The self-developed temperature-controlled large interface static shear apparatus comprises two primary components: an external environmental temperature chamber and an internal static shear system, as depicted in Figure 1. The static shear system is housed within the temperature chamber, which enables precise temperature regulation for interface shear tests over a broad range from −50°C to 300°C, with durations of up to 7 days. The servo load’s normal stress ranges from 0 kPa to 400 kPa, the speed range is 0.1–50 mm/min, and the normal stress accuracy is 1 kPa. The shear loading servo control system can apply stress or displacement-controlled shear loading to the lower shear box through the shear stress loading lever, with the maximum shear force and displacement of 50 kN and 200 mm, respectively, the displacement resolution of 0.001 mm and the accuracy of shear stress of ±0.02%. Static shear tests on the SG interface under controlled temperatures are now possible, thanks to this device.

The geogrid sample used in the experiment was 280 mm wide and 460 mm long, and it was fixed to the front edge of the lower shear box in the static shear apparatus. The static shear test was performed along the longitudinal direction of the geogrid sample. The upper shear box was loaded with marine SS in five layers using a layered compaction method. The height of each layer was 24 mm, totalling 12 mm, and 16 tampers compacted each layer. After compacting the upper shear box fill, cover the rigid bearing plate, apply normal stress, and turn on the temperature control switch for temperature adjustment. The consolidation time of the marine SS sample was 2 hours, and the temperature adjustment time was 2 hours. The examination was performed in a strain-controlled mode, utilizing a shear rate of 1 mm/min. The horizontal tension and shear displacement were recorded automatically throughout the shear process. The test ceases automatically upon the attainment of a shear displacement of 65 mm.

Geogrid is commonly used in marine engineering to reinforce the foundation, the general cover soil thickness is 2.5 m–12.5 m, which exerts about 50 kPa–250 kPa normal stress on the SG interface. Based on the practical application of geogrids in marine engineering, three experimental normal stress values of 50 kPa, 150 kPa, and 250 kPa were selected in order to simulate the stress conditions at different depths that may be encountered in actual reinforcement projects. The normal stress of 50 kPa represents shallow reinforcement applications such as coastal dykes and roadbeds. The normal stress of 150 kPa corresponds to a medium-depth reinforcement scheme, which is geogrid in the engineered fill layer. The normal stress of 250 kPa corresponds to deep foundation applications. In this study, a temperature interval from −5°C to 80°C was selected based on the range of temperature variations common in marine engineering. Intermediate temperatures (0°C, 20°C, 40°C, 50°C, 60°C, and 70°C) were also selected to systematically analyze the effects of temperature variations on the SG interface, considering the value of the experimental results for engineering applications and workload balance. Choosing reasonable temperature intervals can effectively reveal the influence of the law of temperature change on the mechanical behaviour of the interface, especially the change of mechanical response in the high and low-temperature intervals. The shear rate of 1 mm/min was selected based on established standards (ASTM D3080, ISO 17892-10) for direct shear testing of soil-geosynthetic interfaces. A lower shear rate ensures quasi-static loading, allowing stress redistribution and preventing inertial effects or strain localization. This rate aligns with practical geotechnical applications and has been widely adopted in previous studies to accurately capture interface mechanical behaviour under static loading. Under undrained consolidation, the static shear test was carried out on the SG interface using constant temperature control technology. A total of 24 testing programs were designed under three normal stresses (50 kPa, 150 kPa, 250 kPa) and eight different constant temperature conditions (−5°C, 0°C, 20°C, 40°C, 50°C, 60°C, 70°C, and 80°C), as shown in Table 3, to study the effects of different temperatures on the SG interface shear mechanical properties.

Figure 5 illustrates the relationship curves between shear stress and shear displacement of the SG interface at various temperatures and normal stresses.

Figure 5 shows that for most temperatures and normal stresses, the hyperbolic pattern is the most common relationship between shear stress and shear displacement at the SG interface. When the shear displacement grows, the shear stress goes up, down, and then flattens out. Elastic, failure and residual shear are the three separate phases that these curves go through.

In the elastic stage, shear stress and displacement exhibit a linear correlation. At the same level of normal stress, curves for different temperatures nearly overlap, suggesting that temperature has minimal influence on the shear mechanical behaviour of the interface in this stage.

In the failure stage, the SG interface shear stress increases and then decreases with the displacement increase, showing apparent strain-softening characteristics. Strain-softening refers to a phenomenon where, after reaching peak shear stress, the shear resistance decreases as deformation continues, often due to the progressive degradation of material interlocking and bonding. The main reason is that the occlusal embedment between soil particles and geogrid is gradually destroyed with the increase of displacement. This leads to a decrease in shear strength, which is manifested as strain-softening characteristics. The shear displacement corresponding to the peak shear stress increases with the increase of normal stress at the same temperature. For example, under a normal stress of 50 kPa, the shear displacement corresponding to the peak shear stress ranges from 8 mm to 12 mm; under 150 kPa, it ranges from 12 mm to 25 mm; and under 250 kPa, it ranges from 20 mm to 35 mm. However, at high normal stresses and elevated temperatures, the interface may exhibit insignificant strain-softening characteristics or even demonstrate strain-hardening characteristics. Strain-hardening describes a response where shear stress continuously increases with displacement, indicating an increasing resistance to deformation due to material densification or strengthening mechanisms. Specifically, at normal stresses of 50 kPa and 150 kPa, the shear stress initially increases and then decreases with increasing displacement, exhibiting strain-softening characteristics. In contrast, at a normal stress of 250 kPa, the shear stress increases with displacement in the temperature range from 40°C to 80°C, showing strain-hardening characteristics. This is due to the fact that the increased embedded locking effect between sand grains occurs at higher normal stresses due to grain compacting and an increase in contact points. At higher temperatures, the geogrid softens due to thermal expansion, and when shear displacement increases, the embedded locking effect between the sand particles and the geogrid pores becomes more noticeable. Shear strength gradually increases as a result of this. Causing the SG interface to exhibit strain-hardening behaviour.

In the residual shear stage, the SG interface shear stress remains relatively constant as displacement increases. At the same temperature, the residual shear stage is less pronounced under normal stress of 250 kPa than normal stresses of 50 kPa and 150 kPa.

In order to investigate the variation in shear strength of the SG interface as a function of temperature, graphical representations illustrating the relationship between interface shear strength and temperature (specifically at 5°C, 0°C, 20°C, 40°C, 50°C, 60°C, 70°C, and 80°C) are provided under three distinct normal stress conditions (50 kPa, 150 kPa, and 250 kPa), as depicted in Figure 6. The maximum shear stress observed on the relationship curve between shear stress and shear displacement of the SG interface represents the shear strength of the interface. In the context of multiple peak stresses, the initial peak shear stress signifies the shear strength of the interface.

According to Figure 6, the interface shear strength exhibits an initial decline, followed by an increase, and subsequently a further decrease as the temperature rises. Within the temperature range of −5°C–40°C, the interface shear strength exhibits a gradual decline as the temperature rises. This reduction can be ascribed to a diminished bond strength between the sand particles and the geogrid interface. The geogrid material demonstrates increased rigidity in low-temperature environments, resulting in a more pronounced locking effect among the sand particles (Lahoori et al., 2021; Ye and Gao, 2024). Between 40°C and 50°C, the interfacial shear strength significantly increases with the elevation of temperature. The observed increase can be attributed to the moderate softening of the geogrid material at elevated temperatures, which facilitates an enhanced interlocking effect between the sand particles and the geogrid interface, thereby augmenting the shear resistance at the interface (Khan and Latha, 2023; Shu et al., 2025). The enhancement in shear strength is especially notable when subjected to a normal stress of 250 kPa. Nevertheless, within the temperature range of 50°C–80°C, the interface shear strength exhibits a decline as the temperature increases. This indicates that the thermal softening of the material at elevated temperatures considerably influences the mechanical properties of the interface. Within the low-temperature range of −5°C–0°C, the shear strength of the SG interface reaches its peak at −5°C.

The interface shear strength change ratio β is introduced to investigate the sensitivity of the SG interface shear strength to normal stress and temperature changes (Ari and Akbulut, 2024). β = τ 2 − τ 1 τ 1 (1) where, τ 1 represents the interface shear strength under the initial conditions, and τ 2 represents the new interface shear strength after a change in conditions.

The interface shear strength change ratio β (Equation 1) reflects its sensitivity to variations in normal stress and temperature. The shear strength change ratio β at different temperatures and normal stress intervals are presented in Tables 4,5.

As shown in Table 4, the variation in interface shear strength initially increases and then decreases with rising temperature, reaching a peak within a specific temperature range. For normal stress variation intervals of 50 kPa–150 kPa and 50 kPa–250 kPa, the interface shear strength change ratio β reaches its highest values at 40°C, measuring 198.28% and 406.19%, respectively. This indicates that at 40°C, the SG interface is most sensitive to changes in normal stress within these ranges. The underlying reason is the moderate thermal softening of the geogrid material at this temperature, which enhances the interlocking effect between sand particles and the geogrid, thereby improving the interface’s responsiveness to normal stress variations. For the normal stress interval of 50 kPa–150 kPa, the interface shear strength change ratio β is highest at 50°C, with a value of 77.04%. This suggests that the sensitivity of the SG interface shear strength to normal stress changes is most pronounced at 50°C within this range. The interface shear strength change ratio β at room temperature (20°C) is more significant than at both low and high temperatures. Specifically, for the normal stress interval of 50 kPa–150 kPa, the interface shear strength change ratio β is 317.03% at −5°C, 328.99% at 80°C, and 348.53% at 20°C. This indicates that the interface shear strength is more sensitive to normal stress variations at room temperature than at low or high temperatures. To better explain this behaviour, photographs of geogrid samples after testing at different temperature conditions (low, room, and high) are attached, as shown in Figure 7. As shown in Figure 7, no significant deformation of the geogrid occurred at low temperatures. This is because at low temperatures, polypropylene geogrids become stiffer, with a marked increase in stiffness, leading to a reduction in their flexibility. This weakens the occlusal embedment between the geogrid and the sand particles, thereby reducing the sensitivity of the interface shear strength to changes in normal stress. Geogrids undergo significant plastic deformation in high-temperature environments due to the thermal softening effect. This material softening phenomenon significantly reduces structural stiffness, weakening the geogrid’s shear resistance. In contrast, at room temperature, geogrids balance rigidity and flexibility. This balance optimizes the occlusal embedment between the geogrid and the sand particles, enhancing the interface’s mechanical response and making its interfacial shear strength more sensitive to normal stress changes. The observed trends are valid for the polypropylene (PP) geogrids used in this study, as PP geogrids exhibit moderate thermal softening behaviour. However, different geogrids made from alternative polymers, such as polyester or polyethene, may exhibit different responses to temperature variations.

Table 5 shows that as the temperature increases, the interface shear strength change ratio β decreases under normal stress of 50 kPa. This trend suggests that, under normal stress of 50 kPa, the SG interface becomes less sensitive to temperature changes as temperatures increase. Between −5°C and 20°C, the sensitivity to temperature changes is at its peak, and between 50°C and 80°C, it is at its lowest. As the temperature rises, the interface shear strength change ratio β grows under a normal stress of 250 kPa. It can be observed that at a normal stress of 250 kPa, the shear strength of the SG interface is highly sensitive to changes in temperature. The sensitivity to temperature change is greatest in the 5–20°C temperature range, and in the 50–80°C range, it is least.

To investigate the SG interface shear characteristics at different temperatures, the least square method plots the envelope curves of interface shear strength and residual shear strength at various constant temperatures, as shown in Figure 8.

Figure 8 illustrates the linear relationship between shear and residual shear strength and normal stress. The correlation coefficients (R²) for shear and residual shear strength in relation to normal stress range from 0.994 to 1.000, as shown in Table 6. This signifies a robust linear correlation between shear strength and residual shear strength in relation to normal stress. The present study employs the Mohr-Coulomb criterion to analyze this linear relationship. The cohesion and internal friction angle, essential parameters of interface shear strength, are ascertained using the Mohr-Coulomb Failure envelope theorem. According to the theorem, the shear strength ( τ ) at the interface is given by the following equation: τ = c + σ · ⁡ tan φ where, τ represents the shear strength, c represents the cohesion, σ represents the normal stress, and φ represents the angle of internal friction.

The cohesion (c) and friction angle ( φ ) were obtained by fitting the linear envelope of the shear stress data (from both peak and residual stages) to the normal stress, as shown in Table 7, Figures 9,10 .

Although marine SS is usually considered to be cohesionless, the present experimental data measure the SG interface, which differs from pure marine SS. The physical interaction between marine SS particles and geogrid can explain the observed cohesion at the SG interface. The geogrid’s rough surface and large pore size structure create an occlusal embedding interaction with the marine SS particles, effectively increasing the interface’s shear strength. Although marine SS does not have cohesion, this occlusal embedding interaction produces measurable cohesion at the interface. In addition, the tested marine SS had a moisture content of 9.45%, which allowed for a moisture film on the surface of the sand particles. The moisture film creates cohesion between the particles through capillary action or surface tension.

Figure 9 demonstrates that the SG interface friction angle varies between 31.05° and 36.43°, whereas the SG interface residual friction angle fluctuates from 30.92° to 36.32°. The interface friction angle and the residual friction angle initially diminish, then augment, and subsequently decline as the temperature escalates. The minimum values for both angles are recorded at 80°C, while the maximum values are noted at 50°C. At 80°C, the interface friction angle is 31.05°, and the residual friction angle is 30.92°. Conversely, at 50°C, the interface friction angle increases to 36.43°, and the residual friction angle attains 36.32°. As the temperature decreases from 20°C to −5°C, the interface friction angle rises by 3.97%, and the residual friction angle escalates by 8.06%. This suggests that the peak residual interface friction angle exhibits greater sensitivity to temperature fluctuations than the interface friction angle when the temperature declines from 20°C to −5°C. Low temperatures considerably affect the rigidity of the interface material, resulting in more pronounced alterations in interface behaviour under residual conditions (Jin et al., 2022; Shoushtari et al., 2023). When the temperature increases from 20°C to 80°C, the interface friction angle decreases by 9.70%, while the interface residual friction angle decreases by 3.49%. The results indicate that, within the temperature range of 20°C–80°C, the interface friction angle is more sensitive to temperature changes than the interface residual friction angle (Jin et al., 2022; Gao and Ye, 2024). The results indicate that the high-temperature softening effect may significantly impact the initial shear stage. In contrast, the mechanical performance changes during the residual stage are relatively minor.

Figure 10 shows that different values of interface cohesion (from 0.89 kPa to 10.57 kPa) and residual interface cohesion (from 0.26 kPa to 12.93 kPa) are observed. As the temperature increases, the interface and residual cohesion show a pattern of decreasing, increasing, and then decreasing again. At 40°C, the values for both parameters are at their lowest, and 5°C, they are at their highest. The residual and interface cohesion are both lower at 20°C than at the two extremes. An increase of 806.18% in interface cohesion and a rise of 144.43% in residual cohesion are observed as the temperature drops from 20°C to −5°C. This suggests that, within this temperature range, changes in interface cohesion are more noticeable than changes in residual cohesion. Similarly, there is a 647.73% increase in interface cohesion and a 40.48% increase in residual cohesion as the temperature rises from 20°C to 80°C. From 20°C to 80°C, this indicates that, like residual cohesion, interface cohesion is more temperature-sensitive.

The relationship curves between normal and shear displacement in different temperatures are drawn in Figure 11.

Figure 11 depicts the correlation between normal and shear displacement at identical normal stress levels across varying temperatures. In Figure 11, negative normal displacement indicates shrinkage behaviour, whereas positive normal displacement signifies expansion behaviour. The vertical deformation of reinforced soil during the shear displacement process can be categorized into three distinct stages. In the initial stage, the shear displacement is relatively small, the normal displacement is negative, and the reinforced soil undergoes vertical shrinkage. At this stage, as the sand grains are rearranged, the voids between the grains are gradually reduced, and the particles are compacted. In addition, the geogrid enhances the embedded locking effect of the particles so that the sand particles do not undergo large-scale slippage at the initial stage, which promotes the compacting of the soil. This leads to a reduction in the volume of the sandy soil and shear shrinkage. In the second stage, the normal displacement becomes positive as the shear displacement increases, and the reinforced soil exhibits dilatancy. The reason for this phenomenon is the sliding of the sand particles on each other and the surface of the geogrid as the shear displacement increases. Increased relative displacement and space between particles lead to swelling of the sandy soil; thus, the reinforced soil exhibits shear swelling. In the third stage, after reaching the maximum shear stress. The geogrid is gradually withdrawn from the soil layer, and its pores and surface will bring out part of the sandy soil. This results in a continuous decrease in the overall volume of the sandy soil and, thus, a continuous decrease in the normal displacement and shear contraction until the maximum shear displacement is reached.

The shear displacements linked to the changes from shear shrinkage to shear dilation and from shear dilation to shear shrinkage follow similar patterns at different temperatures, as shown in Figure 10, when subjected to the same normal stress. The horizontal displacement linked to the first stage’s completed shear shrinkage grows steadily larger as the normal stress grows. As the second stage nears its end, the horizontal displacement linked to the shear dilation’s end also increases with time. As an example, when normal stress was raised, the horizontal displacement at the end of the first stage of shear shrinkage increased from 5 mm to 8 mm and then to 10 mm. Likewise, as normal stress increased, the horizontal displacement of 15 mm, 20 mm, and 30 mm corresponding to the end of shear dilation in the second stage grew. The soil sample’s shear dilatancy in the second stage became less noticeable or even vanished when subjected to high normal stress and elevated temperatures. Soil samples sheared and vanished in the second stage when subjected to normal stresses of 250 kPa and 80°C; this behaviour persisted at 60°C and 70°C as well.

The initial stage shear shrinkage of the SS sample is negligible or nonexistent at a normal stress of 250 kPa. The dilatancy in the second stage is more noticeable as temperature increases under low normal stress. The normal displacement in the second stage increases by 0.38 mm at −5°C, 0.4 mm at 0°C, 0.45 mm at 20°C, and 0.6 mm at 80°C, according to a normal stress of 50 kPa. The normal displacement turns positive at a maximum normal stress of 250 kPa and a low temperature of −5°C. This is probably because the sand density increases and the pore water ice volume increases under high stress (Chao et al., 2025; Gao et al., 2024).