Research on strength characteristics and mechanism of cement stabilized phosphogypsum materials under dry and wet cycles

Abstract

Phosphogypsum can be used for mine filling, ecological restoration materials, road construction materials, soil conditioning agent, cement retarder, etc., the comprehensive utilization rate of only about 40%, less than 10% for road construction materials, for the huge demand for raw materials road projects, obviously a drop in the bucket. In order to solve the accumulation problem of phosphogypsum, this paper closely combines the major needs of national highway construction, actively responds to the call of Guizhou Provincial People’s Government and Guiyang Municipal People’s Government on the comprehensive utilization of phosphogypsum resources, and researches the strength decay law of cement-stabilized phosphogypsum materials under dry–wet cycle, considering the complex climatic characteristics of dry–wet cycle with the prerequisite of high phosphogypsum mixing, and reveals the mechanism of strength decay, and establishes a strength regression model, and discusses the suitability as pavement base (sub-base) materials. The strength regression model was established, and the suitability of cement stabilized phosphogypsum as the base layer (sub-base layer) of pavement was discussed. The results show that the unconfined compressive strength of cement-stabilized phosphogypsum material decreases with the increase of the number of dry and wet cycles, and the decay is faster in the first four dry and wet cycles, and then gradually tends to level off. The unconfined compressive strength and the number of wet and dry cycles can be fitted with a negative exponential function, and the fitting parameters are all greater than 0.98 . Cement dosage had the most significant effect on unconfined compressive strength with F-value = 988.407, followed by the number of wet and dry cycles and compaction had the least effect. The maximum values of RMSE and MAE were 0.0862 and 0.0776 for the multifactorial regression equation of the unconfined compressive strength of cement-stabilized phosphogypsum materials, which indicated that the regression model had a high reliability. Through the research in this paper, cement stabilized phosphogypsum materials can be used as the base layer (sub-base layer) for highway pavements of various grades under different cement mixing amounts and different traffic volume conditions.

Introduction

Phosphogypsum is a by-product produced during the production of wet process phosphoric acid¹. According to the study, for every 1t of phosphoric acid generated, there are 5t of by-product phosphogypsum², for example, 10⁶t of phosphoramidite per year in Guizhou, there are about 5 × 10⁶t of phosphogypsum emissions. Such a large amount of phosphogypsum emissions will not only pollute the natural environment, but also adversely affect human health. Phosphogypsum resources comprehensive utilization of the main ways of mine filling and ecological restoration materials, road materials, soil conditioning agent, cement retarder, gypsum-based building materials, sulfuric acid, chemical fillers, etc., the comprehensive utilization rate of only about 40%, in a variety of ways of use, phosphogypsum for road materials less than 10%, and for the raw material demand for the huge road project, is obviously a drop in the bucket. To this end, various parts of the country and the Guizhou Provincial People’s Government have issued a series of documents to accelerate the policy measures of phosphogypsum. For example, in March 2024, the Ministry of Industry and Information Technology and other seven departments issued the Action Plan for Comprehensive Utilization of Phosphogypsum. Guizhou Provincial People’s Government on Comprehensively Strengthening the Comprehensive Utilization of Phosphogypsum to Promote the Green Development of Phosphogypsum Phosphorus Chemical Industry (Qianfu Fa [2024] No. 5).

Road engineering has a high demand for raw materials, and the mechanical properties of phosphogypsum mixtures have been studied relatively early and abundantly. Tingting Du et al.³conducted 7d, 28d unconfined compressive strength tests on cemented phosphogypsum materials modified with different dosages of curing agents and liquids concluded that the early strength formation of the mix, in addition to the mosaic effect between the crystals, is mainly provided by the hydration reaction of the cement that occurs to form the cementite strength skeleton. Li Xia⁴ proposed through the study of cement phosphogypsum stabilized materials that phosphogypsum formed a three-dimensional skeleton with Si atoms in cement through cross-linking reaction and consolidation reaction and generated high strength crystals, which led to an increase in the strength of the mixture. Zhang Xiaoqiao et al.⁵ found through the road performance study of phosphogypsum-based pavement base material that under the condition of equal cement dosage, the 7d unconfined compressive strength of cemented phosphogypsum-stabilized gravel increased by 24% compared with that of ordinary cement-stabilized gravel, which indicates that phosphogypsum has the property of enhancing the strength of the mixture. Zuo Yajing et al.⁶ revealed the main reason for the strength formation of phosphogypsum-modified filler through microscopic tests: a denser microstructure, fewer microfractures, and the production of hydrated calcium silicate and calcium alunite. Meskini (2021) et al.⁷ investigated that for different dosages of phosphogypsum, lime, mixing it with fly ash to obtain phosphogypsum-lime-fly ash mixtures, and mechanical properties and mineralogical tests were carried out for different setting cycles. T. P. MASHIFANA (2019) et al.⁸ studied phosphogypsum with lower content of 20% and 30% and modified with fly ash and lime with alkaline oxygen slag. Their role in the evolution of particle size distribution (PSD) on unconfined compressive strength was explored.

The above studies mainly show the strength characteristics of phosphogypsum mixtures, in addition to this, scholars have also explored the feasibility of applying phosphogypsum in highway construction. Huang Wendong et al.⁹ developed a new empirical dynamic constitutive model to study the dynamic properties of phosphogypsum stabilized soil and verified that the dynamic long-term resilient modulus of the mix meets the relevant specification requirements. Chao Liu et al.¹⁰ prepared a cement-phosphogypsum stabilized crushed stone material by partially replacing fine aggregates with phosphogypsum, systematically analyzed the influencing factors of the material’s water-stability and crack-resistance properties, and verified the feasibility of the ratios by experimental road. Chen Kaisheng et al.¹¹for the waste of resources of phosphogypsum and the special engineering characteristics of red clay, cement (lime) as a curing agent, the fatigue characteristics and improvement mechanism of phosphogypsum stabilized soil using SDT-20 dynamic triaxial test to study the fatigue characteristics of phosphogypsum stabilized soil, and the establishment of the cumulative deformation model and the critical dynamic stress model to get the mixture dynamic strength index. Sun, Qiqi¹² investigated the mechanical properties and durability of a new phosphogypsum (PG) pavement material, which was confirmed by experimental results of compressive strength, splitting strength, dry and wet cycles, and freeze and thaw cycles that the mixture could meet the design standards of base material for highways and primary roads. Phosphogypsum processed road construction materials used in highway roadbed pavement at home and abroad are successful test sections. Such as Yunqiqiao to Luojiawan first-class highway on the bed of the quality of the percentage of phosphogypsum: fly ash: gravel = 15:15:70; that is, the quality of the second ash and gravel percentage of 30:70. Different proportioning schemes were adopted for the test section of the He-Tong highway in Shaanxi Province. Namely, lime:phosphogypsum:soil = 5:7:88,lime:phosphogypsum:soil = 6:9:85,lime:phosphogypsum:soil = 7:10:83. In addition to the above research on phosphogypsum soil improvement techniques, other cement, fly ash, and lime foundation improvement techniques have also been widely used in engineering practice. Chompoorat et al.^13,14 experimentally investigated the effect of substitution ratio of fly ash to cement (15% to 60%) on the strength, stiffness and shrinkage potential of improved soft soils, and the results showed that optimization of the fly ash and cement content can significantly improve the mechanical properties of soft soils. Further studies found that ordinary Portland cement or slaked lime-treated coal mine dredging swollen clay showed good correlation between macroscopic mechanical properties and microstructures, which provided a theoretical basis for the engineering application of chemically stabilized swollen clay. Further, Chompoorat et al.¹⁵ also explored the use of agricultural by-product cupro lump rubber (CLR) as a component of controlled low strength material (CLSM) for road construction. The study developed two CLSM blends based on cement and alkali activation and evaluated their workability, mechanical properties and microstructure. In addition, geosynthetics can also enhance the stability and load-bearing capacity of foundations. Chompoorat et al.¹⁶ investigated the effect of palm fibers on the flexural properties and microscopic characterization of CTSF specimens and concluded that palm fiber inclusions are suitable for use as base and sub-base for bonded pavement structures in civil and pavement engineering. Buathong, Phermphorn et al.¹⁷ by studying the unconfined compression characteristics of palm fiber reinforced cement treated sand concluded that palm fibers were most effective in improving the toughness of cemented specimens due to their stiffness and bridging ability. The optimum fiber content and length for cement treated sand were 1.0% and 40 mm respectively.

In summary, phosphogypsum application in highway construction is mostly confined to be used as some auxiliary materials, which has not yet opened up the market, and has not realized engineering and large-scale application. An important reason technical support capacity is insufficient, there is no formation of phosphogypsum in road engineering and other aspects of the standard system. Secondly, the durability research is insufficient, and no in-depth research has been done on the strength decay law of phosphogypsum mixtures under dry and wet cycling conditions, which leads to early lesions on the roadbed pavement, and even poses a hidden danger to the safety of the project^18,19. Therefore, in order to improve the utilization efficiency of phosphogypsum, solve the current situation of “oversupply” of phosphogypsum, and realize the recycling of phosphogypsum, this paper investigates the attenuation law of cement stabilized phosphogypsum material in dry and wet cycles without lateral limit compressive strength. The strength decay mechanism was revealed by SEM and XRD tests. Orthogonal tests were used to analyze the significance of the effect of each factor on strength²⁰. A regression model was established using nonlinear regression analysis and model validation was carried out to explore the suitability of cement-stabilized phosphogypsum materials as pavement base (subbase) materials.

Raw material properties

The phosphogypsum used in this experiment was taken from the urn-fu phosphorus mine dump in Fuquan, Guizhou, which has a grayish-brown surface and is mostly in powder form, as shown in Fig. 1(a). After the phosphogypsum was collected, the gradation curve of phosphogypsum particles was plotted by sieving method as shown in Fig. 2, from which it can be seen that the characteristic particle sizes of phosphogypsum specimens were D₆₀ = 0.0145, D₃₀ = 0.0078, and D₁₀ = 0.0047, respectively, and according to the formulas of the coefficient of inhomogeneity and curvature coefficient, we can get that the C_U was 3.09, which was less than 5, and the curvature coefficient C_C was 0.89, and the phosphogypsum grading is uniform and poorly graded. Further the phosphogypsum was tested for heavy metal content, basic physical properties and chemical composition, and the results are shown in Tables 1, 2, 3. And the cement was purchased from a manufacturer in Guiyang, mainly PC32.5 ordinary silicate cement, gray and dry, as shown in Fig. 1(b). Its basic parameters are shown in Table 4.

Fig. 1

Raw material preparation.

Fig. 2

Phosphogypsum particle grading curve.

Table 1 Physical mechanics of phosphogypsum and chemical composition test.

Table 2 Phosphogypsum radioactivity test results.

Table 3 Phosphogypsum heavy metal content test results.

Table 4 Basic Cement Parameters.

Sample preparation

There have been a lot of studies on the proportion of cement phosphogypsum materials at home and abroad, and according to the recommendation in the Technical Rules for Construction of Highway Pavement Base Level (JTG/T F20-2015)²¹, when cement stabilized soil is used as sub-base level and the plasticity index of the soil is greater than 12, the recommended cement dosage is 6% to 14%.Dong Mansheng, Tang Qingqian and other scholars^22,23 carried out an in-depth study on cement stabilized crushed stone grass-roots level, and found that the 7d unconfined compressive strength of the mixture is the largest when the cement mixing amount is 5% and cement: phosphogypsum = 1:3, which can meet the pavement grass-roots strength requirements. Zhou Mingkai et al.²⁴ carried out the unconfined compressive strength of cemented phosphogypsum stabilized crushed stone and concluded that the compressive strength reaches its maximum when the cement dosage reaches 5%. Qian Zhengfu et al.²⁵concluded that the long-term strength of cement phosphogypsum stabilized material modified by curing agent is better through the unconfined compressive strength test with 10% cement modified phosphogypsum. Combined with the relevant specifications and previous research results, this paper selected seven cement dosage of 5%−15%, three compaction degree of 94%−98%, three different water content of optimal water content of −5%, optimal water content and optimal water content of + 5%, respectively, to conduct compaction tests on the mixture. The method of the compaction test refers to “Highway Geotechnical Testing Procedure” (JTG3430-2020)²⁶, and the calculation of cement admixture is expressed as a percentage of the mass of cement to the mass of all dry stabilized materials in accordance with the “Technical Rules for Highway Pavement Subgrade Construction” (JTG/T F20-2015) . The results of the compaction are shown in Table 5. Based on the results of the compaction test, the sample ratios and test conditions were designed as shown in Table 6 and used to perform the unconfined compressive strength test.

Table 5 Cement phosphogypsum compaction test results.

Table 6 Cement phosphogypsum ratios.

The specimen preparation process, according to the “Highway Engineering Inorganic Binding Material Stabilization Material Test Procedure” (JTG 3411–2024)²⁷, was made by the static method, and the experimental process was (see Fig. 3): phosphogypsum crushing → sieving (2 mm) → first mixing (add water) → stewing → second mixing (add cement) → hydraulic molding → demolding → finished product.

Fig. 3

Flow chart of sample making.

Pilot program

Unconfined compressive strength test program

Put the specimen at the end of maintenance into the lifting plate of the unconfined compressive strength tester, adjust the lifting plate to rise, and adjust the force gauge percentage meter to zero. During the experiment, the loading rate was set to 1 mm/min, and when the test damage was observed, the maximum reading indicated by the dial pointer was recorded, and the value of unconfined compressive strength was subsequently calculated according to the following equation.

$$R_{C} = \frac{P}{A}$$

(1)

where: $R_{C}$ is the unconfined compressive strength of the specimen (MPa); $P$ is the maximum pressure at time of destruction (N); A is the cross-sectional area of the specimen (mm²).

Wet/dry cycle program

As the roadbed soil is subject to rising water table, rainfall infiltration, solar irradiation and other factors, it will undergo changes in wet and dry cycles, and the continuous wet and dry cycles make the internal damage of the roadbed structure, and ultimately lead to a reduction in the strength of the roadbed and cracking of the roadbed, which will result in safety hazards. Therefore, an in-depth study of the mechanism and influence law of dry and wet cycles is of great significance for optimizing the design of roadbase and guaranteeing the safety of the project. Yangsukkasem N et al.²⁸ In order to investigate the effect of dry and wet cycling on the mechanical properties of the samples, the experiments were carried out according to the ASTM D559/D559-15 standard for dry and wet cycling tests. The samples were soaked in water for 5 h, then dried in an oven at 71 ± 3 °C for 42 h, and finally air-dried at room temperature for 1 h to complete the cycle. Hoy, Menglim et al.²⁹conducted w-d cyclic durability test according to ASTM standard. The three test phases consisted of a cycle of immersing the samples in controlled drinking water at 25 °C for 5 h, then drying the samples in an oven at approximately 70 °C for 42 h, and finally air-drying the samples for at least 1 h at room temperature. The weight of the samples was measured after the immersion and drying stages to check the percentage weight loss and water absorption of the samples. By studying phosphogypsum stabilized soils, Chen Kaisheng^30,31et al. found that dry–wet cycle of wetting the soil first and then drying it resulted in more significant strength decay compared to the process of drying and then wetting the soil. Therefore, considering the most unfavorable conditions in engineering practice, it was decided in this paper to test the specimens using the wet-dry cycle scheme of wetting first and then drying. In addition, the research of scholars such as Que Yun, Hu Zhi, Li Peile and others^32,33,34showed that under the influence of natural climate, the water content of the roadbed shows periodic or non-periodic fluctuations around the “equilibrium moisture content” (EMC), and the range of fluctuations is roughly EMC ± 5%. Combined with the research results of Zhou Hao, Chen Kaisheng and other scholars^35,36, that is, the strength of the soil body tends to stabilize after 5 ~ 6 times of wet and dry cycles, this paper formulated a wet and dry cycle scheme (shown in Fig. 4(a)) with 6 cycles³⁷ and 10% wet and dry amplitude to simulate and analyze the strength change of the mixes under the action of wet and dry cycles, and the specific operation flow is shown in Fig. 4(b). shown: First, the specimens whose curing was completed in 7d, 14d and 28d were placed in the curing box and the humidifier was turned on to humidify them. Remove the specimen at regular intervals and monitor its moisture content by weighing, and stop humidification when its moisture content reaches the upper limit of humidification moisture content (initial moisture content + 5%). The specimen is then sealed and left to stand for 24 h to allow for uniform internal moisture, after which the specimen is placed into an oven at a set temperature of 40 °C for drying, and a cycle is completed by continually weighing the specimen until a predetermined drying weight is reached. This process was repeated until six wet and dry cycles were completed.

Fig. 4

Dry and wet cycle test methods.

Orthogonal test

In order to further explore the effects of each factor on the unconfined compressive strength of the mix, orthogonal tests were used to analyze the results of the significant effects of each factor on the 7d unconfined compressive strength. The test factor levels are shown in Table 7.

Table 7 Experimental factors and levels.

Microtesting program

The SEM (Scanning Electron Microscopy) and XRD (X-ray Photoelectron Spectroscopy) diffraction tests were used to design the experimental program as shown in Table 8. The mechanism of strength development under different water content, different compaction degree, different age and different number of wet and dry cycles were analyzed from the particle morphology, arrangement characteristics and chemical composition of the specimens (pores). The effect of the factors can be derived by comparing the microscopic results of the different groups. The different proportioned mixes in the table were prepared by hydrostatic compression, removed after standard curing and subsequently the specimens were soaked in anhydrous ethanol to stop the hydration reaction. It was then placed in an oven set at 40 °C to dry the free water until the weight remained the same. Take out the specimen broken grinding length and width to 1–2 cm (generally not less than 1 cm), the thickness of not more than 10 mm particles placed in the German ZEISS GeminiSEM 300 scanning electron microscope to observe the specimen cross-section morphology characteristics. And part of the broken specimen grinding to a particle size of about 0.05 mm powder for X-ray diffraction test to analyze its composition in different states.

Table 8 Microscopic experimental program.

Analysis of results

Unconfined compressive strength results

(1) Effect of moisture content.

The relationship curve between the unconfined compressive strength (7d) and moisture content of the mix is shown in Fig. 5, from which it can be seen that the unconfined compressive strength of the mix shows an increase and then a decrease during the gradual increase of the moisture content of the mix. Since the relationship between the unconfined compressive strength of mixes with different cement dosage and water content has the same law, only the strength values of different water content at 5% cement dosage are shown in Table 9, from the results in the table, it can be seen that the maximum values of the strengths are found at the optimum water content³⁸. Comparing the SEM scanning electron microscope Figs. 6 (a)-(c), the mixes at the water content of OMC-5 (Fig. a) show weaker hydration reaction, loose particles, and less cementation. At the optimal moisture content (OMC) (Fig. b), the hydration reaction is sufficient, the effective contact between the particles is tight, the structure is denser, and the most cement is generated with the greatest strength. At a water content of OMC + 5 (Fig. c), the presence of excessive free water in the mix weakens the adhesion between the particles, resulting in a looser overall structure and reduced strength.

Fig. 5

Unconfined compressive strength (7d) versus moisture content.

Table 9 Statistics of unconfined compressive strength of mixes with different water contents.

Fig. 6

Scanning electron micrographs at different water contents.

(2) Effect of cement dosing.

As can be seen from Fig. 7, the 7d unconfined compressive strength of the mixes under the same compaction condition with different number of wet and dry cycles increased with the increase of cement dosage, and the slope of the curve was more jittery, indicating that the cement dosage had a greater effect on the strength of the mixes^39,40. As an example, in Fig. 7(a), for 0 wet and dry cycles, the unconfined compressive strength of the mix increased from 1.4 MPa to 3.3 MPa at 98% compaction, with growth rates of: 18.57%, 19.28%, 32.83%, 10.27%, and 14.14%. Due to space constraints, the XRD diffraction analysis of Fig. 8(a) is therefore counted as a histogram shown in Fig. 8(b), from which it can be seen that the main reason for this phenomenon is the increase in calcium alumina with the increase in cement dosage. Calcium alumina is the source of strength of the mix, when adding water and mixing with phosphogypsum, the main mineral components of the cement will undergo a hydration reaction to form hydrated calcium silicate (C-S–H)⁴¹, and these hydration products have high cement strength and filling properties, which are able to tightly bond the mix particles together, and then these hydration intermediates react with the phosphogypsum in (CaSO₄·2H₂O) in phosphogypsum to form calcium alumina (Ca₆(Al(OH)₆)₂(SO₄)₃(H₂O)₂₆, abbreviated as AFt) to play the role of skeleton support (shown in Eq. 2–3). From the microscopic scanning electron microscope, observe Figs. 9(a)-(c), where the rod-like material is the calomelite crystals. Comparing the three pictures it can be seen that as the cement dosing increases the internally generated calomelite increases, resulting in a denser reticular structure.

$$\begin{array}{*{20}c} {3Ca\left( {OH} \right)_{2} + Al_{2} O_{3} + \left( {n - 3} \right)H_{2} O \to 3CaAl_{2} O_{3} nH_{2} O} \\ \end{array}$$

(2)

$$\begin{array}{*{20}c} {3CaOAl_{2} O_{3} + 3\left( {CaSO_{4} 2H_{2} O} \right) + 2Ca\left( {OH} \right)_{2} + 24 H_{2} O \to 3 CaOAl_{2} O_{3} 3CaSO_{4} 32H_{2} O} \\ \end{array}$$

(3)

Fig. 7

Unconfined compressive strength (7d) as a function of cement mixing amount.

Fig. 8

Mineralogical composition of mixes with different cement dosages.

Fig. 9

SEM scanning electron micrographs of different cement dosages.

(3) Effect of compaction.

As can be seen from Fig. 10, the unconfined compressive strength of the mix at 7d increases with the increase of compaction under the condition of the same cement dosage for different number of wet and dry cycles. As an example, in Fig. 10(a) for 0 wet and dry cycles, the unconfined compressive strength of the mix increased from 1.52 MPa to 1.98 MPa with a growth rate of 20.39%−8.19% under the condition of 9% cement dosing and compaction from 94 to 98%. The main reasons for this are: loose structure with low compaction, large pores and low strength (Fig. 11a). With the increase of compaction, the contact between particles is closer and the structure is denser, which makes the hydration reaction between cement and phosphogypsum more adequate, more cementation is formed on the surface of the particles (Fig. 11b), and the strength is increased⁴².

Fig. 10

Unconfined compressive strength (7d) versus compaction.

Fig. 11

Microscopic electron microscope scans at different compaction levels.

(4) Impact of age of conservation.

As can be seen from Fig. 12, the 7d unconfined compressive strength of the mix increases with the age of curing for the same cement admixture at different number of wet and dry cycles⁴³. Taking the 15% cement admixture for 0 wet and dry cycles as an example, the 7d, 14d and 28d strengths were 3.15 MPa, 4.50 MPa, and 5.13Mpa strength growth was obvious, respectively. From the SEM scans (Fig. 13a-b), this is due to the large amount of cementitious material attached to the surface of the specimens at the age of 14d, and the hydration products wrapped the structure tightly. Further observation of the change in the relative mineral content in Fig. 14shows that the reaction between the cement hydration products and phosphogypsum was more adequate with the extension of the curing time, generating more calcite, which led to a significant increase in the strength of the mix⁴⁴.

Fig. 12

Relationship between unconfined compressive strength and age under different number of wet and dry cycles.

Fig. 13

SEM scans at different conservation ages.

Fig. 14

Mineral composition of mixes at different maintenance ages.

(5) Effect of the number of wet and dry cycles.

As can be seen from Fig. 15, under the condition of the same compaction degree with different cement dosage, the unconfined compressive strength (7d) decreased with the increase of the number of wet and dry cycles, and the first four wet and dry cycles decayed faster, and then gradually leveled off. This is mainly due to the fact that the mixture undergoes the effects of two key components, humidification and drying, under the alternating action of the wet and dry cycles. During the humidification phase, the mix is subjected to seepage water, which induces inter-particle migration, a process that directly contributes to the weakening of the specimen strength. At the same time, the presence of moisture also acts as a lubricant, which further reduces the coefficient of friction between the particles, thus exacerbating the decrease in the overall strength of the specimen⁴⁵. During the drying stage, the uneven internal and external heating leads to the generation of tensile stresses, which causes this unfavorable phenomenon to continue to expand when the number of cycles increases, resulting in a continuous decrease in the strength of the specimen. However, the mix strength will eventually level off, indicating that the main influence of specimen strength will shift to other factors after a certain dry–wet alternation⁴⁶. From the microscopic scanning Fig. 16, the mix is less porous, denser in structure, more cemented and stronger under 0 wet and dry cycles (Fig. a). Under four wet and dry cycles (Fig. b), the mix undergoes repeated drying shrinkage and water-absorbing expansion leading to interfacial bond failure, resulting in a rapid increase in pore size, a rapid decrease in cementation, and a large attenuation of the strength between particles. Under six wet and dry cycles (Fig. c), there was only a further slight increase in porosity, and the strength decay was mainly due to the decrease in cementation, which changed the original reticulation to a laminated stacked structure⁴⁷. This finding is consistent with the rapid decrease and then stabilization of specimen strength values with the increase in the number of wet and dry cycles. Further observation of the relative mineral content of the mixture under different wet and dry cycles shows that (Fig. 17), the ${\text{Ca}}\left( {{\text{OH}}} \right)_{2}$ in the mixture disappeared after 4–6 wet and dry cycles, the AFt content decreased, and the ${\text{CaSO}}_{4} \cdot2{\text{H}}_{2} {\text{O}}$ content increased. This is due to the fact that some of the AFt decomposes during the drying process into calcium oxide (CaO), sulfur trioxide (SO₃), and other substances. During the humidification process, SO₃ dissolves in water to form sulfuric acid (${\text{H}}_{2} {\text{SO}}_{4}$), and at the same time, phosphogypsum also contains a large amount of ${\text{SO}}_{4}^{2 - }$, which makes the system acidic and accelerates the decomposition of AFt. This is one of the main reasons for the decrease of AFt in the specimen after the dry–wet cycle. In summary, the main mechanisms of strength decay of the mixture under the action of dry and wet cycles are: (1) Structural: dry and wet cycles increase the porosity inside the specimen, reduce the gelling material between the plates, and the connecting effect is weakened, reducing the integrity of the specimen. (2) Chemically: dry and wet cycles make AFt partially decompose and the reinforcing effect of AFt is weakened.

Fig. 15

Unconfined compressive strength (7d) versus number of wet and dry cycles.

Fig. 16

Scanning electron microscopy of mixes under different dry and wet cycles.

Fig.17

Mineral composition statistics under different dry and wet cycles.

In order to further investigate the intrinsic relationship between the unconfined compressive strength (7d) and the number of dry and wet cycles, the unconfined compressive strength (7d) was fitted to the number of dry and wet cycles, and the results are shown in Fig. 14 and Table 10. It was found that the unconfined compressive strength (7d) was in a negative exponential function with the number of dry and wet cycles (Eq. 1) and the correlation coefficients were all greater than 0.98, which indicated that the fit was good.

$$R_{C} = {\text{a}} - {\text{b}}*{\text{c}}^{ \wedge } {\text{N}}$$

(4)

where: $R_{C}$ is the unconfined compressive strength; N is the number of wet and dry cycles; a, b, c are fitting parameters.

Table 10 Parameters for fitting the number of dry and wet cycles to the unconfined compressive strength.

Significance analysis

Table 11 shows the results of orthogonal tests and Table 12shows the results of ANOVA. The F-value in the table can reflect the size of the P-value, specifically, the larger the F-value, the smaller the P-value, indicating the higher significance of the model⁴⁸. From the data in the table, it can be seen that the P-values of the three factors, cement dosage, compaction, and the number of wet and dry cycles, are all below 0.05, implying that all of them have a significant effect on the strength of the mix. By comparing the F-values, it can be found that the order of the influence of each factor is: cement dosage > number of wet and dry cycles > compaction, which indicates that the cement dosage has the most significant influence on the strength of the mix, followed by the number of wet and dry cycles. This shows that there is a high necessity for the study of dry and wet cycles.

Table 11 Response surface analysis design scheme and experimental results.

Table 12 Results of orthogonal analysis.

Multifactorial analysis

A multifactor nonlinear regression fit leads to the regression equation shown in Eq. 5, which is only applicable to strength prediction for cement dosage in the range of 9–13%, compaction 94–98%, and the number of wet and dry cycles 0–6. The various correlation coefficients of the model are shown in Table 13, from which it can be seen that the model has R-Squared = 0.993 and Adj R-Squared = 0.9962, which indicates that the model has a high degree of fit, and that its regression equation is able to explain 99.62% of the variation in the response values, with a variance of only 0.007%. Finally, the C.V. (coefficient of variation) in the table is 1.55%, which is further evidence that the model is accurate and reliable⁴⁹. The interaction between the factors is shown in Fig. 18. From Fig. 18(c), it can be seen that under the influence of wet and dry cycles, the unconfined compressive strength will be attenuated, but in the same number of wet and dry cycles to improve the compaction for the unconfined compressive strength of the mixture will be improved, and from the figure can be seen that the greater the degree of compaction to enhance the effect is more pronounced, which suggests that improving the degree of compaction can alleviate the attenuation of the mixture in the wet and dry cycles. As an example, for a mix with 7% cement dosing and 94% compaction, after one and six wet and dry cycles, the unconfined compressive strength values were 1.38 MPa and 0.8 MPa, whereas when the compaction was increased to 98%, the unconfined compressive strength values of the mix were 1.66 MPa and 0.99 MPa, which was an improvement of about 20%. From Fig. 18(b), it can be seen that again the unconfined compressive strength of the mix decays under the action of wet and dry cycling, but it can be seen from the figure that when the cement dosage is increased it helps to mitigate this trend, specifically the strength of the specimen increases by at least 10% or more for every 3% increase in cement at K = 96 and 6 wet and dry cycles. Therefore, in summary, in the actual operation process, in order to reduce the impact of dry and wet cycle can be appropriate to enhance the cement mixing or compression.

Table 13 Model reliability analysis.

Fig.18

Three-dimensional corresponding surface plot of the effect of the interaction between the number of wet and dry cycles, cement dosing, and compaction on unconfined compressive strength.

Regression equation:

$$\begin{gathered} {\text{Y}}\, = \, - {72}.{89187}\, + \,{1}.{34375}*{\text{A}}\, + \,{1}.{29875}*{\text{B}} - 0.0{475}00*{\text{C}} - {6}.{25}000{\text{E}} - 00{4}*{\text{A}}*{\text{B}}\, + \,{9}.{\text{58333E}} - 00{3}*{\text{A}}*{\text{C}} - {3}.{\text{33333E}} - 00{3}*{\text{B}}* \hfill \\ {\text{C}} - 0.0{48125}*{\text{A}}^{{2}} - {6}.{25}000{\text{E}} - 00{3}*{\text{B}}^{{2}} \, + \,0.0{23333}*{\text{ C}}^{{2}} \hfill \\ \end{gathered}$$

(5)

Note: A is cement admixture (%); B is compaction (%); C is number of wet and dry cycles (N).

Validation of regression equations

In order to analyze the error in the prediction results of the regression model, this paper plots the degree of deviation of the prediction results from the true values for different cement dosages as shown in Fig. 19. As can be seen in Fig. 19, the predicted values of the regression model are basically very close to the y = x line (predicted values for true values), indicating that the predicted values fit the true values better. Table 14 shows the results of the assessment of the Root Mean Square Error (RMSE) and the Absolute Mean Error (MAE) of the model, from which the maximum values of the RMSE and MAE of the prediction model are obtained as 0.0862 and 0.0776, respectively. This further deepens the reliability of the data and the accuracy of the model.

Fig.19

Comparison of predicted and real values.

Table 14 Regression model error analysis.

Discussion on the suitability of cement phosphogypsum as pavement base layer (sub-base)

The technical rules of highway pavement grass-roots construction JTC/T-2015 for cement-stabilized phosphogypsum material’s 7d age unconfined compressive strength standard is shown in Table 15.

Table 15 Unconfined compressive strength criteria for cement stabilized materials at 7d age Rd(MPa).

The 7d unconfined compressive strength of cement stabilized phosphogypsum material is shown in Table 16.

Table 16 7d unconfined compressive strength of mixes with different ratios at optimum water content.

In order to explore the suitability of cement-stabilized phosphogypsum materials for use as pavement base layer (sub-base layer), it can be seen by comparing the 7d unconfined compressive strength of different ratio mixes in Table 16 with Table 15, “Technical Rules for Construction of Highway Pavement Base Layers,” JTC/T-2015 standard: (1) It can be used as the sub-base of pavement: 5%−9% cement mix, medium and light traffic of Grade 2 and below; 11%−13% cement mix, medium, light and heavy traffic of highway and Grade 1 highway; while 15% cement mix fully meets all the requirements of sub-base material. (2) It can be used as pavement base: 11% cement dosing, medium and light traffic for secondary and lower secondary roads; 15% cement dosing, heavy traffic for secondary and lower secondary roads, and medium and light traffic for highways and primary roads.

Conclusion

In this paper, the mechanical properties and microstructural characteristics of cemented phosphogypsum mixes were investigated using unconfined compressive strength tests, scanning electron microscope (SEM) analysis and X-ray diffraction (XRD) tests, and the following conclusions were drawn:

(1) The unconfined compressive strength of cement stabilized phosphogypsum materials showed an increasing and then decreasing trend with increasing moisture content, reaching a maximum value at the optimal moisture content. It increases with the increase of compaction degree, cement mixing amount, and curing age. Decrease with the increase of the number of wet and dry cycles, the first 4 wet and dry cycles decay faster, and then gradually leveled off. The unconfined compressive strength and the number of wet and dry cycles can be fitted in the form of a negative exponential function, and the fits are all greater than 0.98, which is a good fit and a reliable model.

(2) The main mechanism of strength decay of mixes under wet and dry cycles was revealed by microscopic tests: wet and dry cycles increased the porosity inside the specimen, reduced the cementitious material between the plates, weakened the connecting effect, and reduced the integrity of the specimen; The dry and wet cycles make AFt partially decompose and the reinforcing effect of AFt is weakened.

(3) Through the orthogonal analysis test, it was concluded that the cement dosage had the most significant effect on the unconfined compressive strength, followed by the number of wet and dry cycles, and the compaction degree had the least effect, which indicates that the research on wet and dry cycles has a strong necessity.

(4) The multi-factor regression equation for the unconfined compressive strength of cement-stabilized phosphogypsum materials was established and verified, and the regression model has high reliability, the maximum values of RMSE and MAE are 0.0862 and 0.0776 respectively, with small errors, which can provide reference for the research of strength characteristics of cement-stabilized phosphogypsum materials.

(5) Comparison of the 7d unconfined compressive strength of the mix with the “Technical Rules for Highway Pavement Subgrade Construction” JTC/T-2015 can be obtained. Under different cement mixing amount and different traffic volume conditions, cement stabilized phosphogypsum mixture can be used as the base layer (sub-base layer) of highway pavement of various grades.

(6) In summary, the cement stabilized phosphogypsum materials proposed in this paper can provide certain reference for the application of phosphogypsum in road engineering, but due to regional differences, the results obtained from the test materials in different regions have some influence, so it is necessary to further carry out a large number of practice and research, from which to summarize the experience and enrich the research content of phosphogypsum.