Introduction

Construction of embankments over soft deposits may encounter issues including excesive settlement, significant lateral deformation, and regional or global instability. Compaction piles, displacement type, and replacement type, as well as vacuum pre-consolidation, the utilisation of pre-fabricated vertical drains, and soil reinforcing, are among the several strategies that can be employed to enhance the strength of cohesive soils1. The utilisation of granular piles, also known as stone columns, for soil stabilisation in to address various constructions, including embankments and buildings, is seeing an increasing trend2,3,4,5,6,7,8,9,10,11,12. Granular columns cause soft soil to consolidate more quickly because they have a shorter drainage channel, higher load bearing capacity, and lower settlement13,14,15,16,17. However, because of low lateral strength, building granular columns having undrained strength (Su) below 15 kPa in soft grounds is nearly difficult18. To eliminate this limitation, one possible solution is to encase the granular column with a geosynthetic material. This additional layer of confinement enhances the shear resistance of the column, hence preventing excessive bulging16,19,20,21,22,23.

However, several pressing problems have arisen as a result of the demographic increase and the expansion of urban infrastructure, including the large global constructing materials requirement that is impacted by the depletion of ecological materials, and the enormous stockpiles of waste which are building up in landfills. Because it has become more complex and expensive to get virgin resources, recycling and reusing garbage is a popular choice among governments and other top decision-makers24. Consequently, several initiatives have been taken to utilise industrial wastes as substitute construction materials. The majority of earlier research centred on using industrial wastes like fly ash, blast furnace slag, red mud, and others in various civil projects25,26,27,28,29,30. One of the most promising waste products, steel slag, has been the subject of several studies31,32,33. Estimates put the steel slag output at 10–15% of that of basic metal steel (US Geological Survey 2020). Data from the World Steel Association34show that only 11.5 million tonnes of the 15.7 million tonnes of steel slag produced as a by-product within Europeare recycled35. The use of steel slag again accounts for around 30% of all steel slag in China, the world’s largest steel producer. In other words, China produces 80 million tonnes of steel slag year annually36. About 60% of the complete steel slag output is utilised as a substitute material in the building sector in industrialised nations, compared to 20% in developing countries37,38. As a soil enhancement material, Poh et al.39 combined steel slag fines with two distinct activators, Na2SiO3 (sodium metasilicate) and CaO (quicklime). The findings indicated that using steel slag fines increased strength and durability while lowering the soil’s propensity for expansion. Shen et al.40 claimed that using steel slag and other solid waste products combination for a base material of road produced specimens with an even better strength for a long time compared to specimens fixed with cement and a higher initial strength than lime-fly ash or lime-soil mixes. Akinwumi et al.41 researched the effects of putting pulverised steel slag over a lateritic soil’s geotechnical characteristics. The findings demonstrated that increasing the amount of steel slag in soil samples from five to ten% reduced the soil’s flexibility and swelling potential while improving its permeability and cured strength. Wu et al.42 developed a superior slag-based composite with better cementation efficacy to alter the expansive soil characteristics. The test findings demonstrated that cementing the slag significantly decreased the possibility of swelling and enhanced the strength properties of the soil. Lang et al.43 examined the efficacy of adding steel slag to stabilise dredging sludge. The findings suggested that a steel slag percentage of between 5 and 10% was ideal for enhancing the shear strength of dredged sludge stabilised with cement. In their study, Gulsen et al.44 examined the potential use of steel slags, specifically those from basic oxygen furnace (BOF) and electric arc furnace (EAF), as backfill material in coastal constructions. The study found that the density, porosity, water absorption, and Los Angeles abrasion ratios of steel slags were consistently greater than those of natural aggregates. It is recommended that EAF slags undergo a minimum ageing period of 6 months, whereas BOF slags require at least 24 months of ageing in open air conditions before they may be utilised as backfill material in coastal constructions. Additionally, the maximum particle size should be lowered to 25 mm. Viorica et al.45 conducted a comprehensive analysis of the chemical, structural, and morphological characteristics of three distinct slag samples that were examined for their potential use in agricultural recycling. The results demonstrate that the combination of landfill slag waste and granulated metallurgical slag can effectively be utilised to restore and preserve damaged areas. Despite all the existing studies, the utilise steel slag as an infill material for column reinforcing the soft soil has not been studied in detail. This paper shows a numerical analysis to understand the working behaviour of ordinary steel slag column (OSSC) reinforcing the soft soil. Additionally, research has been done to understand how encased steel slag column (ESSC) behaves. A comparison was made to determine the response of virgin soil, OSSC and ESSC on various parameters like settlement, excess pore water pressure (Excess PWP), stress concentration ratio (SCR) and lateral deformation of columns. All the said analyses were done for the total construction period of the embankment i.e., 60 days.

Fig. 1
figure 1

3D cut representation of the embankment supported by encased stone columns and sitting on soft clay.

Full size image

Numerical analysis

Geometry

In order to simplify the actual geometry of the problem, which typically requires a moderately large number of columns and to solve this problem, a basic geometrical model was used as shown in Fig. 1. Cross-section and plan views of the embankment slice lying on ESSC reinforced soft clay are displayed in Fig. 2, along with numerous monitoring points that were utilised to analyse the simulation findings. The problem is symmetrical, so one half of the embankment cross-section is modelled (in this case the right half is chosen). In the model, a soft clay bed was made on a plane having a breadth 48 m, a thickness of 2.5 m, & a depth of 10 m. A 5 m embankment is created in three layers on top of soil bed having a thickness 2.5 m. The height of the first layer and second layer is 2 m each and height of the third layer is 1 m. The width of the embankment crest is 8 m and width of the whole embankment up to the toe of embankment from origin is 18 m. The slope of embankment has an inclination of 1 V:2 H. Stone columns are provided below the embankment for the entire embankment’s depth in clay bed i.e., 10 m deeper. Stone column’s diameter is considered as 0.75 m (radius = 0.375 m) and a 2.5 m centre-to-centre distance. The first column is at a distance of 1.25 m from the origin and last column is at distance of 16.25 m from origin.

Fig. 2
figure 2

Details displaying cross section & plan of the embankment lying on soft clay improves with ESSC.

Full size image

Material properties and meshing

The constitutive behaviour of the granular columns & embankment soil is simulated in the current investigation using an elastic perfectly plastic Mohr-Coulomb model. A soft soil model was used to model the foundation clay. Similar analysis was adopted by various previous researchers8,23. The water level is considered to be at surface level throughout the investigation, which is typical for very soft clay soils. The material model and related parameters utilised in the investigation are shown in Table 1. Using a linearly elastic model, the geotextile employed for the encasement was modelled as a thin element that could only withstand tensile force. Ten nodded tetrahedral components were used to represent the soft clay, the embankment, & the sand platform, while six nodded triangular surface elements were used to model the geotextile. Twelve nodded interface components of insignificant thickness with a 0.8 strength reduction factor were used to mimic the soil-geosynthetic interaction.

For meshing the whole geometry, a coarse element distribution type was chosen having a global scale factor of 1.20 and a minimum element size factor of 5 × 10−3. For virgin soil total number of generated nodes, stress points and elements were 814, 1400 and 350 respectively. Similarly, for the soft clay bed reinforced with OSSC the total number of nodes, stress points and elements were 23,719, 65,992 and 16,498 respectively. Also, when the same soft clay bed was reinforced with ESSC the total number of nodes, stress points and elements were 32,693, 66,244 and 16,561 respectively. The final meshed diagram for each of the three types i.e., virgin soil, OSSC, and ESSC are represented in Fig. 3.

Model sequence

Model is constructed for three cases (Virgin Soil, OSSC, ESSC). In all cases firstly a soft clay bed is created then embankment is created on the top of clay layer using a three-stage construction sequence to achieve 5 m of height. First two stage was comprised of two substages, one was used for the construction of the embankment (each 2 m height) and another as consolidation stage. The third stage was constructed for a height of 1 m. Initially, first two layers of embankment is constructed for 10 days each and was allowed to be consolidated for 15 days each. Finally, third layer of the embankment is constructed and time interval for the construction of third layer is 10 days.

Table 1 Material properties used in the numerical analyses.
Full size table
Fig. 3
figure 3

Generated Mesh for (a) Virgin Soil without any reinforcement (b) OSSC (c) ESSC.

Full size image

In the virgin soil model after the construction of the embankment, points are selected for the curve for various analyses. In ordinary steel slag stone column (OSSC) model after construction of the embankment, stone columns are provided below the embankment for the whole depth of the embankment in clay bed and then points are selected for the curve. In encased steel slag stone column (ESSC) model, after construction of the embankment, stone columns are provided below the embankment for the whole depth in the clay bed & then geogrid is used to surround stone columns for the complete column’s length, afterwards points are selected for the curve.

To reduce the boundary impact, the boundary’s distance from the toe of the embankment was determined to be 48 m using sensitivity analysis. The displacement and hydraulic boundaries are chosen in accordance with symmetry. The model is given a full-fixity condition at its base and zero displacements in the direction orthogonal to the symmetric plane. A no-flow boundary state was set up on the plane of symmetry and only vertical flow was allowed through the model’s base at the model’s core.

Numerical results and discussions

Displacement analysis

To ascertain the displacement behaviour in relation to time, the head of the first (A point in Fig. 2) & last column (B point in Fig. 2) were selected for the analysis of all three case of Encased steel slag column (ESSC), Ordinary steel slag column (OSSC), soil with no reinforcement (Virgin Soil) and the corresponding nodal points were chosen. The mean resulting displacement on the column’s tops were calculated, permitting the time-displacement graph up to the construction period (60 days) to be drawn accordingly, which is given in Fig. 4.

Fig. 4
figure 4

Time-Displacement graph at the head of (a) 1st column (b) 7th column (c)1st and 7th column.

Full size image

From Fig. 4(a), it is found that, by using OSSC, at the head of the first column, displacement decreases by 1.48 times than that of virgin soil, and by using ESSC, displacement decreased by 1.23 times than that of OSSC and by 2.43 times than that of virgin soil. From Fig. 4(b), at the head of the last column, it is found by using OSSC, displacement decreases by 1.85 times than that of virgin soil and by using ESSC, displacement decreased by 1.59 times than that of OSSC and by 2.35 times than that of virgin soil.

By reducing the embankment load transmitted to the clay layer and improving the stiffness of the stone column, our findings show how the geosynthetic encasement may significantly minimise the settling of the soft ground. Various other researchers also saw the same enhancement (Hosseinpour et al. 2015; Murugesan and Rajagopal 2010).

Stress investigation

To ascertain the stress behaviour in relation to time, the centre of the first column at the mid-depth (point C in Fig. 2) was selected and the investigation was done to study the response of excess pore water pressure (Excess PWP). The resultant excess PWP on the said point for virgin soil, OSSC and ESSC due to the embankment loading (in the last stage of embankment construction that is, 60 days) is represented in Fig. 5.

Fig. 5
figure 5

Generation of Excess Pore Water Pressure in final stage of building phase for virgin soil, OSSC and ESSC.

Full size image

Figure 5 shows the temporal variation of Excess PWP response, and it is found that by including the OSSC, the generated excess PWP was reduced by 1.15 times than unreinforced case and by further encasing the OSSC i.e., ESSC the excess PWP further reduces by 1.08 times as compared to OSSC and by 1.25 times than that of unreinforced case. Additionally, the addition of OSSC and ESSC accelerated the Excess Pore Water Pressure dissipation rate.

At the conclusion of the embankment building stage, Fig. 6 shows the distribution of developed Excess PWP owing to embankment loading for various situations of unreinforced, OSSC and ESSC reinforced soils. As anticipated, there is non-uniformity in the distribution of produced surplus pore pressure at the completion of the building stage. In each case, the highest pore pressure was found below the embankment crown. Additionally, there is greater pore pressure creation in the area between the first two columns under the crown. Beyond the embankment’s toe, the region of clayey soils does not suffer much of an increase in pore pressure production.

Stress concentration ratio

An index called Stress Concentration Ratio (SCR) is a measure of transmission of embankment loading through stone columns. The SCR is the relation of stress in foundation soil to stress in the stone column at an identical location. The more rigid stone columns bear greater weight compared to the fragile soil in a ground supported by stone columns.

To ascertain the impact of OSSC & ESSC on stress concentration behaviour with respect to time, a point at a mid depth of the first column and another point just adjacent to it in the clay bed at an identical location (point C & point D respectively in Fig. 2) were chosen, time-stress concentration curves were drawn accordingly and is represented in Fig. 7. In Fig. 7, the Stress Concentration Ratio for both ESSC and OSSC vary over time. The SCR is not uniform i.e., changes over time for each of the situations taken into consideration, as shown. For the first 10 days or so after building an embankment, the SCR for these circumstances stays constant (very close to one), but with time, it drops because columns share more stresses than the nearby soft clay. Intriguingly, a rise in OSSC can be seen at the conclusion of the embankment building period, indicating that significantly larger pressures were transmitted to the nearby soil. But when columns are covered in geosynthetic material, SCR drops far less than OSSC. It is found that by using OSSC stress is reduced by 1.15 times than that of virgin Soil and by using ESSC stress is reduced by 1.08 times than that of OSSC and by 1.25 times than that of virgin soil.

The collapse or plastic points for clayey soil strengthened with ESSC of various configurations are shown in Fig. 8 at the conclusion of the building stage. Plastic points can be seen in the embankment structure, at the column junction, and throughout the construction period. Plastic points are more prevalent around the column and in the embankment’s body, but they are comparatively less prevalent for columns that are close to the toe.

Figure 9 illustrates how the geogrid encasement affects the principal effective stress produced on the column material and the nearby soft clay beneath the embankment. The findings of the principal effective stress at the head location of each column at the conclusion of the construction stage are displayed. The principal effective stress of unreinforced soft clay increases at the conclusion of the consolidation stage due to an extra rise in geostatic stress caused by embankment construction, however this increase differs based on the geometry of the embankment. It is most intense underneath the embankment’s crest and gradually lessens on the sloped face of the embankment. When OSSC are introduced, column material is subjected to greater effective stress than the nearby soft clay. And, on further addition of encasement to the OSSC (i.e., ESSC) results in higher generation of the effective stress. Principal effective stress mobilised in soft clay with the addition of ESSC was substantially lower for equivalent locations than OSSC and for soft clay without any reinforcement. As it is evident from Fig. 9, the first three columns which are below the crest of the embankment have higher principal stresses as compared to other columns.

Fig. 6
figure 6

Flow of the developed excess pore pressure at final stage of embankment building (60 days) for (a) virgin soil (b) OSSC (c) ESSC.

Full size image
Fig. 7
figure 7

Influence of OSSC and ESSC on stress concentration ratio.

Full size image
Fig. 8
figure 8

Developed plastic points at final construction stage (a) Unreinforced (b) OSSC (c) ESSC.

Full size image
Fig. 9
figure 9

Comparison of variation of principal effective stress below the embankment for the unreinforced case, OSSC and ESSC.

Full size image

Lateral displacement of columns

At the conclusion of the consolidation period, Fig. 10 depicts the patterns of column deformation under embankment stress. Columns are resting on soft clay strengthened by OSSC and ESSC. It is seen that columns beneath the embankment’s crest bulge, but throughout the sloping area, column bending appears to predominate. It can be observed from Fig. 10, that the lateral deformation of all the columns keeps on increasing as we move away from the crest portion to the sloped region. This variation was observed in both OSSCs and ESSCs. To comprehend the impact of OSSC and ESSC given the variation in deformation patterns, columns 2 & 7’s lateral deformation has been examined which is being shown in Fig. 11. From Fig. 11, it is found that by using ESSC the maximum lateral displacement decreased by 1.09 times than that of OSSC at the second column and by 1.03 times for 7th column. However, by encasing the stone column, the lateral bulging has significantly reduced, mostly because of the increased constrainment the encasing provides, as also suggested by other studies23,46.

Fig. 10
figure 10

Lateral displacement of all columns for (a) OSSC (b) ESSC.

Full size image
Fig. 11
figure 11

Comparison of lateral deflection of column under OSSC and ESSC for (a) 2nd column (b) 7th column.

Full size image

Comparison of current work with previous research

While it is not feasible to directly compare this study with existing literature due to variations in soil type, model setup, loading criteria, and material parameters, an effort has been made to assess the results to validate this study’s effectiveness in relation to previous research. As there are not much study available on the use of steel slag as a granular column under embankment loading where time vs. settlement graphs can be compared, therefore a comparison has been made from the data of geosynthetic encased granular column to understand the trend. Figure 12 represents the comparison of load-settlement variation of the current study with the existing literature.

Rajesh and Jain47, reported the findings of a series of numerical investigations conducted to assess the time-dependent performance of ordinary stone columns (OSC) and geosynthetic-encased stone columns (GESC). The tests were conducted with an embankment height of 6 m. The substrata utilised in the study consisted of a layer of soft clay with a thickness of 10 m, which was afterwards followed by a layer of hard rock strata. For both the Overlying Sand Cover (OSC) and the Geosynthetic Encased Sand Column (GESC) methods, a 1-meter-thick sand mat or drainage blanket is positioned on the surface of the soft clay layer. The groundwater table was located at the uppermost surface of the clay layer. The stone column, with a diameter of 400 mm, is installed by driving it from the sand mat and extending it to the complete depth of the soft clay. This results in a length of the stone column, denoted as LS, measuring 11 m. In this work, a replacement ratio of 9% is utilised, with the diameter of the axisymmetric unit cell set at 1.338 m. The study determined that the ultimate settlement of OSC and GESC exhibited a reduction of 43% and 61%, respectively, in comparison to soft clay. A significant decrease in the duration required for the consolidation of ground treated with GESC was seen in comparison to ground treated with OSC.

Yoo48 did the analysis which pertains to a 50-meter-wide embankment featuring side slopes with a 1:1 vertical-to-horizontal ratio. The embankment is constructed using sandy soil and is situated above a homogenous clay layer that is 10 m thick. Below the clay layer, there exists a stiff layer. A square grid-shaped GESCs with a 0.8 m diameter constructed of crushed stone is installed. Before the embankment fill is put, a sand mat with a thickness of 1 m is installed for column drainage. The findings suggest that the inclusion of geosynthetic encasement leads to increased stiffness of the stone column and a decrease in the amount of embankment load transferred to the soft ground. Consequently, this results in a reduction in overall settlement.

The detailed methodology, soil type, loading conditions, material properties and other parameters can be obtained from the referred study. As observed, a higher settlement is observed for untreated soil which reduces when it is being treated with ordinary steel slag column/granular column and further reduces when reinforced with encased steel slag column/granular column. The findings of the current study align with and can be compared to the patterns documented in prior investigations.

Fig. 12
figure 12

Comparison of load-settlement behaviour of present study with existing literature.

Full size image

Conclusion

In the current study, it was looked into how an embankment sitting on soft clay and reinforced with steel slag columns would behave over time both with and without encasing. To calculate the influence of the reinforcing method against embankment loading, a numerical model was created. A parametric analysis was performed to comprehend how OSSC and ESSC behave on various outputs such as settlement, excess PWP, SCR and lateral deformation of columns. The following conclusions from the current investigation can be made based on the observed results:

  1. 1.

    The settlement of soft clay at the end of the embankment construction phase has been significantly higher for the chosen configuration and model characteristics, compared to OSSC. Enclosing the column in a suitably stiffer geosynthetic material will result in a further reduction in settlement.

  2. 2.

    Underneath the embankment, where soft clay produces the maximum excess PWP, the amount of excess PWP decreases towards the embankment slopes. Excess PWP created on the untreated soft clay at the conclusion of the embankment building phase has been greater than OSSC for the selected monitoring location. When ESSC was employed, the excess PWP value was further decreased.

  3. 3.

    Stress concentration ratio (SCR) defined as stress in soft soil to the column at equal elevation decreases for OSSC as compared to the unreinforced soil. This decrease is even more pronounced for ESSC, indicating that ESSC is more successful than OSSC as a means of reinforcement.

  4. 4.

    Principal effective stress increases over the column compared to the surrounding soil when ESSC is used as compared to OSSC and unreinforced soil. Maximum stresses were experienced on the columns under the crest of the embankment.

  5. 5.

    The load capacity and rigidity of steel slag columns can be increased by completely encasing them in geosynthetic material. It is found that the lateral bulging of the stone columns is decreased and that the geosynthetic encasement contains them.