Investigation of Geotechnical Seismic Isolation Systems Based on Recycled Tire Rubber–Sand Mixtures

Abstract

In geotechnical earthquake engineering, enhancing the seismic properties of foundation soil to modify the characteristics of earthquake waves transmitted to structures presents a viable solution. This study investigates the effect of placing an isolation layer, composed of a mixture of recycled tire rubber and sand, beneath structures to mitigate seismic forces acting on buildings situated on soil layers with high amplification potential. In other words, the role of a soil layer functioning as a seismic isolator is examined. To achieve this objective, the seismic behavior of building-type structures is analyzed through numerical simulations, supplemented by laboratory experiments available in the literature. The numerical analyses are performed in the frequency domain using the finite element method within a one-dimensional (1D) framework. To validate the feasibility of the proposed isolation layer based on parametric analysis results, comparisons are made with laboratory tests available. In the literature, seismic isolation applications with thicknesses ranging from 1 to 3 m resulted in reductions of 6.8% to 16.17% in response spectral accelerations measured at the surface, while improvements in Fourier amplitude ratios ranged between 12.03% and 13.98%. This approach aims to provide an economical and efficient solution for earthquake-resistant structures while simultaneously promoting sustainability by recycling waste tires, contributing both to environmental conservation and economic benefits.

1. Introduction

1.1. The Concept of Geotechnical Seismic Isolation

Conventional isolation systems are technically applicable to significant structures such as schools, hospitals, critical transportation facilities, and airports due to the challenges posed by their design, implementation complexity, and high expenditure requirements. Nevertheless, underestimating the potential destruction that earthquakes can inflict on standard buildings would be catastrophic negligence. Furthermore, earthquake-induced fatalities, injuries, and damage to infrastructure and superstructure present devastating repercussions for especially developing countries, which are often technologically and economically constrained and thus less able to absorb and rectify such losses [1]. Therefore, there is an emerging need for a cost-effective and practical geotechnical seismic isolation (GSI) system that can be employed in the construction of low- and mid-rise buildings.

Geotechnical Seismic Isolation (GSI) is a novel seismic isolation system that addresses humanity’s enduring challenge of earthquake resistance and structural reliability. It has been met with excitement and enthusiasm, particularly in countries that regard and experience earthquakes as existential threats, owing to its cost-effectiveness and simplicity of implementation [2]. The fundamental principles of GSI are derived from traditional isolation systems, which typically consist of flexible or sliding interfaces placed between the foundation and vertical load-bearing elements to decouple the structure–foundation interaction. To put it succinctly, these systems separate the lateral movement of the ground from the structure, thereby reducing earthquake-induced shear and torsional forces on the building, making it more resilient against seismic events and their consequent deterioration [3].

In friction-based systems, synthetic layers are customarily used as isolators. The primary objective of using these layers is to create a stable isolation layer where a fraction of the seismic energy generated by earthquakes is dissipated through friction between the two materials. In damping-based systems, one or more improvement techniques are applied beneath the foundation to enhance the damping capacity of the underlying soil, thereby creating a geotechnical isolation layer focused on damping. This layer absorbs a portion of the earthquake’s energy, thus reducing the amount of energy transmitted to the structure.

In synthetic liner ground isolation method, referred to as ground isolation, a smooth synthetic liner is placed within the soil at a depth beneath the foundations to prevent seismic energy from reaching the building or ground surface. This approach isolates the soil layer above the liner from the underlying soil layer [4].

In damping-type geotechnical seismic isolation, energy dissipation is one of the primary mechanisms for reducing seismic ground motions. Rubber is known for its excellent energy absorption capabilities; hence, it is widely used in applications such as vibration control and damping in automotive parts. Rubber and soil particles complement each other functionally. Compared to conventional soils, rubber-reinforced soil demonstrates a significant increase in shear strength [5] and, more importantly, a substantial enhancement in energy dissipation capability.

From a sustainability perspective, the reuse and recycling of waste tires are preferred over energy recovery. Tire fragments can be utilized in civil engineering applications such as road embankments, landslide stabilization, and as backfill material for retaining walls and bridge abutments.

In a study conducted by Tsang [6], numerical analyses were carried out to probe the use of a rubber–soil mixture as a seismic isolator for a 10-story building with a width of 40 m. In this study, a 10 m thick sand–rubber mixture was used as the foundation soil. The hypothetical setup used in this research serves as a representative model for the application of rubber–soil mixtures for geotechnical seismic isolation purposes.

1.2. Research on Geotechnical Seismic Isolation

Tsang [6] proposed a rubber and sand mixture comprising shredded vehicle tires (RSM) as seismic isolation material. This method, termed geotechnical seismic isolation (GSI), involves placing the mixture beneath the building foundation to mitigate lateral and vertical shaking during earthquakes, as demonstrated through experiments and analyses. Xiong and Li [2] claimed that the most critical parameters of GSI method proposed by Tsang’s [7] were the structure’s weight and soil layer thickness. They determined that a mixture containing 35% rubber content provided sufficient isolation but noted that higher proportions could induce stability issues in the superstructure. Wu et al. [8] examined the application of rubber–sand mixture (RSM) pads in GSI systems, demonstrating that RSM pads have high seismic energy dissipation capacity and can be an effective method for structural protection. Yin et al. [9] explored the potential that GSI technology propose in rural areas of China, finding that rubber–sand mixtures (RSM) offered limited isolation during minor earthquakes but achieve heightened effectiveness during major seismic events. Yıldız [10] assessed the seismic performance of a rubber–sand mixture (RSM) isolation layer placed beneath a 10-story reinforced concrete building model, showing that the RSM layer reduced accelerations experienced by the structure by up to 48% during shaking. Pitilakis et al. [11] investigated the impact of rubber–sand mixtures (RSM) on the seismic performance of concrete structures’ foundation layers, finding that these mixtures decreased base shear forces by up to 40%, particularly in high-rise buildings.

Nakhaei et al. [12] investigated the dynamic properties of soils by mixing granular soils with various proportions of granulated rubber. The results indicated that as the rubber content increased, the shear modulus decreased, while an increase in confining pressure led to an increase in the shear modulus. It was observed that in soil without rubber, the damping ratio increased as confining pressure decreased, whereas soil containing rubber exhibited the opposite behavior. Finally, new relationships were defined for maximum shear modulus (G_max) and normalized shear modulus (G/G_max) as a function of confining pressure and granulated rubber content that in this study, G represents the shear modulus at a given strain level, while G_max denotes the maximum shear modulus at very small strains, corresponding to the elastic behavior of the material before any significant nonlinearity occurs. Edil and Bosscher [5] analyzed the compressibility, strength, and hydraulic conductivity of shredded waste tires both as a standalone material and mixed with soil, aiming to use it as lightweight fill material. The study concluded that this material holds significant potential to address economic and technical challenges in applications such as road embankments and landfill cover materials. Feng and Sutter [13] found that granulated rubber–soil mixtures are effective in reducing vibration under seismic loads due to their high damping capacity. Their study demonstrated that rubber-inclusive mixtures positively affect the shear modulus. Edincliler and Yildiz [14] examined the use of waste tires as an additive in soils located in earthquake-prone areas. The study investigated the effects of tire particles in granular and fiber forms, produced through various processing techniques, on dynamic properties. Their findings established that the type of processing technique exerts a considerable influence on dynamic properties based on the rubber material’s particle size, aspect ratio, and content. In a study by Pistolas et al. [15], soil material properties were presented with G/G_max and D/D_max curves within the scope of seismic design and site seismic response analysis, emphasizing their noteworthiness in numerical analysis methods. The dynamic properties of granular soils mixed with granulated rubber were experimentally scrutinized in terms of G_max and D_max. The results reveal that the average particle size ratio of the mixture profoundly impacted the outcomes, and appropriate analytical expressions were proposed for engineering applications.

Yu et al. [16] conducted one-dimensional ground response analyses with the aid of the DeepSoil v.7.0 [17] and Cyclic 1D v.1.4 [18] programs at an offshore wind farm development site in the Taiwan Strait. This study employed equivalent linear and nonlinear analysis approaches, with the results subsequently compared to PLAXIS 2D 2015 [19]. The findings provided indispensable insights into peak ground acceleration, acceleration–time histories, and design response spectra. Niazi [20] studied soil and soil–structure interaction using PLAXIS 2D, DeepSoil, and EERA 2000 programs [21]. This study highlighted that PLAXIS 2D yields more reliable results than other methods, suggesting a more accurate observation of soil–structure interaction. Bildik et al. [22] investigated site performance post-soil improvement utilizing both experimental and numerical methods. Soil behavior under dynamic loads was analyzed through the use of DeepSoil and PLAXIS 2D, and it was found that soil improvement pronouncedly enhanced overall behavior. Silahtar [23] performed one-dimensional nonlinear ground response analyses of the Isparta plain using DeepSoil to assess ground response. The study evaluated the seismic performance of soil profiles in the study area and produced maps showing surface ground acceleration and spectral acceleration distribution. Timur [24] used the DeepSoil program to review amplification effects at sites within the Bayraklı region. Soil profiles at various sites were analyzed, revealing that amplification effects were distinctively associated with damping at lower energy levels. Jin et al. [25] investigated the dynamic properties between soil layers utilizing a shaking table test. The study conducted a one-dimensional ground response analysis with DeepSoil software, and the results were compared with experimental data to evaluate amplification differences and dynamic properties across soil layers. Hong and Kim [26] evaluated the performance of resilient-friction base isolation (R-FBI) systems for multi-story structures. Their analyses found that such isolation systems significantly reduce seismic acceleration and displacement responses, enhancing the earthquake resilience of buildings. Gajan and Saravanathiiban [27] analyzed energy dissipation in structural devices and foundation soil under seismic loading. Their findings demonstrated that foundation rocking could serve as an effective energy dissipation mechanism. This study highlights the role of rocking behavior in dissipating seismic energy. Darendeli [28] conducted extensive laboratory testing to develop a new family of normalized modulus reduction and material damping curves. The results enabled the creation of parametric models that can be used to predict dynamic response in different soil types. Hardin and Drnevich [29] developed empirical relationships and curves for shear modulus and damping ratio in soils. Their study showed that the shear modulus increases under pressure but significantly decreases at higher strains. The findings provide a critical resource for predicting shear modulus and damping in engineering applications. Brennan and Madabhushi [30] investigated seismic acceleration amplification at slope crests using centrifuge modeling. Their findings demonstrated that topographic amplification is a function of frequency and becomes particularly significant at higher loading frequencies. The study emphasized that topographic amplification can create substantial tensile stresses in structures, necessitating consideration in seismic design. Göktepe et al. [31] conducted numerical and experimental analyses to determine an appropriate geometric scaling coefficient for small-scale shaking table tests. The seismic behavior of mid-rise reinforced concrete buildings on silty sand soil was modeled using the 2D finite element method. The results indicated that an appropriately selected scale coefficient could represent real field conditions with sufficient accuracy.

Despite extensive research on geotechnical seismic isolation (GSI) methods, significant knowledge gaps remain regarding soil-structure interaction and the dynamic behavior of different materials. Das [32] provides a fundamental understanding of soil properties, water movement in soil, and stress-strain relationships, serving as a crucial reference for comprehending the mechanical properties of seismic isolation materials. Similarly, Wolf, J.P. [33] offers an in-depth analysis of soil-structure interaction under dynamic loads, providing critical insights for integrating GSI systems into engineering design. Meanwhile, Naeim, F. and Kelly, J.M. [34] explores the theoretical foundations and practical applications of seismic isolation systems, making it an essential guide for understanding the impact of GSI methods on structural safety. While these sources are not directly used in experimental studies, they are of immense importance in understanding geotechnical seismic isolation and dynamic soil behavior.

Despite the extensive research conducted on geotechnical seismic isolation (GSI) methods, significant gaps remain in our understanding of the dynamic behavior of rubber–sand mixtures (RSM) under varying confining pressures and earthquake excitations. Previous studies, such as those by Tsang [6] and Xiong [2], have demonstrated the feasibility of RSM in seismic isolation but have primarily focused on empirical approaches and numerical simulations without extensive experimental validation. Similarly, Nakhaei [12] investigated the influence of rubber inclusion on shear modulus and damping ratio but did not address its implications for seismic isolation efficiency across different site conditions.

This study fills this research gap by integrating both numerical and experimental analyses to evaluate the effectiveness of RSM layers as seismic isolators. Unlike previous works, which often assume fixed dynamic properties, this study incorporates the strain-dependent variations in the shear modulus and damping ratio, ensuring a more realistic representation of soil behavior under seismic loading. Furthermore, this research expands upon the work of Wu [8] by systematically analyzing the impact of isolation layer thickness, granular rubber content, and confining pressure on seismic response. The findings provide a deeper insight into the trade-offs between isolation performance and material stability, an aspect that has been underexplored in earlier studies.

This paper is structured into four sections to guide readers through the study. Section 1 introduces the background, objectives, and significance of the research. Section 2 outlines the materials and methods employed, including the numerical modeling techniques and parametric analyses conducted. Section 3 presents the results and discussions, focusing on the seismic performance of granular rubber–sand mixtures under various conditions. Lastly, Section 4 provides the conclusions and recommendations, emphasizing the practical applications and sustainability of the proposed isolation system.

2. Materials and Methods

In this study, the dynamic properties of granular rubber–sand mixtures were investigated through numerical analyses and laboratory experiments carried out by [12]. The laboratory component primarily relied on consolidated undrained cyclic triaxial tests to determine the shear modulus and damping ratio under various confining pressures. Specimens were prepared using well-graded sand mixed with granular rubber particles derived from recycled tires. The granular rubber content varied between 0%, 5%, and 10%, while confining pressures of 50, 100, 200, and 300 kPa were applied to assess the influence of overburden pressure on dynamic behavior [12].

A GDS Instruments cyclic triaxial testing system was employed for the experiment procedures. This system was equipped with an automated load frame capable of applying cyclic axial loads at a controlled frequency range of 0.1–10 Hz. The system recorded stress–strain responses using high-precision displacement transducers and pressure sensors, allowing for the derivation of modulus reduction and damping curves. The hysteresis loops obtained from cyclic loading were analyzed to compute the shear modulus and damping ratio, which were subsequently normalized to generate G/G_max and damping ratio curves shown in Figure 1.

The experimental data was subsequently used to calibrate numerical models developed in the DeepSoil software. The validation of the numerical framework was achieved through a comparative analysis of response spectra, Fourier amplitude ratios, and acceleration time histories. This integration of laboratory testing with numerical modeling enhances the reliability of the results and provides a more comprehensive understanding of the seismic isolation performance of granular rubber–sand mixtures.

In the dissection of most soil dynamics problems, determining the shear modulus and damping ratio based on shear strain amplitude is a critical parameter for defining dynamic properties or estimating stiffness and energy absorption ability. Soil–structure interaction response to dynamic loads, such as earthquakes, vibrating equipment foundations, explosive forces, and the response of marine structures to ocean waves, is exceptionally convoluted. Under dynamic loading, stress–strain behavior is represented by hysteresis loops. The slope of lines connecting the peaks of these loops indicate stiffness, while the area enclosed by the loops represents material damping [12].

In Figure 1, experimental results on damping ratio and shear modulus reduction were presented. The graph demonstrates the influence of varying rubber content (0%, 8%, 10%, and 14%) and confining pressures (50, 100, 200, and 300 kPa) on dynamic soil properties, such as damping ratio and shear strain amplitude. These experimental data complement the numerical results discussed in this study and validate the trends observed under seismic loading conditions.

In the research presented herein, the research conducted by A. Nakhaei and S.M. Marandi [12] on consolidated undrained cyclic triaxial testing was utilized to obtain the shear modulus and damping ratio of the isolation soil. As shown in

Table 1. Properties of the prepared samples.

Granular Rubber Content (%)	Moist Unit Weight (kN/m3)	Dry Unit Weight (kN/m3)	Void Ratio	Relative Density Dr (%)
0.0	22.0	20.64	0.259	91.47
5.0	20.84	19.42	0.253	91.96
10.0	19.47	17.96	0.266	91.14

, samples were mixed in various proportions with granular rubber of a specific particle size and specific granular soil. Models were introduced to estimate the maximum shear modulus (G_max) and normalized shear modulus (G/G_max) for different confining pressures (σ₃) and granular rubber percentages (R). This research inquiries into how the dynamic properties of soil vary with the supplement of granular rubber and how augmenting the rubber content influences soil stiffness and damping capacity.

Shear modulus–shear strain amplitude curves for soil–granular rubber mixtures are presented in Figure 2 for 50 and 100 kPa confining pressures and granular rubber content ratios of 0.0%, 5%, and 10%. The results indicate that, under constant confining pressure, the shear modulus diminishes as the percentage of granular rubber increases. This outcome can be anticipated by comparing the stiffness of rubber particles to that of soil grains. Addedly, replacing soil grains with granular rubber softens the mixture, resulting in a decrease in shear modulus. In this study, the type of rubber considered is granulated tire rubber derived from waste tires that were mechanically chopped through multiple chopping steps. The granulated rubber particles used in the experiments have nonspherical shapes and dimensions ranging from 0.15 to 9.5 mm. The specific gravity of the granulated rubber was found to be 1.1 at a temperature of 20 °C, following ASTM standards.

Damping ratio–shear strain amplitude curves for soil–granular rubber mixtures are presented in Figure 3 for confining pressures of 50 and 100 kPa, with granular rubber additions of 0.0%, 5%, and 10%. The results show that as the amount of rubber increases, the damping ratio decreases for both 50 and 100 kPa confining pressures. This difference can be attributed to the high elastic deformation capacity of rubber particles, leading to increased elastic deformation and a subsequent decline in damping ratio at lower confining pressures (50 and 100 kPa).

Damping ratio–shear strain amplitude curves for soil–granular rubber mixtures are presented in Figure 3 for granular rubber contents of 0.0%, 5%, and 10% and confining pressures of 50 and 100 kPa. In soils without rubber addition, an increase in confining pressure results in a decline in the damping ratio. This behavior can be explained by the increase in interparticle friction as confining pressure rises, leading to reduced plastic deformation. As a consequence, the width of the hysteresis loop decreases, causing a reduction in the damping ratio.

Figure 3 also shows that in soils with granular rubber, the damping ratio increases with higher confining pressure, while the opposite trend is observed in soils without rubber. This difference can be attributed to the compression of rubber particles under higher confining pressure, which reduces their elasticity; as a result, the sliding of soil particles over each other increases when deviator stress is applied. Consequently, plastic deformation increases, leading to a rise in the damping ratio.

Additionally, the results signify that with 5% and 10% rubber inclusions, not only does the damping ratio increase with confining pressure, but the effect of confining pressure on the damping ratio also intensifies with higher rubber content. This phenomenon can be explained by the increased plastic deformation under high confining pressures and greater elastic deformation at low confining pressures when higher proportions of rubber are used. Therefore, increasing rubber content leads to greater differences in damping ratios under diffirent confining pressures.

Figure 4, Figure 5 and Figure 6 depict the G/G_max values, represented by curves, obtained from laboratory tests on soils selected for the isolation layer. The results suggest that for a given percentage of rubber, G/G_max values increase as confining pressure rises, indicating an increase in soil stiffness with higher confining pressure.

3. Results and Discussions

3.1. Models’ Description

The results obtained are presented below in tables and graphs. For the analyses, the soil profile created using the soil parameters provided in

Table 2. Soil type and parameters used in the geotechnical model.

Soil Type	z (m)	γ (kN/m3)	Vs (m/s)	σ3 (kPa)
Soft Clay	20	16	200	160
Sand with 10% Rubber Content	1–3	17.96	152.14	100
Sand with 5% Rubber Content	1–3	19.415	196.09	100
Sand with 0% Rubber Content	1–3	20.64	243.15	100
Sand with 10% Rubber Content	1–3	17.96	120.61	50
Sand with 5% Rubber Content	1–3	19.415	155.45	50
Sand with 0% Rubber Content	1–3	20.64	192.77	50

.

The process of creating the DeepSoil model involves several sequential steps, as illustrated in Figure 7. These steps encompass defining the analysis type, setting soil properties, characterizing soil layers, selecting earthquake motions, configuring analysis settings, and ultimately executing the analysis to obtain results. This systematic approach ensures the comprehensive evaluation of soil behavior under seismic loading conditions, facilitating precise and reliable numerical simulations. The flowchart provides a clear visual representation of the methodology adopted in this study, enhancing the understanding of the modeling process.

In this model, a 20 m thick soil layer overlying rock (V_s = 1000 m/s and γ = 24 kN/m³) was assumed, and frequency domain solutions were performed applying the equivalent linear analysis method. Of the seven hypothetical soil types, soft clay represents the foundation soil, while the remaining six represent isolation layers with different confining pressures/depths and rubber content shown in Figure 8 and Figure 9.

Isolation soils modeled at depths of 1 m, 2 m, and 3 m with 0%, 5%, and 10% granular rubber content under confining pressures of 50 kPa and 100 kPa were analyzed and evaluated using DeepSoil software based on their performance characteristics.

3.2. Ground Motions Selection

After determining the soil models to be used for numerical analyses, four different earthquake records were chosen, as listed in

Table 3. The values obtained from the analysis of the earthquakes defined in the model with clay soil.

Earthquakes	Time (s)	Magnitude (MG)	Acceleration (g)	Arias Intensity (m/s)
ChiChi	59.00	7.60	1.57	4.13
Coyote	26.83	5.74	0.92	0.57
ImperialValley	63.73	6.50	1.16	3.76
Kocaeli	30.00	7.50	1.18	0.68

. For each aforementioned soil description, analyses were conducted without applying isolation. The resulting earthquake duration, magnitude, peak ground acceleration measured at the surface, and peak Arias intensity values measured at the surface are provided in Table 3.

The earthquakes listed in Table 3 were selected from well-documented seismic events widely recognized in the literature as reference ground motions by various analytical software programs. The modeling and analyses conducted in this study utilized data obtained from earthquake records available within the DeepSoil software.

Data obtained from analytical analyses obtained with DeepSoil software are presented in graphs and tables:

Ground response spectra for the isolation layer containing 10% granular rubber and subjected to a confining pressure of 100 kPa, applied at thicknesses of 1, 2, and 3 m are shown in Figure 10, with a summary of these results provided in

Table 4. Measured maximum response spectrum acceleration in the condition where the isolation layer is composed of sand containing 10% rubber content (σ₀ = 100 kPa).

Shear Wave Velocity of Clay Soil Layer (m/s)	200
Profile Depth (m)	20	19		18		17
Isolation (m)	0	1		2		3
Shear Wave Velocity of Isolation Layer (m/s)	-	152.14		152.14		152.14
Initial Damping Ratio (%)	-	9.91		9.91		9.91
Natural Frequency (Hz)	2.5	2.461		2.424		2.387
Natural Period (s)	0.4	0.4063		0.4126		0.4189
Earthquake	EL PSA (g)	%10–100 kPa 1 m PSA (g)	Damping Ratio (%)	%10–100 kPa 2 m PSA (g)	Damping Ratio (%)	%10–100 kPa 3 m PSA (g)	Damping Ratio (%)
ChiChi	1.574	1.395	−11.37%	1.380	−12.33%	1.377	−12.52%
Coyote	0.915	0.869	−5.03%	0.846	−7.54%	0.825	−9.84%
Imp. Valley	1.162	1.116	−3.96%	1.083	−6.80%	1.052	−9.47%
Kocaeli	1.176	1.067	−9.27%	1.067	−9.27%	1.055	−10.29%
Mean Value			−7.41%		−8.98%		−10.53%

.

The ground response spectra for the isolation layer containing 10% granular rubber and subjected to a confining pressure of 50 kPa, applied at thicknesses of 1, 2, and 3 m are shown in Figure 11, with a summary of these results provided in

Table 5. Measured maximum response spectrum acceleration in the condition where the isolation layer is composed of sand containing 10% rubber content (σ₀ = 50 kPa).

Shear Wave Velocity of Clay Soil Layer (m/s)	200
Profile Depth (m)	20	19		18		17
Isolation (m)	0	1		2		3
Shear Wave Velocity of Isolation Layer (m/s)	-	120.61		120.61		120.61
Initial Damping Ratio (%)	-	4.07		4.07		4.07
Natural Frequency (Hz)	2.5	2.42		2.346		2.275
Natural Period (s)	0.4	0.4132		0.4263		0.4395
Earthquake	EL PSA (g)	%10–50 kPa 1 m PSA (g)	Damping Ratio (%)	%10–50 kPa 2 m PSA (g)	Damping Ratio (%)	%10–50 kPa 3 m PSA (g)	Damping Ratio (%)
ChiChi	1.574	1.397	−11.25%	1.388	−11.82%	1.399	−11.12%
Coyote	0.915	0.872	−4.70%	0.854	−6.67%	0.842	−7.98%
Imp. Valley	1.162	1.119	−3.70%	1.092	−6.02%	1.072	−7.73%
Kocaeli	1.176	1.087	−7.57%	1.141	−2.98%	1.205	2.47%
Mean Value			−6.80%		−6.87%		−6.09%

.

Ground response spectra for the isolation layer containing 5% granular rubber and subjected to a confining pressure of 100 kPa, applied at thicknesses of 1, 2, and 3 m are shown in Figure 12, with a summary of these results provided in

Table 6. Measured maximum response spectrum acceleration in the condition where the isolation layer is composed of sand containing 5% rubber content (σ₀ = 100 kPa).

Shear Wave Velocity of Clay Soil Layer (m/s)	200
Profile Depth (m)	20	19		18		17
Isolation (m)	0	1		2		3
Shear Wave Velocity of Isolation Layer (m/s)	-	196.09		196.09		196.09
Initial Damping Ratio (%)	-	6.59		6.59		6.59
Natural Frequency (Hz)	2.5	2.498		2.495		2.493
Natural Period (s)	0.4	0.4004		0.4008		0.4012
Earthquake	EL PSA (g)	%5–100 kPa 1 m PSA (g)	Damping Ratio (%)	%5–100 kPa 2 m PSA (g)	Damping Ratio (%)	%5–100 kPa 3 m PSA (g)	Damping Ratio (%)
ChiChi	1.574	1.384	−12.07%	1.374	−12.71%	1.370	−12.96%
Coyote	0.915	0.857	−6.34%	0.819	−10.49%	0.789	−13.77%
Imp. Valley	1.162	1.100	−5.34%	1.045	−10.07%	1.005	−13.51%
Kocaeli	1.176	1.049	−10.80%	1.009	−14.20%	0.975	−17.09%
Mean Value			−8.64%		−11.87%		−14.33%

.

Ground response spectra for the isolation layer containing 5% granular rubber and subjected to a confining pressure of 50 kPa, applied at thicknesses of 1, 2, and 3 m are shown in Figure 13, with a summary of these results provided in

Table 7. Measured maximum response spectrum acceleration in the condition where the isolation layer is composed of sand containing 5% rubber content (σ₀ = 50 kPa).

Shear Wave Velocity of Clay Soil Layer (m/s)	200
Profile Depth (m)	20	19		18		17
Isolation (m)	0	1		2		3
Shear Wave Velocity of Isolation Layer (m/s)	-	155.45		155.45		155.45
Initial Damping Ratio (%)	-	5.80		5.80		5.80
Natural Frequency (Hz)	2.5	2.465		2.43		2.397
Natural Period (s)	0.4	0.4057		0.4115		0.4175
Earthquake	EL PSA (g)	%5–50 kPa 1 m PSA (g)	Damping Ratio (%)	%5–50 kPa 2 m PSA (g)	Damping Ratio (%)	%5–50 kPa 3 m PSA (g)	Damping Ratio (%)
ChiChi	1.574	1.385	−12.01%	1.381	−12.26%	1.378	−12.45%
Coyote	0.915	0.859	−6.12%	0.825	−9.84%	0.797	−12.90%
Imp. Valley	1.162	1.101	−5.25%	1.054	−9.29%	1.013	−12.82%
Kocaeli	1.176	1.060	−9.86%	1.048	−10.88%	1.039	−11.65%
Mean Value			−8.31%		−10.57%		−12.46%

.

Ground response spectra for the isolation layer containing 0% granular rubber and subjected to a confining pressure of 100 kPa, applied at thicknesses of 1, 2, and 3 m are shown in Figure 14, with a summary of these results provided in

Table 8. Measured maximum response spectrum acceleration in the condition where the isolation layer is composed of sand containing 0% rubber content (σ₀ = 100 kPa).

Shear Wave Velocity of Clay Soil Layer (m/s)	200
Profile Depth (m)	20	19		18		17
Isolation (m)	0	1		2		3
Shear Wave Velocity of Isolation Layer (m/s)	-	243.15		243.15		243.15
Initial Damping Ratio (%)	-	9.38		9.38		9.38
Natural Frequency (Hz)	2.5	2.522		2.545		2.568
Natural Period (s)	0.4	0.3965		0.3929		0.3894
Earthquake	EL PSA (g)	%0–100 kPa 1 m PSA (g)	Damping Ratio (%)	%0–100 kPa 2 m PSA (g)	Damping Ratio (%)	%0–100 kPa 3 m PSA (g)	Damping Ratio (%)
ChiChi	1.574	1.382	−12.20%	1.378	−12.45%	1.366	−13.21%
Coyote	0.915	0.848	−7.32%	0.804	−12.13%	0.767	−16.17%
Imp. Valley	1.162	1.087	−6.45%	1.026	−11.70%	0.996	−14.29%
Kocaeli	1.176	1.038	−11.73%	0.983	−16.41%	0.929	−21.00%
Mean Value			−9.43%		−13.17%		−16.17%

.

Ground response spectra for the isolation layer containing 0% granular rubber and subjected to a confining pressure of 50 kPa, applied at thicknesses of 1, 2, and 3 m are shown in Figure 15, with a summary of these results provided in

Table 9. Measured maximum response spectrum acceleration in the condition where the isolation layer is composed of sand containing 0% rubber content (σ₀ = 50 kPa).

Shear Wave Velocity of Clay Soil Layer (m/s)	200
Profile Depth (m)	20	19		18		17
Isolation (m)	0	1		2		3
Shear Wave Velocity of Isolation Layer (m/s)	-	192.77		192.77		192.77
Initial Damping Ratio (%)	-	6.23		6.23		6.23
Natural Frequency (Hz)	2.5	2.495		2.491		2.486
Natural Period (s)	0.4	0.4008		0.4015		0.4023
Earthquake	EL PSA (g)	%0–50 kPa 1 m PSA (g)	Damping Ratio (%)	%0–50 kPa 2 m PSA (g)	Damping Ratio (%)	%0–50 kPa 3 m PSA (g)	Damping Ratio (%)
ChiChi	1.574	1.382	−12.20%	1.373	−12.77%	1.373	−12.77%
Coyote	0.915	0.849	−7.21%	0.793	−13.33%	0.773	−15.52%
Imp. Valley	1.162	1.088	−6.37%	1.007	−13.34%	1.001	−13.86%
Kocaeli	1.176	1.043	−11.31%	0.985	−16.24%	0.968	−17.69%
Mean Value			−9.27%		−13.92%		−14.96%

.

The Fourier amplitude ratio for the isolation layer containing 10% granular rubber and subjected to a confining pressure of 100 kPa, applied at thicknesses of 1, 2, and 3 m is shown in Figure 16, with a summary of these results provided in

Table 10. Table of dominant frequencies based on the measured Fourier amplitude ratio for the isolation layer composed of sand containing 10% rubber content (σ₀ = 100 kPa).

Shear Wave Velocity of Clay Soil Layer (m/s)	200
Profile Depth (m)	20	19		18		17
Isolation (m)	0	1		2		3
Shear Wave Velocity of Isolation Layer (m/s)	-	152.14		152.14		152.14
Initial Damping Ratio (%)	-	9.91		9.91		9.91
Natural Frequency (Hz)	2.5	2.461		2.424		2.387
Natural Period (s)	0.4	0.4063		0.4126		0.4189
Earthquake	EL (Hz)	%10–100 kPa 1 m Amp. Ratio (Hz)	Damping Ratio (%)	%10–100 kPa 2 m Amp. Ratio (Hz)	Damping Ratio (%)	%10–100 kPa 3 m Amp. Ratio (Hz)	Damping Ratio (%)
ChiChi	3.702	3.212	−13.24%	3.199	−13.59%	3.179	−14.13%
Coyote	4.056	3.567	−12.06%	3.554	−12.38%	3.534	−12.87%
Imp. Valley	3.873	3.383	−12.65%	3.364	−13.14%	3.343	−13.68%
Kocaeli	4.200	3.745	−10.83%	3.770	−10.24%	3.790	−9.76%
Mean Value			−12.19%		−12.34%		−12.61%

.

The Fourier amplitude ratio for the isolation layer containing 10% granular rubber and subjected to a confining pressure of 50 kPa, applied at thicknesses of 1, 2, and 3 m is shown in Figure 17, with a summary of these results provided in

Table 11. Table of dominant frequencies based on the measured Fourier amplitude ratio for the isolation layer composed of sand containing 10% rubber content (σ₀ = 50 kPa).

Shear Wave Velocity of Clay Soil Layer (m/s)	200
Profile Depth (m)	20	19		18		17
Isolation (m)	0	1		2		3
Shear Wave Velocity of Isolation Layer (m/s)	-	120.61		120.61		120.61
Initial Damping Ratio (%)	-	4.07		4.07		4.07
Natural Frequency (Hz)	2.5	2.42		2.346		2.275
Natural Period (s)	0.4	0.4132		0.4263		0.4395
Earthquake	EL (Hz)	%10–50 kPa 1 m Amp. Ratio (Hz)	Damping Ratio (%)	%10–50 kPa 2 m Amp. Ratio (Hz)	Damping Ratio (%)	%10–50 kPa 3 m Amp. Ratio (Hz)	Damping Ratio (%)
ChiChi	3.702	3.217	−13.10%	3.220	−13.02%	3.232	−12.70%
Coyote	4.056	3.574	−11.88%	3.580	−11.74%	3.603	−11.17%
Imp. Valley	3.873	3.389	−12.50%	3.388	−12.52%	3.403	−12.14%
Kocaeli	4.200	3.753	−10.64%	3.808	−9.33%	3.899	−7.17%
Mean Value			−12.03%		−11.65%		−10.79%

.

The Fourier amplitude ratio for the isolation layer containing 5% granular rubber and subjected to a confining pressure of 100 kPa, applied at thicknesses of 1, 2, and 3 m is shown in Figure 18, with a summary of these results provided in

Table 12. Table of dominant frequencies based on the measured Fourier amplitude ratio for the isolation layer composed of sand containing 5% rubber content (σ₀ = 100 kPa).

Shear Wave Velocity of Clay Soil Layer (m/s)	200
Profile Depth (m)	20	19		18		17
Isolation (m)	0	1		2		3
Shear Wave Velocity of Isolation Layer (m/s)	-	196.09		196.09		196.09
Initial Damping Ratio (%)	-	6.59		6.59		6.59
Natural Frequency (Hz)	2.5	2.498		2.495		2.493
Natural Period (s)	0.4	0.4004		0.4008		0.4012
Earthquake	EL (Hz)	%5–100 kPa 1 m Amp. Ratio (Hz)	Damping Ratio (%)	%5–100 kPa 2 m Amp. Ratio (Hz)	Damping Ratio (%)	%5–100 kPa 3 m Amp. Ratio (Hz)	Damping Ratio (%)
ChiChi	3.702	3.207	−13.37%	3.166	−14.48%	3.141	−15.15%
Coyote	4.056	3.562	−12.18%	3.527	−13.04%	3.506	−13.56%
Imp. Valley	3.873	3.376	−12.83%	3.334	−13.92%	3.313	−14.46%
Kocaeli	4.200	3.744	−10.86%	3.747	−10.79%	3.755	−10.60%
Mean Value			−12.31%		−13.06%		−13.44%

.

The Fourier amplitude ratio for the isolation layer containing 5% granular rubber and subjected to a confining pressure of 50 kPa, applied at thicknesses of 1, 2, and 3 m is shown in Figure 19, with a summary of these results provided in

Table 13. Table of dominant frequencies based on the measured Fourier amplitude ratio for the isolation layer composed of sand containing 5% rubber content (σ₀ = 50 kPa).

Shear Wave Velocity of Clay Soil Layer (m/s)	200
Profile Depth (m)	20	19		18		17
Isolation (m)	0	1		2		3
Shear Wave Velocity of Isolation Layer (m/s)	-	155.45		155.45		155.45
Initial Damping Ratio (%)	-	5.80		5.80		5.80
Natural Frequency (Hz)	2.5	2.465		2.43		2.397
Natural Period (s)	0.4	0.4057		0.4115		0.4175
Earthquake	EL (Hz)	%5–50 kPa 1 m Amp. Ratio (Hz)	Damping Ratio (%)	%5–50 kPa 2 m Amp. Ratio (Hz)	Damping Ratio (%)	%5–50 kPa 3 m Amp. Ratio (Hz)	Damping Ratio (%)
ChiChi	3.702	3.210	−13.29%	3.187	−13.91%	3.158	−14.69%
Coyote	4.056	3.566	−12.08%	3.547	−12.55%	3.533	−12.89%
Imp. Valley	3.873	3.379	−12.75%	3.355	−13.37%	3.336	−13.87%
Kocaeli	4.200	3.748	−10.76%	3.774	−10.14%	3.802	−9.48%
Mean Value			−12.22%		−12.49%		−12.73%

.

The Fourier amplitude ratio for the isolation layer containing 0% granular rubber and subjected to a confining pressure of 100 kPa, applied at thicknesses of 1, 2, and 3 m is shown in Figure 20, with a summary of these results provided in

Table 14. Table of dominant frequencies based on the measured Fourier amplitude ratio for the isolation layer composed of sand containing 0% rubber content (σ₀ = 100 kPa).

Shear Wave Velocity of Clay Soil Layer (m/s)	200
Profile Depth (m)	20	19		18		17
Isolation (m)	0	1		2		3
Shear Wave Velocity of Isolation Layer (m/s)	-	243.15		243.15		243.15
Initial Damping Ratio (%)	-	9.38		9.38		9.38
Natural Frequency (Hz)	2.5	2.522		2.545		2.568
Natural Period (s)	0.4	0.3965		0.3929		0.3894
Earthquake	EL (Hz)	%0–100 kPa 1 m Amp. Ratio (Hz)	Damping Ratio (%)	%0–100 kPa 2 m Amp. Ratio (Hz)	Damping Ratio (%)	%0–100 kPa 3 m Amp. Ratio (Hz)	Damping Ratio (%)
ChiChi	3.702	3.206	−13.40%	3.169	−14.40%	3.106	−16.10%
Coyote	4.056	3.561	−12.20%	3.530	−12.97%	3.490	−13.95%
Imp. Valley	3.873	3.373	−12.91%	3.339	−13.79%	3.297	−14.87%
Kocaeli	4.200	3.745	−10.83%	3.752	−10.67%	3.738	−11.00%
Mean Value			−12.34%		−12.96%		−13.98%

.

The Fourier amplitude ratio for the isolation layer containing 0% granular rubber and subjected to a confining pressure of 50 kPa, applied at thicknesses of 1, 2, and 3 m is shown in Figure 21, with a summary of these results provided in

Table 15. Table of dominant frequencies based on the measured Fourier amplitude ratio for the isolation layer composed of sand containing 0% rubber content (σ₀ = 50 kPa).

Shear Wave Velocity of Clay Soil Layer (m/s)	200
Profile Depth (m)	20	19		18		17
Isolation (m)	0	1		2		3
Shear Wave Velocity of Isolation Layer (m/s)	-	192.77		192.77		192.77
Initial Damping Ratio (%)	-	6.23		6.23		6.23
Natural Frequency (Hz)	2.5	2.495		2.491		2.486
Natural Period (s)	0.4	0.4008		0.4015		0.4023
Earthquake	EL (Hz)	%0–50 kPa 1 m Amp. Ratio (Hz)	Damping Ratio (%)	%0–50 kPa 2 m Amp. Ratio (Hz)	Damping Ratio (%)	%0–50 kPa 3 m Amp. Ratio (Hz)	Damping Ratio (%)
ChiChi	3.702	3.208	−13.34%	3.164	−14.53%	3.165	−14.51%
Coyote	4.056	3.563	−12.15%	3.512	−13.41%	3.511	−13.44%
Imp. Valley	3.873	3.375	−12.86%	3.319	−14.30%	3.316	−14.38%
Kocaeli	4.200	3.748	−10.76%	3.738	−11.00%	3.772	−10.19%
Mean Value			−12.28%		−13.31%		−13.13%

.

Table 4, Table 5, Table 6, Table 7, Table 8, Table 9, Table 10, Table 11, Table 12, Table 13, Table 14 and Table 15 were derived from a combination of numerical simulations and experimental data available in the literature. The response spectral accelerations, Fourier amplitude ratios, and damping behavior were computed using the frequency-domain analysis method, where acceleration time histories obtained from numerical simulations were transformed via Fast Fourier Transform (FFT). The Fourier amplitude ratio was then determined as the ratio of the spectral amplitudes at the surface to those at the base of the isolation layer.

The Fourier amplitude ratios presented in Table 10, Table 11, Table 12, Table 13, Table 14 and Table 15 were derived based on frequency-domain analysis using the Fourier Transform of the acceleration time histories obtained from the numerical simulations. Each Fourier amplitude ratio represents the ratio of the Fourier amplitudes of the surface acceleration to the corresponding Fourier amplitudes at the base level for a given frequency.

To compute the Fourier amplitude ratio, the acceleration time histories from the top and bottom of the soil layers were transformed into the frequency domain using a Fast Fourier Transform (FFT) algorithm. The Fourier amplitudes at each frequency were then normalized by their respective maximum values to facilitate a consistent comparison across different scenarios. This approach allows for a clearer understanding of the amplification behavior across varying frequencies and material configurations.

Regarding the damping behavior, the damping ratios were calculated iteratively based on the equivalent linearization method. This method employs the initial modulus reduction and damping curves for the respective materials (sand and rubber–soil mixtures) under cyclic loading conditions. The calculations account for the strain-dependent nonlinear behavior of the materials and adjust the shear modulus and damping ratio accordingly during each iteration until convergence is achieved.

These assumptions and methodologies ensure that the Fourier amplitude ratios and damping behavior reflect realistic and reliable conditions under seismic loading. By incorporating these details, the results in Table 10, Table 11, Table 12, Table 13, Table 14 and Table 15 provide valuable insights into the dynamic performance of the analyzed isolation layers.

4. Conclusions and Recommendation

The results of this study reveal key differences and similarities with previous studies. While prior studies, such as those by Yin [9], primarily investigated seismic isolation effects in controlled laboratory settings, the present study extends this analysis by incorporating real earthquake records and site-specific ground motion characteristics. The comparative analysis of Fourier amplitude spectra and damping behavior indicates that the inclusion of RSM significantly reduces peak accelerations, particularly in the horizontal direction, a finding consistent with the numerical studies of Pitilakis [11]. However, unlike previous research, which mainly considered shallow isolation layers, this study highlights the importance of confinement depth, demonstrating that the effectiveness of RSM increases with depth up to a certain threshold, beyond which additional thickness provides diminishing returns.

Moreover, the present study challenges some earlier assumptions regarding the relationship between rubber content and damping capacity. While Feng and Sutter [13] suggested a linear increase in damping with higher rubber content, our findings indicate a nonlinear trend where excessive rubber inclusion may lead to a decrease in damping efficiency at lower confining pressures. This discrepancy underscores the necessity of optimizing rubber content based on site-specific conditions rather than applying a uniform design approach.

In summary, this study not only validates and extends previous findings but also introduces a novel approach for optimizing seismic isolation systems using RSM. By addressing limitations in prior research and presenting a comprehensive analysis of geotechnical seismic isolation, this study provides valuable insights for both theoretical advancements and practical applications in earthquake-resistant design.

In this study, the seismic performance of granular rubber–sand mixtures were investigated. For this purpose, pure sand (0% rubber) and sand mixtures containing 5% and 10% granular rubber were utilized. The experimental results reviewed during the literature survey revealed that the seismic isolation performance of granular rubber–sand mixtures is significantly influenced by both the rubber content and the confining pressure. This observation was also confirmed in the numerical analyses conducted during the study, where it was determined that the damping ratio tends to increase as the rubber content in the isolation layer decreases. However, the literature review further highlighted that under higher confining pressures, such as 200–300 kPa, this trend is expected to reverse, as observed in experimental studies in Figure 1.

In the study, it was observed that the damping ratio decreased as the rubber content increased under relatively low confining pressures (50 and 100 kPa). However, the literature review suggests that this trend is expected to reverse at higher confining pressures, such as 200–300 kPa shown in Figure 1. Given its potential to enhance performance and provide economic benefits, this phenomenon remains an open topic for more detailed investigations.

In isolation layers formed with sand–rubber mixtures, it was observed that the damping ratio increased with confining pressure. Similarly, under the same loading conditions, analyses conducted for layers without rubber content revealed that the damping ratio also increased as the confining pressure increased.

Rubber-reinforced soils demonstrated an enhanced damping capacity under dynamic loads. However, at low confining pressures, higher rubber content was found to reduce the elastic modulus, thereby negatively impacting soil performance. Under confining pressures of 50–100 kPa, the reduction in rubber content demonstrated a positive influence on isolation layers, effectively mitigating soil–structure interaction risks.

When evaluated in terms of economic and sustainability aspects, the numerical analyses conducted in this study, combined with the experimental results reviewed in the literature, suggest that isolation layers composed of rubber-enhanced sand (5–10% rubber) and pure sand provide a cost-effective and sustainable seismic isolation solution. However, higher rubber content may negatively affect soil stability, increasing deformation risks. Thus, the careful optimization of rubber content is crucial during the design phase.

The increase in the thickness of the isolation layer significantly reduces amplification ratios. Generally, as the thickness of the seismic isolation layer increases, the damping ratio also increases. However, it has been observed that 1 m thick isolation layers are more effective in preventing seismic loads from being transmitted to the structure compared to 2 m and 3 m thick isolation layers. This is because the damping ratios provided by thicker isolation layers do not increase proportionally; instead, their effects are characterized by relatively smaller increments. Based on the results for isolation layers of 1 m, 2 m, and 3 m thickness, the percentage reductions in the peak accelerations of the soil response spectrum are as follows as shown in

Table 16. The percentage of reductions in the peak accelerations of the response spectrum.

Earthquakes	1 m	2 m	3 m
10% rubber at 50 kPa	%6.8	%6.87	%6.09
10% rubber at 100 kPa	%7.41	%8.98	%10.53
5% rubber at 50 kPa	%8.31	%10.57	%12.46
5% rubber at 100 kPa	%8.64	%11.87	%14.33
0% rubber at 50 kPa	%9.27	%13.92	%14.96
0% rubber at 100 kPa	%9.43	%13.17	%16.17

:

It has been observed that isolation layers influence the natural frequency and amplitude behavior of the soil. The natural frequency of the untreated clay soil with a shear wave velocity of 200 m/s was measured to be 2.5 Hz. However, based on the results of applying isolation layers of 1 m, 2 m, and 3 m thickness, the natural frequencies of the soil were determined as follows as shown in

Table 17. Natural frequency and amplitude behavior of the isolated soil.

Earthquakes	1 m	2 m	3 m
10% rubber at 50 kPa	2.420 Hz	2.346 Hz	2.275 Hz
10% rubber at 100 kPa	2.461 Hz	2.424 Hz	2.387 Hz
5% rubber at 50 kPa	2.465 Hz	2.430 Hz	2.397 Hz
5% rubber at 100 kPa	2.498 Hz	2.495 Hz	2.493 Hz
0% rubber at 50 kPa	2.495 Hz	2.491 Hz	2.486 Hz
0% rubber at 100 kPa	2.522 Hz	2.545 Hz	2.568 Hz

.

Numerical analyses conducted using DeepSoil software were found to be in strong agreement with experimental results. The experimental findings indicated that reducing the rubber content under low confining pressures led to more controlled and reliable dynamic behavior.

Rubber-enhanced soils, produced using recycled granular rubber materials, offer an economic and environmentally friendly solution. These materials, particularly when considering budget constraints in developing countries, stand out as an innovative seismic isolation approach for several key reasons:

Granular rubber is derived from recycled materials such as discarded vehicle tires. This not only addresses waste management challenges but also promotes a sustainable construction approach by reducing environmental impacts. Repurposing end-of-life tires into valuable materials minimizes landfill requirements and decreases the risks of environmental pollution.

Recycling rubber materials helps conserve raw material resources. This reduces the risk of depleting natural resources and promotes a sustainable circular economy model, ensuring the efficient utilization of materials.