Evaluating the Shear Strength of Subbase-subgrade Interface Using Large Scale Direct Shear Test

The inclusion of geogrid in road pavements can improve pavement performance through increasing the lateral confinement, bearing capacity, and overall rigidity of the pavement, as well as reducing the vertical and lateral pavement deformations. The materials used in the present study are: subbase granular materials Type B, two types of subgrade soil; clay and sandy soil, and two nonwoven biaxial geogrids (G1 and G2) used as reinforcing materials. Direct shear testing was adopted by manufacturing a large-scale direct shear apparatus consisted of an upper, square box of size 20 cm × 20 cm × 10 cm, and a lower, rectangular box of size 200 mm × 250 mm × 100 mm is used in the present study. The results show that, for the four normal stresses equal to 25, 50, 75 and 100 kPa, the interface shear stress curves increased and followed similar trend. For clay-subbase interface, installation of geogrid decreases the apparent cohesion of the material from 16.5 kPa (without reinforcement) to be 8 kPa and 13.5 kPa for G1 and G2, respectively. At sand-subbase interface, using geogrid leads to increase the cohesion of the material from 3.5 kPa to be 15.5 and 16 kPa for G1 and G2, respectively. The friction angle increases slightly from 30o (without reinforcement) to be 35o for G1 and G2 when the interface is subbase over clay. While, it decreased from 35.8o to be 32.1o for G1 and G2 at sand-subbase interface. The interaction coefficient for G1 and G2 increased when the normal strength increased at the clay-subbase interface. Otherwise, the behavior of interaction coefficient of the sand-subbase interface appears deferent trend, where increasing normal stress leads to decrease the interaction.


Introduction
Demand on transportation including its miscellaneous modes increases rapidly with the increase of global population (Hatem et al, 2018;Mosa et al, 2021a;Mosa et al, 2022a) to ensure smooth movement of persons and goods among different places in different countries (Banyhussan et al, 2020;Mohammed et al, 2019a;Mohammed et al, 2019b;. Among these transport modes, highways are the most important due their flexibility and efficiency for different users (Mohammed et al, 2018a;Mohammed et al, 2018b;Mosa, 2015;Mosa et al, 2011a). Therefore, providing efficient network of highways is an essential concept (Mosa, 2017a;Mosa et al, 2011b;Mosa et al, 2013a;Mosa et al, 2013b;Mosa et al, 2013c). Pavement can be considered as the major component in highways (Mosa et al, 2018;Mosa, 2017b). Efficient pavements can provide safe and comfortable riding for the drivers. To ensure that, several considerations must be taken. Tightness of underlying layers (subgrade and subbase) is a major factor that ensures the stability of the pavements system (Mosa, 2017c;Mosa, 2017d;Mosa et al, 2022b;. Therefore, investigation of the properties of these layers can improve the efficiency of the pavements system (Mosa, 2017e;Mosa et al, 2021b;Mosa et al, 2017;Salem et al, 2018). Several techniques are adopted for reinforcement of the underlying layers of pavements (Al-Dahlaki & Mosa, 2016). Geogrid reinforcements are used in pavement engineering in two major applications: (i) base reinforcement (embedded within the base or subbase layers), and (ii) subgrade stabilization (embedded between the subgrade and the granular base course). The inclusion of geogrid in road pavements can improve pavement performance through increasing the lateral confinement, bearing capacity, and overall rigidity of the pavement, as well as reducing the vertical and lateral pavement deformations (Liu et al, 2009;Shukla, 2017). Moreover, using geogrid can also prevent reflective cracking and reduce crack propagation, and anti-reflective cracking systems may contain multiple layers of geosynthetic materials in the interlayer zone. From an economical point of view, it can also: (i) decrease the required thickness of base course, (ii) reduce road maintenance costs, and (iii) increase road pavement life. Also the settlement result for soil reinforced by geonet is lower than that for unreinforced soil because the geonet layer strengthen the soil (Nareeman & Fattah, 2012). Pavements require a strong and stable foundation to be built upon, which is not possible in every environment. Some locations necessitating road construction have very poor quality subgrade, and therefore require improvement prior to pavement construction. Failure to improve the subbase to a strong and stable condition results in pavement rutting and permanent deformation (Giroud & Han, 2004). The soil-reinforcement interface properties depend on many factors such as the interaction mechanism between soil and reinforcement (direct shear mode or pullout mode), the physical and mechanical properties of soil (density, particle shape and size, particle-size distribution, and moisture content), and properties of the reinforcement (shape, geometry, tensile stiffness, and strength). Triaxial and direct shear tests are conventionally used to determine the shear strength parameters, however, direct shear apparatus is commonly adopted to determine the interface shear properties of reinforcement and fill materials. Some researchers have used direct shear apparatus and obtained interface properties using a device in which a rigid block was kept as substratum particularly when geotextile was placed at the interface (Choudhary & Krishna, 2014;Kandolkar & Mandal, 2013) whereas other researchers have tested interface properties using a device in which both parts of the shear box were filled with soil and the geogrid was placed at the horizontal shear plane to simulate the field conditions (Arulrajah et al, 2015). Table (1) gives the details of the studies conducted by different researchers using large-scale direct shear device, where the large-size direct shear apparatus was used keeping both the halves of a shear box filled with pavement material, and interface tests were performed with reinforcement placed at the middle coinciding with the shearing plane. Geogrids have been widely used to construct stable subgrade foundations and to provide a working platform for construction over weak and soft soils. Use of geogrid reinforcement in a pavement system ensures a longlasting pavement structure by reducing excessive deformation and cracking. In this study a series of largescale direct shear tests were performed to evaluate the mechanical interaction between subgrade soil and an aggregate sub base layer with and without a geogrid at the interface.

Materials Used Subbase Material
The study used subbase granular materials (SGM) Type B. This type of SGM has been obtained from Sabeaa Al-bour area north of Baghdad city in Iraq. Table (2) illustrates the gradation of subbase granular materials. The physical properties and results of chemical analysis of SGM are listed in Table (3) complied with (SCRB/R6, 2003).

Clay Soil
The clay soil is taken from Al-Muthanaa Airport location in Baghdad city.

Sandy Soil for Subgrade Layer
The sandy soil is taken from Shatt Al-Taji location north of Baghdad city. Table (5) shows the soil properties for sandy soil.

Geogrid
Two nonwoven biaxial geogrids were used as reinforcing materials. Table (6) provides the properties of reinforcing materials. Two geogrid reinforcement types with similar stiffness, but with different aperture sizes, were used in the study.

Direct Shear Testing
The manufactured large-scale direct shear apparatus consisted of an upper, square box of size 200 mm × 200 mm × 100 mm, and a lower, rectangular box of size 200 mm × 250 mm × 100 mm is used in the present study. The size of the lower box was kept larger than that of the upper box in order to maintain a constant shearing area during the tests. The test materials were prepared at their optimum moisture contents.

Interface Testing Program
A total of 24 interface tests were performed on specimens with and without reinforcement.  (2014)] test procedure on interface test stipulates the box size to be at least five times the opening size of the reinforcement. In this study, this ratio was equal to 20 for G1 and 13 for G2 geogrid reinforcements.

Test Setup and Procedure
In this study, manufactured large-size direct shear apparatus of box size equal to 200 mm × 200 mm × 200 mm in length, width, and height was used to measure the interface properties with geogrid and without geogrid. The test apparatus consisted of horizontal load cells of capacity equal to 50 kN used to perform the tests (Figure 1), and two linear variable differential transducers (LVDTs) with a range of ± 50 mm. LVDTs were used to measure the horizontal and vertical deformations of the sample during the shearing process. The measurements were automated through a Data Acquisition System (DAQ). During interface testing, the reinforcement was placed at the interface between the lower and upper.

Results and Discussion
Interface shear testing of two geogrids of different aperture sizes were performed under four normal stresses equal to 25, 50, 75 and 100 kPa to determine their interface shear strength. A total number of 24 experiments were conducted on reinforced interfaces alone. Figures (2a and 2b) show the photographs of two types of reinforcements fixed to the shear box at the shearing plane.
a. b.

Shear Strength
For clay-subbase interface, Figures (3a-3c) show the variation of interface shear stress with a horizontal displacement corresponding to the four normal stresses equal to 25, 50,75 and 100 kPa, with and without reinforcement interfaces, respectively. The interface shear stress curves followed similar trend for the four cases under consideration. The figures show that increasing the normal strength leads to increase shear stresses for all cases. Furthermore, Figures (4a-4c) illustrate the shear stress for sand-subbase interface with different normal stress. Figures (5a-5c) show the Mohr-Coulomb shear strength envelopes at peak for the four cases under consideration. The Interface friction angle and adhesion intercept for clay-subbase interface were found to be 30° and 16 kPa, respectively. It can be noticed that installation of geogrid decreases the apparent cohesion of the material from 8 and 13.5 kPa for G1 and G2, respectively; however, it increases the friction angle slightly to be 35 o for G1 and G2. Table 2 shows shear properties of different interfaces. On the other hand, for sand-subbase interface, the angle of friction and adhesion found to be 35.75° and 3.5 kPa without reinforcing. By using geogrid, it leads to increase the apparent cohesion to be 15.5 and 16 kPa for G1 and G2, respectively; however, it decrease the angle of friction slightly to be 32.1o for G1 and G2 as shown in Figures (6a-6c). Tables (7) and (8)

Interaction Coefficient of Reinforcement
The interaction coefficient of reinforcement with soil is defined as the ratio of the shear strength at the soilreinforcement interface to the shear strength of the soil at the same overburden condition (normal stress).
Equation (1) is used to evaluate the interaction coefficient of the reinforcement.

=
(1) where η is the interaction coefficient of reinforcement with soil at a specified normal stress, τ reinforced is the shear strength of reinforced soil, and τ unreinforced is the shear strength of unreinforced soil. An interaction coefficient exceeding unity represents the beneficial effect of reinforcement with effective interlocking in reinforced pavement systems. For clay-subbase interface, Table (9) presents the interaction coefficients of different reinforcements at normal stress equal to 25, 50, 75 and 100 kPa. The interaction coefficient for G1 reinforced with subbase and clay soil interface observed to be in the range of 0.822-1.04, and from 0.99 to 1.10 for G2 reinforced interface at selected normal stresses, this indicates that effective interlocking and better shear stress transfer with G2 in comparison with G1. Otherwise, the behavior of interaction coefficient of the sand-subbase interface appear deferent trend, where increasing normal stress leads to decrease the interaction coefficient it may be attributed to rearrangement of the soil particles. The interaction coefficient is shown in Table (10).

Conclusions
Based on extensive testing of selected subbase over different subgrade soil types; clay and sand soil, and a series of large-scale direct with two types of reinforcement G1 and G2, the following conclusions are found:  For the four normal stresses equal to 25, 50, 75 and 100 kPa, the interface shear stress curves increased and followed similar trend.  For clay-subbase interface, installation of geogrid decreases the apparent cohesion of the material from 16.5 kPa (without reinforcement) to be 8 kPa and 13.5 kPa for G1 and G2, respectively.  For sand-subbase interface, using geogrid leads to increase the cohesion of the material from 3.5 kPa (without reinforcement) to be 15.5 and 16 kPa for G1 and G2, respectively.  The friction angle increases slightly from 30 (without reinforcement) to be 35 o for G1 and G2 when the interface is subbase over clay. While, it decreased from 35.8 o to be 32.1 o for G1 and G2 at sandsubbase interface.  The interaction coefficient for G1 and G2 increased when the normal strength increased at the claysubbase interface. Otherwise, the behavior of interaction coefficient of the sand-subbase interface appears deferent trend, where increasing normal stress leads to decrease the interaction.