Various test methods were accomplished to analyze the physicochemical characteristics of the two materials of red mud and fly ash. Their composition and microstructure were studied, which provided a basis for the theoretical analysis of geopolymer formation.

Table 1 clearly shows that red mud and fly ash widely exist in silica and alumina, which provides a theoretical basis for their preparation of geopolymers. At the same time, red mud with strong alkali properties promotes the production of geopolymers.

Geopolymers are inorganic polymer structures with binding effects and are made of alkalinous reagents (e.g., sodium hydroxide and potassium hydroxide) and aluminum silicate (such as fly ash). During the formation of the geopolymer, the silica oxygen bond and the aluminum oxide bond of aluminate break under the action of alkali solution, forming a series of silicon oxygen tetrahedron and aluminum oxygen tetrahedron units in an oligomeric state. In the process of the reaction, these oligomeric tetrahedron units are gradually dehydrated and repolymerized to form the geopolymer. The reaction process includes the hardening of the inorganic polymer structure, the precipitation of hydration products and the dissolution and polymerization of aluminosilicate material.

The formation mechanism of geopolymers in red mud is different from that of traditional geopolymers. After the mixture of red mud and a kind of fly ash material is excited by sodium hydroxide solution, fly ash particles dissolve in alkaline solution, and the active silicon and active aluminum react to form a gel. However, research shows that SiO₂ in red mud remains in a quartz style, and red mud is mainly used as an inert filler. Through the polymerization reaction of fly ash particles, the active particles and inactive fillers are combined to form a cemented matrix. The polymerization theory points out that the polymer network structure is mainly a stable crystal structure formed by Si, Al and O, while the red mud geopolymer product is not a pure crystalline polymer but a composite material of inert filler. Research shows that the geopolymer formed by fly ash material and red mud can obtain greater strength under certain reaction conditions, and it can be used as a pile material to strengthen red mud foundations.

According to existing studies (Bai and Shi, 2017; Hu et al., 2018a; Bai et al., 2019b), the composition and mix ratio of raw materials, including the concentration of activator and curing condition, could affect the strength characteristics of red mud geopolymers. In addition, the amount of water in the geopolymer can also affect its properties. However, the understanding of the water effect of geopolymers is insufficient. Research suggests that water only provides a reaction environment and dissolves the geopolymer mixtures, while others suggest that water could participate in geopolymer reactions instead of a dissolution solvent (Nie et al., 2019).

Furthermore, the type and optimum concentration of activator should be determined, and an appropriate water-cement ratio should be adjusted for mixing to ensure the curing time and achieve the required strength.

To study the feasibility of red mud geopolymers as pile materials, strength tests of geopolymers generated under different conditions were carried out in the laboratory. A Bayer red mud sample and a kind of first-grade fly ash from the field were used in the experiment. NaOH alkali solution or NaOH alkali solution mixed with water glass (NaSiO₃) was selected as the activator. In fact, the effect of temperature cannot be ignored (Bai et al., 2014; Yang and Bai, 2019; Bai et al., 2021c). For this, the normal temperature (e.g., 20°C) and a high temperature (e.g., 60°C) were selected as different curing conditions (Bai, 2006; Bai et al., 2017b; Bai et al., 2021a).

The test processes were as follows: the red mud was crushed and screened and mixed with fly ash at a mass ratio of 1:1 for 10 min, and the activator was added into the mixture and stirred for 10 min. To guarantee the uniformity of the mixture, the liquid/solid ratio of the test was determined to be 0.5 for economic purposes. During the experiment, the concentration of alkali solution was adjusted from 0 to 10 mol/L (0, 2.5, 5, 7.5, and 10) to determine the optimal concentration. When using a compound of sodium silicate as an activator, the ratio of sodium silicate to sodium hydroxide solution is 2.5. The geopolymer was poured into the test molds and then carefully cured at temperatures of 20 and 60°C for 24 h (Bai and Su, 2012; Bai et al., 2017b). The prepared specimens were removed from the molds and cured at a constant temperature (namely, 20°C). The unconfined compressive strength of the test specimens after reaching the specified age is measured.

As seen from Tables 2, 3, when NaOH alkali solution is used as the activator, the strength of the geopolymers made with red mud and fly ash is low at room temperature after 28 days, and a higher strength can be obtained after 2 months. When the concentration of NaOH alkali solution is 5 mol/L, the unconfined compressive strength of the geopolymers can reach more than 35 MPa. During high-temperature curing, the compressive strength of the geopolymer formed by fly ash material and red mud develops rapidly. When the concentration of NaOH alkali solution is above 5 mol/L, a strength of approximately 15 MPa can be obtained in 7 days, and the strength can reach more than 20 MPa in 28 days. In this case, the optimal concentration of NaOH alkali solution is 7.5 mol/L. It can be interpreted that elevated temperature promotes the alkaline reactivity of the raw materials and enhances the polycondensation reaction of geopolymer, resulting in the rapid growth of early strength of geopolymer under high-temperature curing.

When the exciter was a mixture of water glass and NaOH (namely mixed activator), the geopolymer could obtain a higher strength at normal temperature (20°C), and the maximum unconfined compressive strength at 28 d could reach more than 40 MPa. High-temperature curing could improve the early compressive strength; however, it had no effect on the compressive strength at 28 d.

Therefore, when the geopolymer composed of red mud and fly ash matter is utilized as pile raw materials, the unconfined compressive strength can reach 2–3 MPa under normal temperature curing with NaOH alkali solution as the activating agent, and its long-term strength is high, which can be used in composite foundations with low initial strength requirements. To obtain geopolymer materials with high early strength, high-temperature heating and curing and an activator consisting of a NaOH alkali solution and sodium silicate can be considered. At this time, geopolymer materials can be used as a variety of building materials in different projects. In summary, the geopolymer formed by red mud and fly ash can meet the strength requirements.

The feasibility test of red mud geopolymer piles was carried out at the construction site (Chen and Chen, 2006). Six geopolymer mixing pile tests were accomplished to measure the bearing capacity of the pile body by in situ static load test. The test piles were constructed by wet mixing method. The solid-liquid ratio of fly ash in NaOH solution premix compound slurry is adopted by the laboratory test. After natural curing for 28 d, static load tests were carried out.

1) Static load test

The test site is located in a red mud yard that needs reinforcement treatment in Jiaokou County, Shanxi Province. As subdams are built on the red mud step by step, the original red mud will be used as the foundation. The stability of the red mud foundation becomes the decisive factor for the safety of the whole red mud storage yard. To ensure the safety of the red mud dam, ground treatment is needed.

According to the strengthening mechanism of composite foundations with mixing piles (Verma et al., 2017; Hu et al., 2018b), geopolymers are formed by fly ash and red mud for mixing single piles as well as composite foundations (Meng et al., 2020; Cui et al., 2021; Yuan B et al., 2021). Using multistage load tests, the bearing characteristics of single piles (SPs) and composite single piles (CPs) in the red mud foundation were analyzed. The pile length was 10 m, and the pile diameter was 0.5 m. Through the pre-installed stress gauges and vertical earth pressure gauges, the stress transfer law of the mixing pile and the soil-pile stress relationship were monitored in the loading process.

The mechanical behavior and load transfer mechanism of a single pile can be analyzed by single pile static load test. A row of steel rebar stress gauges is arranged in the center of the pile section. The layout scheme of the stress gauges is shown in Figure 1, where the test single piles were denoted as TSP1, TSP2 and TSP3. The stress gauges were numbered according to the test pile No. and the buried depth. For example, stress gauges setup in TSP1 were numbered with depth as TSP1-SG0, TSP1-SG2, TSP1-SG4, TSP1-SG6, and TSP1-SG8, where TSP1 means the test pile No., SG means the stress gauge, and the last digit means the buried depth of the stress gauge. The vertical load is imposed to the pile top. In the single pile test, the area of the applied load is equal to the area of the pile top. The stress gauge is tested once for each loading stage until the pile is damaged (or the maximum load that can be applied is reached). The axial force of the pile can be obtained by processing the data collected from stress gauges during the static load test. The axial stress of the pile and the soil stress are proportional functions with the same change trend. The distribution of the pile axial force can be obtained by analyzing the change trend of the stress meter force, and the distribution law of the side friction resistance of the soil layer around the pile can be obtained.

Through the static load test of composite single piles, the soil between piles and the vertical stress of the pile body in the loading process of a single pile composite ground is discussed, and the ratio of pile-soil stress in the red mud composite ground is analyzed. To allow the pile and surrounding soil work together, a sand cushion with a thickness of 30 cm was laid on the pile top and surrounding soil surface. In the static load test for composite single pile, the load was applied through a circular steel plate, of 20 mm thick with a diameter of 0.8 m (0.5 m² in area), on the cushion. Stress gauges were installed in the pile center similar to the single pile test, as shown in Figure 1. The test composite single piles were denoted as TCP1, TCP2 and TCP3. Besides, earth pressure gauges (EGs) were setup on the pile top, along the pile-soil interface and in the soil between piles, whose distance from the pile center were 0, 0.4 and 0.8 m, respectively. Earth pressure gauges were numbered according to the buried depth and distance from pile center. In general, the layout of the stress gauges and earth pressure gauges are illustrated in Figure 1.

2) Pile strength analysis

Figure 2 shows the settlement processes of SPs and CPs obtained from the static load test. It should be noted that the unit of the applied load to CPs in Figure 2B has been transformed from kN into kPa according to the loading plate area (0.5 m²), for the sake of comparing and analyzing of the foundation bearing capacity before and after treatment. Furthermore, due to the lack of jack stroke, the maximum loading value (P _u) of TCP1 was 600 kPa, which did not reach the ultimate load. Meanwhile, the vertical ultimate bearing capacity (Q _u) of the red mud geopolymer mixing single pile (i.e., SP) in red mud ground is approximately 300 kN, while it is approximately 780 kPa for the composite foundation (i.e., CPs). According to the characteristic value of the bearing capacity of a single pile, R _a (R _a = Q _u/K = 300/2 kN = 150 kN, K denotes the assurance coefficient), the back-calculated required compressive strength of the pile material, f _cu (f _cu ≥ ηR _a/A _p, A _p denotes the cross-sectional area of pile and η = 4) should be more than 3 MPa, which is consistent with the results obtained by laboratory tests (Table 2). In addition, the bearing capacity of the composite foundation is significantly enhanced compared with that of the red mud ground before treatment. The bearing capacity of the red mud ground before treatment is only 60∼100 kPa. By comparison, the bearing capacity of the red mud ground after the geopolymer mixing pile treatment is increased by more than 4 times. In summary, the mixing piles that made of geopolymers formed by fly ash and red mud can satisfy the bearing capacity requirements, and the composite foundation reinforcement effect is significant.

3) Reinforcement effect

Standard penetration tests (SPT) were carried out on red mud ground before and after the pile installation. Besides, the strength of mixing piles was also tested by executing SPT on samples drilling from the pile body. The results indicate that before the mixing pile installation, the average SPT N-value of the unmodified red mud was 3. After curing treatment, the average SPT N-value of the red mud between piles reached 7, while the average SPT N-value of the mixing pile drilling core could reach 15. According to the comparison of the above tests, the strength of the unaltered red mud was very low, and the bearing capacity was very small. The strength of the red mud after solidification treatment was improved. Meanwhile, the strength of the red mud geopolymer pile has been improved more obviously than that of the unaltered red mud. Therefore, the formation of geopolymers composed of fly ash and red mud significantly improves the strength of red mud and has a good reinforcement effect. The bearing capacity and the overall stability of the red mud dam are better improved.

A comparative analysis of the red mud between piles before/after the treatment was accomplished by laboratory tests. The results show that the mechanical characteristics of the red mud between piles after the reinforcement have obvious changes compared with the original state of red mud. After reinforcement of red mud with fly ash, the red mud consolidation degree increased significantly, the moisture content decreased (12–18%), the porosity ratio increased (12–20%), the density increased (4–7%), the mechanical properties also obviously increased, the compression coefficient decreased significantly (approximately 27–32%), and the compression modulus increased (approximately 30–36%). The cohesion is obviously increased (increased by more than three times), the friction angle is slightly reduced, and the compressive strength is generally increased.

The red mud raw material and the red mud between piles after reinforcement were tested by cone penetration tests (CPT). The comparative results show that the side friction and cone tip resistance of original red mud are relatively low. After reinforcement, the side friction and cone tip resistance of red mud between piles obviously increases, and the average cone tip resistance increases from 0.665 to 1.55 MPa, with a growth up to 133%, and the average pile side friction rose from 15.15 to 29.43 kPa, up to 92.3%. It is demonstrated that the strength of red mud ground can be greatly improved with geopolymer mixing piles.

From the above test results obtained before and after the geopolymer mixing pile construction, one can see that the physical and mechanical properties of the red mud between piles are obviously improved, the strength and bearing capacity of the red mud foundation are greatly enhanced. This verifies the applicability and effectiveness of the proposed ground treatment method.

As the load is relatively small, the deformation of a single pile is not different from that of a single pile composite foundation, and the settlement curve is relatively close. At this time, the load is primarily undertaken by the pile. As the applied load gradually increases, the settlement of the single pile composite foundation is lower than that of the single pile test under the same loading. At this moment, the pile side resistance had basically played a full role and more load will be undertaken by the soil layer between piles. The pressure-settlement relationship (i.e., p−s curve) of the single pile composite foundation slowly continued to develop until the final limit load is reached. At the same time, as the ultimate load is reached, the deformation curve of the single pile composite foundation is gentler owing to the effect of soil between piles.

Figure 3 shows the variation process of stress in the pile body during the loading process of TSP2. The results show that the stress at different depths increases with the applied load, and its growth rate also has a trend of increasing gradually. The stress in the pile body will not continue to increase when the load increases to a certain value, presenting a stable stress state. As the load gradually decreases, the stress decreases correspondingly, and the decreasing rate is relatively stable and less than the loading rate.

In fact, the pile settlement will gradually increase with the applied load, and the pile axial force will gradually change from the elastic state to the elastoplastic state. With increasing applied load, the pile deformation is only plastic deformation, and the growth rate of the pile axial force shows an increasing trend. In the process of unloading, the pile has certain recovery deformation, which will produce a certain upward displacement. After the load is completely removable, there will be some residual stress in the pile in a short time.

For the single pile load test, Figure 4 shows the variation properties of stress in a single pile along the pile depth during the loading process. Specifically, the pile axial stress is mainly concentrated in the upper part of the pile body (approximately 0–4 m). In addition, the axial force of the pile top is the largest. With increasing pile depth, the pile axial stress decreases rapidly, and the greater the load is, the more obvious the decreasing trend is. The test results clearly show that the bearing capacity of a single pile with a length less than 8 m is relatively small. Moreover, the axial force of the pile under 8 m is very small in the red mud ground.

For mixing piles in composite foundation, Figure 5 shows that the stress in the pile body changes with the applied load. The higher the applied load is, the greater the stress in the pile is. This is consistent with the variation rule in the load test of single pile. However, due to the load sharing effect of soil between piles, its growth rate is less than that of a single pile. At this time, the pile and the soil around the pile produce a certain compression deformation, and the combined action produces the ability to bear an external load.

Figure 6 shows the variation properties of stress in the pile body with depth during the load test of the composite foundation. The pile stress gradually decreases with increasing depth. This is similar to the variation trend of pile stress in the load test of a single pile. This effect is more obvious with increasing load. When the depth is below 8 m, the stress in the pile is not large. That is, the effective length of the pile in the red mud composite foundation with a mixing pile is approximately 8 m. It can be inferred that when the applied load achieves the pile failure load, the increase in the axial force of the pile below 8 m is also very limited. For the mixed pile composite foundation in this red mud ground, it is suggested that the pile length should not exceed 10 m.

Figure 7 indicates the variation process of the vertical stress at the pile top and the pile side (0.4 m away from the pile center) and soil between piles (0.8 m away from the pile center) in the loading process. The higher the applied external load is, the greater the stress at the pile top and at different depths of the pile side and soil between piles. In the process of unloading, the stress at the pile top diminishes with decreasing applied external load, while the stress in the pile side and soil among the piles remains almost unchanged. In the process of loading, the soil settles and incurs unrecoverable compression deformation. Therefore, the soil stress at the settlement site below the pile top will remain almost unchanged.

By comparing the soil stress of the same depth at the pile side with that among the piles, it can be seen that the vertical stress of the soil between piles slightly increases with the increase of the distance from the pile center in the lower soil.

Due to the existence of the pile body, the soil deformation can be restrained to a certain extent and can bear more external load. In addition, on account of the downward diffusion of vertical stress, the deformation and stress of the soil between piles are slightly larger than those of the soil beside piles. However, this trend will gradually weaken with the increase of depth, and the final stresses in soil beside pile and between piles are basically the same.

Under the same applied load, the vertical stress of the pile top is much higher than that of the pile side soil. Take TCP1 as an example, under the maximum load of 300 kN, the vertical stress at pile top (TCP1-EG10) is 711.2 kPa, while that at pile side soil (TCP1-EG20) is 34.2 kPa, the stress ratio of pile top to pile side soil is up to 20.8%. This shows that the pile in the red mud composite foundation undergoes the main load, and the contribution of the pile to the bearing capacity is very significant. Under the action of an applied external load (Figure 8), the vertical stress of the soil layer around the pile decreases rapidly with increasing pile depth. The soil stress at zero depth is much larger than that at depths of 2, 4 and 6 m below the top surface. In the red mud composite foundation, the soil stress is generally limited to a depth of 2 m under the ground surface. In the depth range of 2 m below the surface, the soil will produce a large settlement deformation, resulting in differential settlement between the surrounding soil layer and the pile. At this time, the soil on the side of the pile will provide downward friction resistance to the pile, namely, negative friction resistance.

With increasing applied external load, the bearing capacity provided by the red mud geopolymer mixing pile body and red mud between piles is gradually exerted, and the stress is gradually transferred to the deep soil layer. In the depth range of 2 m below the ground surface, the soil between piles may have negative friction resistance. However, when the surface depth is below 2 m, the soil stress at the pile side is very small due to the external load. In this case, the formation of the side friction resistance of the red mud foundation soil at the pile side mainly depends on its gravity stress. The effective pile length of the red mud geopolymer mixing pile in the red mud composite foundation is approximately 8 m. Hence, the pile stress is mainly contributed by the pile side friction resistance.

In addition, the red mud geopolymer composite foundation significantly improves the bearing capacity. As the pile bears the majority of the load and is restrained by red mud, the soil stress between piles increases. This further enhances the effective stress of the ground and the shear strength of the soil layer between piles and pile side friction.

Generally speaking, the bearing characteristics of the composite foundation with mixing piles on the red mud ground are basically consistent with those on the conventional saturated soft clay ground. On the other hand, the geopolymer formed by fly ash and red mud as the pile material has similar properties to the traditional cement soil pile material, which provides support for its engineering applicability.