Introduction

High-steep slopes are widely distributed in mountainous regions, where strong seismic activity, complex geological conditions, and large elevation differences frequently induce severe instability problems1,2. Earthquake-triggered landslides and collapses are among the most destructive secondary hazards in seismic areas, repeatedly causing substantial casualties and economic losses3. With the rapid expansion of transportation and infrastructure into mountainous seismic regions, enhancing the seismic reliability of slope-support systems has become a critical engineering challenge4,5.

Sheet-pile wall (SPW) systems are widely used to stabilize high-steep slopes due to their simple construction, strong lateral resistance, and effectiveness in controlling sliding deformation6,7. Post-earthquake field investigations following major earthquakes, such as the Wenchuan and Lushan events, indicate that SPWs generally exhibit better seismic integrity than conventional gravity retaining structures8. Consequently, extensive experimental and numerical studies have investigated the seismic response of SPWs, focusing on acceleration amplification, dynamic earth pressure, bending moment distribution, and failure modes9,10,11,12,13. However, these studies consistently report that conventional reinforced concrete (RC) SPWs are prone to stiffness degradation, brittle cracking, and limited energy dissipation under strong seismic excitation14.

To improve seismic performance, anchor cables are often incorporated into SPW systems, forming anchored sheet-pile wall (ASPW) configurations. The introduction of anchors provides additional lateral restraint, modifies the internal force distribution of piles, and alters soil-structure load transfer mechanisms15,16,17,18,19. Recent studies have demonstrated that ASPW systems outperform unanchored SPWs in terms of deformation control, dynamic earth pressure reduction, and overall seismic stability20,21,22,23. Nevertheless, experimental observations and post-earthquake damage reports indicate that even anchored RC piles may still experience cracking or brittle failure under strong shaking, suggesting that structural optimization alone cannot fully resolve material-related seismic vulnerabilities24.

In parallel, engineered cementitious composites (ECC) have received increasing attention in seismic engineering due to their high tensile ductility, strain-hardening behavior, and superior crack-width control25,26,27,28,29. Compared with conventional RC, ECC exhibits multiple micro-cracking and sustained tensile capacity under cyclic loading, leading to improved deformation compatibility and energy dissipation30,31,32. While ECC has been widely studied in structural and foundation applications, its use in slope-support systems remains limited. Existing studies mainly focus on unanchored ECC-SPW systems and demonstrate that replacing RC with ECC can improve crack resistance and deformation capacity under seismic loading33. However, these studies primarily address material effects and do not consider how ECC interacts with structural reinforcement measures such as anchorage.

Notably, although both anchor cables and ECC materials have been shown individually to enhance seismic performance, their combined use in ASPW systems has not been systematically investigated from a mechanism perspective. Specifically, it remains unclear how material ductility enhancement (ECC) and structural constraint modification (anchors) interact to influence load redistribution, damage evolution, and deformation compatibility during seismic excitation. Existing ASPW studies predominantly focus on anchor forces or global response indicators21,34, while ECC-related studies emphasize material-level improvements, leaving the synergistic seismic improvement mechanism between ECC and anchorage largely unexplored.

To address this gap, this study conducts shaking table tests on high-steep slopes supported by ECC and RC based anchored sheet-pile wall systems. Rather than focusing on a simple material comparison, the study aims to clarify the synergistic mechanisms arising from the combined application of ECC materials and anchor cables. By systematically analyzing acceleration response, dynamic earth pressure, pile bending moments, and residual deformation, and by comparing anchored and unanchored configurations reported in previous studies, this research elucidates how material ductility enhancement and structural restraint interact to improve seismic performance. The findings provide new experimental insight into the coupled material-structural optimization of ASPW systems and contribute to the development of resilient slope-support strategies in mountainous seismic regions.

Test methodology

Testing devices and scaling laws

The comparative shaking table tests on ECC and RC anchored sheet-pile wall (ASPW) supported slopes were conducted on the ANCO seismic platform at the Geotechnical Engineering Laboratory of Chongqing University4. The shaking table has a loading surface of 1.2 m × 1.2 m, supports excitation frequencies from 0 to 50 Hz, and provides a maximum payload capacity of 1 t35. A rigid model box with external dimensions of 0.8 m × 0.6 m × 1.5 m (length × width × height) was anchored to the table using high-strength bolts to ensure stable boundary conditions throughout testing.

To ensure dynamic similarity between the scaled model and its prototype, the model design followed established similitude principles36. The Buckingham π theorem, widely adopted in geotechnical shaking table modelling, was applied in this study to derive the governing scaling laws37,38. Geometric length (l), gravitational acceleration (g), and equivalent density (ρ) were selected as the fundamental governing parameters39. Considering equipment capacity, a geometric scale factor of 1:10 (model: prototype) was used, while density and acceleration were kept at a similarity ratio of 1:1 to directly reproduce prototype inertial effects (Table 1).

Table 1 Model similarity ratio of shaking table test.
Full size table
Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

(a) Arrangement of shaking table test; (b) Layout diagram of sensors.

The physical model consisted of four main components: the sliding mass, bedrock, anchor cable, and the sheet-pile wall retaining system. Due to the limited table size, the dimensions of sheet-pile wall retaining system were scaled consistently with the adopted similarity ratios and followed the configuration reported in Ma et al.33. Specifically, square anti-slide piles with a cross-section of 10 cm × 10 cm (equivalent to 1 m in prototype) and an embedded length of 110 cm (scaled to 11 m in prototype) were used. The facing slab was cast as a 2 cm-thick reinforced panel (scaled to 0.2 m in prototype) with a clear spacing of 30 cm (3 m in prototype) between piles, as illustrated in Fig. 1.

Material properties and instrumentation layout

All materials adopted in the physical model were prepared to satisfy the required similarity relationships. C25 concrete was used to simulate the bedrock, providing a rigid base for the embedded portion of the sheet-pile wall retaining system. To ensure mechanical similarity of the sliding mass, its density and shear strength parameters were adjusted according to similitude requirements40. The sliding mass was formulated using a mixture of river sand, clay, and water in a ratio of 23:2:1, determined through multiple direct shear tests. The resulting material exhibited a density of 2 g/cm³, cohesion of 8 kPa, and internal friction angle of 28°.

Considering that prototype concrete typically exhibits an elastic modulus of 30 ~ 45 GPa, its scaled value for the model was set within 3 ~ 4.5 GPa in accordance with the 1:10 geometric similarity ratio. To ensure mechanical similarity of the retaining system, micro-concrete was used to model the RC sheet-pile wall system, while a low-strength ECC mixture was adopted to represent the ECC sheet-pile wall system. The micro-concrete, prepared using a cement: water: standard sand ratio of 1:0.9:5, exhibited a compressive strength of 8.5 MPa and an elastic modulus of 4.2 GPa. The ECC material, produced from cement, fly ash, quartz sand, and water at a mass ratio of 0.15:0.85:0.264:0.37 with 0.93% water reducer and 2% PVA fiber41, achieved a compressive strength of 7.2 MPa and an elastic modulus of 3.5 GPa. The comparatively lower stiffness of ECC resulted from the absence of coarse aggregate, consistent with its material characteristics42. Reinforcement within the structural components was simulated using galvanized iron wires. The anchor cable was modeled with 3 mm steel strands. The anti-slip piles, concrete slabs, and anchor cables collectively constituted the ASPW retaining system used in the shaking table tests, as illustrated in Fig. 2a.

A comprehensive instrumentation scheme was implemented to monitor the seismic response and deformation characteristics of the slope-structure system. Accelerometer A11 was mounted on the shaking table to record input motions, while additional accelerometers were distributed within the slope to measure vertical and horizontal acceleration propagation. Six soil pressure cells were affixed to the rear surface of the central pile to capture dynamic earth pressure variations. Four laser displacement sensors were installed to measure lateral displacements of the retaining structure. Furthermore, seven pairs of strain gauges were symmetrically embedded along the pile shaft to quantify bending moment development during seismic loading.

Test cases and input motions

Two shaking table tests were performed to assess and compare the seismic performance of slopes supported by ECC and RC anchored sheet-pile wall (ASPW) supporting systems. The input excitation consisted of a 5 Hz sinusoidal wave with a 20 s duration43,44, and the peak acceleration was incrementally increased from 0.1 to 1.2 g. To evaluate the evolution of the structural dynamic characteristics, white-noise excitations covering 0 ~ 30 Hz were applied for 20 s both before and after the seismic loading sequence45. These tests enabled monitoring of variations in the natural frequency of the slope-structure system throughout the shaking process.

Results

Failure characteristics and structural stiffness degradation

Figure 2 presents the arrangement of anchor cable and compares the failure modes of slopes supported by RC-ASPW and ECC ASPW supporting systems. When the amplitude of the input seismic motion was relatively small (PGA < 0.4 g), both slopes remained within the elastic range, and the pile-soil-anchor system exhibited coordinated deformation. With increasing input seismic motion, horizontal tensile cracks first appeared at the slope crest and gradually propagated, while the supporting structures developed minor tilting yet maintained overall stability. This progressive damage pattern is consistent with earlier shaking table studies46,47. It is worth noting that the differences between the two supporting systems became more pronounced with the increase of input peak acceleration. As for the RC-ASPW supported slope (Fig. 2c), a network of intersecting cracks formed at the crest, and three major through-going cracks developed in the mid-slope, indicating the nearing failure of the slope. In contrast, the ECC -ASPW supported slope sustained less damage, exhibiting only a localized through-crack near the upper slope (Fig. 2d) while maintaining overall structural integrity. Post-test excavation further confirmed these differences. The connections between anchors and the sheet-pile wall system remained intact, with no signs of pull-out or anchorage failure. However, the RC pile exhibited a pronounced through-going cracks at the fixed section, whereas the ECC pile exhibited no such cracking and retained excellent structural integrity, as shown in Fig. 2b. The superior performance of ECC is attributed to the fiber-bridging mechanism, which suppresses crack propagation and enhances ductility and energy dissipation26,48.

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

Anchor layout and slope failure characteristics: (a) anchor cable arrangement; (b) failure at pile anchorage zone; (c) cracking in RC-ASPW slope; (d) cracking in ECC-ASPW slope.

The natural frequencies of the slope-structure systems were identified using the peak response amplitudes obtained from white noise excitation, which enables the evaluation of stiffness degradation induced by seismic loading49. The first-order natural frequency and damping ratio were determined from the transfer functions computed at measurement points A2 and A1139,44. Figure 3 presents the evolution of these dynamic parameters for both anchored sheet-pile wall (ASPW) and unanchored sheet-pile wall (SPW) systems. Prior to seismic loading, the ECC-ASPW and RC-ASPW slopes exhibited nearly identical initial natural frequencies (15.3 and 15.4 Hz) and damping ratios (4.8% and 5.4%), attributable to their comparable compressive strength and elastic modulus. As the input PGA increased, all models demonstrated a monotonic decrease in natural frequency accompanied by an increase in damping ratio, indicating progressive nonlinear behavior and cumulative damage within the soil-structure system. This trend is consistent with previously reported observations in seismic loading45,50.

When PGA ≤ 0.6 g, the ECC-ASPW and RC-ASPW models exhibited nearly identical degradation patterns, suggesting that the supporting structure remained primarily within the elastic regime, and the observed changes in dynamic parameters were governed by the progressive loosening of the soil particle-pore structure and accumulation of internal damage within the slope. However, when the PGA exceeded 0.6 g, the RC-ASPW system showed accelerated frequency reduction and damping ratio growth, indicating the onset of elastoplastic deformation and more pronounced structural deterioration. Finally, the natural frequency and damping ratio of the ECC-ASPW system reached 13.0 Hz and 18.5%, respectively, whereas those of the RC-ASPW system declined to 12.2 Hz and increased to 22.7%. The slower degradation of the ECC-supported slope demonstrates its superior resistance to stiffness degradation and cumulative damage under strong seismic excitation48.

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

Evolution of dynamic characteristics under progressive seismic loading: (a) natural frequency; (b) damping ratio.

Figure 3a further presents a comparison of the natural frequency evolution of anchored sheet-pile wall (ASPW) systems investigated in this study with the unanchored SPW systems reported by Ma et al.33. Owing to their identical geometric configurations and material properties, all slope models exhibited comparable initial natural frequencies, ranging from 15.28 to 15.42 Hz, indicating a consistent initial stiffness level among the four support systems. As the seismic intensity increased, however, pronounced differences in stiffness degradation emerged. At a peak ground acceleration (PGA) of 0.9 g, the natural frequencies of the ECC-ASPW, RC-ASPW, ECC-SPW, and RC-SPW systems decreased to 13.97 Hz, 12.94 Hz, 13.20 Hz, and 11.90 Hz, corresponding to stiffness reductions of 8.69%, 15.97%, 14.39%, and 22.12%, respectively. These results demonstrate that both material optimization and structural optimization play important roles in mitigating seismic-induced stiffness degradation.

A comparative analysis indicates that replacing RC with ECC reduced stiffness degradation by approximately 7.50%, while the introduction of anchor cables alleviated stiffness degradation by about 5.93%. At this seismic intensity level, the benefit provided by ECC was more pronounced than that of the anchors, suggesting that material ductility and crack control dominate the early-to-moderate stages of seismic damage accumulation. In contrast, the restraining effect of anchor cables had not yet been fully mobilized under this level of ground motion. Overall, the observed ranking of stiffness degradation (ECC-ASPW < ECC-SPW < RC-ASPW < RC-SPW) clearly confirms that the combined use of ECC material and anchor cables offers the most effective resistance against seismic-induced stiffness deterioration. The superior performance of ECC-supported systems can be attributed to two complementary mechanisms: (i) anchor cables provide additional lateral restraint and contribute to energy dissipation, thereby limiting excessive deformation and stiffness loss22; and (ii) the fiber-bridging and strain-hardening characteristics of ECC effectively suppress crack initiation and propagation, reducing cumulative structural damage and enhancing stiffness retention under strong shaking51.

Acceleration response

Seismic excitation induces inertial forces within the slope, which may trigger deformation accumulation and ultimately lead to failure of the retaining structure. Figure 4 compares the acceleration histories at key measurement points (A6 and A10) under PGA levels of 0.1 g and 1.0 g. Owing to the sinusoidal input motion, the measured accelerations also exhibited sinusoidal waveforms. Measurement point A10, located within the “bedrock” zone, recorded peak accelerations close to the input amplitude, whereas A6, located within the landslide mass, recorded larger amplitudes, demonstrating stronger seismic responses within the sliding mass and an upward amplification with elevation. At PGA = 0.1 g, the peak accelerations at A6 and A10 were similar for the ECC- and RC-supported slopes, reflecting their comparable elastic modulus and density while both slopes remained in the elastic stage. However, at PGA = 1.0 g, the peak accelerations of the RC-supported slope exceeded those of the ECC-supported slope, with the discrepancy more pronounced at upper slope locations (e.g., A6). This suggests that acceleration amplification is more sensitive to elevation for the RC-supported slope.

Fig. 4
Fig. 4The alternative text for this image may have been generated using AI.
Full size image

Comparison of acceleration time histories for ECC- and RC-ASPW supported slopes at measuring points A6 and A10 under different PGA levels: (a) A6, PGA = 0.1 g; (b) A10, PGA = 0.1 g; (c) A6, PGA = 1.0 g; (d) A10, PGA = 1.0 g.

Figure 5a presents the variation of peak acceleration with elevation for both slope models. H1 represents the model height, and h1 denotes the installation height of the acceleration sensor. A nonlinear increase in peak acceleration was observed for both systems, with the maximum occurring near the slope crest. This trend is attributed to the superficial dynamic effect at the slope surface, where seismic waves undergo reflection and superposition at the free surface, thereby amplifying the motion of surface soil52. As seismic intensity increased, accumulated slope damage reduced overall structural integrity and caused shallow overburden failure, decreasing the burial depth of sensors and further enhancing near-surface amplification, thereby strengthening the nonlinear elevation-dependent growth of peak acceleration. Under low-level excitations, the acceleration responses of both supporting systems were nearly identical. At higher PGA levels, however, the RC-supported slope exhibited larger peak accelerations, and the difference grew progressively with increasing seismic intensity. This indicates that ECC-ASPW structures possess superior energy dissipation capacity, which effectively suppresses the amplification of seismic motion within the slope29.

Fig. 5
Fig. 5The alternative text for this image may have been generated using AI.
Full size image

Comparison of acceleration responses along elevation for ECC- and RC-ASPW supported slopes: (a) peak acceleration; (b) acceleration amplification factor (AAF).

Figure 5b illustrates the distribution of the acceleration amplification factor (AAF) with elevation. The maximum AAF occurred at the slope crest and increased notably with higher PGA. Across all seismic intensities, the RC-supported slope exhibited higher AAF values compared with the ECC-supported slope. This behavior can be attributed to several factors: (1) ECC materials exhibit fuller hysteresis loops in the elastoplastic stage, providing greater energy dissipation; (2) the ECC-supported slope experiences less structural damage and maintains higher integrity, reducing near-surface nonlinear amplification; (3) seismic waves propagate more uniformly through the ECC-supported slope, minimizing superposition and reflection effects that contribute to acceleration amplification. Overall, ECC-ASPW supporting systems effectively suppress seismic amplification within the sliding mass, thereby offering superior seismic resistance for high-steep slopes.

Figure 6 compares the peak acceleration distributions along elevation for slopes supported by the sheet-pile wall (SPW, dashed lines) and the anchored sheet-pile wall (ASPW, solid lines) in this study. Regardless of whether RC or ECC was used, the presence of anchors exerted limited influence on peak acceleration under low seismic input (PGA ≤ 0.3 g), as the anchor force had not yet been fully mobilized. As the input seismic intensity increased, the restraining and energy-dissipating effects of the anchors became progressively more significant. Consequently, the peak accelerations of the ASPW-supported slopes became lower than those of the SPW-supported slopes. Notably, the influence of anchors was minor at lower elevations but increased with height, reaching the greatest reduction at the slope crest. This amplification suppression is attributed to the anchors enhancing the pile-soil interaction, limiting the lateral vibration of the upper slope, and effectively mitigating the near-surface dynamic amplification effect. These observations underscore the critical role of anchors in improving the seismic stability of steep slopes.

Fig. 6
Fig. 6The alternative text for this image may have been generated using AI.
Full size image

Comparison of peak acceleration responses of ASPW- and SPW-supported slopes: (a) RC systems; (b) ECC systems.

The peak acceleration amplification factor (AAF) was defined as the ratio of the peak acceleration at each measuring point to the peak ground acceleration of A11 at the shaking Tables18,53,54. Figure 7 illustrates the comparison of AAF along elevation for the two supporting systems. The results demonstrate that the presence of anchors alters the AAF distribution. Across all elevations, the AAF of the ASPW-supported slope is lower than that of the SPW-supported slope, and this difference becomes more pronounced at higher locations and under stronger seismic input. With increasing elevation and seismic demand, the additional confinement provided by anchors effectively suppresses strain concentration and reduces wave amplification, thereby lowering the AAF of the upper slope. Overall, the anchors not only enhance the structural stiffness and energy dissipation capacity of the sheet-pile wall system but also play a crucial role in reducing the seismic amplification in the upper slope, highlighting their essential contribution to seismic slope stabilization16,34.

Fig. 7
Fig. 7The alternative text for this image may have been generated using AI.
Full size image

Comparison of AAF for ASPW- and SPW-supported slopes: (a) RC systems; (b) ECC systems.

Dynamic earth pressure

Figure 8 presents the distribution of peak dynamic earth pressure along the height of the four types of support systems (ASPW and SPW; ECC and RC) under increasing seismic input. Before each seismic excitation, the soil pressure sensors were reset to zero, measuring only the dynamic earth pressure changes during a single loading process without accounting for static earth pressure effects13. For the anchored sheet-pile wall (ASPW), the dynamic earth pressure consistently increased with elevation, with the maximum occurring at the pile head, forming a distinct inverted triangular pattern16,55. Under low seismic excitation, the sliding tendency of the slope is negligible, and the differences in material properties combined with the free-surface and whiplash effects lead to intensified relative motion between the upper soil and the structure, thereby producing larger earth pressures at higher elevations56. Notably, even under strong seismic input (PGA = 0.9 g), the ASPW maintained this distribution pattern, indicating that the anchors effectively preserved the structural integrity and prevented changes in the load-transfer mechanism, unlike SPW systems which developed altered pressure patterns under intense seismic.

Fig. 8
Fig. 8The alternative text for this image may have been generated using AI.
Full size image

Comparison of Peak dynamic earth pressure distribution along wall height for ASPW- and SPW-supported slopes: (a) RC systems at PGA = 0.1–0.3 g; (b) RC systems at PGA = 0.7–0.9 g; (c) ECC systems at PGA = 0.1–0.3 g; (d) ECC systems at PGA = 0.7–0.9 g.

Material properties also played an important role. Although the ECC-ASPW and RC-ASPW exhibited comparable peak pressures at the pile head, their distribution along depth differed. For RC-ASPW, significant dynamic earth pressure was detected only above approximately 0.45 H₂, whereas ECC-ASPW exhibited obvious pressure responses above 0.2 H₂, indicating a higher resultant force application point. The underlying mechanism is twofold: (1) The higher stiffness of RC limits relative deformation in the lower and middle portions, reducing earth pressure mobilization; (2) The greater deformability and strain capacity of ECC enables more continuous soil-structure interaction over a larger depth, resulting in a broader distribution of dynamic earth pressure.

A comparison with the SPW results from Ma et al.33 further highlights the role of anchors. When PGA ≤ 0.3 g, the ASPW and SPW exhibited similar inverted triangular earth pressure patterns; however, the peak pressures of ASPW were lower, indicating that anchors enhanced overall slope stability and reduced soil-structure interaction forces16. As the input intensity increased, ASPW systems showed a proportional increase in dynamic earth pressure but maintained their distribution pattern. In contrast, SPW systems experienced structural damage under strong excitation, leading to a transition from an inverted triangular pattern to a Coulomb-like double-peak distribution. This divergence confirms that anchors not only increase the stiffness and energy dissipation capacity of the sheet-pile wall system but also delay the development of soil arching-dominated failure mechanisms57,58.

Bending moment of anti-sliding pile

Figure 9 presents the distribution of dynamic bending moments along the depth of anti-sliding piles in the ASPW supporting system under various seismic input levels. The maximum bending moment consistently occurred at the pile base, and the bending moment profile shifted toward the active earth pressure side with increasing seismic intensity, indicating progressive nonlinear deformation of the pile-soil system. However, a distinctive feature of ASPW is the emergence of a clear moment reversal at the anchor connection, accompanied by a significant reduction of the base moment compared with SPW. This demonstrates that the anchor provides a lateral restraint, transforming the pile behavior from a classical cantilever subjected to triangular earth pressure (SPW) into a partially restrained beam (ASPW), thereby redistributing the internal forces and effectively reducing moment concentration at the base.

Fig. 9
Fig. 9The alternative text for this image may have been generated using AI.
Full size image

Comparison of peak dynamic bending moment distribution of ASPW- and SPW-supported slopes: (a) RC systems; (b) ECC systems.

Material-dependent differences also became pronounced as the input PGA increased. At low seismic intensities (PGA ≤ 0.3 g), the moment profiles of ECC-ASPW and RC-ASPW were nearly identical due to their similar stiffness and strength in the elastic range. As seismic intensity increased, however, the RC-ASPW exhibited larger bending moments than its ECC counterpart. This difference can be attributed to several mechanisms: (i) stronger acceleration amplification in RC-supported slopes, which elevates both the magnitude of dynamic earth pressure and the height of its resultant force; (ii) lower ductility of RC, which leads to faster development of plastic hinges and greater concentration of bending demand at the pile base and anchor junction; (iii) superior energy dissipation and crack control of ECC, which delays stiffness degradation and mitigates bending moment growth under strong ground motion59.

Comparison with the SPW results further highlights the role of anchors: the ASPW system redistributed internal forces more effectively, reduced the bending moment demand at the fixed base, and postponed structural failure. The inherent ductility and strain-hardening characteristics of ECC amplified this beneficial effect, enabling the ECC-ASPW system to maintain lower bending moments and greater seismic resilience under high-intensity shaking.

Lateral displacement of support structures

The lateral displacement of the support structure is a critical parameter for evaluating system stability and serves as a key control indicator in the performance-based design framework60. Figure 10 presents the residual displacements of the anchored sheet-pile wall (ASPW) systems under varying seismic intensities. In both ECC-ASPW and RC-ASPW systems, the deformation pattern is characteristic of a laterally loaded cantilever: negligible displacement occurs within the embedded zone, and the maximum displacement develops at the pile head. Under low seismic excitation, the support structures remained largely elastic, resulting in minimal residual deformation. As the input PGA increases, however, accumulated dynamic earth pressure and the transition to an elastic-plastic state led to a pronounced increase in residual displacement. It is worth noting that the RC-ASPW system exhibited larger residual displacements than the ECC-ASPW system, with the ECC pile head displacement only 69.3% of that of the RC pile at PGA = 1.2 g. This difference arises primarily from two mechanisms: (i) The strain-hardening behaviour and fiber-bridging effect of ECC, which effectively suppress crack propagation and delay stiffness degradation48; (ii) Stronger acceleration amplification in RC-supported slopes, which generates larger dynamic earth pressures and causes more severe cumulative structural damage.

Fig. 10
Fig. 10The alternative text for this image may have been generated using AI.
Full size image

Comparison of residual displacement of ECC- and RC-ASPW supporting structures: (a) PGA = 0.1 ~ 0.6 g; (b) PGA = 0.7 ~ 1.2 g.

Figure 11 compares the residual displacement responses of ASPW systems with the unanchored sheet-pile wall (SPW) systems. Regardless of material type, the ASPW system exhibited smaller residual displacements than the SPW system, and the discrepancy increased with seismic intensity. For example, at PGA = 0.9 g, the anchor reduced the peak residual displacement from 32.9 mm to 7.05 mm in RC-SPW, and from 19.39 mm to 5.29 mm in ECC-SPW. This confirms that anchors provide essential lateral restraint, altering the load-transfer mechanism and greatly reducing deformation demand. Material improvements also contribute to deformation reduction. Replacing RC with ECC reduced the peak residual displacement from 32.9 mm to 19.39 mm in the SPW structure and from 7.05 mm to 5.29 mm in the ASPW structure. Overall, the results indicate that structural enhancement (adding anchors) has the most significant effect on displacement reduction, by modifying the deformation mode15. Material enhancement (using ECC) is particularly effective under strong shaking, due to its superior energy-dissipation capability26.

Fig. 11
Fig. 11The alternative text for this image may have been generated using AI.
Full size image

Residual lateral displacement of ASPW and SPW systems: (a) RC structures at PGA = 0.1 ~ 0.3 g; (b) RC structures at PGA = 0.7 ~ 0.9 g; (c) ECC structures at PGA = 0.1 ~ 0.3 g; (d) ECC structures at PGA = 0.7 ~ 0.9 g.

Figure 12 illustrates that pile-head residual displacement increased nonlinearly with seismic intensity. The overall trend follows: RC-SPW > ECC-SPW > RC-ASPW > ECC-ASPW. These results demonstrated that both the anchoring system and ECC material enhance displacement control, with the anchor-induced lateral restraint contributing the largest improvement, while the superior seismic resilience of ECC becomes increasingly evident under high-intensity ground motion.

Fig. 12
Fig. 12The alternative text for this image may have been generated using AI.
Full size image

Evolution of residual displacement at pile top with input peak acceleration.

Discussion

Improvement effect of ECC material and anchor cable on seismic performance of pile-sheet wall supported slopes

The results of shaking table tests reveal that both ECC materials and anchor cables can enhance the seismic performance of pile-sheet wall supported slopes, yet their improvement mechanisms differ. ECC contributes primarily through its superior ductility and crack-control capacity27,59. The fiber-bridging and strain-hardening behavior effectively suppress crack initiation and propagation, delay stiffness degradation, and maintain structural integrity under strong cyclic loading25,51. These material characteristics account for the slower reduction in natural frequency, smaller increase in damping ratio, and reduced acceleration amplification observed in ECC-supported systems. As a result, ECC piles exhibit lower dynamic earth pressures at the upper slope, smaller bending moments, and reduced residual displacements compared with RC counterparts, especially under strong seismic excitation28,29.

In contrast, the improvement provided by anchor cables arises from modification of the load-transfer mechanism of the supporting system22. By introducing a lateral restraint above the pile mid-height, the anchored configuration alters the bending moment distribution, generating a reverse moment near the anchorage and reducing the maximum bending demand at the pile base23. This geometric constraint also increases the global stiffness of the system, limits relative soil-pile movement, andsuppresses the development of dynamic earth pressure and surface amplification effects15. The difference between anchored and unanchored systems becomes more pronounced with increasing seismic intensity, indicating the efficiency of anchor cables in preventing slope rocking and controlling deformation7,17.

When ECC material and anchor cables are used in combination, their synergistic effects yield the best overall performance, showing the lowest acceleration response, most stable dynamic characteristics, and smallest residual deformation. Quantitative comparisons in Table 2 illustrate this synergy. At point A6 under PGA = 0.7 g, acceleration reductions relative to RC-SPW reach 5.3% (ECC-SPW) and 18.6% (RC-ASPW), whereas the combined ECC-ASPW achieves a larger 29.4% reduction. A consistent advantage is also observed in natural-frequency retention under both 0.3 g and 0.7 g excitations. Mechanically, anchors primarily reduce global deformation demand by redistributing bending and restraining kinematics, while ECC enhances local deformation capacity by delaying cracking and stiffness loss. Anchorage-induced demand reduction enables more effective mobilization of ECC ductility under strong shaking, resulting in higher effective stiffness and improved deformation compatibility. This interaction explains the superior seismic resilience of the ECC-ASPW system compared with configurations employing either material or structural optimization alone.

Table 2 Natural frequency and peak acceleration of four slope systems at PGA = 0.3 g/0.7 g.
Full size table

Advantages and limitations of ECC material and anchor cable in engineering application

From an engineering perspective, both ECC and anchor cables exhibit clear advantages but also possess inherent limitations. ECC offers enhanced ductility, superior damage tolerance, and improved durability, making it suitable for high-seismicity regions where traditional RC structures may experience brittle failure30,31,32. However, ECC construction requires strict quality control, specialized mixing procedures, and higher material cost, which may limit its large-scale adoption in certain projects51. Anchor cables provide rapid and effective structural stiffness enhancement and play a key role in restraining displacement, but they introduce additional construction complexity and long-term maintenance concerns, particularly regarding corrosion protection and prestress loss in harsh mountain environments20,22. Therefore, while both measures improve seismic performance, their practical implementation should consider cost, durability requirements, and site constraints.

Study limitations and future prospects

Several limitations of this study should be acknowledged. First, the shaking table results are based on single representative tests for each configuration. Due to soil heterogeneity and experimental constraints, replicate tests and statistical indicators were not available. Accordingly, the findings should be interpreted as trend-based and mechanism-oriented, rather than statistically definitive. Future studies should include repeated tests to quantify uncertainty and variability. Second, the seismic input was limited to a single 5 Hz sinusoidal wave, which does not represent the broadband and irregular nature of real earthquakes. While this controlled input facilitated direct comparison between ECC/RC and anchored/unanchored systems, the conclusions are conditional on the adopted excitation. The use of recorded, multi-frequency ground motions is recommended in future work. Third, scaling and modeling limitations inherent to 1 g tests remain. The sand-clay mixture cannot fully reproduce stress-dependent soil behavior, and the low-strength ECC, although preserving strain-hardening and crack-control characteristics, cannot achieve exact material similitude. In addition, boundary effects from the rigid model box may influence absolute response amplitudes. These factors limit direct quantitative extrapolation to prototype scale. Overall, this study primarily reveals comparative seismic response trends and underlying mechanisms.

Future research should therefore explore larger-scale or full-scale tests to reduce scaling effects, incorporate real and multi-directional earthquake records to improve applicability, and investigate the influence of anchor configuration, prestress level, and connection details on overall system performance. Furthermore, the long-term behavior of ECC and anchored systems, including environmental durability, fatigue resistance, and lifecycle performance, should be systematically evaluated to support their broader engineering application.

Conclusions

This study conducted two shaking table tests to systematically evaluate the seismic performance of high-steep slopes supported by ECC and RC anchored sheet-pile wall (ASPW) systems, with comparisons to previously tested unanchored SPW systems. The dynamic response mechanisms, including natural frequency evolution, acceleration response, dynamic earth pressure, bending moment distribution, and residual deformation, were comprehensively analyzed. The main conclusions are summarized as follows:

  1. (1)

    Replacing RC with ECC mitigates stiffness degradation, cracking damage, and residual deformation of supported slopes under seismic loading, owing to ECC’s superior crack control and energy dissipation capacities, and demonstrating the effectiveness of material optimization in enhancing seismic resilience.

  2. (2)

    Compared with SPW systems, ASPW-supported slopes showed notably lower peak acceleration, acceleration amplification factors, and residual displacements. The lateral restraint provided by anchors mitigated surface amplification effects and stabilized the overall dynamic response.

  3. (3)

    ECC-ASPW systems demonstrated the smallest acceleration response, bending moments, dynamic soil pressures, and residual deformations compared with RC-ASPW systems. The anchor-induced counteracting moment near the anchorage point lowered the maximum bending moment at the pile base, thereby enhancing structural safety margins.

  4. (4)

    While ECC and RC systems displayed similar responses under weak seismic intensity, RC structures exhibited rapidly increasing acceleration, bending moments, and permanent deformation at higher intensities. In contrast, ECC structures maintained better ductility and lower irreversible deformation, highlighting their superior performance under high-intensity earthquakes.