An Experimental Study on the Interface Characteristics of Geogrid-Reinforced Construction and Demolition (C&D) Waste Recycled Aggregate Based on Pullout Tests

Abstract

China generates substantial construction and demolition (C&D) waste, owing to rapid urbanization. However, the resource utilization rate of C&D waste remains low. This work is devoted to promoting the application of C&D waste in reinforced soil structures. In this research, the physical and mechanical properties of C&D waste recycled aggregate, biaxial geogrids and triaxial geogrids were first clarified. Then, a series of pullout tests were carried out based on the large-size pullout test setup. With the help of macroscopic indicators, including pullout resistance, horizontal displacement and interface friction coefficient, the effects of normal stress, pullout rate and reinforcement type on the characteristics of the reinforcement–C&D waste recycled aggregate interface were clarified. The test results show that normal stress has the greatest influence on pullout resistance. The pullout rate has the lowest effect on pullout resistance. In addition, the interface effect between the triaxial geogrid and the C&D waste recycled aggregate is more significant than that in biaxial geogrid–C&D waste recycled aggregate. The interface friction angle of triaxial geogrids is 18.1% higher than that of biaxial geogrids (11.6° vs. 9.82°), correlating with an enhanced particle interlocking mechanism.

1. Introduction

The rapid advancement of industrialization and urbanization has precipitated a substantial surge in construction and demolition (C&D) waste, which has emerged as a critical environmental challenge demanding urgent scholarly attention and policy intervention. C&D waste not only occupies substantial land resources but also causes environmental pollution. Given the severity of this issue, numerous nations have implemented measures to address the challenges posed by C&D waste [1]. China’s ongoing expansion in transportation infrastructure has precipitated a significant demand for premium-grade fill materials for motorway and railway networks. However, environmental protection imperatives have imposed stringent restrictions on extracting conventional fill materials, including sand and gravel. This constraint has precipitated a supply–demand imbalance in premium-grade fillers and transportation infrastructure requirements [2]. Within the global context, industrialized nations have advocated for advanced recycling methodologies that systematically reclassify C&D waste as renewable material resources [3]. This suggestion demonstrates similar applicability to the sustainable advancement of China’s transportation infrastructure sector. Given the escalating scarcity of land and natural resources in China, the scientific management and integrated utilization of C&D waste have emerged as crucial imperatives. Effectively achieving waste minimization, resource recovery, and environmental neutralization now constitutes a critical challenge and demands urgent resolution in engineering construction practices [4].

C&D waste generally comprises brick, concrete fragments, steel, and timber [5]. C&D waste recycled aggregate can be obtained through the systematic processing of demolition debris. This procedure involves the removal of metals, glass fragments, wood, and plastic, followed by mechanical crushing and purification. C&D waste recycled aggregate predominantly comprises inorganic constituents, primarily concrete fragments and brick. These materials exhibit chemical stability under acidic/alkaline conditions, hydrophobic properties, and hydraulic conductivity. Their key mechanical characteristics are defined by the coarse particle gradation, low plasticity indices, and dimensional stability [6]. C&D waste recycled aggregate is extensively employed in transportation infrastructure components, specifically for embankment construction [7], subgrade stabilization layers [8], and pavement [9]. Empirical evidence confirms that C&D waste recycled aggregate demonstrates viability in motorway embankments when subjected to rigorous processing protocols. Recycled brick and concrete aggregate demonstrate geotechnical properties equivalent to, and even superior to, conventional fill materials. However, crushed brick and concrete aggregate are required to blend with quality aggregate and stabilizing agents for motorway pavement subgrade applications. Jiménez et al. [10] documented that C&D waste recycled aggregate exhibits acceptable structural capacity in low-volume roadway applications. Current research demonstrates that C&D waste recycled aggregate exhibits significant potential as recycled aggregates in transportation infrastructure applications. Recycled concrete aggregate exhibits lower compressive strength compared to natural aggregates. Nevertheless, its mechanical performance conforms to established engineering parameters for structural applications [11]. Recycled concrete aggregate processed from C&D waste can only serve as a partial substitute for natural aggregates in both cement concrete [12] and asphalt mixtures [13]

Although C&D waste recycled aggregate has been extensively utilized in various engineering applications, its incorporation into reinforced soil structures remains underexplored. A reinforced soil structure constitutes an engineered composite system incorporating facing panels, compacted fills, and reinforcement elements [14]. This technology has been widely implemented across transportation infrastructure projects, yielding measurable socioeconomic and ecological benefits [15]. Granular materials are commonly recommended as filling materials for reinforced soil structures. This preference stems from their superior shear strength and effective pore pressure dissipation characteristics [16]. In geotechnical construction practice, engineers frequently adopt locally available fill materials. High-quality fillers are difficult to obtain. In reinforced soil structures, clay-based backfill materials with low permeability are occasionally employed. Such soil selections can compromise geotechnical performance parameters, potentially jeopardizing the structural integrity of the wall system. Investigating the viability of C&D waste recycled materials in reinforced soil structures thus becomes imperative.

The mechanical performance of reinforced soil structures is predominantly governed by the interface mechanism between filling materials and geosynthetic reinforcements [17]. Interfacial shear behavior critically influences load transfer efficiency and stress redistribution patterns. The interfacial characteristics between the fill and reinforcement form an essential theoretical foundation for the mechanical behavior analysis and structural design of reinforced soil structures [18]. The direct shear test and the pullout test serve as the primary methods for investigating fill–reinforcement interface mechanisms [19]. This experimental approach quantifies geogrid displacement behavior under controlled overburden pressures and elucidates frictional resistance, mechanical interlock, and confinement effects within composite systems [20]. Vieira et al. used three commercially available geosynthetics for direct shear tests and pullout tests: an extruded uniaxial HDPE geogrid, a laid uniaxial PET geogrid and a high-strength composite geotextile. C&D waste recycled aggregate exhibited shear strength similar to that of traditional backfill when properly compacted [21]. Arulrajah et al. investigated the characteristics of the reinforced C&D waste recycled aggregate reinforced with biaxial geogrid and triaxial geogrid by direct shear tests. The results indicated that the triaxial geogrid showed satisfactory performance [22]. The above studies aimed to evaluate the reinforcement effect of different types of geosynthetics on C&D waste recycled aggregate. Sarkar et al. assessed the interaction of biaxial geogrids with C&D waste recycled aggregate, steel slag and the conventional backfill material, sand, using direct shear tests and pullout tests. The results showed that the interaction was better with steel slag and C&D waste recycled aggregate than with sand [23].

Soil–geosynthetic interface behavior has been extensively investigated. However, research addressing interactions between geosynthetics and C&D waste recycled aggregate remains limited. Existing studies primarily focus on the reinforcement of C&D waste recycled aggregate using uniaxial and biaxial geogrids. Investigations into triaxial geogrid-reinforced C&D waste recycled aggregate are scarce. This study employs pullout tests to examine the interfacial behavior between C&D waste recycled aggregate and reinforcement. The derived macro-mechanical parameters establish fundamental design criteria for implementing these sustainable composites in reinforced soil structures. This study begins with the clarification of the physical and mechanical properties of geosynthetic reinforcements and C&D waste recycled aggregate. Pullout tests subsequently analyze normal stress magnitude, reinforcement classification, and displacement rate impacts on C&D waste recycled aggregate–reinforcement interface mechanisms. This research direction addresses critical challenges in sustainable geotechnical engineering practices.

2. Experimental Details

2.1. Test Setup

The test setup employed for determining the interface characteristics of C&D waste recycled aggregate–geogrid comprises four components, a pullout test box, a normal loading system, a horizontal pullout control system, and a test control and data acquisition system, as shown in Figure 1. To reduce the rigid boundary effect on the front wall, two 400 mm × 50 mm sleeves are fitted adjacent to the sidewall openings. The loading system consists of a hydraulic jack with a calibrated maximum capacity of 150 kN, a 150 kN pressure sensor, and a reaction frame. The working principle of the normal loader is as follows: the load is applied to the reaction frame by the hydraulic loader and the filler is loaded by the reaction force generated. The horizontal loading device consists of a servo system and a fixture. The force measuring system has a nominal range of 100 kN to measure the horizontal pullout resistance at any time during the pullout process. The measuring range of the front-and-rear displacement transducers is 150 mm, and the displacement accuracy is ±0.02mm.

2.2. Material

(1)

C&D waste recycled aggregate

The fill material is C&D waste recycled aggregate collected from the demolition of rural houses. C&D waste cannot be directly used in this research because of its complex composition, poor particle gradation, and large particle size. C&D waste underwent a series of treatments before being used for pullout tests, including the removal of contaminants such as wood, metal, and plastic, mechanical crushing, and cleaning.

Bricks, concrete, and ceramic tiles are the main materials in the C&D waste recycled aggregate used in pullout tests. Their mass ratio is 6:3:1. The screening test and compaction test of C&D waste recycled aggregate were performed according to the Highway Geotechnical Test Regulations (JTG 3430-2020) [24]. The particle size distribution and compaction curve are shown in Figure 2. The physical properties of C&D waste recycled aggregate are shown in

Table 1. The physical and mechanical properties of C&D waste recycled aggregate.

D60/mm	D30/mm	D10/mm	Optimum Moisture Content/%	Maximum Dry Density /kg·m−3
20.00	9.75	0.65	9.26	1720

.

(2)

Geogrid

The geogrids employed in this test are biaxial geogrids and triaxial geogrids, which are commercially available, as shown in Figure 3. To determine the mechanical properties of the two geogrids, a series of tensile tests were conducted according to Geosynthetics—wide-width tensile test (GB/T 15788-2017) [25]. The size of the specimen used in tensile tests of the geogrid was 200 mm × 300 mm, the clamp spacing was 200 mm, and the tensile rate was 40 mm/min. The ultimate tensile strength, ultimate elongation, and tensile modulus of geogrids under varying strain conditions are presented in

Table 2. Technical indices of geogrid.

Type	Mesh Size/mm	Ultimate Tensile Strength/kN·m−1	Ultimate Elongation/%	Tensile Modulus/kPa
				2%	5%
Biaxial geogrid	40 × 40	39.56	8.92	386,500	283,200
Triaxial geogrid	40 × 40 × 40	35.360	8.38	317,500	230,360

.

To maintain constant frictional area of geogrids in fill during testing, the effective length was set at 60 cm. This specification was determined based on the test box dimensions (60 cm × 40 cm). Baykal and Dadasbilge [17] proposed that a length-to-width ratio of approximately 2 more effectively eliminates boundary effects. The effects originate from side friction during experimental procedures. Wang [26] proposed that large-dimension rectangular grid specimens are employed in pullout testing. This configuration enhances shear strength development stability during reinforced soil interface testing. Additionally, it improves test result analyzability and consistency with field conditions. Based on these considerations, the grid specimen width was determined as 30 cm. This configuration yields an effective geogrid sample size of 60 cm × 30 cm in the test box.

2.3. Specimen Preparation

Two pullout specimen configurations are investigated in this research.

(1)

Biaxial geogrid-reinforced C&D waste recycled aggregate.

(2)

Triaxial geogrid-reinforced C&D waste recycled aggregate.

The pullout test specimens were fabricated within the pullout test box. The specimens comprised four layers of C&D waste recycled aggregate and one reinforcement layer. The fill stratification and reinforcement placement are depicted in Figure 4. Before filling the sample, lubricant was evenly applied to the inner wall of the box. During the test, the compaction degree of C&D waste recycled aggregate was controlled by the mass control method. The height and mass of each packing layer were fixed, respectively, at 125 mm and 48.50 kg, resulting in a 94% compaction degree. The pressure plate and normal-direction loading device facilitate accurate compaction. A layer of geogrid was horizontally positioned on top of the second fill layer when the filling height of the C&D waste recycled aggregate reached 25 cm. The loading end of the geogrid was secured to the traction device via a clamping mechanism, while the free end remained unrestrained. The free end of the geogrid is ≥70 mm. This method prevented alterations to the reinforced soil interface’s effective area during pullout testing. Ultimately, the third and fourth layers were constructed with the C&D waste recycled aggregate.

2.4. Test Program

The effect of the reinforcement type, normal stress, and pullout rate on the interface behavior between the geogrid and C&D waste recycled aggregate were investigated. Experimental configurations are systematically documented in

Table 3. Pullout test scheme.

Code	Geogrid	Pullout Rates/mm·min−1	Normal Stress/kPa
P1	Biaxial geogrid	1	25, 50, 75 and 100
P2		2	25, 50, 75 and 100
P3		3	25, 50, 75 and 100
P4	Triaxial geogrid	2	25, 50, 75 and 100

. Normal stress levels of 25, 50, 75, and 100 kPa were applied. Strain-controlled testing employed pullout rates of 1.0, 2.0, and 3.0 mm/min. Test termination conditions vary considerably in criteria for pullout tests among scholars. Wang [27], Agarwal [28], Tajabadipour [19], and Prasad [29] selected maximum clamp displacements of 80 mm, 90 mm, 40 mm, and 110 mm as the termination conditions of pullout tests. At the same time, excessive wall horizontal displacement is not permitted by guidelines for mechanically stabilized earth walls. BS 8006 and the NGG standard adopt upper limits of 0.5% wall height and 0.1%–0.3% wall height, respectively. Base on the above previous studies and the change in pullout resistance during the pullout testing, a clamp displacement of 70 mm was selected as the termination criteria in this research.

Moreover, the interface friction between filler material and confinement walls may adversely impact pullout test outcomes. A portion of the normal load is dissipated through wall frictional resistance, diminishing the effective stress transmitted to reinforcement layers. Therefore, the inner walls of the test box were smeared with Vaseline.

2.5. Test Responses

Pullout tests were conducted in strict compliance with the code of Test Methods of Geosynthetics for Highway Engineering (JTG E50-2006) [30]. Three key parameters are monitored during the testing process: the pullout resistance, the displacement of geogrids, and normal stress exerted on the vertical loading plate.

Based on the pullout tests results, the interface friction coefficient is determined using Equation (1). The interface friction coefficient characterizes the frictional behavior and dilatancy properties of the material interface. Moraci et al. [31] demonstrated that the interface friction coefficient of soil–reinforcement interfaces at peak and residual states depends on both normal pressure and grid geometric characteristics.

$${\mu }_{s/GSY}={\displaystyle \frac{P_R}{2L_e{\sigma }_v^{\prime }}}$$

(1)

where

${\mu }_{s/GSY}$: The interface friction coefficient between geogrids and fill (dimensionless).

$P_R$: The maximum pullout resistance per unit width (kN/m).

$L_e$: The effective length/length of the reinforcement in the resistant zone (m).

${\sigma }_v^{\prime }$: Effective confining stress at the fill–geosynthetic interface (kN/m²).

The pullout test directly records pullout resistance and the corresponding displacement. The growth rate of pullout resistance is calculated using Equation (2) during the pullout test.

$$\xi ={\displaystyle \frac{P_j-P_{j-1}}{P_{j-1}}}$$

(2)

where

$j$: The serial number of the data points (from 2 to n);

$P_j$: The j-th value of pullout resistance data;

$P_{j-1}$: The j-1-th value of pullout resistance data.

3. Pullout Test Results

3.1. Influence of Confining Stress

The pullout behavior of biaxial geogrid-reinforced C&D waste recycled aggregate was evaluated under normal stresses of 25, 50, 75, and 100 kPa and a pullout rate of 2 mm/min, as shown in Figure 5a. Figure 5a demonstrates that pullout resistance versus displacement curves exhibit analogous behavioral patterns across varying normal stress conditions. The pullout resistance evolution exhibited two distinct phases during testing. Initially, pullout resistance rapidly increased with increasing displacement. Subsequently, a deceleration in pullout resistance development preceding ultimate stabilization occurred at maximum resistance. Meanwhile, Figure 5a reveals discrepancies among the four pullout test groups. At identical reinforcement displacements, pullout resistance increases proportionally with applied normal stress.

Figure 5b quantifies pullout resistance increase rates across different displacement ranges. The mean incremental rates within three displacement intervals (0–5 mm, 5–30 mm, and 30 mm to end) are comparatively analyzed. Figure 5b demonstrates a negative correlation between the pullout resistance development rate and reinforcement displacement. A marked rate reduction occurs after the 5 mm displacement threshold. During the initial 0–5 mm reinforcement displacement phase, the increasing rates of pullout resistance under normal stresses of 25–100 kPa are recorded as 0.334 (5 kPa), 0.342 (50 kPa), 0.354 (75 kPa), and 0.369 (100 kPa). During the 5–30 mm reinforcement displacement phase, the increasing rates of pullout resistance under normal stresses of 25–100 kPa are recorded as 0.025 (25 kPa), 0.026 (50 kPa), 0.030 (75 kPa), and 0.030 (100 kPa).

The load transfer mechanism in geogrid reinforcement involves progressive stress propagation. This stress redistribution originates from the loaded end and extends toward the anchored extremity. During initial pullout testing, geogrid displacement induces particle rearrangement in adjacent C&D waste recycled aggregate. Simultaneously, frictional resistance and particle interlocking mechanisms between the geogrid-reinforced interface and C&D waste recycled aggregate are instantaneously activated. Enhanced normal stress intensifies interfacial frictional resistance and particle interlocking mechanisms at the geogrid–C&D waste recycled aggregate interface. Enhanced normal stress leads to a proportional escalation of pullout resistance. During the final pullout phase, near-loaded geogrid segments achieve full stress mobilization. Distal reinforcement regions exhibit limited stress activation capacity, resulting in a progressive reduction in rates of increase in pullout resistance.

Pullout test-derived displacement–pullout resistance relationship curves enabled the quantification of interfacial peak shear stress evolution patterns. The shear strength characteristics and corresponding interface friction coefficient variations under different normal stresses are presented in Figure 5c,d. Peak shear stress exhibits proportional escalation with increasing normal stress. Concurrently, interface frictional resistance demonstrates a positive correlation with normal stress magnitude.

3.2. Influence of Reinforcements Types

Triaxial geogrid-reinforced C&D waste recycled aggregate was tested under varying normal stress conditions. The corresponding pullout test results are comprehensively illustrated in Figure 6a. Figure 6a reveals the following conclusions: (1) At a normal stress of 25 kPa, biaxial geogrid-reinforced C&D waste recycled aggregate exhibited superior pullout resistance compared to triaxial configurations under equivalent displacement conditions. This differential persists across matched interface displacement levels. (2) At normal stresses of 50–100 kPa, triaxial geogrid-reinforced C&D waste recycled aggregate demonstrated superior mechanical performance compared to biaxial counterparts under equivalent displacement conditions. This interfacial strength differential persists across matched displacement parameters.

Figure 6b,c quantify the peak shear stress evolution and interface friction coefficient variations under different normal stress conditions, respectively. Figure 6b,c demonstrate the effect of geogrid type on mechanical parameters: pullout resistance, peak shear stress, and the interface friction coefficient. Biaxial geogrid-reinforced C&D waste recycled aggregate exhibited enhanced peak shear stress and interface friction coefficient values under low normal stress conditions compared to triaxial configurations. However, triaxial geogrid-reinforced C&D waste recycled aggregate exhibited superior peak shear stress and interface friction coefficient values under higher normal stress conditions compared to biaxial configurations.

Pullout resistance originates from two fundamental mechanisms in geosynthetic reinforcement systems: frictional resistance and particle–geogrid interlocking mechanisms [32]. Under low normal stress conditions, particle–geogrid interlocking exhibits minimal activation. Pullout resistance is predominantly governed by frictional resistance under such stress states. Elevated normal stress conditions intensify particle–geogrid interlocking. Particles at geosynthetic–reinforcement interfaces undergo rotational displacement, inducing pronounced particle dilatancy [33]. Pullout resistance in geosynthetic-reinforced C&D waste recycled aggregate is primarily governed by particle–geogrid interlocking. Compared to triaxial geogrids, biaxial geogrids feature larger apertures and thicker rib elements. This configuration enhances the interfacial contact area with C&D waste recycled aggregate. In contrast, triaxial geogrids possess greater junction density and finer ribs. The proliferation of junctions coupled with finer rib dimensions may compartmentalize filler materials into discrete cellular units, thereby reducing the continuous, effective C&D waste recycled aggregate contact surface area along the rib–backfill interface. Consequently, under low confining pressures, biaxial geogrids with superior frictional contact areas exhibit enhanced reinforcement performance. Conversely, under high confining stress conditions, the structural superiority of triaxial geogrids manifests. Triaxial geogrid geometries induce more significant particle displacement fields in granular media than rectangular counterparts at fill–geosynthetic interfaces. This enhanced particle rearrangement amplifies dilatancy effects, thereby optimizing interfacial frictional resistance through improved interlocking mechanisms. The maximum values correspond to the predominant interfacial shear stress and pullout resistance observed under most operational conditions. Triaxial geogrid-reinforced C&D waste recycled aggregate exhibits the highest recorded interfacial resistance parameters.

3.3. Influence of Pullout Rates

Pullout tests were conducted under four normal stress levels (25, 50, 75, and 100 kPa) with displacement rates of 1.0, 2.0, and 3.0 mm/min. The pullout resistance–displacement relationship curves are presented in Figure 7. Figure 8 illustrates the correlation between normal stress and both peak shear stress and the interface friction coefficient. These parameters are derived from horizontal displacement–pullout resistance curves. The findings are summarized as follows: (1) As illustrated in Figure 7, during the initial phase of the pullout process, the influence of the displacement rate on pullout resistance is minimal. The curves from the pullout tests exhibit nearly complete overlap at the normal stress levels of 25, 50, 75, and 100 kPa. (2) Figure 7 demonstrates that the effect of the displacement rate on pullout resistance becomes progressively more pronounced with the increasing horizontal displacement of reinforcement. (3) During the advanced phase of the pullout process, the significance of the displacement rate on pullout resistance becomes progressively more pronounced as the magnitude of normal stress escalates. (4) Figure 8 demonstrates that the discrepancy between peak shear stress and the interface friction coefficient exhibits a proportional relationship with normal stress enhancement across varying displacement rate conditions.

This phenomenon may result from higher pullout rates compromising interfacial properties between the reinforcement and C&D waste recycled aggregate. This mechanism can be described as follows: Pullout resistance is initially applied at the loading end and progressively transferred away from this point. Increasing the pullout rate accelerates force transmission to the geogrid’s non-bearing end. This rapid transfer leaves insufficient time for coordinated deformation between the reinforcement and C&D waste recycled aggregate at the interface. Consequently, C&D waste recycled aggregate adjacent to the reinforcement cannot undergo effective rearrangement. During the initial phase of the pullout test, the traction-induced pullout resistance remains relatively low. The frictional component constitutes the primary source of pullout resistance. During the latter phase of the pullout test, as pullout resistance propagates along the geogrid, the transmission distance progressively increases. This process activates enhanced interlocking mechanisms and fixation effects between the constituent materials within the C&D waste recycled aggregate. The pullout rate exhibits a positive correlation with both the mobilization of C&D waste recycled aggregate and the resultant pullout resistance during testing. This phenomenon demonstrates proportional amplification under elevated normal stress conditions.

4. Conclusions

The interface characteristics of biaxial/triaxial geogrid-reinforced C&D waste recycle aggregate are revealed based on a series of pullout tests. The main conclusions are as follows:

(1)

Pullout resistance versus displacement curves exhibit analogous behavioral patterns across varying normal stress, geogrid types, and pullout rates. Initially, rapid pullout resistance escalation occurs with increasing displacement, especially during the pullout displacement of 0–5 mm. The subsequent development of pullout resistance decelerates and eventually stabilizes.

(2)

The normal stress and the type of reinforcement have a great influence on pullout resistance and the interface friction coefficient, while the pullout rate has little effect. In addition, the influence of the reinforcement type and pullout rate on the mechanical characteristics of the interface between the reinforcement and C&D waste recycled aggregate increases with an increase in normal stress.

(3)

The biaxial geogrid shows a better reinforcement effect compared to the triaxial geogrid under low normal stress (≤50 kPa). Conversely, the triaxial geogrid shows a better reinforcement effect under high normal stress (≥75 kPa). The selection of reinforcement in reinforced soil structure should fully consider the important indexes such as the type of filler and the height of the structure.

The superior performance of triaxial geogrids fundamentally stems from their triangular geometry-driven structural stability and isotropic mechanical behavior. This combination makes them an optimal solution under high-stress conditions and complex loading scenarios, where they significantly outperform conventional biaxial geogrids, particularly in embankment zones exceeding 4 m fill heights or confined construction spaces such as steep slopes.