-tunnelling-induced
testing data overestimated the reinforcement effect of grouted changes to clay stiffness. Tunn Urban Environ 37–44. https://​doi.​
clay compared with that induced by the TC test. org/​10.​1680/​tue.​63778.​037
Fang Q, Tai Q, Zhang D, Wong LNY (2016) Ground surface settle-
Currently, the IDS tests are also applied in the evaluations ments due to construction of closely-spaced twin tunnels with
of other mechanical properties, e.g., antipulling properties different geometric arrangements. Tunn Undergr Sp Technol
of the grouted soil nail. The conclusions inferred based on 51:144–151. https://​doi.​org/​10.​1016/j.​tust.​2015.​10.​031
this study indicate that the IDS tests usually overestimate Farrell RP (2015) Tunnelling and compensation grouting at Bond
Street, London. Proc Inst Civ Eng Geotech Eng 168:471–482.
the shear strength of the slurry–soil interface. Overoptimis- https://​doi.​org/​10.​1680/​jgeen.​14.​00162
tic evaluation indirectly increases the engineering risk that Gong F, Luo S, Lin G, Li X (2020) Evaluation of shear strength param-
requires attention by engineers. eters of rocks by preset angle shear, direct shear and triaxial com-
pression tests. Rock Mech Rock Eng 53:2505–2519. https://​doi.​
Acknowledgements The research reported in this manuscript was org/​10.​1007/​s00603-​020-​02050-1
funded by the National Natural Science Foundation of China (Grant Haofeng X, Feng X, Feng Z (2018) Improvement for the strength of salt-
No. U1706223) and by the Fundamental Research Funds for the Central rich soft soil reinforced by cement. Mar Georesources Geotechnol
Universities (Grant No. 2019GN079). 36:38–42. https://​doi.​org/​10.​1080/​10641​19X.​2016.​12780​64
Hossain MA, Yin J-H (2014) Behavior of a pressure-grouted soil–
cement interface in direct shear tests. Int J Geomech 14:101–109.
https://​doi.​org/​10.​1061/​(asce)​gm.​1943-​5622.​00003​01
Hossain MA, Yin JH (2011) Influence of grouting pressure on the
behavior of an unsaturated soil-cement interface. J Geotech
References Geoenvironmental Eng 138:193–202. https://​doi.​org/​10.​1061/​
(ASCE)​GT.​1943-​5606.​00005​85
Ajorloo AM, Mroueh H, Lancelot L (2012) Experimental investigation Ismail MA, Joer HA, Sim WH, Randolph MF (2002) Effect of cement
of cement treated sand behavior under triaxial test. Geotech Geol type on shear behavior of cemented calcareous soil. J Geotech
Eng 30:129–143. https://​doi.​org/​10.​1007/​s10706-​011-​9455-4 Geoenvironmental Eng 128:520–529. https://​doi.​org/​10.​1061/​
Alejano LR, Carranza-Torres C (2011) An empirical approach for estimat- (asce)​1090-​0241(2002)​128:​6(520)
ing shear strength of decomposed granites in Galicia, Spain. Eng Jinpeng Z, Limin L, Futao Z, Junzhi C (2018) Development and appli-
Geol 120:91–102. https://​doi.​org/​10.​1016/j.​enggeo.​2011.​04.​003 cation of new composite grouting material for sealing groundwa-
An Q, Zhang Q, Zhang X, Zhang J (2021) Bridging the gap between ter inflow and reinforcing wall rock in deep mine. Sci Rep 8:1–11.
engineering properties and grouting reinforcement mechanisms https://​doi.​org/​10.​1038/​s41598-​018-​23995-y
for loess in eastern China: taking Jinan loess as an example. Kahraman S, Altun H, Tezekici BS, Fener M (2006) Sawability pre-
Bull Eng Geol Environ 80:4125–4141. https://​doi.​org/​10.​1007/​ diction of carbonate rocks from shear strength parameters using
s10064-​021-​02201-w artificial neural networks. Int J Rock Mech Min Sci 43:157–164.
ASTM D4767–11 (2020) Standard Test Method for Consolidated https://​doi.​org/​10.​1016/j.​ijrmms.​2005.​04.​007
Undrained Triaxial Compression Test for Cohesive Soils. ASTM Kakavand A, Dabiri R (2018) Experimental study of applying colloi-
International, West Conshohocken, PA dal nano Silica in improving sand-silt mixtures. Int J Nano Dimens
ASTM D6528–17 (2017) Standard Test Method for Consolidated Und- 9:357–373
rained Direct Simple Shear Testing of Fine Grain Soils. ASTM Lee JS, Han WJ, Kim SY, Byun YH (2020) Shear strength and interface
International, West Conshohocken, PA friction characteristics of expandable foam grout. Constr Build Mater
Ata A, Vipulanandan C (1999) Factors affecting mechnical and creep 249:118719. https://​doi.​org/​10.​1016/j.​conbu​ildmat.​2020.​118719
properties of silicate-grouted sands. J Geotech Geoenviron Eng Li P, Zhang Q, Zhang X et al (2017) Grouting diffusion characteristics in
125:868–876. https://​doi.​org/​10.​1061/​(ASCE)​1090-​0241(1999)​ faults considering the interaction of multiple grouting. Int J Geomech
125:​10(868) 17:1–10. https://​doi.​org/​10.​1061/​(ASCE)​GM.​1943-​5622.​00008​15
Attewell PB, Farmer IW (1974) Ground deformations resulting from Markou IN, Atmatzidis DK (2003) Mechanical behavior of a pulverized
shield tunnelling in London clay. Can Geotech J 11:380–395. fly ash grouted sand. Geotech Test J 26:450–460. https://​doi.​org/​
https://​doi.​org/​10.​1139/​t74-​039 10.​1520/​gtj11​252j
Chakeri H, Hasanpour R, Hindistan MA, Ünver B (2011) Analysis Meng F, Chen R, Wu H, Cheng H (2020) Evaluating closely spaced
of interaction between tunnels in soft ground by 3D numerical twin-tunnelling-induced shear stiffness change by subsurface set-
modeling. Bull Eng Geol Environ 70:439–448. https://​doi.​org/​ tlement in clayey ground. Environ Earth Sci 13:326. https://​doi.​
10.​1007/​s10064-​010-​0333-8 org/​10.​1007/​s12665-​020-​09087-z

13
261 Page 16 of 16 Bulletin of Engineering Geology and the Environment (2022) 81: 261

Moayed RZ, Hosseinali M, Shirkhorshidi SM, Sheibani J (2019) Wang YH, Leung SC (2008) Characterization of cemented sand by
Experimental investigation and constitutive modeling of grout– experimental and numerical investigations. J Geotech Geoenviron-
sand interface. Int J Geomech 19:04019024. https://​doi.​org/​10.​ mental Eng 134:992–1004. https://​doi.​org/​10.​1061/​(asce)​1090-​
1061/​(asce)​gm.​1943-​5622.​00013​84 0241(2008)​134:​7(992)
Murdoch LC (1993) Hydraulic fracturing of soil during laboratory Wang W, Pan J, Jin F (2019) Mechanical behavior of cemented gran-
experiments Part 1. Methods and Observations Geotechnique ular aggregates under uniaxial compression. J Mater Civ Eng
43:255–265. https://​doi.​org/​10.​1680/​geot.​1993.​43.2.​255 31:04019047. https://​doi.​org/​10.​1061/​(asce)​mt.​1943-​5533.​00026​81
Nouri Delavar I, Noorzad R (2020) Drained shear strength parameters Yuan S, Han G (2020) Combined drilling methods to install grout
of silty sand grouted by colloidal silica. Int J Geotech Eng 14:1–8. curtains in a deep underground mine: A case study in Southwest
https://​doi.​org/​10.​1080/​19386​362.​2017.​13803​69 China. Mine Water Environ 39:902–909. https://​doi.​org/​10.​1007/​
Packer M, Prangley C, Newman R, Heath I (2018) Permeation grout- s10230-​020-​00701-x
ing and excavation at Victoria station, London. Proc Inst Civ Eng Zhang DM, Huang ZK, Wang RL et al (2018) Grouting-based treat-
Geotech Eng 171:267–281. https://​doi.​org/​10.​1680/​jgeen.​17.​00115 ment of tunnel settlement: Practice in Shanghai. Tunn Undergr Sp
Park DS, Oh J (2018) Permeation grouting for remediation of dam Technol 80:181–196. https://​doi.​org/​10.​1016/j.​tust.​2018.​06.​017
cores. Eng Geol 233:63–75. https://​doi.​org/​10.​1016/j.​enggeo.​ Zhang JP, Liu LM, Li QH et al (2019) Development of cement-based
2017.​12.​011 self-stress composite grouting material for reinforcing rock mass
Widmann R (1996) International Commission Society for Rock on and engineering application. Constr Build Mater 201:314–327.
Rock Grouting. Int J Rock Mech Min Sci Geomech Abstr 33:803– https://​doi.​org/​10.​1016/j.​conbu​ildmat.​2018.​12.​143
847. https://​doi.​org/​10.​1016/​S0148-​9062(96)​00015-0 Zhang Q, Li P, Zhang X et al (2015) Model test of grouting strength-
Saiyouri N, Alaiwa AA, Hicher P-Y (2011) Permeability and porosity ening mechanism for fault gouge of tunnel. Yanshilixue Yu
improvement of grouted sand. Eur J Environ Civ Eng 15:77–97. Gongcheng Xuebao/Chinese J Rock Mech Eng 34:924–934.
https://​doi.​org/​10.​1080/​19648​189.​2011.​96933​07 https://​doi.​org/​10.​13722/j.​cnki.​jrme.​2014.​0667
Schnaid F, Prietto PDM, Consoli NC (2001) Characterization of Zhao J (2000) Applicability of Mohr-Coulomb and Hoek-Brown
cemented sand in triaxial compression. J Geotech Geoenviron- strength criteria to the dynamic strength of brittle rock. Int J Rock
mental Eng 127:857–868. https://​doi.​org/​10.​1061/​(ASCE)​1090-​ Mech Min Sci 37:1115–1121. https://​doi.​org/​10.​1016/​S1365-​
0241(2001)​127:​10(857) 1609(00)​00049-6
Suits LD, Sheahan TC, El Mohtar CS, Rugg DA (2011) New three-way Zhou D, Zhao Z, Li B et al (2020) Permeability evolution of grout
split mold design and experimental procedure for testing soft, grouted infilled fractures subjected to triaxial compression with low con-
soils. Geotech Test J 34:103645. https://​doi.​org/​10.​1520/​GTJ10​3645 fining pressure. Tunn Undergr Sp Technol 104.https://​doi.​org/​10.​
Todaro C (2021) Grouting of cohesionless soils by means of colloidal 1016/j.​tust.​2020.​103539
nanosilica. Case Stud Constr Mater 15:e00577

13