Consolidation Behavior and Undrained Shear Strength of Soft Soil Reinforced with a Crushed Waste Glass Granular Column

Abstract

Soft soils are characterized by low bearing capacity, high compressibility, and susceptibility to excessive settlement. Granular columns are commonly used to improve such soils; however, conventional column materials such as sand, gravel, and crushed stone are increasingly depleted. As a sustainable alternative, crushed waste glass (CWG) has been identified as a potential granular column material due to its physical and chemical properties being comparable to those of natural aggregates. Despite this potential, limited studies have investigated how key design parameters, such as penetration ratio (PR) and CWG gradation, affect the consolidation behavior and undrained shear strength of reinforced soft soils. This study evaluates the performance of CWG granular columns installed in soft soil represented by kaolin clay. The floating and end-bearing CWG columns with varying gradations were investigated under undrained and consolidation loading conditions. Consolidation and shear strength responses were assessed to quantify the effect of the PR and CWG gradation on soil performance. The results indicate that the CWG column significantly reduces settlement and soil compressibility while improving drainage characteristics. Among the tested configurations, the end-bearing well-graded CWG column provided the greatest improvement, demonstrating a high reduction in total settlement and fast consolidation due to enhanced vertical drainage. These findings highlight the potential of crushed waste glass as an alternative recycled material for granular column reinforcement in soft soil improvement.

1. Introduction

Soft soils, such as clay, have a low bearing capacity, high compressibility, and a tendency to experience significant settlement [1]. Local shear failure planes are also often observed in soft soils underneath embankments, which indicates that soft soils are more sensitive to larger settlements at low bearing loads [2]. Thus, excessive settlement may lead to instability of the embankment and, consequently, structural failure. To alleviate the excessive settlement and enhance the bearing capacity of soft soils, granular columns, which are vertical elements that contain stiffer material aggregates, are embedded into the soft soil using heavy machinery via vibro-displacement, replacement or the rammed method [3,4].

According to Jamal et al. [5] and Guetif et al. [6], the improvement of the soft soil using granular columns can be attributed to three main factors, which are the stiffness of the granular material of the column, the compaction of the surrounding soil, and the function of the column as a vertical drain of the excess pore water that accelerates the consolidation process. Thus, the design of granular column reinforcements, such as the gradation of the granular material and column length, will affect the effectiveness of ground improvement. In the study conducted by Bareither et al. [7], samples with varying material gradations resulted in different friction angles and apparent cohesions, which affected the shear strength of the granular columns. Meanwhile, Najjar et al. [8] explored the effects of varying the penetration ratio (PR) of sand granular columns used for the improvement of the undrained shear strength of the clay. The PR is defined as the ratio of the length penetrated by the granular column to the overall depth of the soft soil bed on top of a hard stratum. Furthermore, it was also inferred that there is an effect on the bearing capacity and settlement of the footings on soft clay reinforced with granular columns if the length-to-diameter (L/d) ratios were varied [9].

Most studies have investigated the design and performance of granular columns constructed using conventional natural aggregates such as sand and gravel. However, the increasing demand for these materials has raised concerns regarding the long-term sustainability of natural aggregate resources. Bendixen et al. [10] mentioned that the most mined materials on Earth are sand, gravel, and crushed stone, resulting in their depletion at a rate that will exceed the rate of natural renewal. Moreover, the continuous extraction of natural aggregates also poses potential environmental and public health risks. According to Tamayo [11], residents living near quarry sites have reported its negative effects, such as the destruction of agricultural lands and respiratory health problems.

In response to these concerns and to effectively recycle waste materials, previous studies have explored sustainable alternatives for granular column materials, including shredded tires, crushed concrete, spent ballast, and crushed waste glass (CWG) [12]. Among these, CWG is particularly promising due to its physical and chemical characteristics that are comparable to those of natural sand. However, only a limited number of studies, such as those by Kazmi et al. [13,14,15], have systematically evaluated its performance as a granular column material. Kazmi et al. [14] demonstrated that end-bearing CWG columns can improve both undrained shear strength and consolidation behavior of soft clay. Their results also highlighted that particle size (gradation) plays a critical role in governing column performance. Coarser CWG particles significantly increased permeability and accelerated consolidation, whereas finer particles enhanced stiffness but were more susceptible to smear effects at the soil–column interface [15]. The smear effect, characterized by a weakened interface zone due to particle intrusion and soil remolding, may reduce load transfer efficiency and compromise column performance.

Despite these insights, existing studies remain limited in two key aspects. First, they predominantly consider uniformly graded CWG and end-bearing column configurations, which do not reflect typical field conditions in which floating columns and mixed gradations are common. Second, the interaction between gradation and penetration ratio (PR), two parameters that directly control drainage efficiency, stress transfer, and failure mechanisms, has not been systematically investigated. Therefore, this study aims to address these gaps by experimental evaluation of floating and end-bearing CWG columns with varying gradations under undrained and consolidation loading conditions.

From an engineering perspective, this limitation is significant. The penetration ratio governs whether the column behaves as end-bearing or floating, directly affecting settlement control and the risk of punching failure. Similarly, gradation determines the balance between permeability and mechanical stability. Poorly graded materials may enhance drainage but increase internal instability, whereas well-graded materials improve stiffness but may reduce drainage capacity. Misjudgment of these parameters can lead to overestimation of bearing capacity, inadequate settlement control, or inefficient drainage design.

This study investigates the behavior of soft soil, represented by kaolin clay, reinforced with crushed waste glass (CWG) granular columns, with particular focus on the combined effects of particle gradation and penetration ratio (PR). The study evaluates both undrained shear strength and consolidation characteristics to capture the short-term and long-term responses of the reinforced system. In contrast to previous studies, which primarily considered end-bearing columns with uniform gradation, this work incorporates both floating and end-bearing configurations and examines well-graded and poorly graded CWG materials. Furthermore, the study analyzes key soil–column interaction mechanisms, including smear effects and potential punching failure, to provide a more mechanistic interpretation of the observed behavior. Thus, CWG column encasements and other confinement techniques were not considered in this study to observe the possible influence of the smear effect.

While classical models, such as Priebe’s method for stone columns [16] and Barron’s radial consolidation theory [17], provide simplified representations of load transfer and drainage behavior, they generally assume idealized material properties and do not explicitly account for variations in column gradation or penetration ratio. In this context, the present study aims to establish quantitative relationships between these design parameters and performance indicators, thereby providing a more rational, physically informed basis for the design and assessment of CWG-reinforced ground systems.

2. Materials

Commercially available dry kaolin powder was used as one of the components of the soft soil to ensure that tests in this study could be consistently replicated with confidence in future studies. Furthermore, there have been several previous studies that have utilized kaolin clay as a soft soil and studied its geotechnical parameters [15,18,19,20,21]. The crushed waste glass (CWG) used in the study was processed from glass bottles at solid waste facilities, which were sorted to ensure that only one type of glass was crushed, thereby eliminating variability in glass quality. The selected relative compaction and relative density of the soft soil and CWG, respectively, is 80% to ensure that the reinforced soil condition is representative of the dense in situ clayey soil.

2.1. Kaolin Clay

The type of the dry kaolin powder is KM20 (Kaolin (Malaysia) Sdn. Bhd., Perak, Malaysia), which has a liquid limit, plastic limit, and plasticity index of 52%, 44%, and 7%, respectively, and a specific gravity of 2.6 [21,22]. The alumina content of KM20 ranges from 17% to 27%, while its silica content ranges from 60% to 70%, which are both obtained using the X-ray fluorescence test method [23]. The maximum dry unit weight of the kaolin clay was determined in accordance with ASTM D698 [24], with a value of 13.81 kN/m³.

The undrained shear strength (S_u) that represents the soft soil condition suitable for the installation of granular column reinforcements ranges from 7 kPa to 50 kPa [25]. In a study by Ambily and Gandhi [26], it was established that varying the clay’s moisture content affects its S_u. Thus, the unconfined compression tests, following ASTM D2166 [27], of kaolin clay with a wide range of moisture content, from 10% to 46%, were conducted in the preliminary phase of this study to determine the moisture content that would represent a soft soil condition with an S_u between 7 and 50 kPa. The average S_u of the specimens and their corresponding moisture content are plotted in Figure 1. The relationship established an inverse relationship wherein an increase in moisture content results in a decrease in S_u. Based on the results, the 11% moisture content was identified as a representative condition that produces an S_u of 9.7 kPa, which falls within the range for soft soils requiring granular column reinforcements. The sensitivity of the results to moisture variation is not explored, which is a limitation of the study.

2.2. Crushed Waste Glass

The CWG was sieved and used to prepare two gradations, well-graded and poorly graded, as shown in Figure 2 and tabulated in

Table 1. Proportion of the well-graded and poorly graded CWG.

Sieve No.	Diameter (mm)	% Passing of CWG
		Well-Graded Distribution	Poorly Graded Distribution
4	4.750	100	100
8	2.360	74	100
10	2.000	55	82
16	1.180	32	37
30	0.600	26	2
40	0.425	20	0
50	0.300	15	0
60	0.250	10	0
100	0.150	4	0
200	0.075	0	0

. Moreover, the effective particle sizes for the two gradations used in the study are tabulated in

Table 2. Effective sizes of the well-graded and poorly graded CWG.

CWG Gradation	Effective Size, D10 (mm)	Coefficient of Uniformity	Coefficient of Curvature
Well-graded	0.2250	9.3099	2.0655
Poorly graded	0.7326	2.1829	0.9664

. The percentage of soil passing a specific sieve, coefficient of uniformity, and coefficient of curvature for a well-graded distribution were based on the study of Hamidi et al. [28], which investigated the impact of gradation on the shear strength of well-graded sand–gravel mixtures. The well-graded distribution of granular materials ensures a wide range of particle sizes, with the larger void spaces filled by smaller particles, thereby reducing the void-to-solid ratio. Meanwhile, the poorly graded distribution of the CWG column material, characterized by particles that are uniform in size with minimal variation that ranges from around 1.0 mm to 1.7 mm, was based on the medium CWG particle size in the study of Kazmi et al. [15]. Since the variation in CWG particle sizes is smaller in a poorly graded distribution, the void spaces in between the uniformly sized particles are not occupied due to the lack of smaller particles that could occupy the voids.

The maximum and minimum dry unit weights of the well-graded and poorly graded CWG were determined using ASTM D4253 [29] and ASTM D4254 [30], respectively. The results, summarized in

Table 3. Maximum and minimum dry unit weights of the well-graded and poorly graded CWG.

Dry Unit Weight	Well-Graded Distribution	Poorly Graded Distribution
Minimum Dry Unit Weight, γdmin (kN/m3)	14.14	12.54
Maximum Dry Unit Weight, γdmax (kN/m3)	17.28	14.81

, were used to calculate the target dry unit weight used for specimen preparation, such that the CWG column samples were compacted to a relative density of 80%.

To ensure that CWG is a permeable material suitable for replacing sand in a granular column, its hydraulic conductivity (k) was determined using the constant-head method (ASTM D2434 [31]). Both the well-graded and poorly graded CWG were compacted to yield a relative density of 80%.

Table 4. Permeability of well-graded and poorly graded CWG columns.

CWG Gradation	Hydraulic Conductivity, k (cm/s)	Degree of Permeability
Well-graded distribution	1.71 × 10−2	Medium
Poorly graded distribution	4.34 × 10−2	Medium

shows the k values of the well-graded and poorly graded CWG, which both have a medium degree of permeability and are comparable to clean sands and fine sands [32,33]. Thus, the drainage potential of the CWG column is similar to that of a conventional granular column composed of sand. It can also be observed that the poorly graded CWG has a higher k due to its larger void spaces. This indicates that the uniformly arranged particles in a poorly graded distribution resulted in a more connected drainage channel. On the other hand, the well-graded CWG exhibited a lower k, indicating a more complex drainage path and reduced overall permeability of the CWG column.

3. Methodology

Unconfined compression and one-dimensional consolidation tests were conducted in accordance with ASTM D2166 [27] and ASTM D2435 [34], respectively, to evaluate the undrained shear strength and consolidation behavior of the specimens. Two specimen configurations were prepared, soft soil without reinforcement and soft soil reinforced with a singular crushed waste glass (CWG) column. The reinforced specimens were designed based on the selected area replacement ratio and penetration ratio (PR) of the CWG column. To accommodate the installation of the CWG column while maintaining the selected area replacement ratio and penetration ratio (PR), the specimen dimensions were increased relative to the standard dimensions specified in the ASTM procedures. Similar modifications have also been adopted in previous laboratory investigations on granular column-reinforced soft soils [8,15,26]. Dimensional scaling was implemented to preserve the geometric relationships between the column and the surrounding soil, which are critical for simulating the stress transfer and deformation mechanisms of the composite soil–column system. By maintaining consistent geometric proportions, such as the area replacement ratio and the length-to-diameter (L/d) ratio, boundary effects were minimized, and comparable failure mechanisms were observed in the laboratory-scale specimens. It is acknowledged that laboratory-scale physical models cannot completely reproduce the in situ stress conditions and scale-dependent behavior observed in field applications, which remains an inherent limitation of small-scale experimental studies on granular column reinforcement.

3.1. Sample Preparation

The specimen dimensions for each test are presented in

Table 5. Dimensions of the soft soil and the CWG column.

Test	Soft Soil Diameter, D (mm)	Soft Soil Height, H (mm)	CWG Column Diameter (mm)	Penetration Ratio (PR)	Distance from the Hard Stratum (mm)	Length-to- Diameter Ratio
Unconfined Compression Test	100	219.5	33	1.0	0	6.65
				0.75	54.88	4.99
				0.5	109.75	3.33
One-dimensional Consolidation Test	151.7	178	50	1.0	0	3.54
				0.5	88.5	1.77

. The area replacement ratio was maintained at 11%, as in a previous study investigating specimens reinforced with end-bearing and long floating columns [9]. The diameter of the CWG column was calculated such that the column area is 11% of the total cross-sectional area of the soft soil specimen. The PR was defined relative to the length of the soft soil specimen, which was assumed to rest on a firm base representing the hard stratum. Both the area replacement ratio and the PR influence the resulting length-to-diameter (L/d) ratio of the CWG column. This L/d ratio is an important parameter commonly considered in the design of granular columns for field applications.

According to Gaber et al. [35], who analyzed the performance of a singular granular column as soft soil reinforcement, the center-to-center spacing, represented by the diameter of the soft soil specimen, should not be greater than three to four times the column diameter to avoid a low improvement of the soft soil. Therefore, in this study, the diameter of all soft soil specimens was ensured to be less than four times the corresponding column diameter tabulated in Table 5. The PRs used in this study were based on Najjar et al. [8], which evaluated the effect of PRs on the shear strength of reinforced soil in sand columns. The schematic diagram showing the PRs is illustrated in Figure 3, which varies from 0.5 to 1.0 in relation to the height of the soft soil specimen. Thus, the PR of 1.0 indicates that the CWG column is an end-bearing column wherein it fully penetrates the kaolin clay, ensuring contact with the hard stratum. A PR of 0.75 corresponds to a long floating CWG column that partially penetrates the soil, while a PR of 0.5 corresponds to a short floating column. The PR of the floating columns was varied to observe possible punching failure during the unconfined compression test.

The specimen without reinforcement at a moisture content of 11%, referred to as soft soil only (SSO), was compacted to a relative compaction of 80% to compare it with the specimens reinforced with a CWG column. To prepare the reinforced specimens, their molds were secured to a square wooden base by placing nails around the outside perimeter of the molds. The square wooden block has a central circular cutout that holds the column mold in place, as shown in Figure 4. After the kaolin clay is compacted around the column mold using a wooden tamper, the column mold is removed through the circular cutout at the base.

To prepare the specimen reinforced with an end-bearing CWG column, the kaolin clay was compacted up to the top of the mold before the column mold was removed. For the specimen reinforced with a floating column, the column mold was removed through the base cutout once the compacted kaolin clay reached the required column height for the specific PR. The removal of the column mold ensures the formation of a true composite specimen, promoting proper interaction between the CWG and the surrounding kaolin clay without any separation or gaps. The CWG particles were manually compacted in layers to achieve a target relative density of 80%. Although this simplified preparation method does not fully replicate field vibro-compaction procedures, it was adopted to provide controlled, repeatable specimen preparation suitable for laboratory-scale evaluation of soil–column interaction mechanisms.

Geosynthetic encasement and other confinement techniques were intentionally excluded from the experimental program to isolate the intrinsic behavior of the CWG column–soil system and directly evaluate the influence of CWG gradation and penetration ratio on the consolidation and shear response of the reinforced specimens. The uncased configuration also enabled clearer observation of mechanisms such as bulging deformation, smear effects, and potential clogging at the soil–column interface without additional restraint from external confinement. Although geosynthetic encasement is commonly adopted in field applications to improve lateral confinement and column stability, its exclusion in the present study was considered necessary to establish the baseline behavior of uncased CWG columns. This simplification is recognized as a limitation of the experimental setup and may affect the direct representation of field conditions.

3.2. Unconfined Compression Test

The unconfined SSO specimen and reinforced specimens were placed on the platform of the electromechanical universal testing machine, wherein an axial load was applied, inducing an axial strain at a rate of 0.5% to 2% per minute. The unconfined compressive strength (UCS) was determined from the stress–strain curve as the stress at 15% axial strain or the peak stress, whichever occurred first. The undrained shear strength was also calculated as half of the UCS. Once testing was concluded, the specimens were cut in half along their longitudinal axis to examine column deformation and assess the failure mode that occurred during testing.

3.3. One-Dimensional Consolidation Test

The one-dimensional consolidation test was used to observe the effect of the CWG column on the consolidation behavior of kaolin clay. The equipment was modified to accommodate a larger specimen measuring 178 mm in height, as shown in Figure 5. The loading frame’s arms were replaced with longer arms (355 mm) to secure the specimen’s top cap, a wooden disk covered with plastic that evenly distributes the applied load. In addition, the post supporting the vertical displacement transducer, which measures deformation, was replaced with a 330 mm post to allow the transducer to reach the upper portion of the loading frame. Before the specimens were placed in the modified front-loading oedometer, they were fully saturated and submerged in water to maintain constant saturation throughout the test. The mold was placed on a wooden block with a circular cutout at its center, allowing water to drain from the bottom of the sample.

A seating pressure of 5 kPa was first applied to each specimen for 1 min to ensure proper contact between the specimen, porous stones, and loading components. The deformation indicator was then adjusted to zero to establish the initial reference reading. During the loading stage, vertical stresses of 12.5 kPa, 25 kPa, 50 kPa, and 100 kPa were applied sequentially to the specimen. Each load increment was maintained for 24 h to allow consolidation. After completion of the loading stage, the unloading stage was carried out by removing the applied loads in reverse order. The stresses were reduced incrementally to 50 kPa, 25 kPa, and 12.5 kPa, with each unloading step maintained for 6 h. Following unloading, a reloading stage was performed to evaluate the recompression behavior of the soil. Vertical stresses of 25 kPa, 50 kPa, and 100 kPa were again applied sequentially to the specimen, with each stress level maintained for 24 h. Throughout all stages of the test, vertical deformation was continuously recorded at 6 s intervals (0.1 min).

The deformation readings obtained from the one-dimensional consolidation tests were plotted against the square root of time to determine the coefficient of consolidation. This is in accordance with Taylor’s square root of time method due to its focus on the initial portion of the curve, since the obtained data from the study is concentrated on the said portion. Moreover, the plot of void ratio against the logarithm of pressure was also used to obtain the λ and κ coefficients which were the slopes of the best-fit lines for the loading and unloading curves, respectively. The λ and κ coefficients were used to determine the compression index, recompression index, and preconsolidation pressure following the Dissipated Strain Energy Method [36]. The coefficient of volume compressibility of the specimens was also determined directly from the consolidation test data. The hydraulic conductivity of the SSO and reinforced specimens was indirectly determined based on Terzaghi’s [37] theory of consolidation using their corresponding coefficient of consolidation and coefficient of volume compressibility, as shown in Equation (1).

$$k=C_v \times m_v \times 0.0000163$$

(1)

where k is the hydraulic conductivity in cm/s, C_v is the coefficient of consolidation in mm²/min, and m_v is the coefficient of volume compressibility in 1/kPa.

4. Results and Discussion

4.1. Unconfined Compressive Strength and Undrained Shear Strength

4.1.1. Stress–Strain Behavior of Reinforced Specimens

The stress–strain responses of the specimens obtained from the unconfined compression tests are presented in Figure 6. The specimens were reinforced with a crushed waste glass (CWG) column having penetration ratios (PRs) of 0.5, 0.75, and 1.0, and gradation conditions classified as well-graded (W) and poorly graded (P).

The results show that specimens reinforced with a floating CWG column (PR = 0.5 and 0.75) exhibited greater initial stiffness, as indicated by the steeper initial slope of the stress–strain curves, compared with the soft soil only (SSO) specimens and specimens reinforced with an end-bearing column (PR = 1.0). This indicates that floating columns can mobilize resistance at lower strain levels, thereby improving the deformation resistance of the unconfined reinforced soil during the early stages of loading. Similar behavior of reinforced soil was reported, wherein the interaction between the granular column and surrounding soil enhances stiffness and redistributes stresses within the composite system [23].

The gradation of the CWG material also influenced the stiffness of the reinforced specimens. Specimens reinforced with a well-graded CWG column generally exhibited slightly higher stiffness than those reinforced with a poorly graded CWG column, as shown in Figure 6. This behavior may be attributed to the improved particle interlocking and packing density of the well-graded CWG. Between the two CWG gradations, the column with a well-graded distribution has smaller voids, evidenced by its smaller permeability. Previous studies have similarly reported that well-graded granular materials, with a smaller amount of voids, provide improved mechanical stability due to the increased number of particle contact points [33,38]. Furthermore, the angularity of the in-contact CWG particles improves their interlocking, enhancing internal friction and load transfer within the granular column reinforcement.

4.1.2. Undrained Shear Strength

The unconfined compressive strength (UCS) and the corresponding undrained shear strength (S_u) are summarized in

Table 6. Summary of the results from the unconfined compression tests.

Specimen	Description	Unconfined Compressive Strength, qu (kPa)	Undrained Shear Strength, Su (kPa)
SSO	Specimen with no reinforcement	67.854	33.927
PR = 1.0, W	Specimen reinforced with an end-bearing well-graded CWG column	38.572	19.286
PR = 1.0, P	Specimen reinforced with an end-bearing poorly graded CWG column	35.303	17.652
PR = 0.75, W	Specimen reinforced with a long floating well-graded CWG column	47.575	23.787
PR = 0.75, P	Specimen reinforced with a long floating poorly graded CWG column	45.304	22.652
PR = 0.5, W	Specimen reinforced with a short floating well-graded CWG column	46.639	23.320
PR = 0.5, P	Specimen reinforced with a short floating poorly graded CWG column	44.017	22.008

. The SSO specimen exhibited the highest UCS of 67.854 kPa, which may be attributed to the cohesive nature of the kaolin clay, which enables a relatively uniform stress distribution under axial loading. In contrast, reinforced specimens contain granular inclusions that introduce heterogeneity within the soil matrix and may promote localized deformation.

Consequently, since S_u is approximately half of the UCS, the SSO specimen also exhibited the highest S_u (33.927 kPa). This observation differs from the findings of Kazmi et al. [15], who reported an increase in shear strength with the inclusion of CWG columns. However, it should be noted that Kazmi et al. [15] evaluated specimens under drained conditions, where the frictional properties of the granular column significantly contribute to load resistance. Under drained conditions, the applied loading provides sufficient time for excess pore water pressure to dissipate, and the shear strength is primarily governed by effective stress parameters. In contrast, the present study focuses on unconfined, undrained loading conditions, which represent short-term behavior in which loading is applied rapidly, preventing pore water dissipation and leading to the development of excess pore water pressure.

Among the reinforced specimens, the highest shear strength was observed in the reinforced specimen with a long floating CWG column (PR = 0.75, W), followed closely by the short floating column configuration (PR = 0.5, W). In contrast, the lowest load-carrying capacity was observed in specimens reinforced with an end-bearing CWG column, particularly those with poorly graded CWG (PR = 1.0, P). The lower S_u observed in the reinforced specimens may be attributed to replacing the cohesive kaolin clay with the cohesionless CWG, thereby reducing the overall cohesion of the specimens under undrained loading conditions. Consequently, the short-term stability of the reinforced specimens remains primarily governed by the cohesive properties of the surrounding clay. A study conducted by Najjar et al. [8] also established the S_u of a soft soil reinforced with a sand column and found improved bearing capacity. The results of this study show an opposite pattern, with the inclusion of soil reinforcement reducing S_u. This observation may be attributed to the lack of lateral confining pressure of the CWG-reinforced specimens during the unconfined compression test, since Najjar et al. [8] applied deviatoric pressure during their consolidated undrained triaxial tests. Thus, it can be inferred that the load-carrying capacity of the granular column also depends on the lateral support provided to the surrounding soil.

The reduction in undrained shear strength observed in the reinforced specimens can therefore be attributed to two factors. First, replacing cohesive clay with cohesionless CWG reduces the overall apparent cohesion of the composite system under undrained conditions. Second, the absence of lateral confinement in the unconfined compression test limits the mobilization of frictional resistance within the granular column. As a result, the contribution of the CWG column to shear strength is not fully developed under these conditions. In contrast, under long-term drained conditions, granular columns are expected to enhance soil performance by increasing effective stress and improving drainage capacity. Therefore, the results indicate that CWG columns primarily contribute to long-term performance improvements, such as settlement reduction and consolidation acceleration, rather than short-term undrained shear strength.

4.1.3. Influence of Column Configuration and Gradation

Specimens reinforced with a well-graded CWG column consistently exhibited slightly higher S_u of 1 kPa to 1.5 kPa than those reinforced with a poorly graded CWG column. This difference may be attributed to the greater particle contact and slightly improved particle interlocking within well-graded CWG, which enhanced the internal stability of the CWG column. In contrast, poorly graded CWG columns tend to have higher void ratios, allowing greater particle rearrangement during loading and resulting in weaker lateral resistance.

The observations considering the PRs are consistent with the findings of Hasan et al. [39], who reported that specimens reinforced with floating columns exhibited improved shear strength compared with specimens reinforced with end-bearing columns. The findings also show that the average S_u of the specimens reinforced with a long and short floating CWG column is relatively small. Thus, the two PRs, 0.5 and 0.75, may be considered as a single category of a floating CWG column. The ranking of the specimens based on their average undrained shear strength is summarized in

Table 7. Undrained shear strength ranking of the specimens.

Rank	Specimen Description	Average Undrained Shear Strength
1 (highest)	Specimen with no reinforcement	33.927 kPa
2	Specimen reinforced with a floating CWG column	22.942 kPa
3 (lowest)	Specimen reinforced with an end-bearing CWG column	18.469 kPa

, where the SSO specimen exhibited the highest S_u, followed by specimens reinforced with a floating CWG column and specimens reinforced with an end-bearing CWG column, respectively.

4.1.4. Deformation and Failure Mechanisms

To further understand the mechanisms governing the strength behavior of the reinforced specimens, their deformation was observed during unconfined compression tests, which predominantly exhibited bulging. It is inferred that the limited confinement provided by the soft clay to the CWG column during axial loading caused bulging deformation, limiting the column’s ability to sustain higher loads [40]. It was also observed that the PR significantly influenced the extent of bulging, as shown in Figure 7. The most pronounced bulging occurred in the end-bearing column, where the lateral expansion reached approximately 1.5 mm. Specimens reinforced with a long floating column exhibited moderate bulging ranging from 0.5 mm to 1.0 mm, whereas very minimal bulging was observed in specimens reinforced with a short floating column. On the other hand, the SSO specimen did not experience bulging and maintained its structural integrity under loading.

The location of bulging in specimens reinforced with end-bearing and long floating columns occurred at depths ranging from approximately 1.5 to 2 times the column diameter from the top of the specimen, which is consistent with the typical bulging pattern reported in granular column studies [41,42]. Furthermore, no significant difference in bulging was observed between well-graded and poorly graded CWG columns, suggesting that gradation had minimal influence on the bulging behavior under the test conditions. In addition, no punching failure was observed in specimens reinforced with a floating CWG column during the unconfined compression test.

Cracks also developed at the top or bottom of the specimens, eventually forming a single vertical split or diagonal shear failure mode, as shown in Figure 8. This behavior was observed in all reinforced specimens, but was more pronounced in specimens reinforced with a short floating column. Although shear cracks developed in the soft soil matrix of the reinforced specimens, the CWG columns themselves remained structurally intact, indicating that the column material did not undergo crushing or internal failure. The microcracks observed in specimens reinforced with an end-bearing and a long floating CWG column may be attributed to lateral expansion of the column during loading, which generates radial stresses in the surrounding soil and leads to localized cracking.

4.1.5. Statistical Analysis and Empirical Prediction Model

To further quantify the influence of column geometry on shear strength, the experimental results on undrained shear strength against the length-to-diameter (L/d) ratio were analyzed using ANOVA at a 95% confidence level. The results showed that the L/d ratio, with a p-value of 0.0044, had a statistically significant effect on S_u, whereas CWG gradation did not, as its p-value was greater than 0.05. Based on these relationships, empirical equations, presented in Equations (2) and (3), were developed to estimate the undrained shear strength of the reinforced specimens without lateral confinement as a function of the L/d ratio explored within this study.

$$S_{u,W}=5.98794+8.39403 \times \left(\mathrm{L}/\mathrm{d} \mathrm{ratio}\right)-0.96282 \times { \left(\mathrm{L}/\mathrm{d} \mathrm{ratio}\right)}^2$$

(2)

$$S_{u,P}=5.11352 +8.29663 \times \left(\mathrm{L}/\mathrm{d} \mathrm{ratio}\right)-0.96282 \times { \left(\mathrm{L}/\mathrm{d} \mathrm{ratio}\right)}^2$$

(3)

where S_u,W is the undrained shear strength of specimens reinforced with a well-graded CWG column in kPa, S_u,P is the undrained shear strength of specimens reinforced with a poorly graded CWG column in kPa, and L/d ratio is the length-to-diameter ratio.

The quadratic form of these equations indicates that S_u initially increases with increasing L/d ratio but decreases beyond a certain value, suggesting the existence of an optimal column length relative to its diameter that maximizes soft soil reinforcement. The positive linear coefficient represents the strengthening contribution of increasing column length, whereas the negative quadratic term indicates diminishing improvement at larger L/d ratios.

Overall, the developed empirical relationships provide a useful predictive tool for estimating the undrained shear strength of soft soils reinforced with a CWG column, as determined by unconfined compression tests, accounting for the influence of column geometry (L/d ratio). The developed empirical relationships provide predictive estimates for the specific range of penetration ratios, gradations, and specimen conditions investigated in this study. Therefore, the proposed equations should be applied only within the experimental domain considered herein unless further validated using additional laboratory or field data.

4.2. Consolidation Behavior

The one-dimensional consolidation tests were conducted on specimens reinforced with a floating column and an end-bearing column to determine the effect of the crushed waste glass (CWG) column length, specifically the presence or absence of contact with the hard stratum, on the consolidation behavior of the reinforced soil. The PR = 0.75 condition was excluded from the consolidation tests to limit experimental duration and complexity. This decision was supported by the unconfined compression results, which showed minimal difference in behavior between PR = 0.75 and PR = 0.5. Therefore, PR = 0.5 was selected as representative of floating column behavior. However, this simplification is acknowledged as a limitation, and future studies should include intermediate PR values for a more comprehensive assessment.

The consolidation tests further evaluate how the CWG column, which provides a stiffer load-carrying element and a preferential drainage path, affects the total settlement, compression index (C_c), recompression index (C_r), preconsolidation pressure (σ′_c), coefficient of consolidation (C_v), and coefficient of volume compressibility (m_v) of the reinforced kaolin clay.

4.2.1. Total Settlement

The total settlement of the specimens is presented in Figure 9. The results indicate that all reinforced specimens exhibited lower total settlement compared with the soft soil only (SSO) specimen, demonstrating the effectiveness of the CWG column in improving the stiffness and stability of soft soils. According to Indraratna et al. [43] and Grizi et al. [44], granular columns with higher stiffness carry a larger share of the applied load than the surrounding soft soil, a phenomenon known as the arching effect. Thus, in specimens reinforced with an end-bearing column, the larger share of the stress applied is carried by the CWG column, in which the stress is transferred directly to the hard stratum. This limits the settlement of the reinforced soil since a smaller share of the stress is carried by the surrounding soft soil.

Among the reinforced specimens, the poorly graded CWG column with end bearing exhibited the lowest total settlement, indicating greater resistance to compression under vertical loading. This suggests that the column effectively transferred a portion of the applied stress to the underlying hard stratum, thereby reducing deformation of the surrounding soft soil. The relatively low settlement observed in the specimen reinforced with a poorly graded CWG end-bearing column. This behavior may be associated with the clogging mechanisms discussed in Section 4.3. The accumulation of clay particles within the column reduces its permeability and increases particle contact among CWG particles, temporarily enhancing the column’s stiffness and limiting compressive deformation. Similar clogging behavior in granular columns has been reported by Tai and Zhou [3].

In contrast, the specimen reinforced with a floating poorly graded CWG column exhibited the largest settlement among the reinforced specimens. This behavior may be attributed to reduced particle contact in poorly graded aggregates, which decreases the column’s internal stability. Furthermore, a floating column transfers the larger share of the applied stress it carries to the underlying soft soil, which possesses lower stiffness and higher compressibility. Consequently, settlement reduction becomes less effective as the PR decreases, consistent with the findings of Ng and Tan [45]. Similar observations were also reported by Al-Kazzaz and Al-Obaydi [46], who noted that short floating columns are less effective in reducing settlement compared with end-bearing columns. Overall, the results indicate that the penetration depth of the CWG column plays a critical role in controlling settlement behavior, with end-bearing columns providing more effective settlement reduction than floating columns.

4.2.2. Compressibility Parameters

The compression index (C_c), recompression index (C_r), and preconsolidation pressure (σ′_c), tabulated in

Table 8. Compression index, recompression index, and preconsolidation pressure of the specimens.

Specimen	Description	Trial	Compression Index, Cc	Recompression Index, Cr	Preconsolidation Pressure, σ′c
SSO	Specimen with no reinforcement	1	0.23046	0.07774	28.207 kPa
		2	0.23092	0.04163	22.219 kPa
PR = 0.5, P	Specimen reinforced with a floating poorly graded CWG column	1	0.28888	0.03496	29.417 kPa
		2	0.29256	0.03519	29.404 kPa
PR = 0.5, W	Specimen reinforced with a floating well-graded CWG column	1	0.19067	0.03289	29.882 kPa
		2	0.14674	0.02277	29.538 kPa
PR = 1.0, P	Specimen reinforced with an end-bearing poorly graded CWG column	1	0.07199	0.02507	39.230 kPa
		2	0.11822	0.02852	35.876 kPa
PR = 1.0, W	Specimen reinforced with an end-bearing well-graded CWG column	1	0.13547	0.05566	34.552 kPa
		2	0.07981	0.02921	34.511 kPa

, were determined to evaluate the compressibility characteristics and stiffness of the reinforced soil system. A negative relationship between σ′_c and C_c was observed, indicating that specimens with higher σ′_c exhibit lower compressibility. The SSO specimen showed the lowest σ′_c and highest C_c, confirming the greater compressibility of the unreinforced kaolin clay. In contrast, specimens reinforced with an end-bearing poorly graded CWG column exhibited the highest σ′_c and the lowest C_c values. This behavior indicates that the CWG column improves the stiffness of the composite soil system and enhances its resistance to plastic deformation under increasing vertical stress.

The analysis of variance (ANOVA) with a confidence level of 95% indicated that the models for σ′_c, represented by Equations (4) and (5), are statistically significant, which denotes that the models can effectively determine the σ′_c for a specific length-to-diameter (L/d) ratio explored within the study. Furthermore, it was also determined that the factor, L/d ratio, with a p-value of 0.0017, had a significant effect on the σ′_c while the CWG gradation did not have a significant effect on the σ′_c. This suggests that variations in penetration ratio (PR), and consequently the L/d ratio, have a substantial influence on the preconsolidation pressure of the reinforced specimens.

$${\sigma }_{c,W}^{\prime }=22.39811 +3.66205 \times \left(\mathrm{L}/\mathrm{d} \mathrm{ratio}\right)$$

(4)

$${\sigma }_{c,P}^{\prime }=23.75916 +3.66205 \times \left(\mathrm{L}/\mathrm{d} \mathrm{ratio}\right)$$

(5)

where σ′_c,W is the preconsolidation pressure of specimens reinforced with a well-graded CWG column in kPa, σ′_c,P is the preconsolidation pressure of specimens reinforced with a poorly graded CWG column in kPa, and L/d ratio is the length-to-diameter ratio.

The empirical model shows that an increase in PR enhances lateral confinement and resistance to plastic deformation by increasing the interaction between the CWG column and the surrounding soil. Similar observations have been reported by Zhai et al. [47] and Hafez et al. [48], who noted that deeper granular columns improve stress transfer and increase the stiffness of the composite soil system. The almost identical slope observed in Equations (4) and (5) indicates that the L/d ratio consistently increases the σ′_c regardless of the CWG gradation. Although the gradation of the CWG material did not have a statistically significant effect on the σ′_c, the slight difference in the intercepts of the linear models suggests that gradation may still influence the baseline strength of the reinforced soil.

4.2.3. Coefficient of Consolidation

The coefficient of consolidation (C_v) provides insight into the rate of excess pore water pressure dissipation within the specimen. Therefore, a higher C_v indicates improved drainage and faster soil mass stabilization. The variation in C_v with the vertical stress is presented in Figure 10. For the SSO specimen, C_v generally decreases with increasing applied vertical stress, consistent with the typical consolidation behavior of cohesive soils, as the reduction in void ratio slows the dissipation of excess pore water pressure [49].

An inflection point near the σ′_c was observed in the C_v curves of specimens reinforced with a CWG column, as shown in Figure 10. This inflection point indicates a transition from elastic compression to plastic deformation of the reinforced kaolin clay, consistent with observations reported in previous studies [36,50]. Beyond this transition, specimens reinforced with a CWG column exhibited an increase in C_v with increasing vertical stress, indicating that the CWG column provides both a stiffer load-carrying element and a preferential drainage path, allowing excess pore water pressure to dissipate more rapidly than in the unreinforced soil [51,52].

For specimens reinforced with an end-bearing CWG column, the C_v continuously increased as the applied vertical stress increased beyond the transition point. This behavior may be attributed to the enhanced load-carrying capacity of the column–soil system as the additional stiffness provided by the underlying hard stratum increases with the increase in applied stress [53]. As a result, the reinforced system becomes more effective in dissipating excess pore water pressure under higher stress levels. In addition, the relatively high C_v observed in the specimen reinforced with an end-bearing, poorly graded CWG column may also be influenced by potential clogging discussed in Section 4.3, allowing the column to carry a larger portion of the applied vertical stress due to increased particle contact. This reduces the generation of excess pore water pressure in the surrounding soil, resulting in a faster apparent rate of consolidation.

A different behavior was observed for specimens reinforced with a floating CWG column. After the transition near σ′_c, C_v initially increased with increasing vertical stress, indicating that the floating CWG column may have gained additional stiffness and contributed to improved drainage within the reinforced layer. However, the C_v eventually reached a peak value followed by a downward trend, suggesting that once the reinforced layer becomes fully consolidated, further consolidation is governed by the underlying unreinforced soft soil layer [45]. The delayed peak observed in specimens reinforced with a floating poorly graded CWG column may be attributed to the larger voids between CWG particles, which enhanced vertical drainage, compared with the well-graded column.

The ANOVA, with a confidence level of 95%, indicated that the column PR in the study had a statistically significant effect on C_v (p-value < 0.0001), as shown by the empirical models in Equations (6)–(9). In contrast, the CWG gradation did not have a statistically significant effect on the C_v.

$$C_{v,1.0,W}=3909.61098-7080.83286 \times \left(\log \sigma \right)+3452.21811 \times { \left(\log \sigma \right)}^2$$

(6)

$$C_{v,1.0,P}=1341.83318-1630.90136 \times \left(\log \sigma \right)+3452.21811 \times {\left(\log \sigma \right)}^2$$

(7)

$$C_{v,0.5,W}=10755.93387-11983.33870 \times \left(\log \sigma \right)+3452.21811 \times {\left(\log \sigma \right)}^2$$

(8)

$$C_{v,0.5,P}=6999.4214-9533.40719 \times \left(\log \sigma \right)+3452.21811 \times {\left(\log \sigma \right)}^2$$

(9)

where C_v,1.0,W is the coefficient of consolidation of specimens reinforced with an end-bearing well-graded CWG column in mm²/min, C_v,1.0,P is the coefficient of consolidation of specimens reinforced with an end-bearing poorly graded CWG column in mm²/min, C_v,0.5,W is the coefficient of consolidation of specimens reinforced with a floating well-graded CWG column in mm²/min, C_v,0.5,P is the coefficient of consolidation of specimens reinforced with a floating poorly graded CWG column in mm²/min, and σ is the applied vertical stress in kPa from 12.5 kPa to 100 kPa.

The proposed empirical relationships for the coefficient of consolidation are applicable only to the range of vertical stresses, penetration ratios, and specimen conditions considered in this study.

4.2.4. Coefficient of Volume Compressibility

The coefficient of volume compressibility (m_v), presented in

Table 9. Coefficient of volume compressibility of the specimens.

Vertical Stress, σ (kPa)	Coefficient of Volume Compressibility, mv (1/kPa)
	SSO	PR = 0.5, P	PR = 0.5, W	PR = 1.0, P	PR = 1.0, W
12.5	1.54 × 10−3	1.085 × 10−5	1.289 × 10−3	2.260 × 10−6	1.441 × 10−3
25	4.16 × 10−3	2.983 × 10−5	2.008 × 10−3	1.944 × 10−4	2.483 × 10−3
50	4.01 × 10−3	1.868 × 10−3	1.533 × 10−3	5.145 × 10−4	2.038 × 10−3
100	3.59 × 10−3	2.511 × 10−3	2.101 × 10−3	6.808 × 10−4	1.854 × 10−3

, represents the change in soil volume per unit increase in effective stress and is commonly used to evaluate soil compressibility during consolidation. Since m_v is inversely related to soil stiffness, lower m_v indicates a stiffer soil system with reduced susceptibility to volumetric deformation.

Among all specimens, the SSO specimen exhibited the highest m_v, as expected, because the soft soil carries the entire applied vertical stress, leading to greater volumetric compression during consolidation. In contrast, specimens reinforced with a CWG column exhibited lower m_v, indicating that the column increased the composite soil system’s stiffness and reduced its susceptibility to volumetric compression. Among the reinforced specimens, the one with an end-bearing poorly graded CWG column exhibited the lowest m_v, which may also be related to the potential clogging of the poorly graded CWG column. Furthermore, the end-bearing CWG column directly transfers a larger portion of the applied vertical stress to the underlying hard stratum, thereby limiting volumetric compression of the surrounding clay.

Overall, the results indicate that CWG columns do not significantly enhance short-term undrained shear strength due to the absence of drainage and confinement. However, under long-term drained conditions, the columns improve soil performance by accelerating consolidation and enhancing drainage capacity. The contrasting behavior observed between the unconfined compression and consolidation tests highlights the importance of loading and drainage conditions in evaluating the effectiveness of CWG columns. Under short-term undrained conditions, the absence of drainage and lateral confinement limits the mobilization of frictional resistance within the CWG column, resulting in lower undrained shear strength. However, under long-term drained conditions, the CWG column functions as both a stiff inclusion and a preferential drainage path, improving consolidation behavior and reducing settlement. Therefore, the effectiveness of CWG columns is more pronounced in applications governed by long-term serviceability and consolidation performance rather than immediate undrained strength.

4.3. Permeability of the Reinforced Soft Soil

The improved consolidation performance observed in the reinforced specimens is closely related to the drainage characteristics of the soil–column system. Thus, it is necessary to evaluate how the crushed waste glass (CWG) column affects the coefficient of permeability or hydraulic conductivity (k) of the reinforced soil mass. The resulting k after the loading stage of the applied vertical stresses is presented in

Table 10. Hydraulic conductivity of the specimens.

Specimen	Description	Hydraulic Conductivity, k
SSO	Specimen with no reinforcement	5.34 × 10−6 cm/s
PR = 0.5, P	Specimen reinforced with a floating poorly graded CWG column	8.18 × 10−5 cm/s
PR = 0.5, W	Specimen reinforced with a floating well-graded CWG column	8.14 × 10−6 cm/s
PR = 1.0, P	Specimen reinforced with an end-bearing poorly graded CWG column	6.48 × 10−5 cm/s
PR = 1.0, W	Specimen reinforced with an end-bearing well-graded CWG column	1.49 × 10−4 cm/s

.

It should be noted that the k was indirectly determined using the coefficient of consolidation (C_v) and the coefficient of volume compressibility (m_v) based on Terzaghi’s consolidation theory [37]. Thus, it is only accurate under the assumption that the specimens are fully saturated during the consolidation process, the water and solid constituents of the specimens are perfectly incompressible, Darcy’s law is valid, the k is constant, and the time lag of consolidation is entirely due to the low permeability of the specimen [37]. Thus, there may be discrepancies between the indirectly determined k and the directly measurable k, since k may not remain constant throughout the process due to variations in specimen density during consolidation. Furthermore, the specimen reinforced with an end-bearing CWG column is a composite specimen that may be treated as a single medium with average properties; however, the individual drainage characteristics of the materials remain distinct. In addition, the specimens reinforced with a CWG floating column are composite specimens with a soft soil layer beneath the CWG column, further increasing the possible variability in k within the composite system. Thus, it is recommended to directly determine the k using the appropriate permeability tests. Nevertheless, the results indicate that all specimens reinforced with a CWG column exhibited a higher estimated k value than the soft soil only (SSO) specimen. This inferred increase in permeability supports the consolidation results discussed in Section 4.2, where reinforced specimens exhibited faster dissipation of excess pore water pressure and improved consolidation behavior due to the CWG column. In general, specimens reinforced with an end-bearing CWG column exhibited a higher estimated k, which may be due to the continuous drainage path along their length to the underlying drainage boundary, allowing excess pore water to dissipate more effectively, unlike floating columns, which terminate within the soft soil layer.

The specimen reinforced with an end-bearing poorly graded CWG column (PR = 1.0, P) did not exhibit the highest estimated k in this study, as expected due to its larger particle voids. Instead, the end-bearing well-graded CWG column (PR = 1.0, W) showed the highest estimated k value, indicating it has the most efficient drainage path for excess pore water. This deviation from the expected trend may be attributed to the possibility that the poorly graded CWG column was clogged by the migration of fine clay particles from the surrounding soil into the larger voids of the poorly graded granular column.

It was also inferred in previous studies that the fine clay particles accumulate in the column during radial drainage, partially blocking the effective pore spaces, which reduces the column’s drainage efficiencies [53,54]. Clogging of granular columns may also intensify during the consolidation process due to the high hydraulic gradient at the soil–column interface [43,44]. Previous studies have also reported that the clogged zone or smear zone may occupy 20% to 50% or more of the column volume, depending on the granular material [3]. Although the properties of the clogged zone were not directly measured in this study, it is inferred that the accumulation of fine particles alters the column’s hydraulic and mechanical properties, reducing its effective drainage radius and permeability [3,54].

The deformation of the CWG column under vertical loading may further contribute to the clogging. When vertical load is applied, the non-rigid granular column tends to bulge laterally into the surrounding soil, while the surrounding soil provides lateral confinement that limits the extent of bulging until equilibrium is reached [55]. Thus, the failure of the reinforced specimen to effectively counteract the lateral expansion of the column under vertical load, as observed in unconfined specimens, is described. It commonly occurs in the upper portion of the column, typically within a critical depth of approximately four times the column diameter [44,56]. The combination of radial drainage and lateral bulging promotes the migration and accumulation of fine clay particles within the column. Consequently, the clogged zone is often concentrated on the upper region of the column where bulging is most pronounced [53]. Observations from the unconfined compression tests indicated that end-bearing CWG columns exhibited more bulging compared with floating columns. Furthermore, poorly graded CWG columns have a higher void ratio than well-graded columns, making them more susceptible to clogging as fine clay particles migrate into the column during radial drainage.

To assess the potential for clogging inferred from the results, a filter retention criterion based on Terzaghi-type design principles was applied, wherein the ratio of the D₁₅ of the CWG column to the D₈₅ of kaolin clay should not exceed 5 to limit the migration of fine particles into the column voids [57,58]. The computed ratios for the well-graded and poorly graded CWG columns are 23.1 and 62.3, respectively, indicating a high potential for clogging due to the migration of fine clay particles. Notably, the ratio for the poorly graded column is nearly three times greater than that of the well-graded column, suggesting a higher susceptibility to clogging. This behavior may explain the relatively lower permeability (k) observed in the poorly graded end-bearing CWG column compared to the well-graded configuration. Both clogging and smearing mechanisms can reduce drainage efficiency by restricting flow paths within and around the column. Although smearing effects during installation were minimized in this study due to the small specimen scale, field installations are likely to induce greater soil disturbance. Consequently, clogging effects in poorly graded CWG columns may be more pronounced under field conditions.

5. Conclusions

This study investigated the consolidation behavior, shear strength, and permeability of soft soil reinforced with a crushed waste glass (CWG) granular column. The effects of penetration ratio (PR) and CWG column gradation on the behavior of the reinforced soil were evaluated through the unconfined compression and consolidation tests. Between the two design factors, column PR plays a more significant role than CWG gradation in controlling the mechanical and hydraulic behavior of the reinforced soil system. Based on the experimental results, the following conclusions can be drawn:

• The reinforced specimens exhibited reduced settlement and lower compressibility compared with the unreinforced soil, demonstrating the effectiveness of the CWG column in improving the stiffness of the composite soil system. In contrast, the inclusion of CWG columns did not improve the undrained shear strength of the specimens under unconfined undrained loading conditions. This behavior was attributed to a reduction in cohesive soil volume and to limited mobilization of frictional resistance within the CWG column, due to the absence of lateral confinement.
• The coefficient of consolidation of reinforced specimens increased when the applied vertical stress exceeded the preconsolidation pressure, indicating the faster dissipation of excess pore water pressure.
• The hydraulic conductivity of the reinforced specimens was greater compared with the unreinforced soil, confirming that the CWG column enhances the drainage capacity of the composite soil. Specimens reinforced with an end-bearing well-graded CWG column exhibited the highest permeability, indicating the most efficient drainage through the granular column. Although poorly graded CWG columns generally contain larger void spaces, they did not exhibit the highest permeability because the potential clogging from the migration of fine clay particles into the column during consolidation can reduce permeability.
• Penetration ratio (PR) had a more pronounced influence than CWG gradation, particularly in controlling settlement reduction, preconsolidation pressure, and consolidation behavior. Considering the combined results for settlement reduction, consolidation response, drainage efficiency, and resistance to potential clogging, the end-bearing well-graded CWG column (PR = 1.0, W) provided the most balanced overall performance.

Overall, the results demonstrate the potential of CWG as a sustainable alternative granular material for soil reinforcement. The findings indicate that CWG columns may be suitable for soft soil improvement applications where settlement reduction and enhanced drainage are the primary design objectives, such as embankments, road subgrades, and lightly loaded foundation systems. Floating columns may still improve stiffness and consolidation behavior, though their settlement reduction performance is lower than that of end-bearing columns. Therefore, the selection of column penetration ratio should consider the required balance between construction cost, settlement control, and drainage performance. Thus, CWG columns show strong potential as an environmentally sustainable substitute for conventional granular column materials derived from natural aggregates.

6. Limitations of the Study and Recommendations for Future Work

This study presents several limitations that should be considered when interpreting the results and guiding future research. The experiments were conducted at a laboratory scale, and scale effects may influence stress distribution, drainage behavior, and soil–column interaction under field conditions. Hence, large-scale laboratory or field validation is recommended to confirm the applicability of the findings.

The use of kaolin clay as a model soil provides controlled and consistent conditions. However, it does not fully represent the variability and complexity of natural soft soils. Future studies should incorporate a wider range of soil types, including organic and highly plastic clays, to improve generalizability. In addition, further research may explore hybrid granular materials, such as CWG–sand mixtures, to enhance column performance while enabling the gradual integration of sustainable CWG with conventional aggregates.

The columns investigated were uncased, whereas geosynthetic encasement is commonly used in practice to enhance confinement and performance. Further research comparing encased and uncased CWG columns is recommended to quantify the benefits of confinement, particularly in very soft soils.

The potential for clogging within the CWG columns was not directly evaluated. Although improved consolidation behavior was observed, long-term drainage performance may still be affected by fine-particle migration. Future work should include permeability and filtration assessments, as well as microstructural characterization techniques such as computed tomography (CT) scanning, to directly evaluate particle migration and clogging mechanisms. In addition, larger-scale and field-representative investigations are recommended to assess the long-term drainage performance and durability of CWG column reinforcements under actual service conditions.

The study considered a limited range of penetration ratios (PRs), which may constrain the understanding of the transition between floating and end-bearing behavior. Expanding the range of PR values, supported by parametric, numerical, or finite element analyses, is recommended to establish more comprehensive design guidelines. Finite element modeling may provide an efficient means of evaluating a broader range of column configurations and PR values, complementing experimental work and enabling rapid assessment of system behavior, as in the approach adopted by Ng and Tan [45].