Performance of wall mesh encased stone columns using recycled rubber concrete and asphalt aggregates for sustainable geotechnics

Abstract

The growing volume of construction and demolition waste, together with discarded tires, highlights the urgent need for sustainable alternatives in soil improvement. Geosynthetic-encased stone columns (GESCs) are widely used to enhance bearing capacity and reduce settlement in soft soils, yet their reliance on natural aggregates contributes to resource depletion. To address this gap, this study evaluates the performance of GESCs constructed with recycled concrete aggregate (RCA), reclaimed asphalt pavement (RAP), and shredded waste rubber as partial replacements for natural gravel. A comprehensive experimental program consisting of 34 large-scale model tests was conducted with constant L/D ratio, two column diameters, and multiple material combinations. The findings show that several recycled mixtures, particularly those incorporating RCA and small rubber fractions, exhibited improved stiffness, enhanced stress–displacement response, and more controlled deformation compared with conventional gravel columns. Quantitatively, recycled-aggregate GESCs achieved up to 2% higher bearing capacity and Stress-to-Displacement Ratio of 12.6 compare to natural gravel columns, demonstrating their viability as an effective and environmentally sustainable ground-improvement solution.

Introduction

With the expansion of urbanization and the development of infrastructure in areas with problematic soils, geotechnical engineers are increasingly confronted with challenges in designing and constructing stable foundations and resilient structures. Weak soils (particularly soft clays) pose serious limitations in construction projects due to their low bearing capacity, significant settlements, and unfavorable consolidation characteristics¹. In this context, ground improvement is recognized as an essential approach to overcome these limitations and to provide a suitable subgrade for structures. Numerous ground improvement techniques have been developed, among which stone columns have attracted considerable attention because of their advantages such as rapid execution, cost-effectiveness, and environmental compatibility².

Stone columns are densely compacted granular inclusions installed within weak soils. They enhance the geotechnical properties of the ground through mechanisms including reinforcement, drainage, and stress redistribution. By creating vertical drainage paths, these columns accelerate the consolidation process. Their higher stiffness compared to the surrounding soil allows them to absorb a substantial portion of applied loads, thereby reducing total settlement and increasing system bearing capacity³. Earlier numerical and experimental investigations have shown that stone column systems develop significant stress concentration due to soil arching, causing a substantial portion of applied loads to be transferred to the stiffer column material⁴. This stress-transfer mechanism plays a central role in settlement reduction and governs the effectiveness of column-supported ground. In addition, stone columns generate a confinement effect in the soil, which in turn enhances the shear strength of the surrounding ground. Despite these benefits, the use of natural materials such as gravel and sand in stone column construction raises significant environmental concerns. Large-scale extraction of aggregates results in habitat destruction, soil erosion, groundwater contamination, and greenhouse gas emissions⁵. Transportation costs of these materials, especially in areas with limited access to stone resources, are also considerable. Meanwhile, the increasing generation of industrial and construction wastes such as demolished concrete, asphalt, and discarded tires has made waste management one of the most pressing global environmental issues⁶. According to the Environmental Protection Agency, more than 600 million tons of construction waste and over 300 million tons of waste tires are produced annually worldwide, and their disposal through landfilling or incineration results in severe environmental consequences⁷.

Against this background, the use of recycled materials in stone column construction has emerged as an innovative solution that simultaneously addresses the challenges of ground improvement and waste management⁸. Replacing natural aggregates with recycled materials such as crushed concrete, reclaimed asphalt, and waste tires not only reduces demand for natural resources but also decreases waste volumes and associated pollution. The use of stone columns can be traced back to the early twentieth century, but systematic development and widespread application mainly took place in the 1970s with the advancement of equipment and construction methods⁹. This ground improvement technique is based on replacing a portion of weak soil with densely compacted granular material of high strength and stiffness, thereby improving geotechnical properties through several simultaneous mechanisms. The main mechanisms include: (a) increasing bearing capacity by transferring stresses to stronger underlying layers, (b) reducing settlement by substituting part of the weak soil with compacted material, (c) accelerating consolidation by providing drainage paths, and (d) improving overall stability by enhancing shear strength. Depending on soil conditions and design requirements, stone columns can be categorized into floating columns, which are installed within the weak layer without reaching the firm stratum, and end-bearing columns, which extend down to a competent bearing layer. Reinforced stone columns, strengthened with geosynthetics, represent a more advanced type used in complex conditions⁹. Selection of the appropriate installation method depends on project requirements, soil properties, and available equipment, and this choice plays a critical role in improving the performance of ground structures¹⁰.

Extensive studies have shown that reinforcement techniques such as geosynthetic encasement can significantly enhance the performance of stone columns¹¹. Research on geogrid-encased stone columns indicates that this approach can increase bearing capacity by up to four times and settlement resistance by up to three times compared with conventional stone columns¹¹. Laboratory and numerical studies, including those by Lou et al.¹², confirm that parameters such as geogrid stiffness, column diameter, column length, and material properties strongly influence the behavior of encased stone columns. In recent decades, rising environmental and economic concerns have accelerated the adoption of recycled materials in geotechnical projects, reflecting an innovative trend in the field¹³. Although the use of recycled materials in geotechnics dates back to the late 1980s, its wider application has expanded in the twenty-first century with advances in recycling technologies and increased environmental awareness¹⁴. Recycled concrete, one of the most commonly used recycled materials in geotechnics, is obtained by crushing concrete from demolished structures and represents a significant portion of construction waste¹⁵. Research has shown that properly graded recycled concrete can exhibit satisfactory geotechnical properties in terms of strength, stiffness, and durability, although it often shows higher water absorption and a different friction angle than natural aggregates⁶. In recent years, several studies have highlighted the promising geotechnical performance of recycled concrete aggregates (RCA) in soft-ground applications, reporting satisfactory strength, stiffness, and drainage characteristics comparable to natural aggregates^16,17. These works demonstrate that carefully processed RCA can serve as a viable infill material for sustainable column-based ground improvement. Moreover, combining recycled materials and using reinforcement techniques such as geosynthetic encasement can provide new approaches for enhancing stone column performance¹⁸. Recent studies on stabilizing engineering sediment with GGBS–EMR–cement binders show that incorporating recycled materials can substantially improve strength and durability in sustainable geotechnical applications¹⁹. Although the additives in that study were cementitious rather than rubber-based, the findings reinforce the broader evidence that well-designed recycled mixtures can enhance mechanical performance, an observation consistent with the use of recycled aggregates and rubber in stone columns and wall-mesh–encased systems.

Recycled rubber, produced by shredding waste tires, has been used in geotechnical applications due to properties such as low weight, high elasticity, and abrasion resistance²⁰. Humphrey and colleagues²¹ demonstrated that mixing recycled rubber with soil can improve dynamic response and reduce settlement. Reclaimed asphalt, obtained by milling or crushing aged pavements, has also been applied in geotechnical works because of its adhesive properties, making it suitable for use in subbase layers, fillers, or blended with other materials²². Geogrid encasement of stone columns is regarded as a major advance in ground improvement technology and has seen significant development over the past two decades²³. This technique provides lateral confinement by wrapping the column with geosynthetic materials such as geotextiles, geogrids, or geomembranes in a cylindrical form Fang et al.²⁴. The main mechanism involves increasing lateral confinement, which enhances shear strength, stiffness, and bearing capacity⁷. When subjected to loading, the stone column tends to expand laterally, but this expansion is restrained by the geogrid, creating confining pressure that improves column strength and stiffness²⁵. Experimental and numerical studies consistently confirm that geogrid encasement substantially enhances confinement and reduces bulging, leading to improved bearing capacity and stiffer load–settlement response; increases of 50–60% in bearing capacity are commonly reported for encased columns compared with ordinary columns^26,27. Different types of geogrids exist with varying tensile strength, stiffness, permeability, and durability. Their proper selection depends on project requirements, soil conditions, and design needs²⁸.

Although extensive research exists on stone columns and on recycled materials in geotechnics, significant research gaps remain in integrating these two areas¹². A review of the literature reveals that studies on recycled materials in geotechnical applications are relatively numerous, but their specific use in stone columns has been limited and often focused on a single type of recycled material²⁹. Few studies have examined in detail the influence of different recycled material contents, such as shredded rubber, on stone column performance. Moreover, only a limited number of investigations have directly compared stone columns constructed with different recycled materials, such as recycled concrete, asphalt, and rubber¹². Furthermore, existing studies typically investigate a single recycled material or focus solely on conventional aggregates, without assessing how different recycled mixtures interact with geogrid encasement to influence stress concentration, bulging behaviour, and overall column performance^17,30. In addition, while the effects of geogrid encasement on conventional stone columns are well documented, the combined influence of encasement and recycled materials requires further study⁷. Most laboratory and numerical research has concentrated on stone columns built with traditional aggregates, with less attention to the behavior of columns constructed with recycled materials in conjunction with encasement techniques.

The rapid growth of construction waste and discarded tires has intensified the need for sustainable approaches in geotechnical engineering. This study evaluates the feasibility of using recycled materials in stone columns through a laboratory program comparing recycled concrete and gravel–rubber mixtures with conventional columns. Beyond assessing bearing capacity and load–settlement behavior, the research addresses key questions regarding load transfer, the role of rubber content, and performance thresholds of alternative mixtures. The novelty lies in three aspects: integrating diverse recycled materials, investigating their interaction with geogrid encasement, and developing practical guidelines that balance engineering performance with environmental sustainability. The use of geosynthetic wall-mesh encasement further enables a refined assessment of reinforced recycled columns, contributing to sustainable ground improvement practices. The experimental program, materials, and testing procedures are first introduced, followed by a presentation of the key results on bearing capacity and load–settlement behavior. The findings are then analyzed to discuss the role of material composition and reinforcement, and the study concludes with recommendations for practical application and directions for future research. Moreover, the literature lacks systematic comparisons of GESCs constructed with different recycled aggregates, such as RCA, RAP, and rubber, and the combined influence of material composition and wall-mesh encasement on load transfer mechanisms remains insufficiently documented. Addressing this gap is essential for developing sustainable, high-performance stone-column systems.

Materials

In this study, both natural and recycled materials were utilized for the construction of stone column specimens. Natural gravel served as the control material, while recycled concrete aggregate (RCA), reclaimed asphalt pavement (RAP), and shredded waste rubber were incorporated as sustainable alternatives. To provide lateral confinement, a geosynthetic wall-mesh was employed. The physical and mechanical characteristics of these materials, as well as their preparation procedures, are described in the following subsections.

Soft clay (Test bed)

Soft clay was used as the laboratory test bed to evaluate the performance of the stone column systems. Owing to its unfavorable geotechnical characteristics, including low undrained shear strength and high compressibility, this soil provided a representative problematic ground condition for assessing the effectiveness of ground improvement. The clay used in this study was sourced from the Chalghi clay mine located in Tangel-e Shur, Mashhad, ensuring material consistency throughout the experimental program. A comprehensive series of laboratory tests was conducted to characterize the physical and mechanical properties of the clay. The soil was classified as CL–CH according to the Unified Soil Classification System (USCS). Key parameters, including Atterberg limits, effective strength properties, stiffness, compressibility, permeability, unit weights, and the undrained shear strength (cu = 30 kPa), were determined and maintained consistently across all tests. These properties are summarized in Table 1, which provides the complete characterization of the soft clay employed in the experimental setup. This detailed soil characterization ensures reproducibility of the testing conditions and provides a reliable basis for interpreting the performance of the improved ground system.

Table 1.

Physical and mechanical properties of the soft clay used in the experimental test bed.

Parameter	Symbol	Value	Unit
Soil classification system	USCS	CL–CH	–
Liquid limit	LL	40	%
Plastic limit	PL	22	%
Plasticity index	PI	18	%
Effective cohesion	c′	15	kPa
Effective internal friction angle	φ′	18	deg
Undrained shear strength	cu	30	kPa
Young’s modulus	E	3	MPa
Poisson’s ratio	ν	0.45	–
Natural unit weight	γ	18	kN/m3
Saturated unit weight	γ_sat	19	kN/m3

Gravel and sand (Control samples)

Rounded natural gravel passing the 1/2-inch sieve and retained on the 3/8-inch sieve was used to construct the control stone columns. The material was selected in accordance with ASTM D2487, considering factors such as internal friction angle, density, and adequate drainage properties. Sieve analysis, compaction, and direct shear tests were performed to determine its physical and mechanical characteristics.

Recycled concrete aggregate

Recycled concrete aggregate was obtained by crushing demolished concrete blocks with an initial compressive strength of 25 MPa. The gradation was adjusted using standard sieves so that the material passed the 1/2-inch sieve and was retained on the 3/8-inch sieve. Physical properties—including density, water absorption, and abrasion resistance—and mechanical properties such as internal friction angle and elastic modulus were measured in accordance with ASTM C33 and ASTM D7012. Relative to natural gravel, the recycled concrete aggregate showed higher water absorption and lower specific gravity.

Recycled rubber

Recycled rubber was sourced from discarded automobile tires. After the removal of steel and fabric components, the rubber was cut into two particle sizes: 1 × 1 cm and 1 × 2 cm. The rubber had a density of 950 kg/m³ with negligible water absorption. The elastic modulus ranged from 5 to 8 MPa, and Poisson’s ratio was measured between 0.45 and 0.49. All rubber particles were thoroughly washed and dried prior to use.

Reclaimed asphalt pavement

Reclaimed asphalt pavement (RAP) was obtained by milling deteriorated highway surfacing layers and then sieved according to ASTM D448 to retain particles in the 4–25 mm size range. The bitumen content, density, water absorption, and mechanical properties such as internal friction angle and cohesion were determined in accordance with ASTM D2172 and ASTM D3080.

Geosynthetic wall-mesh (Encasement)

A high-strength geosynthetic wall-mesh was used to encase the stone columns. Its properties, including tensile strength in both principal directions, were measured both before and after exposure to an alkaline environment, following ASTM D4595 and ASTM D5262.

At the end of this section, Table 2 presents a summary of the experimental matrix, including the volumetric composition of gravel, recycled concrete aggregate, reclaimed asphalt pavement, and recycled rubber, along with the selected rubber particle sizes and the corresponding internal friction angles and dry unit weights of the materials used in the stone column specimens.

Table 2.

Test material properties of the stone column specimens.

Test No.	Constituent materials	Gravel content (% by volume)	Recycled concrete content (% by volume)	Recycled asphalt content (% by volume)	Recycled rubber content (% by volume)	Rubber particle size	Internal friction angle φ (deg)	Max dry unit weight γmax (kN/m3)	Min dry unit weight γmin (kN/m3)
1, 18	Gravel + Rubber	80	0	0	20	1*1	21	20.5	16.5
2, 19	Gravel + Rubber	90	0	0	10	1*1	31	20.5	16.5
3, 20	Gravel + Rubber	80	0	0	20	1*2	18	20.5	16.5
4, 21	Gravel + Rubber	90	0	0	10	1*2	21	20.5	16.5
5, 22	Gravel	100	0	0	0	0	40	20.5	16.5
6, 23	Concrete + Rubber	0	80	0	20	1*1	31	19.5	16
7, 24	Concrete + Rubber	0	90	0	10	1*1	22	19.5	16
8, 25	Concrete + Rubber	0	80	0	20	1*2	18	19.5	16
9, 26	Concrete + Rubber	0	90	0	10	1*2	15	19.5	16
10, 27	Concrete	0	100	0	0	0	17	19.5	16
11, 28	Asphalt + Rubber	0	0	80	20	1*1	13	18.5	15
12, 29	Asphalt + Rubber	0	0	90	10	1*1	15	18.5	15
13, 30	Asphalt + Rubber	0	0	80	20	1*2	8	18.5	15
14, 31	Asphalt + Rubber	0	0	90	10	1*2	9	18.5	15
15, 32	Asphalt	0	0	100	0	0	16	18.5	15
16, 33	Gravel + Concrete	80	20	0	0	0	31	20	16.5
17, 34	Gravel + Concrete	20	80	0	0	0	10	20	16.5

Laboratory methods

The experimental program was designed to systematically evaluate the performance of stone columns constructed with different combinations of natural and recycled materials. A comprehensive testing procedure was established, including the definition of independent and dependent variables, sample preparation, column installation, and axial loading tests under controlled laboratory conditions. The details of these procedures are presented in the following subsections to ensure clarity and reproducibility.

Definition of experimental variables

In this study, four main independent variables were considered: (1) the type of column material (natural gravel, recycled concrete, recycled asphalt, and their mixtures with rubber), (2) the volumetric percentage of recycled rubber (0%, 10%, and 20%), (3) the size of rubber particles (1 × 1 cm and 1 × 2 cm), and (4) the geometric dimensions of the stone columns (lengths of 80 and 100 cm, and diameters of 16 and 20 cm). The dependent variables included the ultimate bearing capacity and the settlement of the columns under axial loading.

Design and preparation of samples

Based on these independent variables, a total of 34 tests were designed. The specimens included control columns (100% gravel, 100% recycled concrete, and 100% recycled asphalt), columns with rubber (80% or 90% primary material mixed with 20% or 10% recycled rubber), and composite columns (80% gravel + 20% recycled concrete, or 20% gravel + 80% recycled concrete). For each material combination, two column sizes (80 × 16 cm and 100 × 20 cm) and two rubber particle sizes (1 × 1 cm and 1 × 2 cm) were considered.

Constant experimental assumptions

To ensure comparability, several fixed conditions were maintained: the same soft clay bed was used as the foundation soil, all columns were constructed as floating columns, complete encasement was provided using identical geosynthetic wall-mesh, each column was tested individually (not in groups), a constant loading rate was applied, and laboratory environmental conditions were controlled (temperature 23 ± 2 °C and relative humidity 50 ± 5%). These measures ensured reliable assessment of the effects of the independent variables.

Equipment and test system

The test chamber was made of steel with a 10 mm wall thickness and internal dimensions of 120 × 120 × 150 cm, providing sufficient clearance to minimize boundary effects. The loading system consisted of a rigid steel frame, a hydraulic jack with a 50 kN capacity, and a loading rate control unit. A load cell with 0.01 kN precision and digital displacement gauges were employed for measurements. All instruments were connected to a data acquisition system for continuous recording of load and settlement. Sensors were calibrated prior to each test.

Overview of experimental procedure

The experimental procedure for constructing and testing wall-mesh–encased stone columns in a soft clay bed was carried out in a stepwise manner, as schematically illustrated in the different subfigures.

Initially, the soft clay was thoroughly prepared and brought to its optimum moisture content to ensure uniform and repeatable testing conditions. In parallel, the granular materials used for constructing the stone columns—including natural gravel, recycled aggregates (concrete and asphalt), and shredded rubber—were prepared according to the predefined mix proportions and stockpiled prior to testing (Fig. 1a).

Fig. 1.

Schematic illustration of the experimental setup and testing procedure for wall-mesh–encased stone columns in soft clay.

Subsequently, the test box was filled with the prepared clay in a layer-by-layer manner, and each layer was compacted using a manual compactor to achieve uniform compaction throughout the clay bed (Fig. 1b). After forming the initial clay bed, the location of the stone column was marked at the center of the test box.

A PVC pipe with a height of 20 cm and the designated diameter was then positioned vertically at the center of the box. The inner surface of the pipe was lined with a geosynthetic wall-mesh, after which the stone column material was carefully placed inside the pipe (Fig. 1c). Clay was then placed and compacted around the pipe to ensure proper confinement and contact between the column and the surrounding soil.

Following each filling stage, the PVC pipe was connected to the upper loading frame using a chain and gradually lifted upward. Immediately after lifting the pipe, the surrounding clay was re-compacted to prevent void formation and to maintain uniform confinement (Fig. 1d). This sequence of column filling, pipe lifting, and surrounding soil compaction was repeated until the stone column reached the target height.

Once the column construction was completed, a steel loading plate with a diameter larger than that of the stone column was placed on the column head to ensure uniform load distribution between the column and the surrounding soil. Vertical loading was then applied using a hydraulic jack equipped with a load cell and a digital displacement gauge, and the test was continued until column failure occurred (Fig. 1e).

After completion of each test, the testing box was completely emptied, and the clay bed was reconstructed for the subsequent experiment. This procedure was repeated for all test specimens to ensure consistent testing conditions across the experimental program (Fig. 1f).

Preparation of the soil bed

The clay bed was first dried, pulverized, and sieved. Its optimum moisture content was determined through standard compaction tests. The clay was then brought to the optimum moisture and stored in a humid chamber for 24 h to ensure uniform moisture distribution. It was placed in the test chamber in 10 cm layers, each compacted with a vibrating hammer to achieve 95% of the maximum dry density determined by the standard Proctor test. Uniformity of compaction was verified using the sand cone test.

Construction and installation of stone columns

A borehole with the designated diameter and depth was drilled at the center of the chamber. A casing pipe of the same diameter was inserted, and the geosynthetic wall-mesh was placed inside in cylindrical form. The column material was then poured in three layers, each compacted with a tamping rod. After completion, the casing pipe was carefully withdrawn, leaving the wall-mesh and column materials in place. A circular steel plate equal to the column diameter was placed on top to ensure uniform load distribution.

Axial loading test

After column construction, the hydraulic jack and load–displacement sensors were installed. Loading was applied at a constant rate of 1 mm/min, and load–settlement data were continuously recorded. Each test continued until a settlement of 50 mm was reached or visible failure occurred. For each material combination, the tests were repeated at least twice to ensure the reliability of the results.

Laboratory setup and experimental documentation

To enhance the clarity of the experimental methodology and provide visual documentation of the execution process, Fig. 2 presents a comprehensive view of the laboratory environment, loading equipment, column construction, and specimen conditions at different stages of testing. The figure serves as a visual complement to the methodological descriptions and illustrates the actual laboratory implementation of the experimental program. Figure 2a shows the axial loading system, including the hydraulic jack, loading plate, and load–displacement measurement devices used to apply a uniform vertical load on the stone column head. Figure 2g provides an overall view of the test box and its surrounding laboratory environment, illustrating the configuration, positioning, and boundary conditions of the experimental chamber. Figure 2b illustrates a wall-mesh–encased stone column after full placement within the clay bed, representing the column geometry and its interaction with the surrounding soil prior to loading. Figure 2h shows the manual compaction of the clay layers around the stone column, which was carried out to ensure uniform density and controlled preparation of the clay bed. Figure 2c presents the stone column after failure, exposing the deformation pattern and bulging mechanism developed under axial loading. Figure 2d–f display the stone column materials prepared with different combinations, including natural gravel, recycled concrete aggregate, reclaimed asphalt pavement, and shredded rubber, documenting the variety of mixtures used in the experimental program.

Fig. 2.

Laboratory setup, loading system, material preparation, and representative views of stone column construction and failure.

Results and discussion

This section presents and analyzes the results of the laboratory tests performed on stone columns encased with geosynthetic wall-mesh and incorporating recycled materials. The main indicators considered are ultimate bearing capacity, settlement, and the stress-to-displacement ratio. After introducing the key parameters and summarizing the overall results, the effects of material type, rubber content, and other experimental factors on column behavior are examined.

Key experimental parameters

Several main parameters were used to evaluate the performance of the stone columns. The ultimate bearing capacity represented the maximum load that a column could sustain and served as a fundamental criterion for comparing different materials. The final settlement indicated the vertical displacement of the column under loading and reflected the system’s deformability. The stress-to-displacement ratio was used to compare the relative stiffness of the columns and to assess the balance between stiffness and ductility across different material combinations. Together, these parameters provided a comprehensive framework for analyzing the mechanical behavior of stone columns constructed with recycled materials.

Grouping of specimens

The test specimens were grouped according to the type and proportion of materials. In this study, stone columns were constructed using diverse combinations of gravel, recycled concrete, recycled asphalt, and varying percentages of rubber. In addition, changes in column dimensions and rubber particle sizes had a measurable influence on the mechanical behavior and relative stiffness of the columns.

Tables 3 and 4 summarize all 34 tests conducted, categorized by primary material, material proportions, column dimensions, and rubber particle sizes. Tests 1 to 17 correspond to Tests 18 to 34, with the only difference being specimen size, while the length-to-diameter ratio remained constant. These tables provide detailed information on the composition of each specimen, along with their mechanical performance. For the rubber-containing specimens, the volumetric content of rubber was set at 10% or 20%, while the proportions of gravel, recycled concrete, or asphalt ranged from 0% to 100%. It should also be noted that the parameters Length (cm), Diameter (cm), and L/D Ratio remain constant within each series and are therefore not repeated in the tables. For Series 1, these values are 80 cm, 16 cm, and 5, respectively, while for Series 2 they are 100 cm, 20 cm, and 5.

Table 3.

Material compositions, geometric dimensions, and mechanical results of stone columns in the first test series (specimens 80 cm length and 16 cm in diameter, L/D = 5).

Test No.	Constituent materials	Gravel content (% by volume)	Recycled concrete content (% by volume)	Recycled asphalt content (% by volume)	Recycled rubber content (% by volume)	Rubber particle size	Maximum stress (kN/m2)	Settlement (mm)	Stress-to-displacement ratio
1	Gravel + Rubber	80	0	0	20	1*1	866.7	167.7	5.17
2	Gravel + Rubber	90	0	0	10	1*1	955.3	75.8	12.6
3	Gravel + Rubber	80	0	0	20	1*2	876.3	169.5	5.17
4	Gravel + Rubber	90	0	0	10	1*2	847.3	167	5.07
5	Gravel	100	0	0	0	0	936.1	138.6	6.75
6	Concrete + Rubber	0	80	0	20	1*1	919.1	83.4	11.02
7	Concrete + Rubber	0	90	0	10	1*1	881.1	140.7	6.26
8	Concrete + Rubber	0	80	0	20	1*2	853.4	168.3	5.07
9	Concrete + Rubber	0	90	0	10	1*2	881.1	151.5	5.81
10	Concrete	0	100	0	0	0	483.9	180.5	2.68
11	Asphalt + Rubber	0	0	80	20	1*1	799.6	134.1	5.96
12	Asphalt + Rubber	0	0	90	10	1*1	829.3	156.8	5.29
13	Asphalt + Rubber	0	0	80	20	1*2	645	68.9	9.36
14	Asphalt + Rubber	0	0	90	10	1*2	704.4	71.3	9.88
15	Asphalt	0	0	100	0	0	839.3	150.3	5.59
16	Gravel + Concrete	80	20	0	0	0	963.3	82	11.74
17	Gravel + Concrete	20	80	0	0	0	706.2	87.8	8.04

Table 4.

Material compositions, geometric dimensions, and mechanical results of stone columns in the second test series (specimens 100 cm length and 20 cm in diameter, L/D = 5).

Test No.	Constituent materials	Gravel content (% by volume)	Recycled concrete content (% by volume)	Recycled asphalt content (% by volume)	Recycled rubber content (% by volume)	Rubber particle size	Maximum stress (kN/m2)	Settlement (mm)	Stress-to-displacement ratio
18	Gravel + Rubber	80	0	0	20	1*1	353.7	147.3	2.4
19	Gravel + Rubber	90	0	0	10	1*1	389.2	75.8	5.13
20	Gravel + Rubber	80	0	0	20	1*2	348.9	167.5	2.08
21	Gravel + Rubber	90	0	0	10	1*2	345.8	167	2.07
22	Gravel	100	0	0	0	0	382.1	136.5	2.8
23	Concrete + Rubber	0	80	0	20	1*1	368.1	169.2	2.18
24	Concrete + Rubber	0	90	0	10	1*1	359.6	142.8	2.52
25	Concrete + Rubber	0	80	0	20	1*2	348	165	2.11
26	Concrete + Rubber	0	90	0	10	1*2	330	163	2.03
27	Concrete	0	100	0	0	0	216.4	172.4	1.26
28	Asphalt + Rubber	0	0	80	20	1*1	322.6	154.6	2.09
29	Asphalt + Rubber	0	0	90	10	1*1	331.8	175.1	1.9
30	Asphalt + Rubber	0	0	80	20	1*2	263.2	77	3.42
31	Asphalt + Rubber	0	0	90	10	1*2	287.5	71.3	4.03
32	Asphalt	0	0	100	0	0	342.5	91	3.76
33	Gravel + Concrete	80	20	0	0	0	389.2	82	4.74
34	Gravel + Concrete	20	80	0	0	0	342.3	87.8	3.9

Comparative analysis of corresponding tests (First and second series)

For each material composition, the tests conducted in the first series (Tests 1–17) were directly compared with their corresponding counterparts in the second series (Tests 18–34) in terms of ultimate bearing capacity, final settlement, and stress-to-displacement response. Each paired comparison (e.g., Tests 1 and 18, 2 and 19, etc.) involved identical material compositions, with the only difference being the column dimensions, where both length and diameter were increased while maintaining a constant length-to-diameter (L/D) ratio.

As illustrated in Fig. 3, the mechanical performance of the tested stone columns is presented comparatively for the two experimental series. The bar charts in Fig. 3a and b show the ultimate bearing capacity and final settlement of each specimen in the first (smaller-scale) and second (larger-scale) series, respectively. Although increasing the column dimensions altered the absolute values of bearing capacity and settlement, the relative performance trends and ranking among different material combinations remained largely unchanged. Material groups that exhibited superior performance in the first series consistently demonstrated similar advantages in the second series.

Fig. 3.

Comparison of ultimate bearing capacity and settlement for stone column specimens in the two experimental series.

In several material combinations, particularly those rich in gravel or recycled concrete, the larger columns exhibited a slightly more deformable response. This behavior can be attributed to the greater mobilization of particle rearrangement and lateral deformation along the column height as the column dimensions increased, even though the length-to-diameter ratio was kept constant. As a result, strain was distributed over a larger volume of material, leading to smoother stress–displacement curves and increased deformation capacity. Importantly, this change in deformation behavior did not alter the dominant load-transfer mechanism or the relative performance ranking of the material combinations, indicating that the observed differences arise from geometric scaling effects rather than changes in the fundamental mechanical response.

These observations indicate that maintaining a constant L/D ratio effectively preserves the dominant failure mechanism and overall mechanical behavior of the stone columns. Consequently, variations in column dimensions primarily influence absolute settlement magnitudes and subtle response features, rather than fundamentally altering stiffness, load-transfer mechanisms, or performance ranking among different material compositions.

Mechanistic interpretation of stress–displacement responses

The stress–displacement responses presented in the revised figures provide insight into the governing mechanisms controlling the behavior of wall-mesh–encased stone columns constructed with different material combinations. Rather than merely reflecting geometric effects, the shape and slope of the curves primarily capture the interaction between material composition, internal friction, compressibility, and the confinement effect induced by the wall-mesh. Consequently, variations in stiffness, ultimate bearing capacity, and post-peak behavior can be directly linked to the mechanical characteristics of the constituent materials and their ability to mobilize confinement under axial loading.

For a more rigorous and comprehensive evaluation, test results corresponding to similar material compositions but different column dimensions are examined side by side. This comparative approach allows the influence of column size to be isolated while keeping material-related parameters constant, thereby enabling a clearer interpretation of scale effects on stress–displacement behavior. Based on this framework, the following discussion systematically analyzes paired experimental results to elucidate how changes in column dimensions affect stiffness, bearing capacity, and deformation characteristics under otherwise comparable conditions.

Figure 4 compares the stress–displacement responses of wall-mesh–encased stone columns constructed with gravel-based mixtures, including the control gravel column and gravel–rubber blends, tested in two geometric series. Figure 4a presents Tests 1–5 (smaller columns), while Fig. 4b shows the corresponding Tests 18–22 (larger columns) with the same material compositions. Pairing the results in this manner isolates the influence of column dimensions while keeping the constituent materials constant (Exp. 1–18, 2–19, 3–20, 4–21, and 5–22), thereby enabling a clearer interpretation of scale-related effects under comparable material conditions. Across both series, the overall ranking and the characteristic shape of the curves remain broadly consistent, indicating that material composition is the dominant factor governing mechanical response, whereas increasing column size primarily affects the absolute stress and displacement levels. The gravel control specimens (Exp. 5 and Exp. 22) provide the reference behavior of a fully granular column where load transfer is governed by particle interlocking and friction mobilization under confinement. Introducing recycled rubber modifies this mechanism by increasing the deformability of the granular skeleton and reducing the efficiency of force chains and inter-particle contact stiffness, which is reflected by a generally softer stress–displacement response relative to the gravel control—particularly when the rubber content is high. A clear effect of rubber content is observed in both series: mixtures with 10% rubber (Exp. 2 and Exp. 4; Exp. 19 and Exp. 21) typically show stiffer responses and higher resistance than their 20% rubber counterparts (Exp. 1 and Exp. 3; Exp. 18 and Exp. 20). Mechanistically, limited rubber inclusion can act as a compliant phase that enhances ductility and improves stress redistribution without severely disrupting the load-bearing granular skeleton. However, when rubber content increases to 20%, the proportion of compressible inclusions becomes sufficiently large to weaken the aggregate-to-aggregate contact network, reduce shear resistance mobilization, and promote larger settlements at comparable stress levels. Consequently, the 20% rubber specimens exhibit lower apparent stiffness and earlier tendency toward softening compared with the 10% rubber specimens. The rubber particle size also contributes to the observed trends. Comparing tests with identical rubber content but different rubber sizes (1 × 1 cm versus 1 × 2 cm) indicates that particle geometry influences packing and interfacial interaction with gravel. Smaller rubber particles are more readily distributed within the void structure of gravel and may allow a more continuous granular framework, whereas larger rubber pieces can interrupt particle interlocking more significantly and promote localized deformation. This difference is manifested in the divergence of the stress–displacement trajectories among the 10% and 20% rubber groups across both series. Overall, Fig. 4 demonstrates that, for gravel-based encased columns, a moderate rubber content (10%) maintains a more effective frictional skeleton and confinement mobilization, whereas higher rubber content (20%) leads to a softer response and larger displacement due to increased compressibility and reduced interlocking. Importantly, the similarity in behavioral ranking between the two geometric series supports the conclusion that the mixture design effects are robust with respect to the tested scale, while absolute response levels remain sensitive to column dimensions and boundary conditions.

Fig. 4.

Stress–displacement responses of wall-mesh–encased gravel-based stone columns with different rubber contents and particle sizes: (a) first series (Exp. 1–5) and (b) second series (Exp. 18–22).

Figure 5 presents the stress–displacement behavior of wall-mesh–encased stone columns constructed with recycled concrete aggregate (RCA) as the primary constituent, including both the RCA control columns and RCA–rubber mixtures, tested in two geometric series. Figure 5a corresponds to Tests 6–10 (smaller columns), while Fig. 5b shows the matching Tests 23–27 (larger columns). As in the previous figure, each pair of tests shares an identical material composition, allowing the effect of column size to be examined independently from material-related factors. In both series, the RCA control specimens (Exp. 10 and Exp. 27) exhibit the stiffest response and the highest resistance among their respective groups, characterized by steeper initial slopes and higher stress levels at comparable displacements. This behavior reflects the relatively rigid nature of recycled concrete particles and their ability to form a load-bearing skeleton under confinement. However, compared with gravel-based columns discussed previously, the overall stiffness and bearing capacity of RCA columns are noticeably lower, indicating reduced inter-particle friction and less effective force-chain development due to the angularity, surface roughness, and internal micro-cracking typically associated with recycled concrete aggregates. The inclusion of recycled rubber markedly alters the mechanical response. Similar to the gravel-based systems, specimens containing 10% rubber (Exp. 7 and Exp. 9; Exp. 24 and Exp. 26) generally demonstrate improved resistance relative to the 20% rubber mixtures (Exp. 6 and Exp. 8; Exp. 23 and Exp. 25). At moderate rubber content, the deformable rubber inclusions appear to enhance stress redistribution and delay localized failure without fully disrupting the RCA skeleton. In contrast, higher rubber content significantly increases compressibility, weakens aggregate-to-aggregate contact, and results in pronounced softening and larger settlements under axial loading. The influence of rubber particle size is also evident within the RCA-based mixtures. Comparisons between 1 × 1 cm and 1 × 2 cm rubber particles at identical rubber contents reveal that larger rubber pieces tend to induce more pronounced deformation and earlier stiffness degradation. This trend suggests that larger rubber particles interrupt the continuity of the RCA framework more severely, reducing the effectiveness of confinement mobilization by the wall-mesh and promoting nonuniform strain localization along the column height. Across both geometric series, the relative ranking of material combinations remains consistent, despite differences in absolute stress and displacement magnitudes. This consistency confirms that, for RCA-based encased columns, material composition—particularly rubber content and particle size—governs the mechanical response more strongly than column dimensions within the tested scale range. Overall, Fig. 5 highlights the inherently softer behavior of RCA systems compared with gravel-based columns and underscores the existence of an optimal rubber content beyond which the load-bearing efficiency of recycled concrete stone columns deteriorates.

Fig. 5.

Stress–displacement responses of wall-mesh–encased stone columns constructed with recycled concrete aggregate and recycled rubber: (a) first series (Exp. 6–10) and (b) second series (Exp. 23–27).

Figure 6 illustrates the stress–displacement responses of wall-mesh–encased stone columns constructed using reclaimed asphalt pavement (RAP) as the primary material, including both RAP-only control columns and RAP–rubber mixtures, tested in two geometric series. Figure 6a presents Tests 11–15 for the smaller columns, while Fig. 6b shows the corresponding Tests 28–32 for the larger columns. As in the previous figures, the paired presentation allows direct comparison between specimens with identical material composition but different column dimensions. Compared with gravel- and RCA-based systems, RAP columns exhibit a noticeably softer mechanical response in both series, characterized by lower initial stiffness and reduced stress levels at comparable displacements. This behavior is primarily attributed to the presence of residual bitumen coating on RAP particles, which reduces effective inter-particle friction and limits the formation of stable force chains under axial loading. Even under geosynthetic confinement, the RAP skeleton mobilizes shear resistance less efficiently, resulting in a more gradual stress–displacement trajectory and earlier onset of deformation. The inclusion of recycled rubber further accentuates this trend. Specimens containing 20% rubber (Exp. 11 and Exp. 13; Exp. 28 and Exp. 30) consistently show the softest response, with pronounced displacement at relatively low stress levels. Increasing rubber content in an already compliant RAP matrix leads to a cumulative reduction in stiffness, as both the bitumen-coated aggregates and the rubber particles act as deformable phases. In contrast, mixtures with 10% rubber (Exp. 12 and Exp. 14; Exp. 29 and Exp. 31) demonstrate marginally improved resistance, although their performance remains inferior to RAP-only control specimens (Exp. 15 and Exp. 32). The effect of rubber particle size is also evident in RAP-based mixtures. Larger rubber particles (1 × 2 cm) tend to induce greater discontinuity within the RAP skeleton, promoting localized deformation and increasing displacement under loading. Smaller rubber particles (1 × 1 cm), while still reducing stiffness, allow relatively better packing and stress redistribution, resulting in slightly steeper stress–displacement curves. This observation underscores the sensitivity of RAP systems to geometric compatibility between aggregate particles and rubber inclusions. Across both series, the relative ranking of material combinations remains consistent, confirming that the influence of RAP and rubber content outweighs the effect of column size within the tested range. Overall, Fig. 6 highlights the limited load-bearing efficiency of RAP-based stone columns, particularly when combined with high rubber content, and suggests that RAP mixtures are better suited for applications where deformation tolerance and sustainability considerations take precedence over maximum bearing capacity.

Fig. 6.

Stress–displacement responses of wall-mesh–encased stone columns constructed with reclaimed asphalt pavement (RAP) and RAP–rubber mixtures: (a) first series (Exp. 11–15) and (b) second series (Exp. 28–32).

Figure 7 presents the stress–displacement responses of wall-mesh–encased stone columns constructed with binary mixtures of gravel and recycled concrete aggregate (RCA), without rubber inclusion. Figure 7a corresponds to Tests 16 and 17 from the first series, and Fig. 7b shows the corresponding Tests 33 and 34 from the second series. In each series, the specimens represent gravel-dominant (80% gravel + 20% RCA) and RCA-dominant (20% gravel + 80% RCA) mixtures, enabling direct evaluation of the effect of blending two granular materials with different stiffness characteristics. In both geometric series, the gravel-dominant columns exhibit higher initial stiffness and greater resistance than the RCA-rich specimens, confirming the positive contribution of natural gravel to load transfer and force-chain development under confinement. However, despite this improved stiffness, both blended configurations display a relatively limited deformation capacity, as evidenced by the abrupt curvature changes and shorter post-peak displacement ranges in the stress–displacement curves. This response indicates low shape-ability (ductility), with failure occurring after a comparatively small increase in displacement once peak resistance is mobilized. When these results are contrasted with the behavior of gravel- or RCA-based columns incorporating a limited proportion of recycled rubber (e.g., 10%), a clear weakness of the gravel–RCA systems becomes evident. While rubber-free blended columns can achieve moderate stiffness, they lack the deformation tolerance observed in rubber-modified mixtures. In rubber-containing columns, the presence of a compliant phase promotes stress redistribution, delays strain localization, and results in a smoother post-peak response. By contrast, the gravel–RCA mixtures rely solely on rigid particle interlocking, making them more prone to brittle-like behavior once confinement is insufficient to accommodate further deformation. Furthermore, the absence of rubber reduces the system’s ability to absorb energy and accommodate differential strains between the column and the surrounding clay. As a result, although gravel–RCA blending offers an improvement over pure RCA columns in terms of stiffness, it remains mechanically inferior to gravel-based mixtures with small rubber additions when both load-bearing capacity and deformation performance are considered together. This comparison highlights that introducing a limited rubber content can be more effective than simply blending two rigid aggregates in enhancing the overall mechanical resilience of encased stone columns. Overall, Fig. 7 demonstrates that gravel–RCA composite columns provide a stiffness-driven improvement but suffer from reduced shape-ability and post-peak deformability. These limitations underscore the importance of incorporating controlled amounts of deformable inclusions, such as recycled rubber, when improved ductility and sustained load-bearing performance are required in ground-improvement applications.

Fig. 7.

Stress–displacement responses of wall-mesh–encased stone columns constructed with gravel–recycled concrete aggregate mixtures (without rubber): (a) first series (Exp. 16–17) and (b) second series (Exp. 33–34).

Influence of materials on ultimate stress and final settlement

The results of the two test series, conducted on specimens with different dimensions but the same length-to-diameter ratio, clearly showed that the type of material and the proportion of recycled constituents were the most decisive factors governing the mechanical performance of stone columns. In both series, columns constructed with gravel combined with a portion of recycled concrete, particularly gravel with 20% recycled concrete, achieved the highest bearing capacity and the lowest settlement. This mixture not only increased ultimate resistance but also minimized excessive deformations, thereby providing stable and reliable performance.

The addition of a small proportion of recycled rubber as part of the aggregate in gravel, concrete, or asphalt mixtures had a noticeable effect in improving bearing capacity and reducing settlement. However, the negative influence of rubber became more pronounced when its content reached 20% by volume; under these conditions, specimens generally exhibited lower mechanical efficiency and greater deformability compared with mixtures without rubber or with smaller amounts. In other words, while the inclusion of 10% rubber in concrete or asphalt occasionally produced a reasonable balance between stiffness and ductility, this benefit was not sustained, and higher rubber content typically resulted in substantial performance losses.

By contrast, specimens composed solely of recycled concrete consistently recorded the lowest bearing capacity and the greatest settlement in both series, underscoring the weakness of this mixture, particularly in applications requiring high strength and stability. Asphalt and asphalt–rubber combinations displayed moderate to relatively good responses, yet they remained consistently inferior to gravel-based mixtures and gravel–concrete blends.

Ratio of maximum stress to final settlement in stone columns

The ratio of maximum stress to final settlement serves as a composite indicator that simultaneously evaluates the mechanical performance of stone columns in terms of both strength (bearing capacity) and deformation (settlement). A high value of this ratio reflects optimal performance, where high load-bearing capacity is accompanied by limited deformation. In contrast, a low ratio typically results from low resistance or excessive settlement, indicating poorer mechanical behavior.

The experimental results revealed that the highest values of this ratio were achieved in mixtures of gravel with recycled rubber, gravel with recycled concrete, and concrete with recycled rubber. In contrast, specimens composed of pure recycled concrete, or concrete with a higher proportion of rubber, generally recorded lower ratios. Moreover, the addition of recycled rubber, particularly at 20% by volume, led to a reduction in this ratio across almost all material groups.

This ratio is a valuable tool for geotechnical engineers, as it helps identify material combinations that achieve the best balance between strength and deformation. Therefore, it can be recommended as a key parameter for selecting optimal mix designs and for guiding both technical and economic evaluations in the application of stone columns.

Discussion of findings

This study was designed to investigate the role and effectiveness of incorporating recycled rubber into the aggregates of stone columns, and to compare its performance with other conventional mixtures. The laboratory findings clearly demonstrated that the type of material and the proportion of recycled constituents, particularly rubber, exert a direct and decisive influence on the mechanical behavior of stone columns, with each combination exhibiting distinct performance characteristics.

One of the most notable outcomes of this research was the improvement observed in the performance index represented by the ratio of maximum stress to final settlement in some mixtures containing a low percentage of rubber. In other words, while adding larger amounts of rubber (20%) generally resulted in reduced bearing capacity and greater settlement, the inclusion of a limited proportion (10%) in gravel, concrete, or asphalt mixtures occasionally produced a desirable balance between stiffness, ductility, and settlement control. This suggests that recycled rubber, when used judiciously, can enhance deformability while maintaining adequate strength, positioning it as an innovative and environmentally responsible approach in geotechnical engineering.

Nevertheless, the results revealed that gravel–concrete mixtures, particularly those with a dominant share of gravel and a smaller proportion of recycled concrete, provided the highest overall performance in terms of both load-bearing capacity and minimal settlement. This combination not only achieved high stiffness, but also limited deformations more effectively compared to other groups, making it a highly suitable option for improving weak soils.

Interestingly, the highest value of the stress-to-settlement ratio in this study was recorded for gravel mixed with a small proportion of rubber. This indicates that the controlled inclusion of rubber in gravel can yield the most favorable balance between ultimate strength and deformability. However, this outcome was not consistent across all mixtures, highlighting the strong dependency of rubber’s performance on the type and proportion of the primary materials.

When compared with previous studies, the present findings align with reports suggesting the positive influence of limited rubber content on the dynamic performance and settlement reduction of geosynthetic-reinforced columns. At the same time, this study further demonstrates that excessive rubber content can impair mechanical behavior. Such discrepancies may stem from variations in gradation, column dimensions, or the intrinsic properties of the base materials, underlining the need for further research to define the optimal limits for rubber use.

From a practical perspective, the results emphasize that the targeted and controlled application of recycled rubber in stone columns can preserve favorable mechanical performance while also contributing to environmental sustainability and waste management. Nevertheless, for critical projects, gravel combined with a small percentage of recycled concrete, or with no more than 10% rubber, should be considered the most reliable option.

Finally, it must be acknowledged that laboratory limitations, such as model scale and the specific type of recycled rubber used, may influence the generalization of the results. Future studies are recommended to explore a broader range of recycled material combinations and to assess long-term field performance, thereby enabling the development of more sustainable and engineering-oriented soil improvement approaches.

Conclusion

This research experimentally evaluated the performance of GESCs constructed with recycled materials, including RCA, RAP, and shredded waste rubber, as a sustainable alternative for ground improvement in soft clay soils. The novelty of the study lies in systematically combining multiple recycled materials with geosynthetic confinement to enhance the mechanical performance of stone columns while reducing environmental impact. A total of 34 laboratory tests were performed on columns with varying material compositions, rubber contents (10% and 20%), and column sizes, maintaining a constant length-to-diameter (L/D) ratio. Ultimate bearing capacity, settlement, and stress-to-displacement ratio were analyzed to assess performance. Based on the comprehensive experimental program and analysis, the key findings of this research are summarized below:

• The results demonstrated that the type and proportion of constituent materials were the primary factors governing the mechanical behavior of geosynthetic-encased stone columns.

• Among all tested mixtures, stone columns with gravel blended with 10% recycled material content exhibited the most favorable overall performance, characterized by higher bearing capacity and reduced settlement compared to conventional gravel columns.

• Introducing a small proportion of recycled rubber (10% by volume) into gravel or RCA mixtures enhanced deformability and improved the stress-to-displacement ratio by up to 16%, indicating that limited rubber inclusion can improve ductility without compromising overall load-bearing performance.

• Increasing the rubber content to 20% resulted in a significant reduction in bearing capacity and a noticeable increase in settlement, identifying a critical threshold beyond which rubber addition negatively impacts the structural performance of stone columns.

• Columns composed entirely of RCA consistently showed the lowest bearing capacity and the greatest settlement, highlighting the necessity of blended mixtures rather than single-material solutions for sustainable and effective soil improvement.

• Stone columns incorporating RAP demonstrated moderate mechanical performance. Although inferior to gravel-based or gravel–RCA combinations, RAP-based mixtures may still be considered for low-load or temporary applications where sustainability objectives take precedence over maximum strength.

The findings confirm that incorporating RCA and controlled rubber content into GESCs is a feasible and environmentally responsible solution for soil improvement. Future research should explore long-term durability under cyclic and seismic loading, field-scale validation, cost-benefit analysis, and integration with advanced modeling techniques to optimize design guidelines and extend practical applications in sustainable geotechnical engineering.

Overall, this study provides a practical and environmentally responsible pathway for integrating recycled aggregates into GESC design, highlighting the substantial potential of reusing waste rubber to reduce material consumption, minimize landfill burden, and advance the broader goals of sustainable geotechnics. From an economic perspective, the incorporation of recycled aggregates can optimize project cost structures by lowering procurement expenditures and reducing dependence on quarried materials, whose extraction and transport represent major cost drivers in ground-improvement works. Moreover, the valorization of waste streams introduces a secondary-resource market that can stabilize supply chains and mitigate price volatility associated with natural aggregates, thereby improving the long-term economic resilience of large-scale geotechnical projects.

This study also carries inherent experimental limitations. The reduced-scale configuration and the selected length-to-diameter ratio provide controlled insight into column behavior but cannot fully replicate field-scale stress conditions. The homogeneous soft clay used in the tests simplifies natural soil variability, and the adjusted aggregate gradation (passing 1/2 inch and retained on 3/8 inch) represents an unavoidable limitation imposed by the reduced-scale configuration. These factors should be taken into account when interpreting and generalizing the results, underscoring the need for future field-scale studies and advanced numerical simulations to validate the findings and more reliably extend their applicability to real-world conditions.

Author contributions

M.H. performed formal analysis, investigation, data curation, experimental development, provided resources, and wrote the original draft. A.Z. contributed to conceptualization, data interpretation, supervision, resources, and writing of the original draft. M.H.R. contributed to supervision and methodology. M.A contributed to Supervision and resources. All authors reviewed the manuscript.

Data availability

All data generated or analyzed during this study are included in this published article. The raw data is also available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.