Introduction

Poorly graded and loose granular soils, which are widely distributed around the globe, are characterised by a lack of cohesion and low strength. Construction of highway networks, railways, airport runways, embankments, and other earthen structures on the sandy subgrade soils can cause problems such as permanent settlement, lateral displacement, and inadequate stability. One method to improve and strengthen the soil is stabilization, using eco-friendly geopolymer materials to improve the mechanical properties of poorly graded sandy soils1. Portland cement and lime are conventional choices for the stabilization of these soils in geotechnical engineering2,3,4,5. However, the production of these materials has caused environmental harm, including carbon dioxide emissions and high energy consumption, which contribute to major climate change and an increase in global warming6,7,8. In recent years, there have been several efforts to stabilize weak soils efficiently and find low-carbon alternatives to Portland cement by utilizing ecologically friendly geopolymer materials resulting from industrial solid waste, which are known for their low carbon footprint, and the extraction and production costs are almost non-existent9,10,11,12,13.

Geopolymer, also known as green cement, is an alkali stabilizer that enhances mechanical strength and controls volume change via the process of polycondensation13,14,15. A geopolymer is formed by mixing an inorganic binder (such as natural pozzolan or industrial waste products) rich in silica (Si2+), alumina (Al3+), and calcium (Ca2+) with an alkaline solution for activation such as sodium hydroxide or sodium silicate for activation, etc.

Previous research on stabilizing different soil types has used many natural raw materials and industrial waste materials, including granulated blast-furnace slag, fly ash, rice husk ash, silica fume, glass powder, and metakaolin to form geopolymers16,17,18,19. These materials have high levels of silicon dioxide (SiO2), aluminium oxide (Al2O3), and calcium oxide (CaO) making them suitable for the geopolymerization process. The alkaline pozzolanic raw materials supply alkali metal cations, thus increasing the pH of the alkaline solution. It accelerates the dissolution of solid precursors (silicon, aluminium and calcium ions) and condenses them, synthesizing the best geopolymer binders20,21. Davidovits observed that the inorganic silico-aluminate based geopolymers had a network composed of SiO4 and AlO4tetrahedra connected by sharing all the oxygens alternately to form (-Si-O-Al-O-) poly(sialate), (-Si-O-Al-O-Si-O-) poly(sialate-siloxo) and (-Si-O-Al-O-Si-O-Si-O-) poly(sialate-disiloxo)22. The oligomers, intermediates in the polymerization procedure, are connected to the geopolymer chain by three-dimensional metallic structures in five-fold coordination. They may be represented by the chemical formula in the ‘’equation (1)’’:

$$\:{M}_{n}{\left[-{\left({SiO}_{2}\right)}_{X}{-AlO}_{n}-\right]}_{n}$$
(1)

Where n is the degree of polymerization, x is the Si/Al molar ratio defined as (1, 2, 3, etc…), and M is the alkali such as sodium (Na+), calcium (Ca2+) or potassium (K+)10,23.

Paper is one of the most industrial waste materials in all countries due to its extensive usage across several areas of life. This waste is currently invested and used in several industries, the most important of which is the production of construction materials. Previous studies have shown that burning paper sludge ash (PSA) contains a high percentage of calcium, which can be considered supplemental cementing material or fly ash-based geopolymers24,25,26,27,28. It is worth noting that the presence of soluble calcium is crucial as it accelerates the hardening process and early strength development due to enhanced aluminosilicate gel formation27,29. Some authors have noted that the final gel formation is mostly determined by the chemical characteristics of the precursor and the alkali activator used. Strong bonds are formed by calcium aluminosilicate hydrate gel (C-A-S-H), resulting from the high CaO content in the precursors components like fly ash and PSA, which is similar to calcium silicate hydrate gel (C–S–H) produced during OPC hydration or sodium aluminate-silicate-hydrate (N-A-S-H) gel13,30,31,32. Finally, after ion exchange (Ca, N)-A-S-H transforms into a gel-like substance that binds with soil particles, reducing porosity, improving cohesion, increasing dry density, and enhancing the soil’s early strength and shear strength33,34,35.

Previous research attempts to stabilize clay soils using geopolymers based on PSA combustion were evaluated as sustainable alternatives to conventional stabilizers36. Nevertheless, there is a scarcity of research on the use of PSA-based geopolymers for assessing their appropriateness in sand soil stabilization applications. Researchers have primarily focused on soil stabilization using PSA, without using PSA-based geopolymer-stabilized soil37,38,39,40,41.

Permanent strain in granular soil is a deformation resulting from an accumulation of permanent strains in the surface soil layers caused by repeated traffic loads in the long term. It is a crucial subject in geotechnical engineering for ensuring the safety and reliability of geotechnical structures subject to repetitive loads. Permanent strain, in turn, changes the properties and behavior of soil and leads to the subsequent erosion of flexible pavement42,43. Despite numerous studies on the stabilization of different types of soil using geopolymers, there is still a need to understand the effect of cyclic loads on the permanent deformation behavior of geopolymer-stabilized soil layers, which constitutes also a knowledge gap that this research attempts to address.

This research examines the effect of using PSA-based geopolymer produced from waste paper, on loose, poorly graded sand soils to enhance their mechanical properties and support sustainable development objectives. To understand the mechanical properties of stabilized sand soil by PSA-based geopolymer, tests such as compaction, compressive strength, California Bearing Ratio, and direct shear test were performed. For this purpose, many parameters that affect the improvement of the strength of stabilized sand soil, including the curing time, the effect of increasing the concentration and type of alkaline solution, and adding different percentages of PSA to the soil, have been investigated. Sodium hydroxide (NaOH) and sodium silicate (Na₂SiO₃) were utilized as alkaline activators to produce a PSA-based geopolymer. Additionally, the microstructural alterations and chemical composition of the stabilized soil samples were investigated using Scanning Electron Microscopy (SEM) images and X-ray diffraction (XRD) analysis. Finally, the permanent strain for both unstabilized and stabilized soils is assessed by developing and fabricating a laboratory testing instrument assembly.

Materials and methods

Materials properties

(Soil). The sandy soil utilized in this research is subgrade soil with almost uniform grain size, located in the southern part of Karbala city with a geographic location of latitude 32.552488° N and longitude 43.995383° E. The gradation curve of this soil is shown in Fig. 1according to ASTM D422 standards44. The soil is classified as poorly graded sand soil (SP) according to the unified soil classification system (USCS)45and type (A-3) based on the American Association of State Highway and Transportation classification system (AASHTO)46. This soil is problematic because it has a poorly graded granularity and a low bearing capacity. Table 1 presents a concise of the soil’s physical and mechanical characteristics.

Fig. 1
Fig. 1
Full size image

Grain distribution curve of the sand soil.

Table 1 Physical and mechanical properties of the studied natural subgrade soil.
Full size table

(Paper sludge Ash). This research uses PSA as a geopolymer waste material for stabilizing sandy soil. The growing paper industry in Iraq and the increasing rate of paper recycling have led to a continuous rise in PSA production. PSA has a specific gravity of 1.69 and a pH value equal to 12.4. The process of preparing a PSA involves three stages: Firstly, A substantial quantity of paper was incinerated in an open-air setting as the principal method of reducing its size, eliminating oxides, and transforming it into ash, as shown in Fig. 2a. Secondly, the ash obtained from the first stage was subjected to a again combustion process in a furnace operating at a temperature of 700 °C for two hours, resulting in its transformation into re-active pozzolanic material48,49, as shown in Fig. 2b. Finally, the produced ash was milled for thirty minutes to simulate the PSA of actual ash. Subsequently, it was passed through a No. 200 sieve, as recommended by Garcia et al.26,50, as shown in Fig. 2c. Table 2displays the chemical composition characteristics of PSA using XRF elemental analysis. The major components of the PSA are calcium, silica, and alumina. This indicates that it has a significant proportion of kaolin and exhibits pozzolanic properties51. Additionally, it has a low concentration of magnesium oxide, iron oxide, potassium oxide, and sodium oxide as secondary components31. The mass of calcium oxide (CaO) in the PSA was around 53%. Noteworthy is that the component makeup of the PSA closely resembled that of ordinary Portland cement (OPC)52. It is important to note that the high loss on ignition (LOI) value of PSA is almost equal to the combined amount of gases generated by the de-carbonation of calcite, and the pyrolysis of the small cellulose fibres (organic portion). This tendency corresponds with the data recorded by Ferreiro et al.53,. Figure 3 presents the XRD study of PSA, which was conducted within the 2Theta range of 5 to 90 degrees. The phases were determined to be calcium carbonate (CaCO3), and C3A-cubic aluminate (Ca9(Al6O18)).

Table 2 Chemical composition of the PSA used. *LOI: loss on ignition.
Full size table
Fig. 2
Fig. 2
Full size image

The procedure of preparing PSA: (a) initial burn phase; (b) PSA after heat treatment; and (c) grinding PSA.

Fig. 3
Fig. 3
Full size image

XRD analysis of PSA.

(Alkaline activator). The alkaline activator facilitates the dissolution of the raw pozzolanic materials and enhances the development of mechanical characteristics in the formed geopolymer. In the current research, two types of alkaline activators, including sodium hydroxide (NaOH) and sodium silicate (Na₂SiO₃), with an alkalinity of SiO₂:Na₂O (1:1), have been used to activate the geopolymerization process. The alkali solutions with different molar concentrations were prepared by dissolving the flakes or powders of these materials in distilled water. A (5 Molar) NaOH solution was prepared by progressively diluting (200 g) of NaOH flakes in (1 L) of distilled water. The dissolution of NaOH in water leads to an increase in temperature. Consequently, it is used to make inspection samples before one day, allowing the solution to attain room temperature and greater homogeneity. In addition, the second solution is created by slowly dissolving a precise amount (366 g) of Na₂SiO₃ powder in one liter of distilled water within a glass container. This process produces a Na₂SiO₃ solution with a concentration of (3 Molar). It is unnecessary to wait one day before using the Na₂SiO₃ solution. It may be used directly to make samples after producing the solvent. The addition of Na₂SiO₃ into the geopolymer system enhances the Si/Al ratio, hence leading to a more efficient alkaline-activator54,55. According to Table 3, NaOH and Na₂SiO₃ molar masses are (39.99 and 122.06 g/mol), respectively. To make and calculate an alkaline solution with a concentration of (x) molar, the grams of alkaline activator are divided by the molar masses and dissolved in a volume of (1 L) of distilled water.

Alkaline activator additions provide an alkaline environment to promote geopolymerization and increase the level of \(\:{\text{S}\text{i}\text{O}}_{4}^{4-}\)and OH−ions available in the sand soil-PSA combination. Consequently, the alkalinity increases the solubility of the siliceous material, so accelerating the pozzolanic reaction and improving the degree of geopolymerization. This leads to the formation of highly cementitious calcium silicates, which are shown as follows56:

CaO + H2O→Ca(OH)2 (Portlandite).

Ca(OH)2 + Na2SiO3.9H2O → CaSiO3 + 2NaOH + 9 H₂O.

Table 3 Chemical and physical characteristics of the alkaline activator.
Full size table

Finally, the section (Determination of alkaline solutions concentration) has identified the optimum concentration of NaOH and Na₂SiO₃ that produces geopolymers with the highest UCS.

Experimental methods

Soil mixture proportions

In the current research, PSA was used as the main source material for geopolymers with a specific mixing percentage (5, 10, and 15%) because researchers circulated these percentages. Soil and a specific percentage of PSA were mixed in a dry state. A quantity of alkaline solution (NaOH or Na₂SiO₃), equivalent to the optimum moisture content (OMC), was added to the soil-PSA mixture. The mixture was stirred for 5 min to achieve a homogeneous consistency, and the stabilized soil mold was cast. Table 4 displays the composition of several soil samples that were stabilized and cured for three different periods: 1, 3, and 7 days.

Table 4 Specifications of all soil specimens.
Full size table

Determination of alkaline solution concentration

The objective and reason were to investigate the effect of both the kind and molarity of alkaline activators on the stabilized soil with geopolymer to achieve the highest UCS at the most optimum concentration. The kind of alkaline solution and its concentration significantly influence the rate at which dissolution occurs57. Various molarities of NaOH and Na₂SiO₃ solution were used, ranging from low to high concentrations. Observed as concentrations increased, there was a noticeable decrease in compressive strength. Alkaline solutions with different molarities of NaOH (2.5, 5, and 7.5 M) and Na₂SiO₃ (1.5, 3, and 4.5 M) were used. Moreover, it is essential to highlight that changes in the chemical composition, namely the CaO, SiO2, and Al2O3 content, of both the geopolymer and the soil might result in variations in the amounts and ratios of alkaline concentrations needed. Hence, it is important to acknowledge that the alkaline concentrations mentioned in prior studies may not be applicable and are susceptible to alteration.

(NaOH solution). The UCS of the soil which has been stabilized using geopolymer produced from a constant proportion of PSA with NaOH at concentrations of (2.5, 5, and 7.5 M), is shown in Fig. 4. The results showed a positive correlation between the NaOH concentration and the UCS. The results obtained for the UCS were 136, 374, and 403 kPa when using NaOH solutions with concentrations of (2.5, 5, and 7.5 M), respectively. The percentage of increase in UCS, or the growth rate of UCS, was highest at a concentration of (5 M) when compared to (2.5 and 7.5 M). This concentration is considered the optimal concentration and was used in this research. In other words, using a lower concentration alkali activator (i.e. 2.5 M NaOH or less), there is lower dissolution of PSA, resulting in a geopolymer with less binding strength. On the other hand, when the concentration of NaOH solution is increased to levels over 5 M, the resulting geopolymer mixes experience a higher temperature, causing the surplus water to evaporate before the formation of a homogeneous geopolymer. Xu & Van Deventer58 have stated that brittleness and efflorescence may occur as a result of elevated concentrations of alkali solution in stabilized samples. As a consequence, the compressive strength of the geopolymer is lower.

Finally, optimal NaOH concentration is essential to promote and activate geopolymerization, which is crucial for developing the mechanical characteristics of soil.

(Na₂SiO₃ solution). Figure 4 illustrates the UCS of stabilized soil by a constant proportion of PSA with different concentrations of the Na₂SiO₃ solutions at (1.5, 3, and 4.5 M). The results displayed that as the Na₂SiO₃ concentration increased, the UCS also increased. The recorded values for the UCS were 427, 983, and 1102 kPa when using Na₂SiO₃ solutions with concentrations of (1.5, 3, and 4.5 M), respectively. The growth rate of UCS showed the highest percentage of improvement at a concentration of (3 M), compared to (1.5 M and 4.5 M). This concentration is considered the optimal concentration and was chosen in this research.

The additional Ca(OH)2(Portlandite) in the PSA and soil undergoes a chemical reaction that permanently bonds the silicates, enhancing resistance. This kind of alkaline activator is crucial in the geopolymerization process58,59. However, as indicated by García-Lodeiro et al.60, the improvement in the compressive strength can be attributed to the ionic exchange of Ca2+ (with an ionic size of 114 pm) for Na+ (with an ionic size of 116 pm). Consequently, (N-A-S-H) gels gradually become (C-A-S-H) gels, which are stable geopolymer systems rich in calcium, similar to the PSA in this research. Consequently, the presence of more geopolymer precursors in Na₂SiO₃ solutions leads to higher compressive strength of the geopolymers compared to those formed with NaOH solutions.

Fig. 4
Fig. 4
Full size image

The UCS results of stabilized soil with different alkaline solution concentrations.

Specimen preparation and laboratory tests

To ensure the correctness of the results, pairs of unstabilized and stabilized soil samples are created using several combinations of PSA stabilizers (at addition percentages of 5, 10, and 15%) and alkaline activators (either NaOH or Na₂SiO₃) for each test, as shown in Fig. 5 in phase one. After molding, samples are kept moist at laboratory temperature using a thin plastic covering to reduce moisture evaporation. The length of the curing process is a crucial factor in enhancing the strength of the samples, as it enables the geopolymerization process to fully occur between the geopolymer material and sandy soil. After the curing periods of 1, 3, and 7 days, the different combinations of stabilizers were subjected to laboratory testing, as shown in Fig. 5 in phase two. In addition, to study the effect of the addition of PSA and concentration of alkaline activator solutions on unconfined compressive strength (UCS), California bearing ratio (CBR), direct shear test (DST), X-ray diffraction (XRD), and scanning electron microscopy (SEM) of stabilized sand soil. Figure 5 provides a concise overview of the test methodology used in this two-phase research.

Fig. 5
Fig. 5
Full size image

Experimental program chart.

Modified proctor test

The modified Proctor compaction test is performed according to the ASTM D1557 standard to determine the optimal moisture content (OMC) and maximum dry density (MDD) for different combinations of soil and geopolymer61.

Unconfined Compression Strength Test (UCS)

The compressive strength and modulus of elasticity of the stabilized soil samples are determined using the UCS test, as specified by ASTM D216662, in order to evaluate the influence of PSA. The UCS test’s objective is to determine the optimal amount of PSA and the concentration of the activator used to stabilize the soil.

California Bearing Ratio Test (CBR)

The California Bearing Ratio (CBR) test is performed according to the specifications described in ASTMD188363. Although the CBR test is often used in pavement design techniques to assess pavement behavior under traffic loads, it is essential to note that preparing soil samples in the lab requires considerable time64.

Direct Shear Test (DST)

The study used the direct shear test, following the ASTMD1883 standard65, to determine the internal friction angle (ϕ), cohesion parameter (C), and ultimate shear strength. The specimen was prepared for testing in a direct shear test, measuring (6 * 6 * 2.5 cm), by including varying proportions of PSA with 5% increments, mainly 5, 10, and 15%, and two types of different alkaline solutions. The proportions were determined based on the OMC and MDD of the stabilized soil, as in the UCS and CBR tests. To facilitate the curing process, the samples placed in the mold of the DST are stored in hermetically sealed plastic bags to prevent the loss of moisture. For this test, researchers produced three identical specimens from each parentage using the direct shear box. The test included subjecting soil samples to three distinct levels of normal stress: (100, 200, and 400 kPa), in order to quantify the horizontal displacements and shear stress of the stabilized soil.

Repetitive Static Plate Load Test (RSPLT)

The repetitive static plate load test was implemented using the equipment and laboratory testing setup illustrated in Fig. 6. The test was carried out on the large laboratory specimen model subjected to cyclic load. The following subsections discuss the laboratory testing setup and large specimen preparation.

Equipment used in laboratory testing setup

The laboratory testing program aimed to assess and predict the accumulated permanent (non-recoverable) strain behavior of stabilized and unstabilized soils under cyclic loading by RSPLT. To accomplish this objective, the test was carried out in a large-laboratory testing setup, which included a steel mold, a loading frame, a hydraulic jack connected to an electric pump, a control panel, a load cell (70 kN), a steel plate (diameter = 200 mm, thickness = 25 mm), linear variable differential transformers sensors (LVDTs) with a sensitivity of 0.01 mm and data acquisition systems (DAQ) input/output device, as shown in Fig. 6a, b. The steel mold was designed with dimensions of 70 cm length × 70 cm width × 60 cm height to emulate the in-situ conditions of subgrade soil. These dimensions were selected to ensure that the applied vertical load is not influenced by the mold’s boundary conditions (i.e., lateral edges and depth), based on recommendations from prior research on similar models66,67.

A DAQ device from National Instruments enables quick communication and control between software and associated hardware (an electric hydraulic pump, LVDTs, and load cell), ensuring superior accuracy at high sampling rates. In addition, the LabVIEW driver (which can offer exact control and accurate simulation) is used to apply repetitive loads at a consistent loading rate.

Fig. 6
Fig. 6
Full size image

Testing system set-up: (a) computer monitor with control panel, DAQ device, (b) repetitive static plate load test assembly.

Soil preparation

The quantities of sand soil, pozzolanic material (derived from optimal mixing ratios of 10% PSA and 90% soil from the UCS test), and alkali solution were determined according to the modified Proctor test (MDD and OMC) and the mold dimensions. The soil and PSA were mixed together in dry conditions utilizing a 250-liter drum mixer. Water and the alkaline solution were then added to the natural and geopolymer-stabilized soil until the optimal moisture content was reached. The steel mold is filled with soil (either natural or geopolymer-stabilized soil) in three layers, each 20 cm high, and compacted thoroughly using a 15 kg hand compactor to achieve the desired density. The sand cone method was carried out to check the compaction degree of the soil layers. The soil moisture of these layers was monitored as well. The dry densities of the geopolymer-stabilized soil ranged from 1.842 to 1.859 gm/cm3, while the degree of compaction varied between 95.0% and 95.9% across the three layers, as shown in Table 5. Consequently, the geopolymer-stabilized soil layer might be characterized as very dense. The testing technique was conducted in the following sequence order:

  1. Step 1.

    After filling the steel mold for the soil sample to a total thickness of 60 cm, ensure the upper surface achieves a consistent and nearly flat upper surface, which is critical for accurate measurement and leaving it during the curing period. The one-day curing period allows the alkaline solution to react chemically, which helps the soil gain sufficient initial strength. For testing, the hydraulic jack and load cell are positioned in the mold center, as shown in Fig. 6b.

  2. Step 2.

    According to Huang’s suggestion68, the triangular-shaped load pulse was repeatedly applied with a constant vertical cyclic load to unstabilized and stabilized soil specimens within steel mold to simulate the wheel load condition above a specific point on the pavement. As shown in Fig. 7, the load cycle duration was 5 s (a four-second loading period and a one-second rest period) for 200 cycles.

  1. Step 3.

    The permanent strain was calculated by recording the average readings of two LVDTs for each load cycle.

Table 5 Specifications for mold testing specimens.
Full size table
Fig. 7
Fig. 7
Full size image

Applied vertical load pulse in the search.

Results and discussion

Compaction modified proctor test

The results showed that the addition of 5, 10, and 15% PSA resulted in a decrease in the MDD, as shown in Fig. 8. Therefore, specific gravity plays an important role and greatly influences the compaction parameters. The decrease in the specific gravity of PSA leads to a corresponding decrease in the MDD of stabilized soil. This tendency corresponds to the results recorded by Gunaydın69. Furthermore, as depicted in Fig. 9, there is a negative correlation between the percentage of PSA and the OMC of the stabilized soil. Previous studies have shown that the application of pozzolanic material stabilization leads to a reduction in OMC in stabilized soil, as reported in the literature70,71,72.

Fig. 8
Fig. 8
Full size image

Compaction curve of different PSA-soil mixes.

Fig. 9
Fig. 9
Full size image

The effect of adding PSA % on OMC and MDD of stabilized soil.

UCS test

The UCS and Es values of all stabilized soils, with different percentages of PSA, exhibited a substantial and consistent improvement at all curing durations, as well as a decrease in strain, as shown in Table 6; Fig. 10. Figures 11 and 12, show the results of the UCS tests conducted to assess the strength and Es of stabilized sandy soil with different proportions 5, 10, and 15% of PSA-based geopolymer and using the alkaline solution after curing for 1, 3, and 7 days. Moreover, the PSA-based geopolymer samples containing Na₂SiO₃ had significantly higher UCS and Es, as shown in Figs. 11b and 12b, than those incorporating NaOH, as demonstrated in Figs. 11a and 12a.

Table 6 UCS test results.
Full size table
Fig. 10
Fig. 10
Full size image

Stress-strain curves of all stabilized soil specimens with PSA-based geopolymer with: a (5 M) NaOH solution at (a) 1 day, (b) 3 days, (c) 7 days; and a (3 M) Na₂SiO₃ solution at (d) 1 day, (e) 3 days, (f) 7 days.

This improvement can be attributed to the rise in pH and the presence of more geopolymer precursors of the alkaline activator content. This resulted in an enhanced leaching process of silicon and aluminium from the amorphous phase of the PSA, leading to increased formation of N-A-S-H and C-A-S-H gels between sand soil particles. Therefore, C-A-S-H is a calcium-rich geopolymer system and is more stable than N-A-S-H, due to the exchange of Ca2+ ions for Na+23,30,31.

On the other hand, in order to assess the effectiveness of enhancing the quality of sandy soil by stability, the UCS results were compared with the specified minimum UCS standard for stabilized soils outlined in ASTM D4609. The Standard Guide for Evaluating the Effectiveness of Admixtures for Fine-grained Soil stabilization suggests that if an increase in UCS of 345 kPa or more due to treatment occurs, then the treatment may be considered effective73. Therefore, as shown in Fig. 11, the UCS of stabilized sandy soil samples for all added percentages of PSA containing Na₂SiO₃ solution and of all curing times showed effective improvement more than NaOH solution in the sand soil stabilization. It was shown that the improvement in compressive strength (compressive strength growth rate) was greater when 10% PSA was added and (3 M) Na2SiO₃ was used, the optimum ratio used in this research.

Fig. 11
Fig. 11
Full size image

The UCS of all stabilized soil specimens with PSA-based geopolymer with a: (a) (5 M) NaOH solution; (b) (3 M) Na₂SiO₃ solution.

Fig. 12
Fig. 12
Full size image

The Es results of all stabilized soil samples with PSA-based geopolymer with a: (a) (5 M) NaOH solution; (b) (3 M) Na₂SiO₃ solution.

CBR test

Soil specimens stabilized with varying percentages of PSA addition have shown varying increases in CBR values compared to unstabilized sand soil. When PSA was added to a NaOH solution at percentages of 5%, 10%, and 15%, the CBR values increased by 17%, 25%, and 35%, respectively. while, using (3 M) Na2SiO3 solution added to PSA at percentages of 5%, 10%, and 15%, the CBR values increased by 94%, 110%, and 130%, respectively. Figure 13 illustrates the impact of adding 5,10, and 15% PSA in sand soil, using an alkaline solution, on the CBR value for control (unstabilized) and stabilized sand soil in soaked conditions. Moreover, it is clear from the results that CBR values increase in stabilized soil samples with PSA-based geopolymer with Na₂SiO₃ compared to the containing NaOH.

Using Na2SiO3 solution in the alkaline activation process led to the acceleration of monocarboaluminate production. The Si and Al ions underwent polymerization with Ca2+ions, resulting in the formation of many gel products that overlapped such as highly cementitious calcium aluminosilicate hydrate23. This process had a positive effect on the enhancement of the strength of the stabilized soil and loading support. The reactions continued to progress even in water. Water facilitates the dissolution of alumina silicates in the soil and stabilizer mixture, enabling the formation of geopolymer gels. It increased the CBR value of sandy soil stabilized with PSA-activated geopolymer, shown in Fig. 13. The research results indicate that using a Na₂SiO₃ solution instead of a NaOH solution to stabilize sandy soils using a PSA-based geopolymer is more effective.

Fig. 13
Fig. 13
Full size image

Changes in the CBR value of soil stabilized with PSA content, alkaline activator, and unstabilized soil.

Direct shear test

The ultimate shear strength of stabilized samples with geopolymer with a (3 M) Na2SiO3 is reached faster within 2.5 to 10 mm (horizontal displacements), compared to stabilized samples with geopolymer with a (5 M) NaOH, which occurs around 6 to 10 mm, as shown in Fig. 14.

Fig. 14
Fig. 14
Full size image

Shear strength vis horizontal displacement was determined for PSA-based geopolymer-stabilized soil using: (a) (5 M) NaOH solution; (b) (3 M) Na₂SiO₃ solution.

Figure 15shows the effect of the addition of the PSA content on the cohesion parameter (C) and internal friction angle (ϕ) of the stabilized and unstabilized soils at 3 days of curing. For the unstabilized samples (natural subgrade soils), C and ϕ were found to be 0 kPa and 35°, respectively. Previous studies showed that adding geopolymer increased the shear strength parameters of stabilized soils74. The results showed that the addition of PSA resulted in an increased C, ϕ, and shear strength of the stabilized soil in both alkaline solutions used, as shown in Table 7. In the stabilized samples, when PSA was added at a rate of 5%, the shear strength exhibited greater values than the other additive ratios (i.e., 10%, 15%) and also compared to the unstabilized samples (control). The findings indicated that shear strength decreases with decreasing dry density of stabilized soil and also with increasing PSA addition ratio.

However, the data clearly shows that increasing the PSA content from 5 to 15% significantly enhanced the cohesiveness (C) of the natural subgrade soil. The recorded cohesion for stabilized soil ranged from 5.7 to 26.0 kPa. The results presented a positive relationship between the %PSA addition and C. Stabilized soil specimens demonstrated an increase in C with both from alkaline solutions. Table 7 shows that the cohesiveness of stabilized sand soil is significantly enhanced by using a PSA-based geopolymer with a NaOH solution, particularly when compared to the Na₂SiO₃ solution. NaOH acts primarily as a soil activator, stimulating the dissolution of soil minerals and promoting ionic exchange processes, which leads to improved molecule bonding. Using Na2SiO3 solution decreased the bulk density of the stabilized samples and the shear stress. This had a clear effect on the stable soil cohesion values.

Table 7 illustrates an inverse correlation between the proportion of PSA added and ϕ. At a PSA addition of 5%, the observed ϕ value was the greatest. This might be attributed to bigger particle sizes exhibiting a greater friction angle than sandy soils with smaller particles that include 10 or 15% PSA75. This is also attributed that the highest MDD for the stabilized soils was obtained with the addition of 5% of PSA. However, when the PSA percentage increased, the ϕ value declined in both alkaline solutions. Furthermore, it was observed that the increase in the ϕ in the stabilized soil samples treated with PSA-based geopolymer using a (3 M) Na₂SiO₃ solution is higher compared to the stabilized soil samples treated with PSA-based geopolymer using a (5 M) NaOH solution.

Table 7 Direct shear test results of unstabilized and stabilized soil specimens of different PSA content.
Full size table
Fig. 15
Fig. 15
Full size image

Direct shear results of natural subgrade and stabilized soil with 5, 10, and 15% PSA-based geopolymer after 3 days of curing time with the utilization of an (a) (5 M) NaOH solution; (b) (3 M) Na₂SiO₃ solution.

Repetitive static plate load test

This study developed a set of testing devices to predict the permanent deformation of natural subgrade and stabilized soils subjected to loading cycles. The results shown in Fig. 16 showed a rapid accumulation of permanent stress at the beginning of loading for both natural subgrade and stabilized soil using PSA-based geopolymers. These obtained curves align with the findings proposed by Werkmeister et al.76, which were a series of cyclic tests with a single cyclic stress level. Cyclic loading in subgrade soils causes particles to accumulate and rearrange, resulting in denser formations and subsequent settlement, which creates more damaging permanent stress of (0.075) compared to stabilized soils77.

On the other hand, adding PSA to natural soil reduced voids and increased stiffness, limiting particle movement, improving soil cohesion, and increasing its resistance to repeated deformations. This resulted in a sharp decrease in permanent strain without significant deterioration (enhanced performance) with the increasing number of cycles. Comparable behavior was observed by Abdallah et al.78, in the initial period of the loading cycles. Therefore, the development trend in the number of cycles of 200 may be a good indicator for predicting permanent deformation in the natural subgrade and geopolymer-stabilized soil layer.

Additionally, the stabilized soil with PSA-based geopolymer using Na₂SiO₃ solution showed a permanent deformation less than that of geopolymer-stabilized soil using NaOH solutions, by (0.025, 0.043), indicating a decrease of 300% and 174% compared to the unstabilized soil, respectively. This can be attributed to the significant increase in geopolymerization processes, which resulted in the production of cementitious products (cementitious gel) between soil particles. These products increased the stiffness of geopolymer-stabilized soil and significantly reduced permanent deformation.

Finally, it was observed that the rate of increase in permanent strain was faster and exhibited greater deterioration in natural soils compared to geopolymer-stabilized soils as the number of cycles increased.

Fig. 16
Fig. 16
Full size image

Permanent strain versus number of load cycles.

XRD analysis

XRD analysis investigates the composition of the mineral material and the distribution of crystalline peaks of its constituent elements resulting from geopolymerization processes in stabilized soil, which arise from the interaction of geopolymer precursors and soil79,80. XRD study was conducted on both the natural subgrade soil and the geopolymer-stabilized soil specimens containing 10% PSA with a (3 M) Na₂SiO₃ to find out the mineral changes in this sample that gave an accelerated growth rate of resistance during the stabilization process. The analysis was performed after 7 days of curing, as shown in Fig. 17. Observations reveal that in soil samples stabilized with 10% PSA, some soil crystals display reduced intensity or complete absence of their crystalline peaks, which are located at an angle of 2 Theta < 70 degrees. This indicates that the alkaline solution (Na₂SiO₃) reactions and PSA and the sandy soils under the action of curing time generated the metastable state are amorphous phases. The increase in CaO content in the precursor (PSA) leads to a noticeable enhancement in strength. This improvement can be explained by the formation of main hydration products, such as calcium-aluminate silicate-hydrate (C-A-S-H) and calcium-silicate-hydrate (C-S-H). These products are formed simultaneously with sodium aluminosilicate hydrate (N-A-S-H) gels during the geopolymerization process32,33,36. The presence of these products results in a strong interfacial bond between the geopolymer and sandy soil.

In addition, the quartz peak during 7 days in the sandy soil also decreased, resulting from its dissolution with PSA as seen in the decreased intensity of the peaks at 2θ 26.6°. The XRD analysis showed that the sandy soil also contributed to the establishment of a robust interfacial connection between PSA-based geopolymers. The type of soil had a notable impact on the stabilizing process.

Fig. 17
Fig. 17
Full size image

XRD analysis for natural subgrade soil and geopolymer-stabilized soil.

SEM images

The SEM microstructure investigation included the examination of the fractured surfaces of the PSA-based geopolymer-stabilized soil samples, which were subjected to a UCS test and cured for seven days. The examination of hydration products was conducted using a high magnification of 1500 times. Figure 18 shows the SEM images of stabilized soil samples containing varying proportions of PSA-based geopolymer, namely 5%, 10%, and 15%, using a Na₂SiO₃ solution. SEM image presented that the use of varying quantities of geopolymer resulted in an important enhancement in the sand soil structure. The 5% geopolymer samples exhibited more cracks and with large pores, leading to diminished cohesion between stabilized soil particles and therefore a reduction in UCS values than the other addition percentages, i.e. 10 and 15%, as shown in Fig. 18a. As the PSA addition percentages increase, as is evident in the Fig. 18b, pore size decreases. The elevated pH of the alkaline solution creates a favorable setting for the interaction of alumina, silica, and other minerals with the CaO contained in PSA and sandy soil, resulting in the formation of cementitious products, such as sodium aluminosilicate hydrate (N-A-S-H) and calcium aluminosilicate hydrate (C-A-S-H) gels, which form between the particles of soil, as shown in Fig. 18b and c. This causes also the Ca-O bonds to become more susceptible to breaking compared to the Si-O and Al-O bonds31,81. As a result, the Ca2+ ions are released more rapidly and undergo re-polymerization, leading to the formation of these two primary hydration products and enhancing the strength of soil stabilized with PSA-based geopolymers. PSA geopolymers have been shown to improve the strength of sandy soil by altering its structure, transforming it from an incohesive material with low or almost no strength unsuitable for road subgrade into a cohesive material with sufficient compressive strength that meets the use requirements.

Fig. 18
Fig. 18
Full size image

SEM images and patterns of soil samples were stabilized using PSA-based geopolymer with addition ratios: (a): 5%, (b): 10% and (c): 15%.

Conclusions

This research examines the use of NaOH and Na₂SiO₃ as alkaline activators in PSA-based geopolymers (by adding 5%, 10%, and 15%) to enhance the mechanical characteristics of poorly graded sandy soil. The key results from this research, which included analyzing compaction characteristics, UCS, CBR, DST, and assessing soil structure on samples treated with PSA-geopolymer, and also studying the permanent deformation of natural subgrade and stabilized soil layers subjected to loading cycles, may be described as follows:

  • It was found that adding PSA into sand soil reduces the maximum dry density value and optimum moisture content.

  • The geopolymer-stabilized sand soil samples with Na₂SiO₃ demonstrated higher unconfined compressive strength, elastic modulus and effective improvement than those with NaOH. The study found that adding 10% of PSA to the weight of the dry sandy soil specimen and utilizing 3 M Na₂SiO₃ is the most optimal ratio.

  • Soil specimens stabilized with varying percentages of PSA addition have shown varying increases in CBR values compared to unstabilized sand soil. CBR values increase and are more effective in soil samples stabilized with PSA-based geopolymer when Na₂SiO₃ is employed compared to NaOH.

  • The results showed an inverse correlation between the proportion of PSA added and ϕ in both alkaline solutions. At a PSA addition of 5%, the observed ϕ value was the greatest. It was observed that the cohesion of stabilized sandy soils is significantly enhanced using PSA-based geopolymers with NaOH solutions, especially when compared to Na₂SiO₃ solutions. The Na₂SiO₃ solution decreased the bulk density and shear stress of the stabilized samples. This significantly impacted the cohesiveness values of the stabilized soil.

  • Microscopic analysis revealed that the addition of 10% PSA led to a denser and more uniform structure, resulting in increased formation of C-A-S-H and N-A-S-H gels and enhancing the mechanical properties of stabilized soil.

  • It was found that soil stabilized using PSA-based geopolymer using Na₂SiO₃ solution showed less permanent deformation accumulation and more stable behavior when subjected to repeated loading cycles compared to unstabilized soil.