Effects of Sand–Cement Columns on Primary Consolidation Settlement

Abstract

The rapid increase in population and the corresponding increase in developments have necessitated the stabilization of areas with poor soil conditions. Due to consolidation settlement, the soft grounds available are deemed unsuitable for such structures. This paper presents the use of cement additives to build sand–cement columns in saturated clayey soils. The approach significantly reduces consolidation settlement and increases the bearing capacity, providing a viable solution to foundation problems. Consolidation tests were conducted on saturated clay specimens and sand–cement columns arranged in various patterns. A 5% cement content by the dry weight of the sand was used in building sand–cement columns. The results showed that the consolidation settlement rate was high due to the extra drainage formed by the widened pores in the sand–cement columns. The extra drainage caused more water to leave the specimen in a given time. However, after full contact between the loading platen and sand–cement columns, the rate of consolidation settlement decreased. At this stage, sand–cement participated in carrying the load. Additionally, the effect of vertical drainage on speeding up consolidation at higher stress levels was minimal, as the widened pores in the sand–cement columns began to close.

1. Introduction

Consolidation settlement in soft clay poses a significant challenge in geotechnical engineering, particularly in the design and construction of foundations. Clay soils are inherently problematic due to their low permeability, low shear strength, high compressibility, and low bearing capacity [1], making them a primary source of foundation-related issues [2]. Soft clays, characterized by weak shear strength and high compressibility, pose particular challenges for light structures, as noted by Banadaki et al. (2012) [3]. Remarkably, the financial impact of these damages surpasses the combined costs of damages caused by earthquakes, hurricanes, and tornadoes [2]. Addressing these challenges require a deep understanding of the geotechnical properties of soft clays and the implementation of effective mitigation strategies to ensure the stability and longevity of structures built on such soils.

Ground improvement for foundation purposes involves enhancing the engineering properties of soil to achieve the desired performance levels [4]. Geotechnical engineers have developed various techniques to address the challenges posed by weak or problematic soils, including vertical drains, preloading, geosynthetics, concrete piles, stone columns, and deep mixing columns [3]. Among these, the deep mixing method has gained significant attention for its effectiveness in mitigating sliding failures, controlling seepage, reducing shear deformation, and enhancing the dynamic performance of soft or contaminated soils, as Bouassida and Porbaha (2004) [5] explained. This method involves adding cement powder to soft ground to create soil–cement columns, which significantly improves the shear strength and compressibility of thick layers of weak soil, as noted by Broms and Boman (1979) [6]. Cement columns are particularly valuable for supporting the foundations of buildings and lightweight structures, with practical applications demonstrating their efficacy. For instance, Hibino (1996) [7] reported that cement columns 1 meter in diameter arranged in square or rectangular patterns successfully supported a five-story building in Japan. These advancements highlight the significance of ground improvement techniques in ensuring the stability and durability of structures constructed in challenging soil conditions.

Chemical stabilization techniques have proven to be one of the most economical and effective methods for stabilizing subgrade material [8]. Clay soil can be treated with cement, fly ash, lime, and gypsum additives [9]. Chai et al. (2006) [10] affirmed that chemical additives improved the consolidation index or rate of time of consolidation and Beeghly (2003) [11] mentioned that using lime and fly ash to pavement subgrade settlement was more inexpensive than the traditional cut-and-fill method. For instance, adding 6% cement to soft mud improved pre-consolidation stress, reduced the secondary consolidation index, increased the coefficient of consolidation, and reduced post-construction settlement [12]. The soil–cement technique, widely recognized for its effectiveness in ground stabilization through deep soil mixing, has achieved global success by leveraging the binding properties of cement to create a cement–soil material.

Application in saturated clayey soils benefits from the dual action of the sand core and cement binder [13,14]. Sand provides drainage and densification, while cement introduces cohesion and long-term strength through pozzolanic reactions [15]. These composite columns further stiffen the soil matrix, creating preferential load paths through high-modulus inclusions and constraining lateral deformation in the surrounding clay [16]. Rajab et al. (2016) [17] demonstrated that cement-stabilized sand columns under footing models improved the bearing capacity by up to 62–90%, while also significantly reducing compressibility and settlement. These improvements hold especially in regions with a high moisture content, where stabilization is critical to mitigating pore pressure buildup and swelling. According to Ni et al. (2019) [16], laboratory and field studies have consistently demonstrated substantial bearing capacity gains within composite sand–cement columns, typically in the order of 70–90%, compared to untreated clay foundations.

Consolidation settlement is a gradual and continuous process in which the soil beneath a foundation compresses under applied loads, reducing its volume over time due to water expulsion. When the settlement exceeds the allowable limits, it can lead to significant structural issues, including cracks, tilting, and other severe deformations in buildings and infrastructure. The financial impact of repairing structures affected by excessive consolidation settlement is often substantial, making it a critical engineering concern. Addressing this challenge requires innovative stabilization techniques to mitigate settlement and ensure the long-term stability of foundations. This research uses sand–cement columns as a stabilization technique to effectively control consolidation settlement in saturated clayey soils, addressing some limitations of conventional ground improvement methods. This study investigates the combined performance of sand–cement columns, particularly their dual capacity to enhance both drainage and stiffness simultaneously. Additionally, the work provides an innovative analysis of how varying proportions of sand and cement influence consolidation behavior and offers practical guidance for optimized design in soft clay foundations.

2. Materials and Methods

2.1. Experiment Overview

The experiment employed a one-dimensional oedometer test setup that incorporated incremental weights ranging from 1 kg to 64 kg, along with a consolidation unit consisting of a specimen ring, filter paper, porous stone, and a loading cap. The test [18] involved conducting the one-dimensional oedometer test on saturated clay specimens with variations in the sand–cement column patterns, including no sand–cement column, one sand–cement column, and two sand–cement column patterns. The deformation rate versus the root of time was recorded under different load applications. Each load was maintained for 24 h. The flow chart in Figure 1 outlines the diverse laboratory procedures, experimental setup, equipment used, materials employed, and the tests conducted.

Kaolinite clay was selected for this study due to its well-known properties [19,20] and frequent use in geotechnical research. The clay was sourced commercially to ensure consistent quality across all tests, and several geotechnical laboratory tests were conducted to characterize the clay’s physical properties, following ASTM standards. These included the Atterberg limits (liquid limit, plastic limit, and plasticity index) [21], specific gravity [22], and bulk density tests [23]. The results from these tests (summarized in

Table 1. Properties of the kaolinite clay.

Parameter	Value
Liquid Limit (%)	49
Plastic Limit (%)	25
Plastic Index (%)	24
Specific Gravity	2.68
Bulk Density (g/cm3)	2.62

) were used to calculate the precise amount of water required to saturate the clay. This calculation ensured a uniform moisture content, a key factor in obtaining consistent and reproducible experimental results. Once the water content was determined, the clay was carefully mixed with the calculated volume of water using a mechanical mortar mixer.

2.2. Sand–Cement Mixture

The sand used in this study was sourced from United States Silica Company to ensure a consistent and high-quality experimental material. Before creating the sand–cement mixture, the sand’s properties were thoroughly analyzed following ASTM procedures. The results from these tests, including the Specific Gravity, Permeability [24], and Bulk Density, are summarized in

Table 2. Properties of the sand.

Parameter	Value
Permeability (m/s)	1.0 × 10−4
Bulk Density (g/cm3)	1.62
Specific Gravity	2.67

. Portland cement was added at a ratio of 5% to the dry weight of the sand, a proportion selected based on its effectiveness in improving soil strength and workability, as highlighted in earlier studies by Yaghoubi et al. (2018) [25] and Feng et al. (2001) [12]. To create the sand–cement mixture, 5% cement by the dry weight of sand was carefully measured and mixed with the sand to ensure a uniform blend. The dry components were placed into a mortar mixer and thoroughly mixed for five minutes to achieve homogeneity. Afterward, water, equivalent to 25% of the sand’s volume, was gradually added to the mixture, transforming it into a mortar paste.

2.3. Specimen Preparation and Experimental Procedures

A measured quantity of kaolinite clay, weighing 700 g, was combined with 270 g of water to create a smooth and uniform mixture. This prepared mixture was then carefully shaped into cylindrical specimens with a diameter of 60 mm and a height of 20 mm, using steel rings to ensure precise dimensions. Templates with 6 mm diameter holes were used to precisely mark the positions where the sand–cement columns would be placed within the clay specimens. A plastic straw was carefully inserted to extract the clay from the marked spots, creating these holes. These holes were filled with the sand–cement mixture (Figure 2).

Three saturated clay specimens were subjected to a standard test method [1] to determine the one-dimensional consolidation properties of soils using incremental loading. The test involved loading the specimen in a rigid oedometer ring under incremental vertical stresses while measuring its deformation over time. Data from timed compression readings during each load step are used to plot settlement curves, enabling an analysis of the consolidation behavior. The tests were conducted using a semi-automated consolidation test apparatus, which facilitated precise data acquisition and analysis. The oedometers, containing the clay specimens, were positioned on a loading frame equipped with a lever arm, weights, and a jack. The jack supported the lever arm, allowing for the controlled application of incremental loads. A loading cap was placed on top of each specimen and connected to transducers, which measured the resulting deformation. These transducers converted the physical deformation data into electrical signals, which were transmitted to a data logger for analysis. The data logger processed the data and displayed the results in the form of deformation versus the square root of time, allowing for a detailed evaluation of the consolidation behavior. The experimental setup is summarized schematically in the accompanying diagram (Figure 3).

2.4. Consolidation Test

Cylindrical saturated clay specimens were prepared using specimen rings measuring 20 mm in height and 60 mm in diameter. The rings were pressed into the clay to extract the specimen, and the ends were trimmed with a spatula to produce smooth, even surfaces. This method was repeated to create three specimens representing the following configurations: one without sand–cement columns, one with a single sand–cement column, and one with two sand–cement columns. Column cavities were formed for specimens requiring sand–cement columns by extracting clay from the designated positions using a 6 mm diameter plastic straw. The sand–cement mixture was prepared by blending dry sand and cement in a mortar mixer for five minutes to ensure uniformity. The prepared mixture was then poured into the cavities, ensuring they were filled evenly to the full 20 mm height of the specimen. Water was added to the dry mix in a controlled proportion of 25% by volume of sand, measured with a syringe to achieve a consistent mortar paste. Porous stones were first saturated by immersing them in distilled water to prepare the consolidation setup and ensure full saturation. In each consolidation mold, a lower porous stone was positioned at the base, followed by a filter paper that had been pre-soaked in distilled water to prevent air entrapment. The saturated clay specimen ring was carefully placed on the filter paper. This assembly was completed by layering another wet filter paper and an upper porous stone with a loading cap on top to ensure uniform load distribution during testing. This procedure was systematically repeated for all three specimens, ensuring consistent preparation across configurations.

The sand–cement columns were cured for seven days to facilitate adequate hydration and hardening. During this phase, the consolidation units containing the prepared specimens were submerged in water, with levels meticulously monitored to ensure continuous immersion, consistent with standard soil–cement stabilization procedures [26]. Each consolidation unit was placed in a one-dimensional oedometer test apparatus capable of conducting three simultaneous tests. Incremental weights of 1 kg, 2 kg, 4 kg, 8 kg, 16 kg, 32 kg, and 64 kg were sequentially applied using a loading hanger, adhering to established protocols for consolidation testing [18]. Transducers affixed to the loading caps measured the deformation of the saturated clay specimens, both with and without sand–cement columns, under the applied stress. The deformation data were transmitted to a data logger, which processed and displayed the results as deformation versus the square root of time, facilitating a detailed analysis of the consolidation behavior.

2.5. Unconfined Compressive Strength (UCS) of Sand–Cement Column

An Unconfined Compressive Strength (UCS) test was conducted to determine the compressive strength of the cured sand–cement columns, following the ASTM D2166/D2166M-13 [27] standard procedure. The average diameter and height of the specimen were 7.0 cm and 15.0 cm, respectively. The cross-sectional area of the column was 38 cm². The specimens were cured for seven days under controlled conditions for sufficient hydration and hardening. After curing, the specimens were subjected to axial compression in a UCS testing apparatus until failure. The peak stress at failure was recorded as the UCS value, providing a quantitative assessment of the compressive strength of the sand–cement columns. This was achieved by applying a seating value of 5 N and a logging displacement increment of 0.1 mm until failure.

3. Results and Discussions

The advantage of this research lies in the experimental work. Therefore, much attention was given to the sand–cement column, which provides stiffness to the saturated clay. The stiffness of the column was also a significant factor in controlling settlement. Following the consolidation tests, the condition of the sand–cement columns was examined, with a focus on identifying patterns of deformation and failure. Specimens containing one (Figure 4a,b) and two sand–cement columns (Figure 4c,d) were analyzed using a binocular microscope. Microscopic observations revealed the widening of pores between particles, leading to the formation of drainage paths within the sand–cement columns during the consolidation process (Figure 4). These findings, along with detailed analyses and evaluations of the observed widened pores and drainage paths, are discussed in subsequent chapters.

3.1. Unconfined Compressive Strength Results

To predict the peak strength or the stress at which failure would occur within the sand–cement column, unconfined compressive strength (UCS) tests were conducted. Figure 5 shows the UCS results of three sand–cement column specimens. The average diameter and height of the specimen were 0.07 m and 0.015 m, respectively. The cross-sectional area of the column was 0.0038 m². The unconfined compression strength of the soil was 255 kPa (Figure 5). These findings reinforce the efficacy of sand–cement columns in providing structural support, aligning with similar studies emphasizing their role in reducing settlement and increasing the load-bearing capacity [13,28].

3.2. Consolidation Test Results

Figure 6, Figure 7, Figure 8, Figure 9, Figure 10, Figure 11 and Figure 12 show the consolidation test results representing the relationship between deformation and time. Figure 6 shows the deformation and time after 1 kg or 32 kPa of vertical stress was applied to the specimen, 24 h later, for specimens with and without sand–cement columns. Interestingly, the graph shows that the specimen with two columns recorded the maximum deformation at 1.06 mm, followed by the specimen with no sand–cement column, which showed a settlement of 1.03 mm. The specimen with one column recorded the least deformation at 0.97 mm.

Figure 7 shows the result after 48 h when a normal stress of 64 kPa was applied. Again, the specimen with two columns followed by one column recorded the maximum deformation. The specimen with no sand–cement column recorded the least deformation at 1.32 mm.

Figure 8, Figure 9 and Figure 10 show a similar trend of deformation versus time, with maximum deformation evident in the one-column specimen, followed by the no-column specimen. A two-column specimen recorded the least deformation after 72 h (Day 3), 96 h (Day 4), and 120 h (Day 5) for applied normal stresses of 128 kPa, 256 kPa, and 512 kPa.

Figure 11 and Figure 12 showed identical trends in terms of deformation; the specimen with two columns recorded the least deformation, followed by the specimen with one column, and the specimen with no column had the most deformation after 144 h (Day 6) and 168 h (Day 7).

4. Discussion of the Results

Figure 6 shows more deformation in the specimen with two columns than in the specimen with no columns. This is because, before a full contact is formed between the loading plate and the sand–cement column, the pores in the sand–cement column provide extra drainage. The extra drainage caused more water to leave the specimen, contributing to the high deformation. The reduced deformation in the one-column specimen could be attributed to the fact that the column stiffness began to contribute to controlling settlement by reducing settlement at the early stage of the test. Figure 7 shows the deformation for the three specimens with and without columns after increasing the stress from 32 kPa to 64 kPa. After forming full contact between the loading plate and sand–cement columns, the specimen with column(s) provided an additional drainage path because of the widened pores developed (Figure 4) in the sand–cement column(s). The extra drainage caused more deformation due to the rapid volume change. The specimen with two columns provided more drainage than the specimen with one column. Therefore, the two-column specimen experienced more deformation than the specimen with one column. Less deformation was observed in the specimen with no column because no extra drainage was created. Figure 8, Figure 9 and Figure 10 exhibit a similar trend, where increasing stress from 64 kPa to 512 kPa results in the formation of bigger pore openings in the sand–cement columns. Observations from the USC results confirmed that the sand–cement column fails at 255 kPa, as evident when the specimens were analyzed under a microscope. However, at this stage, the stiffness of the columns contributed to reducing deformation as stress increased, and the drainage effect had little influence on the deformation. Therefore, less deformation was observed in the specimen with two columns compared to the specimens with one column and no column. The one-column specimen also experienced the most deformation because of the stress concentration on the single column.

As the applied stress continued to increase, the confining pressure around the columns forced these openings to progressively close up over time, effectively reducing internal voids or pores. This closure enhances interparticle contact and friction, resulting in an increase in stiffness and load resistance. Figure 11 and Figure 12 also show an identical trend, where specimens with two columns experienced the least deformation due to the stiffness of the columns, combined with the stress distribution among them. The stiffness provided by the one column also increased with an increase in stress as the bigger pores within the columns closed. Hence, less deformation was observed in the specimen with a sand–cement column than in the specimen without one.

5. Conclusions

The saturated clay specimen was prepared for the consolidation test by adding 270 g of water to saturate 700 g of clay. The sand–cement columns were built by adding 5% cement to the sand by the dry weight of sand and 25% of water by the volume of sand to make the sand–cement column. The consolidation test conducted on the saturated clay specimen with and without sand–cement revealed that, before forming full contact between the loading plate and the sand–cement column(s), the extra drainage because of the widened pores in the sand–cement columns speeds up the consolidation settlement. After full contact was formed between the loading plate and the sand–cement columns, the settlement was controlled by the stiffness of the sand–cement columns.

The sand–cement columns speed up the consolidation settlement at the early stages of loading. However, as the load increases, some bigger openings start to develop in the sand–cement columns. At this stage, the stiffness of the sand–cement columns contributes to carrying the load and controlling settlement, and the effect of the extra drainage on settlement is minimal; therefore, the consolidation settlement is less than in the case with no sand–cement columns. After full contact is developed between the loading plate and the sand–cement columns, the rate of settlement is reduced.